diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,193746 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 3.0, + "eval_steps": 500, + "global_step": 27672, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00010841283607979184, + "grad_norm": 52.482452392578125, + "learning_rate": 2.0000000000000002e-07, + "loss": 2.8821, + "step": 1 + }, + { + "epoch": 0.0002168256721595837, + "grad_norm": 50.96417236328125, + "learning_rate": 4.0000000000000003e-07, + "loss": 2.8459, + "step": 2 + }, + { + "epoch": 0.00032523850823937554, + "grad_norm": 58.878318786621094, + "learning_rate": 6.000000000000001e-07, + "loss": 2.8672, + "step": 3 + }, + { + "epoch": 0.0004336513443191674, + "grad_norm": 49.31254959106445, + "learning_rate": 8.000000000000001e-07, + "loss": 2.8479, + "step": 4 + }, + { + "epoch": 0.0005420641803989592, + "grad_norm": 47.3974494934082, + "learning_rate": 1.0000000000000002e-06, + "loss": 2.963, + "step": 5 + }, + { + "epoch": 0.0006504770164787511, + "grad_norm": 50.50360107421875, + "learning_rate": 1.2000000000000002e-06, + "loss": 2.8872, + "step": 6 + }, + { + "epoch": 0.000758889852558543, + "grad_norm": 52.41767120361328, + "learning_rate": 1.4000000000000001e-06, + "loss": 2.8418, + "step": 7 + }, + { + "epoch": 0.0008673026886383347, + "grad_norm": 48.43821716308594, + "learning_rate": 1.6000000000000001e-06, + "loss": 2.8884, + "step": 8 + }, + { + "epoch": 0.0009757155247181266, + "grad_norm": 50.77457046508789, + "learning_rate": 1.8000000000000001e-06, + "loss": 2.8652, + "step": 9 + }, + { + "epoch": 0.0010841283607979184, + "grad_norm": 49.59326934814453, + "learning_rate": 2.0000000000000003e-06, + "loss": 2.7935, + "step": 10 + }, + { + "epoch": 0.0011925411968777104, + "grad_norm": 49.99949264526367, + "learning_rate": 2.2e-06, + "loss": 2.8278, + "step": 11 + }, + { + "epoch": 0.0013009540329575022, + "grad_norm": 57.374717712402344, + "learning_rate": 2.4000000000000003e-06, + "loss": 2.8308, + "step": 12 + }, + { + "epoch": 0.001409366869037294, + "grad_norm": 47.822635650634766, + "learning_rate": 2.6e-06, + "loss": 2.6684, + "step": 13 + }, + { + "epoch": 0.001517779705117086, + "grad_norm": 45.97601318359375, + "learning_rate": 2.8000000000000003e-06, + "loss": 2.7504, + "step": 14 + }, + { + "epoch": 0.0016261925411968777, + "grad_norm": 45.66847610473633, + "learning_rate": 3e-06, + "loss": 2.3727, + "step": 15 + }, + { + "epoch": 0.0017346053772766695, + "grad_norm": 49.50860595703125, + "learning_rate": 3.2000000000000003e-06, + "loss": 2.4081, + "step": 16 + }, + { + "epoch": 0.0018430182133564615, + "grad_norm": 46.31549072265625, + "learning_rate": 3.4000000000000005e-06, + "loss": 2.197, + "step": 17 + }, + { + "epoch": 0.0019514310494362533, + "grad_norm": 45.704437255859375, + "learning_rate": 3.6000000000000003e-06, + "loss": 2.2501, + "step": 18 + }, + { + "epoch": 0.002059843885516045, + "grad_norm": 46.57144546508789, + "learning_rate": 3.8000000000000005e-06, + "loss": 2.1068, + "step": 19 + }, + { + "epoch": 0.002168256721595837, + "grad_norm": 36.14391326904297, + "learning_rate": 4.000000000000001e-06, + "loss": 1.7364, + "step": 20 + }, + { + "epoch": 0.0022766695576756286, + "grad_norm": 35.68048095703125, + "learning_rate": 4.2000000000000004e-06, + "loss": 1.5302, + "step": 21 + }, + { + "epoch": 0.002385082393755421, + "grad_norm": 31.292877197265625, + "learning_rate": 4.4e-06, + "loss": 1.5205, + "step": 22 + }, + { + "epoch": 0.0024934952298352126, + "grad_norm": 32.74424362182617, + "learning_rate": 4.600000000000001e-06, + "loss": 1.4001, + "step": 23 + }, + { + "epoch": 0.0026019080659150044, + "grad_norm": 29.111459732055664, + "learning_rate": 4.800000000000001e-06, + "loss": 1.2814, + "step": 24 + }, + { + "epoch": 0.002710320901994796, + "grad_norm": 30.326162338256836, + "learning_rate": 5e-06, + "loss": 1.2465, + "step": 25 + }, + { + "epoch": 0.002818733738074588, + "grad_norm": 27.826641082763672, + "learning_rate": 5.2e-06, + "loss": 0.9693, + "step": 26 + }, + { + "epoch": 0.0029271465741543797, + "grad_norm": 26.075820922851562, + "learning_rate": 5.400000000000001e-06, + "loss": 0.7454, + "step": 27 + }, + { + "epoch": 0.003035559410234172, + "grad_norm": 33.648681640625, + "learning_rate": 5.600000000000001e-06, + "loss": 0.7843, + "step": 28 + }, + { + "epoch": 0.0031439722463139637, + "grad_norm": 20.305675506591797, + "learning_rate": 5.8e-06, + "loss": 0.5237, + "step": 29 + }, + { + "epoch": 0.0032523850823937554, + "grad_norm": 21.991626739501953, + "learning_rate": 6e-06, + "loss": 0.522, + "step": 30 + }, + { + "epoch": 0.0033607979184735472, + "grad_norm": 15.066353797912598, + "learning_rate": 6.200000000000001e-06, + "loss": 0.3914, + "step": 31 + }, + { + "epoch": 0.003469210754553339, + "grad_norm": 15.034688949584961, + "learning_rate": 6.4000000000000006e-06, + "loss": 0.3601, + "step": 32 + }, + { + "epoch": 0.0035776235906331308, + "grad_norm": 12.905313491821289, + "learning_rate": 6.600000000000001e-06, + "loss": 0.289, + "step": 33 + }, + { + "epoch": 0.003686036426712923, + "grad_norm": 12.697845458984375, + "learning_rate": 6.800000000000001e-06, + "loss": 0.2615, + "step": 34 + }, + { + "epoch": 0.0037944492627927148, + "grad_norm": 8.485983848571777, + "learning_rate": 7e-06, + "loss": 0.2166, + "step": 35 + }, + { + "epoch": 0.0039028620988725065, + "grad_norm": 5.852139949798584, + "learning_rate": 7.2000000000000005e-06, + "loss": 0.188, + "step": 36 + }, + { + "epoch": 0.004011274934952298, + "grad_norm": 5.829102993011475, + "learning_rate": 7.4e-06, + "loss": 0.1485, + "step": 37 + }, + { + "epoch": 0.00411968777103209, + "grad_norm": 4.062847137451172, + "learning_rate": 7.600000000000001e-06, + "loss": 0.1384, + "step": 38 + }, + { + "epoch": 0.004228100607111882, + "grad_norm": 5.279386043548584, + "learning_rate": 7.800000000000002e-06, + "loss": 0.1387, + "step": 39 + }, + { + "epoch": 0.004336513443191674, + "grad_norm": 3.165452003479004, + "learning_rate": 8.000000000000001e-06, + "loss": 0.1445, + "step": 40 + }, + { + "epoch": 0.004444926279271465, + "grad_norm": 3.205404043197632, + "learning_rate": 8.2e-06, + "loss": 0.1435, + "step": 41 + }, + { + "epoch": 0.004553339115351257, + "grad_norm": 2.1782217025756836, + "learning_rate": 8.400000000000001e-06, + "loss": 0.1004, + "step": 42 + }, + { + "epoch": 0.00466175195143105, + "grad_norm": 3.22420334815979, + "learning_rate": 8.6e-06, + "loss": 0.1096, + "step": 43 + }, + { + "epoch": 0.004770164787510842, + "grad_norm": 4.1656880378723145, + "learning_rate": 8.8e-06, + "loss": 0.1335, + "step": 44 + }, + { + "epoch": 0.004878577623590633, + "grad_norm": 4.064053058624268, + "learning_rate": 9e-06, + "loss": 0.1349, + "step": 45 + }, + { + "epoch": 0.004986990459670425, + "grad_norm": 4.505921363830566, + "learning_rate": 9.200000000000002e-06, + "loss": 0.1356, + "step": 46 + }, + { + "epoch": 0.005095403295750217, + "grad_norm": 6.455411911010742, + "learning_rate": 9.4e-06, + "loss": 0.1515, + "step": 47 + }, + { + "epoch": 0.005203816131830009, + "grad_norm": 4.459105014801025, + "learning_rate": 9.600000000000001e-06, + "loss": 0.1224, + "step": 48 + }, + { + "epoch": 0.0053122289679098005, + "grad_norm": 7.928279399871826, + "learning_rate": 9.800000000000001e-06, + "loss": 0.1297, + "step": 49 + }, + { + "epoch": 0.005420641803989592, + "grad_norm": 6.529297828674316, + "learning_rate": 1e-05, + "loss": 0.1426, + "step": 50 + }, + { + "epoch": 0.005529054640069384, + "grad_norm": 4.190447807312012, + "learning_rate": 1.02e-05, + "loss": 0.1387, + "step": 51 + }, + { + "epoch": 0.005637467476149176, + "grad_norm": 3.4309422969818115, + "learning_rate": 1.04e-05, + "loss": 0.1421, + "step": 52 + }, + { + "epoch": 0.005745880312228968, + "grad_norm": 3.8757483959198, + "learning_rate": 1.0600000000000002e-05, + "loss": 0.1315, + "step": 53 + }, + { + "epoch": 0.005854293148308759, + "grad_norm": 1.9450610876083374, + "learning_rate": 1.0800000000000002e-05, + "loss": 0.112, + "step": 54 + }, + { + "epoch": 0.005962705984388552, + "grad_norm": 3.8325629234313965, + "learning_rate": 1.1000000000000001e-05, + "loss": 0.1468, + "step": 55 + }, + { + "epoch": 0.006071118820468344, + "grad_norm": 1.3910635709762573, + "learning_rate": 1.1200000000000001e-05, + "loss": 0.1059, + "step": 56 + }, + { + "epoch": 0.0061795316565481356, + "grad_norm": 5.278274059295654, + "learning_rate": 1.14e-05, + "loss": 0.1702, + "step": 57 + }, + { + "epoch": 0.006287944492627927, + "grad_norm": 2.068516254425049, + "learning_rate": 1.16e-05, + "loss": 0.1052, + "step": 58 + }, + { + "epoch": 0.006396357328707719, + "grad_norm": 1.7592467069625854, + "learning_rate": 1.18e-05, + "loss": 0.1286, + "step": 59 + }, + { + "epoch": 0.006504770164787511, + "grad_norm": 2.104170083999634, + "learning_rate": 1.2e-05, + "loss": 0.129, + "step": 60 + }, + { + "epoch": 0.006613183000867303, + "grad_norm": 5.306731224060059, + "learning_rate": 1.22e-05, + "loss": 0.5374, + "step": 61 + }, + { + "epoch": 0.0067215958369470944, + "grad_norm": 1.0200713872909546, + "learning_rate": 1.2400000000000002e-05, + "loss": 0.1071, + "step": 62 + }, + { + "epoch": 0.006830008673026886, + "grad_norm": 1.4894516468048096, + "learning_rate": 1.2600000000000001e-05, + "loss": 0.1034, + "step": 63 + }, + { + "epoch": 0.006938421509106678, + "grad_norm": 3.1711857318878174, + "learning_rate": 1.2800000000000001e-05, + "loss": 0.1574, + "step": 64 + }, + { + "epoch": 0.00704683434518647, + "grad_norm": 1.9187594652175903, + "learning_rate": 1.3000000000000001e-05, + "loss": 0.1304, + "step": 65 + }, + { + "epoch": 0.0071552471812662615, + "grad_norm": 2.2690370082855225, + "learning_rate": 1.3200000000000002e-05, + "loss": 0.1383, + "step": 66 + }, + { + "epoch": 0.007263660017346054, + "grad_norm": 2.383378267288208, + "learning_rate": 1.3400000000000002e-05, + "loss": 0.1391, + "step": 67 + }, + { + "epoch": 0.007372072853425846, + "grad_norm": 1.892323613166809, + "learning_rate": 1.3600000000000002e-05, + "loss": 0.1413, + "step": 68 + }, + { + "epoch": 0.007480485689505638, + "grad_norm": 2.1305816173553467, + "learning_rate": 1.38e-05, + "loss": 0.1347, + "step": 69 + }, + { + "epoch": 0.0075888985255854295, + "grad_norm": 3.0918381214141846, + "learning_rate": 1.4e-05, + "loss": 0.1094, + "step": 70 + }, + { + "epoch": 0.007697311361665221, + "grad_norm": 1.73595130443573, + "learning_rate": 1.4200000000000001e-05, + "loss": 0.1393, + "step": 71 + }, + { + "epoch": 0.007805724197745013, + "grad_norm": 7.937272548675537, + "learning_rate": 1.4400000000000001e-05, + "loss": 0.2171, + "step": 72 + }, + { + "epoch": 0.007914137033824806, + "grad_norm": 2.835125207901001, + "learning_rate": 1.46e-05, + "loss": 0.1354, + "step": 73 + }, + { + "epoch": 0.008022549869904597, + "grad_norm": 1.5085129737854004, + "learning_rate": 1.48e-05, + "loss": 0.1097, + "step": 74 + }, + { + "epoch": 0.00813096270598439, + "grad_norm": 1.7230224609375, + "learning_rate": 1.5000000000000002e-05, + "loss": 0.1148, + "step": 75 + }, + { + "epoch": 0.00823937554206418, + "grad_norm": 2.175292491912842, + "learning_rate": 1.5200000000000002e-05, + "loss": 0.1138, + "step": 76 + }, + { + "epoch": 0.008347788378143973, + "grad_norm": 3.3230648040771484, + "learning_rate": 1.54e-05, + "loss": 0.1245, + "step": 77 + }, + { + "epoch": 0.008456201214223764, + "grad_norm": 3.8972442150115967, + "learning_rate": 1.5600000000000003e-05, + "loss": 0.1408, + "step": 78 + }, + { + "epoch": 0.008564614050303556, + "grad_norm": 2.6898117065429688, + "learning_rate": 1.58e-05, + "loss": 0.1208, + "step": 79 + }, + { + "epoch": 0.008673026886383347, + "grad_norm": 2.5461599826812744, + "learning_rate": 1.6000000000000003e-05, + "loss": 0.1006, + "step": 80 + }, + { + "epoch": 0.00878143972246314, + "grad_norm": 3.4649367332458496, + "learning_rate": 1.62e-05, + "loss": 0.1016, + "step": 81 + }, + { + "epoch": 0.00888985255854293, + "grad_norm": 3.6255903244018555, + "learning_rate": 1.64e-05, + "loss": 0.1121, + "step": 82 + }, + { + "epoch": 0.008998265394622723, + "grad_norm": 3.6274213790893555, + "learning_rate": 1.66e-05, + "loss": 0.1406, + "step": 83 + }, + { + "epoch": 0.009106678230702514, + "grad_norm": 2.4610044956207275, + "learning_rate": 1.6800000000000002e-05, + "loss": 0.1033, + "step": 84 + }, + { + "epoch": 0.009215091066782307, + "grad_norm": 4.4305267333984375, + "learning_rate": 1.7e-05, + "loss": 0.1252, + "step": 85 + }, + { + "epoch": 0.0093235039028621, + "grad_norm": 4.74913215637207, + "learning_rate": 1.72e-05, + "loss": 0.1572, + "step": 86 + }, + { + "epoch": 0.00943191673894189, + "grad_norm": 1.56508207321167, + "learning_rate": 1.7400000000000003e-05, + "loss": 0.0918, + "step": 87 + }, + { + "epoch": 0.009540329575021683, + "grad_norm": 1.743727684020996, + "learning_rate": 1.76e-05, + "loss": 0.1101, + "step": 88 + }, + { + "epoch": 0.009648742411101474, + "grad_norm": 3.2393627166748047, + "learning_rate": 1.7800000000000002e-05, + "loss": 0.1388, + "step": 89 + }, + { + "epoch": 0.009757155247181267, + "grad_norm": 2.405870199203491, + "learning_rate": 1.8e-05, + "loss": 0.1029, + "step": 90 + }, + { + "epoch": 0.009865568083261058, + "grad_norm": 5.1828203201293945, + "learning_rate": 1.8200000000000002e-05, + "loss": 0.1224, + "step": 91 + }, + { + "epoch": 0.00997398091934085, + "grad_norm": 3.903477668762207, + "learning_rate": 1.8400000000000003e-05, + "loss": 0.12, + "step": 92 + }, + { + "epoch": 0.010082393755420641, + "grad_norm": 4.98646879196167, + "learning_rate": 1.86e-05, + "loss": 0.1023, + "step": 93 + }, + { + "epoch": 0.010190806591500434, + "grad_norm": 22.051128387451172, + "learning_rate": 1.88e-05, + "loss": 0.1664, + "step": 94 + }, + { + "epoch": 0.010299219427580225, + "grad_norm": 4.599372386932373, + "learning_rate": 1.9e-05, + "loss": 0.1473, + "step": 95 + }, + { + "epoch": 0.010407632263660017, + "grad_norm": 4.740875244140625, + "learning_rate": 1.9200000000000003e-05, + "loss": 0.0862, + "step": 96 + }, + { + "epoch": 0.010516045099739808, + "grad_norm": 5.410498142242432, + "learning_rate": 1.94e-05, + "loss": 0.139, + "step": 97 + }, + { + "epoch": 0.010624457935819601, + "grad_norm": 4.7862982749938965, + "learning_rate": 1.9600000000000002e-05, + "loss": 0.1472, + "step": 98 + }, + { + "epoch": 0.010732870771899394, + "grad_norm": 4.430928707122803, + "learning_rate": 1.98e-05, + "loss": 0.1383, + "step": 99 + }, + { + "epoch": 0.010841283607979185, + "grad_norm": 6.088566780090332, + "learning_rate": 2e-05, + "loss": 0.1163, + "step": 100 + }, + { + "epoch": 0.010949696444058977, + "grad_norm": 3.0314104557037354, + "learning_rate": 1.9999274626432614e-05, + "loss": 0.0963, + "step": 101 + }, + { + "epoch": 0.011058109280138768, + "grad_norm": 3.663384437561035, + "learning_rate": 1.9998549252865226e-05, + "loss": 0.1292, + "step": 102 + }, + { + "epoch": 0.01116652211621856, + "grad_norm": 2.9560699462890625, + "learning_rate": 1.9997823879297838e-05, + "loss": 0.1034, + "step": 103 + }, + { + "epoch": 0.011274934952298352, + "grad_norm": 2.6167972087860107, + "learning_rate": 1.9997098505730454e-05, + "loss": 0.0962, + "step": 104 + }, + { + "epoch": 0.011383347788378144, + "grad_norm": 10.235919952392578, + "learning_rate": 1.9996373132163066e-05, + "loss": 0.1172, + "step": 105 + }, + { + "epoch": 0.011491760624457935, + "grad_norm": 3.884672164916992, + "learning_rate": 1.9995647758595678e-05, + "loss": 0.1369, + "step": 106 + }, + { + "epoch": 0.011600173460537728, + "grad_norm": 5.411236763000488, + "learning_rate": 1.999492238502829e-05, + "loss": 0.1222, + "step": 107 + }, + { + "epoch": 0.011708586296617519, + "grad_norm": 2.467689275741577, + "learning_rate": 1.9994197011460902e-05, + "loss": 0.1123, + "step": 108 + }, + { + "epoch": 0.011816999132697311, + "grad_norm": 4.753982067108154, + "learning_rate": 1.9993471637893518e-05, + "loss": 0.1021, + "step": 109 + }, + { + "epoch": 0.011925411968777104, + "grad_norm": 3.258648633956909, + "learning_rate": 1.999274626432613e-05, + "loss": 0.1079, + "step": 110 + }, + { + "epoch": 0.012033824804856895, + "grad_norm": 4.9031662940979, + "learning_rate": 1.9992020890758742e-05, + "loss": 0.1315, + "step": 111 + }, + { + "epoch": 0.012142237640936688, + "grad_norm": 7.437492847442627, + "learning_rate": 1.9991295517191358e-05, + "loss": 0.1856, + "step": 112 + }, + { + "epoch": 0.012250650477016478, + "grad_norm": 2.8113110065460205, + "learning_rate": 1.999057014362397e-05, + "loss": 0.0895, + "step": 113 + }, + { + "epoch": 0.012359063313096271, + "grad_norm": 3.9714345932006836, + "learning_rate": 1.9989844770056582e-05, + "loss": 0.0923, + "step": 114 + }, + { + "epoch": 0.012467476149176062, + "grad_norm": 4.005449295043945, + "learning_rate": 1.9989119396489194e-05, + "loss": 0.1261, + "step": 115 + }, + { + "epoch": 0.012575888985255855, + "grad_norm": 8.382218360900879, + "learning_rate": 1.9988394022921806e-05, + "loss": 0.0929, + "step": 116 + }, + { + "epoch": 0.012684301821335646, + "grad_norm": 2.9554052352905273, + "learning_rate": 1.9987668649354418e-05, + "loss": 0.0866, + "step": 117 + }, + { + "epoch": 0.012792714657415438, + "grad_norm": 2.808568000793457, + "learning_rate": 1.998694327578703e-05, + "loss": 0.0726, + "step": 118 + }, + { + "epoch": 0.01290112749349523, + "grad_norm": 19.925622940063477, + "learning_rate": 1.9986217902219646e-05, + "loss": 0.1562, + "step": 119 + }, + { + "epoch": 0.013009540329575022, + "grad_norm": 6.193617343902588, + "learning_rate": 1.9985492528652258e-05, + "loss": 0.1061, + "step": 120 + }, + { + "epoch": 0.013117953165654813, + "grad_norm": 4.706057548522949, + "learning_rate": 1.998476715508487e-05, + "loss": 0.0636, + "step": 121 + }, + { + "epoch": 0.013226366001734605, + "grad_norm": 8.763757705688477, + "learning_rate": 1.9984041781517482e-05, + "loss": 0.1625, + "step": 122 + }, + { + "epoch": 0.013334778837814398, + "grad_norm": 7.199608325958252, + "learning_rate": 1.9983316407950094e-05, + "loss": 0.0936, + "step": 123 + }, + { + "epoch": 0.013443191673894189, + "grad_norm": 5.027112007141113, + "learning_rate": 1.9982591034382707e-05, + "loss": 0.0965, + "step": 124 + }, + { + "epoch": 0.013551604509973982, + "grad_norm": 6.421952724456787, + "learning_rate": 1.9981865660815322e-05, + "loss": 0.0982, + "step": 125 + }, + { + "epoch": 0.013660017346053772, + "grad_norm": 12.643232345581055, + "learning_rate": 1.9981140287247934e-05, + "loss": 0.1347, + "step": 126 + }, + { + "epoch": 0.013768430182133565, + "grad_norm": 7.623237133026123, + "learning_rate": 1.9980414913680546e-05, + "loss": 0.0992, + "step": 127 + }, + { + "epoch": 0.013876843018213356, + "grad_norm": 6.297314643859863, + "learning_rate": 1.9979689540113162e-05, + "loss": 0.1269, + "step": 128 + }, + { + "epoch": 0.013985255854293149, + "grad_norm": 5.98291015625, + "learning_rate": 1.9978964166545774e-05, + "loss": 0.0887, + "step": 129 + }, + { + "epoch": 0.01409366869037294, + "grad_norm": 4.467803478240967, + "learning_rate": 1.9978238792978386e-05, + "loss": 0.1243, + "step": 130 + }, + { + "epoch": 0.014202081526452732, + "grad_norm": 5.704503059387207, + "learning_rate": 1.9977513419411e-05, + "loss": 0.1196, + "step": 131 + }, + { + "epoch": 0.014310494362532523, + "grad_norm": 3.0214226245880127, + "learning_rate": 1.997678804584361e-05, + "loss": 0.1118, + "step": 132 + }, + { + "epoch": 0.014418907198612316, + "grad_norm": 2.774113416671753, + "learning_rate": 1.9976062672276223e-05, + "loss": 0.1192, + "step": 133 + }, + { + "epoch": 0.014527320034692108, + "grad_norm": 2.3147518634796143, + "learning_rate": 1.9975337298708838e-05, + "loss": 0.0938, + "step": 134 + }, + { + "epoch": 0.0146357328707719, + "grad_norm": 2.7756924629211426, + "learning_rate": 1.997461192514145e-05, + "loss": 0.1115, + "step": 135 + }, + { + "epoch": 0.014744145706851692, + "grad_norm": 2.1733040809631348, + "learning_rate": 1.9973886551574063e-05, + "loss": 0.1101, + "step": 136 + }, + { + "epoch": 0.014852558542931483, + "grad_norm": 2.083955764770508, + "learning_rate": 1.9973161178006675e-05, + "loss": 0.0998, + "step": 137 + }, + { + "epoch": 0.014960971379011275, + "grad_norm": 1.7569630146026611, + "learning_rate": 1.9972435804439287e-05, + "loss": 0.0951, + "step": 138 + }, + { + "epoch": 0.015069384215091066, + "grad_norm": 2.7598555088043213, + "learning_rate": 1.99717104308719e-05, + "loss": 0.1109, + "step": 139 + }, + { + "epoch": 0.015177797051170859, + "grad_norm": 2.4213955402374268, + "learning_rate": 1.997098505730451e-05, + "loss": 0.0973, + "step": 140 + }, + { + "epoch": 0.01528620988725065, + "grad_norm": 7.738302707672119, + "learning_rate": 1.9970259683737127e-05, + "loss": 0.1196, + "step": 141 + }, + { + "epoch": 0.015394622723330443, + "grad_norm": 2.914405345916748, + "learning_rate": 1.996953431016974e-05, + "loss": 0.1213, + "step": 142 + }, + { + "epoch": 0.015503035559410233, + "grad_norm": 4.754631996154785, + "learning_rate": 1.996880893660235e-05, + "loss": 0.1172, + "step": 143 + }, + { + "epoch": 0.015611448395490026, + "grad_norm": 7.275913715362549, + "learning_rate": 1.9968083563034966e-05, + "loss": 0.1161, + "step": 144 + }, + { + "epoch": 0.015719861231569817, + "grad_norm": 5.1431684494018555, + "learning_rate": 1.996735818946758e-05, + "loss": 0.1289, + "step": 145 + }, + { + "epoch": 0.01582827406764961, + "grad_norm": 4.455898761749268, + "learning_rate": 1.996663281590019e-05, + "loss": 0.1035, + "step": 146 + }, + { + "epoch": 0.015936686903729402, + "grad_norm": 2.776719570159912, + "learning_rate": 1.9965907442332803e-05, + "loss": 0.0947, + "step": 147 + }, + { + "epoch": 0.016045099739809193, + "grad_norm": 2.612802028656006, + "learning_rate": 1.9965182068765415e-05, + "loss": 0.0753, + "step": 148 + }, + { + "epoch": 0.016153512575888984, + "grad_norm": 6.958138942718506, + "learning_rate": 1.996445669519803e-05, + "loss": 0.1386, + "step": 149 + }, + { + "epoch": 0.01626192541196878, + "grad_norm": 5.706098556518555, + "learning_rate": 1.9963731321630643e-05, + "loss": 0.092, + "step": 150 + }, + { + "epoch": 0.01637033824804857, + "grad_norm": 5.760374069213867, + "learning_rate": 1.9963005948063255e-05, + "loss": 0.1177, + "step": 151 + }, + { + "epoch": 0.01647875108412836, + "grad_norm": 2.5484039783477783, + "learning_rate": 1.9962280574495867e-05, + "loss": 0.0702, + "step": 152 + }, + { + "epoch": 0.01658716392020815, + "grad_norm": 6.108762741088867, + "learning_rate": 1.996155520092848e-05, + "loss": 0.1077, + "step": 153 + }, + { + "epoch": 0.016695576756287946, + "grad_norm": 3.482775926589966, + "learning_rate": 1.996082982736109e-05, + "loss": 0.0803, + "step": 154 + }, + { + "epoch": 0.016803989592367737, + "grad_norm": 3.0431551933288574, + "learning_rate": 1.9960104453793703e-05, + "loss": 0.0835, + "step": 155 + }, + { + "epoch": 0.016912402428447527, + "grad_norm": 5.382293701171875, + "learning_rate": 1.995937908022632e-05, + "loss": 0.0983, + "step": 156 + }, + { + "epoch": 0.01702081526452732, + "grad_norm": 2.927158832550049, + "learning_rate": 1.995865370665893e-05, + "loss": 0.0604, + "step": 157 + }, + { + "epoch": 0.017129228100607113, + "grad_norm": 5.493638038635254, + "learning_rate": 1.9957928333091543e-05, + "loss": 0.0863, + "step": 158 + }, + { + "epoch": 0.017237640936686904, + "grad_norm": 11.64525032043457, + "learning_rate": 1.9957202959524155e-05, + "loss": 0.1083, + "step": 159 + }, + { + "epoch": 0.017346053772766695, + "grad_norm": 6.874933242797852, + "learning_rate": 1.9956477585956768e-05, + "loss": 0.1435, + "step": 160 + }, + { + "epoch": 0.01745446660884649, + "grad_norm": 3.414384365081787, + "learning_rate": 1.9955752212389383e-05, + "loss": 0.111, + "step": 161 + }, + { + "epoch": 0.01756287944492628, + "grad_norm": 4.065157890319824, + "learning_rate": 1.9955026838821995e-05, + "loss": 0.0704, + "step": 162 + }, + { + "epoch": 0.01767129228100607, + "grad_norm": 3.0878212451934814, + "learning_rate": 1.9954301465254607e-05, + "loss": 0.0733, + "step": 163 + }, + { + "epoch": 0.01777970511708586, + "grad_norm": 4.246405601501465, + "learning_rate": 1.9953576091687223e-05, + "loss": 0.1056, + "step": 164 + }, + { + "epoch": 0.017888117953165656, + "grad_norm": 8.949536323547363, + "learning_rate": 1.9952850718119835e-05, + "loss": 0.1116, + "step": 165 + }, + { + "epoch": 0.017996530789245447, + "grad_norm": 5.4146599769592285, + "learning_rate": 1.9952125344552447e-05, + "loss": 0.1092, + "step": 166 + }, + { + "epoch": 0.018104943625325238, + "grad_norm": 5.006230354309082, + "learning_rate": 1.995139997098506e-05, + "loss": 0.0912, + "step": 167 + }, + { + "epoch": 0.01821335646140503, + "grad_norm": 11.402140617370605, + "learning_rate": 1.995067459741767e-05, + "loss": 0.1088, + "step": 168 + }, + { + "epoch": 0.018321769297484823, + "grad_norm": 4.234574794769287, + "learning_rate": 1.9949949223850284e-05, + "loss": 0.1116, + "step": 169 + }, + { + "epoch": 0.018430182133564614, + "grad_norm": 5.938980579376221, + "learning_rate": 1.9949223850282896e-05, + "loss": 0.0777, + "step": 170 + }, + { + "epoch": 0.018538594969644405, + "grad_norm": 4.712377071380615, + "learning_rate": 1.994849847671551e-05, + "loss": 0.1172, + "step": 171 + }, + { + "epoch": 0.0186470078057242, + "grad_norm": 4.0525102615356445, + "learning_rate": 1.9947773103148123e-05, + "loss": 0.0905, + "step": 172 + }, + { + "epoch": 0.01875542064180399, + "grad_norm": 2.756293773651123, + "learning_rate": 1.9947047729580736e-05, + "loss": 0.0771, + "step": 173 + }, + { + "epoch": 0.01886383347788378, + "grad_norm": 4.854081630706787, + "learning_rate": 1.9946322356013348e-05, + "loss": 0.1218, + "step": 174 + }, + { + "epoch": 0.018972246313963572, + "grad_norm": 5.267223358154297, + "learning_rate": 1.994559698244596e-05, + "loss": 0.0935, + "step": 175 + }, + { + "epoch": 0.019080659150043366, + "grad_norm": 3.7825586795806885, + "learning_rate": 1.9944871608878572e-05, + "loss": 0.0993, + "step": 176 + }, + { + "epoch": 0.019189071986123157, + "grad_norm": 4.204143047332764, + "learning_rate": 1.9944146235311187e-05, + "loss": 0.1027, + "step": 177 + }, + { + "epoch": 0.019297484822202948, + "grad_norm": 3.0048840045928955, + "learning_rate": 1.99434208617438e-05, + "loss": 0.0732, + "step": 178 + }, + { + "epoch": 0.01940589765828274, + "grad_norm": 3.6845614910125732, + "learning_rate": 1.9942695488176412e-05, + "loss": 0.0886, + "step": 179 + }, + { + "epoch": 0.019514310494362534, + "grad_norm": 5.5694661140441895, + "learning_rate": 1.9941970114609027e-05, + "loss": 0.1106, + "step": 180 + }, + { + "epoch": 0.019622723330442324, + "grad_norm": 4.959323883056641, + "learning_rate": 1.994124474104164e-05, + "loss": 0.0804, + "step": 181 + }, + { + "epoch": 0.019731136166522115, + "grad_norm": 17.514890670776367, + "learning_rate": 1.994051936747425e-05, + "loss": 0.1091, + "step": 182 + }, + { + "epoch": 0.01983954900260191, + "grad_norm": 6.392334461212158, + "learning_rate": 1.9939793993906864e-05, + "loss": 0.112, + "step": 183 + }, + { + "epoch": 0.0199479618386817, + "grad_norm": 6.940232276916504, + "learning_rate": 1.9939068620339476e-05, + "loss": 0.0919, + "step": 184 + }, + { + "epoch": 0.02005637467476149, + "grad_norm": 5.047093391418457, + "learning_rate": 1.9938343246772088e-05, + "loss": 0.1223, + "step": 185 + }, + { + "epoch": 0.020164787510841282, + "grad_norm": 5.190830707550049, + "learning_rate": 1.9937617873204704e-05, + "loss": 0.1189, + "step": 186 + }, + { + "epoch": 0.020273200346921077, + "grad_norm": 4.248964309692383, + "learning_rate": 1.9936892499637316e-05, + "loss": 0.0934, + "step": 187 + }, + { + "epoch": 0.020381613183000868, + "grad_norm": 2.779627561569214, + "learning_rate": 1.9936167126069928e-05, + "loss": 0.0704, + "step": 188 + }, + { + "epoch": 0.02049002601908066, + "grad_norm": 5.834138870239258, + "learning_rate": 1.993544175250254e-05, + "loss": 0.1063, + "step": 189 + }, + { + "epoch": 0.02059843885516045, + "grad_norm": 4.719006538391113, + "learning_rate": 1.9934716378935152e-05, + "loss": 0.1231, + "step": 190 + }, + { + "epoch": 0.020706851691240244, + "grad_norm": 4.380193710327148, + "learning_rate": 1.9933991005367764e-05, + "loss": 0.0934, + "step": 191 + }, + { + "epoch": 0.020815264527320035, + "grad_norm": 3.0285990238189697, + "learning_rate": 1.9933265631800376e-05, + "loss": 0.0734, + "step": 192 + }, + { + "epoch": 0.020923677363399826, + "grad_norm": 4.102128505706787, + "learning_rate": 1.9932540258232992e-05, + "loss": 0.0862, + "step": 193 + }, + { + "epoch": 0.021032090199479617, + "grad_norm": 6.654309272766113, + "learning_rate": 1.9931814884665604e-05, + "loss": 0.087, + "step": 194 + }, + { + "epoch": 0.02114050303555941, + "grad_norm": 3.760178804397583, + "learning_rate": 1.9931089511098216e-05, + "loss": 0.087, + "step": 195 + }, + { + "epoch": 0.021248915871639202, + "grad_norm": 3.010437488555908, + "learning_rate": 1.9930364137530832e-05, + "loss": 0.0829, + "step": 196 + }, + { + "epoch": 0.021357328707718993, + "grad_norm": 3.431546449661255, + "learning_rate": 1.9929638763963444e-05, + "loss": 0.0853, + "step": 197 + }, + { + "epoch": 0.021465741543798787, + "grad_norm": 3.2244746685028076, + "learning_rate": 1.9928913390396056e-05, + "loss": 0.0892, + "step": 198 + }, + { + "epoch": 0.021574154379878578, + "grad_norm": 4.963876724243164, + "learning_rate": 1.9928188016828668e-05, + "loss": 0.1093, + "step": 199 + }, + { + "epoch": 0.02168256721595837, + "grad_norm": 5.741057872772217, + "learning_rate": 1.992746264326128e-05, + "loss": 0.0608, + "step": 200 + }, + { + "epoch": 0.02179098005203816, + "grad_norm": 6.672482013702393, + "learning_rate": 1.9926737269693896e-05, + "loss": 0.0974, + "step": 201 + }, + { + "epoch": 0.021899392888117954, + "grad_norm": 3.389791250228882, + "learning_rate": 1.9926011896126508e-05, + "loss": 0.0681, + "step": 202 + }, + { + "epoch": 0.022007805724197745, + "grad_norm": 7.305804252624512, + "learning_rate": 1.992528652255912e-05, + "loss": 0.0818, + "step": 203 + }, + { + "epoch": 0.022116218560277536, + "grad_norm": 3.285996913909912, + "learning_rate": 1.9924561148991732e-05, + "loss": 0.0637, + "step": 204 + }, + { + "epoch": 0.022224631396357327, + "grad_norm": 5.247317314147949, + "learning_rate": 1.9923835775424344e-05, + "loss": 0.0655, + "step": 205 + }, + { + "epoch": 0.02233304423243712, + "grad_norm": 6.186617374420166, + "learning_rate": 1.9923110401856957e-05, + "loss": 0.1013, + "step": 206 + }, + { + "epoch": 0.022441457068516912, + "grad_norm": 5.714375019073486, + "learning_rate": 1.992238502828957e-05, + "loss": 0.1227, + "step": 207 + }, + { + "epoch": 0.022549869904596703, + "grad_norm": 5.364121437072754, + "learning_rate": 1.9921659654722184e-05, + "loss": 0.0718, + "step": 208 + }, + { + "epoch": 0.022658282740676498, + "grad_norm": 3.6896731853485107, + "learning_rate": 1.9920934281154796e-05, + "loss": 0.0836, + "step": 209 + }, + { + "epoch": 0.02276669557675629, + "grad_norm": 3.2458622455596924, + "learning_rate": 1.992020890758741e-05, + "loss": 0.0697, + "step": 210 + }, + { + "epoch": 0.02287510841283608, + "grad_norm": 2.9845306873321533, + "learning_rate": 1.991948353402002e-05, + "loss": 0.0699, + "step": 211 + }, + { + "epoch": 0.02298352124891587, + "grad_norm": 5.974741458892822, + "learning_rate": 1.9918758160452633e-05, + "loss": 0.0715, + "step": 212 + }, + { + "epoch": 0.023091934084995665, + "grad_norm": 2.3922653198242188, + "learning_rate": 1.991803278688525e-05, + "loss": 0.0563, + "step": 213 + }, + { + "epoch": 0.023200346921075456, + "grad_norm": 2.8166370391845703, + "learning_rate": 1.991730741331786e-05, + "loss": 0.0808, + "step": 214 + }, + { + "epoch": 0.023308759757155247, + "grad_norm": 6.965699672698975, + "learning_rate": 1.9916582039750473e-05, + "loss": 0.061, + "step": 215 + }, + { + "epoch": 0.023417172593235037, + "grad_norm": 3.534987449645996, + "learning_rate": 1.9915856666183088e-05, + "loss": 0.058, + "step": 216 + }, + { + "epoch": 0.023525585429314832, + "grad_norm": 4.579822540283203, + "learning_rate": 1.99151312926157e-05, + "loss": 0.1366, + "step": 217 + }, + { + "epoch": 0.023633998265394623, + "grad_norm": 4.121971607208252, + "learning_rate": 1.9914405919048312e-05, + "loss": 0.0789, + "step": 218 + }, + { + "epoch": 0.023742411101474414, + "grad_norm": 5.954920768737793, + "learning_rate": 1.9913680545480925e-05, + "loss": 0.0836, + "step": 219 + }, + { + "epoch": 0.023850823937554208, + "grad_norm": 3.6247189044952393, + "learning_rate": 1.9912955171913537e-05, + "loss": 0.0665, + "step": 220 + }, + { + "epoch": 0.023959236773634, + "grad_norm": 2.6035425662994385, + "learning_rate": 1.991222979834615e-05, + "loss": 0.0539, + "step": 221 + }, + { + "epoch": 0.02406764960971379, + "grad_norm": 3.3900630474090576, + "learning_rate": 1.991150442477876e-05, + "loss": 0.093, + "step": 222 + }, + { + "epoch": 0.02417606244579358, + "grad_norm": 5.95042610168457, + "learning_rate": 1.9910779051211377e-05, + "loss": 0.0783, + "step": 223 + }, + { + "epoch": 0.024284475281873375, + "grad_norm": 7.11301326751709, + "learning_rate": 1.991005367764399e-05, + "loss": 0.0736, + "step": 224 + }, + { + "epoch": 0.024392888117953166, + "grad_norm": 7.502345561981201, + "learning_rate": 1.99093283040766e-05, + "loss": 0.0895, + "step": 225 + }, + { + "epoch": 0.024501300954032957, + "grad_norm": 3.5790371894836426, + "learning_rate": 1.9908602930509213e-05, + "loss": 0.057, + "step": 226 + }, + { + "epoch": 0.024609713790112748, + "grad_norm": 6.983285903930664, + "learning_rate": 1.9907877556941825e-05, + "loss": 0.1158, + "step": 227 + }, + { + "epoch": 0.024718126626192542, + "grad_norm": 4.131753921508789, + "learning_rate": 1.9907152183374437e-05, + "loss": 0.0735, + "step": 228 + }, + { + "epoch": 0.024826539462272333, + "grad_norm": 4.747294902801514, + "learning_rate": 1.9906426809807053e-05, + "loss": 0.0552, + "step": 229 + }, + { + "epoch": 0.024934952298352124, + "grad_norm": 2.5750882625579834, + "learning_rate": 1.9905701436239665e-05, + "loss": 0.0503, + "step": 230 + }, + { + "epoch": 0.02504336513443192, + "grad_norm": 9.119868278503418, + "learning_rate": 1.9904976062672277e-05, + "loss": 0.1401, + "step": 231 + }, + { + "epoch": 0.02515177797051171, + "grad_norm": 3.492645740509033, + "learning_rate": 1.9904250689104893e-05, + "loss": 0.0594, + "step": 232 + }, + { + "epoch": 0.0252601908065915, + "grad_norm": 6.061630725860596, + "learning_rate": 1.9903525315537505e-05, + "loss": 0.0727, + "step": 233 + }, + { + "epoch": 0.02536860364267129, + "grad_norm": 5.648778438568115, + "learning_rate": 1.9902799941970117e-05, + "loss": 0.0785, + "step": 234 + }, + { + "epoch": 0.025477016478751086, + "grad_norm": 4.189158916473389, + "learning_rate": 1.990207456840273e-05, + "loss": 0.05, + "step": 235 + }, + { + "epoch": 0.025585429314830876, + "grad_norm": 14.1614990234375, + "learning_rate": 1.990134919483534e-05, + "loss": 0.1067, + "step": 236 + }, + { + "epoch": 0.025693842150910667, + "grad_norm": 7.106956481933594, + "learning_rate": 1.9900623821267953e-05, + "loss": 0.0777, + "step": 237 + }, + { + "epoch": 0.02580225498699046, + "grad_norm": 5.662889003753662, + "learning_rate": 1.989989844770057e-05, + "loss": 0.0573, + "step": 238 + }, + { + "epoch": 0.025910667823070253, + "grad_norm": 8.461828231811523, + "learning_rate": 1.989917307413318e-05, + "loss": 0.1177, + "step": 239 + }, + { + "epoch": 0.026019080659150044, + "grad_norm": 5.564363956451416, + "learning_rate": 1.9898447700565793e-05, + "loss": 0.1045, + "step": 240 + }, + { + "epoch": 0.026127493495229834, + "grad_norm": 3.3390278816223145, + "learning_rate": 1.9897722326998405e-05, + "loss": 0.0684, + "step": 241 + }, + { + "epoch": 0.026235906331309625, + "grad_norm": 6.8690876960754395, + "learning_rate": 1.9896996953431017e-05, + "loss": 0.0758, + "step": 242 + }, + { + "epoch": 0.02634431916738942, + "grad_norm": 6.28677225112915, + "learning_rate": 1.989627157986363e-05, + "loss": 0.108, + "step": 243 + }, + { + "epoch": 0.02645273200346921, + "grad_norm": 2.8103504180908203, + "learning_rate": 1.9895546206296242e-05, + "loss": 0.0541, + "step": 244 + }, + { + "epoch": 0.026561144839549, + "grad_norm": 7.805689811706543, + "learning_rate": 1.9894820832728857e-05, + "loss": 0.1114, + "step": 245 + }, + { + "epoch": 0.026669557675628796, + "grad_norm": 4.685619354248047, + "learning_rate": 1.989409545916147e-05, + "loss": 0.0837, + "step": 246 + }, + { + "epoch": 0.026777970511708587, + "grad_norm": 3.780235528945923, + "learning_rate": 1.9893370085594085e-05, + "loss": 0.0593, + "step": 247 + }, + { + "epoch": 0.026886383347788378, + "grad_norm": 5.1039042472839355, + "learning_rate": 1.9892644712026697e-05, + "loss": 0.0576, + "step": 248 + }, + { + "epoch": 0.02699479618386817, + "grad_norm": 3.365614652633667, + "learning_rate": 1.989191933845931e-05, + "loss": 0.048, + "step": 249 + }, + { + "epoch": 0.027103209019947963, + "grad_norm": 10.01582145690918, + "learning_rate": 1.989119396489192e-05, + "loss": 0.0863, + "step": 250 + }, + { + "epoch": 0.027211621856027754, + "grad_norm": 4.323184967041016, + "learning_rate": 1.9890468591324534e-05, + "loss": 0.0458, + "step": 251 + }, + { + "epoch": 0.027320034692107545, + "grad_norm": 5.044948101043701, + "learning_rate": 1.9889743217757146e-05, + "loss": 0.0574, + "step": 252 + }, + { + "epoch": 0.027428447528187336, + "grad_norm": 7.713884353637695, + "learning_rate": 1.988901784418976e-05, + "loss": 0.071, + "step": 253 + }, + { + "epoch": 0.02753686036426713, + "grad_norm": 5.134047985076904, + "learning_rate": 1.9888292470622373e-05, + "loss": 0.0762, + "step": 254 + }, + { + "epoch": 0.02764527320034692, + "grad_norm": 6.494691848754883, + "learning_rate": 1.9887567097054986e-05, + "loss": 0.1069, + "step": 255 + }, + { + "epoch": 0.027753686036426712, + "grad_norm": 5.587743759155273, + "learning_rate": 1.9886841723487598e-05, + "loss": 0.1161, + "step": 256 + }, + { + "epoch": 0.027862098872506506, + "grad_norm": 15.214599609375, + "learning_rate": 1.988611634992021e-05, + "loss": 0.0632, + "step": 257 + }, + { + "epoch": 0.027970511708586297, + "grad_norm": 5.785831928253174, + "learning_rate": 1.9885390976352822e-05, + "loss": 0.1291, + "step": 258 + }, + { + "epoch": 0.028078924544666088, + "grad_norm": 5.567306041717529, + "learning_rate": 1.9884665602785434e-05, + "loss": 0.1238, + "step": 259 + }, + { + "epoch": 0.02818733738074588, + "grad_norm": 6.673053741455078, + "learning_rate": 1.988394022921805e-05, + "loss": 0.0767, + "step": 260 + }, + { + "epoch": 0.028295750216825673, + "grad_norm": 5.135158061981201, + "learning_rate": 1.9883214855650662e-05, + "loss": 0.0718, + "step": 261 + }, + { + "epoch": 0.028404163052905464, + "grad_norm": 3.3646395206451416, + "learning_rate": 1.9882489482083274e-05, + "loss": 0.058, + "step": 262 + }, + { + "epoch": 0.028512575888985255, + "grad_norm": 4.626059055328369, + "learning_rate": 1.9881764108515886e-05, + "loss": 0.116, + "step": 263 + }, + { + "epoch": 0.028620988725065046, + "grad_norm": 2.355438232421875, + "learning_rate": 1.9881038734948498e-05, + "loss": 0.0463, + "step": 264 + }, + { + "epoch": 0.02872940156114484, + "grad_norm": 4.319096088409424, + "learning_rate": 1.9880313361381114e-05, + "loss": 0.0792, + "step": 265 + }, + { + "epoch": 0.02883781439722463, + "grad_norm": 2.70794415473938, + "learning_rate": 1.9879587987813726e-05, + "loss": 0.067, + "step": 266 + }, + { + "epoch": 0.028946227233304422, + "grad_norm": 1.8219363689422607, + "learning_rate": 1.9878862614246338e-05, + "loss": 0.0637, + "step": 267 + }, + { + "epoch": 0.029054640069384217, + "grad_norm": 5.3458051681518555, + "learning_rate": 1.9878137240678954e-05, + "loss": 0.0679, + "step": 268 + }, + { + "epoch": 0.029163052905464008, + "grad_norm": 4.09788179397583, + "learning_rate": 1.9877411867111566e-05, + "loss": 0.0669, + "step": 269 + }, + { + "epoch": 0.0292714657415438, + "grad_norm": 3.6953063011169434, + "learning_rate": 1.9876686493544178e-05, + "loss": 0.0731, + "step": 270 + }, + { + "epoch": 0.02937987857762359, + "grad_norm": 6.547903060913086, + "learning_rate": 1.987596111997679e-05, + "loss": 0.1161, + "step": 271 + }, + { + "epoch": 0.029488291413703384, + "grad_norm": 2.5795278549194336, + "learning_rate": 1.9875235746409402e-05, + "loss": 0.0637, + "step": 272 + }, + { + "epoch": 0.029596704249783175, + "grad_norm": 6.786388397216797, + "learning_rate": 1.9874510372842014e-05, + "loss": 0.0774, + "step": 273 + }, + { + "epoch": 0.029705117085862966, + "grad_norm": 5.611429691314697, + "learning_rate": 1.9873784999274626e-05, + "loss": 0.1341, + "step": 274 + }, + { + "epoch": 0.029813529921942757, + "grad_norm": 6.21265172958374, + "learning_rate": 1.9873059625707242e-05, + "loss": 0.0981, + "step": 275 + }, + { + "epoch": 0.02992194275802255, + "grad_norm": 4.911186218261719, + "learning_rate": 1.9872334252139854e-05, + "loss": 0.1015, + "step": 276 + }, + { + "epoch": 0.030030355594102342, + "grad_norm": 7.124551773071289, + "learning_rate": 1.9871608878572466e-05, + "loss": 0.0959, + "step": 277 + }, + { + "epoch": 0.030138768430182133, + "grad_norm": 6.597419261932373, + "learning_rate": 1.987088350500508e-05, + "loss": 0.0813, + "step": 278 + }, + { + "epoch": 0.030247181266261924, + "grad_norm": 2.1101720333099365, + "learning_rate": 1.987015813143769e-05, + "loss": 0.0463, + "step": 279 + }, + { + "epoch": 0.030355594102341718, + "grad_norm": 3.5698509216308594, + "learning_rate": 1.9869432757870303e-05, + "loss": 0.0747, + "step": 280 + }, + { + "epoch": 0.03046400693842151, + "grad_norm": 3.067415952682495, + "learning_rate": 1.9868707384302918e-05, + "loss": 0.0616, + "step": 281 + }, + { + "epoch": 0.0305724197745013, + "grad_norm": 3.879551649093628, + "learning_rate": 1.986798201073553e-05, + "loss": 0.0789, + "step": 282 + }, + { + "epoch": 0.030680832610581094, + "grad_norm": 4.427328586578369, + "learning_rate": 1.9867256637168142e-05, + "loss": 0.101, + "step": 283 + }, + { + "epoch": 0.030789245446660885, + "grad_norm": 5.5939412117004395, + "learning_rate": 1.9866531263600758e-05, + "loss": 0.0579, + "step": 284 + }, + { + "epoch": 0.030897658282740676, + "grad_norm": 3.7970330715179443, + "learning_rate": 1.986580589003337e-05, + "loss": 0.1213, + "step": 285 + }, + { + "epoch": 0.031006071118820467, + "grad_norm": 4.122864246368408, + "learning_rate": 1.9865080516465982e-05, + "loss": 0.0489, + "step": 286 + }, + { + "epoch": 0.03111448395490026, + "grad_norm": 5.919523239135742, + "learning_rate": 1.9864355142898594e-05, + "loss": 0.0934, + "step": 287 + }, + { + "epoch": 0.031222896790980052, + "grad_norm": 6.688137531280518, + "learning_rate": 1.9863629769331207e-05, + "loss": 0.0788, + "step": 288 + }, + { + "epoch": 0.03133130962705984, + "grad_norm": 7.52555513381958, + "learning_rate": 1.986290439576382e-05, + "loss": 0.0861, + "step": 289 + }, + { + "epoch": 0.031439722463139634, + "grad_norm": 2.768979549407959, + "learning_rate": 1.9862179022196434e-05, + "loss": 0.0671, + "step": 290 + }, + { + "epoch": 0.031548135299219425, + "grad_norm": 3.162796974182129, + "learning_rate": 1.9861453648629046e-05, + "loss": 0.0708, + "step": 291 + }, + { + "epoch": 0.03165654813529922, + "grad_norm": 3.768096923828125, + "learning_rate": 1.986072827506166e-05, + "loss": 0.0515, + "step": 292 + }, + { + "epoch": 0.031764960971379014, + "grad_norm": 5.256648063659668, + "learning_rate": 1.986000290149427e-05, + "loss": 0.0997, + "step": 293 + }, + { + "epoch": 0.031873373807458805, + "grad_norm": 6.693769931793213, + "learning_rate": 1.9859277527926883e-05, + "loss": 0.1028, + "step": 294 + }, + { + "epoch": 0.031981786643538596, + "grad_norm": 6.232283115386963, + "learning_rate": 1.9858552154359495e-05, + "loss": 0.094, + "step": 295 + }, + { + "epoch": 0.032090199479618386, + "grad_norm": 3.2436363697052, + "learning_rate": 1.9857826780792107e-05, + "loss": 0.1082, + "step": 296 + }, + { + "epoch": 0.03219861231569818, + "grad_norm": 5.87486457824707, + "learning_rate": 1.9857101407224723e-05, + "loss": 0.0748, + "step": 297 + }, + { + "epoch": 0.03230702515177797, + "grad_norm": 4.977159023284912, + "learning_rate": 1.9856376033657335e-05, + "loss": 0.0837, + "step": 298 + }, + { + "epoch": 0.03241543798785776, + "grad_norm": 2.2429819107055664, + "learning_rate": 1.985565066008995e-05, + "loss": 0.0783, + "step": 299 + }, + { + "epoch": 0.03252385082393756, + "grad_norm": 4.399456977844238, + "learning_rate": 1.9854925286522562e-05, + "loss": 0.0792, + "step": 300 + }, + { + "epoch": 0.03263226366001735, + "grad_norm": 6.3295488357543945, + "learning_rate": 1.9854199912955175e-05, + "loss": 0.0974, + "step": 301 + }, + { + "epoch": 0.03274067649609714, + "grad_norm": 2.066760778427124, + "learning_rate": 1.9853474539387787e-05, + "loss": 0.0445, + "step": 302 + }, + { + "epoch": 0.03284908933217693, + "grad_norm": 2.716750144958496, + "learning_rate": 1.98527491658204e-05, + "loss": 0.0678, + "step": 303 + }, + { + "epoch": 0.03295750216825672, + "grad_norm": 5.346451759338379, + "learning_rate": 1.985202379225301e-05, + "loss": 0.0947, + "step": 304 + }, + { + "epoch": 0.03306591500433651, + "grad_norm": 3.0387864112854004, + "learning_rate": 1.9851298418685627e-05, + "loss": 0.065, + "step": 305 + }, + { + "epoch": 0.0331743278404163, + "grad_norm": 3.2657992839813232, + "learning_rate": 1.985057304511824e-05, + "loss": 0.0665, + "step": 306 + }, + { + "epoch": 0.0332827406764961, + "grad_norm": 3.493032455444336, + "learning_rate": 1.984984767155085e-05, + "loss": 0.0848, + "step": 307 + }, + { + "epoch": 0.03339115351257589, + "grad_norm": 3.8406379222869873, + "learning_rate": 1.9849122297983463e-05, + "loss": 0.0593, + "step": 308 + }, + { + "epoch": 0.03349956634865568, + "grad_norm": 2.161555528640747, + "learning_rate": 1.9848396924416075e-05, + "loss": 0.0541, + "step": 309 + }, + { + "epoch": 0.03360797918473547, + "grad_norm": 5.9680705070495605, + "learning_rate": 1.9847671550848687e-05, + "loss": 0.0766, + "step": 310 + }, + { + "epoch": 0.033716392020815264, + "grad_norm": 4.919120788574219, + "learning_rate": 1.98469461772813e-05, + "loss": 0.0886, + "step": 311 + }, + { + "epoch": 0.033824804856895055, + "grad_norm": 4.991422653198242, + "learning_rate": 1.9846220803713915e-05, + "loss": 0.0727, + "step": 312 + }, + { + "epoch": 0.033933217692974846, + "grad_norm": 5.512183666229248, + "learning_rate": 1.9845495430146527e-05, + "loss": 0.048, + "step": 313 + }, + { + "epoch": 0.03404163052905464, + "grad_norm": 7.753085136413574, + "learning_rate": 1.984477005657914e-05, + "loss": 0.1139, + "step": 314 + }, + { + "epoch": 0.034150043365134435, + "grad_norm": 2.4199397563934326, + "learning_rate": 1.984404468301175e-05, + "loss": 0.0519, + "step": 315 + }, + { + "epoch": 0.034258456201214225, + "grad_norm": 3.676190137863159, + "learning_rate": 1.9843319309444367e-05, + "loss": 0.0279, + "step": 316 + }, + { + "epoch": 0.034366869037294016, + "grad_norm": 9.513651847839355, + "learning_rate": 1.984259393587698e-05, + "loss": 0.1094, + "step": 317 + }, + { + "epoch": 0.03447528187337381, + "grad_norm": 1.0867362022399902, + "learning_rate": 1.984186856230959e-05, + "loss": 0.0272, + "step": 318 + }, + { + "epoch": 0.0345836947094536, + "grad_norm": 1.8090728521347046, + "learning_rate": 1.9841143188742203e-05, + "loss": 0.0229, + "step": 319 + }, + { + "epoch": 0.03469210754553339, + "grad_norm": 8.054291725158691, + "learning_rate": 1.984041781517482e-05, + "loss": 0.0746, + "step": 320 + }, + { + "epoch": 0.03480052038161318, + "grad_norm": 11.40935230255127, + "learning_rate": 1.983969244160743e-05, + "loss": 0.0362, + "step": 321 + }, + { + "epoch": 0.03490893321769298, + "grad_norm": 11.964215278625488, + "learning_rate": 1.9838967068040043e-05, + "loss": 0.0634, + "step": 322 + }, + { + "epoch": 0.03501734605377277, + "grad_norm": 9.50049877166748, + "learning_rate": 1.9838241694472655e-05, + "loss": 0.0211, + "step": 323 + }, + { + "epoch": 0.03512575888985256, + "grad_norm": 8.241752624511719, + "learning_rate": 1.9837516320905267e-05, + "loss": 0.0287, + "step": 324 + }, + { + "epoch": 0.03523417172593235, + "grad_norm": 7.938928127288818, + "learning_rate": 1.983679094733788e-05, + "loss": 0.1318, + "step": 325 + }, + { + "epoch": 0.03534258456201214, + "grad_norm": 12.558661460876465, + "learning_rate": 1.9836065573770492e-05, + "loss": 0.0659, + "step": 326 + }, + { + "epoch": 0.03545099739809193, + "grad_norm": 11.722322463989258, + "learning_rate": 1.9835340200203107e-05, + "loss": 0.0958, + "step": 327 + }, + { + "epoch": 0.03555941023417172, + "grad_norm": 2.9864628314971924, + "learning_rate": 1.983461482663572e-05, + "loss": 0.0847, + "step": 328 + }, + { + "epoch": 0.03566782307025152, + "grad_norm": 15.963151931762695, + "learning_rate": 1.983388945306833e-05, + "loss": 0.1643, + "step": 329 + }, + { + "epoch": 0.03577623590633131, + "grad_norm": 10.570533752441406, + "learning_rate": 1.9833164079500944e-05, + "loss": 0.0838, + "step": 330 + }, + { + "epoch": 0.0358846487424111, + "grad_norm": 2.6846280097961426, + "learning_rate": 1.9832438705933556e-05, + "loss": 0.0195, + "step": 331 + }, + { + "epoch": 0.035993061578490894, + "grad_norm": 6.996094226837158, + "learning_rate": 1.9831713332366168e-05, + "loss": 0.0571, + "step": 332 + }, + { + "epoch": 0.036101474414570685, + "grad_norm": 6.028219699859619, + "learning_rate": 1.9830987958798784e-05, + "loss": 0.0834, + "step": 333 + }, + { + "epoch": 0.036209887250650476, + "grad_norm": 8.418713569641113, + "learning_rate": 1.9830262585231396e-05, + "loss": 0.1099, + "step": 334 + }, + { + "epoch": 0.03631830008673027, + "grad_norm": 8.341958999633789, + "learning_rate": 1.9829537211664008e-05, + "loss": 0.0387, + "step": 335 + }, + { + "epoch": 0.03642671292281006, + "grad_norm": 13.83373737335205, + "learning_rate": 1.9828811838096623e-05, + "loss": 0.0798, + "step": 336 + }, + { + "epoch": 0.036535125758889855, + "grad_norm": 7.802219867706299, + "learning_rate": 1.9828086464529235e-05, + "loss": 0.1124, + "step": 337 + }, + { + "epoch": 0.036643538594969646, + "grad_norm": 4.741801738739014, + "learning_rate": 1.9827361090961848e-05, + "loss": 0.087, + "step": 338 + }, + { + "epoch": 0.03675195143104944, + "grad_norm": 3.0522842407226562, + "learning_rate": 1.982663571739446e-05, + "loss": 0.055, + "step": 339 + }, + { + "epoch": 0.03686036426712923, + "grad_norm": 8.003628730773926, + "learning_rate": 1.9825910343827072e-05, + "loss": 0.0995, + "step": 340 + }, + { + "epoch": 0.03696877710320902, + "grad_norm": 6.298898696899414, + "learning_rate": 1.9825184970259684e-05, + "loss": 0.0808, + "step": 341 + }, + { + "epoch": 0.03707718993928881, + "grad_norm": 3.477252721786499, + "learning_rate": 1.98244595966923e-05, + "loss": 0.0408, + "step": 342 + }, + { + "epoch": 0.0371856027753686, + "grad_norm": 4.183280944824219, + "learning_rate": 1.9823734223124912e-05, + "loss": 0.096, + "step": 343 + }, + { + "epoch": 0.0372940156114484, + "grad_norm": 4.141633987426758, + "learning_rate": 1.9823008849557524e-05, + "loss": 0.0582, + "step": 344 + }, + { + "epoch": 0.03740242844752819, + "grad_norm": 7.375187873840332, + "learning_rate": 1.9822283475990136e-05, + "loss": 0.101, + "step": 345 + }, + { + "epoch": 0.03751084128360798, + "grad_norm": 3.477104663848877, + "learning_rate": 1.9821558102422748e-05, + "loss": 0.0621, + "step": 346 + }, + { + "epoch": 0.03761925411968777, + "grad_norm": 4.132571697235107, + "learning_rate": 1.982083272885536e-05, + "loss": 0.0814, + "step": 347 + }, + { + "epoch": 0.03772766695576756, + "grad_norm": 3.9397168159484863, + "learning_rate": 1.9820107355287976e-05, + "loss": 0.0759, + "step": 348 + }, + { + "epoch": 0.03783607979184735, + "grad_norm": 3.8670706748962402, + "learning_rate": 1.9819381981720588e-05, + "loss": 0.0641, + "step": 349 + }, + { + "epoch": 0.037944492627927144, + "grad_norm": 3.641822576522827, + "learning_rate": 1.98186566081532e-05, + "loss": 0.0571, + "step": 350 + }, + { + "epoch": 0.038052905464006935, + "grad_norm": 3.788414478302002, + "learning_rate": 1.9817931234585816e-05, + "loss": 0.0879, + "step": 351 + }, + { + "epoch": 0.03816131830008673, + "grad_norm": 3.0895161628723145, + "learning_rate": 1.9817205861018428e-05, + "loss": 0.0521, + "step": 352 + }, + { + "epoch": 0.038269731136166524, + "grad_norm": 3.770545721054077, + "learning_rate": 1.981648048745104e-05, + "loss": 0.0421, + "step": 353 + }, + { + "epoch": 0.038378143972246315, + "grad_norm": 2.915156364440918, + "learning_rate": 1.9815755113883652e-05, + "loss": 0.063, + "step": 354 + }, + { + "epoch": 0.038486556808326106, + "grad_norm": 4.046220302581787, + "learning_rate": 1.9815029740316264e-05, + "loss": 0.063, + "step": 355 + }, + { + "epoch": 0.038594969644405897, + "grad_norm": 2.3924717903137207, + "learning_rate": 1.9814304366748876e-05, + "loss": 0.0568, + "step": 356 + }, + { + "epoch": 0.03870338248048569, + "grad_norm": 5.0907158851623535, + "learning_rate": 1.9813578993181492e-05, + "loss": 0.0815, + "step": 357 + }, + { + "epoch": 0.03881179531656548, + "grad_norm": 7.582966327667236, + "learning_rate": 1.9812853619614104e-05, + "loss": 0.1031, + "step": 358 + }, + { + "epoch": 0.038920208152645276, + "grad_norm": 3.1146609783172607, + "learning_rate": 1.9812128246046716e-05, + "loss": 0.0766, + "step": 359 + }, + { + "epoch": 0.03902862098872507, + "grad_norm": 4.987020969390869, + "learning_rate": 1.981140287247933e-05, + "loss": 0.1124, + "step": 360 + }, + { + "epoch": 0.03913703382480486, + "grad_norm": 5.227577209472656, + "learning_rate": 1.981067749891194e-05, + "loss": 0.0722, + "step": 361 + }, + { + "epoch": 0.03924544666088465, + "grad_norm": 2.244252920150757, + "learning_rate": 1.9809952125344553e-05, + "loss": 0.0416, + "step": 362 + }, + { + "epoch": 0.03935385949696444, + "grad_norm": 3.5093610286712646, + "learning_rate": 1.9809226751777165e-05, + "loss": 0.0612, + "step": 363 + }, + { + "epoch": 0.03946227233304423, + "grad_norm": 7.728423595428467, + "learning_rate": 1.980850137820978e-05, + "loss": 0.0826, + "step": 364 + }, + { + "epoch": 0.03957068516912402, + "grad_norm": 6.105544090270996, + "learning_rate": 1.9807776004642392e-05, + "loss": 0.111, + "step": 365 + }, + { + "epoch": 0.03967909800520382, + "grad_norm": 2.4621012210845947, + "learning_rate": 1.9807050631075005e-05, + "loss": 0.0644, + "step": 366 + }, + { + "epoch": 0.03978751084128361, + "grad_norm": 4.910451889038086, + "learning_rate": 1.9806325257507617e-05, + "loss": 0.0932, + "step": 367 + }, + { + "epoch": 0.0398959236773634, + "grad_norm": 6.199349403381348, + "learning_rate": 1.9805599883940232e-05, + "loss": 0.0868, + "step": 368 + }, + { + "epoch": 0.04000433651344319, + "grad_norm": 3.958813190460205, + "learning_rate": 1.9804874510372844e-05, + "loss": 0.0777, + "step": 369 + }, + { + "epoch": 0.04011274934952298, + "grad_norm": 3.3823471069335938, + "learning_rate": 1.9804149136805457e-05, + "loss": 0.0442, + "step": 370 + }, + { + "epoch": 0.040221162185602774, + "grad_norm": 3.0546321868896484, + "learning_rate": 1.980342376323807e-05, + "loss": 0.0596, + "step": 371 + }, + { + "epoch": 0.040329575021682565, + "grad_norm": 5.694910049438477, + "learning_rate": 1.9802698389670684e-05, + "loss": 0.0984, + "step": 372 + }, + { + "epoch": 0.040437987857762356, + "grad_norm": 1.9845010042190552, + "learning_rate": 1.9801973016103296e-05, + "loss": 0.0391, + "step": 373 + }, + { + "epoch": 0.040546400693842154, + "grad_norm": 2.035719156265259, + "learning_rate": 1.980124764253591e-05, + "loss": 0.0391, + "step": 374 + }, + { + "epoch": 0.040654813529921945, + "grad_norm": 6.13949728012085, + "learning_rate": 1.980052226896852e-05, + "loss": 0.1316, + "step": 375 + }, + { + "epoch": 0.040763226366001735, + "grad_norm": 4.683819770812988, + "learning_rate": 1.9799796895401133e-05, + "loss": 0.0841, + "step": 376 + }, + { + "epoch": 0.040871639202081526, + "grad_norm": 5.2730889320373535, + "learning_rate": 1.9799071521833745e-05, + "loss": 0.1194, + "step": 377 + }, + { + "epoch": 0.04098005203816132, + "grad_norm": 2.07407808303833, + "learning_rate": 1.9798346148266357e-05, + "loss": 0.0348, + "step": 378 + }, + { + "epoch": 0.04108846487424111, + "grad_norm": 6.538453578948975, + "learning_rate": 1.9797620774698973e-05, + "loss": 0.098, + "step": 379 + }, + { + "epoch": 0.0411968777103209, + "grad_norm": 6.024566173553467, + "learning_rate": 1.9796895401131585e-05, + "loss": 0.1334, + "step": 380 + }, + { + "epoch": 0.0413052905464007, + "grad_norm": 3.00349497795105, + "learning_rate": 1.9796170027564197e-05, + "loss": 0.0708, + "step": 381 + }, + { + "epoch": 0.04141370338248049, + "grad_norm": 5.827469348907471, + "learning_rate": 1.979544465399681e-05, + "loss": 0.0747, + "step": 382 + }, + { + "epoch": 0.04152211621856028, + "grad_norm": 6.623713493347168, + "learning_rate": 1.979471928042942e-05, + "loss": 0.0418, + "step": 383 + }, + { + "epoch": 0.04163052905464007, + "grad_norm": 2.954470634460449, + "learning_rate": 1.9793993906862033e-05, + "loss": 0.0534, + "step": 384 + }, + { + "epoch": 0.04173894189071986, + "grad_norm": 4.468087196350098, + "learning_rate": 1.979326853329465e-05, + "loss": 0.0831, + "step": 385 + }, + { + "epoch": 0.04184735472679965, + "grad_norm": 9.48854923248291, + "learning_rate": 1.979254315972726e-05, + "loss": 0.0782, + "step": 386 + }, + { + "epoch": 0.04195576756287944, + "grad_norm": 4.304708003997803, + "learning_rate": 1.9791817786159873e-05, + "loss": 0.07, + "step": 387 + }, + { + "epoch": 0.04206418039895923, + "grad_norm": 3.810044288635254, + "learning_rate": 1.979109241259249e-05, + "loss": 0.0509, + "step": 388 + }, + { + "epoch": 0.04217259323503903, + "grad_norm": 5.7801899909973145, + "learning_rate": 1.97903670390251e-05, + "loss": 0.0804, + "step": 389 + }, + { + "epoch": 0.04228100607111882, + "grad_norm": 4.124120712280273, + "learning_rate": 1.9789641665457713e-05, + "loss": 0.0773, + "step": 390 + }, + { + "epoch": 0.04238941890719861, + "grad_norm": 3.7026572227478027, + "learning_rate": 1.9788916291890325e-05, + "loss": 0.0693, + "step": 391 + }, + { + "epoch": 0.042497831743278404, + "grad_norm": 5.677544116973877, + "learning_rate": 1.9788190918322937e-05, + "loss": 0.0833, + "step": 392 + }, + { + "epoch": 0.042606244579358195, + "grad_norm": 2.901461362838745, + "learning_rate": 1.978746554475555e-05, + "loss": 0.0884, + "step": 393 + }, + { + "epoch": 0.042714657415437986, + "grad_norm": 4.6867451667785645, + "learning_rate": 1.9786740171188165e-05, + "loss": 0.1124, + "step": 394 + }, + { + "epoch": 0.04282307025151778, + "grad_norm": 4.082432746887207, + "learning_rate": 1.9786014797620777e-05, + "loss": 0.0887, + "step": 395 + }, + { + "epoch": 0.042931483087597574, + "grad_norm": 4.943412780761719, + "learning_rate": 1.978528942405339e-05, + "loss": 0.1167, + "step": 396 + }, + { + "epoch": 0.043039895923677365, + "grad_norm": 2.115537643432617, + "learning_rate": 1.9784564050486e-05, + "loss": 0.0683, + "step": 397 + }, + { + "epoch": 0.043148308759757156, + "grad_norm": 6.901070594787598, + "learning_rate": 1.9783838676918613e-05, + "loss": 0.0836, + "step": 398 + }, + { + "epoch": 0.04325672159583695, + "grad_norm": 2.825960159301758, + "learning_rate": 1.9783113303351226e-05, + "loss": 0.0784, + "step": 399 + }, + { + "epoch": 0.04336513443191674, + "grad_norm": 3.2459776401519775, + "learning_rate": 1.978238792978384e-05, + "loss": 0.1007, + "step": 400 + }, + { + "epoch": 0.04347354726799653, + "grad_norm": 3.9339890480041504, + "learning_rate": 1.9781662556216453e-05, + "loss": 0.0671, + "step": 401 + }, + { + "epoch": 0.04358196010407632, + "grad_norm": 2.077697515487671, + "learning_rate": 1.9780937182649065e-05, + "loss": 0.0547, + "step": 402 + }, + { + "epoch": 0.04369037294015612, + "grad_norm": 2.504938840866089, + "learning_rate": 1.978021180908168e-05, + "loss": 0.0737, + "step": 403 + }, + { + "epoch": 0.04379878577623591, + "grad_norm": 2.990108013153076, + "learning_rate": 1.9779486435514293e-05, + "loss": 0.1002, + "step": 404 + }, + { + "epoch": 0.0439071986123157, + "grad_norm": 3.30985951423645, + "learning_rate": 1.9778761061946905e-05, + "loss": 0.1042, + "step": 405 + }, + { + "epoch": 0.04401561144839549, + "grad_norm": 2.16667103767395, + "learning_rate": 1.9778035688379517e-05, + "loss": 0.0527, + "step": 406 + }, + { + "epoch": 0.04412402428447528, + "grad_norm": 1.982399821281433, + "learning_rate": 1.977731031481213e-05, + "loss": 0.0594, + "step": 407 + }, + { + "epoch": 0.04423243712055507, + "grad_norm": 2.3149046897888184, + "learning_rate": 1.9776584941244742e-05, + "loss": 0.0592, + "step": 408 + }, + { + "epoch": 0.04434084995663486, + "grad_norm": 6.455365180969238, + "learning_rate": 1.9775859567677357e-05, + "loss": 0.1155, + "step": 409 + }, + { + "epoch": 0.044449262792714654, + "grad_norm": 1.767106533050537, + "learning_rate": 1.977513419410997e-05, + "loss": 0.0403, + "step": 410 + }, + { + "epoch": 0.04455767562879445, + "grad_norm": 3.998359203338623, + "learning_rate": 1.977440882054258e-05, + "loss": 0.0968, + "step": 411 + }, + { + "epoch": 0.04466608846487424, + "grad_norm": 4.433526515960693, + "learning_rate": 1.9773683446975194e-05, + "loss": 0.093, + "step": 412 + }, + { + "epoch": 0.044774501300954034, + "grad_norm": 6.254210472106934, + "learning_rate": 1.9772958073407806e-05, + "loss": 0.0568, + "step": 413 + }, + { + "epoch": 0.044882914137033825, + "grad_norm": 4.305883884429932, + "learning_rate": 1.9772232699840418e-05, + "loss": 0.0574, + "step": 414 + }, + { + "epoch": 0.044991326973113616, + "grad_norm": 7.75441312789917, + "learning_rate": 1.977150732627303e-05, + "loss": 0.0846, + "step": 415 + }, + { + "epoch": 0.045099739809193407, + "grad_norm": 7.497109413146973, + "learning_rate": 1.9770781952705646e-05, + "loss": 0.0741, + "step": 416 + }, + { + "epoch": 0.0452081526452732, + "grad_norm": 6.334640979766846, + "learning_rate": 1.9770056579138258e-05, + "loss": 0.0519, + "step": 417 + }, + { + "epoch": 0.045316565481352995, + "grad_norm": 4.485202789306641, + "learning_rate": 1.976933120557087e-05, + "loss": 0.0817, + "step": 418 + }, + { + "epoch": 0.045424978317432786, + "grad_norm": 3.873929023742676, + "learning_rate": 1.9768605832003482e-05, + "loss": 0.0519, + "step": 419 + }, + { + "epoch": 0.04553339115351258, + "grad_norm": 6.467846870422363, + "learning_rate": 1.9767880458436098e-05, + "loss": 0.0481, + "step": 420 + }, + { + "epoch": 0.04564180398959237, + "grad_norm": 5.588794708251953, + "learning_rate": 1.976715508486871e-05, + "loss": 0.076, + "step": 421 + }, + { + "epoch": 0.04575021682567216, + "grad_norm": 9.920389175415039, + "learning_rate": 1.9766429711301322e-05, + "loss": 0.0495, + "step": 422 + }, + { + "epoch": 0.04585862966175195, + "grad_norm": 7.464017868041992, + "learning_rate": 1.9765704337733934e-05, + "loss": 0.0736, + "step": 423 + }, + { + "epoch": 0.04596704249783174, + "grad_norm": 5.463119029998779, + "learning_rate": 1.976497896416655e-05, + "loss": 0.0825, + "step": 424 + }, + { + "epoch": 0.04607545533391154, + "grad_norm": 5.545544624328613, + "learning_rate": 1.976425359059916e-05, + "loss": 0.0411, + "step": 425 + }, + { + "epoch": 0.04618386816999133, + "grad_norm": 4.828194618225098, + "learning_rate": 1.9763528217031774e-05, + "loss": 0.0583, + "step": 426 + }, + { + "epoch": 0.04629228100607112, + "grad_norm": 5.939785003662109, + "learning_rate": 1.9762802843464386e-05, + "loss": 0.081, + "step": 427 + }, + { + "epoch": 0.04640069384215091, + "grad_norm": 2.1533069610595703, + "learning_rate": 1.9762077469896998e-05, + "loss": 0.0509, + "step": 428 + }, + { + "epoch": 0.0465091066782307, + "grad_norm": 13.268999099731445, + "learning_rate": 1.976135209632961e-05, + "loss": 0.0952, + "step": 429 + }, + { + "epoch": 0.04661751951431049, + "grad_norm": 3.7655632495880127, + "learning_rate": 1.9760626722762222e-05, + "loss": 0.0735, + "step": 430 + }, + { + "epoch": 0.046725932350390284, + "grad_norm": 6.520329475402832, + "learning_rate": 1.9759901349194838e-05, + "loss": 0.0749, + "step": 431 + }, + { + "epoch": 0.046834345186470075, + "grad_norm": 6.873312950134277, + "learning_rate": 1.975917597562745e-05, + "loss": 0.0694, + "step": 432 + }, + { + "epoch": 0.04694275802254987, + "grad_norm": 4.098426342010498, + "learning_rate": 1.9758450602060062e-05, + "loss": 0.0706, + "step": 433 + }, + { + "epoch": 0.047051170858629664, + "grad_norm": 7.48695707321167, + "learning_rate": 1.9757725228492674e-05, + "loss": 0.0562, + "step": 434 + }, + { + "epoch": 0.047159583694709455, + "grad_norm": 4.258584499359131, + "learning_rate": 1.9756999854925287e-05, + "loss": 0.0494, + "step": 435 + }, + { + "epoch": 0.047267996530789246, + "grad_norm": 6.777398109436035, + "learning_rate": 1.97562744813579e-05, + "loss": 0.0566, + "step": 436 + }, + { + "epoch": 0.047376409366869036, + "grad_norm": 8.245614051818848, + "learning_rate": 1.9755549107790514e-05, + "loss": 0.0579, + "step": 437 + }, + { + "epoch": 0.04748482220294883, + "grad_norm": 3.2472918033599854, + "learning_rate": 1.9754823734223126e-05, + "loss": 0.0945, + "step": 438 + }, + { + "epoch": 0.04759323503902862, + "grad_norm": 7.802513599395752, + "learning_rate": 1.975409836065574e-05, + "loss": 0.0622, + "step": 439 + }, + { + "epoch": 0.047701647875108416, + "grad_norm": 5.366639137268066, + "learning_rate": 1.9753372987088354e-05, + "loss": 0.0554, + "step": 440 + }, + { + "epoch": 0.04781006071118821, + "grad_norm": 5.972220420837402, + "learning_rate": 1.9752647613520966e-05, + "loss": 0.0989, + "step": 441 + }, + { + "epoch": 0.047918473547268, + "grad_norm": 6.57342004776001, + "learning_rate": 1.9751922239953578e-05, + "loss": 0.0392, + "step": 442 + }, + { + "epoch": 0.04802688638334779, + "grad_norm": 4.104686737060547, + "learning_rate": 1.975119686638619e-05, + "loss": 0.0738, + "step": 443 + }, + { + "epoch": 0.04813529921942758, + "grad_norm": 3.742647409439087, + "learning_rate": 1.9750471492818803e-05, + "loss": 0.0742, + "step": 444 + }, + { + "epoch": 0.04824371205550737, + "grad_norm": 4.948635101318359, + "learning_rate": 1.9749746119251415e-05, + "loss": 0.0874, + "step": 445 + }, + { + "epoch": 0.04835212489158716, + "grad_norm": 4.6274237632751465, + "learning_rate": 1.974902074568403e-05, + "loss": 0.0777, + "step": 446 + }, + { + "epoch": 0.04846053772766695, + "grad_norm": 2.1932971477508545, + "learning_rate": 1.9748295372116642e-05, + "loss": 0.0654, + "step": 447 + }, + { + "epoch": 0.04856895056374675, + "grad_norm": 3.6179111003875732, + "learning_rate": 1.9747569998549255e-05, + "loss": 0.0656, + "step": 448 + }, + { + "epoch": 0.04867736339982654, + "grad_norm": 4.794009685516357, + "learning_rate": 1.9746844624981867e-05, + "loss": 0.0439, + "step": 449 + }, + { + "epoch": 0.04878577623590633, + "grad_norm": 1.480925440788269, + "learning_rate": 1.974611925141448e-05, + "loss": 0.0404, + "step": 450 + }, + { + "epoch": 0.04889418907198612, + "grad_norm": 2.873109817504883, + "learning_rate": 1.974539387784709e-05, + "loss": 0.1014, + "step": 451 + }, + { + "epoch": 0.049002601908065914, + "grad_norm": 2.3841991424560547, + "learning_rate": 1.9744668504279707e-05, + "loss": 0.0712, + "step": 452 + }, + { + "epoch": 0.049111014744145705, + "grad_norm": 2.099700689315796, + "learning_rate": 1.974394313071232e-05, + "loss": 0.0354, + "step": 453 + }, + { + "epoch": 0.049219427580225496, + "grad_norm": 4.729235649108887, + "learning_rate": 1.974321775714493e-05, + "loss": 0.0852, + "step": 454 + }, + { + "epoch": 0.049327840416305294, + "grad_norm": 2.9683735370635986, + "learning_rate": 1.9742492383577546e-05, + "loss": 0.0536, + "step": 455 + }, + { + "epoch": 0.049436253252385085, + "grad_norm": 3.2920548915863037, + "learning_rate": 1.974176701001016e-05, + "loss": 0.0347, + "step": 456 + }, + { + "epoch": 0.049544666088464875, + "grad_norm": 4.203485488891602, + "learning_rate": 1.974104163644277e-05, + "loss": 0.0545, + "step": 457 + }, + { + "epoch": 0.049653078924544666, + "grad_norm": 6.228875160217285, + "learning_rate": 1.9740316262875383e-05, + "loss": 0.0449, + "step": 458 + }, + { + "epoch": 0.04976149176062446, + "grad_norm": 8.839595794677734, + "learning_rate": 1.9739590889307995e-05, + "loss": 0.1109, + "step": 459 + }, + { + "epoch": 0.04986990459670425, + "grad_norm": 3.2218027114868164, + "learning_rate": 1.9738865515740607e-05, + "loss": 0.0777, + "step": 460 + }, + { + "epoch": 0.04997831743278404, + "grad_norm": 2.272939443588257, + "learning_rate": 1.9738140142173223e-05, + "loss": 0.0394, + "step": 461 + }, + { + "epoch": 0.05008673026886384, + "grad_norm": 5.112960338592529, + "learning_rate": 1.9737414768605835e-05, + "loss": 0.0412, + "step": 462 + }, + { + "epoch": 0.05019514310494363, + "grad_norm": 6.045676231384277, + "learning_rate": 1.9736689395038447e-05, + "loss": 0.1132, + "step": 463 + }, + { + "epoch": 0.05030355594102342, + "grad_norm": 4.716992378234863, + "learning_rate": 1.973596402147106e-05, + "loss": 0.0638, + "step": 464 + }, + { + "epoch": 0.05041196877710321, + "grad_norm": 8.401906967163086, + "learning_rate": 1.973523864790367e-05, + "loss": 0.0733, + "step": 465 + }, + { + "epoch": 0.050520381613183, + "grad_norm": 2.9713199138641357, + "learning_rate": 1.9734513274336283e-05, + "loss": 0.0566, + "step": 466 + }, + { + "epoch": 0.05062879444926279, + "grad_norm": 6.73055362701416, + "learning_rate": 1.9733787900768895e-05, + "loss": 0.0396, + "step": 467 + }, + { + "epoch": 0.05073720728534258, + "grad_norm": 3.666384696960449, + "learning_rate": 1.973306252720151e-05, + "loss": 0.0409, + "step": 468 + }, + { + "epoch": 0.05084562012142237, + "grad_norm": 3.186624765396118, + "learning_rate": 1.9732337153634123e-05, + "loss": 0.0324, + "step": 469 + }, + { + "epoch": 0.05095403295750217, + "grad_norm": 6.8318681716918945, + "learning_rate": 1.9731611780066735e-05, + "loss": 0.1143, + "step": 470 + }, + { + "epoch": 0.05106244579358196, + "grad_norm": 6.481635570526123, + "learning_rate": 1.9730886406499347e-05, + "loss": 0.0623, + "step": 471 + }, + { + "epoch": 0.05117085862966175, + "grad_norm": 4.126763343811035, + "learning_rate": 1.9730161032931963e-05, + "loss": 0.0471, + "step": 472 + }, + { + "epoch": 0.051279271465741544, + "grad_norm": 6.825005531311035, + "learning_rate": 1.9729435659364575e-05, + "loss": 0.0471, + "step": 473 + }, + { + "epoch": 0.051387684301821335, + "grad_norm": 2.9673898220062256, + "learning_rate": 1.9728710285797187e-05, + "loss": 0.0324, + "step": 474 + }, + { + "epoch": 0.051496097137901126, + "grad_norm": 6.466978549957275, + "learning_rate": 1.97279849122298e-05, + "loss": 0.0504, + "step": 475 + }, + { + "epoch": 0.05160450997398092, + "grad_norm": 2.8388566970825195, + "learning_rate": 1.9727259538662415e-05, + "loss": 0.066, + "step": 476 + }, + { + "epoch": 0.051712922810060714, + "grad_norm": 5.945403099060059, + "learning_rate": 1.9726534165095027e-05, + "loss": 0.0639, + "step": 477 + }, + { + "epoch": 0.051821335646140505, + "grad_norm": 3.2252559661865234, + "learning_rate": 1.972580879152764e-05, + "loss": 0.0532, + "step": 478 + }, + { + "epoch": 0.051929748482220296, + "grad_norm": 7.376084804534912, + "learning_rate": 1.972508341796025e-05, + "loss": 0.0983, + "step": 479 + }, + { + "epoch": 0.05203816131830009, + "grad_norm": 1.6793725490570068, + "learning_rate": 1.9724358044392863e-05, + "loss": 0.0299, + "step": 480 + }, + { + "epoch": 0.05214657415437988, + "grad_norm": 1.5764626264572144, + "learning_rate": 1.9723632670825476e-05, + "loss": 0.0151, + "step": 481 + }, + { + "epoch": 0.05225498699045967, + "grad_norm": 5.8865275382995605, + "learning_rate": 1.9722907297258088e-05, + "loss": 0.1303, + "step": 482 + }, + { + "epoch": 0.05236339982653946, + "grad_norm": 3.4569175243377686, + "learning_rate": 1.9722181923690703e-05, + "loss": 0.0681, + "step": 483 + }, + { + "epoch": 0.05247181266261925, + "grad_norm": 6.050510883331299, + "learning_rate": 1.9721456550123315e-05, + "loss": 0.1042, + "step": 484 + }, + { + "epoch": 0.05258022549869905, + "grad_norm": 6.143428325653076, + "learning_rate": 1.9720731176555928e-05, + "loss": 0.0747, + "step": 485 + }, + { + "epoch": 0.05268863833477884, + "grad_norm": 6.87505578994751, + "learning_rate": 1.972000580298854e-05, + "loss": 0.0834, + "step": 486 + }, + { + "epoch": 0.05279705117085863, + "grad_norm": 6.587521076202393, + "learning_rate": 1.9719280429421152e-05, + "loss": 0.039, + "step": 487 + }, + { + "epoch": 0.05290546400693842, + "grad_norm": 2.0380136966705322, + "learning_rate": 1.9718555055853764e-05, + "loss": 0.0446, + "step": 488 + }, + { + "epoch": 0.05301387684301821, + "grad_norm": 2.178363800048828, + "learning_rate": 1.971782968228638e-05, + "loss": 0.0607, + "step": 489 + }, + { + "epoch": 0.053122289679098, + "grad_norm": 8.855755805969238, + "learning_rate": 1.971710430871899e-05, + "loss": 0.1173, + "step": 490 + }, + { + "epoch": 0.053230702515177794, + "grad_norm": 7.235390663146973, + "learning_rate": 1.9716378935151604e-05, + "loss": 0.0718, + "step": 491 + }, + { + "epoch": 0.05333911535125759, + "grad_norm": 3.0307109355926514, + "learning_rate": 1.971565356158422e-05, + "loss": 0.0504, + "step": 492 + }, + { + "epoch": 0.05344752818733738, + "grad_norm": 4.262338161468506, + "learning_rate": 1.971492818801683e-05, + "loss": 0.0484, + "step": 493 + }, + { + "epoch": 0.053555941023417174, + "grad_norm": 6.928067684173584, + "learning_rate": 1.9714202814449444e-05, + "loss": 0.4035, + "step": 494 + }, + { + "epoch": 0.053664353859496965, + "grad_norm": 4.0397186279296875, + "learning_rate": 1.9713477440882056e-05, + "loss": 0.0385, + "step": 495 + }, + { + "epoch": 0.053772766695576756, + "grad_norm": 2.876046657562256, + "learning_rate": 1.9712752067314668e-05, + "loss": 0.0652, + "step": 496 + }, + { + "epoch": 0.053881179531656546, + "grad_norm": 7.431445121765137, + "learning_rate": 1.971202669374728e-05, + "loss": 0.07, + "step": 497 + }, + { + "epoch": 0.05398959236773634, + "grad_norm": 4.297962665557861, + "learning_rate": 1.9711301320179896e-05, + "loss": 0.074, + "step": 498 + }, + { + "epoch": 0.054098005203816135, + "grad_norm": 3.5256924629211426, + "learning_rate": 1.9710575946612508e-05, + "loss": 0.0545, + "step": 499 + }, + { + "epoch": 0.054206418039895926, + "grad_norm": 4.3503642082214355, + "learning_rate": 1.970985057304512e-05, + "loss": 0.0667, + "step": 500 + }, + { + "epoch": 0.05431483087597572, + "grad_norm": 4.249512672424316, + "learning_rate": 1.9709125199477732e-05, + "loss": 0.0558, + "step": 501 + }, + { + "epoch": 0.05442324371205551, + "grad_norm": 4.847731590270996, + "learning_rate": 1.9708399825910344e-05, + "loss": 0.0855, + "step": 502 + }, + { + "epoch": 0.0545316565481353, + "grad_norm": 6.177563190460205, + "learning_rate": 1.9707674452342956e-05, + "loss": 0.0944, + "step": 503 + }, + { + "epoch": 0.05464006938421509, + "grad_norm": 2.8983423709869385, + "learning_rate": 1.9706949078775572e-05, + "loss": 0.043, + "step": 504 + }, + { + "epoch": 0.05474848222029488, + "grad_norm": 2.1377108097076416, + "learning_rate": 1.9706223705208184e-05, + "loss": 0.0397, + "step": 505 + }, + { + "epoch": 0.05485689505637467, + "grad_norm": 6.598865032196045, + "learning_rate": 1.9705498331640796e-05, + "loss": 0.1077, + "step": 506 + }, + { + "epoch": 0.05496530789245447, + "grad_norm": 5.815805435180664, + "learning_rate": 1.970477295807341e-05, + "loss": 0.054, + "step": 507 + }, + { + "epoch": 0.05507372072853426, + "grad_norm": 3.6762139797210693, + "learning_rate": 1.9704047584506024e-05, + "loss": 0.041, + "step": 508 + }, + { + "epoch": 0.05518213356461405, + "grad_norm": 3.452237606048584, + "learning_rate": 1.9703322210938636e-05, + "loss": 0.0734, + "step": 509 + }, + { + "epoch": 0.05529054640069384, + "grad_norm": 1.8070402145385742, + "learning_rate": 1.9702596837371248e-05, + "loss": 0.0405, + "step": 510 + }, + { + "epoch": 0.05539895923677363, + "grad_norm": 2.4397130012512207, + "learning_rate": 1.970187146380386e-05, + "loss": 0.0394, + "step": 511 + }, + { + "epoch": 0.055507372072853424, + "grad_norm": 3.5867652893066406, + "learning_rate": 1.9701146090236472e-05, + "loss": 0.0656, + "step": 512 + }, + { + "epoch": 0.055615784908933215, + "grad_norm": 3.70146107673645, + "learning_rate": 1.9700420716669088e-05, + "loss": 0.0601, + "step": 513 + }, + { + "epoch": 0.05572419774501301, + "grad_norm": 1.9665242433547974, + "learning_rate": 1.96996953431017e-05, + "loss": 0.0325, + "step": 514 + }, + { + "epoch": 0.055832610581092804, + "grad_norm": 2.203073740005493, + "learning_rate": 1.9698969969534312e-05, + "loss": 0.0356, + "step": 515 + }, + { + "epoch": 0.055941023417172595, + "grad_norm": 12.787723541259766, + "learning_rate": 1.9698244595966924e-05, + "loss": 0.0565, + "step": 516 + }, + { + "epoch": 0.056049436253252385, + "grad_norm": 5.371872425079346, + "learning_rate": 1.9697519222399536e-05, + "loss": 0.0429, + "step": 517 + }, + { + "epoch": 0.056157849089332176, + "grad_norm": 3.1282191276550293, + "learning_rate": 1.969679384883215e-05, + "loss": 0.0672, + "step": 518 + }, + { + "epoch": 0.05626626192541197, + "grad_norm": 3.5776376724243164, + "learning_rate": 1.969606847526476e-05, + "loss": 0.0295, + "step": 519 + }, + { + "epoch": 0.05637467476149176, + "grad_norm": 1.6574004888534546, + "learning_rate": 1.9695343101697376e-05, + "loss": 0.0305, + "step": 520 + }, + { + "epoch": 0.05648308759757155, + "grad_norm": 5.77610445022583, + "learning_rate": 1.969461772812999e-05, + "loss": 0.0716, + "step": 521 + }, + { + "epoch": 0.05659150043365135, + "grad_norm": 3.301579713821411, + "learning_rate": 1.96938923545626e-05, + "loss": 0.0438, + "step": 522 + }, + { + "epoch": 0.05669991326973114, + "grad_norm": 3.5058047771453857, + "learning_rate": 1.9693166980995213e-05, + "loss": 0.0729, + "step": 523 + }, + { + "epoch": 0.05680832610581093, + "grad_norm": 3.928373336791992, + "learning_rate": 1.9692441607427828e-05, + "loss": 0.0487, + "step": 524 + }, + { + "epoch": 0.05691673894189072, + "grad_norm": 14.293484687805176, + "learning_rate": 1.969171623386044e-05, + "loss": 0.0399, + "step": 525 + }, + { + "epoch": 0.05702515177797051, + "grad_norm": 5.035843372344971, + "learning_rate": 1.9690990860293053e-05, + "loss": 0.0341, + "step": 526 + }, + { + "epoch": 0.0571335646140503, + "grad_norm": 5.328527927398682, + "learning_rate": 1.9690265486725665e-05, + "loss": 0.0835, + "step": 527 + }, + { + "epoch": 0.05724197745013009, + "grad_norm": 7.580070972442627, + "learning_rate": 1.968954011315828e-05, + "loss": 0.0493, + "step": 528 + }, + { + "epoch": 0.05735039028620989, + "grad_norm": 8.82111644744873, + "learning_rate": 1.9688814739590892e-05, + "loss": 0.0659, + "step": 529 + }, + { + "epoch": 0.05745880312228968, + "grad_norm": 2.095811605453491, + "learning_rate": 1.9688089366023505e-05, + "loss": 0.0252, + "step": 530 + }, + { + "epoch": 0.05756721595836947, + "grad_norm": 13.131628036499023, + "learning_rate": 1.9687363992456117e-05, + "loss": 0.1016, + "step": 531 + }, + { + "epoch": 0.05767562879444926, + "grad_norm": 4.160408020019531, + "learning_rate": 1.968663861888873e-05, + "loss": 0.0245, + "step": 532 + }, + { + "epoch": 0.057784041630529054, + "grad_norm": 8.497014045715332, + "learning_rate": 1.968591324532134e-05, + "loss": 0.0669, + "step": 533 + }, + { + "epoch": 0.057892454466608845, + "grad_norm": 2.419959306716919, + "learning_rate": 1.9685187871753953e-05, + "loss": 0.047, + "step": 534 + }, + { + "epoch": 0.058000867302688636, + "grad_norm": 7.627657413482666, + "learning_rate": 1.968446249818657e-05, + "loss": 0.1549, + "step": 535 + }, + { + "epoch": 0.058109280138768434, + "grad_norm": 8.268096923828125, + "learning_rate": 1.968373712461918e-05, + "loss": 0.136, + "step": 536 + }, + { + "epoch": 0.058217692974848224, + "grad_norm": 3.2126543521881104, + "learning_rate": 1.9683011751051793e-05, + "loss": 0.0578, + "step": 537 + }, + { + "epoch": 0.058326105810928015, + "grad_norm": 4.175128936767578, + "learning_rate": 1.9682286377484405e-05, + "loss": 0.0822, + "step": 538 + }, + { + "epoch": 0.058434518647007806, + "grad_norm": 4.5583648681640625, + "learning_rate": 1.9681561003917017e-05, + "loss": 0.0806, + "step": 539 + }, + { + "epoch": 0.0585429314830876, + "grad_norm": 5.13988733291626, + "learning_rate": 1.968083563034963e-05, + "loss": 0.0588, + "step": 540 + }, + { + "epoch": 0.05865134431916739, + "grad_norm": 6.541275978088379, + "learning_rate": 1.9680110256782245e-05, + "loss": 0.0775, + "step": 541 + }, + { + "epoch": 0.05875975715524718, + "grad_norm": 3.2457268238067627, + "learning_rate": 1.9679384883214857e-05, + "loss": 0.0749, + "step": 542 + }, + { + "epoch": 0.05886816999132697, + "grad_norm": 5.326140880584717, + "learning_rate": 1.9678659509647473e-05, + "loss": 0.0734, + "step": 543 + }, + { + "epoch": 0.05897658282740677, + "grad_norm": 4.660572528839111, + "learning_rate": 1.9677934136080085e-05, + "loss": 0.0905, + "step": 544 + }, + { + "epoch": 0.05908499566348656, + "grad_norm": 2.801358222961426, + "learning_rate": 1.9677208762512697e-05, + "loss": 0.0645, + "step": 545 + }, + { + "epoch": 0.05919340849956635, + "grad_norm": 4.029997825622559, + "learning_rate": 1.967648338894531e-05, + "loss": 0.0454, + "step": 546 + }, + { + "epoch": 0.05930182133564614, + "grad_norm": 1.1576186418533325, + "learning_rate": 1.967575801537792e-05, + "loss": 0.0285, + "step": 547 + }, + { + "epoch": 0.05941023417172593, + "grad_norm": 2.6291046142578125, + "learning_rate": 1.9675032641810533e-05, + "loss": 0.0551, + "step": 548 + }, + { + "epoch": 0.05951864700780572, + "grad_norm": 3.2775402069091797, + "learning_rate": 1.9674307268243145e-05, + "loss": 0.0436, + "step": 549 + }, + { + "epoch": 0.05962705984388551, + "grad_norm": 5.1007771492004395, + "learning_rate": 1.967358189467576e-05, + "loss": 0.0872, + "step": 550 + }, + { + "epoch": 0.05973547267996531, + "grad_norm": 1.9447747468948364, + "learning_rate": 1.9672856521108373e-05, + "loss": 0.0423, + "step": 551 + }, + { + "epoch": 0.0598438855160451, + "grad_norm": 1.9295045137405396, + "learning_rate": 1.9672131147540985e-05, + "loss": 0.0588, + "step": 552 + }, + { + "epoch": 0.05995229835212489, + "grad_norm": 4.722122669219971, + "learning_rate": 1.9671405773973597e-05, + "loss": 0.0619, + "step": 553 + }, + { + "epoch": 0.060060711188204684, + "grad_norm": 4.106877326965332, + "learning_rate": 1.967068040040621e-05, + "loss": 0.0667, + "step": 554 + }, + { + "epoch": 0.060169124024284475, + "grad_norm": 3.848353624343872, + "learning_rate": 1.966995502683882e-05, + "loss": 0.07, + "step": 555 + }, + { + "epoch": 0.060277536860364266, + "grad_norm": 7.349127769470215, + "learning_rate": 1.9669229653271437e-05, + "loss": 0.0861, + "step": 556 + }, + { + "epoch": 0.060385949696444056, + "grad_norm": 7.409067630767822, + "learning_rate": 1.966850427970405e-05, + "loss": 0.0944, + "step": 557 + }, + { + "epoch": 0.06049436253252385, + "grad_norm": 1.8027660846710205, + "learning_rate": 1.966777890613666e-05, + "loss": 0.0473, + "step": 558 + }, + { + "epoch": 0.060602775368603645, + "grad_norm": 3.5219364166259766, + "learning_rate": 1.9667053532569277e-05, + "loss": 0.0538, + "step": 559 + }, + { + "epoch": 0.060711188204683436, + "grad_norm": 3.2544519901275635, + "learning_rate": 1.966632815900189e-05, + "loss": 0.0576, + "step": 560 + }, + { + "epoch": 0.06081960104076323, + "grad_norm": 1.6488791704177856, + "learning_rate": 1.96656027854345e-05, + "loss": 0.0331, + "step": 561 + }, + { + "epoch": 0.06092801387684302, + "grad_norm": 1.6038893461227417, + "learning_rate": 1.9664877411867113e-05, + "loss": 0.0265, + "step": 562 + }, + { + "epoch": 0.06103642671292281, + "grad_norm": 1.450395107269287, + "learning_rate": 1.9664152038299726e-05, + "loss": 0.0297, + "step": 563 + }, + { + "epoch": 0.0611448395490026, + "grad_norm": 4.047382831573486, + "learning_rate": 1.9663426664732338e-05, + "loss": 0.0376, + "step": 564 + }, + { + "epoch": 0.06125325238508239, + "grad_norm": 4.518980979919434, + "learning_rate": 1.9662701291164953e-05, + "loss": 0.058, + "step": 565 + }, + { + "epoch": 0.06136166522116219, + "grad_norm": 1.532928228378296, + "learning_rate": 1.9661975917597565e-05, + "loss": 0.0257, + "step": 566 + }, + { + "epoch": 0.06147007805724198, + "grad_norm": 3.4170846939086914, + "learning_rate": 1.9661250544030178e-05, + "loss": 0.0788, + "step": 567 + }, + { + "epoch": 0.06157849089332177, + "grad_norm": 5.719178676605225, + "learning_rate": 1.966052517046279e-05, + "loss": 0.0424, + "step": 568 + }, + { + "epoch": 0.06168690372940156, + "grad_norm": 11.48940658569336, + "learning_rate": 1.9659799796895402e-05, + "loss": 0.0941, + "step": 569 + }, + { + "epoch": 0.06179531656548135, + "grad_norm": 4.028711318969727, + "learning_rate": 1.9659074423328014e-05, + "loss": 0.0847, + "step": 570 + }, + { + "epoch": 0.06190372940156114, + "grad_norm": 2.0734031200408936, + "learning_rate": 1.9658349049760626e-05, + "loss": 0.0436, + "step": 571 + }, + { + "epoch": 0.062012142237640934, + "grad_norm": 3.995292901992798, + "learning_rate": 1.965762367619324e-05, + "loss": 0.0455, + "step": 572 + }, + { + "epoch": 0.06212055507372073, + "grad_norm": 2.1835110187530518, + "learning_rate": 1.9656898302625854e-05, + "loss": 0.0472, + "step": 573 + }, + { + "epoch": 0.06222896790980052, + "grad_norm": 6.980388641357422, + "learning_rate": 1.9656172929058466e-05, + "loss": 0.0809, + "step": 574 + }, + { + "epoch": 0.062337380745880314, + "grad_norm": 6.883688926696777, + "learning_rate": 1.9655447555491078e-05, + "loss": 0.0919, + "step": 575 + }, + { + "epoch": 0.062445793581960105, + "grad_norm": 4.599647045135498, + "learning_rate": 1.9654722181923694e-05, + "loss": 0.0494, + "step": 576 + }, + { + "epoch": 0.0625542064180399, + "grad_norm": 5.772831916809082, + "learning_rate": 1.9653996808356306e-05, + "loss": 0.0752, + "step": 577 + }, + { + "epoch": 0.06266261925411969, + "grad_norm": 9.321843147277832, + "learning_rate": 1.9653271434788918e-05, + "loss": 0.0886, + "step": 578 + }, + { + "epoch": 0.06277103209019948, + "grad_norm": 7.057833194732666, + "learning_rate": 1.965254606122153e-05, + "loss": 0.0433, + "step": 579 + }, + { + "epoch": 0.06287944492627927, + "grad_norm": 2.2003297805786133, + "learning_rate": 1.9651820687654146e-05, + "loss": 0.0279, + "step": 580 + }, + { + "epoch": 0.06298785776235906, + "grad_norm": 3.8917267322540283, + "learning_rate": 1.9651095314086758e-05, + "loss": 0.0674, + "step": 581 + }, + { + "epoch": 0.06309627059843885, + "grad_norm": 6.756973743438721, + "learning_rate": 1.965036994051937e-05, + "loss": 0.0923, + "step": 582 + }, + { + "epoch": 0.06320468343451864, + "grad_norm": 4.395269393920898, + "learning_rate": 1.9649644566951982e-05, + "loss": 0.0407, + "step": 583 + }, + { + "epoch": 0.06331309627059845, + "grad_norm": 2.5474274158477783, + "learning_rate": 1.9648919193384594e-05, + "loss": 0.0504, + "step": 584 + }, + { + "epoch": 0.06342150910667824, + "grad_norm": 6.1852922439575195, + "learning_rate": 1.9648193819817206e-05, + "loss": 0.0598, + "step": 585 + }, + { + "epoch": 0.06352992194275803, + "grad_norm": 2.851414203643799, + "learning_rate": 1.964746844624982e-05, + "loss": 0.0525, + "step": 586 + }, + { + "epoch": 0.06363833477883782, + "grad_norm": 6.873074054718018, + "learning_rate": 1.9646743072682434e-05, + "loss": 0.1039, + "step": 587 + }, + { + "epoch": 0.06374674761491761, + "grad_norm": 5.0939249992370605, + "learning_rate": 1.9646017699115046e-05, + "loss": 0.0709, + "step": 588 + }, + { + "epoch": 0.0638551604509974, + "grad_norm": 3.9062387943267822, + "learning_rate": 1.9645292325547658e-05, + "loss": 0.0639, + "step": 589 + }, + { + "epoch": 0.06396357328707719, + "grad_norm": 3.3961434364318848, + "learning_rate": 1.964456695198027e-05, + "loss": 0.0796, + "step": 590 + }, + { + "epoch": 0.06407198612315698, + "grad_norm": 3.2446420192718506, + "learning_rate": 1.9643841578412883e-05, + "loss": 0.0707, + "step": 591 + }, + { + "epoch": 0.06418039895923677, + "grad_norm": 4.133655548095703, + "learning_rate": 1.9643116204845495e-05, + "loss": 0.0671, + "step": 592 + }, + { + "epoch": 0.06428881179531656, + "grad_norm": 5.322088241577148, + "learning_rate": 1.964239083127811e-05, + "loss": 0.077, + "step": 593 + }, + { + "epoch": 0.06439722463139635, + "grad_norm": 4.160154819488525, + "learning_rate": 1.9641665457710722e-05, + "loss": 0.0279, + "step": 594 + }, + { + "epoch": 0.06450563746747615, + "grad_norm": 4.265829086303711, + "learning_rate": 1.9640940084143338e-05, + "loss": 0.0446, + "step": 595 + }, + { + "epoch": 0.06461405030355594, + "grad_norm": 4.991794586181641, + "learning_rate": 1.964021471057595e-05, + "loss": 0.0795, + "step": 596 + }, + { + "epoch": 0.06472246313963573, + "grad_norm": 5.683764457702637, + "learning_rate": 1.9639489337008562e-05, + "loss": 0.087, + "step": 597 + }, + { + "epoch": 0.06483087597571552, + "grad_norm": 1.5644044876098633, + "learning_rate": 1.9638763963441174e-05, + "loss": 0.0194, + "step": 598 + }, + { + "epoch": 0.06493928881179532, + "grad_norm": 2.947310209274292, + "learning_rate": 1.9638038589873786e-05, + "loss": 0.0276, + "step": 599 + }, + { + "epoch": 0.06504770164787511, + "grad_norm": 1.1565451622009277, + "learning_rate": 1.96373132163064e-05, + "loss": 0.0178, + "step": 600 + }, + { + "epoch": 0.0651561144839549, + "grad_norm": 5.5906548500061035, + "learning_rate": 1.963658784273901e-05, + "loss": 0.096, + "step": 601 + }, + { + "epoch": 0.0652645273200347, + "grad_norm": 4.409996032714844, + "learning_rate": 1.9635862469171626e-05, + "loss": 0.066, + "step": 602 + }, + { + "epoch": 0.06537294015611449, + "grad_norm": 4.0190911293029785, + "learning_rate": 1.963513709560424e-05, + "loss": 0.0462, + "step": 603 + }, + { + "epoch": 0.06548135299219428, + "grad_norm": 7.38273286819458, + "learning_rate": 1.963441172203685e-05, + "loss": 0.0714, + "step": 604 + }, + { + "epoch": 0.06558976582827407, + "grad_norm": 2.9760684967041016, + "learning_rate": 1.9633686348469463e-05, + "loss": 0.0496, + "step": 605 + }, + { + "epoch": 0.06569817866435386, + "grad_norm": 9.336134910583496, + "learning_rate": 1.9632960974902075e-05, + "loss": 0.1198, + "step": 606 + }, + { + "epoch": 0.06580659150043365, + "grad_norm": 3.734377145767212, + "learning_rate": 1.9632235601334687e-05, + "loss": 0.0349, + "step": 607 + }, + { + "epoch": 0.06591500433651344, + "grad_norm": 3.8217499256134033, + "learning_rate": 1.9631510227767303e-05, + "loss": 0.051, + "step": 608 + }, + { + "epoch": 0.06602341717259323, + "grad_norm": 4.1807403564453125, + "learning_rate": 1.9630784854199915e-05, + "loss": 0.0631, + "step": 609 + }, + { + "epoch": 0.06613183000867302, + "grad_norm": 6.578866481781006, + "learning_rate": 1.9630059480632527e-05, + "loss": 0.0602, + "step": 610 + }, + { + "epoch": 0.06624024284475281, + "grad_norm": 7.531875133514404, + "learning_rate": 1.9629334107065142e-05, + "loss": 0.059, + "step": 611 + }, + { + "epoch": 0.0663486556808326, + "grad_norm": 3.490719795227051, + "learning_rate": 1.9628608733497754e-05, + "loss": 0.0709, + "step": 612 + }, + { + "epoch": 0.0664570685169124, + "grad_norm": 4.482875823974609, + "learning_rate": 1.9627883359930367e-05, + "loss": 0.0411, + "step": 613 + }, + { + "epoch": 0.0665654813529922, + "grad_norm": 11.614676475524902, + "learning_rate": 1.962715798636298e-05, + "loss": 0.0344, + "step": 614 + }, + { + "epoch": 0.06667389418907199, + "grad_norm": 4.560874938964844, + "learning_rate": 1.962643261279559e-05, + "loss": 0.0742, + "step": 615 + }, + { + "epoch": 0.06678230702515178, + "grad_norm": 4.274447441101074, + "learning_rate": 1.9625707239228203e-05, + "loss": 0.0802, + "step": 616 + }, + { + "epoch": 0.06689071986123157, + "grad_norm": 4.349006175994873, + "learning_rate": 1.962498186566082e-05, + "loss": 0.0221, + "step": 617 + }, + { + "epoch": 0.06699913269731136, + "grad_norm": 3.0241899490356445, + "learning_rate": 1.962425649209343e-05, + "loss": 0.0301, + "step": 618 + }, + { + "epoch": 0.06710754553339116, + "grad_norm": 5.013612747192383, + "learning_rate": 1.9623531118526043e-05, + "loss": 0.1, + "step": 619 + }, + { + "epoch": 0.06721595836947095, + "grad_norm": 7.860844612121582, + "learning_rate": 1.9622805744958655e-05, + "loss": 0.0672, + "step": 620 + }, + { + "epoch": 0.06732437120555074, + "grad_norm": 5.331148624420166, + "learning_rate": 1.9622080371391267e-05, + "loss": 0.0274, + "step": 621 + }, + { + "epoch": 0.06743278404163053, + "grad_norm": 3.92598295211792, + "learning_rate": 1.962135499782388e-05, + "loss": 0.0564, + "step": 622 + }, + { + "epoch": 0.06754119687771032, + "grad_norm": 6.045600414276123, + "learning_rate": 1.962062962425649e-05, + "loss": 0.0387, + "step": 623 + }, + { + "epoch": 0.06764960971379011, + "grad_norm": 5.325209140777588, + "learning_rate": 1.9619904250689107e-05, + "loss": 0.0469, + "step": 624 + }, + { + "epoch": 0.0677580225498699, + "grad_norm": 4.774067401885986, + "learning_rate": 1.961917887712172e-05, + "loss": 0.0795, + "step": 625 + }, + { + "epoch": 0.06786643538594969, + "grad_norm": 3.8158862590789795, + "learning_rate": 1.961845350355433e-05, + "loss": 0.0422, + "step": 626 + }, + { + "epoch": 0.06797484822202948, + "grad_norm": 6.8761725425720215, + "learning_rate": 1.9617728129986947e-05, + "loss": 0.0593, + "step": 627 + }, + { + "epoch": 0.06808326105810927, + "grad_norm": 4.7520036697387695, + "learning_rate": 1.961700275641956e-05, + "loss": 0.0684, + "step": 628 + }, + { + "epoch": 0.06819167389418908, + "grad_norm": 3.5900464057922363, + "learning_rate": 1.961627738285217e-05, + "loss": 0.0576, + "step": 629 + }, + { + "epoch": 0.06830008673026887, + "grad_norm": 4.124595642089844, + "learning_rate": 1.9615552009284783e-05, + "loss": 0.0682, + "step": 630 + }, + { + "epoch": 0.06840849956634866, + "grad_norm": 4.10090970993042, + "learning_rate": 1.9614826635717395e-05, + "loss": 0.0408, + "step": 631 + }, + { + "epoch": 0.06851691240242845, + "grad_norm": 6.631425380706787, + "learning_rate": 1.961410126215001e-05, + "loss": 0.0479, + "step": 632 + }, + { + "epoch": 0.06862532523850824, + "grad_norm": 4.835718154907227, + "learning_rate": 1.9613375888582623e-05, + "loss": 0.1433, + "step": 633 + }, + { + "epoch": 0.06873373807458803, + "grad_norm": 2.898359537124634, + "learning_rate": 1.9612650515015235e-05, + "loss": 0.0435, + "step": 634 + }, + { + "epoch": 0.06884215091066782, + "grad_norm": 4.94704532623291, + "learning_rate": 1.9611925141447847e-05, + "loss": 0.0708, + "step": 635 + }, + { + "epoch": 0.06895056374674761, + "grad_norm": 3.466096878051758, + "learning_rate": 1.961119976788046e-05, + "loss": 0.0304, + "step": 636 + }, + { + "epoch": 0.0690589765828274, + "grad_norm": 2.4028866291046143, + "learning_rate": 1.961047439431307e-05, + "loss": 0.0446, + "step": 637 + }, + { + "epoch": 0.0691673894189072, + "grad_norm": 4.439962863922119, + "learning_rate": 1.9609749020745684e-05, + "loss": 0.0831, + "step": 638 + }, + { + "epoch": 0.06927580225498699, + "grad_norm": 4.1488189697265625, + "learning_rate": 1.96090236471783e-05, + "loss": 0.0985, + "step": 639 + }, + { + "epoch": 0.06938421509106678, + "grad_norm": 3.0562660694122314, + "learning_rate": 1.960829827361091e-05, + "loss": 0.0444, + "step": 640 + }, + { + "epoch": 0.06949262792714657, + "grad_norm": 3.1314749717712402, + "learning_rate": 1.9607572900043524e-05, + "loss": 0.0554, + "step": 641 + }, + { + "epoch": 0.06960104076322636, + "grad_norm": 2.3239996433258057, + "learning_rate": 1.9606847526476136e-05, + "loss": 0.0489, + "step": 642 + }, + { + "epoch": 0.06970945359930615, + "grad_norm": 2.0652356147766113, + "learning_rate": 1.9606122152908748e-05, + "loss": 0.0386, + "step": 643 + }, + { + "epoch": 0.06981786643538596, + "grad_norm": 0.9675468802452087, + "learning_rate": 1.960539677934136e-05, + "loss": 0.0173, + "step": 644 + }, + { + "epoch": 0.06992627927146575, + "grad_norm": 4.094545364379883, + "learning_rate": 1.9604671405773976e-05, + "loss": 0.0922, + "step": 645 + }, + { + "epoch": 0.07003469210754554, + "grad_norm": 4.655760765075684, + "learning_rate": 1.9603946032206588e-05, + "loss": 0.0889, + "step": 646 + }, + { + "epoch": 0.07014310494362533, + "grad_norm": 4.994489669799805, + "learning_rate": 1.9603220658639203e-05, + "loss": 0.0307, + "step": 647 + }, + { + "epoch": 0.07025151777970512, + "grad_norm": 7.304422378540039, + "learning_rate": 1.9602495285071815e-05, + "loss": 0.0519, + "step": 648 + }, + { + "epoch": 0.07035993061578491, + "grad_norm": 6.299842357635498, + "learning_rate": 1.9601769911504427e-05, + "loss": 0.0323, + "step": 649 + }, + { + "epoch": 0.0704683434518647, + "grad_norm": 5.3501296043396, + "learning_rate": 1.960104453793704e-05, + "loss": 0.0396, + "step": 650 + }, + { + "epoch": 0.07057675628794449, + "grad_norm": 6.396124362945557, + "learning_rate": 1.9600319164369652e-05, + "loss": 0.1072, + "step": 651 + }, + { + "epoch": 0.07068516912402428, + "grad_norm": 8.612509727478027, + "learning_rate": 1.9599593790802264e-05, + "loss": 0.099, + "step": 652 + }, + { + "epoch": 0.07079358196010407, + "grad_norm": 2.4586727619171143, + "learning_rate": 1.9598868417234876e-05, + "loss": 0.0282, + "step": 653 + }, + { + "epoch": 0.07090199479618386, + "grad_norm": 6.449411392211914, + "learning_rate": 1.959814304366749e-05, + "loss": 0.0413, + "step": 654 + }, + { + "epoch": 0.07101040763226366, + "grad_norm": 3.265376091003418, + "learning_rate": 1.9597417670100104e-05, + "loss": 0.0292, + "step": 655 + }, + { + "epoch": 0.07111882046834345, + "grad_norm": 4.9886345863342285, + "learning_rate": 1.9596692296532716e-05, + "loss": 0.0352, + "step": 656 + }, + { + "epoch": 0.07122723330442324, + "grad_norm": 8.438525199890137, + "learning_rate": 1.9595966922965328e-05, + "loss": 0.0989, + "step": 657 + }, + { + "epoch": 0.07133564614050304, + "grad_norm": 3.970309257507324, + "learning_rate": 1.959524154939794e-05, + "loss": 0.0162, + "step": 658 + }, + { + "epoch": 0.07144405897658283, + "grad_norm": 4.0921149253845215, + "learning_rate": 1.9594516175830552e-05, + "loss": 0.0569, + "step": 659 + }, + { + "epoch": 0.07155247181266262, + "grad_norm": 4.593786716461182, + "learning_rate": 1.9593790802263168e-05, + "loss": 0.0565, + "step": 660 + }, + { + "epoch": 0.07166088464874242, + "grad_norm": 4.3889241218566895, + "learning_rate": 1.959306542869578e-05, + "loss": 0.0615, + "step": 661 + }, + { + "epoch": 0.0717692974848222, + "grad_norm": 3.177832841873169, + "learning_rate": 1.9592340055128392e-05, + "loss": 0.0265, + "step": 662 + }, + { + "epoch": 0.071877710320902, + "grad_norm": 5.3990864753723145, + "learning_rate": 1.9591614681561008e-05, + "loss": 0.0145, + "step": 663 + }, + { + "epoch": 0.07198612315698179, + "grad_norm": 4.929172515869141, + "learning_rate": 1.959088930799362e-05, + "loss": 0.0772, + "step": 664 + }, + { + "epoch": 0.07209453599306158, + "grad_norm": 6.951450347900391, + "learning_rate": 1.9590163934426232e-05, + "loss": 0.0583, + "step": 665 + }, + { + "epoch": 0.07220294882914137, + "grad_norm": 4.97042989730835, + "learning_rate": 1.9589438560858844e-05, + "loss": 0.0703, + "step": 666 + }, + { + "epoch": 0.07231136166522116, + "grad_norm": 3.735499143600464, + "learning_rate": 1.9588713187291456e-05, + "loss": 0.0617, + "step": 667 + }, + { + "epoch": 0.07241977450130095, + "grad_norm": 5.815539360046387, + "learning_rate": 1.958798781372407e-05, + "loss": 0.0963, + "step": 668 + }, + { + "epoch": 0.07252818733738074, + "grad_norm": 1.9129610061645508, + "learning_rate": 1.9587262440156684e-05, + "loss": 0.0429, + "step": 669 + }, + { + "epoch": 0.07263660017346053, + "grad_norm": 2.310994863510132, + "learning_rate": 1.9586537066589296e-05, + "loss": 0.08, + "step": 670 + }, + { + "epoch": 0.07274501300954032, + "grad_norm": 4.386163711547852, + "learning_rate": 1.9585811693021908e-05, + "loss": 0.0545, + "step": 671 + }, + { + "epoch": 0.07285342584562012, + "grad_norm": 4.929885387420654, + "learning_rate": 1.958508631945452e-05, + "loss": 0.0246, + "step": 672 + }, + { + "epoch": 0.07296183868169992, + "grad_norm": 3.256256103515625, + "learning_rate": 1.9584360945887132e-05, + "loss": 0.0832, + "step": 673 + }, + { + "epoch": 0.07307025151777971, + "grad_norm": 4.34868049621582, + "learning_rate": 1.9583635572319745e-05, + "loss": 0.0717, + "step": 674 + }, + { + "epoch": 0.0731786643538595, + "grad_norm": 2.8049278259277344, + "learning_rate": 1.9582910198752357e-05, + "loss": 0.0559, + "step": 675 + }, + { + "epoch": 0.07328707718993929, + "grad_norm": 4.723699569702148, + "learning_rate": 1.9582184825184972e-05, + "loss": 0.0828, + "step": 676 + }, + { + "epoch": 0.07339549002601908, + "grad_norm": 2.4702048301696777, + "learning_rate": 1.9581459451617584e-05, + "loss": 0.043, + "step": 677 + }, + { + "epoch": 0.07350390286209887, + "grad_norm": 2.7687535285949707, + "learning_rate": 1.9580734078050197e-05, + "loss": 0.0723, + "step": 678 + }, + { + "epoch": 0.07361231569817867, + "grad_norm": 3.6039464473724365, + "learning_rate": 1.9580008704482812e-05, + "loss": 0.0568, + "step": 679 + }, + { + "epoch": 0.07372072853425846, + "grad_norm": 2.100511074066162, + "learning_rate": 1.9579283330915424e-05, + "loss": 0.0655, + "step": 680 + }, + { + "epoch": 0.07382914137033825, + "grad_norm": 2.8085885047912598, + "learning_rate": 1.9578557957348036e-05, + "loss": 0.0443, + "step": 681 + }, + { + "epoch": 0.07393755420641804, + "grad_norm": 2.5448920726776123, + "learning_rate": 1.957783258378065e-05, + "loss": 0.0344, + "step": 682 + }, + { + "epoch": 0.07404596704249783, + "grad_norm": 4.2892961502075195, + "learning_rate": 1.957710721021326e-05, + "loss": 0.0586, + "step": 683 + }, + { + "epoch": 0.07415437987857762, + "grad_norm": 3.1418113708496094, + "learning_rate": 1.9576381836645876e-05, + "loss": 0.0836, + "step": 684 + }, + { + "epoch": 0.07426279271465741, + "grad_norm": 3.2205545902252197, + "learning_rate": 1.957565646307849e-05, + "loss": 0.0489, + "step": 685 + }, + { + "epoch": 0.0743712055507372, + "grad_norm": 1.3204724788665771, + "learning_rate": 1.95749310895111e-05, + "loss": 0.029, + "step": 686 + }, + { + "epoch": 0.07447961838681699, + "grad_norm": 4.444270133972168, + "learning_rate": 1.9574205715943713e-05, + "loss": 0.0801, + "step": 687 + }, + { + "epoch": 0.0745880312228968, + "grad_norm": 4.183277130126953, + "learning_rate": 1.9573480342376325e-05, + "loss": 0.0829, + "step": 688 + }, + { + "epoch": 0.07469644405897659, + "grad_norm": 3.303800582885742, + "learning_rate": 1.9572754968808937e-05, + "loss": 0.0645, + "step": 689 + }, + { + "epoch": 0.07480485689505638, + "grad_norm": 2.8308680057525635, + "learning_rate": 1.957202959524155e-05, + "loss": 0.0412, + "step": 690 + }, + { + "epoch": 0.07491326973113617, + "grad_norm": 3.5221476554870605, + "learning_rate": 1.9571304221674165e-05, + "loss": 0.0482, + "step": 691 + }, + { + "epoch": 0.07502168256721596, + "grad_norm": 8.178550720214844, + "learning_rate": 1.9570578848106777e-05, + "loss": 0.0854, + "step": 692 + }, + { + "epoch": 0.07513009540329575, + "grad_norm": 6.3867998123168945, + "learning_rate": 1.956985347453939e-05, + "loss": 0.0443, + "step": 693 + }, + { + "epoch": 0.07523850823937554, + "grad_norm": 2.6286418437957764, + "learning_rate": 1.9569128100972e-05, + "loss": 0.0238, + "step": 694 + }, + { + "epoch": 0.07534692107545533, + "grad_norm": 2.024155855178833, + "learning_rate": 1.9568402727404613e-05, + "loss": 0.0292, + "step": 695 + }, + { + "epoch": 0.07545533391153512, + "grad_norm": 2.844961643218994, + "learning_rate": 1.9567677353837225e-05, + "loss": 0.0329, + "step": 696 + }, + { + "epoch": 0.07556374674761492, + "grad_norm": 8.691267967224121, + "learning_rate": 1.956695198026984e-05, + "loss": 0.0823, + "step": 697 + }, + { + "epoch": 0.0756721595836947, + "grad_norm": 17.36857795715332, + "learning_rate": 1.9566226606702453e-05, + "loss": 0.0646, + "step": 698 + }, + { + "epoch": 0.0757805724197745, + "grad_norm": 3.0775399208068848, + "learning_rate": 1.956550123313507e-05, + "loss": 0.0235, + "step": 699 + }, + { + "epoch": 0.07588898525585429, + "grad_norm": 5.886754989624023, + "learning_rate": 1.956477585956768e-05, + "loss": 0.0442, + "step": 700 + }, + { + "epoch": 0.07599739809193408, + "grad_norm": 6.7042412757873535, + "learning_rate": 1.9564050486000293e-05, + "loss": 0.0923, + "step": 701 + }, + { + "epoch": 0.07610581092801387, + "grad_norm": 3.9600915908813477, + "learning_rate": 1.9563325112432905e-05, + "loss": 0.0508, + "step": 702 + }, + { + "epoch": 0.07621422376409367, + "grad_norm": 6.631053447723389, + "learning_rate": 1.9562599738865517e-05, + "loss": 0.0766, + "step": 703 + }, + { + "epoch": 0.07632263660017347, + "grad_norm": 4.456844329833984, + "learning_rate": 1.956187436529813e-05, + "loss": 0.0861, + "step": 704 + }, + { + "epoch": 0.07643104943625326, + "grad_norm": 4.406484603881836, + "learning_rate": 1.956114899173074e-05, + "loss": 0.0639, + "step": 705 + }, + { + "epoch": 0.07653946227233305, + "grad_norm": 5.980234146118164, + "learning_rate": 1.9560423618163357e-05, + "loss": 0.0594, + "step": 706 + }, + { + "epoch": 0.07664787510841284, + "grad_norm": 4.073902130126953, + "learning_rate": 1.955969824459597e-05, + "loss": 0.0378, + "step": 707 + }, + { + "epoch": 0.07675628794449263, + "grad_norm": 5.269154071807861, + "learning_rate": 1.955897287102858e-05, + "loss": 0.0722, + "step": 708 + }, + { + "epoch": 0.07686470078057242, + "grad_norm": 9.442144393920898, + "learning_rate": 1.9558247497461193e-05, + "loss": 0.0391, + "step": 709 + }, + { + "epoch": 0.07697311361665221, + "grad_norm": 6.282106399536133, + "learning_rate": 1.9557522123893806e-05, + "loss": 0.0906, + "step": 710 + }, + { + "epoch": 0.077081526452732, + "grad_norm": 2.9227778911590576, + "learning_rate": 1.9556796750326418e-05, + "loss": 0.0255, + "step": 711 + }, + { + "epoch": 0.07718993928881179, + "grad_norm": 3.933335781097412, + "learning_rate": 1.9556071376759033e-05, + "loss": 0.0504, + "step": 712 + }, + { + "epoch": 0.07729835212489158, + "grad_norm": 4.967467308044434, + "learning_rate": 1.9555346003191645e-05, + "loss": 0.1046, + "step": 713 + }, + { + "epoch": 0.07740676496097137, + "grad_norm": 1.938075304031372, + "learning_rate": 1.9554620629624257e-05, + "loss": 0.0265, + "step": 714 + }, + { + "epoch": 0.07751517779705117, + "grad_norm": 2.6778995990753174, + "learning_rate": 1.9553895256056873e-05, + "loss": 0.0593, + "step": 715 + }, + { + "epoch": 0.07762359063313096, + "grad_norm": 6.621160507202148, + "learning_rate": 1.9553169882489485e-05, + "loss": 0.0232, + "step": 716 + }, + { + "epoch": 0.07773200346921076, + "grad_norm": 6.589027404785156, + "learning_rate": 1.9552444508922097e-05, + "loss": 0.0375, + "step": 717 + }, + { + "epoch": 0.07784041630529055, + "grad_norm": 4.7608561515808105, + "learning_rate": 1.955171913535471e-05, + "loss": 0.1245, + "step": 718 + }, + { + "epoch": 0.07794882914137034, + "grad_norm": 4.735029220581055, + "learning_rate": 1.955099376178732e-05, + "loss": 0.049, + "step": 719 + }, + { + "epoch": 0.07805724197745013, + "grad_norm": 3.155494213104248, + "learning_rate": 1.9550268388219934e-05, + "loss": 0.0718, + "step": 720 + }, + { + "epoch": 0.07816565481352993, + "grad_norm": 4.056248664855957, + "learning_rate": 1.954954301465255e-05, + "loss": 0.0635, + "step": 721 + }, + { + "epoch": 0.07827406764960972, + "grad_norm": 2.7396605014801025, + "learning_rate": 1.954881764108516e-05, + "loss": 0.0441, + "step": 722 + }, + { + "epoch": 0.0783824804856895, + "grad_norm": 8.752630233764648, + "learning_rate": 1.9548092267517774e-05, + "loss": 0.0851, + "step": 723 + }, + { + "epoch": 0.0784908933217693, + "grad_norm": 5.171670913696289, + "learning_rate": 1.9547366893950386e-05, + "loss": 0.0935, + "step": 724 + }, + { + "epoch": 0.07859930615784909, + "grad_norm": 3.7545554637908936, + "learning_rate": 1.9546641520382998e-05, + "loss": 0.1034, + "step": 725 + }, + { + "epoch": 0.07870771899392888, + "grad_norm": 3.7631537914276123, + "learning_rate": 1.954591614681561e-05, + "loss": 0.0326, + "step": 726 + }, + { + "epoch": 0.07881613183000867, + "grad_norm": 1.9414000511169434, + "learning_rate": 1.9545190773248222e-05, + "loss": 0.0409, + "step": 727 + }, + { + "epoch": 0.07892454466608846, + "grad_norm": 4.3137898445129395, + "learning_rate": 1.9544465399680838e-05, + "loss": 0.0892, + "step": 728 + }, + { + "epoch": 0.07903295750216825, + "grad_norm": 2.1176934242248535, + "learning_rate": 1.954374002611345e-05, + "loss": 0.0476, + "step": 729 + }, + { + "epoch": 0.07914137033824804, + "grad_norm": 3.6086196899414062, + "learning_rate": 1.9543014652546062e-05, + "loss": 0.087, + "step": 730 + }, + { + "epoch": 0.07924978317432783, + "grad_norm": 2.7040531635284424, + "learning_rate": 1.9542289278978677e-05, + "loss": 0.0461, + "step": 731 + }, + { + "epoch": 0.07935819601040764, + "grad_norm": 2.2338058948516846, + "learning_rate": 1.954156390541129e-05, + "loss": 0.0229, + "step": 732 + }, + { + "epoch": 0.07946660884648743, + "grad_norm": 3.052525281906128, + "learning_rate": 1.9540838531843902e-05, + "loss": 0.0516, + "step": 733 + }, + { + "epoch": 0.07957502168256722, + "grad_norm": 5.599859714508057, + "learning_rate": 1.9540113158276514e-05, + "loss": 0.0694, + "step": 734 + }, + { + "epoch": 0.07968343451864701, + "grad_norm": 5.666347026824951, + "learning_rate": 1.9539387784709126e-05, + "loss": 0.0934, + "step": 735 + }, + { + "epoch": 0.0797918473547268, + "grad_norm": 2.592864990234375, + "learning_rate": 1.953866241114174e-05, + "loss": 0.0608, + "step": 736 + }, + { + "epoch": 0.0799002601908066, + "grad_norm": 2.248353958129883, + "learning_rate": 1.9537937037574354e-05, + "loss": 0.0408, + "step": 737 + }, + { + "epoch": 0.08000867302688638, + "grad_norm": 1.9514923095703125, + "learning_rate": 1.9537211664006966e-05, + "loss": 0.0506, + "step": 738 + }, + { + "epoch": 0.08011708586296618, + "grad_norm": 3.634967565536499, + "learning_rate": 1.9536486290439578e-05, + "loss": 0.0417, + "step": 739 + }, + { + "epoch": 0.08022549869904597, + "grad_norm": 2.1217386722564697, + "learning_rate": 1.953576091687219e-05, + "loss": 0.053, + "step": 740 + }, + { + "epoch": 0.08033391153512576, + "grad_norm": 3.934887409210205, + "learning_rate": 1.9535035543304802e-05, + "loss": 0.0378, + "step": 741 + }, + { + "epoch": 0.08044232437120555, + "grad_norm": 2.942495346069336, + "learning_rate": 1.9534310169737414e-05, + "loss": 0.0392, + "step": 742 + }, + { + "epoch": 0.08055073720728534, + "grad_norm": 6.2960100173950195, + "learning_rate": 1.953358479617003e-05, + "loss": 0.0669, + "step": 743 + }, + { + "epoch": 0.08065915004336513, + "grad_norm": 1.0936157703399658, + "learning_rate": 1.9532859422602642e-05, + "loss": 0.0149, + "step": 744 + }, + { + "epoch": 0.08076756287944492, + "grad_norm": 3.2995800971984863, + "learning_rate": 1.9532134049035254e-05, + "loss": 0.0685, + "step": 745 + }, + { + "epoch": 0.08087597571552471, + "grad_norm": 4.69568395614624, + "learning_rate": 1.9531408675467866e-05, + "loss": 0.0482, + "step": 746 + }, + { + "epoch": 0.08098438855160452, + "grad_norm": 5.384230613708496, + "learning_rate": 1.953068330190048e-05, + "loss": 0.0577, + "step": 747 + }, + { + "epoch": 0.08109280138768431, + "grad_norm": 2.9350576400756836, + "learning_rate": 1.9529957928333094e-05, + "loss": 0.0467, + "step": 748 + }, + { + "epoch": 0.0812012142237641, + "grad_norm": 5.120463848114014, + "learning_rate": 1.9529232554765706e-05, + "loss": 0.0454, + "step": 749 + }, + { + "epoch": 0.08130962705984389, + "grad_norm": 8.016936302185059, + "learning_rate": 1.952850718119832e-05, + "loss": 0.0921, + "step": 750 + }, + { + "epoch": 0.08141803989592368, + "grad_norm": 9.568730354309082, + "learning_rate": 1.9527781807630934e-05, + "loss": 0.1025, + "step": 751 + }, + { + "epoch": 0.08152645273200347, + "grad_norm": 6.874969005584717, + "learning_rate": 1.9527056434063546e-05, + "loss": 0.0565, + "step": 752 + }, + { + "epoch": 0.08163486556808326, + "grad_norm": 2.116212844848633, + "learning_rate": 1.9526331060496158e-05, + "loss": 0.0324, + "step": 753 + }, + { + "epoch": 0.08174327840416305, + "grad_norm": 13.28122615814209, + "learning_rate": 1.952560568692877e-05, + "loss": 0.055, + "step": 754 + }, + { + "epoch": 0.08185169124024284, + "grad_norm": 6.001333236694336, + "learning_rate": 1.9524880313361382e-05, + "loss": 0.1068, + "step": 755 + }, + { + "epoch": 0.08196010407632263, + "grad_norm": 3.454289436340332, + "learning_rate": 1.9524154939793995e-05, + "loss": 0.0574, + "step": 756 + }, + { + "epoch": 0.08206851691240243, + "grad_norm": 8.321687698364258, + "learning_rate": 1.9523429566226607e-05, + "loss": 0.0876, + "step": 757 + }, + { + "epoch": 0.08217692974848222, + "grad_norm": 1.405375599861145, + "learning_rate": 1.9522704192659222e-05, + "loss": 0.0087, + "step": 758 + }, + { + "epoch": 0.08228534258456201, + "grad_norm": 4.916119575500488, + "learning_rate": 1.9521978819091834e-05, + "loss": 0.0371, + "step": 759 + }, + { + "epoch": 0.0823937554206418, + "grad_norm": 3.6181840896606445, + "learning_rate": 1.9521253445524447e-05, + "loss": 0.0263, + "step": 760 + }, + { + "epoch": 0.08250216825672159, + "grad_norm": 3.174304246902466, + "learning_rate": 1.952052807195706e-05, + "loss": 0.0319, + "step": 761 + }, + { + "epoch": 0.0826105810928014, + "grad_norm": 2.2908027172088623, + "learning_rate": 1.951980269838967e-05, + "loss": 0.0228, + "step": 762 + }, + { + "epoch": 0.08271899392888118, + "grad_norm": 5.456090927124023, + "learning_rate": 1.9519077324822283e-05, + "loss": 0.0792, + "step": 763 + }, + { + "epoch": 0.08282740676496098, + "grad_norm": 6.451651573181152, + "learning_rate": 1.95183519512549e-05, + "loss": 0.0349, + "step": 764 + }, + { + "epoch": 0.08293581960104077, + "grad_norm": 2.8808646202087402, + "learning_rate": 1.951762657768751e-05, + "loss": 0.0461, + "step": 765 + }, + { + "epoch": 0.08304423243712056, + "grad_norm": 6.510561466217041, + "learning_rate": 1.9516901204120123e-05, + "loss": 0.0864, + "step": 766 + }, + { + "epoch": 0.08315264527320035, + "grad_norm": 4.662564754486084, + "learning_rate": 1.951617583055274e-05, + "loss": 0.0384, + "step": 767 + }, + { + "epoch": 0.08326105810928014, + "grad_norm": 6.593279838562012, + "learning_rate": 1.951545045698535e-05, + "loss": 0.0718, + "step": 768 + }, + { + "epoch": 0.08336947094535993, + "grad_norm": 6.118373394012451, + "learning_rate": 1.9514725083417963e-05, + "loss": 0.0497, + "step": 769 + }, + { + "epoch": 0.08347788378143972, + "grad_norm": 4.9401655197143555, + "learning_rate": 1.9513999709850575e-05, + "loss": 0.0336, + "step": 770 + }, + { + "epoch": 0.08358629661751951, + "grad_norm": 10.064708709716797, + "learning_rate": 1.9513274336283187e-05, + "loss": 0.1036, + "step": 771 + }, + { + "epoch": 0.0836947094535993, + "grad_norm": 7.177042007446289, + "learning_rate": 1.95125489627158e-05, + "loss": 0.0924, + "step": 772 + }, + { + "epoch": 0.0838031222896791, + "grad_norm": 2.413504123687744, + "learning_rate": 1.9511823589148415e-05, + "loss": 0.0341, + "step": 773 + }, + { + "epoch": 0.08391153512575888, + "grad_norm": 2.3939058780670166, + "learning_rate": 1.9511098215581027e-05, + "loss": 0.0342, + "step": 774 + }, + { + "epoch": 0.08401994796183868, + "grad_norm": 4.678165435791016, + "learning_rate": 1.951037284201364e-05, + "loss": 0.0321, + "step": 775 + }, + { + "epoch": 0.08412836079791847, + "grad_norm": 4.58705472946167, + "learning_rate": 1.950964746844625e-05, + "loss": 0.0516, + "step": 776 + }, + { + "epoch": 0.08423677363399827, + "grad_norm": 2.5956778526306152, + "learning_rate": 1.9508922094878863e-05, + "loss": 0.0453, + "step": 777 + }, + { + "epoch": 0.08434518647007806, + "grad_norm": 9.918566703796387, + "learning_rate": 1.9508196721311475e-05, + "loss": 0.1001, + "step": 778 + }, + { + "epoch": 0.08445359930615785, + "grad_norm": 3.666691780090332, + "learning_rate": 1.9507471347744087e-05, + "loss": 0.0664, + "step": 779 + }, + { + "epoch": 0.08456201214223764, + "grad_norm": 1.7953053712844849, + "learning_rate": 1.9506745974176703e-05, + "loss": 0.0448, + "step": 780 + }, + { + "epoch": 0.08467042497831744, + "grad_norm": 2.8222339153289795, + "learning_rate": 1.9506020600609315e-05, + "loss": 0.0277, + "step": 781 + }, + { + "epoch": 0.08477883781439723, + "grad_norm": 2.783782482147217, + "learning_rate": 1.9505295227041927e-05, + "loss": 0.0497, + "step": 782 + }, + { + "epoch": 0.08488725065047702, + "grad_norm": 5.656794548034668, + "learning_rate": 1.9504569853474543e-05, + "loss": 0.034, + "step": 783 + }, + { + "epoch": 0.08499566348655681, + "grad_norm": 3.9180924892425537, + "learning_rate": 1.9503844479907155e-05, + "loss": 0.07, + "step": 784 + }, + { + "epoch": 0.0851040763226366, + "grad_norm": 8.324980735778809, + "learning_rate": 1.9503119106339767e-05, + "loss": 0.0842, + "step": 785 + }, + { + "epoch": 0.08521248915871639, + "grad_norm": 1.5601868629455566, + "learning_rate": 1.950239373277238e-05, + "loss": 0.033, + "step": 786 + }, + { + "epoch": 0.08532090199479618, + "grad_norm": 6.1452131271362305, + "learning_rate": 1.950166835920499e-05, + "loss": 0.0313, + "step": 787 + }, + { + "epoch": 0.08542931483087597, + "grad_norm": 1.8361932039260864, + "learning_rate": 1.9500942985637607e-05, + "loss": 0.0264, + "step": 788 + }, + { + "epoch": 0.08553772766695576, + "grad_norm": 5.725986480712891, + "learning_rate": 1.950021761207022e-05, + "loss": 0.0387, + "step": 789 + }, + { + "epoch": 0.08564614050303555, + "grad_norm": 4.696506023406982, + "learning_rate": 1.949949223850283e-05, + "loss": 0.0518, + "step": 790 + }, + { + "epoch": 0.08575455333911536, + "grad_norm": 6.10752534866333, + "learning_rate": 1.9498766864935443e-05, + "loss": 0.0896, + "step": 791 + }, + { + "epoch": 0.08586296617519515, + "grad_norm": 2.61262583732605, + "learning_rate": 1.9498041491368055e-05, + "loss": 0.054, + "step": 792 + }, + { + "epoch": 0.08597137901127494, + "grad_norm": 3.795004367828369, + "learning_rate": 1.9497316117800668e-05, + "loss": 0.0334, + "step": 793 + }, + { + "epoch": 0.08607979184735473, + "grad_norm": 2.7687923908233643, + "learning_rate": 1.949659074423328e-05, + "loss": 0.0328, + "step": 794 + }, + { + "epoch": 0.08618820468343452, + "grad_norm": 2.8980627059936523, + "learning_rate": 1.9495865370665895e-05, + "loss": 0.0406, + "step": 795 + }, + { + "epoch": 0.08629661751951431, + "grad_norm": 4.9509077072143555, + "learning_rate": 1.9495139997098507e-05, + "loss": 0.0707, + "step": 796 + }, + { + "epoch": 0.0864050303555941, + "grad_norm": 1.9342420101165771, + "learning_rate": 1.949441462353112e-05, + "loss": 0.018, + "step": 797 + }, + { + "epoch": 0.0865134431916739, + "grad_norm": 5.715460300445557, + "learning_rate": 1.9493689249963732e-05, + "loss": 0.0706, + "step": 798 + }, + { + "epoch": 0.08662185602775369, + "grad_norm": 3.3059585094451904, + "learning_rate": 1.9492963876396344e-05, + "loss": 0.0522, + "step": 799 + }, + { + "epoch": 0.08673026886383348, + "grad_norm": 5.206146717071533, + "learning_rate": 1.949223850282896e-05, + "loss": 0.0665, + "step": 800 + }, + { + "epoch": 0.08683868169991327, + "grad_norm": 8.02914810180664, + "learning_rate": 1.949151312926157e-05, + "loss": 0.0737, + "step": 801 + }, + { + "epoch": 0.08694709453599306, + "grad_norm": 2.2402868270874023, + "learning_rate": 1.9490787755694184e-05, + "loss": 0.0448, + "step": 802 + }, + { + "epoch": 0.08705550737207285, + "grad_norm": 4.337299823760986, + "learning_rate": 1.94900623821268e-05, + "loss": 0.0369, + "step": 803 + }, + { + "epoch": 0.08716392020815264, + "grad_norm": 2.704327344894409, + "learning_rate": 1.948933700855941e-05, + "loss": 0.026, + "step": 804 + }, + { + "epoch": 0.08727233304423243, + "grad_norm": 1.1834079027175903, + "learning_rate": 1.9488611634992024e-05, + "loss": 0.0142, + "step": 805 + }, + { + "epoch": 0.08738074588031224, + "grad_norm": 5.068702220916748, + "learning_rate": 1.9487886261424636e-05, + "loss": 0.0936, + "step": 806 + }, + { + "epoch": 0.08748915871639203, + "grad_norm": 4.188436985015869, + "learning_rate": 1.9487160887857248e-05, + "loss": 0.0669, + "step": 807 + }, + { + "epoch": 0.08759757155247182, + "grad_norm": 6.857095241546631, + "learning_rate": 1.948643551428986e-05, + "loss": 0.1121, + "step": 808 + }, + { + "epoch": 0.08770598438855161, + "grad_norm": 6.99936580657959, + "learning_rate": 1.9485710140722472e-05, + "loss": 0.0787, + "step": 809 + }, + { + "epoch": 0.0878143972246314, + "grad_norm": 7.110266208648682, + "learning_rate": 1.9484984767155088e-05, + "loss": 0.057, + "step": 810 + }, + { + "epoch": 0.08792281006071119, + "grad_norm": 1.164623737335205, + "learning_rate": 1.94842593935877e-05, + "loss": 0.0127, + "step": 811 + }, + { + "epoch": 0.08803122289679098, + "grad_norm": 3.7436797618865967, + "learning_rate": 1.9483534020020312e-05, + "loss": 0.0578, + "step": 812 + }, + { + "epoch": 0.08813963573287077, + "grad_norm": 5.587394714355469, + "learning_rate": 1.9482808646452924e-05, + "loss": 0.059, + "step": 813 + }, + { + "epoch": 0.08824804856895056, + "grad_norm": 7.35496187210083, + "learning_rate": 1.9482083272885536e-05, + "loss": 0.0499, + "step": 814 + }, + { + "epoch": 0.08835646140503035, + "grad_norm": 4.177717208862305, + "learning_rate": 1.948135789931815e-05, + "loss": 0.0632, + "step": 815 + }, + { + "epoch": 0.08846487424111014, + "grad_norm": 3.167919635772705, + "learning_rate": 1.9480632525750764e-05, + "loss": 0.084, + "step": 816 + }, + { + "epoch": 0.08857328707718994, + "grad_norm": 6.313333988189697, + "learning_rate": 1.9479907152183376e-05, + "loss": 0.057, + "step": 817 + }, + { + "epoch": 0.08868169991326973, + "grad_norm": 3.391019344329834, + "learning_rate": 1.9479181778615988e-05, + "loss": 0.0416, + "step": 818 + }, + { + "epoch": 0.08879011274934952, + "grad_norm": 4.037591457366943, + "learning_rate": 1.9478456405048604e-05, + "loss": 0.0371, + "step": 819 + }, + { + "epoch": 0.08889852558542931, + "grad_norm": 2.3596301078796387, + "learning_rate": 1.9477731031481216e-05, + "loss": 0.0283, + "step": 820 + }, + { + "epoch": 0.08900693842150911, + "grad_norm": 2.5450851917266846, + "learning_rate": 1.9477005657913828e-05, + "loss": 0.0419, + "step": 821 + }, + { + "epoch": 0.0891153512575889, + "grad_norm": 4.595979690551758, + "learning_rate": 1.947628028434644e-05, + "loss": 0.0732, + "step": 822 + }, + { + "epoch": 0.0892237640936687, + "grad_norm": 2.130807399749756, + "learning_rate": 1.9475554910779052e-05, + "loss": 0.0267, + "step": 823 + }, + { + "epoch": 0.08933217692974849, + "grad_norm": 3.706829786300659, + "learning_rate": 1.9474829537211664e-05, + "loss": 0.0611, + "step": 824 + }, + { + "epoch": 0.08944058976582828, + "grad_norm": 1.443727731704712, + "learning_rate": 1.947410416364428e-05, + "loss": 0.0331, + "step": 825 + }, + { + "epoch": 0.08954900260190807, + "grad_norm": 4.366591453552246, + "learning_rate": 1.9473378790076892e-05, + "loss": 0.0686, + "step": 826 + }, + { + "epoch": 0.08965741543798786, + "grad_norm": 0.700121283531189, + "learning_rate": 1.9472653416509504e-05, + "loss": 0.0075, + "step": 827 + }, + { + "epoch": 0.08976582827406765, + "grad_norm": 2.330325126647949, + "learning_rate": 1.9471928042942116e-05, + "loss": 0.0273, + "step": 828 + }, + { + "epoch": 0.08987424111014744, + "grad_norm": 4.880648136138916, + "learning_rate": 1.947120266937473e-05, + "loss": 0.1022, + "step": 829 + }, + { + "epoch": 0.08998265394622723, + "grad_norm": 4.901490688323975, + "learning_rate": 1.947047729580734e-05, + "loss": 0.0152, + "step": 830 + }, + { + "epoch": 0.09009106678230702, + "grad_norm": 4.2005815505981445, + "learning_rate": 1.9469751922239953e-05, + "loss": 0.0977, + "step": 831 + }, + { + "epoch": 0.09019947961838681, + "grad_norm": 4.863993167877197, + "learning_rate": 1.946902654867257e-05, + "loss": 0.0376, + "step": 832 + }, + { + "epoch": 0.0903078924544666, + "grad_norm": 6.803310871124268, + "learning_rate": 1.946830117510518e-05, + "loss": 0.0839, + "step": 833 + }, + { + "epoch": 0.0904163052905464, + "grad_norm": 3.235945224761963, + "learning_rate": 1.9467575801537793e-05, + "loss": 0.0474, + "step": 834 + }, + { + "epoch": 0.09052471812662619, + "grad_norm": 6.838757038116455, + "learning_rate": 1.9466850427970408e-05, + "loss": 0.0586, + "step": 835 + }, + { + "epoch": 0.09063313096270599, + "grad_norm": 4.037535667419434, + "learning_rate": 1.946612505440302e-05, + "loss": 0.0345, + "step": 836 + }, + { + "epoch": 0.09074154379878578, + "grad_norm": 9.277021408081055, + "learning_rate": 1.9465399680835632e-05, + "loss": 0.0592, + "step": 837 + }, + { + "epoch": 0.09084995663486557, + "grad_norm": 3.920079231262207, + "learning_rate": 1.9464674307268245e-05, + "loss": 0.0808, + "step": 838 + }, + { + "epoch": 0.09095836947094536, + "grad_norm": 1.3412655591964722, + "learning_rate": 1.9463948933700857e-05, + "loss": 0.0278, + "step": 839 + }, + { + "epoch": 0.09106678230702515, + "grad_norm": 4.649803638458252, + "learning_rate": 1.9463223560133472e-05, + "loss": 0.0772, + "step": 840 + }, + { + "epoch": 0.09117519514310495, + "grad_norm": 2.7432143688201904, + "learning_rate": 1.9462498186566084e-05, + "loss": 0.0578, + "step": 841 + }, + { + "epoch": 0.09128360797918474, + "grad_norm": 1.7694340944290161, + "learning_rate": 1.9461772812998697e-05, + "loss": 0.0439, + "step": 842 + }, + { + "epoch": 0.09139202081526453, + "grad_norm": 3.9077839851379395, + "learning_rate": 1.946104743943131e-05, + "loss": 0.06, + "step": 843 + }, + { + "epoch": 0.09150043365134432, + "grad_norm": 2.9284846782684326, + "learning_rate": 1.946032206586392e-05, + "loss": 0.0431, + "step": 844 + }, + { + "epoch": 0.09160884648742411, + "grad_norm": 3.464989423751831, + "learning_rate": 1.9459596692296533e-05, + "loss": 0.0507, + "step": 845 + }, + { + "epoch": 0.0917172593235039, + "grad_norm": 2.230135440826416, + "learning_rate": 1.9458871318729145e-05, + "loss": 0.0476, + "step": 846 + }, + { + "epoch": 0.09182567215958369, + "grad_norm": 3.9420626163482666, + "learning_rate": 1.945814594516176e-05, + "loss": 0.0588, + "step": 847 + }, + { + "epoch": 0.09193408499566348, + "grad_norm": 10.60716438293457, + "learning_rate": 1.9457420571594373e-05, + "loss": 0.0559, + "step": 848 + }, + { + "epoch": 0.09204249783174327, + "grad_norm": 2.4616222381591797, + "learning_rate": 1.9456695198026985e-05, + "loss": 0.0713, + "step": 849 + }, + { + "epoch": 0.09215091066782308, + "grad_norm": 3.7037386894226074, + "learning_rate": 1.9455969824459597e-05, + "loss": 0.0721, + "step": 850 + }, + { + "epoch": 0.09225932350390287, + "grad_norm": 2.914976119995117, + "learning_rate": 1.945524445089221e-05, + "loss": 0.0647, + "step": 851 + }, + { + "epoch": 0.09236773633998266, + "grad_norm": 3.9968464374542236, + "learning_rate": 1.9454519077324825e-05, + "loss": 0.0424, + "step": 852 + }, + { + "epoch": 0.09247614917606245, + "grad_norm": 4.507113456726074, + "learning_rate": 1.9453793703757437e-05, + "loss": 0.0985, + "step": 853 + }, + { + "epoch": 0.09258456201214224, + "grad_norm": 3.3729567527770996, + "learning_rate": 1.945306833019005e-05, + "loss": 0.0724, + "step": 854 + }, + { + "epoch": 0.09269297484822203, + "grad_norm": 3.737800121307373, + "learning_rate": 1.9452342956622665e-05, + "loss": 0.059, + "step": 855 + }, + { + "epoch": 0.09280138768430182, + "grad_norm": 3.814436674118042, + "learning_rate": 1.9451617583055277e-05, + "loss": 0.0745, + "step": 856 + }, + { + "epoch": 0.09290980052038161, + "grad_norm": 1.4286221265792847, + "learning_rate": 1.945089220948789e-05, + "loss": 0.0283, + "step": 857 + }, + { + "epoch": 0.0930182133564614, + "grad_norm": 4.6080546379089355, + "learning_rate": 1.94501668359205e-05, + "loss": 0.0784, + "step": 858 + }, + { + "epoch": 0.0931266261925412, + "grad_norm": 3.515054225921631, + "learning_rate": 1.9449441462353113e-05, + "loss": 0.0746, + "step": 859 + }, + { + "epoch": 0.09323503902862099, + "grad_norm": 3.4896130561828613, + "learning_rate": 1.9448716088785725e-05, + "loss": 0.063, + "step": 860 + }, + { + "epoch": 0.09334345186470078, + "grad_norm": 7.852476596832275, + "learning_rate": 1.9447990715218337e-05, + "loss": 0.055, + "step": 861 + }, + { + "epoch": 0.09345186470078057, + "grad_norm": 2.2720658779144287, + "learning_rate": 1.9447265341650953e-05, + "loss": 0.0403, + "step": 862 + }, + { + "epoch": 0.09356027753686036, + "grad_norm": 4.165477752685547, + "learning_rate": 1.9446539968083565e-05, + "loss": 0.0861, + "step": 863 + }, + { + "epoch": 0.09366869037294015, + "grad_norm": 2.324749708175659, + "learning_rate": 1.9445814594516177e-05, + "loss": 0.0311, + "step": 864 + }, + { + "epoch": 0.09377710320901995, + "grad_norm": 0.9276569485664368, + "learning_rate": 1.944508922094879e-05, + "loss": 0.015, + "step": 865 + }, + { + "epoch": 0.09388551604509975, + "grad_norm": 5.809394359588623, + "learning_rate": 1.94443638473814e-05, + "loss": 0.0716, + "step": 866 + }, + { + "epoch": 0.09399392888117954, + "grad_norm": 3.603482246398926, + "learning_rate": 1.9443638473814014e-05, + "loss": 0.0744, + "step": 867 + }, + { + "epoch": 0.09410234171725933, + "grad_norm": 1.3665783405303955, + "learning_rate": 1.944291310024663e-05, + "loss": 0.013, + "step": 868 + }, + { + "epoch": 0.09421075455333912, + "grad_norm": 4.7585601806640625, + "learning_rate": 1.944218772667924e-05, + "loss": 0.0789, + "step": 869 + }, + { + "epoch": 0.09431916738941891, + "grad_norm": 2.2458834648132324, + "learning_rate": 1.9441462353111853e-05, + "loss": 0.0237, + "step": 870 + }, + { + "epoch": 0.0944275802254987, + "grad_norm": 4.481781959533691, + "learning_rate": 1.944073697954447e-05, + "loss": 0.045, + "step": 871 + }, + { + "epoch": 0.09453599306157849, + "grad_norm": 2.6370999813079834, + "learning_rate": 1.944001160597708e-05, + "loss": 0.0293, + "step": 872 + }, + { + "epoch": 0.09464440589765828, + "grad_norm": 7.849095821380615, + "learning_rate": 1.9439286232409693e-05, + "loss": 0.0823, + "step": 873 + }, + { + "epoch": 0.09475281873373807, + "grad_norm": 5.444472312927246, + "learning_rate": 1.9438560858842305e-05, + "loss": 0.0395, + "step": 874 + }, + { + "epoch": 0.09486123156981786, + "grad_norm": 10.290040016174316, + "learning_rate": 1.9437835485274918e-05, + "loss": 0.0992, + "step": 875 + }, + { + "epoch": 0.09496964440589765, + "grad_norm": 1.7177259922027588, + "learning_rate": 1.943711011170753e-05, + "loss": 0.02, + "step": 876 + }, + { + "epoch": 0.09507805724197745, + "grad_norm": 7.122033596038818, + "learning_rate": 1.9436384738140145e-05, + "loss": 0.041, + "step": 877 + }, + { + "epoch": 0.09518647007805724, + "grad_norm": 23.860782623291016, + "learning_rate": 1.9435659364572757e-05, + "loss": 0.0923, + "step": 878 + }, + { + "epoch": 0.09529488291413703, + "grad_norm": 7.782773971557617, + "learning_rate": 1.943493399100537e-05, + "loss": 0.094, + "step": 879 + }, + { + "epoch": 0.09540329575021683, + "grad_norm": 10.300323486328125, + "learning_rate": 1.9434208617437982e-05, + "loss": 0.1085, + "step": 880 + }, + { + "epoch": 0.09551170858629662, + "grad_norm": 2.5205280780792236, + "learning_rate": 1.9433483243870594e-05, + "loss": 0.0447, + "step": 881 + }, + { + "epoch": 0.09562012142237641, + "grad_norm": 5.659046173095703, + "learning_rate": 1.9432757870303206e-05, + "loss": 0.1012, + "step": 882 + }, + { + "epoch": 0.0957285342584562, + "grad_norm": 5.026376247406006, + "learning_rate": 1.9432032496735818e-05, + "loss": 0.0414, + "step": 883 + }, + { + "epoch": 0.095836947094536, + "grad_norm": 4.072183609008789, + "learning_rate": 1.9431307123168434e-05, + "loss": 0.0472, + "step": 884 + }, + { + "epoch": 0.09594535993061579, + "grad_norm": 4.766936779022217, + "learning_rate": 1.9430581749601046e-05, + "loss": 0.0501, + "step": 885 + }, + { + "epoch": 0.09605377276669558, + "grad_norm": 2.551861524581909, + "learning_rate": 1.9429856376033658e-05, + "loss": 0.0218, + "step": 886 + }, + { + "epoch": 0.09616218560277537, + "grad_norm": 3.479288339614868, + "learning_rate": 1.9429131002466273e-05, + "loss": 0.1043, + "step": 887 + }, + { + "epoch": 0.09627059843885516, + "grad_norm": 3.93548583984375, + "learning_rate": 1.9428405628898886e-05, + "loss": 0.0622, + "step": 888 + }, + { + "epoch": 0.09637901127493495, + "grad_norm": 4.046957969665527, + "learning_rate": 1.9427680255331498e-05, + "loss": 0.0711, + "step": 889 + }, + { + "epoch": 0.09648742411101474, + "grad_norm": 3.1159775257110596, + "learning_rate": 1.942695488176411e-05, + "loss": 0.0277, + "step": 890 + }, + { + "epoch": 0.09659583694709453, + "grad_norm": 3.037130355834961, + "learning_rate": 1.9426229508196722e-05, + "loss": 0.0727, + "step": 891 + }, + { + "epoch": 0.09670424978317432, + "grad_norm": 4.866814136505127, + "learning_rate": 1.9425504134629338e-05, + "loss": 0.0316, + "step": 892 + }, + { + "epoch": 0.09681266261925411, + "grad_norm": 6.422913551330566, + "learning_rate": 1.942477876106195e-05, + "loss": 0.0422, + "step": 893 + }, + { + "epoch": 0.0969210754553339, + "grad_norm": 2.4080233573913574, + "learning_rate": 1.9424053387494562e-05, + "loss": 0.022, + "step": 894 + }, + { + "epoch": 0.09702948829141371, + "grad_norm": 2.543135166168213, + "learning_rate": 1.9423328013927174e-05, + "loss": 0.0381, + "step": 895 + }, + { + "epoch": 0.0971379011274935, + "grad_norm": 2.5774388313293457, + "learning_rate": 1.9422602640359786e-05, + "loss": 0.0509, + "step": 896 + }, + { + "epoch": 0.09724631396357329, + "grad_norm": 3.6625630855560303, + "learning_rate": 1.9421877266792398e-05, + "loss": 0.0738, + "step": 897 + }, + { + "epoch": 0.09735472679965308, + "grad_norm": 3.029111862182617, + "learning_rate": 1.942115189322501e-05, + "loss": 0.0332, + "step": 898 + }, + { + "epoch": 0.09746313963573287, + "grad_norm": 2.0200536251068115, + "learning_rate": 1.9420426519657626e-05, + "loss": 0.0288, + "step": 899 + }, + { + "epoch": 0.09757155247181266, + "grad_norm": 4.205935478210449, + "learning_rate": 1.9419701146090238e-05, + "loss": 0.1081, + "step": 900 + }, + { + "epoch": 0.09767996530789246, + "grad_norm": 3.498061418533325, + "learning_rate": 1.941897577252285e-05, + "loss": 0.0568, + "step": 901 + }, + { + "epoch": 0.09778837814397225, + "grad_norm": 2.9726603031158447, + "learning_rate": 1.9418250398955462e-05, + "loss": 0.0369, + "step": 902 + }, + { + "epoch": 0.09789679098005204, + "grad_norm": 8.703743934631348, + "learning_rate": 1.9417525025388075e-05, + "loss": 0.103, + "step": 903 + }, + { + "epoch": 0.09800520381613183, + "grad_norm": 5.044507026672363, + "learning_rate": 1.941679965182069e-05, + "loss": 0.0962, + "step": 904 + }, + { + "epoch": 0.09811361665221162, + "grad_norm": 5.264965534210205, + "learning_rate": 1.9416074278253302e-05, + "loss": 0.1005, + "step": 905 + }, + { + "epoch": 0.09822202948829141, + "grad_norm": 3.860732316970825, + "learning_rate": 1.9415348904685914e-05, + "loss": 0.0927, + "step": 906 + }, + { + "epoch": 0.0983304423243712, + "grad_norm": 3.550140857696533, + "learning_rate": 1.941462353111853e-05, + "loss": 0.0741, + "step": 907 + }, + { + "epoch": 0.09843885516045099, + "grad_norm": 2.726102352142334, + "learning_rate": 1.9413898157551142e-05, + "loss": 0.0362, + "step": 908 + }, + { + "epoch": 0.09854726799653078, + "grad_norm": 2.345214366912842, + "learning_rate": 1.9413172783983754e-05, + "loss": 0.0309, + "step": 909 + }, + { + "epoch": 0.09865568083261059, + "grad_norm": 2.9278666973114014, + "learning_rate": 1.9412447410416366e-05, + "loss": 0.0769, + "step": 910 + }, + { + "epoch": 0.09876409366869038, + "grad_norm": 4.313033580780029, + "learning_rate": 1.941172203684898e-05, + "loss": 0.0484, + "step": 911 + }, + { + "epoch": 0.09887250650477017, + "grad_norm": 3.8208224773406982, + "learning_rate": 1.941099666328159e-05, + "loss": 0.0951, + "step": 912 + }, + { + "epoch": 0.09898091934084996, + "grad_norm": 4.208080768585205, + "learning_rate": 1.9410271289714203e-05, + "loss": 0.0606, + "step": 913 + }, + { + "epoch": 0.09908933217692975, + "grad_norm": 2.8162248134613037, + "learning_rate": 1.9409545916146818e-05, + "loss": 0.053, + "step": 914 + }, + { + "epoch": 0.09919774501300954, + "grad_norm": 1.9112223386764526, + "learning_rate": 1.940882054257943e-05, + "loss": 0.0353, + "step": 915 + }, + { + "epoch": 0.09930615784908933, + "grad_norm": 2.1684110164642334, + "learning_rate": 1.9408095169012043e-05, + "loss": 0.0443, + "step": 916 + }, + { + "epoch": 0.09941457068516912, + "grad_norm": 6.147825717926025, + "learning_rate": 1.9407369795444655e-05, + "loss": 0.0724, + "step": 917 + }, + { + "epoch": 0.09952298352124891, + "grad_norm": 2.0954818725585938, + "learning_rate": 1.9406644421877267e-05, + "loss": 0.0321, + "step": 918 + }, + { + "epoch": 0.0996313963573287, + "grad_norm": 7.385792255401611, + "learning_rate": 1.940591904830988e-05, + "loss": 0.0579, + "step": 919 + }, + { + "epoch": 0.0997398091934085, + "grad_norm": 6.7389020919799805, + "learning_rate": 1.9405193674742495e-05, + "loss": 0.0748, + "step": 920 + }, + { + "epoch": 0.09984822202948829, + "grad_norm": 4.492173671722412, + "learning_rate": 1.9404468301175107e-05, + "loss": 0.0467, + "step": 921 + }, + { + "epoch": 0.09995663486556808, + "grad_norm": 4.384671688079834, + "learning_rate": 1.940374292760772e-05, + "loss": 0.0733, + "step": 922 + }, + { + "epoch": 0.10006504770164787, + "grad_norm": 3.6825428009033203, + "learning_rate": 1.9403017554040334e-05, + "loss": 0.0518, + "step": 923 + }, + { + "epoch": 0.10017346053772767, + "grad_norm": 6.540402889251709, + "learning_rate": 1.9402292180472947e-05, + "loss": 0.0548, + "step": 924 + }, + { + "epoch": 0.10028187337380746, + "grad_norm": 2.8940436840057373, + "learning_rate": 1.940156680690556e-05, + "loss": 0.0476, + "step": 925 + }, + { + "epoch": 0.10039028620988726, + "grad_norm": 4.17808723449707, + "learning_rate": 1.940084143333817e-05, + "loss": 0.0693, + "step": 926 + }, + { + "epoch": 0.10049869904596705, + "grad_norm": 3.6034393310546875, + "learning_rate": 1.9400116059770783e-05, + "loss": 0.0445, + "step": 927 + }, + { + "epoch": 0.10060711188204684, + "grad_norm": 5.5795674324035645, + "learning_rate": 1.9399390686203395e-05, + "loss": 0.0439, + "step": 928 + }, + { + "epoch": 0.10071552471812663, + "grad_norm": 7.5579328536987305, + "learning_rate": 1.939866531263601e-05, + "loss": 0.1098, + "step": 929 + }, + { + "epoch": 0.10082393755420642, + "grad_norm": 4.861972332000732, + "learning_rate": 1.9397939939068623e-05, + "loss": 0.0838, + "step": 930 + }, + { + "epoch": 0.10093235039028621, + "grad_norm": 3.757972240447998, + "learning_rate": 1.9397214565501235e-05, + "loss": 0.1265, + "step": 931 + }, + { + "epoch": 0.101040763226366, + "grad_norm": 1.843143105506897, + "learning_rate": 1.9396489191933847e-05, + "loss": 0.0231, + "step": 932 + }, + { + "epoch": 0.10114917606244579, + "grad_norm": 2.621896743774414, + "learning_rate": 1.939576381836646e-05, + "loss": 0.0315, + "step": 933 + }, + { + "epoch": 0.10125758889852558, + "grad_norm": 5.764852046966553, + "learning_rate": 1.939503844479907e-05, + "loss": 0.0846, + "step": 934 + }, + { + "epoch": 0.10136600173460537, + "grad_norm": 3.2244865894317627, + "learning_rate": 1.9394313071231687e-05, + "loss": 0.0429, + "step": 935 + }, + { + "epoch": 0.10147441457068516, + "grad_norm": 3.521711826324463, + "learning_rate": 1.93935876976643e-05, + "loss": 0.0382, + "step": 936 + }, + { + "epoch": 0.10158282740676496, + "grad_norm": 2.9273717403411865, + "learning_rate": 1.939286232409691e-05, + "loss": 0.0723, + "step": 937 + }, + { + "epoch": 0.10169124024284475, + "grad_norm": 3.9563660621643066, + "learning_rate": 1.9392136950529527e-05, + "loss": 0.0385, + "step": 938 + }, + { + "epoch": 0.10179965307892455, + "grad_norm": 2.2446417808532715, + "learning_rate": 1.939141157696214e-05, + "loss": 0.0463, + "step": 939 + }, + { + "epoch": 0.10190806591500434, + "grad_norm": 5.669506072998047, + "learning_rate": 1.939068620339475e-05, + "loss": 0.0952, + "step": 940 + }, + { + "epoch": 0.10201647875108413, + "grad_norm": 6.788668155670166, + "learning_rate": 1.9389960829827363e-05, + "loss": 0.0588, + "step": 941 + }, + { + "epoch": 0.10212489158716392, + "grad_norm": 3.998284339904785, + "learning_rate": 1.9389235456259975e-05, + "loss": 0.0843, + "step": 942 + }, + { + "epoch": 0.10223330442324371, + "grad_norm": 4.536159038543701, + "learning_rate": 1.9388510082692587e-05, + "loss": 0.058, + "step": 943 + }, + { + "epoch": 0.1023417172593235, + "grad_norm": 2.644707679748535, + "learning_rate": 1.9387784709125203e-05, + "loss": 0.0512, + "step": 944 + }, + { + "epoch": 0.1024501300954033, + "grad_norm": 1.351309895515442, + "learning_rate": 1.9387059335557815e-05, + "loss": 0.026, + "step": 945 + }, + { + "epoch": 0.10255854293148309, + "grad_norm": 3.109466314315796, + "learning_rate": 1.9386333961990427e-05, + "loss": 0.0398, + "step": 946 + }, + { + "epoch": 0.10266695576756288, + "grad_norm": 3.9789836406707764, + "learning_rate": 1.938560858842304e-05, + "loss": 0.0574, + "step": 947 + }, + { + "epoch": 0.10277536860364267, + "grad_norm": 3.6100268363952637, + "learning_rate": 1.938488321485565e-05, + "loss": 0.0561, + "step": 948 + }, + { + "epoch": 0.10288378143972246, + "grad_norm": 4.309483528137207, + "learning_rate": 1.9384157841288264e-05, + "loss": 0.0583, + "step": 949 + }, + { + "epoch": 0.10299219427580225, + "grad_norm": 2.3464438915252686, + "learning_rate": 1.9383432467720876e-05, + "loss": 0.0422, + "step": 950 + }, + { + "epoch": 0.10310060711188204, + "grad_norm": 3.363886594772339, + "learning_rate": 1.938270709415349e-05, + "loss": 0.0514, + "step": 951 + }, + { + "epoch": 0.10320901994796183, + "grad_norm": 3.3630259037017822, + "learning_rate": 1.9381981720586103e-05, + "loss": 0.0499, + "step": 952 + }, + { + "epoch": 0.10331743278404162, + "grad_norm": 4.765928745269775, + "learning_rate": 1.9381256347018716e-05, + "loss": 0.0628, + "step": 953 + }, + { + "epoch": 0.10342584562012143, + "grad_norm": 5.115732669830322, + "learning_rate": 1.9380530973451328e-05, + "loss": 0.0542, + "step": 954 + }, + { + "epoch": 0.10353425845620122, + "grad_norm": 5.754235744476318, + "learning_rate": 1.937980559988394e-05, + "loss": 0.0834, + "step": 955 + }, + { + "epoch": 0.10364267129228101, + "grad_norm": 9.645397186279297, + "learning_rate": 1.9379080226316555e-05, + "loss": 0.1054, + "step": 956 + }, + { + "epoch": 0.1037510841283608, + "grad_norm": 3.0257928371429443, + "learning_rate": 1.9378354852749168e-05, + "loss": 0.0664, + "step": 957 + }, + { + "epoch": 0.10385949696444059, + "grad_norm": 2.86427903175354, + "learning_rate": 1.937762947918178e-05, + "loss": 0.0601, + "step": 958 + }, + { + "epoch": 0.10396790980052038, + "grad_norm": 4.439101696014404, + "learning_rate": 1.9376904105614395e-05, + "loss": 0.0417, + "step": 959 + }, + { + "epoch": 0.10407632263660017, + "grad_norm": 5.0267720222473145, + "learning_rate": 1.9376178732047007e-05, + "loss": 0.025, + "step": 960 + }, + { + "epoch": 0.10418473547267997, + "grad_norm": 3.085222005844116, + "learning_rate": 1.937545335847962e-05, + "loss": 0.0284, + "step": 961 + }, + { + "epoch": 0.10429314830875976, + "grad_norm": 2.9868335723876953, + "learning_rate": 1.937472798491223e-05, + "loss": 0.0417, + "step": 962 + }, + { + "epoch": 0.10440156114483955, + "grad_norm": 2.347395420074463, + "learning_rate": 1.9374002611344844e-05, + "loss": 0.0229, + "step": 963 + }, + { + "epoch": 0.10450997398091934, + "grad_norm": 2.7579517364501953, + "learning_rate": 1.9373277237777456e-05, + "loss": 0.0709, + "step": 964 + }, + { + "epoch": 0.10461838681699913, + "grad_norm": 8.904274940490723, + "learning_rate": 1.9372551864210068e-05, + "loss": 0.0692, + "step": 965 + }, + { + "epoch": 0.10472679965307892, + "grad_norm": 4.2662553787231445, + "learning_rate": 1.9371826490642684e-05, + "loss": 0.0544, + "step": 966 + }, + { + "epoch": 0.10483521248915871, + "grad_norm": 2.289334297180176, + "learning_rate": 1.9371101117075296e-05, + "loss": 0.0483, + "step": 967 + }, + { + "epoch": 0.1049436253252385, + "grad_norm": 2.5926222801208496, + "learning_rate": 1.9370375743507908e-05, + "loss": 0.0242, + "step": 968 + }, + { + "epoch": 0.1050520381613183, + "grad_norm": 2.504258394241333, + "learning_rate": 1.936965036994052e-05, + "loss": 0.0315, + "step": 969 + }, + { + "epoch": 0.1051604509973981, + "grad_norm": 5.883917808532715, + "learning_rate": 1.9368924996373132e-05, + "loss": 0.0739, + "step": 970 + }, + { + "epoch": 0.10526886383347789, + "grad_norm": 2.3552908897399902, + "learning_rate": 1.9368199622805744e-05, + "loss": 0.0258, + "step": 971 + }, + { + "epoch": 0.10537727666955768, + "grad_norm": 6.904064178466797, + "learning_rate": 1.936747424923836e-05, + "loss": 0.0995, + "step": 972 + }, + { + "epoch": 0.10548568950563747, + "grad_norm": 4.934479713439941, + "learning_rate": 1.9366748875670972e-05, + "loss": 0.0545, + "step": 973 + }, + { + "epoch": 0.10559410234171726, + "grad_norm": 1.5419316291809082, + "learning_rate": 1.9366023502103584e-05, + "loss": 0.0349, + "step": 974 + }, + { + "epoch": 0.10570251517779705, + "grad_norm": 5.002201557159424, + "learning_rate": 1.93652981285362e-05, + "loss": 0.0519, + "step": 975 + }, + { + "epoch": 0.10581092801387684, + "grad_norm": 2.1225991249084473, + "learning_rate": 1.9364572754968812e-05, + "loss": 0.0206, + "step": 976 + }, + { + "epoch": 0.10591934084995663, + "grad_norm": 2.218329429626465, + "learning_rate": 1.9363847381401424e-05, + "loss": 0.0344, + "step": 977 + }, + { + "epoch": 0.10602775368603642, + "grad_norm": 8.355992317199707, + "learning_rate": 1.9363122007834036e-05, + "loss": 0.0797, + "step": 978 + }, + { + "epoch": 0.10613616652211622, + "grad_norm": 4.072930812835693, + "learning_rate": 1.9362396634266648e-05, + "loss": 0.0589, + "step": 979 + }, + { + "epoch": 0.106244579358196, + "grad_norm": 2.1519525051116943, + "learning_rate": 1.936167126069926e-05, + "loss": 0.0178, + "step": 980 + }, + { + "epoch": 0.1063529921942758, + "grad_norm": 3.530991554260254, + "learning_rate": 1.9360945887131876e-05, + "loss": 0.0647, + "step": 981 + }, + { + "epoch": 0.10646140503035559, + "grad_norm": 4.495023727416992, + "learning_rate": 1.9360220513564488e-05, + "loss": 0.0639, + "step": 982 + }, + { + "epoch": 0.10656981786643538, + "grad_norm": 3.267656087875366, + "learning_rate": 1.93594951399971e-05, + "loss": 0.0473, + "step": 983 + }, + { + "epoch": 0.10667823070251518, + "grad_norm": 2.2635738849639893, + "learning_rate": 1.9358769766429712e-05, + "loss": 0.037, + "step": 984 + }, + { + "epoch": 0.10678664353859497, + "grad_norm": 5.31442403793335, + "learning_rate": 1.9358044392862325e-05, + "loss": 0.0762, + "step": 985 + }, + { + "epoch": 0.10689505637467477, + "grad_norm": 6.006298065185547, + "learning_rate": 1.9357319019294937e-05, + "loss": 0.056, + "step": 986 + }, + { + "epoch": 0.10700346921075456, + "grad_norm": 3.397385835647583, + "learning_rate": 1.9356593645727552e-05, + "loss": 0.0205, + "step": 987 + }, + { + "epoch": 0.10711188204683435, + "grad_norm": 4.285634994506836, + "learning_rate": 1.9355868272160164e-05, + "loss": 0.0439, + "step": 988 + }, + { + "epoch": 0.10722029488291414, + "grad_norm": 3.9945361614227295, + "learning_rate": 1.9355142898592776e-05, + "loss": 0.0412, + "step": 989 + }, + { + "epoch": 0.10732870771899393, + "grad_norm": 2.134857654571533, + "learning_rate": 1.9354417525025392e-05, + "loss": 0.0416, + "step": 990 + }, + { + "epoch": 0.10743712055507372, + "grad_norm": 4.743936061859131, + "learning_rate": 1.9353692151458004e-05, + "loss": 0.0648, + "step": 991 + }, + { + "epoch": 0.10754553339115351, + "grad_norm": 4.222489833831787, + "learning_rate": 1.9352966777890616e-05, + "loss": 0.0561, + "step": 992 + }, + { + "epoch": 0.1076539462272333, + "grad_norm": 2.2362682819366455, + "learning_rate": 1.935224140432323e-05, + "loss": 0.035, + "step": 993 + }, + { + "epoch": 0.10776235906331309, + "grad_norm": 5.947559833526611, + "learning_rate": 1.935151603075584e-05, + "loss": 0.0429, + "step": 994 + }, + { + "epoch": 0.10787077189939288, + "grad_norm": 4.9501447677612305, + "learning_rate": 1.9350790657188453e-05, + "loss": 0.1276, + "step": 995 + }, + { + "epoch": 0.10797918473547267, + "grad_norm": 3.2228729724884033, + "learning_rate": 1.9350065283621068e-05, + "loss": 0.0204, + "step": 996 + }, + { + "epoch": 0.10808759757155247, + "grad_norm": 2.6675901412963867, + "learning_rate": 1.934933991005368e-05, + "loss": 0.0259, + "step": 997 + }, + { + "epoch": 0.10819601040763227, + "grad_norm": 8.700155258178711, + "learning_rate": 1.9348614536486293e-05, + "loss": 0.0597, + "step": 998 + }, + { + "epoch": 0.10830442324371206, + "grad_norm": 4.360569953918457, + "learning_rate": 1.9347889162918905e-05, + "loss": 0.077, + "step": 999 + }, + { + "epoch": 0.10841283607979185, + "grad_norm": 4.094634056091309, + "learning_rate": 1.9347163789351517e-05, + "loss": 0.0588, + "step": 1000 + }, + { + "epoch": 0.10852124891587164, + "grad_norm": 2.0986618995666504, + "learning_rate": 1.934643841578413e-05, + "loss": 0.0303, + "step": 1001 + }, + { + "epoch": 0.10862966175195143, + "grad_norm": 4.143651008605957, + "learning_rate": 1.934571304221674e-05, + "loss": 0.0377, + "step": 1002 + }, + { + "epoch": 0.10873807458803122, + "grad_norm": 2.4457180500030518, + "learning_rate": 1.9344987668649357e-05, + "loss": 0.0324, + "step": 1003 + }, + { + "epoch": 0.10884648742411102, + "grad_norm": 3.7569382190704346, + "learning_rate": 1.934426229508197e-05, + "loss": 0.0672, + "step": 1004 + }, + { + "epoch": 0.1089549002601908, + "grad_norm": 1.4398865699768066, + "learning_rate": 1.934353692151458e-05, + "loss": 0.0237, + "step": 1005 + }, + { + "epoch": 0.1090633130962706, + "grad_norm": 1.5539031028747559, + "learning_rate": 1.9342811547947193e-05, + "loss": 0.0232, + "step": 1006 + }, + { + "epoch": 0.10917172593235039, + "grad_norm": 4.406874656677246, + "learning_rate": 1.934208617437981e-05, + "loss": 0.0904, + "step": 1007 + }, + { + "epoch": 0.10928013876843018, + "grad_norm": 3.2183852195739746, + "learning_rate": 1.934136080081242e-05, + "loss": 0.0253, + "step": 1008 + }, + { + "epoch": 0.10938855160450997, + "grad_norm": 0.7787014245986938, + "learning_rate": 1.9340635427245033e-05, + "loss": 0.0105, + "step": 1009 + }, + { + "epoch": 0.10949696444058976, + "grad_norm": 2.2766926288604736, + "learning_rate": 1.9339910053677645e-05, + "loss": 0.0348, + "step": 1010 + }, + { + "epoch": 0.10960537727666955, + "grad_norm": 3.8529248237609863, + "learning_rate": 1.933918468011026e-05, + "loss": 0.052, + "step": 1011 + }, + { + "epoch": 0.10971379011274934, + "grad_norm": 9.804343223571777, + "learning_rate": 1.9338459306542873e-05, + "loss": 0.0702, + "step": 1012 + }, + { + "epoch": 0.10982220294882915, + "grad_norm": 5.389828681945801, + "learning_rate": 1.9337733932975485e-05, + "loss": 0.0387, + "step": 1013 + }, + { + "epoch": 0.10993061578490894, + "grad_norm": 5.1237335205078125, + "learning_rate": 1.9337008559408097e-05, + "loss": 0.0773, + "step": 1014 + }, + { + "epoch": 0.11003902862098873, + "grad_norm": 3.614504814147949, + "learning_rate": 1.933628318584071e-05, + "loss": 0.0486, + "step": 1015 + }, + { + "epoch": 0.11014744145706852, + "grad_norm": 4.617439270019531, + "learning_rate": 1.933555781227332e-05, + "loss": 0.0772, + "step": 1016 + }, + { + "epoch": 0.11025585429314831, + "grad_norm": 3.8516197204589844, + "learning_rate": 1.9334832438705933e-05, + "loss": 0.0316, + "step": 1017 + }, + { + "epoch": 0.1103642671292281, + "grad_norm": 3.2416605949401855, + "learning_rate": 1.933410706513855e-05, + "loss": 0.0506, + "step": 1018 + }, + { + "epoch": 0.1104726799653079, + "grad_norm": 5.142822742462158, + "learning_rate": 1.933338169157116e-05, + "loss": 0.0381, + "step": 1019 + }, + { + "epoch": 0.11058109280138768, + "grad_norm": 4.753759860992432, + "learning_rate": 1.9332656318003773e-05, + "loss": 0.107, + "step": 1020 + }, + { + "epoch": 0.11068950563746748, + "grad_norm": 2.148019313812256, + "learning_rate": 1.9331930944436385e-05, + "loss": 0.0421, + "step": 1021 + }, + { + "epoch": 0.11079791847354727, + "grad_norm": 2.937791109085083, + "learning_rate": 1.9331205570868998e-05, + "loss": 0.0187, + "step": 1022 + }, + { + "epoch": 0.11090633130962706, + "grad_norm": 3.8443591594696045, + "learning_rate": 1.933048019730161e-05, + "loss": 0.0646, + "step": 1023 + }, + { + "epoch": 0.11101474414570685, + "grad_norm": 9.496718406677246, + "learning_rate": 1.9329754823734225e-05, + "loss": 0.0614, + "step": 1024 + }, + { + "epoch": 0.11112315698178664, + "grad_norm": 3.1218972206115723, + "learning_rate": 1.9329029450166837e-05, + "loss": 0.0586, + "step": 1025 + }, + { + "epoch": 0.11123156981786643, + "grad_norm": 3.0111422538757324, + "learning_rate": 1.932830407659945e-05, + "loss": 0.0348, + "step": 1026 + }, + { + "epoch": 0.11133998265394622, + "grad_norm": 1.5309468507766724, + "learning_rate": 1.9327578703032065e-05, + "loss": 0.0123, + "step": 1027 + }, + { + "epoch": 0.11144839549002603, + "grad_norm": 4.8348236083984375, + "learning_rate": 1.9326853329464677e-05, + "loss": 0.0582, + "step": 1028 + }, + { + "epoch": 0.11155680832610582, + "grad_norm": 2.1481330394744873, + "learning_rate": 1.932612795589729e-05, + "loss": 0.0345, + "step": 1029 + }, + { + "epoch": 0.11166522116218561, + "grad_norm": 1.5070487260818481, + "learning_rate": 1.93254025823299e-05, + "loss": 0.02, + "step": 1030 + }, + { + "epoch": 0.1117736339982654, + "grad_norm": 9.66238784790039, + "learning_rate": 1.9324677208762514e-05, + "loss": 0.1353, + "step": 1031 + }, + { + "epoch": 0.11188204683434519, + "grad_norm": 6.9368367195129395, + "learning_rate": 1.9323951835195126e-05, + "loss": 0.0467, + "step": 1032 + }, + { + "epoch": 0.11199045967042498, + "grad_norm": 2.0223896503448486, + "learning_rate": 1.932322646162774e-05, + "loss": 0.0292, + "step": 1033 + }, + { + "epoch": 0.11209887250650477, + "grad_norm": 5.592133045196533, + "learning_rate": 1.9322501088060353e-05, + "loss": 0.0573, + "step": 1034 + }, + { + "epoch": 0.11220728534258456, + "grad_norm": 5.635114669799805, + "learning_rate": 1.9321775714492966e-05, + "loss": 0.0584, + "step": 1035 + }, + { + "epoch": 0.11231569817866435, + "grad_norm": 4.665574550628662, + "learning_rate": 1.9321050340925578e-05, + "loss": 0.0608, + "step": 1036 + }, + { + "epoch": 0.11242411101474414, + "grad_norm": 4.536787033081055, + "learning_rate": 1.932032496735819e-05, + "loss": 0.0705, + "step": 1037 + }, + { + "epoch": 0.11253252385082393, + "grad_norm": 4.839252948760986, + "learning_rate": 1.9319599593790802e-05, + "loss": 0.05, + "step": 1038 + }, + { + "epoch": 0.11264093668690373, + "grad_norm": 6.85059928894043, + "learning_rate": 1.9318874220223418e-05, + "loss": 0.0911, + "step": 1039 + }, + { + "epoch": 0.11274934952298352, + "grad_norm": 2.6016476154327393, + "learning_rate": 1.931814884665603e-05, + "loss": 0.0392, + "step": 1040 + }, + { + "epoch": 0.11285776235906331, + "grad_norm": 5.200837135314941, + "learning_rate": 1.9317423473088642e-05, + "loss": 0.0945, + "step": 1041 + }, + { + "epoch": 0.1129661751951431, + "grad_norm": 5.648915767669678, + "learning_rate": 1.9316698099521257e-05, + "loss": 0.0515, + "step": 1042 + }, + { + "epoch": 0.1130745880312229, + "grad_norm": 5.452339172363281, + "learning_rate": 1.931597272595387e-05, + "loss": 0.0507, + "step": 1043 + }, + { + "epoch": 0.1131830008673027, + "grad_norm": 7.344969272613525, + "learning_rate": 1.931524735238648e-05, + "loss": 0.0624, + "step": 1044 + }, + { + "epoch": 0.11329141370338248, + "grad_norm": 2.1254425048828125, + "learning_rate": 1.9314521978819094e-05, + "loss": 0.0374, + "step": 1045 + }, + { + "epoch": 0.11339982653946228, + "grad_norm": 1.939237117767334, + "learning_rate": 1.9313796605251706e-05, + "loss": 0.0385, + "step": 1046 + }, + { + "epoch": 0.11350823937554207, + "grad_norm": 5.643937587738037, + "learning_rate": 1.9313071231684318e-05, + "loss": 0.062, + "step": 1047 + }, + { + "epoch": 0.11361665221162186, + "grad_norm": 3.1810386180877686, + "learning_rate": 1.9312345858116934e-05, + "loss": 0.0804, + "step": 1048 + }, + { + "epoch": 0.11372506504770165, + "grad_norm": 2.302656412124634, + "learning_rate": 1.9311620484549546e-05, + "loss": 0.0362, + "step": 1049 + }, + { + "epoch": 0.11383347788378144, + "grad_norm": 2.6449615955352783, + "learning_rate": 1.9310895110982158e-05, + "loss": 0.0563, + "step": 1050 + }, + { + "epoch": 0.11394189071986123, + "grad_norm": 3.0949959754943848, + "learning_rate": 1.931016973741477e-05, + "loss": 0.0441, + "step": 1051 + }, + { + "epoch": 0.11405030355594102, + "grad_norm": 1.9239611625671387, + "learning_rate": 1.9309444363847382e-05, + "loss": 0.0298, + "step": 1052 + }, + { + "epoch": 0.11415871639202081, + "grad_norm": 6.173488140106201, + "learning_rate": 1.9308718990279994e-05, + "loss": 0.0995, + "step": 1053 + }, + { + "epoch": 0.1142671292281006, + "grad_norm": 3.695042848587036, + "learning_rate": 1.9307993616712606e-05, + "loss": 0.0794, + "step": 1054 + }, + { + "epoch": 0.1143755420641804, + "grad_norm": 2.723478078842163, + "learning_rate": 1.9307268243145222e-05, + "loss": 0.0402, + "step": 1055 + }, + { + "epoch": 0.11448395490026018, + "grad_norm": 3.781778335571289, + "learning_rate": 1.9306542869577834e-05, + "loss": 0.0489, + "step": 1056 + }, + { + "epoch": 0.11459236773633999, + "grad_norm": 5.718045711517334, + "learning_rate": 1.9305817496010446e-05, + "loss": 0.0282, + "step": 1057 + }, + { + "epoch": 0.11470078057241978, + "grad_norm": 1.3789453506469727, + "learning_rate": 1.930509212244306e-05, + "loss": 0.0177, + "step": 1058 + }, + { + "epoch": 0.11480919340849957, + "grad_norm": 3.658168315887451, + "learning_rate": 1.9304366748875674e-05, + "loss": 0.101, + "step": 1059 + }, + { + "epoch": 0.11491760624457936, + "grad_norm": 2.8674256801605225, + "learning_rate": 1.9303641375308286e-05, + "loss": 0.0271, + "step": 1060 + }, + { + "epoch": 0.11502601908065915, + "grad_norm": 2.503366470336914, + "learning_rate": 1.9302916001740898e-05, + "loss": 0.0516, + "step": 1061 + }, + { + "epoch": 0.11513443191673894, + "grad_norm": 3.7374610900878906, + "learning_rate": 1.930219062817351e-05, + "loss": 0.0588, + "step": 1062 + }, + { + "epoch": 0.11524284475281873, + "grad_norm": 3.543269634246826, + "learning_rate": 1.9301465254606126e-05, + "loss": 0.0428, + "step": 1063 + }, + { + "epoch": 0.11535125758889853, + "grad_norm": 5.421008586883545, + "learning_rate": 1.9300739881038738e-05, + "loss": 0.0692, + "step": 1064 + }, + { + "epoch": 0.11545967042497832, + "grad_norm": 2.765151262283325, + "learning_rate": 1.930001450747135e-05, + "loss": 0.0445, + "step": 1065 + }, + { + "epoch": 0.11556808326105811, + "grad_norm": 4.077737808227539, + "learning_rate": 1.9299289133903962e-05, + "loss": 0.0779, + "step": 1066 + }, + { + "epoch": 0.1156764960971379, + "grad_norm": 2.987248420715332, + "learning_rate": 1.9298563760336574e-05, + "loss": 0.0611, + "step": 1067 + }, + { + "epoch": 0.11578490893321769, + "grad_norm": 3.1041789054870605, + "learning_rate": 1.9297838386769187e-05, + "loss": 0.0312, + "step": 1068 + }, + { + "epoch": 0.11589332176929748, + "grad_norm": 3.9829394817352295, + "learning_rate": 1.92971130132018e-05, + "loss": 0.1076, + "step": 1069 + }, + { + "epoch": 0.11600173460537727, + "grad_norm": 1.7150590419769287, + "learning_rate": 1.9296387639634414e-05, + "loss": 0.0159, + "step": 1070 + }, + { + "epoch": 0.11611014744145706, + "grad_norm": 2.5707173347473145, + "learning_rate": 1.9295662266067026e-05, + "loss": 0.0235, + "step": 1071 + }, + { + "epoch": 0.11621856027753687, + "grad_norm": 3.6098828315734863, + "learning_rate": 1.929493689249964e-05, + "loss": 0.0642, + "step": 1072 + }, + { + "epoch": 0.11632697311361666, + "grad_norm": 4.495543956756592, + "learning_rate": 1.929421151893225e-05, + "loss": 0.0552, + "step": 1073 + }, + { + "epoch": 0.11643538594969645, + "grad_norm": 4.3741254806518555, + "learning_rate": 1.9293486145364863e-05, + "loss": 0.0398, + "step": 1074 + }, + { + "epoch": 0.11654379878577624, + "grad_norm": 5.540560722351074, + "learning_rate": 1.9292760771797475e-05, + "loss": 0.0194, + "step": 1075 + }, + { + "epoch": 0.11665221162185603, + "grad_norm": 2.8695201873779297, + "learning_rate": 1.929203539823009e-05, + "loss": 0.0199, + "step": 1076 + }, + { + "epoch": 0.11676062445793582, + "grad_norm": 3.3501651287078857, + "learning_rate": 1.9291310024662703e-05, + "loss": 0.0373, + "step": 1077 + }, + { + "epoch": 0.11686903729401561, + "grad_norm": 3.3605923652648926, + "learning_rate": 1.9290584651095315e-05, + "loss": 0.057, + "step": 1078 + }, + { + "epoch": 0.1169774501300954, + "grad_norm": 4.166085720062256, + "learning_rate": 1.928985927752793e-05, + "loss": 0.044, + "step": 1079 + }, + { + "epoch": 0.1170858629661752, + "grad_norm": 4.550034999847412, + "learning_rate": 1.9289133903960543e-05, + "loss": 0.0913, + "step": 1080 + }, + { + "epoch": 0.11719427580225499, + "grad_norm": 7.368113994598389, + "learning_rate": 1.9288408530393155e-05, + "loss": 0.0329, + "step": 1081 + }, + { + "epoch": 0.11730268863833478, + "grad_norm": 3.454575777053833, + "learning_rate": 1.9287683156825767e-05, + "loss": 0.035, + "step": 1082 + }, + { + "epoch": 0.11741110147441457, + "grad_norm": 4.425349712371826, + "learning_rate": 1.928695778325838e-05, + "loss": 0.0442, + "step": 1083 + }, + { + "epoch": 0.11751951431049436, + "grad_norm": 6.804541110992432, + "learning_rate": 1.928623240969099e-05, + "loss": 0.0757, + "step": 1084 + }, + { + "epoch": 0.11762792714657415, + "grad_norm": 8.344995498657227, + "learning_rate": 1.9285507036123607e-05, + "loss": 0.0394, + "step": 1085 + }, + { + "epoch": 0.11773633998265394, + "grad_norm": 7.425122261047363, + "learning_rate": 1.928478166255622e-05, + "loss": 0.0539, + "step": 1086 + }, + { + "epoch": 0.11784475281873374, + "grad_norm": 2.7670373916625977, + "learning_rate": 1.928405628898883e-05, + "loss": 0.0555, + "step": 1087 + }, + { + "epoch": 0.11795316565481354, + "grad_norm": 2.5674545764923096, + "learning_rate": 1.9283330915421443e-05, + "loss": 0.0526, + "step": 1088 + }, + { + "epoch": 0.11806157849089333, + "grad_norm": 4.032062530517578, + "learning_rate": 1.9282605541854055e-05, + "loss": 0.0554, + "step": 1089 + }, + { + "epoch": 0.11816999132697312, + "grad_norm": 0.8893749713897705, + "learning_rate": 1.9281880168286667e-05, + "loss": 0.0108, + "step": 1090 + }, + { + "epoch": 0.11827840416305291, + "grad_norm": 2.441518783569336, + "learning_rate": 1.9281154794719283e-05, + "loss": 0.0562, + "step": 1091 + }, + { + "epoch": 0.1183868169991327, + "grad_norm": 5.568926811218262, + "learning_rate": 1.9280429421151895e-05, + "loss": 0.0769, + "step": 1092 + }, + { + "epoch": 0.11849522983521249, + "grad_norm": 3.22296142578125, + "learning_rate": 1.9279704047584507e-05, + "loss": 0.0503, + "step": 1093 + }, + { + "epoch": 0.11860364267129228, + "grad_norm": 4.154721260070801, + "learning_rate": 1.9278978674017123e-05, + "loss": 0.0312, + "step": 1094 + }, + { + "epoch": 0.11871205550737207, + "grad_norm": 3.4323151111602783, + "learning_rate": 1.9278253300449735e-05, + "loss": 0.0629, + "step": 1095 + }, + { + "epoch": 0.11882046834345186, + "grad_norm": 3.5999412536621094, + "learning_rate": 1.9277527926882347e-05, + "loss": 0.0511, + "step": 1096 + }, + { + "epoch": 0.11892888117953165, + "grad_norm": 4.333664417266846, + "learning_rate": 1.927680255331496e-05, + "loss": 0.0561, + "step": 1097 + }, + { + "epoch": 0.11903729401561144, + "grad_norm": 7.672799587249756, + "learning_rate": 1.927607717974757e-05, + "loss": 0.1194, + "step": 1098 + }, + { + "epoch": 0.11914570685169124, + "grad_norm": 3.4288275241851807, + "learning_rate": 1.9275351806180183e-05, + "loss": 0.0514, + "step": 1099 + }, + { + "epoch": 0.11925411968777103, + "grad_norm": 12.751933097839355, + "learning_rate": 1.92746264326128e-05, + "loss": 0.0774, + "step": 1100 + }, + { + "epoch": 0.11936253252385082, + "grad_norm": 3.451085329055786, + "learning_rate": 1.927390105904541e-05, + "loss": 0.0524, + "step": 1101 + }, + { + "epoch": 0.11947094535993062, + "grad_norm": 3.49931263923645, + "learning_rate": 1.9273175685478023e-05, + "loss": 0.0421, + "step": 1102 + }, + { + "epoch": 0.11957935819601041, + "grad_norm": 2.350374221801758, + "learning_rate": 1.9272450311910635e-05, + "loss": 0.0264, + "step": 1103 + }, + { + "epoch": 0.1196877710320902, + "grad_norm": 2.845515251159668, + "learning_rate": 1.9271724938343248e-05, + "loss": 0.0412, + "step": 1104 + }, + { + "epoch": 0.11979618386817, + "grad_norm": 1.8426282405853271, + "learning_rate": 1.927099956477586e-05, + "loss": 0.0419, + "step": 1105 + }, + { + "epoch": 0.11990459670424979, + "grad_norm": 3.3670520782470703, + "learning_rate": 1.9270274191208472e-05, + "loss": 0.0493, + "step": 1106 + }, + { + "epoch": 0.12001300954032958, + "grad_norm": 6.303942680358887, + "learning_rate": 1.9269548817641087e-05, + "loss": 0.072, + "step": 1107 + }, + { + "epoch": 0.12012142237640937, + "grad_norm": 1.8817834854125977, + "learning_rate": 1.92688234440737e-05, + "loss": 0.0245, + "step": 1108 + }, + { + "epoch": 0.12022983521248916, + "grad_norm": 2.3913910388946533, + "learning_rate": 1.926809807050631e-05, + "loss": 0.0338, + "step": 1109 + }, + { + "epoch": 0.12033824804856895, + "grad_norm": 4.235208034515381, + "learning_rate": 1.9267372696938924e-05, + "loss": 0.0469, + "step": 1110 + }, + { + "epoch": 0.12044666088464874, + "grad_norm": 15.647963523864746, + "learning_rate": 1.926664732337154e-05, + "loss": 0.048, + "step": 1111 + }, + { + "epoch": 0.12055507372072853, + "grad_norm": 6.44443941116333, + "learning_rate": 1.926592194980415e-05, + "loss": 0.0865, + "step": 1112 + }, + { + "epoch": 0.12066348655680832, + "grad_norm": 4.950728416442871, + "learning_rate": 1.9265196576236764e-05, + "loss": 0.0842, + "step": 1113 + }, + { + "epoch": 0.12077189939288811, + "grad_norm": 3.9263930320739746, + "learning_rate": 1.9264471202669376e-05, + "loss": 0.0531, + "step": 1114 + }, + { + "epoch": 0.1208803122289679, + "grad_norm": 7.719465255737305, + "learning_rate": 1.926374582910199e-05, + "loss": 0.0569, + "step": 1115 + }, + { + "epoch": 0.1209887250650477, + "grad_norm": 3.9861080646514893, + "learning_rate": 1.9263020455534603e-05, + "loss": 0.0514, + "step": 1116 + }, + { + "epoch": 0.1210971379011275, + "grad_norm": 1.3499609231948853, + "learning_rate": 1.9262295081967216e-05, + "loss": 0.0144, + "step": 1117 + }, + { + "epoch": 0.12120555073720729, + "grad_norm": 2.170278549194336, + "learning_rate": 1.9261569708399828e-05, + "loss": 0.0458, + "step": 1118 + }, + { + "epoch": 0.12131396357328708, + "grad_norm": 4.0117058753967285, + "learning_rate": 1.926084433483244e-05, + "loss": 0.0491, + "step": 1119 + }, + { + "epoch": 0.12142237640936687, + "grad_norm": 3.3749916553497314, + "learning_rate": 1.9260118961265052e-05, + "loss": 0.0479, + "step": 1120 + }, + { + "epoch": 0.12153078924544666, + "grad_norm": 3.1554503440856934, + "learning_rate": 1.9259393587697664e-05, + "loss": 0.0724, + "step": 1121 + }, + { + "epoch": 0.12163920208152645, + "grad_norm": 5.36997127532959, + "learning_rate": 1.925866821413028e-05, + "loss": 0.0829, + "step": 1122 + }, + { + "epoch": 0.12174761491760624, + "grad_norm": 1.2115401029586792, + "learning_rate": 1.9257942840562892e-05, + "loss": 0.0072, + "step": 1123 + }, + { + "epoch": 0.12185602775368604, + "grad_norm": 3.3908801078796387, + "learning_rate": 1.9257217466995504e-05, + "loss": 0.036, + "step": 1124 + }, + { + "epoch": 0.12196444058976583, + "grad_norm": 2.031550168991089, + "learning_rate": 1.9256492093428116e-05, + "loss": 0.0376, + "step": 1125 + }, + { + "epoch": 0.12207285342584562, + "grad_norm": 2.4098358154296875, + "learning_rate": 1.9255766719860728e-05, + "loss": 0.0477, + "step": 1126 + }, + { + "epoch": 0.12218126626192541, + "grad_norm": 7.45513916015625, + "learning_rate": 1.925504134629334e-05, + "loss": 0.1483, + "step": 1127 + }, + { + "epoch": 0.1222896790980052, + "grad_norm": 7.267603397369385, + "learning_rate": 1.9254315972725956e-05, + "loss": 0.0462, + "step": 1128 + }, + { + "epoch": 0.12239809193408499, + "grad_norm": 2.3340814113616943, + "learning_rate": 1.9253590599158568e-05, + "loss": 0.0361, + "step": 1129 + }, + { + "epoch": 0.12250650477016478, + "grad_norm": 5.4939069747924805, + "learning_rate": 1.9252865225591184e-05, + "loss": 0.0454, + "step": 1130 + }, + { + "epoch": 0.12261491760624459, + "grad_norm": 2.782931327819824, + "learning_rate": 1.9252139852023796e-05, + "loss": 0.0474, + "step": 1131 + }, + { + "epoch": 0.12272333044232438, + "grad_norm": 4.235967636108398, + "learning_rate": 1.9251414478456408e-05, + "loss": 0.051, + "step": 1132 + }, + { + "epoch": 0.12283174327840417, + "grad_norm": 4.353731632232666, + "learning_rate": 1.925068910488902e-05, + "loss": 0.0254, + "step": 1133 + }, + { + "epoch": 0.12294015611448396, + "grad_norm": 3.295511484146118, + "learning_rate": 1.9249963731321632e-05, + "loss": 0.0494, + "step": 1134 + }, + { + "epoch": 0.12304856895056375, + "grad_norm": 9.447793960571289, + "learning_rate": 1.9249238357754244e-05, + "loss": 0.0649, + "step": 1135 + }, + { + "epoch": 0.12315698178664354, + "grad_norm": 4.1348443031311035, + "learning_rate": 1.9248512984186856e-05, + "loss": 0.0511, + "step": 1136 + }, + { + "epoch": 0.12326539462272333, + "grad_norm": 4.470040321350098, + "learning_rate": 1.9247787610619472e-05, + "loss": 0.0536, + "step": 1137 + }, + { + "epoch": 0.12337380745880312, + "grad_norm": 6.68125057220459, + "learning_rate": 1.9247062237052084e-05, + "loss": 0.0661, + "step": 1138 + }, + { + "epoch": 0.12348222029488291, + "grad_norm": 3.75789213180542, + "learning_rate": 1.9246336863484696e-05, + "loss": 0.0706, + "step": 1139 + }, + { + "epoch": 0.1235906331309627, + "grad_norm": 3.3418922424316406, + "learning_rate": 1.924561148991731e-05, + "loss": 0.0254, + "step": 1140 + }, + { + "epoch": 0.1236990459670425, + "grad_norm": 3.1340177059173584, + "learning_rate": 1.924488611634992e-05, + "loss": 0.0577, + "step": 1141 + }, + { + "epoch": 0.12380745880312229, + "grad_norm": 3.968933343887329, + "learning_rate": 1.9244160742782533e-05, + "loss": 0.0421, + "step": 1142 + }, + { + "epoch": 0.12391587163920208, + "grad_norm": 2.801959991455078, + "learning_rate": 1.9243435369215148e-05, + "loss": 0.0328, + "step": 1143 + }, + { + "epoch": 0.12402428447528187, + "grad_norm": 5.56210994720459, + "learning_rate": 1.924270999564776e-05, + "loss": 0.0917, + "step": 1144 + }, + { + "epoch": 0.12413269731136166, + "grad_norm": 4.366122245788574, + "learning_rate": 1.9241984622080372e-05, + "loss": 0.0412, + "step": 1145 + }, + { + "epoch": 0.12424111014744146, + "grad_norm": 6.115045547485352, + "learning_rate": 1.9241259248512988e-05, + "loss": 0.1081, + "step": 1146 + }, + { + "epoch": 0.12434952298352125, + "grad_norm": 3.006269693374634, + "learning_rate": 1.92405338749456e-05, + "loss": 0.0339, + "step": 1147 + }, + { + "epoch": 0.12445793581960105, + "grad_norm": 3.735149383544922, + "learning_rate": 1.9239808501378212e-05, + "loss": 0.0486, + "step": 1148 + }, + { + "epoch": 0.12456634865568084, + "grad_norm": 1.2028173208236694, + "learning_rate": 1.9239083127810824e-05, + "loss": 0.0137, + "step": 1149 + }, + { + "epoch": 0.12467476149176063, + "grad_norm": 1.0271188020706177, + "learning_rate": 1.9238357754243437e-05, + "loss": 0.0156, + "step": 1150 + }, + { + "epoch": 0.12478317432784042, + "grad_norm": 2.590895414352417, + "learning_rate": 1.923763238067605e-05, + "loss": 0.0363, + "step": 1151 + }, + { + "epoch": 0.12489158716392021, + "grad_norm": 4.647332191467285, + "learning_rate": 1.9236907007108664e-05, + "loss": 0.0362, + "step": 1152 + }, + { + "epoch": 0.125, + "grad_norm": 3.507371187210083, + "learning_rate": 1.9236181633541276e-05, + "loss": 0.0676, + "step": 1153 + }, + { + "epoch": 0.1251084128360798, + "grad_norm": 3.258491039276123, + "learning_rate": 1.923545625997389e-05, + "loss": 0.0543, + "step": 1154 + }, + { + "epoch": 0.12521682567215958, + "grad_norm": 3.2764813899993896, + "learning_rate": 1.92347308864065e-05, + "loss": 0.022, + "step": 1155 + }, + { + "epoch": 0.12532523850823937, + "grad_norm": 2.902959108352661, + "learning_rate": 1.9234005512839113e-05, + "loss": 0.0635, + "step": 1156 + }, + { + "epoch": 0.12543365134431916, + "grad_norm": 2.2020761966705322, + "learning_rate": 1.9233280139271725e-05, + "loss": 0.0213, + "step": 1157 + }, + { + "epoch": 0.12554206418039895, + "grad_norm": 3.9264585971832275, + "learning_rate": 1.9232554765704337e-05, + "loss": 0.0571, + "step": 1158 + }, + { + "epoch": 0.12565047701647875, + "grad_norm": 4.83601188659668, + "learning_rate": 1.9231829392136953e-05, + "loss": 0.0173, + "step": 1159 + }, + { + "epoch": 0.12575888985255854, + "grad_norm": 7.664961814880371, + "learning_rate": 1.9231104018569565e-05, + "loss": 0.0768, + "step": 1160 + }, + { + "epoch": 0.12586730268863833, + "grad_norm": 3.6537721157073975, + "learning_rate": 1.9230378645002177e-05, + "loss": 0.0314, + "step": 1161 + }, + { + "epoch": 0.12597571552471812, + "grad_norm": 6.88771390914917, + "learning_rate": 1.922965327143479e-05, + "loss": 0.0703, + "step": 1162 + }, + { + "epoch": 0.1260841283607979, + "grad_norm": 2.8854568004608154, + "learning_rate": 1.9228927897867405e-05, + "loss": 0.0119, + "step": 1163 + }, + { + "epoch": 0.1261925411968777, + "grad_norm": 4.49363374710083, + "learning_rate": 1.9228202524300017e-05, + "loss": 0.0523, + "step": 1164 + }, + { + "epoch": 0.1263009540329575, + "grad_norm": 2.388197660446167, + "learning_rate": 1.922747715073263e-05, + "loss": 0.0189, + "step": 1165 + }, + { + "epoch": 0.12640936686903728, + "grad_norm": 2.0625431537628174, + "learning_rate": 1.922675177716524e-05, + "loss": 0.0349, + "step": 1166 + }, + { + "epoch": 0.12651777970511707, + "grad_norm": 4.242441177368164, + "learning_rate": 1.9226026403597857e-05, + "loss": 0.0395, + "step": 1167 + }, + { + "epoch": 0.1266261925411969, + "grad_norm": 8.8588285446167, + "learning_rate": 1.922530103003047e-05, + "loss": 0.055, + "step": 1168 + }, + { + "epoch": 0.12673460537727668, + "grad_norm": 6.846457004547119, + "learning_rate": 1.922457565646308e-05, + "loss": 0.0693, + "step": 1169 + }, + { + "epoch": 0.12684301821335647, + "grad_norm": 5.095306873321533, + "learning_rate": 1.9223850282895693e-05, + "loss": 0.0727, + "step": 1170 + }, + { + "epoch": 0.12695143104943626, + "grad_norm": 6.443818092346191, + "learning_rate": 1.9223124909328305e-05, + "loss": 0.0771, + "step": 1171 + }, + { + "epoch": 0.12705984388551606, + "grad_norm": 3.7412984371185303, + "learning_rate": 1.9222399535760917e-05, + "loss": 0.0233, + "step": 1172 + }, + { + "epoch": 0.12716825672159585, + "grad_norm": 2.121922016143799, + "learning_rate": 1.922167416219353e-05, + "loss": 0.0227, + "step": 1173 + }, + { + "epoch": 0.12727666955767564, + "grad_norm": 2.920807361602783, + "learning_rate": 1.9220948788626145e-05, + "loss": 0.0262, + "step": 1174 + }, + { + "epoch": 0.12738508239375543, + "grad_norm": 4.795711517333984, + "learning_rate": 1.9220223415058757e-05, + "loss": 0.0506, + "step": 1175 + }, + { + "epoch": 0.12749349522983522, + "grad_norm": 3.7042770385742188, + "learning_rate": 1.921949804149137e-05, + "loss": 0.0634, + "step": 1176 + }, + { + "epoch": 0.127601908065915, + "grad_norm": 5.959484100341797, + "learning_rate": 1.921877266792398e-05, + "loss": 0.1195, + "step": 1177 + }, + { + "epoch": 0.1277103209019948, + "grad_norm": 2.860337018966675, + "learning_rate": 1.9218047294356594e-05, + "loss": 0.0549, + "step": 1178 + }, + { + "epoch": 0.1278187337380746, + "grad_norm": 4.377027988433838, + "learning_rate": 1.9217321920789206e-05, + "loss": 0.03, + "step": 1179 + }, + { + "epoch": 0.12792714657415438, + "grad_norm": 3.5321907997131348, + "learning_rate": 1.921659654722182e-05, + "loss": 0.04, + "step": 1180 + }, + { + "epoch": 0.12803555941023417, + "grad_norm": 14.747949600219727, + "learning_rate": 1.9215871173654433e-05, + "loss": 0.0905, + "step": 1181 + }, + { + "epoch": 0.12814397224631396, + "grad_norm": 3.3942835330963135, + "learning_rate": 1.921514580008705e-05, + "loss": 0.0541, + "step": 1182 + }, + { + "epoch": 0.12825238508239376, + "grad_norm": 1.9659472703933716, + "learning_rate": 1.921442042651966e-05, + "loss": 0.0546, + "step": 1183 + }, + { + "epoch": 0.12836079791847355, + "grad_norm": 2.5037877559661865, + "learning_rate": 1.9213695052952273e-05, + "loss": 0.0433, + "step": 1184 + }, + { + "epoch": 0.12846921075455334, + "grad_norm": 2.7070236206054688, + "learning_rate": 1.9212969679384885e-05, + "loss": 0.0583, + "step": 1185 + }, + { + "epoch": 0.12857762359063313, + "grad_norm": 4.4624924659729, + "learning_rate": 1.9212244305817497e-05, + "loss": 0.0686, + "step": 1186 + }, + { + "epoch": 0.12868603642671292, + "grad_norm": 5.3482842445373535, + "learning_rate": 1.921151893225011e-05, + "loss": 0.0869, + "step": 1187 + }, + { + "epoch": 0.1287944492627927, + "grad_norm": 2.244420051574707, + "learning_rate": 1.9210793558682722e-05, + "loss": 0.0451, + "step": 1188 + }, + { + "epoch": 0.1289028620988725, + "grad_norm": 2.146698474884033, + "learning_rate": 1.9210068185115337e-05, + "loss": 0.0486, + "step": 1189 + }, + { + "epoch": 0.1290112749349523, + "grad_norm": 0.7233756184577942, + "learning_rate": 1.920934281154795e-05, + "loss": 0.0077, + "step": 1190 + }, + { + "epoch": 0.12911968777103208, + "grad_norm": 3.402175188064575, + "learning_rate": 1.920861743798056e-05, + "loss": 0.059, + "step": 1191 + }, + { + "epoch": 0.12922810060711187, + "grad_norm": 2.35970401763916, + "learning_rate": 1.9207892064413174e-05, + "loss": 0.043, + "step": 1192 + }, + { + "epoch": 0.12933651344319166, + "grad_norm": 5.3700103759765625, + "learning_rate": 1.9207166690845786e-05, + "loss": 0.0619, + "step": 1193 + }, + { + "epoch": 0.12944492627927146, + "grad_norm": 2.009695291519165, + "learning_rate": 1.9206441317278398e-05, + "loss": 0.0237, + "step": 1194 + }, + { + "epoch": 0.12955333911535125, + "grad_norm": 3.188138961791992, + "learning_rate": 1.9205715943711014e-05, + "loss": 0.0654, + "step": 1195 + }, + { + "epoch": 0.12966175195143104, + "grad_norm": 4.251428127288818, + "learning_rate": 1.9204990570143626e-05, + "loss": 0.0682, + "step": 1196 + }, + { + "epoch": 0.12977016478751083, + "grad_norm": 3.8546833992004395, + "learning_rate": 1.9204265196576238e-05, + "loss": 0.0266, + "step": 1197 + }, + { + "epoch": 0.12987857762359065, + "grad_norm": 3.813593864440918, + "learning_rate": 1.9203539823008853e-05, + "loss": 0.0466, + "step": 1198 + }, + { + "epoch": 0.12998699045967044, + "grad_norm": 1.9764575958251953, + "learning_rate": 1.9202814449441466e-05, + "loss": 0.032, + "step": 1199 + }, + { + "epoch": 0.13009540329575023, + "grad_norm": 1.8420960903167725, + "learning_rate": 1.9202089075874078e-05, + "loss": 0.0246, + "step": 1200 + }, + { + "epoch": 0.13020381613183002, + "grad_norm": 6.076582431793213, + "learning_rate": 1.920136370230669e-05, + "loss": 0.0911, + "step": 1201 + }, + { + "epoch": 0.1303122289679098, + "grad_norm": 5.6047749519348145, + "learning_rate": 1.9200638328739302e-05, + "loss": 0.0407, + "step": 1202 + }, + { + "epoch": 0.1304206418039896, + "grad_norm": 12.654923439025879, + "learning_rate": 1.9199912955171914e-05, + "loss": 0.0501, + "step": 1203 + }, + { + "epoch": 0.1305290546400694, + "grad_norm": 3.767827033996582, + "learning_rate": 1.919918758160453e-05, + "loss": 0.0354, + "step": 1204 + }, + { + "epoch": 0.13063746747614918, + "grad_norm": 7.049373149871826, + "learning_rate": 1.9198462208037142e-05, + "loss": 0.0433, + "step": 1205 + }, + { + "epoch": 0.13074588031222897, + "grad_norm": 3.1705260276794434, + "learning_rate": 1.9197736834469754e-05, + "loss": 0.0219, + "step": 1206 + }, + { + "epoch": 0.13085429314830876, + "grad_norm": 5.505399703979492, + "learning_rate": 1.9197011460902366e-05, + "loss": 0.0469, + "step": 1207 + }, + { + "epoch": 0.13096270598438856, + "grad_norm": 3.4204533100128174, + "learning_rate": 1.9196286087334978e-05, + "loss": 0.0714, + "step": 1208 + }, + { + "epoch": 0.13107111882046835, + "grad_norm": 4.899256229400635, + "learning_rate": 1.919556071376759e-05, + "loss": 0.0483, + "step": 1209 + }, + { + "epoch": 0.13117953165654814, + "grad_norm": 1.1855862140655518, + "learning_rate": 1.9194835340200202e-05, + "loss": 0.0209, + "step": 1210 + }, + { + "epoch": 0.13128794449262793, + "grad_norm": 4.0755534172058105, + "learning_rate": 1.9194109966632818e-05, + "loss": 0.0971, + "step": 1211 + }, + { + "epoch": 0.13139635732870772, + "grad_norm": 5.384300231933594, + "learning_rate": 1.919338459306543e-05, + "loss": 0.0655, + "step": 1212 + }, + { + "epoch": 0.1315047701647875, + "grad_norm": 2.134268283843994, + "learning_rate": 1.9192659219498042e-05, + "loss": 0.0364, + "step": 1213 + }, + { + "epoch": 0.1316131830008673, + "grad_norm": 2.888718605041504, + "learning_rate": 1.9191933845930654e-05, + "loss": 0.0633, + "step": 1214 + }, + { + "epoch": 0.1317215958369471, + "grad_norm": 4.0023274421691895, + "learning_rate": 1.919120847236327e-05, + "loss": 0.0394, + "step": 1215 + }, + { + "epoch": 0.13183000867302688, + "grad_norm": 8.737356185913086, + "learning_rate": 1.9190483098795882e-05, + "loss": 0.0779, + "step": 1216 + }, + { + "epoch": 0.13193842150910667, + "grad_norm": 4.052062511444092, + "learning_rate": 1.9189757725228494e-05, + "loss": 0.0265, + "step": 1217 + }, + { + "epoch": 0.13204683434518646, + "grad_norm": 2.58437442779541, + "learning_rate": 1.9189032351661106e-05, + "loss": 0.023, + "step": 1218 + }, + { + "epoch": 0.13215524718126626, + "grad_norm": 5.418023109436035, + "learning_rate": 1.9188306978093722e-05, + "loss": 0.1046, + "step": 1219 + }, + { + "epoch": 0.13226366001734605, + "grad_norm": 3.569227695465088, + "learning_rate": 1.9187581604526334e-05, + "loss": 0.0597, + "step": 1220 + }, + { + "epoch": 0.13237207285342584, + "grad_norm": 1.9888674020767212, + "learning_rate": 1.9186856230958946e-05, + "loss": 0.0241, + "step": 1221 + }, + { + "epoch": 0.13248048568950563, + "grad_norm": 3.534299373626709, + "learning_rate": 1.918613085739156e-05, + "loss": 0.0284, + "step": 1222 + }, + { + "epoch": 0.13258889852558542, + "grad_norm": 5.777711391448975, + "learning_rate": 1.918540548382417e-05, + "loss": 0.12, + "step": 1223 + }, + { + "epoch": 0.1326973113616652, + "grad_norm": 4.002889156341553, + "learning_rate": 1.9184680110256783e-05, + "loss": 0.0725, + "step": 1224 + }, + { + "epoch": 0.132805724197745, + "grad_norm": 4.106054782867432, + "learning_rate": 1.9183954736689395e-05, + "loss": 0.0387, + "step": 1225 + }, + { + "epoch": 0.1329141370338248, + "grad_norm": 2.8488829135894775, + "learning_rate": 1.918322936312201e-05, + "loss": 0.0281, + "step": 1226 + }, + { + "epoch": 0.1330225498699046, + "grad_norm": 3.6946754455566406, + "learning_rate": 1.9182503989554622e-05, + "loss": 0.043, + "step": 1227 + }, + { + "epoch": 0.1331309627059844, + "grad_norm": 2.817021131515503, + "learning_rate": 1.9181778615987235e-05, + "loss": 0.0676, + "step": 1228 + }, + { + "epoch": 0.1332393755420642, + "grad_norm": 5.295990943908691, + "learning_rate": 1.9181053242419847e-05, + "loss": 0.0883, + "step": 1229 + }, + { + "epoch": 0.13334778837814398, + "grad_norm": 4.259591102600098, + "learning_rate": 1.918032786885246e-05, + "loss": 0.0549, + "step": 1230 + }, + { + "epoch": 0.13345620121422377, + "grad_norm": 5.479467868804932, + "learning_rate": 1.917960249528507e-05, + "loss": 0.1177, + "step": 1231 + }, + { + "epoch": 0.13356461405030357, + "grad_norm": 16.1825008392334, + "learning_rate": 1.9178877121717687e-05, + "loss": 0.0643, + "step": 1232 + }, + { + "epoch": 0.13367302688638336, + "grad_norm": 3.397393226623535, + "learning_rate": 1.91781517481503e-05, + "loss": 0.0747, + "step": 1233 + }, + { + "epoch": 0.13378143972246315, + "grad_norm": 4.764841556549072, + "learning_rate": 1.9177426374582914e-05, + "loss": 0.0774, + "step": 1234 + }, + { + "epoch": 0.13388985255854294, + "grad_norm": 4.359391689300537, + "learning_rate": 1.9176701001015526e-05, + "loss": 0.0704, + "step": 1235 + }, + { + "epoch": 0.13399826539462273, + "grad_norm": 4.405508041381836, + "learning_rate": 1.917597562744814e-05, + "loss": 0.0936, + "step": 1236 + }, + { + "epoch": 0.13410667823070252, + "grad_norm": 4.095656394958496, + "learning_rate": 1.917525025388075e-05, + "loss": 0.0532, + "step": 1237 + }, + { + "epoch": 0.1342150910667823, + "grad_norm": 3.731290102005005, + "learning_rate": 1.9174524880313363e-05, + "loss": 0.0337, + "step": 1238 + }, + { + "epoch": 0.1343235039028621, + "grad_norm": 2.2770798206329346, + "learning_rate": 1.9173799506745975e-05, + "loss": 0.0254, + "step": 1239 + }, + { + "epoch": 0.1344319167389419, + "grad_norm": 3.4548773765563965, + "learning_rate": 1.9173074133178587e-05, + "loss": 0.0435, + "step": 1240 + }, + { + "epoch": 0.13454032957502168, + "grad_norm": 4.440578460693359, + "learning_rate": 1.9172348759611203e-05, + "loss": 0.0702, + "step": 1241 + }, + { + "epoch": 0.13464874241110147, + "grad_norm": 3.797715425491333, + "learning_rate": 1.9171623386043815e-05, + "loss": 0.0661, + "step": 1242 + }, + { + "epoch": 0.13475715524718127, + "grad_norm": 2.139801263809204, + "learning_rate": 1.9170898012476427e-05, + "loss": 0.0356, + "step": 1243 + }, + { + "epoch": 0.13486556808326106, + "grad_norm": 3.6093244552612305, + "learning_rate": 1.917017263890904e-05, + "loss": 0.0614, + "step": 1244 + }, + { + "epoch": 0.13497398091934085, + "grad_norm": 1.344985842704773, + "learning_rate": 1.916944726534165e-05, + "loss": 0.0291, + "step": 1245 + }, + { + "epoch": 0.13508239375542064, + "grad_norm": 1.5274993181228638, + "learning_rate": 1.9168721891774263e-05, + "loss": 0.0244, + "step": 1246 + }, + { + "epoch": 0.13519080659150043, + "grad_norm": 2.4597344398498535, + "learning_rate": 1.916799651820688e-05, + "loss": 0.0315, + "step": 1247 + }, + { + "epoch": 0.13529921942758022, + "grad_norm": 3.920032262802124, + "learning_rate": 1.916727114463949e-05, + "loss": 0.0497, + "step": 1248 + }, + { + "epoch": 0.13540763226366, + "grad_norm": 5.935805320739746, + "learning_rate": 1.9166545771072103e-05, + "loss": 0.0491, + "step": 1249 + }, + { + "epoch": 0.1355160450997398, + "grad_norm": 6.125152111053467, + "learning_rate": 1.916582039750472e-05, + "loss": 0.0578, + "step": 1250 + }, + { + "epoch": 0.1356244579358196, + "grad_norm": 2.6976401805877686, + "learning_rate": 1.916509502393733e-05, + "loss": 0.0513, + "step": 1251 + }, + { + "epoch": 0.13573287077189938, + "grad_norm": 6.341214656829834, + "learning_rate": 1.9164369650369943e-05, + "loss": 0.0434, + "step": 1252 + }, + { + "epoch": 0.13584128360797917, + "grad_norm": 4.453490257263184, + "learning_rate": 1.9163644276802555e-05, + "loss": 0.0553, + "step": 1253 + }, + { + "epoch": 0.13594969644405897, + "grad_norm": 2.932094097137451, + "learning_rate": 1.9162918903235167e-05, + "loss": 0.0512, + "step": 1254 + }, + { + "epoch": 0.13605810928013876, + "grad_norm": 5.09404182434082, + "learning_rate": 1.916219352966778e-05, + "loss": 0.0607, + "step": 1255 + }, + { + "epoch": 0.13616652211621855, + "grad_norm": 2.961238145828247, + "learning_rate": 1.9161468156100395e-05, + "loss": 0.0587, + "step": 1256 + }, + { + "epoch": 0.13627493495229837, + "grad_norm": 4.191768646240234, + "learning_rate": 1.9160742782533007e-05, + "loss": 0.0525, + "step": 1257 + }, + { + "epoch": 0.13638334778837816, + "grad_norm": 9.107038497924805, + "learning_rate": 1.916001740896562e-05, + "loss": 0.0809, + "step": 1258 + }, + { + "epoch": 0.13649176062445795, + "grad_norm": 9.804431915283203, + "learning_rate": 1.915929203539823e-05, + "loss": 0.0606, + "step": 1259 + }, + { + "epoch": 0.13660017346053774, + "grad_norm": 5.491832256317139, + "learning_rate": 1.9158566661830844e-05, + "loss": 0.0721, + "step": 1260 + }, + { + "epoch": 0.13670858629661753, + "grad_norm": 5.758472442626953, + "learning_rate": 1.9157841288263456e-05, + "loss": 0.0594, + "step": 1261 + }, + { + "epoch": 0.13681699913269732, + "grad_norm": 4.133118152618408, + "learning_rate": 1.9157115914696068e-05, + "loss": 0.0475, + "step": 1262 + }, + { + "epoch": 0.1369254119687771, + "grad_norm": 3.4755754470825195, + "learning_rate": 1.9156390541128683e-05, + "loss": 0.0238, + "step": 1263 + }, + { + "epoch": 0.1370338248048569, + "grad_norm": 3.0394980907440186, + "learning_rate": 1.9155665167561295e-05, + "loss": 0.0337, + "step": 1264 + }, + { + "epoch": 0.1371422376409367, + "grad_norm": 2.565267324447632, + "learning_rate": 1.9154939793993908e-05, + "loss": 0.0391, + "step": 1265 + }, + { + "epoch": 0.13725065047701648, + "grad_norm": 3.4451560974121094, + "learning_rate": 1.915421442042652e-05, + "loss": 0.0458, + "step": 1266 + }, + { + "epoch": 0.13735906331309627, + "grad_norm": 1.6359723806381226, + "learning_rate": 1.9153489046859135e-05, + "loss": 0.0112, + "step": 1267 + }, + { + "epoch": 0.13746747614917607, + "grad_norm": 10.937738418579102, + "learning_rate": 1.9152763673291747e-05, + "loss": 0.1579, + "step": 1268 + }, + { + "epoch": 0.13757588898525586, + "grad_norm": 4.8218278884887695, + "learning_rate": 1.915203829972436e-05, + "loss": 0.0907, + "step": 1269 + }, + { + "epoch": 0.13768430182133565, + "grad_norm": 4.522794723510742, + "learning_rate": 1.9151312926156972e-05, + "loss": 0.0716, + "step": 1270 + }, + { + "epoch": 0.13779271465741544, + "grad_norm": 5.80670166015625, + "learning_rate": 1.9150587552589587e-05, + "loss": 0.0886, + "step": 1271 + }, + { + "epoch": 0.13790112749349523, + "grad_norm": 4.143108367919922, + "learning_rate": 1.91498621790222e-05, + "loss": 0.0529, + "step": 1272 + }, + { + "epoch": 0.13800954032957502, + "grad_norm": 2.5175864696502686, + "learning_rate": 1.914913680545481e-05, + "loss": 0.0203, + "step": 1273 + }, + { + "epoch": 0.1381179531656548, + "grad_norm": 5.595544338226318, + "learning_rate": 1.9148411431887424e-05, + "loss": 0.0278, + "step": 1274 + }, + { + "epoch": 0.1382263660017346, + "grad_norm": 3.6583337783813477, + "learning_rate": 1.9147686058320036e-05, + "loss": 0.041, + "step": 1275 + }, + { + "epoch": 0.1383347788378144, + "grad_norm": 3.0859262943267822, + "learning_rate": 1.9146960684752648e-05, + "loss": 0.0433, + "step": 1276 + }, + { + "epoch": 0.13844319167389418, + "grad_norm": 2.4804317951202393, + "learning_rate": 1.914623531118526e-05, + "loss": 0.0322, + "step": 1277 + }, + { + "epoch": 0.13855160450997397, + "grad_norm": 6.616519451141357, + "learning_rate": 1.9145509937617876e-05, + "loss": 0.0415, + "step": 1278 + }, + { + "epoch": 0.13866001734605377, + "grad_norm": 3.2764809131622314, + "learning_rate": 1.9144784564050488e-05, + "loss": 0.0511, + "step": 1279 + }, + { + "epoch": 0.13876843018213356, + "grad_norm": 9.8041410446167, + "learning_rate": 1.91440591904831e-05, + "loss": 0.028, + "step": 1280 + }, + { + "epoch": 0.13887684301821335, + "grad_norm": 2.251103401184082, + "learning_rate": 1.9143333816915712e-05, + "loss": 0.0421, + "step": 1281 + }, + { + "epoch": 0.13898525585429314, + "grad_norm": 4.330857276916504, + "learning_rate": 1.9142608443348324e-05, + "loss": 0.0938, + "step": 1282 + }, + { + "epoch": 0.13909366869037293, + "grad_norm": 1.3638548851013184, + "learning_rate": 1.9141883069780936e-05, + "loss": 0.0202, + "step": 1283 + }, + { + "epoch": 0.13920208152645272, + "grad_norm": 1.5806902647018433, + "learning_rate": 1.9141157696213552e-05, + "loss": 0.0295, + "step": 1284 + }, + { + "epoch": 0.1393104943625325, + "grad_norm": 3.395331382751465, + "learning_rate": 1.9140432322646164e-05, + "loss": 0.0479, + "step": 1285 + }, + { + "epoch": 0.1394189071986123, + "grad_norm": 3.142953872680664, + "learning_rate": 1.913970694907878e-05, + "loss": 0.1107, + "step": 1286 + }, + { + "epoch": 0.13952732003469212, + "grad_norm": 5.696978569030762, + "learning_rate": 1.9138981575511392e-05, + "loss": 0.0757, + "step": 1287 + }, + { + "epoch": 0.1396357328707719, + "grad_norm": 2.7593417167663574, + "learning_rate": 1.9138256201944004e-05, + "loss": 0.0531, + "step": 1288 + }, + { + "epoch": 0.1397441457068517, + "grad_norm": 2.416794776916504, + "learning_rate": 1.9137530828376616e-05, + "loss": 0.0618, + "step": 1289 + }, + { + "epoch": 0.1398525585429315, + "grad_norm": 4.085274696350098, + "learning_rate": 1.9136805454809228e-05, + "loss": 0.1076, + "step": 1290 + }, + { + "epoch": 0.13996097137901128, + "grad_norm": 2.637925148010254, + "learning_rate": 1.913608008124184e-05, + "loss": 0.0264, + "step": 1291 + }, + { + "epoch": 0.14006938421509108, + "grad_norm": 3.4670825004577637, + "learning_rate": 1.9135354707674452e-05, + "loss": 0.0563, + "step": 1292 + }, + { + "epoch": 0.14017779705117087, + "grad_norm": 4.077402114868164, + "learning_rate": 1.9134629334107068e-05, + "loss": 0.0832, + "step": 1293 + }, + { + "epoch": 0.14028620988725066, + "grad_norm": 5.148631572723389, + "learning_rate": 1.913390396053968e-05, + "loss": 0.0488, + "step": 1294 + }, + { + "epoch": 0.14039462272333045, + "grad_norm": 2.0921261310577393, + "learning_rate": 1.9133178586972292e-05, + "loss": 0.0282, + "step": 1295 + }, + { + "epoch": 0.14050303555941024, + "grad_norm": 6.211324214935303, + "learning_rate": 1.9132453213404904e-05, + "loss": 0.1028, + "step": 1296 + }, + { + "epoch": 0.14061144839549003, + "grad_norm": 2.3986079692840576, + "learning_rate": 1.9131727839837517e-05, + "loss": 0.0539, + "step": 1297 + }, + { + "epoch": 0.14071986123156982, + "grad_norm": 1.732200264930725, + "learning_rate": 1.913100246627013e-05, + "loss": 0.0356, + "step": 1298 + }, + { + "epoch": 0.1408282740676496, + "grad_norm": 3.2988412380218506, + "learning_rate": 1.9130277092702744e-05, + "loss": 0.0695, + "step": 1299 + }, + { + "epoch": 0.1409366869037294, + "grad_norm": 0.8782435655593872, + "learning_rate": 1.9129551719135356e-05, + "loss": 0.015, + "step": 1300 + }, + { + "epoch": 0.1410450997398092, + "grad_norm": 4.786044597625732, + "learning_rate": 1.912882634556797e-05, + "loss": 0.0475, + "step": 1301 + }, + { + "epoch": 0.14115351257588898, + "grad_norm": 1.5948207378387451, + "learning_rate": 1.9128100972000584e-05, + "loss": 0.0328, + "step": 1302 + }, + { + "epoch": 0.14126192541196878, + "grad_norm": 5.253811836242676, + "learning_rate": 1.9127375598433196e-05, + "loss": 0.0717, + "step": 1303 + }, + { + "epoch": 0.14137033824804857, + "grad_norm": 2.587606430053711, + "learning_rate": 1.912665022486581e-05, + "loss": 0.0217, + "step": 1304 + }, + { + "epoch": 0.14147875108412836, + "grad_norm": 6.60851526260376, + "learning_rate": 1.912592485129842e-05, + "loss": 0.073, + "step": 1305 + }, + { + "epoch": 0.14158716392020815, + "grad_norm": 4.501598358154297, + "learning_rate": 1.9125199477731033e-05, + "loss": 0.0588, + "step": 1306 + }, + { + "epoch": 0.14169557675628794, + "grad_norm": 3.047086477279663, + "learning_rate": 1.9124474104163645e-05, + "loss": 0.038, + "step": 1307 + }, + { + "epoch": 0.14180398959236773, + "grad_norm": 3.3635518550872803, + "learning_rate": 1.912374873059626e-05, + "loss": 0.0251, + "step": 1308 + }, + { + "epoch": 0.14191240242844752, + "grad_norm": 3.1848623752593994, + "learning_rate": 1.9123023357028872e-05, + "loss": 0.032, + "step": 1309 + }, + { + "epoch": 0.1420208152645273, + "grad_norm": 2.8914778232574463, + "learning_rate": 1.9122297983461485e-05, + "loss": 0.0406, + "step": 1310 + }, + { + "epoch": 0.1421292281006071, + "grad_norm": 1.4939366579055786, + "learning_rate": 1.9121572609894097e-05, + "loss": 0.0322, + "step": 1311 + }, + { + "epoch": 0.1422376409366869, + "grad_norm": 9.386131286621094, + "learning_rate": 1.912084723632671e-05, + "loss": 0.0621, + "step": 1312 + }, + { + "epoch": 0.14234605377276668, + "grad_norm": 11.125380516052246, + "learning_rate": 1.912012186275932e-05, + "loss": 0.0521, + "step": 1313 + }, + { + "epoch": 0.14245446660884648, + "grad_norm": 4.714948654174805, + "learning_rate": 1.9119396489191933e-05, + "loss": 0.0423, + "step": 1314 + }, + { + "epoch": 0.14256287944492627, + "grad_norm": 8.295570373535156, + "learning_rate": 1.911867111562455e-05, + "loss": 0.0782, + "step": 1315 + }, + { + "epoch": 0.14267129228100608, + "grad_norm": 4.206087112426758, + "learning_rate": 1.911794574205716e-05, + "loss": 0.0306, + "step": 1316 + }, + { + "epoch": 0.14277970511708588, + "grad_norm": 3.9826769828796387, + "learning_rate": 1.9117220368489773e-05, + "loss": 0.0984, + "step": 1317 + }, + { + "epoch": 0.14288811795316567, + "grad_norm": 11.265451431274414, + "learning_rate": 1.911649499492239e-05, + "loss": 0.08, + "step": 1318 + }, + { + "epoch": 0.14299653078924546, + "grad_norm": 3.6267502307891846, + "learning_rate": 1.9115769621355e-05, + "loss": 0.066, + "step": 1319 + }, + { + "epoch": 0.14310494362532525, + "grad_norm": 2.9988083839416504, + "learning_rate": 1.9115044247787613e-05, + "loss": 0.0528, + "step": 1320 + }, + { + "epoch": 0.14321335646140504, + "grad_norm": 3.094489336013794, + "learning_rate": 1.9114318874220225e-05, + "loss": 0.0319, + "step": 1321 + }, + { + "epoch": 0.14332176929748483, + "grad_norm": 4.910242557525635, + "learning_rate": 1.9113593500652837e-05, + "loss": 0.1168, + "step": 1322 + }, + { + "epoch": 0.14343018213356462, + "grad_norm": 5.280900001525879, + "learning_rate": 1.9112868127085453e-05, + "loss": 0.0728, + "step": 1323 + }, + { + "epoch": 0.1435385949696444, + "grad_norm": 0.748489260673523, + "learning_rate": 1.9112142753518065e-05, + "loss": 0.0083, + "step": 1324 + }, + { + "epoch": 0.1436470078057242, + "grad_norm": 2.2958035469055176, + "learning_rate": 1.9111417379950677e-05, + "loss": 0.025, + "step": 1325 + }, + { + "epoch": 0.143755420641804, + "grad_norm": 2.6413047313690186, + "learning_rate": 1.911069200638329e-05, + "loss": 0.048, + "step": 1326 + }, + { + "epoch": 0.14386383347788378, + "grad_norm": 3.8399250507354736, + "learning_rate": 1.91099666328159e-05, + "loss": 0.0468, + "step": 1327 + }, + { + "epoch": 0.14397224631396358, + "grad_norm": 5.787071704864502, + "learning_rate": 1.9109241259248513e-05, + "loss": 0.0533, + "step": 1328 + }, + { + "epoch": 0.14408065915004337, + "grad_norm": 3.6964662075042725, + "learning_rate": 1.9108515885681125e-05, + "loss": 0.058, + "step": 1329 + }, + { + "epoch": 0.14418907198612316, + "grad_norm": 3.139359951019287, + "learning_rate": 1.910779051211374e-05, + "loss": 0.0576, + "step": 1330 + }, + { + "epoch": 0.14429748482220295, + "grad_norm": 3.284067392349243, + "learning_rate": 1.9107065138546353e-05, + "loss": 0.0442, + "step": 1331 + }, + { + "epoch": 0.14440589765828274, + "grad_norm": 2.5978217124938965, + "learning_rate": 1.9106339764978965e-05, + "loss": 0.0463, + "step": 1332 + }, + { + "epoch": 0.14451431049436253, + "grad_norm": 2.2016916275024414, + "learning_rate": 1.9105614391411577e-05, + "loss": 0.0275, + "step": 1333 + }, + { + "epoch": 0.14462272333044232, + "grad_norm": 2.539658784866333, + "learning_rate": 1.910488901784419e-05, + "loss": 0.0303, + "step": 1334 + }, + { + "epoch": 0.1447311361665221, + "grad_norm": 1.4900686740875244, + "learning_rate": 1.9104163644276802e-05, + "loss": 0.0286, + "step": 1335 + }, + { + "epoch": 0.1448395490026019, + "grad_norm": 6.9648823738098145, + "learning_rate": 1.9103438270709417e-05, + "loss": 0.0397, + "step": 1336 + }, + { + "epoch": 0.1449479618386817, + "grad_norm": 7.110777378082275, + "learning_rate": 1.910271289714203e-05, + "loss": 0.0416, + "step": 1337 + }, + { + "epoch": 0.14505637467476148, + "grad_norm": 1.1598669290542603, + "learning_rate": 1.9101987523574645e-05, + "loss": 0.0222, + "step": 1338 + }, + { + "epoch": 0.14516478751084128, + "grad_norm": 2.3936827182769775, + "learning_rate": 1.9101262150007257e-05, + "loss": 0.051, + "step": 1339 + }, + { + "epoch": 0.14527320034692107, + "grad_norm": 2.870990514755249, + "learning_rate": 1.910053677643987e-05, + "loss": 0.069, + "step": 1340 + }, + { + "epoch": 0.14538161318300086, + "grad_norm": 3.486985921859741, + "learning_rate": 1.909981140287248e-05, + "loss": 0.0742, + "step": 1341 + }, + { + "epoch": 0.14549002601908065, + "grad_norm": 3.3636605739593506, + "learning_rate": 1.9099086029305093e-05, + "loss": 0.0465, + "step": 1342 + }, + { + "epoch": 0.14559843885516044, + "grad_norm": 4.358847141265869, + "learning_rate": 1.9098360655737706e-05, + "loss": 0.0627, + "step": 1343 + }, + { + "epoch": 0.14570685169124023, + "grad_norm": 1.3642498254776, + "learning_rate": 1.9097635282170318e-05, + "loss": 0.0207, + "step": 1344 + }, + { + "epoch": 0.14581526452732002, + "grad_norm": 8.7581787109375, + "learning_rate": 1.9096909908602933e-05, + "loss": 0.0971, + "step": 1345 + }, + { + "epoch": 0.14592367736339984, + "grad_norm": 4.465841770172119, + "learning_rate": 1.9096184535035545e-05, + "loss": 0.0658, + "step": 1346 + }, + { + "epoch": 0.14603209019947963, + "grad_norm": 4.988814830780029, + "learning_rate": 1.9095459161468158e-05, + "loss": 0.0323, + "step": 1347 + }, + { + "epoch": 0.14614050303555942, + "grad_norm": 10.725008010864258, + "learning_rate": 1.909473378790077e-05, + "loss": 0.0277, + "step": 1348 + }, + { + "epoch": 0.1462489158716392, + "grad_norm": 3.725013494491577, + "learning_rate": 1.9094008414333382e-05, + "loss": 0.0379, + "step": 1349 + }, + { + "epoch": 0.146357328707719, + "grad_norm": 3.4554786682128906, + "learning_rate": 1.9093283040765994e-05, + "loss": 0.0373, + "step": 1350 + }, + { + "epoch": 0.1464657415437988, + "grad_norm": 5.1305389404296875, + "learning_rate": 1.909255766719861e-05, + "loss": 0.0327, + "step": 1351 + }, + { + "epoch": 0.14657415437987859, + "grad_norm": 1.7641351222991943, + "learning_rate": 1.9091832293631222e-05, + "loss": 0.0187, + "step": 1352 + }, + { + "epoch": 0.14668256721595838, + "grad_norm": 4.2398762702941895, + "learning_rate": 1.9091106920063834e-05, + "loss": 0.0327, + "step": 1353 + }, + { + "epoch": 0.14679098005203817, + "grad_norm": 2.6192681789398193, + "learning_rate": 1.909038154649645e-05, + "loss": 0.0549, + "step": 1354 + }, + { + "epoch": 0.14689939288811796, + "grad_norm": 2.4994962215423584, + "learning_rate": 1.908965617292906e-05, + "loss": 0.0326, + "step": 1355 + }, + { + "epoch": 0.14700780572419775, + "grad_norm": 2.278627634048462, + "learning_rate": 1.9088930799361674e-05, + "loss": 0.0564, + "step": 1356 + }, + { + "epoch": 0.14711621856027754, + "grad_norm": 10.054217338562012, + "learning_rate": 1.9088205425794286e-05, + "loss": 0.0928, + "step": 1357 + }, + { + "epoch": 0.14722463139635733, + "grad_norm": 3.2799479961395264, + "learning_rate": 1.9087480052226898e-05, + "loss": 0.0451, + "step": 1358 + }, + { + "epoch": 0.14733304423243712, + "grad_norm": 2.972947120666504, + "learning_rate": 1.908675467865951e-05, + "loss": 0.0412, + "step": 1359 + }, + { + "epoch": 0.1474414570685169, + "grad_norm": 2.43233060836792, + "learning_rate": 1.9086029305092126e-05, + "loss": 0.0408, + "step": 1360 + }, + { + "epoch": 0.1475498699045967, + "grad_norm": 5.660552978515625, + "learning_rate": 1.9085303931524738e-05, + "loss": 0.0647, + "step": 1361 + }, + { + "epoch": 0.1476582827406765, + "grad_norm": 0.6046493053436279, + "learning_rate": 1.908457855795735e-05, + "loss": 0.0054, + "step": 1362 + }, + { + "epoch": 0.14776669557675629, + "grad_norm": 8.790339469909668, + "learning_rate": 1.9083853184389962e-05, + "loss": 0.0333, + "step": 1363 + }, + { + "epoch": 0.14787510841283608, + "grad_norm": 5.034299373626709, + "learning_rate": 1.9083127810822574e-05, + "loss": 0.0493, + "step": 1364 + }, + { + "epoch": 0.14798352124891587, + "grad_norm": 5.6883625984191895, + "learning_rate": 1.9082402437255186e-05, + "loss": 0.0898, + "step": 1365 + }, + { + "epoch": 0.14809193408499566, + "grad_norm": 8.74362564086914, + "learning_rate": 1.90816770636878e-05, + "loss": 0.0589, + "step": 1366 + }, + { + "epoch": 0.14820034692107545, + "grad_norm": 3.723947525024414, + "learning_rate": 1.9080951690120414e-05, + "loss": 0.0427, + "step": 1367 + }, + { + "epoch": 0.14830875975715524, + "grad_norm": 2.7374167442321777, + "learning_rate": 1.9080226316553026e-05, + "loss": 0.0536, + "step": 1368 + }, + { + "epoch": 0.14841717259323503, + "grad_norm": 8.243258476257324, + "learning_rate": 1.9079500942985638e-05, + "loss": 0.08, + "step": 1369 + }, + { + "epoch": 0.14852558542931482, + "grad_norm": 4.988433837890625, + "learning_rate": 1.9078775569418254e-05, + "loss": 0.0489, + "step": 1370 + }, + { + "epoch": 0.1486339982653946, + "grad_norm": 3.0338940620422363, + "learning_rate": 1.9078050195850866e-05, + "loss": 0.0393, + "step": 1371 + }, + { + "epoch": 0.1487424111014744, + "grad_norm": 5.599745750427246, + "learning_rate": 1.9077324822283478e-05, + "loss": 0.1188, + "step": 1372 + }, + { + "epoch": 0.1488508239375542, + "grad_norm": 2.737233877182007, + "learning_rate": 1.907659944871609e-05, + "loss": 0.0551, + "step": 1373 + }, + { + "epoch": 0.14895923677363399, + "grad_norm": 5.024610996246338, + "learning_rate": 1.9075874075148702e-05, + "loss": 0.0905, + "step": 1374 + }, + { + "epoch": 0.1490676496097138, + "grad_norm": 5.028563022613525, + "learning_rate": 1.9075148701581318e-05, + "loss": 0.0388, + "step": 1375 + }, + { + "epoch": 0.1491760624457936, + "grad_norm": 2.2288427352905273, + "learning_rate": 1.907442332801393e-05, + "loss": 0.0214, + "step": 1376 + }, + { + "epoch": 0.14928447528187339, + "grad_norm": 4.5979437828063965, + "learning_rate": 1.9073697954446542e-05, + "loss": 0.0921, + "step": 1377 + }, + { + "epoch": 0.14939288811795318, + "grad_norm": 2.0612266063690186, + "learning_rate": 1.9072972580879154e-05, + "loss": 0.0541, + "step": 1378 + }, + { + "epoch": 0.14950130095403297, + "grad_norm": 3.7006547451019287, + "learning_rate": 1.9072247207311767e-05, + "loss": 0.0487, + "step": 1379 + }, + { + "epoch": 0.14960971379011276, + "grad_norm": 2.2059624195098877, + "learning_rate": 1.907152183374438e-05, + "loss": 0.0392, + "step": 1380 + }, + { + "epoch": 0.14971812662619255, + "grad_norm": 2.5265614986419678, + "learning_rate": 1.907079646017699e-05, + "loss": 0.0692, + "step": 1381 + }, + { + "epoch": 0.14982653946227234, + "grad_norm": 2.8920867443084717, + "learning_rate": 1.9070071086609606e-05, + "loss": 0.0422, + "step": 1382 + }, + { + "epoch": 0.14993495229835213, + "grad_norm": 4.199863433837891, + "learning_rate": 1.906934571304222e-05, + "loss": 0.0622, + "step": 1383 + }, + { + "epoch": 0.15004336513443192, + "grad_norm": 3.221388816833496, + "learning_rate": 1.906862033947483e-05, + "loss": 0.0704, + "step": 1384 + }, + { + "epoch": 0.1501517779705117, + "grad_norm": 4.24530553817749, + "learning_rate": 1.9067894965907443e-05, + "loss": 0.0586, + "step": 1385 + }, + { + "epoch": 0.1502601908065915, + "grad_norm": 7.189826965332031, + "learning_rate": 1.9067169592340055e-05, + "loss": 0.0502, + "step": 1386 + }, + { + "epoch": 0.1503686036426713, + "grad_norm": 1.6530559062957764, + "learning_rate": 1.9066444218772667e-05, + "loss": 0.0625, + "step": 1387 + }, + { + "epoch": 0.15047701647875109, + "grad_norm": 3.1176302433013916, + "learning_rate": 1.9065718845205283e-05, + "loss": 0.0663, + "step": 1388 + }, + { + "epoch": 0.15058542931483088, + "grad_norm": 2.0447535514831543, + "learning_rate": 1.9064993471637895e-05, + "loss": 0.0459, + "step": 1389 + }, + { + "epoch": 0.15069384215091067, + "grad_norm": 1.815861463546753, + "learning_rate": 1.906426809807051e-05, + "loss": 0.041, + "step": 1390 + }, + { + "epoch": 0.15080225498699046, + "grad_norm": 3.1534264087677, + "learning_rate": 1.9063542724503122e-05, + "loss": 0.0703, + "step": 1391 + }, + { + "epoch": 0.15091066782307025, + "grad_norm": 1.354274034500122, + "learning_rate": 1.9062817350935735e-05, + "loss": 0.0208, + "step": 1392 + }, + { + "epoch": 0.15101908065915004, + "grad_norm": 1.731326937675476, + "learning_rate": 1.9062091977368347e-05, + "loss": 0.0528, + "step": 1393 + }, + { + "epoch": 0.15112749349522983, + "grad_norm": 2.741539239883423, + "learning_rate": 1.906136660380096e-05, + "loss": 0.039, + "step": 1394 + }, + { + "epoch": 0.15123590633130962, + "grad_norm": 3.953502655029297, + "learning_rate": 1.906064123023357e-05, + "loss": 0.0576, + "step": 1395 + }, + { + "epoch": 0.1513443191673894, + "grad_norm": 5.129430294036865, + "learning_rate": 1.9059915856666183e-05, + "loss": 0.037, + "step": 1396 + }, + { + "epoch": 0.1514527320034692, + "grad_norm": 3.0686709880828857, + "learning_rate": 1.90591904830988e-05, + "loss": 0.0493, + "step": 1397 + }, + { + "epoch": 0.151561144839549, + "grad_norm": 3.869480609893799, + "learning_rate": 1.905846510953141e-05, + "loss": 0.0439, + "step": 1398 + }, + { + "epoch": 0.15166955767562879, + "grad_norm": 2.019409656524658, + "learning_rate": 1.9057739735964023e-05, + "loss": 0.0625, + "step": 1399 + }, + { + "epoch": 0.15177797051170858, + "grad_norm": 9.078502655029297, + "learning_rate": 1.9057014362396635e-05, + "loss": 0.1215, + "step": 1400 + }, + { + "epoch": 0.15188638334778837, + "grad_norm": 4.4946980476379395, + "learning_rate": 1.9056288988829247e-05, + "loss": 0.0504, + "step": 1401 + }, + { + "epoch": 0.15199479618386816, + "grad_norm": 7.747179985046387, + "learning_rate": 1.905556361526186e-05, + "loss": 0.0531, + "step": 1402 + }, + { + "epoch": 0.15210320901994795, + "grad_norm": 1.9182164669036865, + "learning_rate": 1.9054838241694475e-05, + "loss": 0.0374, + "step": 1403 + }, + { + "epoch": 0.15221162185602774, + "grad_norm": 3.8675484657287598, + "learning_rate": 1.9054112868127087e-05, + "loss": 0.0701, + "step": 1404 + }, + { + "epoch": 0.15232003469210756, + "grad_norm": 5.543450832366943, + "learning_rate": 1.90533874945597e-05, + "loss": 0.0616, + "step": 1405 + }, + { + "epoch": 0.15242844752818735, + "grad_norm": 4.734589099884033, + "learning_rate": 1.9052662120992315e-05, + "loss": 0.0459, + "step": 1406 + }, + { + "epoch": 0.15253686036426714, + "grad_norm": 6.562049865722656, + "learning_rate": 1.9051936747424927e-05, + "loss": 0.119, + "step": 1407 + }, + { + "epoch": 0.15264527320034693, + "grad_norm": 3.715792417526245, + "learning_rate": 1.905121137385754e-05, + "loss": 0.0337, + "step": 1408 + }, + { + "epoch": 0.15275368603642672, + "grad_norm": 2.505403995513916, + "learning_rate": 1.905048600029015e-05, + "loss": 0.0185, + "step": 1409 + }, + { + "epoch": 0.1528620988725065, + "grad_norm": 1.8966223001480103, + "learning_rate": 1.9049760626722763e-05, + "loss": 0.033, + "step": 1410 + }, + { + "epoch": 0.1529705117085863, + "grad_norm": 3.4252076148986816, + "learning_rate": 1.9049035253155375e-05, + "loss": 0.0586, + "step": 1411 + }, + { + "epoch": 0.1530789245446661, + "grad_norm": 5.586063861846924, + "learning_rate": 1.904830987958799e-05, + "loss": 0.085, + "step": 1412 + }, + { + "epoch": 0.15318733738074589, + "grad_norm": 2.6108486652374268, + "learning_rate": 1.9047584506020603e-05, + "loss": 0.0637, + "step": 1413 + }, + { + "epoch": 0.15329575021682568, + "grad_norm": 7.295788764953613, + "learning_rate": 1.9046859132453215e-05, + "loss": 0.0438, + "step": 1414 + }, + { + "epoch": 0.15340416305290547, + "grad_norm": 2.5528883934020996, + "learning_rate": 1.9046133758885827e-05, + "loss": 0.0232, + "step": 1415 + }, + { + "epoch": 0.15351257588898526, + "grad_norm": 5.93359899520874, + "learning_rate": 1.904540838531844e-05, + "loss": 0.0317, + "step": 1416 + }, + { + "epoch": 0.15362098872506505, + "grad_norm": 4.751140594482422, + "learning_rate": 1.904468301175105e-05, + "loss": 0.0884, + "step": 1417 + }, + { + "epoch": 0.15372940156114484, + "grad_norm": 1.532470703125, + "learning_rate": 1.9043957638183664e-05, + "loss": 0.0251, + "step": 1418 + }, + { + "epoch": 0.15383781439722463, + "grad_norm": 1.6168566942214966, + "learning_rate": 1.904323226461628e-05, + "loss": 0.0505, + "step": 1419 + }, + { + "epoch": 0.15394622723330442, + "grad_norm": 4.626262664794922, + "learning_rate": 1.904250689104889e-05, + "loss": 0.1021, + "step": 1420 + }, + { + "epoch": 0.1540546400693842, + "grad_norm": 2.5690221786499023, + "learning_rate": 1.9041781517481504e-05, + "loss": 0.0216, + "step": 1421 + }, + { + "epoch": 0.154163052905464, + "grad_norm": 2.542947292327881, + "learning_rate": 1.904105614391412e-05, + "loss": 0.0255, + "step": 1422 + }, + { + "epoch": 0.1542714657415438, + "grad_norm": 14.359969139099121, + "learning_rate": 1.904033077034673e-05, + "loss": 0.0619, + "step": 1423 + }, + { + "epoch": 0.15437987857762359, + "grad_norm": 5.369576454162598, + "learning_rate": 1.9039605396779343e-05, + "loss": 0.0873, + "step": 1424 + }, + { + "epoch": 0.15448829141370338, + "grad_norm": 4.918118476867676, + "learning_rate": 1.9038880023211956e-05, + "loss": 0.0665, + "step": 1425 + }, + { + "epoch": 0.15459670424978317, + "grad_norm": 4.1678690910339355, + "learning_rate": 1.9038154649644568e-05, + "loss": 0.0371, + "step": 1426 + }, + { + "epoch": 0.15470511708586296, + "grad_norm": 2.7524492740631104, + "learning_rate": 1.9037429276077183e-05, + "loss": 0.0371, + "step": 1427 + }, + { + "epoch": 0.15481352992194275, + "grad_norm": 5.768860340118408, + "learning_rate": 1.9036703902509795e-05, + "loss": 0.0528, + "step": 1428 + }, + { + "epoch": 0.15492194275802254, + "grad_norm": 6.024853229522705, + "learning_rate": 1.9035978528942408e-05, + "loss": 0.0625, + "step": 1429 + }, + { + "epoch": 0.15503035559410233, + "grad_norm": 4.533526420593262, + "learning_rate": 1.903525315537502e-05, + "loss": 0.0936, + "step": 1430 + }, + { + "epoch": 0.15513876843018212, + "grad_norm": 5.0021891593933105, + "learning_rate": 1.9034527781807632e-05, + "loss": 0.0488, + "step": 1431 + }, + { + "epoch": 0.1552471812662619, + "grad_norm": 2.8361029624938965, + "learning_rate": 1.9033802408240244e-05, + "loss": 0.0285, + "step": 1432 + }, + { + "epoch": 0.1553555941023417, + "grad_norm": 2.3860177993774414, + "learning_rate": 1.9033077034672856e-05, + "loss": 0.079, + "step": 1433 + }, + { + "epoch": 0.15546400693842152, + "grad_norm": 3.968482494354248, + "learning_rate": 1.903235166110547e-05, + "loss": 0.0684, + "step": 1434 + }, + { + "epoch": 0.1555724197745013, + "grad_norm": 2.5141711235046387, + "learning_rate": 1.9031626287538084e-05, + "loss": 0.0323, + "step": 1435 + }, + { + "epoch": 0.1556808326105811, + "grad_norm": 3.928802013397217, + "learning_rate": 1.9030900913970696e-05, + "loss": 0.0608, + "step": 1436 + }, + { + "epoch": 0.1557892454466609, + "grad_norm": 4.70235013961792, + "learning_rate": 1.9030175540403308e-05, + "loss": 0.0674, + "step": 1437 + }, + { + "epoch": 0.1558976582827407, + "grad_norm": 3.8693861961364746, + "learning_rate": 1.902945016683592e-05, + "loss": 0.0515, + "step": 1438 + }, + { + "epoch": 0.15600607111882048, + "grad_norm": 10.97087574005127, + "learning_rate": 1.9028724793268536e-05, + "loss": 0.068, + "step": 1439 + }, + { + "epoch": 0.15611448395490027, + "grad_norm": 2.815279483795166, + "learning_rate": 1.9027999419701148e-05, + "loss": 0.0337, + "step": 1440 + }, + { + "epoch": 0.15622289679098006, + "grad_norm": 3.1995246410369873, + "learning_rate": 1.902727404613376e-05, + "loss": 0.0264, + "step": 1441 + }, + { + "epoch": 0.15633130962705985, + "grad_norm": 2.766683340072632, + "learning_rate": 1.9026548672566376e-05, + "loss": 0.0561, + "step": 1442 + }, + { + "epoch": 0.15643972246313964, + "grad_norm": 2.139742612838745, + "learning_rate": 1.9025823298998988e-05, + "loss": 0.0437, + "step": 1443 + }, + { + "epoch": 0.15654813529921943, + "grad_norm": 3.9948182106018066, + "learning_rate": 1.90250979254316e-05, + "loss": 0.0737, + "step": 1444 + }, + { + "epoch": 0.15665654813529922, + "grad_norm": 3.4528250694274902, + "learning_rate": 1.9024372551864212e-05, + "loss": 0.0382, + "step": 1445 + }, + { + "epoch": 0.156764960971379, + "grad_norm": 3.7926218509674072, + "learning_rate": 1.9023647178296824e-05, + "loss": 0.0332, + "step": 1446 + }, + { + "epoch": 0.1568733738074588, + "grad_norm": 5.890145778656006, + "learning_rate": 1.9022921804729436e-05, + "loss": 0.0635, + "step": 1447 + }, + { + "epoch": 0.1569817866435386, + "grad_norm": 6.709765434265137, + "learning_rate": 1.902219643116205e-05, + "loss": 0.0567, + "step": 1448 + }, + { + "epoch": 0.1570901994796184, + "grad_norm": 2.7219574451446533, + "learning_rate": 1.9021471057594664e-05, + "loss": 0.022, + "step": 1449 + }, + { + "epoch": 0.15719861231569818, + "grad_norm": 4.49485445022583, + "learning_rate": 1.9020745684027276e-05, + "loss": 0.0643, + "step": 1450 + }, + { + "epoch": 0.15730702515177797, + "grad_norm": 3.7122771739959717, + "learning_rate": 1.9020020310459888e-05, + "loss": 0.0369, + "step": 1451 + }, + { + "epoch": 0.15741543798785776, + "grad_norm": 6.594842910766602, + "learning_rate": 1.90192949368925e-05, + "loss": 0.0533, + "step": 1452 + }, + { + "epoch": 0.15752385082393755, + "grad_norm": 3.408493757247925, + "learning_rate": 1.9018569563325113e-05, + "loss": 0.0355, + "step": 1453 + }, + { + "epoch": 0.15763226366001734, + "grad_norm": 2.797823190689087, + "learning_rate": 1.9017844189757725e-05, + "loss": 0.0547, + "step": 1454 + }, + { + "epoch": 0.15774067649609713, + "grad_norm": 3.413397789001465, + "learning_rate": 1.901711881619034e-05, + "loss": 0.0323, + "step": 1455 + }, + { + "epoch": 0.15784908933217692, + "grad_norm": 2.7756707668304443, + "learning_rate": 1.9016393442622952e-05, + "loss": 0.0448, + "step": 1456 + }, + { + "epoch": 0.1579575021682567, + "grad_norm": 1.5777277946472168, + "learning_rate": 1.9015668069055565e-05, + "loss": 0.0094, + "step": 1457 + }, + { + "epoch": 0.1580659150043365, + "grad_norm": 2.0687036514282227, + "learning_rate": 1.901494269548818e-05, + "loss": 0.0197, + "step": 1458 + }, + { + "epoch": 0.1581743278404163, + "grad_norm": 5.375793933868408, + "learning_rate": 1.9014217321920792e-05, + "loss": 0.0555, + "step": 1459 + }, + { + "epoch": 0.1582827406764961, + "grad_norm": 7.84032678604126, + "learning_rate": 1.9013491948353404e-05, + "loss": 0.0799, + "step": 1460 + }, + { + "epoch": 0.15839115351257588, + "grad_norm": 4.240949630737305, + "learning_rate": 1.9012766574786016e-05, + "loss": 0.0845, + "step": 1461 + }, + { + "epoch": 0.15849956634865567, + "grad_norm": 5.634677410125732, + "learning_rate": 1.901204120121863e-05, + "loss": 0.1043, + "step": 1462 + }, + { + "epoch": 0.15860797918473546, + "grad_norm": 3.201732635498047, + "learning_rate": 1.901131582765124e-05, + "loss": 0.0263, + "step": 1463 + }, + { + "epoch": 0.15871639202081528, + "grad_norm": 5.208518028259277, + "learning_rate": 1.9010590454083856e-05, + "loss": 0.0435, + "step": 1464 + }, + { + "epoch": 0.15882480485689507, + "grad_norm": 4.092844486236572, + "learning_rate": 1.900986508051647e-05, + "loss": 0.0505, + "step": 1465 + }, + { + "epoch": 0.15893321769297486, + "grad_norm": 6.988559722900391, + "learning_rate": 1.900913970694908e-05, + "loss": 0.0576, + "step": 1466 + }, + { + "epoch": 0.15904163052905465, + "grad_norm": 3.8218111991882324, + "learning_rate": 1.9008414333381693e-05, + "loss": 0.0512, + "step": 1467 + }, + { + "epoch": 0.15915004336513444, + "grad_norm": 8.209327697753906, + "learning_rate": 1.9007688959814305e-05, + "loss": 0.0743, + "step": 1468 + }, + { + "epoch": 0.15925845620121423, + "grad_norm": 2.2789840698242188, + "learning_rate": 1.9006963586246917e-05, + "loss": 0.0616, + "step": 1469 + }, + { + "epoch": 0.15936686903729402, + "grad_norm": 4.1193695068359375, + "learning_rate": 1.900623821267953e-05, + "loss": 0.0342, + "step": 1470 + }, + { + "epoch": 0.15947528187337381, + "grad_norm": 3.912226438522339, + "learning_rate": 1.9005512839112145e-05, + "loss": 0.0652, + "step": 1471 + }, + { + "epoch": 0.1595836947094536, + "grad_norm": 4.0427165031433105, + "learning_rate": 1.9004787465544757e-05, + "loss": 0.0845, + "step": 1472 + }, + { + "epoch": 0.1596921075455334, + "grad_norm": 2.541106700897217, + "learning_rate": 1.900406209197737e-05, + "loss": 0.0254, + "step": 1473 + }, + { + "epoch": 0.1598005203816132, + "grad_norm": 1.8129290342330933, + "learning_rate": 1.9003336718409985e-05, + "loss": 0.033, + "step": 1474 + }, + { + "epoch": 0.15990893321769298, + "grad_norm": 2.202108860015869, + "learning_rate": 1.9002611344842597e-05, + "loss": 0.036, + "step": 1475 + }, + { + "epoch": 0.16001734605377277, + "grad_norm": 5.670005798339844, + "learning_rate": 1.900188597127521e-05, + "loss": 0.0577, + "step": 1476 + }, + { + "epoch": 0.16012575888985256, + "grad_norm": 4.660211086273193, + "learning_rate": 1.900116059770782e-05, + "loss": 0.0637, + "step": 1477 + }, + { + "epoch": 0.16023417172593235, + "grad_norm": 2.8273959159851074, + "learning_rate": 1.9000435224140433e-05, + "loss": 0.0356, + "step": 1478 + }, + { + "epoch": 0.16034258456201214, + "grad_norm": 4.46704626083374, + "learning_rate": 1.899970985057305e-05, + "loss": 0.0429, + "step": 1479 + }, + { + "epoch": 0.16045099739809193, + "grad_norm": 1.7266044616699219, + "learning_rate": 1.899898447700566e-05, + "loss": 0.0302, + "step": 1480 + }, + { + "epoch": 0.16055941023417172, + "grad_norm": 4.764721393585205, + "learning_rate": 1.8998259103438273e-05, + "loss": 0.0692, + "step": 1481 + }, + { + "epoch": 0.16066782307025151, + "grad_norm": 5.9035820960998535, + "learning_rate": 1.8997533729870885e-05, + "loss": 0.0296, + "step": 1482 + }, + { + "epoch": 0.1607762359063313, + "grad_norm": 2.984504461288452, + "learning_rate": 1.8996808356303497e-05, + "loss": 0.0138, + "step": 1483 + }, + { + "epoch": 0.1608846487424111, + "grad_norm": 3.061575174331665, + "learning_rate": 1.899608298273611e-05, + "loss": 0.0559, + "step": 1484 + }, + { + "epoch": 0.1609930615784909, + "grad_norm": 3.9240634441375732, + "learning_rate": 1.899535760916872e-05, + "loss": 0.0943, + "step": 1485 + }, + { + "epoch": 0.16110147441457068, + "grad_norm": 14.190691947937012, + "learning_rate": 1.8994632235601337e-05, + "loss": 0.0572, + "step": 1486 + }, + { + "epoch": 0.16120988725065047, + "grad_norm": 2.4709761142730713, + "learning_rate": 1.899390686203395e-05, + "loss": 0.0696, + "step": 1487 + }, + { + "epoch": 0.16131830008673026, + "grad_norm": 2.2248220443725586, + "learning_rate": 1.899318148846656e-05, + "loss": 0.0284, + "step": 1488 + }, + { + "epoch": 0.16142671292281005, + "grad_norm": 6.889444828033447, + "learning_rate": 1.8992456114899173e-05, + "loss": 0.0891, + "step": 1489 + }, + { + "epoch": 0.16153512575888984, + "grad_norm": 3.3065218925476074, + "learning_rate": 1.8991730741331786e-05, + "loss": 0.0663, + "step": 1490 + }, + { + "epoch": 0.16164353859496963, + "grad_norm": 4.450634956359863, + "learning_rate": 1.89910053677644e-05, + "loss": 0.0582, + "step": 1491 + }, + { + "epoch": 0.16175195143104942, + "grad_norm": 2.452845335006714, + "learning_rate": 1.8990279994197013e-05, + "loss": 0.0621, + "step": 1492 + }, + { + "epoch": 0.16186036426712924, + "grad_norm": 11.593733787536621, + "learning_rate": 1.8989554620629625e-05, + "loss": 0.0467, + "step": 1493 + }, + { + "epoch": 0.16196877710320903, + "grad_norm": 3.786621332168579, + "learning_rate": 1.898882924706224e-05, + "loss": 0.0536, + "step": 1494 + }, + { + "epoch": 0.16207718993928882, + "grad_norm": 2.861039400100708, + "learning_rate": 1.8988103873494853e-05, + "loss": 0.0667, + "step": 1495 + }, + { + "epoch": 0.16218560277536861, + "grad_norm": 1.2861016988754272, + "learning_rate": 1.8987378499927465e-05, + "loss": 0.0225, + "step": 1496 + }, + { + "epoch": 0.1622940156114484, + "grad_norm": 3.7424843311309814, + "learning_rate": 1.8986653126360077e-05, + "loss": 0.0653, + "step": 1497 + }, + { + "epoch": 0.1624024284475282, + "grad_norm": 3.357684373855591, + "learning_rate": 1.898592775279269e-05, + "loss": 0.0749, + "step": 1498 + }, + { + "epoch": 0.162510841283608, + "grad_norm": 3.54040789604187, + "learning_rate": 1.89852023792253e-05, + "loss": 0.0441, + "step": 1499 + }, + { + "epoch": 0.16261925411968778, + "grad_norm": 3.804035186767578, + "learning_rate": 1.8984477005657914e-05, + "loss": 0.0803, + "step": 1500 + }, + { + "epoch": 0.16272766695576757, + "grad_norm": 3.6055357456207275, + "learning_rate": 1.898375163209053e-05, + "loss": 0.0416, + "step": 1501 + }, + { + "epoch": 0.16283607979184736, + "grad_norm": 1.571051001548767, + "learning_rate": 1.898302625852314e-05, + "loss": 0.0269, + "step": 1502 + }, + { + "epoch": 0.16294449262792715, + "grad_norm": 3.3273074626922607, + "learning_rate": 1.8982300884955754e-05, + "loss": 0.0572, + "step": 1503 + }, + { + "epoch": 0.16305290546400694, + "grad_norm": 2.691787004470825, + "learning_rate": 1.8981575511388366e-05, + "loss": 0.0407, + "step": 1504 + }, + { + "epoch": 0.16316131830008673, + "grad_norm": 3.878589630126953, + "learning_rate": 1.8980850137820978e-05, + "loss": 0.1055, + "step": 1505 + }, + { + "epoch": 0.16326973113616652, + "grad_norm": 3.112067699432373, + "learning_rate": 1.898012476425359e-05, + "loss": 0.0314, + "step": 1506 + }, + { + "epoch": 0.16337814397224631, + "grad_norm": 4.4850850105285645, + "learning_rate": 1.8979399390686206e-05, + "loss": 0.0838, + "step": 1507 + }, + { + "epoch": 0.1634865568083261, + "grad_norm": 6.463613033294678, + "learning_rate": 1.8978674017118818e-05, + "loss": 0.0298, + "step": 1508 + }, + { + "epoch": 0.1635949696444059, + "grad_norm": 3.9606616497039795, + "learning_rate": 1.897794864355143e-05, + "loss": 0.0584, + "step": 1509 + }, + { + "epoch": 0.1637033824804857, + "grad_norm": 2.0103092193603516, + "learning_rate": 1.8977223269984045e-05, + "loss": 0.037, + "step": 1510 + }, + { + "epoch": 0.16381179531656548, + "grad_norm": 1.9894403219223022, + "learning_rate": 1.8976497896416658e-05, + "loss": 0.0419, + "step": 1511 + }, + { + "epoch": 0.16392020815264527, + "grad_norm": 2.161423444747925, + "learning_rate": 1.897577252284927e-05, + "loss": 0.0367, + "step": 1512 + }, + { + "epoch": 0.16402862098872506, + "grad_norm": 4.373612403869629, + "learning_rate": 1.8975047149281882e-05, + "loss": 0.0445, + "step": 1513 + }, + { + "epoch": 0.16413703382480485, + "grad_norm": 4.163061141967773, + "learning_rate": 1.8974321775714494e-05, + "loss": 0.0457, + "step": 1514 + }, + { + "epoch": 0.16424544666088464, + "grad_norm": 2.576779842376709, + "learning_rate": 1.8973596402147106e-05, + "loss": 0.0656, + "step": 1515 + }, + { + "epoch": 0.16435385949696443, + "grad_norm": 4.855607032775879, + "learning_rate": 1.897287102857972e-05, + "loss": 0.0835, + "step": 1516 + }, + { + "epoch": 0.16446227233304422, + "grad_norm": 4.651301383972168, + "learning_rate": 1.8972145655012334e-05, + "loss": 0.0895, + "step": 1517 + }, + { + "epoch": 0.16457068516912401, + "grad_norm": 3.4440813064575195, + "learning_rate": 1.8971420281444946e-05, + "loss": 0.0391, + "step": 1518 + }, + { + "epoch": 0.1646790980052038, + "grad_norm": 0.8277063369750977, + "learning_rate": 1.8970694907877558e-05, + "loss": 0.0092, + "step": 1519 + }, + { + "epoch": 0.1647875108412836, + "grad_norm": 2.3471028804779053, + "learning_rate": 1.896996953431017e-05, + "loss": 0.0543, + "step": 1520 + }, + { + "epoch": 0.1648959236773634, + "grad_norm": 3.502725124359131, + "learning_rate": 1.8969244160742782e-05, + "loss": 0.0745, + "step": 1521 + }, + { + "epoch": 0.16500433651344318, + "grad_norm": 5.2802605628967285, + "learning_rate": 1.8968518787175394e-05, + "loss": 0.0741, + "step": 1522 + }, + { + "epoch": 0.165112749349523, + "grad_norm": 4.682559013366699, + "learning_rate": 1.896779341360801e-05, + "loss": 0.0458, + "step": 1523 + }, + { + "epoch": 0.1652211621856028, + "grad_norm": 4.8749542236328125, + "learning_rate": 1.8967068040040622e-05, + "loss": 0.0548, + "step": 1524 + }, + { + "epoch": 0.16532957502168258, + "grad_norm": 3.2133419513702393, + "learning_rate": 1.8966342666473234e-05, + "loss": 0.026, + "step": 1525 + }, + { + "epoch": 0.16543798785776237, + "grad_norm": 3.0849015712738037, + "learning_rate": 1.896561729290585e-05, + "loss": 0.0262, + "step": 1526 + }, + { + "epoch": 0.16554640069384216, + "grad_norm": 2.9927237033843994, + "learning_rate": 1.8964891919338462e-05, + "loss": 0.0389, + "step": 1527 + }, + { + "epoch": 0.16565481352992195, + "grad_norm": 2.0693767070770264, + "learning_rate": 1.8964166545771074e-05, + "loss": 0.0419, + "step": 1528 + }, + { + "epoch": 0.16576322636600174, + "grad_norm": 2.448169469833374, + "learning_rate": 1.8963441172203686e-05, + "loss": 0.0293, + "step": 1529 + }, + { + "epoch": 0.16587163920208153, + "grad_norm": 4.364576816558838, + "learning_rate": 1.89627157986363e-05, + "loss": 0.0717, + "step": 1530 + }, + { + "epoch": 0.16598005203816132, + "grad_norm": 4.345706462860107, + "learning_rate": 1.8961990425068914e-05, + "loss": 0.0289, + "step": 1531 + }, + { + "epoch": 0.16608846487424112, + "grad_norm": 2.0837996006011963, + "learning_rate": 1.8961265051501526e-05, + "loss": 0.0209, + "step": 1532 + }, + { + "epoch": 0.1661968777103209, + "grad_norm": 6.011240482330322, + "learning_rate": 1.8960539677934138e-05, + "loss": 0.0684, + "step": 1533 + }, + { + "epoch": 0.1663052905464007, + "grad_norm": 2.726104974746704, + "learning_rate": 1.895981430436675e-05, + "loss": 0.0314, + "step": 1534 + }, + { + "epoch": 0.1664137033824805, + "grad_norm": 5.517116069793701, + "learning_rate": 1.8959088930799363e-05, + "loss": 0.0833, + "step": 1535 + }, + { + "epoch": 0.16652211621856028, + "grad_norm": 2.411252498626709, + "learning_rate": 1.8958363557231975e-05, + "loss": 0.0386, + "step": 1536 + }, + { + "epoch": 0.16663052905464007, + "grad_norm": 3.573645830154419, + "learning_rate": 1.8957638183664587e-05, + "loss": 0.0571, + "step": 1537 + }, + { + "epoch": 0.16673894189071986, + "grad_norm": 0.9550082087516785, + "learning_rate": 1.8956912810097202e-05, + "loss": 0.0153, + "step": 1538 + }, + { + "epoch": 0.16684735472679965, + "grad_norm": 2.346034526824951, + "learning_rate": 1.8956187436529814e-05, + "loss": 0.0208, + "step": 1539 + }, + { + "epoch": 0.16695576756287944, + "grad_norm": 3.800478458404541, + "learning_rate": 1.8955462062962427e-05, + "loss": 0.0289, + "step": 1540 + }, + { + "epoch": 0.16706418039895923, + "grad_norm": 7.366245269775391, + "learning_rate": 1.895473668939504e-05, + "loss": 0.0951, + "step": 1541 + }, + { + "epoch": 0.16717259323503902, + "grad_norm": 3.553497552871704, + "learning_rate": 1.895401131582765e-05, + "loss": 0.0215, + "step": 1542 + }, + { + "epoch": 0.16728100607111882, + "grad_norm": 2.3422164916992188, + "learning_rate": 1.8953285942260266e-05, + "loss": 0.0504, + "step": 1543 + }, + { + "epoch": 0.1673894189071986, + "grad_norm": 4.016317844390869, + "learning_rate": 1.895256056869288e-05, + "loss": 0.0364, + "step": 1544 + }, + { + "epoch": 0.1674978317432784, + "grad_norm": 3.5720202922821045, + "learning_rate": 1.895183519512549e-05, + "loss": 0.0423, + "step": 1545 + }, + { + "epoch": 0.1676062445793582, + "grad_norm": 4.768626689910889, + "learning_rate": 1.8951109821558106e-05, + "loss": 0.0863, + "step": 1546 + }, + { + "epoch": 0.16771465741543798, + "grad_norm": 5.077347278594971, + "learning_rate": 1.895038444799072e-05, + "loss": 0.0869, + "step": 1547 + }, + { + "epoch": 0.16782307025151777, + "grad_norm": 8.78687572479248, + "learning_rate": 1.894965907442333e-05, + "loss": 0.0902, + "step": 1548 + }, + { + "epoch": 0.16793148308759756, + "grad_norm": 8.628527641296387, + "learning_rate": 1.8948933700855943e-05, + "loss": 0.0663, + "step": 1549 + }, + { + "epoch": 0.16803989592367735, + "grad_norm": 6.884238243103027, + "learning_rate": 1.8948208327288555e-05, + "loss": 0.0333, + "step": 1550 + }, + { + "epoch": 0.16814830875975714, + "grad_norm": 5.254474639892578, + "learning_rate": 1.8947482953721167e-05, + "loss": 0.022, + "step": 1551 + }, + { + "epoch": 0.16825672159583693, + "grad_norm": 0.8603410124778748, + "learning_rate": 1.894675758015378e-05, + "loss": 0.009, + "step": 1552 + }, + { + "epoch": 0.16836513443191675, + "grad_norm": 4.83103084564209, + "learning_rate": 1.8946032206586395e-05, + "loss": 0.0668, + "step": 1553 + }, + { + "epoch": 0.16847354726799654, + "grad_norm": 4.190361976623535, + "learning_rate": 1.8945306833019007e-05, + "loss": 0.0774, + "step": 1554 + }, + { + "epoch": 0.16858196010407633, + "grad_norm": 7.190981388092041, + "learning_rate": 1.894458145945162e-05, + "loss": 0.0703, + "step": 1555 + }, + { + "epoch": 0.16869037294015612, + "grad_norm": 4.967663764953613, + "learning_rate": 1.894385608588423e-05, + "loss": 0.0998, + "step": 1556 + }, + { + "epoch": 0.16879878577623592, + "grad_norm": 2.0865249633789062, + "learning_rate": 1.8943130712316843e-05, + "loss": 0.0416, + "step": 1557 + }, + { + "epoch": 0.1689071986123157, + "grad_norm": 4.4371113777160645, + "learning_rate": 1.8942405338749455e-05, + "loss": 0.0463, + "step": 1558 + }, + { + "epoch": 0.1690156114483955, + "grad_norm": 4.807804584503174, + "learning_rate": 1.894167996518207e-05, + "loss": 0.0866, + "step": 1559 + }, + { + "epoch": 0.1691240242844753, + "grad_norm": 3.212339162826538, + "learning_rate": 1.8940954591614683e-05, + "loss": 0.04, + "step": 1560 + }, + { + "epoch": 0.16923243712055508, + "grad_norm": 4.979723930358887, + "learning_rate": 1.8940229218047295e-05, + "loss": 0.0736, + "step": 1561 + }, + { + "epoch": 0.16934084995663487, + "grad_norm": 4.364170551300049, + "learning_rate": 1.893950384447991e-05, + "loss": 0.0679, + "step": 1562 + }, + { + "epoch": 0.16944926279271466, + "grad_norm": 3.074873924255371, + "learning_rate": 1.8938778470912523e-05, + "loss": 0.0704, + "step": 1563 + }, + { + "epoch": 0.16955767562879445, + "grad_norm": 4.4319610595703125, + "learning_rate": 1.8938053097345135e-05, + "loss": 0.0662, + "step": 1564 + }, + { + "epoch": 0.16966608846487424, + "grad_norm": 2.8051998615264893, + "learning_rate": 1.8937327723777747e-05, + "loss": 0.0389, + "step": 1565 + }, + { + "epoch": 0.16977450130095403, + "grad_norm": 2.069972276687622, + "learning_rate": 1.893660235021036e-05, + "loss": 0.0548, + "step": 1566 + }, + { + "epoch": 0.16988291413703382, + "grad_norm": 5.181976795196533, + "learning_rate": 1.893587697664297e-05, + "loss": 0.0852, + "step": 1567 + }, + { + "epoch": 0.16999132697311362, + "grad_norm": 4.19508695602417, + "learning_rate": 1.8935151603075587e-05, + "loss": 0.053, + "step": 1568 + }, + { + "epoch": 0.1700997398091934, + "grad_norm": 3.12092661857605, + "learning_rate": 1.89344262295082e-05, + "loss": 0.0706, + "step": 1569 + }, + { + "epoch": 0.1702081526452732, + "grad_norm": 3.174760103225708, + "learning_rate": 1.893370085594081e-05, + "loss": 0.0385, + "step": 1570 + }, + { + "epoch": 0.170316565481353, + "grad_norm": 1.8067735433578491, + "learning_rate": 1.8932975482373423e-05, + "loss": 0.0313, + "step": 1571 + }, + { + "epoch": 0.17042497831743278, + "grad_norm": 2.914015054702759, + "learning_rate": 1.8932250108806036e-05, + "loss": 0.0756, + "step": 1572 + }, + { + "epoch": 0.17053339115351257, + "grad_norm": 2.2515478134155273, + "learning_rate": 1.8931524735238648e-05, + "loss": 0.0615, + "step": 1573 + }, + { + "epoch": 0.17064180398959236, + "grad_norm": 2.6249241828918457, + "learning_rate": 1.8930799361671263e-05, + "loss": 0.0509, + "step": 1574 + }, + { + "epoch": 0.17075021682567215, + "grad_norm": 2.0797367095947266, + "learning_rate": 1.8930073988103875e-05, + "loss": 0.0448, + "step": 1575 + }, + { + "epoch": 0.17085862966175194, + "grad_norm": 1.204756498336792, + "learning_rate": 1.8929348614536488e-05, + "loss": 0.0227, + "step": 1576 + }, + { + "epoch": 0.17096704249783173, + "grad_norm": 3.047987937927246, + "learning_rate": 1.89286232409691e-05, + "loss": 0.0858, + "step": 1577 + }, + { + "epoch": 0.17107545533391152, + "grad_norm": 1.9909757375717163, + "learning_rate": 1.8927897867401715e-05, + "loss": 0.0481, + "step": 1578 + }, + { + "epoch": 0.17118386816999132, + "grad_norm": 4.044754981994629, + "learning_rate": 1.8927172493834327e-05, + "loss": 0.0858, + "step": 1579 + }, + { + "epoch": 0.1712922810060711, + "grad_norm": 2.195007562637329, + "learning_rate": 1.892644712026694e-05, + "loss": 0.0401, + "step": 1580 + }, + { + "epoch": 0.1714006938421509, + "grad_norm": 2.298278331756592, + "learning_rate": 1.892572174669955e-05, + "loss": 0.0444, + "step": 1581 + }, + { + "epoch": 0.17150910667823072, + "grad_norm": 1.4183281660079956, + "learning_rate": 1.8924996373132164e-05, + "loss": 0.0422, + "step": 1582 + }, + { + "epoch": 0.1716175195143105, + "grad_norm": 1.681492567062378, + "learning_rate": 1.892427099956478e-05, + "loss": 0.0385, + "step": 1583 + }, + { + "epoch": 0.1717259323503903, + "grad_norm": 1.47900390625, + "learning_rate": 1.892354562599739e-05, + "loss": 0.0412, + "step": 1584 + }, + { + "epoch": 0.1718343451864701, + "grad_norm": 1.302433729171753, + "learning_rate": 1.8922820252430004e-05, + "loss": 0.0305, + "step": 1585 + }, + { + "epoch": 0.17194275802254988, + "grad_norm": 2.248971700668335, + "learning_rate": 1.8922094878862616e-05, + "loss": 0.0265, + "step": 1586 + }, + { + "epoch": 0.17205117085862967, + "grad_norm": 3.738004207611084, + "learning_rate": 1.8921369505295228e-05, + "loss": 0.0352, + "step": 1587 + }, + { + "epoch": 0.17215958369470946, + "grad_norm": 1.6536388397216797, + "learning_rate": 1.892064413172784e-05, + "loss": 0.0273, + "step": 1588 + }, + { + "epoch": 0.17226799653078925, + "grad_norm": 3.62646746635437, + "learning_rate": 1.8919918758160452e-05, + "loss": 0.0321, + "step": 1589 + }, + { + "epoch": 0.17237640936686904, + "grad_norm": 7.016899108886719, + "learning_rate": 1.8919193384593068e-05, + "loss": 0.0678, + "step": 1590 + }, + { + "epoch": 0.17248482220294883, + "grad_norm": 5.682413578033447, + "learning_rate": 1.891846801102568e-05, + "loss": 0.0943, + "step": 1591 + }, + { + "epoch": 0.17259323503902863, + "grad_norm": 2.1257219314575195, + "learning_rate": 1.8917742637458292e-05, + "loss": 0.0467, + "step": 1592 + }, + { + "epoch": 0.17270164787510842, + "grad_norm": 2.3222742080688477, + "learning_rate": 1.8917017263890904e-05, + "loss": 0.047, + "step": 1593 + }, + { + "epoch": 0.1728100607111882, + "grad_norm": 4.139674663543701, + "learning_rate": 1.8916291890323516e-05, + "loss": 0.0434, + "step": 1594 + }, + { + "epoch": 0.172918473547268, + "grad_norm": 4.4856038093566895, + "learning_rate": 1.8915566516756132e-05, + "loss": 0.0433, + "step": 1595 + }, + { + "epoch": 0.1730268863833478, + "grad_norm": 1.9930256605148315, + "learning_rate": 1.8914841143188744e-05, + "loss": 0.0471, + "step": 1596 + }, + { + "epoch": 0.17313529921942758, + "grad_norm": 2.2420432567596436, + "learning_rate": 1.8914115769621356e-05, + "loss": 0.0749, + "step": 1597 + }, + { + "epoch": 0.17324371205550737, + "grad_norm": 7.250087738037109, + "learning_rate": 1.891339039605397e-05, + "loss": 0.0266, + "step": 1598 + }, + { + "epoch": 0.17335212489158716, + "grad_norm": 9.011466979980469, + "learning_rate": 1.8912665022486584e-05, + "loss": 0.0367, + "step": 1599 + }, + { + "epoch": 0.17346053772766695, + "grad_norm": 6.610247611999512, + "learning_rate": 1.8911939648919196e-05, + "loss": 0.0686, + "step": 1600 + }, + { + "epoch": 0.17356895056374674, + "grad_norm": 2.6023764610290527, + "learning_rate": 1.8911214275351808e-05, + "loss": 0.0519, + "step": 1601 + }, + { + "epoch": 0.17367736339982653, + "grad_norm": 5.568012714385986, + "learning_rate": 1.891048890178442e-05, + "loss": 0.0562, + "step": 1602 + }, + { + "epoch": 0.17378577623590633, + "grad_norm": 2.2880818843841553, + "learning_rate": 1.8909763528217032e-05, + "loss": 0.0335, + "step": 1603 + }, + { + "epoch": 0.17389418907198612, + "grad_norm": 3.5323147773742676, + "learning_rate": 1.8909038154649644e-05, + "loss": 0.0115, + "step": 1604 + }, + { + "epoch": 0.1740026019080659, + "grad_norm": 2.883795976638794, + "learning_rate": 1.890831278108226e-05, + "loss": 0.0262, + "step": 1605 + }, + { + "epoch": 0.1741110147441457, + "grad_norm": 4.790670871734619, + "learning_rate": 1.8907587407514872e-05, + "loss": 0.0857, + "step": 1606 + }, + { + "epoch": 0.1742194275802255, + "grad_norm": 3.5948009490966797, + "learning_rate": 1.8906862033947484e-05, + "loss": 0.0304, + "step": 1607 + }, + { + "epoch": 0.17432784041630528, + "grad_norm": 2.237062692642212, + "learning_rate": 1.8906136660380096e-05, + "loss": 0.0613, + "step": 1608 + }, + { + "epoch": 0.17443625325238507, + "grad_norm": 4.420834064483643, + "learning_rate": 1.890541128681271e-05, + "loss": 0.0566, + "step": 1609 + }, + { + "epoch": 0.17454466608846486, + "grad_norm": 3.896394968032837, + "learning_rate": 1.890468591324532e-05, + "loss": 0.0458, + "step": 1610 + }, + { + "epoch": 0.17465307892454465, + "grad_norm": 0.6443367600440979, + "learning_rate": 1.8903960539677936e-05, + "loss": 0.0117, + "step": 1611 + }, + { + "epoch": 0.17476149176062447, + "grad_norm": 1.1357245445251465, + "learning_rate": 1.890323516611055e-05, + "loss": 0.0146, + "step": 1612 + }, + { + "epoch": 0.17486990459670426, + "grad_norm": 1.027443766593933, + "learning_rate": 1.890250979254316e-05, + "loss": 0.028, + "step": 1613 + }, + { + "epoch": 0.17497831743278405, + "grad_norm": 4.700620174407959, + "learning_rate": 1.8901784418975776e-05, + "loss": 0.0375, + "step": 1614 + }, + { + "epoch": 0.17508673026886384, + "grad_norm": 6.264010906219482, + "learning_rate": 1.8901059045408388e-05, + "loss": 0.0938, + "step": 1615 + }, + { + "epoch": 0.17519514310494363, + "grad_norm": 3.3641364574432373, + "learning_rate": 1.8900333671841e-05, + "loss": 0.0473, + "step": 1616 + }, + { + "epoch": 0.17530355594102343, + "grad_norm": 3.6987128257751465, + "learning_rate": 1.8899608298273612e-05, + "loss": 0.0298, + "step": 1617 + }, + { + "epoch": 0.17541196877710322, + "grad_norm": 4.161534786224365, + "learning_rate": 1.8898882924706225e-05, + "loss": 0.1013, + "step": 1618 + }, + { + "epoch": 0.175520381613183, + "grad_norm": 4.2809977531433105, + "learning_rate": 1.8898157551138837e-05, + "loss": 0.0455, + "step": 1619 + }, + { + "epoch": 0.1756287944492628, + "grad_norm": 5.248568534851074, + "learning_rate": 1.8897432177571452e-05, + "loss": 0.0864, + "step": 1620 + }, + { + "epoch": 0.1757372072853426, + "grad_norm": 4.806812286376953, + "learning_rate": 1.8896706804004064e-05, + "loss": 0.0539, + "step": 1621 + }, + { + "epoch": 0.17584562012142238, + "grad_norm": 3.618622064590454, + "learning_rate": 1.8895981430436677e-05, + "loss": 0.061, + "step": 1622 + }, + { + "epoch": 0.17595403295750217, + "grad_norm": 3.2319347858428955, + "learning_rate": 1.889525605686929e-05, + "loss": 0.0514, + "step": 1623 + }, + { + "epoch": 0.17606244579358196, + "grad_norm": 3.314210891723633, + "learning_rate": 1.88945306833019e-05, + "loss": 0.0388, + "step": 1624 + }, + { + "epoch": 0.17617085862966175, + "grad_norm": 2.7897868156433105, + "learning_rate": 1.8893805309734513e-05, + "loss": 0.0369, + "step": 1625 + }, + { + "epoch": 0.17627927146574154, + "grad_norm": 0.9827674627304077, + "learning_rate": 1.889307993616713e-05, + "loss": 0.0269, + "step": 1626 + }, + { + "epoch": 0.17638768430182133, + "grad_norm": 3.2353875637054443, + "learning_rate": 1.889235456259974e-05, + "loss": 0.0894, + "step": 1627 + }, + { + "epoch": 0.17649609713790113, + "grad_norm": 3.3815250396728516, + "learning_rate": 1.8891629189032353e-05, + "loss": 0.0178, + "step": 1628 + }, + { + "epoch": 0.17660450997398092, + "grad_norm": 1.818088173866272, + "learning_rate": 1.889090381546497e-05, + "loss": 0.0201, + "step": 1629 + }, + { + "epoch": 0.1767129228100607, + "grad_norm": 3.7387378215789795, + "learning_rate": 1.889017844189758e-05, + "loss": 0.0667, + "step": 1630 + }, + { + "epoch": 0.1768213356461405, + "grad_norm": 5.588094234466553, + "learning_rate": 1.8889453068330193e-05, + "loss": 0.0681, + "step": 1631 + }, + { + "epoch": 0.1769297484822203, + "grad_norm": 2.9530086517333984, + "learning_rate": 1.8888727694762805e-05, + "loss": 0.0258, + "step": 1632 + }, + { + "epoch": 0.17703816131830008, + "grad_norm": 6.043403148651123, + "learning_rate": 1.8888002321195417e-05, + "loss": 0.0851, + "step": 1633 + }, + { + "epoch": 0.17714657415437987, + "grad_norm": 2.6778223514556885, + "learning_rate": 1.888727694762803e-05, + "loss": 0.0479, + "step": 1634 + }, + { + "epoch": 0.17725498699045966, + "grad_norm": 3.615062713623047, + "learning_rate": 1.8886551574060645e-05, + "loss": 0.0439, + "step": 1635 + }, + { + "epoch": 0.17736339982653945, + "grad_norm": 4.561100959777832, + "learning_rate": 1.8885826200493257e-05, + "loss": 0.0658, + "step": 1636 + }, + { + "epoch": 0.17747181266261924, + "grad_norm": 3.288841962814331, + "learning_rate": 1.888510082692587e-05, + "loss": 0.0358, + "step": 1637 + }, + { + "epoch": 0.17758022549869903, + "grad_norm": 5.455854892730713, + "learning_rate": 1.888437545335848e-05, + "loss": 0.0383, + "step": 1638 + }, + { + "epoch": 0.17768863833477883, + "grad_norm": 1.1051223278045654, + "learning_rate": 1.8883650079791093e-05, + "loss": 0.0181, + "step": 1639 + }, + { + "epoch": 0.17779705117085862, + "grad_norm": 2.7602505683898926, + "learning_rate": 1.8882924706223705e-05, + "loss": 0.0729, + "step": 1640 + }, + { + "epoch": 0.17790546400693844, + "grad_norm": 4.024786472320557, + "learning_rate": 1.8882199332656317e-05, + "loss": 0.0969, + "step": 1641 + }, + { + "epoch": 0.17801387684301823, + "grad_norm": 2.602782964706421, + "learning_rate": 1.8881473959088933e-05, + "loss": 0.0295, + "step": 1642 + }, + { + "epoch": 0.17812228967909802, + "grad_norm": 6.9056782722473145, + "learning_rate": 1.8880748585521545e-05, + "loss": 0.0274, + "step": 1643 + }, + { + "epoch": 0.1782307025151778, + "grad_norm": 2.397981643676758, + "learning_rate": 1.8880023211954157e-05, + "loss": 0.0389, + "step": 1644 + }, + { + "epoch": 0.1783391153512576, + "grad_norm": 1.6600570678710938, + "learning_rate": 1.887929783838677e-05, + "loss": 0.0177, + "step": 1645 + }, + { + "epoch": 0.1784475281873374, + "grad_norm": 3.128948211669922, + "learning_rate": 1.887857246481938e-05, + "loss": 0.0457, + "step": 1646 + }, + { + "epoch": 0.17855594102341718, + "grad_norm": 7.045431613922119, + "learning_rate": 1.8877847091251997e-05, + "loss": 0.0696, + "step": 1647 + }, + { + "epoch": 0.17866435385949697, + "grad_norm": 6.870469570159912, + "learning_rate": 1.887712171768461e-05, + "loss": 0.0852, + "step": 1648 + }, + { + "epoch": 0.17877276669557676, + "grad_norm": 2.587506055831909, + "learning_rate": 1.887639634411722e-05, + "loss": 0.041, + "step": 1649 + }, + { + "epoch": 0.17888117953165655, + "grad_norm": 3.781921863555908, + "learning_rate": 1.8875670970549837e-05, + "loss": 0.0454, + "step": 1650 + }, + { + "epoch": 0.17898959236773634, + "grad_norm": 4.5128865242004395, + "learning_rate": 1.887494559698245e-05, + "loss": 0.053, + "step": 1651 + }, + { + "epoch": 0.17909800520381614, + "grad_norm": 4.971549987792969, + "learning_rate": 1.887422022341506e-05, + "loss": 0.0283, + "step": 1652 + }, + { + "epoch": 0.17920641803989593, + "grad_norm": 2.654900312423706, + "learning_rate": 1.8873494849847673e-05, + "loss": 0.0319, + "step": 1653 + }, + { + "epoch": 0.17931483087597572, + "grad_norm": 3.813060998916626, + "learning_rate": 1.8872769476280286e-05, + "loss": 0.052, + "step": 1654 + }, + { + "epoch": 0.1794232437120555, + "grad_norm": 2.6613776683807373, + "learning_rate": 1.8872044102712898e-05, + "loss": 0.0417, + "step": 1655 + }, + { + "epoch": 0.1795316565481353, + "grad_norm": 6.1018548011779785, + "learning_rate": 1.887131872914551e-05, + "loss": 0.0668, + "step": 1656 + }, + { + "epoch": 0.1796400693842151, + "grad_norm": 0.6697196364402771, + "learning_rate": 1.8870593355578125e-05, + "loss": 0.0121, + "step": 1657 + }, + { + "epoch": 0.17974848222029488, + "grad_norm": 4.51483678817749, + "learning_rate": 1.8869867982010737e-05, + "loss": 0.067, + "step": 1658 + }, + { + "epoch": 0.17985689505637467, + "grad_norm": 4.708706378936768, + "learning_rate": 1.886914260844335e-05, + "loss": 0.0936, + "step": 1659 + }, + { + "epoch": 0.17996530789245446, + "grad_norm": 2.0856494903564453, + "learning_rate": 1.8868417234875962e-05, + "loss": 0.0433, + "step": 1660 + }, + { + "epoch": 0.18007372072853425, + "grad_norm": 1.6569619178771973, + "learning_rate": 1.8867691861308574e-05, + "loss": 0.0354, + "step": 1661 + }, + { + "epoch": 0.18018213356461404, + "grad_norm": 5.131792068481445, + "learning_rate": 1.8866966487741186e-05, + "loss": 0.0897, + "step": 1662 + }, + { + "epoch": 0.18029054640069384, + "grad_norm": 1.8183211088180542, + "learning_rate": 1.88662411141738e-05, + "loss": 0.0348, + "step": 1663 + }, + { + "epoch": 0.18039895923677363, + "grad_norm": 1.7585818767547607, + "learning_rate": 1.8865515740606414e-05, + "loss": 0.039, + "step": 1664 + }, + { + "epoch": 0.18050737207285342, + "grad_norm": 2.3351151943206787, + "learning_rate": 1.8864790367039026e-05, + "loss": 0.0539, + "step": 1665 + }, + { + "epoch": 0.1806157849089332, + "grad_norm": 3.1367995738983154, + "learning_rate": 1.886406499347164e-05, + "loss": 0.0327, + "step": 1666 + }, + { + "epoch": 0.180724197745013, + "grad_norm": 1.92263925075531, + "learning_rate": 1.8863339619904254e-05, + "loss": 0.0451, + "step": 1667 + }, + { + "epoch": 0.1808326105810928, + "grad_norm": 4.8627028465271, + "learning_rate": 1.8862614246336866e-05, + "loss": 0.1727, + "step": 1668 + }, + { + "epoch": 0.18094102341717258, + "grad_norm": 4.610968589782715, + "learning_rate": 1.8861888872769478e-05, + "loss": 0.0812, + "step": 1669 + }, + { + "epoch": 0.18104943625325237, + "grad_norm": 3.43160343170166, + "learning_rate": 1.886116349920209e-05, + "loss": 0.0325, + "step": 1670 + }, + { + "epoch": 0.1811578490893322, + "grad_norm": 6.437473297119141, + "learning_rate": 1.8860438125634702e-05, + "loss": 0.0721, + "step": 1671 + }, + { + "epoch": 0.18126626192541198, + "grad_norm": 3.348219394683838, + "learning_rate": 1.8859712752067318e-05, + "loss": 0.0824, + "step": 1672 + }, + { + "epoch": 0.18137467476149177, + "grad_norm": 3.5057053565979004, + "learning_rate": 1.885898737849993e-05, + "loss": 0.0607, + "step": 1673 + }, + { + "epoch": 0.18148308759757156, + "grad_norm": 5.580626010894775, + "learning_rate": 1.8858262004932542e-05, + "loss": 0.04, + "step": 1674 + }, + { + "epoch": 0.18159150043365135, + "grad_norm": 7.861453056335449, + "learning_rate": 1.8857536631365154e-05, + "loss": 0.0917, + "step": 1675 + }, + { + "epoch": 0.18169991326973114, + "grad_norm": 3.02555775642395, + "learning_rate": 1.8856811257797766e-05, + "loss": 0.018, + "step": 1676 + }, + { + "epoch": 0.18180832610581094, + "grad_norm": 1.7066056728363037, + "learning_rate": 1.885608588423038e-05, + "loss": 0.0448, + "step": 1677 + }, + { + "epoch": 0.18191673894189073, + "grad_norm": 4.367844104766846, + "learning_rate": 1.8855360510662994e-05, + "loss": 0.0608, + "step": 1678 + }, + { + "epoch": 0.18202515177797052, + "grad_norm": 2.042638063430786, + "learning_rate": 1.8854635137095606e-05, + "loss": 0.0356, + "step": 1679 + }, + { + "epoch": 0.1821335646140503, + "grad_norm": 3.0485424995422363, + "learning_rate": 1.8853909763528218e-05, + "loss": 0.1099, + "step": 1680 + }, + { + "epoch": 0.1822419774501301, + "grad_norm": 7.577797889709473, + "learning_rate": 1.8853184389960834e-05, + "loss": 0.0752, + "step": 1681 + }, + { + "epoch": 0.1823503902862099, + "grad_norm": 1.8915272951126099, + "learning_rate": 1.8852459016393446e-05, + "loss": 0.0459, + "step": 1682 + }, + { + "epoch": 0.18245880312228968, + "grad_norm": 4.293701171875, + "learning_rate": 1.8851733642826058e-05, + "loss": 0.0859, + "step": 1683 + }, + { + "epoch": 0.18256721595836947, + "grad_norm": 3.2357537746429443, + "learning_rate": 1.885100826925867e-05, + "loss": 0.0862, + "step": 1684 + }, + { + "epoch": 0.18267562879444926, + "grad_norm": 5.4621686935424805, + "learning_rate": 1.8850282895691282e-05, + "loss": 0.0505, + "step": 1685 + }, + { + "epoch": 0.18278404163052905, + "grad_norm": 0.5618569850921631, + "learning_rate": 1.8849557522123894e-05, + "loss": 0.0112, + "step": 1686 + }, + { + "epoch": 0.18289245446660884, + "grad_norm": 2.283707618713379, + "learning_rate": 1.884883214855651e-05, + "loss": 0.0425, + "step": 1687 + }, + { + "epoch": 0.18300086730268864, + "grad_norm": 6.350940227508545, + "learning_rate": 1.8848106774989122e-05, + "loss": 0.0549, + "step": 1688 + }, + { + "epoch": 0.18310928013876843, + "grad_norm": 2.9755406379699707, + "learning_rate": 1.8847381401421734e-05, + "loss": 0.0383, + "step": 1689 + }, + { + "epoch": 0.18321769297484822, + "grad_norm": 1.3410348892211914, + "learning_rate": 1.8846656027854346e-05, + "loss": 0.0305, + "step": 1690 + }, + { + "epoch": 0.183326105810928, + "grad_norm": 2.7654316425323486, + "learning_rate": 1.884593065428696e-05, + "loss": 0.0312, + "step": 1691 + }, + { + "epoch": 0.1834345186470078, + "grad_norm": 2.461711883544922, + "learning_rate": 1.884520528071957e-05, + "loss": 0.0358, + "step": 1692 + }, + { + "epoch": 0.1835429314830876, + "grad_norm": 2.542790174484253, + "learning_rate": 1.8844479907152183e-05, + "loss": 0.0824, + "step": 1693 + }, + { + "epoch": 0.18365134431916738, + "grad_norm": 5.688746929168701, + "learning_rate": 1.88437545335848e-05, + "loss": 0.0409, + "step": 1694 + }, + { + "epoch": 0.18375975715524717, + "grad_norm": 3.844144344329834, + "learning_rate": 1.884302916001741e-05, + "loss": 0.0568, + "step": 1695 + }, + { + "epoch": 0.18386816999132696, + "grad_norm": 7.010463237762451, + "learning_rate": 1.8842303786450023e-05, + "loss": 0.0337, + "step": 1696 + }, + { + "epoch": 0.18397658282740675, + "grad_norm": 3.494554042816162, + "learning_rate": 1.8841578412882635e-05, + "loss": 0.0637, + "step": 1697 + }, + { + "epoch": 0.18408499566348654, + "grad_norm": 1.4030448198318481, + "learning_rate": 1.8840853039315247e-05, + "loss": 0.0303, + "step": 1698 + }, + { + "epoch": 0.18419340849956634, + "grad_norm": 3.9195590019226074, + "learning_rate": 1.8840127665747862e-05, + "loss": 0.0575, + "step": 1699 + }, + { + "epoch": 0.18430182133564615, + "grad_norm": 2.0403895378112793, + "learning_rate": 1.8839402292180475e-05, + "loss": 0.0257, + "step": 1700 + }, + { + "epoch": 0.18441023417172595, + "grad_norm": 6.183429718017578, + "learning_rate": 1.8838676918613087e-05, + "loss": 0.1266, + "step": 1701 + }, + { + "epoch": 0.18451864700780574, + "grad_norm": 2.8405542373657227, + "learning_rate": 1.8837951545045702e-05, + "loss": 0.0286, + "step": 1702 + }, + { + "epoch": 0.18462705984388553, + "grad_norm": 4.902461051940918, + "learning_rate": 1.8837226171478314e-05, + "loss": 0.0289, + "step": 1703 + }, + { + "epoch": 0.18473547267996532, + "grad_norm": 1.403379201889038, + "learning_rate": 1.8836500797910927e-05, + "loss": 0.0145, + "step": 1704 + }, + { + "epoch": 0.1848438855160451, + "grad_norm": 4.432432651519775, + "learning_rate": 1.883577542434354e-05, + "loss": 0.0422, + "step": 1705 + }, + { + "epoch": 0.1849522983521249, + "grad_norm": 2.782142162322998, + "learning_rate": 1.883505005077615e-05, + "loss": 0.0662, + "step": 1706 + }, + { + "epoch": 0.1850607111882047, + "grad_norm": 1.6690765619277954, + "learning_rate": 1.8834324677208763e-05, + "loss": 0.01, + "step": 1707 + }, + { + "epoch": 0.18516912402428448, + "grad_norm": 1.8579723834991455, + "learning_rate": 1.8833599303641375e-05, + "loss": 0.0289, + "step": 1708 + }, + { + "epoch": 0.18527753686036427, + "grad_norm": 6.519844055175781, + "learning_rate": 1.883287393007399e-05, + "loss": 0.0418, + "step": 1709 + }, + { + "epoch": 0.18538594969644406, + "grad_norm": 4.237362861633301, + "learning_rate": 1.8832148556506603e-05, + "loss": 0.0971, + "step": 1710 + }, + { + "epoch": 0.18549436253252385, + "grad_norm": 2.3279385566711426, + "learning_rate": 1.8831423182939215e-05, + "loss": 0.0253, + "step": 1711 + }, + { + "epoch": 0.18560277536860365, + "grad_norm": 3.731497049331665, + "learning_rate": 1.8830697809371827e-05, + "loss": 0.0674, + "step": 1712 + }, + { + "epoch": 0.18571118820468344, + "grad_norm": 4.264967441558838, + "learning_rate": 1.882997243580444e-05, + "loss": 0.0355, + "step": 1713 + }, + { + "epoch": 0.18581960104076323, + "grad_norm": 3.0228326320648193, + "learning_rate": 1.882924706223705e-05, + "loss": 0.0548, + "step": 1714 + }, + { + "epoch": 0.18592801387684302, + "grad_norm": 3.6098814010620117, + "learning_rate": 1.8828521688669667e-05, + "loss": 0.0448, + "step": 1715 + }, + { + "epoch": 0.1860364267129228, + "grad_norm": 2.304701805114746, + "learning_rate": 1.882779631510228e-05, + "loss": 0.0351, + "step": 1716 + }, + { + "epoch": 0.1861448395490026, + "grad_norm": 5.994250297546387, + "learning_rate": 1.882707094153489e-05, + "loss": 0.0515, + "step": 1717 + }, + { + "epoch": 0.1862532523850824, + "grad_norm": 1.7432891130447388, + "learning_rate": 1.8826345567967507e-05, + "loss": 0.0338, + "step": 1718 + }, + { + "epoch": 0.18636166522116218, + "grad_norm": 2.9971251487731934, + "learning_rate": 1.882562019440012e-05, + "loss": 0.0434, + "step": 1719 + }, + { + "epoch": 0.18647007805724197, + "grad_norm": 4.52812385559082, + "learning_rate": 1.882489482083273e-05, + "loss": 0.0712, + "step": 1720 + }, + { + "epoch": 0.18657849089332176, + "grad_norm": 5.600216865539551, + "learning_rate": 1.8824169447265343e-05, + "loss": 0.0276, + "step": 1721 + }, + { + "epoch": 0.18668690372940155, + "grad_norm": 2.2482779026031494, + "learning_rate": 1.8823444073697955e-05, + "loss": 0.0308, + "step": 1722 + }, + { + "epoch": 0.18679531656548135, + "grad_norm": 3.535977363586426, + "learning_rate": 1.8822718700130567e-05, + "loss": 0.0619, + "step": 1723 + }, + { + "epoch": 0.18690372940156114, + "grad_norm": 6.940131664276123, + "learning_rate": 1.8821993326563183e-05, + "loss": 0.0451, + "step": 1724 + }, + { + "epoch": 0.18701214223764093, + "grad_norm": 6.754718780517578, + "learning_rate": 1.8821267952995795e-05, + "loss": 0.1821, + "step": 1725 + }, + { + "epoch": 0.18712055507372072, + "grad_norm": 3.0747547149658203, + "learning_rate": 1.8820542579428407e-05, + "loss": 0.049, + "step": 1726 + }, + { + "epoch": 0.1872289679098005, + "grad_norm": 5.839418888092041, + "learning_rate": 1.881981720586102e-05, + "loss": 0.0469, + "step": 1727 + }, + { + "epoch": 0.1873373807458803, + "grad_norm": 1.5219831466674805, + "learning_rate": 1.881909183229363e-05, + "loss": 0.0257, + "step": 1728 + }, + { + "epoch": 0.1874457935819601, + "grad_norm": 1.9641693830490112, + "learning_rate": 1.8818366458726244e-05, + "loss": 0.0252, + "step": 1729 + }, + { + "epoch": 0.1875542064180399, + "grad_norm": 3.6159844398498535, + "learning_rate": 1.881764108515886e-05, + "loss": 0.088, + "step": 1730 + }, + { + "epoch": 0.1876626192541197, + "grad_norm": 2.336268186569214, + "learning_rate": 1.881691571159147e-05, + "loss": 0.0391, + "step": 1731 + }, + { + "epoch": 0.1877710320901995, + "grad_norm": 2.295180082321167, + "learning_rate": 1.8816190338024084e-05, + "loss": 0.0327, + "step": 1732 + }, + { + "epoch": 0.18787944492627928, + "grad_norm": 4.717994689941406, + "learning_rate": 1.88154649644567e-05, + "loss": 0.071, + "step": 1733 + }, + { + "epoch": 0.18798785776235907, + "grad_norm": 2.7705461978912354, + "learning_rate": 1.881473959088931e-05, + "loss": 0.0434, + "step": 1734 + }, + { + "epoch": 0.18809627059843886, + "grad_norm": 4.760623455047607, + "learning_rate": 1.8814014217321923e-05, + "loss": 0.0591, + "step": 1735 + }, + { + "epoch": 0.18820468343451865, + "grad_norm": 1.862112283706665, + "learning_rate": 1.8813288843754535e-05, + "loss": 0.019, + "step": 1736 + }, + { + "epoch": 0.18831309627059845, + "grad_norm": 3.1854305267333984, + "learning_rate": 1.8812563470187148e-05, + "loss": 0.0453, + "step": 1737 + }, + { + "epoch": 0.18842150910667824, + "grad_norm": 2.6007816791534424, + "learning_rate": 1.881183809661976e-05, + "loss": 0.0377, + "step": 1738 + }, + { + "epoch": 0.18852992194275803, + "grad_norm": 4.488107204437256, + "learning_rate": 1.8811112723052375e-05, + "loss": 0.0309, + "step": 1739 + }, + { + "epoch": 0.18863833477883782, + "grad_norm": 2.914067506790161, + "learning_rate": 1.8810387349484987e-05, + "loss": 0.0493, + "step": 1740 + }, + { + "epoch": 0.1887467476149176, + "grad_norm": 5.8611931800842285, + "learning_rate": 1.88096619759176e-05, + "loss": 0.0952, + "step": 1741 + }, + { + "epoch": 0.1888551604509974, + "grad_norm": 2.3738512992858887, + "learning_rate": 1.8808936602350212e-05, + "loss": 0.0264, + "step": 1742 + }, + { + "epoch": 0.1889635732870772, + "grad_norm": 2.2231087684631348, + "learning_rate": 1.8808211228782824e-05, + "loss": 0.0242, + "step": 1743 + }, + { + "epoch": 0.18907198612315698, + "grad_norm": 3.431577682495117, + "learning_rate": 1.8807485855215436e-05, + "loss": 0.055, + "step": 1744 + }, + { + "epoch": 0.18918039895923677, + "grad_norm": 2.0954394340515137, + "learning_rate": 1.8806760481648048e-05, + "loss": 0.0344, + "step": 1745 + }, + { + "epoch": 0.18928881179531656, + "grad_norm": 1.8626524209976196, + "learning_rate": 1.8806035108080664e-05, + "loss": 0.0199, + "step": 1746 + }, + { + "epoch": 0.18939722463139635, + "grad_norm": 3.2162554264068604, + "learning_rate": 1.8805309734513276e-05, + "loss": 0.0373, + "step": 1747 + }, + { + "epoch": 0.18950563746747615, + "grad_norm": 2.6284189224243164, + "learning_rate": 1.8804584360945888e-05, + "loss": 0.0327, + "step": 1748 + }, + { + "epoch": 0.18961405030355594, + "grad_norm": 3.0379700660705566, + "learning_rate": 1.88038589873785e-05, + "loss": 0.0642, + "step": 1749 + }, + { + "epoch": 0.18972246313963573, + "grad_norm": 3.2878620624542236, + "learning_rate": 1.8803133613811116e-05, + "loss": 0.0707, + "step": 1750 + }, + { + "epoch": 0.18983087597571552, + "grad_norm": 5.988759517669678, + "learning_rate": 1.8802408240243728e-05, + "loss": 0.069, + "step": 1751 + }, + { + "epoch": 0.1899392888117953, + "grad_norm": 3.353227138519287, + "learning_rate": 1.880168286667634e-05, + "loss": 0.0438, + "step": 1752 + }, + { + "epoch": 0.1900477016478751, + "grad_norm": 4.371632099151611, + "learning_rate": 1.8800957493108952e-05, + "loss": 0.0344, + "step": 1753 + }, + { + "epoch": 0.1901561144839549, + "grad_norm": 2.410648822784424, + "learning_rate": 1.8800232119541568e-05, + "loss": 0.0236, + "step": 1754 + }, + { + "epoch": 0.19026452732003468, + "grad_norm": 2.760805606842041, + "learning_rate": 1.879950674597418e-05, + "loss": 0.0378, + "step": 1755 + }, + { + "epoch": 0.19037294015611447, + "grad_norm": 3.5627646446228027, + "learning_rate": 1.8798781372406792e-05, + "loss": 0.0597, + "step": 1756 + }, + { + "epoch": 0.19048135299219426, + "grad_norm": 4.546586036682129, + "learning_rate": 1.8798055998839404e-05, + "loss": 0.0507, + "step": 1757 + }, + { + "epoch": 0.19058976582827405, + "grad_norm": 2.1553761959075928, + "learning_rate": 1.8797330625272016e-05, + "loss": 0.0233, + "step": 1758 + }, + { + "epoch": 0.19069817866435385, + "grad_norm": 2.840252161026001, + "learning_rate": 1.879660525170463e-05, + "loss": 0.065, + "step": 1759 + }, + { + "epoch": 0.19080659150043366, + "grad_norm": 2.8432862758636475, + "learning_rate": 1.879587987813724e-05, + "loss": 0.0481, + "step": 1760 + }, + { + "epoch": 0.19091500433651346, + "grad_norm": 4.807224750518799, + "learning_rate": 1.8795154504569856e-05, + "loss": 0.0669, + "step": 1761 + }, + { + "epoch": 0.19102341717259325, + "grad_norm": 2.9164187908172607, + "learning_rate": 1.8794429131002468e-05, + "loss": 0.0409, + "step": 1762 + }, + { + "epoch": 0.19113183000867304, + "grad_norm": 3.7512547969818115, + "learning_rate": 1.879370375743508e-05, + "loss": 0.0435, + "step": 1763 + }, + { + "epoch": 0.19124024284475283, + "grad_norm": 4.094489574432373, + "learning_rate": 1.8792978383867692e-05, + "loss": 0.0354, + "step": 1764 + }, + { + "epoch": 0.19134865568083262, + "grad_norm": 8.347036361694336, + "learning_rate": 1.8792253010300305e-05, + "loss": 0.0412, + "step": 1765 + }, + { + "epoch": 0.1914570685169124, + "grad_norm": 3.3814139366149902, + "learning_rate": 1.8791527636732917e-05, + "loss": 0.0814, + "step": 1766 + }, + { + "epoch": 0.1915654813529922, + "grad_norm": 2.867457866668701, + "learning_rate": 1.8790802263165532e-05, + "loss": 0.0169, + "step": 1767 + }, + { + "epoch": 0.191673894189072, + "grad_norm": 6.168741703033447, + "learning_rate": 1.8790076889598144e-05, + "loss": 0.0547, + "step": 1768 + }, + { + "epoch": 0.19178230702515178, + "grad_norm": 2.916579484939575, + "learning_rate": 1.878935151603076e-05, + "loss": 0.0438, + "step": 1769 + }, + { + "epoch": 0.19189071986123157, + "grad_norm": 3.5368149280548096, + "learning_rate": 1.8788626142463372e-05, + "loss": 0.0587, + "step": 1770 + }, + { + "epoch": 0.19199913269731136, + "grad_norm": 6.071336269378662, + "learning_rate": 1.8787900768895984e-05, + "loss": 0.0857, + "step": 1771 + }, + { + "epoch": 0.19210754553339116, + "grad_norm": 49.05857849121094, + "learning_rate": 1.8787175395328596e-05, + "loss": 0.0702, + "step": 1772 + }, + { + "epoch": 0.19221595836947095, + "grad_norm": 4.38391637802124, + "learning_rate": 1.878645002176121e-05, + "loss": 0.0647, + "step": 1773 + }, + { + "epoch": 0.19232437120555074, + "grad_norm": 2.99184250831604, + "learning_rate": 1.878572464819382e-05, + "loss": 0.0676, + "step": 1774 + }, + { + "epoch": 0.19243278404163053, + "grad_norm": 2.9388511180877686, + "learning_rate": 1.8784999274626433e-05, + "loss": 0.0518, + "step": 1775 + }, + { + "epoch": 0.19254119687771032, + "grad_norm": 2.239208936691284, + "learning_rate": 1.878427390105905e-05, + "loss": 0.0439, + "step": 1776 + }, + { + "epoch": 0.1926496097137901, + "grad_norm": 4.3146491050720215, + "learning_rate": 1.878354852749166e-05, + "loss": 0.0598, + "step": 1777 + }, + { + "epoch": 0.1927580225498699, + "grad_norm": 2.7052741050720215, + "learning_rate": 1.8782823153924273e-05, + "loss": 0.0371, + "step": 1778 + }, + { + "epoch": 0.1928664353859497, + "grad_norm": 3.319444179534912, + "learning_rate": 1.8782097780356885e-05, + "loss": 0.0366, + "step": 1779 + }, + { + "epoch": 0.19297484822202948, + "grad_norm": 2.036278009414673, + "learning_rate": 1.8781372406789497e-05, + "loss": 0.0219, + "step": 1780 + }, + { + "epoch": 0.19308326105810927, + "grad_norm": 4.2786664962768555, + "learning_rate": 1.878064703322211e-05, + "loss": 0.0461, + "step": 1781 + }, + { + "epoch": 0.19319167389418906, + "grad_norm": 2.195392370223999, + "learning_rate": 1.8779921659654725e-05, + "loss": 0.0387, + "step": 1782 + }, + { + "epoch": 0.19330008673026886, + "grad_norm": 2.185108184814453, + "learning_rate": 1.8779196286087337e-05, + "loss": 0.0293, + "step": 1783 + }, + { + "epoch": 0.19340849956634865, + "grad_norm": 2.4690842628479004, + "learning_rate": 1.877847091251995e-05, + "loss": 0.0165, + "step": 1784 + }, + { + "epoch": 0.19351691240242844, + "grad_norm": 2.613454818725586, + "learning_rate": 1.8777745538952564e-05, + "loss": 0.0564, + "step": 1785 + }, + { + "epoch": 0.19362532523850823, + "grad_norm": 5.967633247375488, + "learning_rate": 1.8777020165385177e-05, + "loss": 0.088, + "step": 1786 + }, + { + "epoch": 0.19373373807458802, + "grad_norm": 4.158743858337402, + "learning_rate": 1.877629479181779e-05, + "loss": 0.1448, + "step": 1787 + }, + { + "epoch": 0.1938421509106678, + "grad_norm": 1.5819330215454102, + "learning_rate": 1.87755694182504e-05, + "loss": 0.0281, + "step": 1788 + }, + { + "epoch": 0.19395056374674763, + "grad_norm": 9.725539207458496, + "learning_rate": 1.8774844044683013e-05, + "loss": 0.0543, + "step": 1789 + }, + { + "epoch": 0.19405897658282742, + "grad_norm": 2.5401506423950195, + "learning_rate": 1.8774118671115625e-05, + "loss": 0.025, + "step": 1790 + }, + { + "epoch": 0.1941673894189072, + "grad_norm": 2.8911032676696777, + "learning_rate": 1.877339329754824e-05, + "loss": 0.0735, + "step": 1791 + }, + { + "epoch": 0.194275802254987, + "grad_norm": 2.785625696182251, + "learning_rate": 1.8772667923980853e-05, + "loss": 0.0444, + "step": 1792 + }, + { + "epoch": 0.1943842150910668, + "grad_norm": 3.958883047103882, + "learning_rate": 1.8771942550413465e-05, + "loss": 0.0878, + "step": 1793 + }, + { + "epoch": 0.19449262792714658, + "grad_norm": 4.698209762573242, + "learning_rate": 1.8771217176846077e-05, + "loss": 0.0846, + "step": 1794 + }, + { + "epoch": 0.19460104076322637, + "grad_norm": 5.090432643890381, + "learning_rate": 1.877049180327869e-05, + "loss": 0.0717, + "step": 1795 + }, + { + "epoch": 0.19470945359930616, + "grad_norm": 1.4053198099136353, + "learning_rate": 1.87697664297113e-05, + "loss": 0.0092, + "step": 1796 + }, + { + "epoch": 0.19481786643538596, + "grad_norm": 3.2640602588653564, + "learning_rate": 1.8769041056143913e-05, + "loss": 0.0172, + "step": 1797 + }, + { + "epoch": 0.19492627927146575, + "grad_norm": 1.3738595247268677, + "learning_rate": 1.876831568257653e-05, + "loss": 0.0181, + "step": 1798 + }, + { + "epoch": 0.19503469210754554, + "grad_norm": 6.742196559906006, + "learning_rate": 1.876759030900914e-05, + "loss": 0.202, + "step": 1799 + }, + { + "epoch": 0.19514310494362533, + "grad_norm": 5.414575576782227, + "learning_rate": 1.8766864935441753e-05, + "loss": 0.0626, + "step": 1800 + }, + { + "epoch": 0.19525151777970512, + "grad_norm": 3.362173318862915, + "learning_rate": 1.8766139561874365e-05, + "loss": 0.0546, + "step": 1801 + }, + { + "epoch": 0.1953599306157849, + "grad_norm": 2.6045122146606445, + "learning_rate": 1.876541418830698e-05, + "loss": 0.0344, + "step": 1802 + }, + { + "epoch": 0.1954683434518647, + "grad_norm": 3.6753523349761963, + "learning_rate": 1.8764688814739593e-05, + "loss": 0.0708, + "step": 1803 + }, + { + "epoch": 0.1955767562879445, + "grad_norm": 2.0674026012420654, + "learning_rate": 1.8763963441172205e-05, + "loss": 0.0251, + "step": 1804 + }, + { + "epoch": 0.19568516912402428, + "grad_norm": 1.6511573791503906, + "learning_rate": 1.8763238067604817e-05, + "loss": 0.0466, + "step": 1805 + }, + { + "epoch": 0.19579358196010407, + "grad_norm": 4.1113386154174805, + "learning_rate": 1.8762512694037433e-05, + "loss": 0.0284, + "step": 1806 + }, + { + "epoch": 0.19590199479618386, + "grad_norm": 4.457508563995361, + "learning_rate": 1.8761787320470045e-05, + "loss": 0.0284, + "step": 1807 + }, + { + "epoch": 0.19601040763226366, + "grad_norm": 3.3705127239227295, + "learning_rate": 1.8761061946902657e-05, + "loss": 0.0508, + "step": 1808 + }, + { + "epoch": 0.19611882046834345, + "grad_norm": 3.964195966720581, + "learning_rate": 1.876033657333527e-05, + "loss": 0.0197, + "step": 1809 + }, + { + "epoch": 0.19622723330442324, + "grad_norm": 5.261170864105225, + "learning_rate": 1.875961119976788e-05, + "loss": 0.0353, + "step": 1810 + }, + { + "epoch": 0.19633564614050303, + "grad_norm": 3.0400314331054688, + "learning_rate": 1.8758885826200494e-05, + "loss": 0.1004, + "step": 1811 + }, + { + "epoch": 0.19644405897658282, + "grad_norm": 2.0966408252716064, + "learning_rate": 1.8758160452633106e-05, + "loss": 0.0135, + "step": 1812 + }, + { + "epoch": 0.1965524718126626, + "grad_norm": 6.234340667724609, + "learning_rate": 1.875743507906572e-05, + "loss": 0.0636, + "step": 1813 + }, + { + "epoch": 0.1966608846487424, + "grad_norm": 2.652523994445801, + "learning_rate": 1.8756709705498333e-05, + "loss": 0.0364, + "step": 1814 + }, + { + "epoch": 0.1967692974848222, + "grad_norm": 2.9469406604766846, + "learning_rate": 1.8755984331930946e-05, + "loss": 0.0389, + "step": 1815 + }, + { + "epoch": 0.19687771032090198, + "grad_norm": 5.904929161071777, + "learning_rate": 1.8755258958363558e-05, + "loss": 0.0955, + "step": 1816 + }, + { + "epoch": 0.19698612315698177, + "grad_norm": 1.7498537302017212, + "learning_rate": 1.875453358479617e-05, + "loss": 0.0171, + "step": 1817 + }, + { + "epoch": 0.19709453599306156, + "grad_norm": 4.314613342285156, + "learning_rate": 1.8753808211228782e-05, + "loss": 0.0393, + "step": 1818 + }, + { + "epoch": 0.19720294882914138, + "grad_norm": 3.00742506980896, + "learning_rate": 1.8753082837661398e-05, + "loss": 0.0801, + "step": 1819 + }, + { + "epoch": 0.19731136166522117, + "grad_norm": 6.87745475769043, + "learning_rate": 1.875235746409401e-05, + "loss": 0.0537, + "step": 1820 + }, + { + "epoch": 0.19741977450130097, + "grad_norm": 3.849153757095337, + "learning_rate": 1.8751632090526625e-05, + "loss": 0.0528, + "step": 1821 + }, + { + "epoch": 0.19752818733738076, + "grad_norm": 1.4793391227722168, + "learning_rate": 1.8750906716959237e-05, + "loss": 0.0139, + "step": 1822 + }, + { + "epoch": 0.19763660017346055, + "grad_norm": 1.9141095876693726, + "learning_rate": 1.875018134339185e-05, + "loss": 0.0171, + "step": 1823 + }, + { + "epoch": 0.19774501300954034, + "grad_norm": 3.065286159515381, + "learning_rate": 1.8749455969824462e-05, + "loss": 0.0422, + "step": 1824 + }, + { + "epoch": 0.19785342584562013, + "grad_norm": 4.203239917755127, + "learning_rate": 1.8748730596257074e-05, + "loss": 0.0973, + "step": 1825 + }, + { + "epoch": 0.19796183868169992, + "grad_norm": 2.597187042236328, + "learning_rate": 1.8748005222689686e-05, + "loss": 0.0639, + "step": 1826 + }, + { + "epoch": 0.1980702515177797, + "grad_norm": 3.170769453048706, + "learning_rate": 1.8747279849122298e-05, + "loss": 0.0725, + "step": 1827 + }, + { + "epoch": 0.1981786643538595, + "grad_norm": 4.923027515411377, + "learning_rate": 1.8746554475554914e-05, + "loss": 0.039, + "step": 1828 + }, + { + "epoch": 0.1982870771899393, + "grad_norm": 2.8209609985351562, + "learning_rate": 1.8745829101987526e-05, + "loss": 0.0346, + "step": 1829 + }, + { + "epoch": 0.19839549002601908, + "grad_norm": 2.6795742511749268, + "learning_rate": 1.8745103728420138e-05, + "loss": 0.0368, + "step": 1830 + }, + { + "epoch": 0.19850390286209887, + "grad_norm": 2.7862327098846436, + "learning_rate": 1.874437835485275e-05, + "loss": 0.0265, + "step": 1831 + }, + { + "epoch": 0.19861231569817867, + "grad_norm": 1.5646905899047852, + "learning_rate": 1.8743652981285362e-05, + "loss": 0.0261, + "step": 1832 + }, + { + "epoch": 0.19872072853425846, + "grad_norm": 1.8656575679779053, + "learning_rate": 1.8742927607717974e-05, + "loss": 0.0245, + "step": 1833 + }, + { + "epoch": 0.19882914137033825, + "grad_norm": 1.547572135925293, + "learning_rate": 1.874220223415059e-05, + "loss": 0.0361, + "step": 1834 + }, + { + "epoch": 0.19893755420641804, + "grad_norm": 4.40411901473999, + "learning_rate": 1.8741476860583202e-05, + "loss": 0.0278, + "step": 1835 + }, + { + "epoch": 0.19904596704249783, + "grad_norm": 4.594245433807373, + "learning_rate": 1.8740751487015814e-05, + "loss": 0.0455, + "step": 1836 + }, + { + "epoch": 0.19915437987857762, + "grad_norm": 2.7350902557373047, + "learning_rate": 1.874002611344843e-05, + "loss": 0.0207, + "step": 1837 + }, + { + "epoch": 0.1992627927146574, + "grad_norm": 7.156994819641113, + "learning_rate": 1.8739300739881042e-05, + "loss": 0.0639, + "step": 1838 + }, + { + "epoch": 0.1993712055507372, + "grad_norm": 3.807030200958252, + "learning_rate": 1.8738575366313654e-05, + "loss": 0.0244, + "step": 1839 + }, + { + "epoch": 0.199479618386817, + "grad_norm": 8.695966720581055, + "learning_rate": 1.8737849992746266e-05, + "loss": 0.071, + "step": 1840 + }, + { + "epoch": 0.19958803122289678, + "grad_norm": 1.757629156112671, + "learning_rate": 1.8737124619178878e-05, + "loss": 0.0394, + "step": 1841 + }, + { + "epoch": 0.19969644405897657, + "grad_norm": 4.8103203773498535, + "learning_rate": 1.873639924561149e-05, + "loss": 0.0957, + "step": 1842 + }, + { + "epoch": 0.19980485689505637, + "grad_norm": 4.458452224731445, + "learning_rate": 1.8735673872044106e-05, + "loss": 0.0662, + "step": 1843 + }, + { + "epoch": 0.19991326973113616, + "grad_norm": 4.402895450592041, + "learning_rate": 1.8734948498476718e-05, + "loss": 0.044, + "step": 1844 + }, + { + "epoch": 0.20002168256721595, + "grad_norm": 3.3460137844085693, + "learning_rate": 1.873422312490933e-05, + "loss": 0.0518, + "step": 1845 + }, + { + "epoch": 0.20013009540329574, + "grad_norm": 5.638411045074463, + "learning_rate": 1.8733497751341942e-05, + "loss": 0.03, + "step": 1846 + }, + { + "epoch": 0.20023850823937553, + "grad_norm": 3.136476516723633, + "learning_rate": 1.8732772377774555e-05, + "loss": 0.0914, + "step": 1847 + }, + { + "epoch": 0.20034692107545535, + "grad_norm": 6.295787811279297, + "learning_rate": 1.8732047004207167e-05, + "loss": 0.0619, + "step": 1848 + }, + { + "epoch": 0.20045533391153514, + "grad_norm": 3.3136167526245117, + "learning_rate": 1.873132163063978e-05, + "loss": 0.0295, + "step": 1849 + }, + { + "epoch": 0.20056374674761493, + "grad_norm": 6.4208664894104, + "learning_rate": 1.8730596257072394e-05, + "loss": 0.096, + "step": 1850 + }, + { + "epoch": 0.20067215958369472, + "grad_norm": 4.348890781402588, + "learning_rate": 1.8729870883505007e-05, + "loss": 0.0459, + "step": 1851 + }, + { + "epoch": 0.2007805724197745, + "grad_norm": 2.4251673221588135, + "learning_rate": 1.872914550993762e-05, + "loss": 0.0362, + "step": 1852 + }, + { + "epoch": 0.2008889852558543, + "grad_norm": 10.581982612609863, + "learning_rate": 1.872842013637023e-05, + "loss": 0.0533, + "step": 1853 + }, + { + "epoch": 0.2009973980919341, + "grad_norm": 3.3690924644470215, + "learning_rate": 1.8727694762802846e-05, + "loss": 0.0376, + "step": 1854 + }, + { + "epoch": 0.20110581092801388, + "grad_norm": 4.309840202331543, + "learning_rate": 1.872696938923546e-05, + "loss": 0.0847, + "step": 1855 + }, + { + "epoch": 0.20121422376409367, + "grad_norm": 2.0666887760162354, + "learning_rate": 1.872624401566807e-05, + "loss": 0.0456, + "step": 1856 + }, + { + "epoch": 0.20132263660017347, + "grad_norm": 1.7131179571151733, + "learning_rate": 1.8725518642100683e-05, + "loss": 0.0292, + "step": 1857 + }, + { + "epoch": 0.20143104943625326, + "grad_norm": 5.355206489562988, + "learning_rate": 1.8724793268533298e-05, + "loss": 0.0693, + "step": 1858 + }, + { + "epoch": 0.20153946227233305, + "grad_norm": 2.871251106262207, + "learning_rate": 1.872406789496591e-05, + "loss": 0.0723, + "step": 1859 + }, + { + "epoch": 0.20164787510841284, + "grad_norm": 8.05436897277832, + "learning_rate": 1.8723342521398523e-05, + "loss": 0.0475, + "step": 1860 + }, + { + "epoch": 0.20175628794449263, + "grad_norm": 4.19611930847168, + "learning_rate": 1.8722617147831135e-05, + "loss": 0.0665, + "step": 1861 + }, + { + "epoch": 0.20186470078057242, + "grad_norm": 2.6933162212371826, + "learning_rate": 1.8721891774263747e-05, + "loss": 0.0556, + "step": 1862 + }, + { + "epoch": 0.2019731136166522, + "grad_norm": 2.96453595161438, + "learning_rate": 1.872116640069636e-05, + "loss": 0.0305, + "step": 1863 + }, + { + "epoch": 0.202081526452732, + "grad_norm": 1.8203763961791992, + "learning_rate": 1.872044102712897e-05, + "loss": 0.0332, + "step": 1864 + }, + { + "epoch": 0.2021899392888118, + "grad_norm": 5.466218948364258, + "learning_rate": 1.8719715653561587e-05, + "loss": 0.0644, + "step": 1865 + }, + { + "epoch": 0.20229835212489158, + "grad_norm": 2.1292977333068848, + "learning_rate": 1.87189902799942e-05, + "loss": 0.0428, + "step": 1866 + }, + { + "epoch": 0.20240676496097137, + "grad_norm": 3.460855484008789, + "learning_rate": 1.871826490642681e-05, + "loss": 0.0349, + "step": 1867 + }, + { + "epoch": 0.20251517779705117, + "grad_norm": 2.0759050846099854, + "learning_rate": 1.8717539532859423e-05, + "loss": 0.0331, + "step": 1868 + }, + { + "epoch": 0.20262359063313096, + "grad_norm": 5.781092166900635, + "learning_rate": 1.8716814159292035e-05, + "loss": 0.0395, + "step": 1869 + }, + { + "epoch": 0.20273200346921075, + "grad_norm": 1.2893396615982056, + "learning_rate": 1.8716088785724647e-05, + "loss": 0.0168, + "step": 1870 + }, + { + "epoch": 0.20284041630529054, + "grad_norm": 2.164724111557007, + "learning_rate": 1.8715363412157263e-05, + "loss": 0.0224, + "step": 1871 + }, + { + "epoch": 0.20294882914137033, + "grad_norm": 3.5560574531555176, + "learning_rate": 1.8714638038589875e-05, + "loss": 0.0712, + "step": 1872 + }, + { + "epoch": 0.20305724197745012, + "grad_norm": 4.1737565994262695, + "learning_rate": 1.871391266502249e-05, + "loss": 0.0735, + "step": 1873 + }, + { + "epoch": 0.2031656548135299, + "grad_norm": 1.3087854385375977, + "learning_rate": 1.8713187291455103e-05, + "loss": 0.0085, + "step": 1874 + }, + { + "epoch": 0.2032740676496097, + "grad_norm": 3.4719834327697754, + "learning_rate": 1.8712461917887715e-05, + "loss": 0.0234, + "step": 1875 + }, + { + "epoch": 0.2033824804856895, + "grad_norm": 3.8513271808624268, + "learning_rate": 1.8711736544320327e-05, + "loss": 0.0611, + "step": 1876 + }, + { + "epoch": 0.20349089332176928, + "grad_norm": 0.8494982123374939, + "learning_rate": 1.871101117075294e-05, + "loss": 0.0144, + "step": 1877 + }, + { + "epoch": 0.2035993061578491, + "grad_norm": 5.388260841369629, + "learning_rate": 1.871028579718555e-05, + "loss": 0.046, + "step": 1878 + }, + { + "epoch": 0.2037077189939289, + "grad_norm": 1.8024333715438843, + "learning_rate": 1.8709560423618163e-05, + "loss": 0.0205, + "step": 1879 + }, + { + "epoch": 0.20381613183000868, + "grad_norm": 5.357097148895264, + "learning_rate": 1.870883505005078e-05, + "loss": 0.1169, + "step": 1880 + }, + { + "epoch": 0.20392454466608848, + "grad_norm": 4.7420806884765625, + "learning_rate": 1.870810967648339e-05, + "loss": 0.0413, + "step": 1881 + }, + { + "epoch": 0.20403295750216827, + "grad_norm": 3.7952468395233154, + "learning_rate": 1.8707384302916003e-05, + "loss": 0.0516, + "step": 1882 + }, + { + "epoch": 0.20414137033824806, + "grad_norm": 4.435917854309082, + "learning_rate": 1.8706658929348615e-05, + "loss": 0.029, + "step": 1883 + }, + { + "epoch": 0.20424978317432785, + "grad_norm": 6.0149407386779785, + "learning_rate": 1.8705933555781228e-05, + "loss": 0.0246, + "step": 1884 + }, + { + "epoch": 0.20435819601040764, + "grad_norm": 4.986382484436035, + "learning_rate": 1.870520818221384e-05, + "loss": 0.0863, + "step": 1885 + }, + { + "epoch": 0.20446660884648743, + "grad_norm": 2.410644292831421, + "learning_rate": 1.8704482808646455e-05, + "loss": 0.0078, + "step": 1886 + }, + { + "epoch": 0.20457502168256722, + "grad_norm": 2.4413464069366455, + "learning_rate": 1.8703757435079067e-05, + "loss": 0.0386, + "step": 1887 + }, + { + "epoch": 0.204683434518647, + "grad_norm": 2.6235342025756836, + "learning_rate": 1.870303206151168e-05, + "loss": 0.0251, + "step": 1888 + }, + { + "epoch": 0.2047918473547268, + "grad_norm": 3.7191531658172607, + "learning_rate": 1.8702306687944295e-05, + "loss": 0.0243, + "step": 1889 + }, + { + "epoch": 0.2049002601908066, + "grad_norm": 5.2519683837890625, + "learning_rate": 1.8701581314376907e-05, + "loss": 0.0675, + "step": 1890 + }, + { + "epoch": 0.20500867302688638, + "grad_norm": 5.6525139808654785, + "learning_rate": 1.870085594080952e-05, + "loss": 0.1015, + "step": 1891 + }, + { + "epoch": 0.20511708586296618, + "grad_norm": 11.325906753540039, + "learning_rate": 1.870013056724213e-05, + "loss": 0.1178, + "step": 1892 + }, + { + "epoch": 0.20522549869904597, + "grad_norm": 4.7969889640808105, + "learning_rate": 1.8699405193674744e-05, + "loss": 0.1066, + "step": 1893 + }, + { + "epoch": 0.20533391153512576, + "grad_norm": 4.931793212890625, + "learning_rate": 1.8698679820107356e-05, + "loss": 0.0336, + "step": 1894 + }, + { + "epoch": 0.20544232437120555, + "grad_norm": 4.156637668609619, + "learning_rate": 1.869795444653997e-05, + "loss": 0.0405, + "step": 1895 + }, + { + "epoch": 0.20555073720728534, + "grad_norm": 2.2087888717651367, + "learning_rate": 1.8697229072972583e-05, + "loss": 0.0257, + "step": 1896 + }, + { + "epoch": 0.20565915004336513, + "grad_norm": 4.767368316650391, + "learning_rate": 1.8696503699405196e-05, + "loss": 0.0403, + "step": 1897 + }, + { + "epoch": 0.20576756287944492, + "grad_norm": 6.103938579559326, + "learning_rate": 1.8695778325837808e-05, + "loss": 0.0527, + "step": 1898 + }, + { + "epoch": 0.2058759757155247, + "grad_norm": 2.9575371742248535, + "learning_rate": 1.869505295227042e-05, + "loss": 0.0464, + "step": 1899 + }, + { + "epoch": 0.2059843885516045, + "grad_norm": 1.820959210395813, + "learning_rate": 1.8694327578703032e-05, + "loss": 0.0099, + "step": 1900 + }, + { + "epoch": 0.2060928013876843, + "grad_norm": 4.641014575958252, + "learning_rate": 1.8693602205135644e-05, + "loss": 0.0973, + "step": 1901 + }, + { + "epoch": 0.20620121422376408, + "grad_norm": 4.912262916564941, + "learning_rate": 1.869287683156826e-05, + "loss": 0.0191, + "step": 1902 + }, + { + "epoch": 0.20630962705984388, + "grad_norm": 2.8701088428497314, + "learning_rate": 1.8692151458000872e-05, + "loss": 0.0746, + "step": 1903 + }, + { + "epoch": 0.20641803989592367, + "grad_norm": 6.419967174530029, + "learning_rate": 1.8691426084433484e-05, + "loss": 0.0411, + "step": 1904 + }, + { + "epoch": 0.20652645273200346, + "grad_norm": 2.7655487060546875, + "learning_rate": 1.8690700710866096e-05, + "loss": 0.0305, + "step": 1905 + }, + { + "epoch": 0.20663486556808325, + "grad_norm": 2.7806143760681152, + "learning_rate": 1.868997533729871e-05, + "loss": 0.0436, + "step": 1906 + }, + { + "epoch": 0.20674327840416307, + "grad_norm": 3.633998155593872, + "learning_rate": 1.8689249963731324e-05, + "loss": 0.0085, + "step": 1907 + }, + { + "epoch": 0.20685169124024286, + "grad_norm": 3.6797947883605957, + "learning_rate": 1.8688524590163936e-05, + "loss": 0.0232, + "step": 1908 + }, + { + "epoch": 0.20696010407632265, + "grad_norm": 2.96687650680542, + "learning_rate": 1.8687799216596548e-05, + "loss": 0.0783, + "step": 1909 + }, + { + "epoch": 0.20706851691240244, + "grad_norm": 3.1921918392181396, + "learning_rate": 1.8687073843029164e-05, + "loss": 0.0545, + "step": 1910 + }, + { + "epoch": 0.20717692974848223, + "grad_norm": 5.884296894073486, + "learning_rate": 1.8686348469461776e-05, + "loss": 0.0277, + "step": 1911 + }, + { + "epoch": 0.20728534258456202, + "grad_norm": 2.016709327697754, + "learning_rate": 1.8685623095894388e-05, + "loss": 0.027, + "step": 1912 + }, + { + "epoch": 0.2073937554206418, + "grad_norm": 4.85611629486084, + "learning_rate": 1.8684897722327e-05, + "loss": 0.049, + "step": 1913 + }, + { + "epoch": 0.2075021682567216, + "grad_norm": 2.6095259189605713, + "learning_rate": 1.8684172348759612e-05, + "loss": 0.0167, + "step": 1914 + }, + { + "epoch": 0.2076105810928014, + "grad_norm": 2.5030972957611084, + "learning_rate": 1.8683446975192224e-05, + "loss": 0.0365, + "step": 1915 + }, + { + "epoch": 0.20771899392888118, + "grad_norm": 4.356866836547852, + "learning_rate": 1.8682721601624836e-05, + "loss": 0.0624, + "step": 1916 + }, + { + "epoch": 0.20782740676496098, + "grad_norm": 4.080116271972656, + "learning_rate": 1.8681996228057452e-05, + "loss": 0.0424, + "step": 1917 + }, + { + "epoch": 0.20793581960104077, + "grad_norm": 2.3239901065826416, + "learning_rate": 1.8681270854490064e-05, + "loss": 0.0363, + "step": 1918 + }, + { + "epoch": 0.20804423243712056, + "grad_norm": 2.500711679458618, + "learning_rate": 1.8680545480922676e-05, + "loss": 0.0231, + "step": 1919 + }, + { + "epoch": 0.20815264527320035, + "grad_norm": 2.3641116619110107, + "learning_rate": 1.867982010735529e-05, + "loss": 0.0289, + "step": 1920 + }, + { + "epoch": 0.20826105810928014, + "grad_norm": 3.1192688941955566, + "learning_rate": 1.86790947337879e-05, + "loss": 0.027, + "step": 1921 + }, + { + "epoch": 0.20836947094535993, + "grad_norm": 1.8687264919281006, + "learning_rate": 1.8678369360220513e-05, + "loss": 0.0412, + "step": 1922 + }, + { + "epoch": 0.20847788378143972, + "grad_norm": 3.9754831790924072, + "learning_rate": 1.8677643986653128e-05, + "loss": 0.0389, + "step": 1923 + }, + { + "epoch": 0.2085862966175195, + "grad_norm": 3.669999599456787, + "learning_rate": 1.867691861308574e-05, + "loss": 0.0207, + "step": 1924 + }, + { + "epoch": 0.2086947094535993, + "grad_norm": 1.1722162961959839, + "learning_rate": 1.8676193239518356e-05, + "loss": 0.0296, + "step": 1925 + }, + { + "epoch": 0.2088031222896791, + "grad_norm": 3.611643075942993, + "learning_rate": 1.8675467865950968e-05, + "loss": 0.0448, + "step": 1926 + }, + { + "epoch": 0.20891153512575888, + "grad_norm": 3.2066054344177246, + "learning_rate": 1.867474249238358e-05, + "loss": 0.0699, + "step": 1927 + }, + { + "epoch": 0.20901994796183868, + "grad_norm": 2.6844985485076904, + "learning_rate": 1.8674017118816192e-05, + "loss": 0.0849, + "step": 1928 + }, + { + "epoch": 0.20912836079791847, + "grad_norm": 6.611279487609863, + "learning_rate": 1.8673291745248805e-05, + "loss": 0.0466, + "step": 1929 + }, + { + "epoch": 0.20923677363399826, + "grad_norm": 6.036165237426758, + "learning_rate": 1.8672566371681417e-05, + "loss": 0.0441, + "step": 1930 + }, + { + "epoch": 0.20934518647007805, + "grad_norm": 5.36448860168457, + "learning_rate": 1.867184099811403e-05, + "loss": 0.0648, + "step": 1931 + }, + { + "epoch": 0.20945359930615784, + "grad_norm": 1.9858298301696777, + "learning_rate": 1.8671115624546644e-05, + "loss": 0.0457, + "step": 1932 + }, + { + "epoch": 0.20956201214223763, + "grad_norm": 1.3628519773483276, + "learning_rate": 1.8670390250979256e-05, + "loss": 0.0216, + "step": 1933 + }, + { + "epoch": 0.20967042497831742, + "grad_norm": 3.826439380645752, + "learning_rate": 1.866966487741187e-05, + "loss": 0.0898, + "step": 1934 + }, + { + "epoch": 0.2097788378143972, + "grad_norm": 0.24337492883205414, + "learning_rate": 1.866893950384448e-05, + "loss": 0.0031, + "step": 1935 + }, + { + "epoch": 0.209887250650477, + "grad_norm": 1.0679067373275757, + "learning_rate": 1.8668214130277093e-05, + "loss": 0.0177, + "step": 1936 + }, + { + "epoch": 0.20999566348655682, + "grad_norm": 2.2936205863952637, + "learning_rate": 1.8667488756709705e-05, + "loss": 0.0402, + "step": 1937 + }, + { + "epoch": 0.2101040763226366, + "grad_norm": 4.4517645835876465, + "learning_rate": 1.866676338314232e-05, + "loss": 0.0517, + "step": 1938 + }, + { + "epoch": 0.2102124891587164, + "grad_norm": 2.4304709434509277, + "learning_rate": 1.8666038009574933e-05, + "loss": 0.0456, + "step": 1939 + }, + { + "epoch": 0.2103209019947962, + "grad_norm": 3.0469889640808105, + "learning_rate": 1.8665312636007545e-05, + "loss": 0.0449, + "step": 1940 + }, + { + "epoch": 0.21042931483087599, + "grad_norm": 3.3002476692199707, + "learning_rate": 1.866458726244016e-05, + "loss": 0.06, + "step": 1941 + }, + { + "epoch": 0.21053772766695578, + "grad_norm": 3.8224587440490723, + "learning_rate": 1.8663861888872773e-05, + "loss": 0.0386, + "step": 1942 + }, + { + "epoch": 0.21064614050303557, + "grad_norm": 4.10772705078125, + "learning_rate": 1.8663136515305385e-05, + "loss": 0.0688, + "step": 1943 + }, + { + "epoch": 0.21075455333911536, + "grad_norm": 3.8953349590301514, + "learning_rate": 1.8662411141737997e-05, + "loss": 0.0852, + "step": 1944 + }, + { + "epoch": 0.21086296617519515, + "grad_norm": 2.5965466499328613, + "learning_rate": 1.866168576817061e-05, + "loss": 0.0424, + "step": 1945 + }, + { + "epoch": 0.21097137901127494, + "grad_norm": 1.255903720855713, + "learning_rate": 1.866096039460322e-05, + "loss": 0.0099, + "step": 1946 + }, + { + "epoch": 0.21107979184735473, + "grad_norm": 4.72607421875, + "learning_rate": 1.8660235021035837e-05, + "loss": 0.084, + "step": 1947 + }, + { + "epoch": 0.21118820468343452, + "grad_norm": 1.3825241327285767, + "learning_rate": 1.865950964746845e-05, + "loss": 0.0401, + "step": 1948 + }, + { + "epoch": 0.2112966175195143, + "grad_norm": 6.333207607269287, + "learning_rate": 1.865878427390106e-05, + "loss": 0.0383, + "step": 1949 + }, + { + "epoch": 0.2114050303555941, + "grad_norm": 3.2201547622680664, + "learning_rate": 1.8658058900333673e-05, + "loss": 0.0581, + "step": 1950 + }, + { + "epoch": 0.2115134431916739, + "grad_norm": 2.9949347972869873, + "learning_rate": 1.8657333526766285e-05, + "loss": 0.0614, + "step": 1951 + }, + { + "epoch": 0.21162185602775369, + "grad_norm": 2.706915855407715, + "learning_rate": 1.8656608153198897e-05, + "loss": 0.0948, + "step": 1952 + }, + { + "epoch": 0.21173026886383348, + "grad_norm": 3.177971601486206, + "learning_rate": 1.865588277963151e-05, + "loss": 0.0403, + "step": 1953 + }, + { + "epoch": 0.21183868169991327, + "grad_norm": 2.443223237991333, + "learning_rate": 1.8655157406064125e-05, + "loss": 0.029, + "step": 1954 + }, + { + "epoch": 0.21194709453599306, + "grad_norm": 4.293915271759033, + "learning_rate": 1.8654432032496737e-05, + "loss": 0.0595, + "step": 1955 + }, + { + "epoch": 0.21205550737207285, + "grad_norm": 3.9469189643859863, + "learning_rate": 1.865370665892935e-05, + "loss": 0.0483, + "step": 1956 + }, + { + "epoch": 0.21216392020815264, + "grad_norm": 2.814251661300659, + "learning_rate": 1.865298128536196e-05, + "loss": 0.0543, + "step": 1957 + }, + { + "epoch": 0.21227233304423243, + "grad_norm": 4.88180685043335, + "learning_rate": 1.8652255911794577e-05, + "loss": 0.0294, + "step": 1958 + }, + { + "epoch": 0.21238074588031222, + "grad_norm": 1.296428918838501, + "learning_rate": 1.865153053822719e-05, + "loss": 0.0217, + "step": 1959 + }, + { + "epoch": 0.212489158716392, + "grad_norm": 3.247399091720581, + "learning_rate": 1.86508051646598e-05, + "loss": 0.056, + "step": 1960 + }, + { + "epoch": 0.2125975715524718, + "grad_norm": 2.1911556720733643, + "learning_rate": 1.8650079791092413e-05, + "loss": 0.031, + "step": 1961 + }, + { + "epoch": 0.2127059843885516, + "grad_norm": 2.2771120071411133, + "learning_rate": 1.864935441752503e-05, + "loss": 0.0238, + "step": 1962 + }, + { + "epoch": 0.21281439722463139, + "grad_norm": 2.515523910522461, + "learning_rate": 1.864862904395764e-05, + "loss": 0.0512, + "step": 1963 + }, + { + "epoch": 0.21292281006071118, + "grad_norm": 3.134568452835083, + "learning_rate": 1.8647903670390253e-05, + "loss": 0.0481, + "step": 1964 + }, + { + "epoch": 0.21303122289679097, + "grad_norm": 6.4638261795043945, + "learning_rate": 1.8647178296822865e-05, + "loss": 0.0873, + "step": 1965 + }, + { + "epoch": 0.21313963573287076, + "grad_norm": 1.8622722625732422, + "learning_rate": 1.8646452923255478e-05, + "loss": 0.0386, + "step": 1966 + }, + { + "epoch": 0.21324804856895058, + "grad_norm": 5.478961944580078, + "learning_rate": 1.864572754968809e-05, + "loss": 0.1142, + "step": 1967 + }, + { + "epoch": 0.21335646140503037, + "grad_norm": 3.9981541633605957, + "learning_rate": 1.8645002176120702e-05, + "loss": 0.0523, + "step": 1968 + }, + { + "epoch": 0.21346487424111016, + "grad_norm": 5.842113018035889, + "learning_rate": 1.8644276802553317e-05, + "loss": 0.0709, + "step": 1969 + }, + { + "epoch": 0.21357328707718995, + "grad_norm": 3.793621063232422, + "learning_rate": 1.864355142898593e-05, + "loss": 0.0922, + "step": 1970 + }, + { + "epoch": 0.21368169991326974, + "grad_norm": 4.428802967071533, + "learning_rate": 1.864282605541854e-05, + "loss": 0.0288, + "step": 1971 + }, + { + "epoch": 0.21379011274934953, + "grad_norm": 2.5223522186279297, + "learning_rate": 1.8642100681851154e-05, + "loss": 0.0612, + "step": 1972 + }, + { + "epoch": 0.21389852558542932, + "grad_norm": 5.794287204742432, + "learning_rate": 1.8641375308283766e-05, + "loss": 0.0524, + "step": 1973 + }, + { + "epoch": 0.2140069384215091, + "grad_norm": 3.0250678062438965, + "learning_rate": 1.8640649934716378e-05, + "loss": 0.0607, + "step": 1974 + }, + { + "epoch": 0.2141153512575889, + "grad_norm": 5.254593372344971, + "learning_rate": 1.8639924561148994e-05, + "loss": 0.0488, + "step": 1975 + }, + { + "epoch": 0.2142237640936687, + "grad_norm": 1.8281348943710327, + "learning_rate": 1.8639199187581606e-05, + "loss": 0.0212, + "step": 1976 + }, + { + "epoch": 0.21433217692974849, + "grad_norm": 3.827528715133667, + "learning_rate": 1.863847381401422e-05, + "loss": 0.052, + "step": 1977 + }, + { + "epoch": 0.21444058976582828, + "grad_norm": 3.8818907737731934, + "learning_rate": 1.8637748440446833e-05, + "loss": 0.0551, + "step": 1978 + }, + { + "epoch": 0.21454900260190807, + "grad_norm": 2.8982789516448975, + "learning_rate": 1.8637023066879446e-05, + "loss": 0.0307, + "step": 1979 + }, + { + "epoch": 0.21465741543798786, + "grad_norm": 1.5206220149993896, + "learning_rate": 1.8636297693312058e-05, + "loss": 0.0108, + "step": 1980 + }, + { + "epoch": 0.21476582827406765, + "grad_norm": 3.958564519882202, + "learning_rate": 1.863557231974467e-05, + "loss": 0.0532, + "step": 1981 + }, + { + "epoch": 0.21487424111014744, + "grad_norm": 3.3957979679107666, + "learning_rate": 1.8634846946177282e-05, + "loss": 0.0672, + "step": 1982 + }, + { + "epoch": 0.21498265394622723, + "grad_norm": 4.52908992767334, + "learning_rate": 1.8634121572609894e-05, + "loss": 0.0731, + "step": 1983 + }, + { + "epoch": 0.21509106678230702, + "grad_norm": 1.953094482421875, + "learning_rate": 1.863339619904251e-05, + "loss": 0.0236, + "step": 1984 + }, + { + "epoch": 0.2151994796183868, + "grad_norm": 4.53131628036499, + "learning_rate": 1.8632670825475122e-05, + "loss": 0.0373, + "step": 1985 + }, + { + "epoch": 0.2153078924544666, + "grad_norm": 6.8083415031433105, + "learning_rate": 1.8631945451907734e-05, + "loss": 0.0778, + "step": 1986 + }, + { + "epoch": 0.2154163052905464, + "grad_norm": 3.2680375576019287, + "learning_rate": 1.8631220078340346e-05, + "loss": 0.0758, + "step": 1987 + }, + { + "epoch": 0.21552471812662619, + "grad_norm": 4.090797424316406, + "learning_rate": 1.8630494704772958e-05, + "loss": 0.0675, + "step": 1988 + }, + { + "epoch": 0.21563313096270598, + "grad_norm": 2.4796693325042725, + "learning_rate": 1.862976933120557e-05, + "loss": 0.0352, + "step": 1989 + }, + { + "epoch": 0.21574154379878577, + "grad_norm": 3.390282154083252, + "learning_rate": 1.8629043957638186e-05, + "loss": 0.049, + "step": 1990 + }, + { + "epoch": 0.21584995663486556, + "grad_norm": 5.113428115844727, + "learning_rate": 1.8628318584070798e-05, + "loss": 0.0494, + "step": 1991 + }, + { + "epoch": 0.21595836947094535, + "grad_norm": 5.374655723571777, + "learning_rate": 1.862759321050341e-05, + "loss": 0.0466, + "step": 1992 + }, + { + "epoch": 0.21606678230702514, + "grad_norm": 3.0746989250183105, + "learning_rate": 1.8626867836936026e-05, + "loss": 0.03, + "step": 1993 + }, + { + "epoch": 0.21617519514310493, + "grad_norm": 0.9672751426696777, + "learning_rate": 1.8626142463368638e-05, + "loss": 0.0116, + "step": 1994 + }, + { + "epoch": 0.21628360797918472, + "grad_norm": 2.043954610824585, + "learning_rate": 1.862541708980125e-05, + "loss": 0.026, + "step": 1995 + }, + { + "epoch": 0.21639202081526454, + "grad_norm": 2.714475154876709, + "learning_rate": 1.8624691716233862e-05, + "loss": 0.0264, + "step": 1996 + }, + { + "epoch": 0.21650043365134433, + "grad_norm": 2.726105213165283, + "learning_rate": 1.8623966342666474e-05, + "loss": 0.0275, + "step": 1997 + }, + { + "epoch": 0.21660884648742412, + "grad_norm": 6.894297122955322, + "learning_rate": 1.8623240969099086e-05, + "loss": 0.0814, + "step": 1998 + }, + { + "epoch": 0.2167172593235039, + "grad_norm": 1.7545040845870972, + "learning_rate": 1.8622515595531702e-05, + "loss": 0.0256, + "step": 1999 + }, + { + "epoch": 0.2168256721595837, + "grad_norm": 3.311152696609497, + "learning_rate": 1.8621790221964314e-05, + "loss": 0.0465, + "step": 2000 + }, + { + "epoch": 0.2169340849956635, + "grad_norm": 6.456691741943359, + "learning_rate": 1.8621064848396926e-05, + "loss": 0.1342, + "step": 2001 + }, + { + "epoch": 0.2170424978317433, + "grad_norm": 5.787269115447998, + "learning_rate": 1.862033947482954e-05, + "loss": 0.0247, + "step": 2002 + }, + { + "epoch": 0.21715091066782308, + "grad_norm": 6.505336761474609, + "learning_rate": 1.861961410126215e-05, + "loss": 0.0406, + "step": 2003 + }, + { + "epoch": 0.21725932350390287, + "grad_norm": 4.184741020202637, + "learning_rate": 1.8618888727694763e-05, + "loss": 0.0504, + "step": 2004 + }, + { + "epoch": 0.21736773633998266, + "grad_norm": 4.996067047119141, + "learning_rate": 1.8618163354127375e-05, + "loss": 0.0324, + "step": 2005 + }, + { + "epoch": 0.21747614917606245, + "grad_norm": 2.5768091678619385, + "learning_rate": 1.861743798055999e-05, + "loss": 0.0332, + "step": 2006 + }, + { + "epoch": 0.21758456201214224, + "grad_norm": 7.272979259490967, + "learning_rate": 1.8616712606992603e-05, + "loss": 0.0756, + "step": 2007 + }, + { + "epoch": 0.21769297484822203, + "grad_norm": 5.70379114151001, + "learning_rate": 1.8615987233425215e-05, + "loss": 0.0534, + "step": 2008 + }, + { + "epoch": 0.21780138768430182, + "grad_norm": 2.5838468074798584, + "learning_rate": 1.861526185985783e-05, + "loss": 0.027, + "step": 2009 + }, + { + "epoch": 0.2179098005203816, + "grad_norm": 4.397194862365723, + "learning_rate": 1.8614536486290442e-05, + "loss": 0.0472, + "step": 2010 + }, + { + "epoch": 0.2180182133564614, + "grad_norm": 4.262328147888184, + "learning_rate": 1.8613811112723054e-05, + "loss": 0.0327, + "step": 2011 + }, + { + "epoch": 0.2181266261925412, + "grad_norm": 3.9662508964538574, + "learning_rate": 1.8613085739155667e-05, + "loss": 0.0491, + "step": 2012 + }, + { + "epoch": 0.218235039028621, + "grad_norm": 6.4381327629089355, + "learning_rate": 1.861236036558828e-05, + "loss": 0.072, + "step": 2013 + }, + { + "epoch": 0.21834345186470078, + "grad_norm": 7.037211894989014, + "learning_rate": 1.8611634992020894e-05, + "loss": 0.0977, + "step": 2014 + }, + { + "epoch": 0.21845186470078057, + "grad_norm": 2.6934518814086914, + "learning_rate": 1.8610909618453506e-05, + "loss": 0.0325, + "step": 2015 + }, + { + "epoch": 0.21856027753686036, + "grad_norm": 3.0950307846069336, + "learning_rate": 1.861018424488612e-05, + "loss": 0.0371, + "step": 2016 + }, + { + "epoch": 0.21866869037294015, + "grad_norm": 2.8593878746032715, + "learning_rate": 1.860945887131873e-05, + "loss": 0.0537, + "step": 2017 + }, + { + "epoch": 0.21877710320901994, + "grad_norm": 2.8225350379943848, + "learning_rate": 1.8608733497751343e-05, + "loss": 0.0237, + "step": 2018 + }, + { + "epoch": 0.21888551604509973, + "grad_norm": 3.5750365257263184, + "learning_rate": 1.8608008124183955e-05, + "loss": 0.0416, + "step": 2019 + }, + { + "epoch": 0.21899392888117952, + "grad_norm": 2.6419548988342285, + "learning_rate": 1.8607282750616567e-05, + "loss": 0.0236, + "step": 2020 + }, + { + "epoch": 0.2191023417172593, + "grad_norm": 3.2869558334350586, + "learning_rate": 1.8606557377049183e-05, + "loss": 0.038, + "step": 2021 + }, + { + "epoch": 0.2192107545533391, + "grad_norm": 4.2810378074646, + "learning_rate": 1.8605832003481795e-05, + "loss": 0.0966, + "step": 2022 + }, + { + "epoch": 0.2193191673894189, + "grad_norm": 1.6446171998977661, + "learning_rate": 1.8605106629914407e-05, + "loss": 0.0474, + "step": 2023 + }, + { + "epoch": 0.2194275802254987, + "grad_norm": 1.6577690839767456, + "learning_rate": 1.860438125634702e-05, + "loss": 0.0316, + "step": 2024 + }, + { + "epoch": 0.21953599306157848, + "grad_norm": 3.2030680179595947, + "learning_rate": 1.860365588277963e-05, + "loss": 0.0317, + "step": 2025 + }, + { + "epoch": 0.2196444058976583, + "grad_norm": 1.1483972072601318, + "learning_rate": 1.8602930509212243e-05, + "loss": 0.0099, + "step": 2026 + }, + { + "epoch": 0.2197528187337381, + "grad_norm": 1.5214630365371704, + "learning_rate": 1.860220513564486e-05, + "loss": 0.0168, + "step": 2027 + }, + { + "epoch": 0.21986123156981788, + "grad_norm": 4.1440300941467285, + "learning_rate": 1.860147976207747e-05, + "loss": 0.0357, + "step": 2028 + }, + { + "epoch": 0.21996964440589767, + "grad_norm": 6.868283748626709, + "learning_rate": 1.8600754388510087e-05, + "loss": 0.0493, + "step": 2029 + }, + { + "epoch": 0.22007805724197746, + "grad_norm": 2.6210732460021973, + "learning_rate": 1.86000290149427e-05, + "loss": 0.0652, + "step": 2030 + }, + { + "epoch": 0.22018647007805725, + "grad_norm": 4.778876304626465, + "learning_rate": 1.859930364137531e-05, + "loss": 0.0591, + "step": 2031 + }, + { + "epoch": 0.22029488291413704, + "grad_norm": 3.0951478481292725, + "learning_rate": 1.8598578267807923e-05, + "loss": 0.036, + "step": 2032 + }, + { + "epoch": 0.22040329575021683, + "grad_norm": 0.8915024399757385, + "learning_rate": 1.8597852894240535e-05, + "loss": 0.0207, + "step": 2033 + }, + { + "epoch": 0.22051170858629662, + "grad_norm": 2.8076250553131104, + "learning_rate": 1.8597127520673147e-05, + "loss": 0.033, + "step": 2034 + }, + { + "epoch": 0.22062012142237641, + "grad_norm": 2.6715571880340576, + "learning_rate": 1.859640214710576e-05, + "loss": 0.0247, + "step": 2035 + }, + { + "epoch": 0.2207285342584562, + "grad_norm": 2.643564462661743, + "learning_rate": 1.8595676773538375e-05, + "loss": 0.047, + "step": 2036 + }, + { + "epoch": 0.220836947094536, + "grad_norm": 1.2828991413116455, + "learning_rate": 1.8594951399970987e-05, + "loss": 0.0456, + "step": 2037 + }, + { + "epoch": 0.2209453599306158, + "grad_norm": 2.4046175479888916, + "learning_rate": 1.85942260264036e-05, + "loss": 0.0324, + "step": 2038 + }, + { + "epoch": 0.22105377276669558, + "grad_norm": 2.3877413272857666, + "learning_rate": 1.859350065283621e-05, + "loss": 0.0304, + "step": 2039 + }, + { + "epoch": 0.22116218560277537, + "grad_norm": 6.567047119140625, + "learning_rate": 1.8592775279268824e-05, + "loss": 0.0567, + "step": 2040 + }, + { + "epoch": 0.22127059843885516, + "grad_norm": 2.3100178241729736, + "learning_rate": 1.8592049905701436e-05, + "loss": 0.0263, + "step": 2041 + }, + { + "epoch": 0.22137901127493495, + "grad_norm": 6.221439361572266, + "learning_rate": 1.859132453213405e-05, + "loss": 0.0824, + "step": 2042 + }, + { + "epoch": 0.22148742411101474, + "grad_norm": 2.85713791847229, + "learning_rate": 1.8590599158566663e-05, + "loss": 0.0315, + "step": 2043 + }, + { + "epoch": 0.22159583694709453, + "grad_norm": 2.9790706634521484, + "learning_rate": 1.8589873784999276e-05, + "loss": 0.0497, + "step": 2044 + }, + { + "epoch": 0.22170424978317432, + "grad_norm": 5.423086166381836, + "learning_rate": 1.858914841143189e-05, + "loss": 0.0787, + "step": 2045 + }, + { + "epoch": 0.22181266261925411, + "grad_norm": 3.6851701736450195, + "learning_rate": 1.8588423037864503e-05, + "loss": 0.0451, + "step": 2046 + }, + { + "epoch": 0.2219210754553339, + "grad_norm": 4.5308756828308105, + "learning_rate": 1.8587697664297115e-05, + "loss": 0.1317, + "step": 2047 + }, + { + "epoch": 0.2220294882914137, + "grad_norm": 2.458209991455078, + "learning_rate": 1.8586972290729728e-05, + "loss": 0.021, + "step": 2048 + }, + { + "epoch": 0.2221379011274935, + "grad_norm": 4.731989860534668, + "learning_rate": 1.858624691716234e-05, + "loss": 0.0168, + "step": 2049 + }, + { + "epoch": 0.22224631396357328, + "grad_norm": 3.0770599842071533, + "learning_rate": 1.8585521543594952e-05, + "loss": 0.0259, + "step": 2050 + }, + { + "epoch": 0.22235472679965307, + "grad_norm": 3.871412992477417, + "learning_rate": 1.8584796170027567e-05, + "loss": 0.041, + "step": 2051 + }, + { + "epoch": 0.22246313963573286, + "grad_norm": 2.3667919635772705, + "learning_rate": 1.858407079646018e-05, + "loss": 0.0354, + "step": 2052 + }, + { + "epoch": 0.22257155247181265, + "grad_norm": 2.5179359912872314, + "learning_rate": 1.858334542289279e-05, + "loss": 0.0333, + "step": 2053 + }, + { + "epoch": 0.22267996530789244, + "grad_norm": 4.471822261810303, + "learning_rate": 1.8582620049325404e-05, + "loss": 0.0974, + "step": 2054 + }, + { + "epoch": 0.22278837814397226, + "grad_norm": 3.0718882083892822, + "learning_rate": 1.8581894675758016e-05, + "loss": 0.0252, + "step": 2055 + }, + { + "epoch": 0.22289679098005205, + "grad_norm": 3.63629412651062, + "learning_rate": 1.8581169302190628e-05, + "loss": 0.0149, + "step": 2056 + }, + { + "epoch": 0.22300520381613184, + "grad_norm": 10.211212158203125, + "learning_rate": 1.858044392862324e-05, + "loss": 0.0849, + "step": 2057 + }, + { + "epoch": 0.22311361665221163, + "grad_norm": 2.597835063934326, + "learning_rate": 1.8579718555055856e-05, + "loss": 0.0484, + "step": 2058 + }, + { + "epoch": 0.22322202948829142, + "grad_norm": 4.884099960327148, + "learning_rate": 1.8578993181488468e-05, + "loss": 0.0244, + "step": 2059 + }, + { + "epoch": 0.22333044232437121, + "grad_norm": 2.2137339115142822, + "learning_rate": 1.857826780792108e-05, + "loss": 0.0296, + "step": 2060 + }, + { + "epoch": 0.223438855160451, + "grad_norm": 3.89328670501709, + "learning_rate": 1.8577542434353696e-05, + "loss": 0.0387, + "step": 2061 + }, + { + "epoch": 0.2235472679965308, + "grad_norm": 2.7492709159851074, + "learning_rate": 1.8576817060786308e-05, + "loss": 0.0974, + "step": 2062 + }, + { + "epoch": 0.2236556808326106, + "grad_norm": 1.0972486734390259, + "learning_rate": 1.857609168721892e-05, + "loss": 0.0101, + "step": 2063 + }, + { + "epoch": 0.22376409366869038, + "grad_norm": 1.7045024633407593, + "learning_rate": 1.8575366313651532e-05, + "loss": 0.0326, + "step": 2064 + }, + { + "epoch": 0.22387250650477017, + "grad_norm": 1.5828524827957153, + "learning_rate": 1.8574640940084144e-05, + "loss": 0.0245, + "step": 2065 + }, + { + "epoch": 0.22398091934084996, + "grad_norm": 8.398812294006348, + "learning_rate": 1.857391556651676e-05, + "loss": 0.0557, + "step": 2066 + }, + { + "epoch": 0.22408933217692975, + "grad_norm": 6.4078192710876465, + "learning_rate": 1.8573190192949372e-05, + "loss": 0.1063, + "step": 2067 + }, + { + "epoch": 0.22419774501300954, + "grad_norm": 2.89272141456604, + "learning_rate": 1.8572464819381984e-05, + "loss": 0.044, + "step": 2068 + }, + { + "epoch": 0.22430615784908933, + "grad_norm": 4.775213241577148, + "learning_rate": 1.8571739445814596e-05, + "loss": 0.0914, + "step": 2069 + }, + { + "epoch": 0.22441457068516912, + "grad_norm": 5.192051887512207, + "learning_rate": 1.8571014072247208e-05, + "loss": 0.0886, + "step": 2070 + }, + { + "epoch": 0.22452298352124891, + "grad_norm": 2.241558074951172, + "learning_rate": 1.857028869867982e-05, + "loss": 0.0302, + "step": 2071 + }, + { + "epoch": 0.2246313963573287, + "grad_norm": 4.123575210571289, + "learning_rate": 1.8569563325112433e-05, + "loss": 0.0583, + "step": 2072 + }, + { + "epoch": 0.2247398091934085, + "grad_norm": 4.421757221221924, + "learning_rate": 1.8568837951545048e-05, + "loss": 0.0686, + "step": 2073 + }, + { + "epoch": 0.2248482220294883, + "grad_norm": 2.293928861618042, + "learning_rate": 1.856811257797766e-05, + "loss": 0.0213, + "step": 2074 + }, + { + "epoch": 0.22495663486556808, + "grad_norm": 4.698619842529297, + "learning_rate": 1.8567387204410272e-05, + "loss": 0.0926, + "step": 2075 + }, + { + "epoch": 0.22506504770164787, + "grad_norm": 6.128471374511719, + "learning_rate": 1.8566661830842884e-05, + "loss": 0.0782, + "step": 2076 + }, + { + "epoch": 0.22517346053772766, + "grad_norm": 1.7482184171676636, + "learning_rate": 1.8565936457275497e-05, + "loss": 0.0313, + "step": 2077 + }, + { + "epoch": 0.22528187337380745, + "grad_norm": 6.501456260681152, + "learning_rate": 1.856521108370811e-05, + "loss": 0.0856, + "step": 2078 + }, + { + "epoch": 0.22539028620988724, + "grad_norm": 2.3651108741760254, + "learning_rate": 1.8564485710140724e-05, + "loss": 0.0445, + "step": 2079 + }, + { + "epoch": 0.22549869904596703, + "grad_norm": 6.274560928344727, + "learning_rate": 1.8563760336573336e-05, + "loss": 0.0758, + "step": 2080 + }, + { + "epoch": 0.22560711188204682, + "grad_norm": 4.913519859313965, + "learning_rate": 1.8563034963005952e-05, + "loss": 0.0752, + "step": 2081 + }, + { + "epoch": 0.22571552471812661, + "grad_norm": 4.2436723709106445, + "learning_rate": 1.8562309589438564e-05, + "loss": 0.0328, + "step": 2082 + }, + { + "epoch": 0.2258239375542064, + "grad_norm": 1.9411355257034302, + "learning_rate": 1.8561584215871176e-05, + "loss": 0.034, + "step": 2083 + }, + { + "epoch": 0.2259323503902862, + "grad_norm": 2.067031145095825, + "learning_rate": 1.856085884230379e-05, + "loss": 0.0265, + "step": 2084 + }, + { + "epoch": 0.22604076322636601, + "grad_norm": 3.4824156761169434, + "learning_rate": 1.85601334687364e-05, + "loss": 0.0779, + "step": 2085 + }, + { + "epoch": 0.2261491760624458, + "grad_norm": 2.748936653137207, + "learning_rate": 1.8559408095169013e-05, + "loss": 0.0192, + "step": 2086 + }, + { + "epoch": 0.2262575888985256, + "grad_norm": 2.7048778533935547, + "learning_rate": 1.8558682721601625e-05, + "loss": 0.0387, + "step": 2087 + }, + { + "epoch": 0.2263660017346054, + "grad_norm": 1.4304436445236206, + "learning_rate": 1.855795734803424e-05, + "loss": 0.0302, + "step": 2088 + }, + { + "epoch": 0.22647441457068518, + "grad_norm": 4.00627326965332, + "learning_rate": 1.8557231974466852e-05, + "loss": 0.0325, + "step": 2089 + }, + { + "epoch": 0.22658282740676497, + "grad_norm": 3.9087963104248047, + "learning_rate": 1.8556506600899465e-05, + "loss": 0.0505, + "step": 2090 + }, + { + "epoch": 0.22669124024284476, + "grad_norm": 3.5343058109283447, + "learning_rate": 1.8555781227332077e-05, + "loss": 0.0333, + "step": 2091 + }, + { + "epoch": 0.22679965307892455, + "grad_norm": 2.4643452167510986, + "learning_rate": 1.855505585376469e-05, + "loss": 0.047, + "step": 2092 + }, + { + "epoch": 0.22690806591500434, + "grad_norm": 2.949193239212036, + "learning_rate": 1.85543304801973e-05, + "loss": 0.0433, + "step": 2093 + }, + { + "epoch": 0.22701647875108413, + "grad_norm": 5.099905014038086, + "learning_rate": 1.8553605106629917e-05, + "loss": 0.0803, + "step": 2094 + }, + { + "epoch": 0.22712489158716392, + "grad_norm": 1.567813515663147, + "learning_rate": 1.855287973306253e-05, + "loss": 0.0231, + "step": 2095 + }, + { + "epoch": 0.22723330442324371, + "grad_norm": 4.601634979248047, + "learning_rate": 1.855215435949514e-05, + "loss": 0.0612, + "step": 2096 + }, + { + "epoch": 0.2273417172593235, + "grad_norm": 2.7646725177764893, + "learning_rate": 1.8551428985927756e-05, + "loss": 0.0365, + "step": 2097 + }, + { + "epoch": 0.2274501300954033, + "grad_norm": 3.543720006942749, + "learning_rate": 1.855070361236037e-05, + "loss": 0.1467, + "step": 2098 + }, + { + "epoch": 0.2275585429314831, + "grad_norm": 3.1206765174865723, + "learning_rate": 1.854997823879298e-05, + "loss": 0.0539, + "step": 2099 + }, + { + "epoch": 0.22766695576756288, + "grad_norm": 2.976992607116699, + "learning_rate": 1.8549252865225593e-05, + "loss": 0.0519, + "step": 2100 + }, + { + "epoch": 0.22777536860364267, + "grad_norm": 5.9177565574646, + "learning_rate": 1.8548527491658205e-05, + "loss": 0.0763, + "step": 2101 + }, + { + "epoch": 0.22788378143972246, + "grad_norm": 6.528652191162109, + "learning_rate": 1.8547802118090817e-05, + "loss": 0.0596, + "step": 2102 + }, + { + "epoch": 0.22799219427580225, + "grad_norm": 4.583979606628418, + "learning_rate": 1.8547076744523433e-05, + "loss": 0.0688, + "step": 2103 + }, + { + "epoch": 0.22810060711188204, + "grad_norm": 2.281949043273926, + "learning_rate": 1.8546351370956045e-05, + "loss": 0.046, + "step": 2104 + }, + { + "epoch": 0.22820901994796183, + "grad_norm": 3.464564085006714, + "learning_rate": 1.8545625997388657e-05, + "loss": 0.0461, + "step": 2105 + }, + { + "epoch": 0.22831743278404162, + "grad_norm": 1.9701035022735596, + "learning_rate": 1.854490062382127e-05, + "loss": 0.0371, + "step": 2106 + }, + { + "epoch": 0.22842584562012141, + "grad_norm": 4.303263187408447, + "learning_rate": 1.854417525025388e-05, + "loss": 0.0739, + "step": 2107 + }, + { + "epoch": 0.2285342584562012, + "grad_norm": 3.180046558380127, + "learning_rate": 1.8543449876686493e-05, + "loss": 0.0376, + "step": 2108 + }, + { + "epoch": 0.228642671292281, + "grad_norm": 2.040503978729248, + "learning_rate": 1.8542724503119106e-05, + "loss": 0.0446, + "step": 2109 + }, + { + "epoch": 0.2287510841283608, + "grad_norm": 1.4239156246185303, + "learning_rate": 1.854199912955172e-05, + "loss": 0.0237, + "step": 2110 + }, + { + "epoch": 0.22885949696444058, + "grad_norm": 4.749313831329346, + "learning_rate": 1.8541273755984333e-05, + "loss": 0.0769, + "step": 2111 + }, + { + "epoch": 0.22896790980052037, + "grad_norm": 3.2654800415039062, + "learning_rate": 1.8540548382416945e-05, + "loss": 0.0505, + "step": 2112 + }, + { + "epoch": 0.22907632263660016, + "grad_norm": 1.3715095520019531, + "learning_rate": 1.853982300884956e-05, + "loss": 0.0553, + "step": 2113 + }, + { + "epoch": 0.22918473547267998, + "grad_norm": 3.35121750831604, + "learning_rate": 1.8539097635282173e-05, + "loss": 0.059, + "step": 2114 + }, + { + "epoch": 0.22929314830875977, + "grad_norm": 4.037988185882568, + "learning_rate": 1.8538372261714785e-05, + "loss": 0.0619, + "step": 2115 + }, + { + "epoch": 0.22940156114483956, + "grad_norm": 2.2971744537353516, + "learning_rate": 1.8537646888147397e-05, + "loss": 0.0481, + "step": 2116 + }, + { + "epoch": 0.22950997398091935, + "grad_norm": 3.3536834716796875, + "learning_rate": 1.853692151458001e-05, + "loss": 0.0764, + "step": 2117 + }, + { + "epoch": 0.22961838681699914, + "grad_norm": 2.701514720916748, + "learning_rate": 1.8536196141012625e-05, + "loss": 0.0208, + "step": 2118 + }, + { + "epoch": 0.22972679965307893, + "grad_norm": 2.667290687561035, + "learning_rate": 1.8535470767445237e-05, + "loss": 0.0332, + "step": 2119 + }, + { + "epoch": 0.22983521248915872, + "grad_norm": 3.8280601501464844, + "learning_rate": 1.853474539387785e-05, + "loss": 0.0766, + "step": 2120 + }, + { + "epoch": 0.22994362532523852, + "grad_norm": 7.877985000610352, + "learning_rate": 1.853402002031046e-05, + "loss": 0.0587, + "step": 2121 + }, + { + "epoch": 0.2300520381613183, + "grad_norm": 1.8060085773468018, + "learning_rate": 1.8533294646743074e-05, + "loss": 0.0339, + "step": 2122 + }, + { + "epoch": 0.2301604509973981, + "grad_norm": 2.8129258155822754, + "learning_rate": 1.8532569273175686e-05, + "loss": 0.0548, + "step": 2123 + }, + { + "epoch": 0.2302688638334779, + "grad_norm": 4.199626922607422, + "learning_rate": 1.8531843899608298e-05, + "loss": 0.0588, + "step": 2124 + }, + { + "epoch": 0.23037727666955768, + "grad_norm": 4.803268909454346, + "learning_rate": 1.8531118526040913e-05, + "loss": 0.0594, + "step": 2125 + }, + { + "epoch": 0.23048568950563747, + "grad_norm": 1.2202386856079102, + "learning_rate": 1.8530393152473526e-05, + "loss": 0.0168, + "step": 2126 + }, + { + "epoch": 0.23059410234171726, + "grad_norm": 2.031320810317993, + "learning_rate": 1.8529667778906138e-05, + "loss": 0.0263, + "step": 2127 + }, + { + "epoch": 0.23070251517779705, + "grad_norm": 2.0476489067077637, + "learning_rate": 1.852894240533875e-05, + "loss": 0.0409, + "step": 2128 + }, + { + "epoch": 0.23081092801387684, + "grad_norm": 2.313424587249756, + "learning_rate": 1.8528217031771362e-05, + "loss": 0.0187, + "step": 2129 + }, + { + "epoch": 0.23091934084995663, + "grad_norm": 6.246420860290527, + "learning_rate": 1.8527491658203977e-05, + "loss": 0.0498, + "step": 2130 + }, + { + "epoch": 0.23102775368603642, + "grad_norm": 4.701159954071045, + "learning_rate": 1.852676628463659e-05, + "loss": 0.0625, + "step": 2131 + }, + { + "epoch": 0.23113616652211622, + "grad_norm": 4.886260032653809, + "learning_rate": 1.8526040911069202e-05, + "loss": 0.0478, + "step": 2132 + }, + { + "epoch": 0.231244579358196, + "grad_norm": 3.8219470977783203, + "learning_rate": 1.8525315537501817e-05, + "loss": 0.0639, + "step": 2133 + }, + { + "epoch": 0.2313529921942758, + "grad_norm": 2.0773422718048096, + "learning_rate": 1.852459016393443e-05, + "loss": 0.0229, + "step": 2134 + }, + { + "epoch": 0.2314614050303556, + "grad_norm": 2.868321418762207, + "learning_rate": 1.852386479036704e-05, + "loss": 0.0238, + "step": 2135 + }, + { + "epoch": 0.23156981786643538, + "grad_norm": 8.036497116088867, + "learning_rate": 1.8523139416799654e-05, + "loss": 0.1157, + "step": 2136 + }, + { + "epoch": 0.23167823070251517, + "grad_norm": 2.2794642448425293, + "learning_rate": 1.8522414043232266e-05, + "loss": 0.0346, + "step": 2137 + }, + { + "epoch": 0.23178664353859496, + "grad_norm": 5.000258445739746, + "learning_rate": 1.8521688669664878e-05, + "loss": 0.0772, + "step": 2138 + }, + { + "epoch": 0.23189505637467475, + "grad_norm": 5.241751670837402, + "learning_rate": 1.852096329609749e-05, + "loss": 0.0586, + "step": 2139 + }, + { + "epoch": 0.23200346921075454, + "grad_norm": 2.9014971256256104, + "learning_rate": 1.8520237922530106e-05, + "loss": 0.0641, + "step": 2140 + }, + { + "epoch": 0.23211188204683433, + "grad_norm": 3.7042312622070312, + "learning_rate": 1.8519512548962718e-05, + "loss": 0.0637, + "step": 2141 + }, + { + "epoch": 0.23222029488291412, + "grad_norm": 2.859504461288452, + "learning_rate": 1.851878717539533e-05, + "loss": 0.0579, + "step": 2142 + }, + { + "epoch": 0.23232870771899392, + "grad_norm": 2.2621424198150635, + "learning_rate": 1.8518061801827942e-05, + "loss": 0.0587, + "step": 2143 + }, + { + "epoch": 0.23243712055507373, + "grad_norm": 1.1506792306900024, + "learning_rate": 1.8517336428260554e-05, + "loss": 0.0259, + "step": 2144 + }, + { + "epoch": 0.23254553339115352, + "grad_norm": 3.80501651763916, + "learning_rate": 1.8516611054693166e-05, + "loss": 0.0429, + "step": 2145 + }, + { + "epoch": 0.23265394622723332, + "grad_norm": 3.64742374420166, + "learning_rate": 1.8515885681125782e-05, + "loss": 0.0525, + "step": 2146 + }, + { + "epoch": 0.2327623590633131, + "grad_norm": 2.932171106338501, + "learning_rate": 1.8515160307558394e-05, + "loss": 0.0412, + "step": 2147 + }, + { + "epoch": 0.2328707718993929, + "grad_norm": 4.015199661254883, + "learning_rate": 1.8514434933991006e-05, + "loss": 0.015, + "step": 2148 + }, + { + "epoch": 0.2329791847354727, + "grad_norm": 2.417578935623169, + "learning_rate": 1.8513709560423622e-05, + "loss": 0.059, + "step": 2149 + }, + { + "epoch": 0.23308759757155248, + "grad_norm": 3.8655083179473877, + "learning_rate": 1.8512984186856234e-05, + "loss": 0.0526, + "step": 2150 + }, + { + "epoch": 0.23319601040763227, + "grad_norm": 1.731955647468567, + "learning_rate": 1.8512258813288846e-05, + "loss": 0.0336, + "step": 2151 + }, + { + "epoch": 0.23330442324371206, + "grad_norm": 4.632339954376221, + "learning_rate": 1.8511533439721458e-05, + "loss": 0.0239, + "step": 2152 + }, + { + "epoch": 0.23341283607979185, + "grad_norm": 2.925224781036377, + "learning_rate": 1.851080806615407e-05, + "loss": 0.026, + "step": 2153 + }, + { + "epoch": 0.23352124891587164, + "grad_norm": 2.5739059448242188, + "learning_rate": 1.8510082692586682e-05, + "loss": 0.0424, + "step": 2154 + }, + { + "epoch": 0.23362966175195143, + "grad_norm": 4.210566997528076, + "learning_rate": 1.8509357319019298e-05, + "loss": 0.0545, + "step": 2155 + }, + { + "epoch": 0.23373807458803122, + "grad_norm": 3.2656993865966797, + "learning_rate": 1.850863194545191e-05, + "loss": 0.0423, + "step": 2156 + }, + { + "epoch": 0.23384648742411102, + "grad_norm": 4.364925384521484, + "learning_rate": 1.8507906571884522e-05, + "loss": 0.0432, + "step": 2157 + }, + { + "epoch": 0.2339549002601908, + "grad_norm": 2.8819713592529297, + "learning_rate": 1.8507181198317134e-05, + "loss": 0.028, + "step": 2158 + }, + { + "epoch": 0.2340633130962706, + "grad_norm": 1.5831679105758667, + "learning_rate": 1.8506455824749747e-05, + "loss": 0.0163, + "step": 2159 + }, + { + "epoch": 0.2341717259323504, + "grad_norm": 1.3475474119186401, + "learning_rate": 1.850573045118236e-05, + "loss": 0.0219, + "step": 2160 + }, + { + "epoch": 0.23428013876843018, + "grad_norm": 2.2622532844543457, + "learning_rate": 1.850500507761497e-05, + "loss": 0.0365, + "step": 2161 + }, + { + "epoch": 0.23438855160450997, + "grad_norm": 2.8649795055389404, + "learning_rate": 1.8504279704047586e-05, + "loss": 0.0436, + "step": 2162 + }, + { + "epoch": 0.23449696444058976, + "grad_norm": 3.298187494277954, + "learning_rate": 1.85035543304802e-05, + "loss": 0.0318, + "step": 2163 + }, + { + "epoch": 0.23460537727666955, + "grad_norm": 4.141169548034668, + "learning_rate": 1.850282895691281e-05, + "loss": 0.0889, + "step": 2164 + }, + { + "epoch": 0.23471379011274934, + "grad_norm": 3.8592381477355957, + "learning_rate": 1.8502103583345426e-05, + "loss": 0.05, + "step": 2165 + }, + { + "epoch": 0.23482220294882913, + "grad_norm": 3.91055965423584, + "learning_rate": 1.850137820977804e-05, + "loss": 0.0493, + "step": 2166 + }, + { + "epoch": 0.23493061578490892, + "grad_norm": 3.67802095413208, + "learning_rate": 1.850065283621065e-05, + "loss": 0.057, + "step": 2167 + }, + { + "epoch": 0.23503902862098872, + "grad_norm": 2.795520544052124, + "learning_rate": 1.8499927462643263e-05, + "loss": 0.05, + "step": 2168 + }, + { + "epoch": 0.2351474414570685, + "grad_norm": 3.4263904094696045, + "learning_rate": 1.8499202089075875e-05, + "loss": 0.0418, + "step": 2169 + }, + { + "epoch": 0.2352558542931483, + "grad_norm": 2.8419666290283203, + "learning_rate": 1.849847671550849e-05, + "loss": 0.0342, + "step": 2170 + }, + { + "epoch": 0.2353642671292281, + "grad_norm": 4.271003246307373, + "learning_rate": 1.8497751341941102e-05, + "loss": 0.0396, + "step": 2171 + }, + { + "epoch": 0.23547267996530788, + "grad_norm": 5.122949600219727, + "learning_rate": 1.8497025968373715e-05, + "loss": 0.0702, + "step": 2172 + }, + { + "epoch": 0.2355810928013877, + "grad_norm": 2.566546678543091, + "learning_rate": 1.8496300594806327e-05, + "loss": 0.0255, + "step": 2173 + }, + { + "epoch": 0.2356895056374675, + "grad_norm": 2.535227060317993, + "learning_rate": 1.849557522123894e-05, + "loss": 0.0583, + "step": 2174 + }, + { + "epoch": 0.23579791847354728, + "grad_norm": 3.217564582824707, + "learning_rate": 1.849484984767155e-05, + "loss": 0.0384, + "step": 2175 + }, + { + "epoch": 0.23590633130962707, + "grad_norm": 5.6305389404296875, + "learning_rate": 1.8494124474104163e-05, + "loss": 0.0439, + "step": 2176 + }, + { + "epoch": 0.23601474414570686, + "grad_norm": 3.2171497344970703, + "learning_rate": 1.849339910053678e-05, + "loss": 0.0788, + "step": 2177 + }, + { + "epoch": 0.23612315698178665, + "grad_norm": 3.6780753135681152, + "learning_rate": 1.849267372696939e-05, + "loss": 0.0543, + "step": 2178 + }, + { + "epoch": 0.23623156981786644, + "grad_norm": 1.7789955139160156, + "learning_rate": 1.8491948353402003e-05, + "loss": 0.0402, + "step": 2179 + }, + { + "epoch": 0.23633998265394623, + "grad_norm": 4.269473075866699, + "learning_rate": 1.8491222979834615e-05, + "loss": 0.0611, + "step": 2180 + }, + { + "epoch": 0.23644839549002603, + "grad_norm": 2.2415401935577393, + "learning_rate": 1.8490497606267227e-05, + "loss": 0.0501, + "step": 2181 + }, + { + "epoch": 0.23655680832610582, + "grad_norm": 3.8455400466918945, + "learning_rate": 1.8489772232699843e-05, + "loss": 0.0624, + "step": 2182 + }, + { + "epoch": 0.2366652211621856, + "grad_norm": 4.838232517242432, + "learning_rate": 1.8489046859132455e-05, + "loss": 0.0402, + "step": 2183 + }, + { + "epoch": 0.2367736339982654, + "grad_norm": 1.5610066652297974, + "learning_rate": 1.8488321485565067e-05, + "loss": 0.0398, + "step": 2184 + }, + { + "epoch": 0.2368820468343452, + "grad_norm": 2.0361995697021484, + "learning_rate": 1.8487596111997683e-05, + "loss": 0.0406, + "step": 2185 + }, + { + "epoch": 0.23699045967042498, + "grad_norm": 2.180983304977417, + "learning_rate": 1.8486870738430295e-05, + "loss": 0.0336, + "step": 2186 + }, + { + "epoch": 0.23709887250650477, + "grad_norm": 1.0783554315567017, + "learning_rate": 1.8486145364862907e-05, + "loss": 0.0317, + "step": 2187 + }, + { + "epoch": 0.23720728534258456, + "grad_norm": 4.532052993774414, + "learning_rate": 1.848541999129552e-05, + "loss": 0.0869, + "step": 2188 + }, + { + "epoch": 0.23731569817866435, + "grad_norm": 3.555368423461914, + "learning_rate": 1.848469461772813e-05, + "loss": 0.0611, + "step": 2189 + }, + { + "epoch": 0.23742411101474414, + "grad_norm": 3.64945650100708, + "learning_rate": 1.8483969244160743e-05, + "loss": 0.0585, + "step": 2190 + }, + { + "epoch": 0.23753252385082393, + "grad_norm": 1.603625774383545, + "learning_rate": 1.8483243870593355e-05, + "loss": 0.0454, + "step": 2191 + }, + { + "epoch": 0.23764093668690373, + "grad_norm": 2.4657857418060303, + "learning_rate": 1.848251849702597e-05, + "loss": 0.0236, + "step": 2192 + }, + { + "epoch": 0.23774934952298352, + "grad_norm": 1.8266812562942505, + "learning_rate": 1.8481793123458583e-05, + "loss": 0.0192, + "step": 2193 + }, + { + "epoch": 0.2378577623590633, + "grad_norm": 5.824270725250244, + "learning_rate": 1.8481067749891195e-05, + "loss": 0.0498, + "step": 2194 + }, + { + "epoch": 0.2379661751951431, + "grad_norm": 3.9635608196258545, + "learning_rate": 1.8480342376323807e-05, + "loss": 0.0765, + "step": 2195 + }, + { + "epoch": 0.2380745880312229, + "grad_norm": 8.797316551208496, + "learning_rate": 1.847961700275642e-05, + "loss": 0.054, + "step": 2196 + }, + { + "epoch": 0.23818300086730268, + "grad_norm": 0.6354877948760986, + "learning_rate": 1.8478891629189032e-05, + "loss": 0.0074, + "step": 2197 + }, + { + "epoch": 0.23829141370338247, + "grad_norm": 2.404409646987915, + "learning_rate": 1.8478166255621647e-05, + "loss": 0.0141, + "step": 2198 + }, + { + "epoch": 0.23839982653946226, + "grad_norm": 3.052698850631714, + "learning_rate": 1.847744088205426e-05, + "loss": 0.0342, + "step": 2199 + }, + { + "epoch": 0.23850823937554205, + "grad_norm": 2.9280178546905518, + "learning_rate": 1.847671550848687e-05, + "loss": 0.0432, + "step": 2200 + }, + { + "epoch": 0.23861665221162184, + "grad_norm": 4.069578647613525, + "learning_rate": 1.8475990134919487e-05, + "loss": 0.147, + "step": 2201 + }, + { + "epoch": 0.23872506504770163, + "grad_norm": 3.047111749649048, + "learning_rate": 1.84752647613521e-05, + "loss": 0.0815, + "step": 2202 + }, + { + "epoch": 0.23883347788378145, + "grad_norm": 1.6489794254302979, + "learning_rate": 1.847453938778471e-05, + "loss": 0.0237, + "step": 2203 + }, + { + "epoch": 0.23894189071986124, + "grad_norm": 2.7998223304748535, + "learning_rate": 1.8473814014217324e-05, + "loss": 0.0315, + "step": 2204 + }, + { + "epoch": 0.23905030355594103, + "grad_norm": 2.6032447814941406, + "learning_rate": 1.8473088640649936e-05, + "loss": 0.0259, + "step": 2205 + }, + { + "epoch": 0.23915871639202083, + "grad_norm": 1.3166917562484741, + "learning_rate": 1.8472363267082548e-05, + "loss": 0.0216, + "step": 2206 + }, + { + "epoch": 0.23926712922810062, + "grad_norm": 2.254208564758301, + "learning_rate": 1.8471637893515163e-05, + "loss": 0.0369, + "step": 2207 + }, + { + "epoch": 0.2393755420641804, + "grad_norm": 1.1404228210449219, + "learning_rate": 1.8470912519947775e-05, + "loss": 0.0204, + "step": 2208 + }, + { + "epoch": 0.2394839549002602, + "grad_norm": 2.5996084213256836, + "learning_rate": 1.8470187146380388e-05, + "loss": 0.029, + "step": 2209 + }, + { + "epoch": 0.23959236773634, + "grad_norm": 3.765835762023926, + "learning_rate": 1.8469461772813e-05, + "loss": 0.0304, + "step": 2210 + }, + { + "epoch": 0.23970078057241978, + "grad_norm": 2.9047136306762695, + "learning_rate": 1.8468736399245612e-05, + "loss": 0.0214, + "step": 2211 + }, + { + "epoch": 0.23980919340849957, + "grad_norm": 6.376235485076904, + "learning_rate": 1.8468011025678224e-05, + "loss": 0.1202, + "step": 2212 + }, + { + "epoch": 0.23991760624457936, + "grad_norm": 2.4999096393585205, + "learning_rate": 1.846728565211084e-05, + "loss": 0.0228, + "step": 2213 + }, + { + "epoch": 0.24002601908065915, + "grad_norm": 1.7705811262130737, + "learning_rate": 1.8466560278543452e-05, + "loss": 0.0225, + "step": 2214 + }, + { + "epoch": 0.24013443191673894, + "grad_norm": 4.346545219421387, + "learning_rate": 1.8465834904976064e-05, + "loss": 0.0401, + "step": 2215 + }, + { + "epoch": 0.24024284475281873, + "grad_norm": 2.5761756896972656, + "learning_rate": 1.8465109531408676e-05, + "loss": 0.0539, + "step": 2216 + }, + { + "epoch": 0.24035125758889853, + "grad_norm": 2.0653493404388428, + "learning_rate": 1.846438415784129e-05, + "loss": 0.0311, + "step": 2217 + }, + { + "epoch": 0.24045967042497832, + "grad_norm": 4.917637348175049, + "learning_rate": 1.8463658784273904e-05, + "loss": 0.0878, + "step": 2218 + }, + { + "epoch": 0.2405680832610581, + "grad_norm": 6.965369701385498, + "learning_rate": 1.8462933410706516e-05, + "loss": 0.072, + "step": 2219 + }, + { + "epoch": 0.2406764960971379, + "grad_norm": 3.7954959869384766, + "learning_rate": 1.8462208037139128e-05, + "loss": 0.0554, + "step": 2220 + }, + { + "epoch": 0.2407849089332177, + "grad_norm": 5.035728931427002, + "learning_rate": 1.846148266357174e-05, + "loss": 0.085, + "step": 2221 + }, + { + "epoch": 0.24089332176929748, + "grad_norm": 2.5386152267456055, + "learning_rate": 1.8460757290004356e-05, + "loss": 0.0506, + "step": 2222 + }, + { + "epoch": 0.24100173460537727, + "grad_norm": 4.233162879943848, + "learning_rate": 1.8460031916436968e-05, + "loss": 0.0446, + "step": 2223 + }, + { + "epoch": 0.24111014744145706, + "grad_norm": 3.0674514770507812, + "learning_rate": 1.845930654286958e-05, + "loss": 0.0585, + "step": 2224 + }, + { + "epoch": 0.24121856027753685, + "grad_norm": 2.6644458770751953, + "learning_rate": 1.8458581169302192e-05, + "loss": 0.0259, + "step": 2225 + }, + { + "epoch": 0.24132697311361664, + "grad_norm": 7.029405117034912, + "learning_rate": 1.8457855795734804e-05, + "loss": 0.0849, + "step": 2226 + }, + { + "epoch": 0.24143538594969643, + "grad_norm": 5.196549415588379, + "learning_rate": 1.8457130422167416e-05, + "loss": 0.0671, + "step": 2227 + }, + { + "epoch": 0.24154379878577623, + "grad_norm": 1.7498167753219604, + "learning_rate": 1.845640504860003e-05, + "loss": 0.0125, + "step": 2228 + }, + { + "epoch": 0.24165221162185602, + "grad_norm": 5.195971965789795, + "learning_rate": 1.8455679675032644e-05, + "loss": 0.0805, + "step": 2229 + }, + { + "epoch": 0.2417606244579358, + "grad_norm": 2.2284107208251953, + "learning_rate": 1.8454954301465256e-05, + "loss": 0.0311, + "step": 2230 + }, + { + "epoch": 0.2418690372940156, + "grad_norm": 1.5413310527801514, + "learning_rate": 1.845422892789787e-05, + "loss": 0.0272, + "step": 2231 + }, + { + "epoch": 0.2419774501300954, + "grad_norm": 1.2614442110061646, + "learning_rate": 1.845350355433048e-05, + "loss": 0.0193, + "step": 2232 + }, + { + "epoch": 0.2420858629661752, + "grad_norm": 1.347915768623352, + "learning_rate": 1.8452778180763093e-05, + "loss": 0.044, + "step": 2233 + }, + { + "epoch": 0.242194275802255, + "grad_norm": 4.775267601013184, + "learning_rate": 1.8452052807195708e-05, + "loss": 0.0408, + "step": 2234 + }, + { + "epoch": 0.2423026886383348, + "grad_norm": 3.578883647918701, + "learning_rate": 1.845132743362832e-05, + "loss": 0.0605, + "step": 2235 + }, + { + "epoch": 0.24241110147441458, + "grad_norm": 1.9900739192962646, + "learning_rate": 1.8450602060060932e-05, + "loss": 0.0336, + "step": 2236 + }, + { + "epoch": 0.24251951431049437, + "grad_norm": 5.437609672546387, + "learning_rate": 1.8449876686493548e-05, + "loss": 0.0691, + "step": 2237 + }, + { + "epoch": 0.24262792714657416, + "grad_norm": 2.6241705417633057, + "learning_rate": 1.844915131292616e-05, + "loss": 0.0559, + "step": 2238 + }, + { + "epoch": 0.24273633998265395, + "grad_norm": 4.606181621551514, + "learning_rate": 1.8448425939358772e-05, + "loss": 0.0504, + "step": 2239 + }, + { + "epoch": 0.24284475281873374, + "grad_norm": 3.3837738037109375, + "learning_rate": 1.8447700565791384e-05, + "loss": 0.0509, + "step": 2240 + }, + { + "epoch": 0.24295316565481354, + "grad_norm": 4.915663242340088, + "learning_rate": 1.8446975192223997e-05, + "loss": 0.0434, + "step": 2241 + }, + { + "epoch": 0.24306157849089333, + "grad_norm": 1.810084581375122, + "learning_rate": 1.844624981865661e-05, + "loss": 0.0338, + "step": 2242 + }, + { + "epoch": 0.24316999132697312, + "grad_norm": 1.6156169176101685, + "learning_rate": 1.844552444508922e-05, + "loss": 0.0344, + "step": 2243 + }, + { + "epoch": 0.2432784041630529, + "grad_norm": 1.9399898052215576, + "learning_rate": 1.8444799071521836e-05, + "loss": 0.0237, + "step": 2244 + }, + { + "epoch": 0.2433868169991327, + "grad_norm": 4.648477077484131, + "learning_rate": 1.844407369795445e-05, + "loss": 0.0699, + "step": 2245 + }, + { + "epoch": 0.2434952298352125, + "grad_norm": 1.3533254861831665, + "learning_rate": 1.844334832438706e-05, + "loss": 0.0346, + "step": 2246 + }, + { + "epoch": 0.24360364267129228, + "grad_norm": 6.56313419342041, + "learning_rate": 1.8442622950819673e-05, + "loss": 0.0817, + "step": 2247 + }, + { + "epoch": 0.24371205550737207, + "grad_norm": 2.7353129386901855, + "learning_rate": 1.8441897577252285e-05, + "loss": 0.0533, + "step": 2248 + }, + { + "epoch": 0.24382046834345186, + "grad_norm": 3.723716974258423, + "learning_rate": 1.8441172203684897e-05, + "loss": 0.0552, + "step": 2249 + }, + { + "epoch": 0.24392888117953165, + "grad_norm": 4.050520420074463, + "learning_rate": 1.8440446830117513e-05, + "loss": 0.0504, + "step": 2250 + }, + { + "epoch": 0.24403729401561144, + "grad_norm": 3.4807403087615967, + "learning_rate": 1.8439721456550125e-05, + "loss": 0.061, + "step": 2251 + }, + { + "epoch": 0.24414570685169124, + "grad_norm": 2.6381382942199707, + "learning_rate": 1.8438996082982737e-05, + "loss": 0.0515, + "step": 2252 + }, + { + "epoch": 0.24425411968777103, + "grad_norm": 2.5182454586029053, + "learning_rate": 1.8438270709415352e-05, + "loss": 0.0407, + "step": 2253 + }, + { + "epoch": 0.24436253252385082, + "grad_norm": 1.291696548461914, + "learning_rate": 1.8437545335847965e-05, + "loss": 0.0195, + "step": 2254 + }, + { + "epoch": 0.2444709453599306, + "grad_norm": 4.694850921630859, + "learning_rate": 1.8436819962280577e-05, + "loss": 0.0697, + "step": 2255 + }, + { + "epoch": 0.2445793581960104, + "grad_norm": 2.4261488914489746, + "learning_rate": 1.843609458871319e-05, + "loss": 0.0249, + "step": 2256 + }, + { + "epoch": 0.2446877710320902, + "grad_norm": 9.08958911895752, + "learning_rate": 1.84353692151458e-05, + "loss": 0.0589, + "step": 2257 + }, + { + "epoch": 0.24479618386816998, + "grad_norm": 6.233621597290039, + "learning_rate": 1.8434643841578413e-05, + "loss": 0.0456, + "step": 2258 + }, + { + "epoch": 0.24490459670424977, + "grad_norm": 10.336219787597656, + "learning_rate": 1.843391846801103e-05, + "loss": 0.0333, + "step": 2259 + }, + { + "epoch": 0.24501300954032956, + "grad_norm": 1.5606348514556885, + "learning_rate": 1.843319309444364e-05, + "loss": 0.0157, + "step": 2260 + }, + { + "epoch": 0.24512142237640935, + "grad_norm": 5.0524444580078125, + "learning_rate": 1.8432467720876253e-05, + "loss": 0.0687, + "step": 2261 + }, + { + "epoch": 0.24522983521248917, + "grad_norm": 3.572355031967163, + "learning_rate": 1.8431742347308865e-05, + "loss": 0.045, + "step": 2262 + }, + { + "epoch": 0.24533824804856896, + "grad_norm": 2.958829402923584, + "learning_rate": 1.8431016973741477e-05, + "loss": 0.0277, + "step": 2263 + }, + { + "epoch": 0.24544666088464875, + "grad_norm": 2.623659372329712, + "learning_rate": 1.843029160017409e-05, + "loss": 0.0155, + "step": 2264 + }, + { + "epoch": 0.24555507372072854, + "grad_norm": 2.115173101425171, + "learning_rate": 1.8429566226606705e-05, + "loss": 0.05, + "step": 2265 + }, + { + "epoch": 0.24566348655680834, + "grad_norm": 3.0897340774536133, + "learning_rate": 1.8428840853039317e-05, + "loss": 0.0268, + "step": 2266 + }, + { + "epoch": 0.24577189939288813, + "grad_norm": 1.4591290950775146, + "learning_rate": 1.842811547947193e-05, + "loss": 0.0214, + "step": 2267 + }, + { + "epoch": 0.24588031222896792, + "grad_norm": 2.4454774856567383, + "learning_rate": 1.842739010590454e-05, + "loss": 0.0283, + "step": 2268 + }, + { + "epoch": 0.2459887250650477, + "grad_norm": 3.7890098094940186, + "learning_rate": 1.8426664732337157e-05, + "loss": 0.0447, + "step": 2269 + }, + { + "epoch": 0.2460971379011275, + "grad_norm": 3.0968828201293945, + "learning_rate": 1.842593935876977e-05, + "loss": 0.0435, + "step": 2270 + }, + { + "epoch": 0.2462055507372073, + "grad_norm": 2.18036150932312, + "learning_rate": 1.842521398520238e-05, + "loss": 0.0121, + "step": 2271 + }, + { + "epoch": 0.24631396357328708, + "grad_norm": 6.648874759674072, + "learning_rate": 1.8424488611634993e-05, + "loss": 0.0502, + "step": 2272 + }, + { + "epoch": 0.24642237640936687, + "grad_norm": 2.2142820358276367, + "learning_rate": 1.8423763238067605e-05, + "loss": 0.037, + "step": 2273 + }, + { + "epoch": 0.24653078924544666, + "grad_norm": 3.6859805583953857, + "learning_rate": 1.842303786450022e-05, + "loss": 0.0614, + "step": 2274 + }, + { + "epoch": 0.24663920208152645, + "grad_norm": 8.858135223388672, + "learning_rate": 1.8422312490932833e-05, + "loss": 0.1113, + "step": 2275 + }, + { + "epoch": 0.24674761491760624, + "grad_norm": 1.9076359272003174, + "learning_rate": 1.8421587117365445e-05, + "loss": 0.0328, + "step": 2276 + }, + { + "epoch": 0.24685602775368604, + "grad_norm": 2.837085723876953, + "learning_rate": 1.8420861743798057e-05, + "loss": 0.0606, + "step": 2277 + }, + { + "epoch": 0.24696444058976583, + "grad_norm": 3.5836000442504883, + "learning_rate": 1.842013637023067e-05, + "loss": 0.0553, + "step": 2278 + }, + { + "epoch": 0.24707285342584562, + "grad_norm": 4.738800525665283, + "learning_rate": 1.8419410996663282e-05, + "loss": 0.0393, + "step": 2279 + }, + { + "epoch": 0.2471812662619254, + "grad_norm": 4.996350288391113, + "learning_rate": 1.8418685623095894e-05, + "loss": 0.044, + "step": 2280 + }, + { + "epoch": 0.2472896790980052, + "grad_norm": 7.546358585357666, + "learning_rate": 1.841796024952851e-05, + "loss": 0.0411, + "step": 2281 + }, + { + "epoch": 0.247398091934085, + "grad_norm": 2.9235641956329346, + "learning_rate": 1.841723487596112e-05, + "loss": 0.0363, + "step": 2282 + }, + { + "epoch": 0.24750650477016478, + "grad_norm": 3.204606771469116, + "learning_rate": 1.8416509502393734e-05, + "loss": 0.0363, + "step": 2283 + }, + { + "epoch": 0.24761491760624457, + "grad_norm": 2.4104573726654053, + "learning_rate": 1.8415784128826346e-05, + "loss": 0.0411, + "step": 2284 + }, + { + "epoch": 0.24772333044232436, + "grad_norm": 4.097807884216309, + "learning_rate": 1.8415058755258958e-05, + "loss": 0.043, + "step": 2285 + }, + { + "epoch": 0.24783174327840415, + "grad_norm": 5.05214786529541, + "learning_rate": 1.8414333381691573e-05, + "loss": 0.0531, + "step": 2286 + }, + { + "epoch": 0.24794015611448394, + "grad_norm": 4.512495517730713, + "learning_rate": 1.8413608008124186e-05, + "loss": 0.0624, + "step": 2287 + }, + { + "epoch": 0.24804856895056374, + "grad_norm": 1.2553125619888306, + "learning_rate": 1.8412882634556798e-05, + "loss": 0.0145, + "step": 2288 + }, + { + "epoch": 0.24815698178664353, + "grad_norm": 4.470438480377197, + "learning_rate": 1.8412157260989413e-05, + "loss": 0.0379, + "step": 2289 + }, + { + "epoch": 0.24826539462272332, + "grad_norm": 1.9157912731170654, + "learning_rate": 1.8411431887422025e-05, + "loss": 0.0452, + "step": 2290 + }, + { + "epoch": 0.2483738074588031, + "grad_norm": 4.219871520996094, + "learning_rate": 1.8410706513854638e-05, + "loss": 0.056, + "step": 2291 + }, + { + "epoch": 0.24848222029488293, + "grad_norm": 5.653114318847656, + "learning_rate": 1.840998114028725e-05, + "loss": 0.0763, + "step": 2292 + }, + { + "epoch": 0.24859063313096272, + "grad_norm": 2.898636817932129, + "learning_rate": 1.8409255766719862e-05, + "loss": 0.0309, + "step": 2293 + }, + { + "epoch": 0.2486990459670425, + "grad_norm": 9.779182434082031, + "learning_rate": 1.8408530393152474e-05, + "loss": 0.1153, + "step": 2294 + }, + { + "epoch": 0.2488074588031223, + "grad_norm": 3.4157638549804688, + "learning_rate": 1.8407805019585086e-05, + "loss": 0.0293, + "step": 2295 + }, + { + "epoch": 0.2489158716392021, + "grad_norm": 6.5012288093566895, + "learning_rate": 1.8407079646017702e-05, + "loss": 0.076, + "step": 2296 + }, + { + "epoch": 0.24902428447528188, + "grad_norm": 4.638744831085205, + "learning_rate": 1.8406354272450314e-05, + "loss": 0.0353, + "step": 2297 + }, + { + "epoch": 0.24913269731136167, + "grad_norm": 4.604453086853027, + "learning_rate": 1.8405628898882926e-05, + "loss": 0.0619, + "step": 2298 + }, + { + "epoch": 0.24924111014744146, + "grad_norm": 2.1598637104034424, + "learning_rate": 1.8404903525315538e-05, + "loss": 0.0195, + "step": 2299 + }, + { + "epoch": 0.24934952298352125, + "grad_norm": 5.405832767486572, + "learning_rate": 1.840417815174815e-05, + "loss": 0.1183, + "step": 2300 + }, + { + "epoch": 0.24945793581960105, + "grad_norm": 2.7371037006378174, + "learning_rate": 1.8403452778180762e-05, + "loss": 0.052, + "step": 2301 + }, + { + "epoch": 0.24956634865568084, + "grad_norm": 1.9935358762741089, + "learning_rate": 1.8402727404613378e-05, + "loss": 0.0273, + "step": 2302 + }, + { + "epoch": 0.24967476149176063, + "grad_norm": 2.362785816192627, + "learning_rate": 1.840200203104599e-05, + "loss": 0.024, + "step": 2303 + }, + { + "epoch": 0.24978317432784042, + "grad_norm": 4.599124431610107, + "learning_rate": 1.8401276657478602e-05, + "loss": 0.0287, + "step": 2304 + }, + { + "epoch": 0.2498915871639202, + "grad_norm": 5.7473063468933105, + "learning_rate": 1.8400551283911218e-05, + "loss": 0.0628, + "step": 2305 + }, + { + "epoch": 0.25, + "grad_norm": 2.418081521987915, + "learning_rate": 1.839982591034383e-05, + "loss": 0.0537, + "step": 2306 + }, + { + "epoch": 0.2501084128360798, + "grad_norm": 1.8798428773880005, + "learning_rate": 1.8399100536776442e-05, + "loss": 0.0175, + "step": 2307 + }, + { + "epoch": 0.2502168256721596, + "grad_norm": 2.9616119861602783, + "learning_rate": 1.8398375163209054e-05, + "loss": 0.0686, + "step": 2308 + }, + { + "epoch": 0.2503252385082394, + "grad_norm": 5.235623359680176, + "learning_rate": 1.8397649789641666e-05, + "loss": 0.0521, + "step": 2309 + }, + { + "epoch": 0.25043365134431916, + "grad_norm": 4.374730110168457, + "learning_rate": 1.839692441607428e-05, + "loss": 0.0787, + "step": 2310 + }, + { + "epoch": 0.250542064180399, + "grad_norm": 1.0317476987838745, + "learning_rate": 1.8396199042506894e-05, + "loss": 0.0143, + "step": 2311 + }, + { + "epoch": 0.25065047701647875, + "grad_norm": 4.177217483520508, + "learning_rate": 1.8395473668939506e-05, + "loss": 0.0457, + "step": 2312 + }, + { + "epoch": 0.25075888985255856, + "grad_norm": 5.521426200866699, + "learning_rate": 1.8394748295372118e-05, + "loss": 0.0582, + "step": 2313 + }, + { + "epoch": 0.2508673026886383, + "grad_norm": 2.3718559741973877, + "learning_rate": 1.839402292180473e-05, + "loss": 0.0092, + "step": 2314 + }, + { + "epoch": 0.25097571552471815, + "grad_norm": 6.355928897857666, + "learning_rate": 1.8393297548237343e-05, + "loss": 0.0859, + "step": 2315 + }, + { + "epoch": 0.2510841283607979, + "grad_norm": 4.130633354187012, + "learning_rate": 1.8392572174669955e-05, + "loss": 0.0501, + "step": 2316 + }, + { + "epoch": 0.2511925411968777, + "grad_norm": 2.536478281021118, + "learning_rate": 1.839184680110257e-05, + "loss": 0.0271, + "step": 2317 + }, + { + "epoch": 0.2513009540329575, + "grad_norm": 4.082536697387695, + "learning_rate": 1.8391121427535182e-05, + "loss": 0.0517, + "step": 2318 + }, + { + "epoch": 0.2514093668690373, + "grad_norm": 4.338755130767822, + "learning_rate": 1.8390396053967795e-05, + "loss": 0.0543, + "step": 2319 + }, + { + "epoch": 0.2515177797051171, + "grad_norm": 4.964118957519531, + "learning_rate": 1.838967068040041e-05, + "loss": 0.0956, + "step": 2320 + }, + { + "epoch": 0.2516261925411969, + "grad_norm": 3.6338744163513184, + "learning_rate": 1.8388945306833022e-05, + "loss": 0.0468, + "step": 2321 + }, + { + "epoch": 0.25173460537727665, + "grad_norm": 2.342283010482788, + "learning_rate": 1.8388219933265634e-05, + "loss": 0.0362, + "step": 2322 + }, + { + "epoch": 0.2518430182133565, + "grad_norm": 2.4626288414001465, + "learning_rate": 1.8387494559698247e-05, + "loss": 0.0418, + "step": 2323 + }, + { + "epoch": 0.25195143104943624, + "grad_norm": 1.2935270071029663, + "learning_rate": 1.838676918613086e-05, + "loss": 0.013, + "step": 2324 + }, + { + "epoch": 0.25205984388551606, + "grad_norm": 5.067253112792969, + "learning_rate": 1.838604381256347e-05, + "loss": 0.0469, + "step": 2325 + }, + { + "epoch": 0.2521682567215958, + "grad_norm": 2.664487600326538, + "learning_rate": 1.8385318438996086e-05, + "loss": 0.0355, + "step": 2326 + }, + { + "epoch": 0.25227666955767564, + "grad_norm": 1.0509015321731567, + "learning_rate": 1.83845930654287e-05, + "loss": 0.026, + "step": 2327 + }, + { + "epoch": 0.2523850823937554, + "grad_norm": 2.146862030029297, + "learning_rate": 1.838386769186131e-05, + "loss": 0.0231, + "step": 2328 + }, + { + "epoch": 0.2524934952298352, + "grad_norm": 2.1589858531951904, + "learning_rate": 1.8383142318293923e-05, + "loss": 0.0647, + "step": 2329 + }, + { + "epoch": 0.252601908065915, + "grad_norm": 4.503753662109375, + "learning_rate": 1.8382416944726535e-05, + "loss": 0.0171, + "step": 2330 + }, + { + "epoch": 0.2527103209019948, + "grad_norm": 4.433594226837158, + "learning_rate": 1.8381691571159147e-05, + "loss": 0.0976, + "step": 2331 + }, + { + "epoch": 0.25281873373807456, + "grad_norm": 4.307727336883545, + "learning_rate": 1.838096619759176e-05, + "loss": 0.0882, + "step": 2332 + }, + { + "epoch": 0.2529271465741544, + "grad_norm": 4.8844499588012695, + "learning_rate": 1.8380240824024375e-05, + "loss": 0.0822, + "step": 2333 + }, + { + "epoch": 0.25303555941023415, + "grad_norm": 2.9252769947052, + "learning_rate": 1.8379515450456987e-05, + "loss": 0.0585, + "step": 2334 + }, + { + "epoch": 0.25314397224631396, + "grad_norm": 3.1684768199920654, + "learning_rate": 1.83787900768896e-05, + "loss": 0.049, + "step": 2335 + }, + { + "epoch": 0.2532523850823938, + "grad_norm": 4.269593238830566, + "learning_rate": 1.837806470332221e-05, + "loss": 0.0696, + "step": 2336 + }, + { + "epoch": 0.25336079791847355, + "grad_norm": 3.4905669689178467, + "learning_rate": 1.8377339329754823e-05, + "loss": 0.0428, + "step": 2337 + }, + { + "epoch": 0.25346921075455336, + "grad_norm": 3.405658721923828, + "learning_rate": 1.837661395618744e-05, + "loss": 0.0467, + "step": 2338 + }, + { + "epoch": 0.2535776235906331, + "grad_norm": 1.9931446313858032, + "learning_rate": 1.837588858262005e-05, + "loss": 0.0181, + "step": 2339 + }, + { + "epoch": 0.25368603642671295, + "grad_norm": 5.777990341186523, + "learning_rate": 1.8375163209052663e-05, + "loss": 0.0598, + "step": 2340 + }, + { + "epoch": 0.2537944492627927, + "grad_norm": 1.1653363704681396, + "learning_rate": 1.837443783548528e-05, + "loss": 0.0167, + "step": 2341 + }, + { + "epoch": 0.25390286209887253, + "grad_norm": 3.1007752418518066, + "learning_rate": 1.837371246191789e-05, + "loss": 0.0379, + "step": 2342 + }, + { + "epoch": 0.2540112749349523, + "grad_norm": 1.9086161851882935, + "learning_rate": 1.8372987088350503e-05, + "loss": 0.0494, + "step": 2343 + }, + { + "epoch": 0.2541196877710321, + "grad_norm": 4.825074672698975, + "learning_rate": 1.8372261714783115e-05, + "loss": 0.0629, + "step": 2344 + }, + { + "epoch": 0.2542281006071119, + "grad_norm": 3.1576387882232666, + "learning_rate": 1.8371536341215727e-05, + "loss": 0.0411, + "step": 2345 + }, + { + "epoch": 0.2543365134431917, + "grad_norm": 6.990752696990967, + "learning_rate": 1.837081096764834e-05, + "loss": 0.0437, + "step": 2346 + }, + { + "epoch": 0.25444492627927146, + "grad_norm": 2.774616241455078, + "learning_rate": 1.837008559408095e-05, + "loss": 0.0152, + "step": 2347 + }, + { + "epoch": 0.2545533391153513, + "grad_norm": 3.625684976577759, + "learning_rate": 1.8369360220513567e-05, + "loss": 0.0392, + "step": 2348 + }, + { + "epoch": 0.25466175195143104, + "grad_norm": 2.2788333892822266, + "learning_rate": 1.836863484694618e-05, + "loss": 0.0566, + "step": 2349 + }, + { + "epoch": 0.25477016478751086, + "grad_norm": 2.0731053352355957, + "learning_rate": 1.836790947337879e-05, + "loss": 0.0285, + "step": 2350 + }, + { + "epoch": 0.2548785776235906, + "grad_norm": 4.582964897155762, + "learning_rate": 1.8367184099811403e-05, + "loss": 0.0618, + "step": 2351 + }, + { + "epoch": 0.25498699045967044, + "grad_norm": 3.2468464374542236, + "learning_rate": 1.8366458726244016e-05, + "loss": 0.0652, + "step": 2352 + }, + { + "epoch": 0.2550954032957502, + "grad_norm": 2.5959665775299072, + "learning_rate": 1.8365733352676628e-05, + "loss": 0.0216, + "step": 2353 + }, + { + "epoch": 0.25520381613183, + "grad_norm": 3.7386362552642822, + "learning_rate": 1.8365007979109243e-05, + "loss": 0.0597, + "step": 2354 + }, + { + "epoch": 0.2553122289679098, + "grad_norm": 2.793792486190796, + "learning_rate": 1.8364282605541855e-05, + "loss": 0.0371, + "step": 2355 + }, + { + "epoch": 0.2554206418039896, + "grad_norm": 6.451307773590088, + "learning_rate": 1.836355723197447e-05, + "loss": 0.0481, + "step": 2356 + }, + { + "epoch": 0.25552905464006936, + "grad_norm": 3.3852760791778564, + "learning_rate": 1.8362831858407083e-05, + "loss": 0.0556, + "step": 2357 + }, + { + "epoch": 0.2556374674761492, + "grad_norm": 6.353251934051514, + "learning_rate": 1.8362106484839695e-05, + "loss": 0.075, + "step": 2358 + }, + { + "epoch": 0.25574588031222895, + "grad_norm": 4.976004600524902, + "learning_rate": 1.8361381111272307e-05, + "loss": 0.0789, + "step": 2359 + }, + { + "epoch": 0.25585429314830876, + "grad_norm": 4.851810932159424, + "learning_rate": 1.836065573770492e-05, + "loss": 0.0751, + "step": 2360 + }, + { + "epoch": 0.2559627059843885, + "grad_norm": 2.0496139526367188, + "learning_rate": 1.835993036413753e-05, + "loss": 0.0244, + "step": 2361 + }, + { + "epoch": 0.25607111882046835, + "grad_norm": 3.0866715908050537, + "learning_rate": 1.8359204990570144e-05, + "loss": 0.0727, + "step": 2362 + }, + { + "epoch": 0.2561795316565481, + "grad_norm": 2.491483211517334, + "learning_rate": 1.835847961700276e-05, + "loss": 0.0578, + "step": 2363 + }, + { + "epoch": 0.25628794449262793, + "grad_norm": 4.325271129608154, + "learning_rate": 1.835775424343537e-05, + "loss": 0.0816, + "step": 2364 + }, + { + "epoch": 0.2563963573287077, + "grad_norm": 2.491676092147827, + "learning_rate": 1.8357028869867984e-05, + "loss": 0.0187, + "step": 2365 + }, + { + "epoch": 0.2565047701647875, + "grad_norm": 2.0914368629455566, + "learning_rate": 1.8356303496300596e-05, + "loss": 0.026, + "step": 2366 + }, + { + "epoch": 0.25661318300086733, + "grad_norm": 2.8731892108917236, + "learning_rate": 1.8355578122733208e-05, + "loss": 0.06, + "step": 2367 + }, + { + "epoch": 0.2567215958369471, + "grad_norm": 3.2463173866271973, + "learning_rate": 1.835485274916582e-05, + "loss": 0.0434, + "step": 2368 + }, + { + "epoch": 0.2568300086730269, + "grad_norm": 2.8686914443969727, + "learning_rate": 1.8354127375598436e-05, + "loss": 0.0755, + "step": 2369 + }, + { + "epoch": 0.2569384215091067, + "grad_norm": 2.388913154602051, + "learning_rate": 1.8353402002031048e-05, + "loss": 0.0379, + "step": 2370 + }, + { + "epoch": 0.2570468343451865, + "grad_norm": 2.4746828079223633, + "learning_rate": 1.835267662846366e-05, + "loss": 0.0213, + "step": 2371 + }, + { + "epoch": 0.25715524718126626, + "grad_norm": 1.5121642351150513, + "learning_rate": 1.8351951254896275e-05, + "loss": 0.0304, + "step": 2372 + }, + { + "epoch": 0.2572636600173461, + "grad_norm": 5.573277950286865, + "learning_rate": 1.8351225881328888e-05, + "loss": 0.0518, + "step": 2373 + }, + { + "epoch": 0.25737207285342584, + "grad_norm": 1.1909584999084473, + "learning_rate": 1.83505005077615e-05, + "loss": 0.0186, + "step": 2374 + }, + { + "epoch": 0.25748048568950566, + "grad_norm": 1.408575177192688, + "learning_rate": 1.8349775134194112e-05, + "loss": 0.0251, + "step": 2375 + }, + { + "epoch": 0.2575888985255854, + "grad_norm": 9.057255744934082, + "learning_rate": 1.8349049760626724e-05, + "loss": 0.0574, + "step": 2376 + }, + { + "epoch": 0.25769731136166524, + "grad_norm": 2.7288334369659424, + "learning_rate": 1.8348324387059336e-05, + "loss": 0.0651, + "step": 2377 + }, + { + "epoch": 0.257805724197745, + "grad_norm": 2.3087258338928223, + "learning_rate": 1.834759901349195e-05, + "loss": 0.0382, + "step": 2378 + }, + { + "epoch": 0.2579141370338248, + "grad_norm": 2.583451747894287, + "learning_rate": 1.8346873639924564e-05, + "loss": 0.0279, + "step": 2379 + }, + { + "epoch": 0.2580225498699046, + "grad_norm": 7.381651878356934, + "learning_rate": 1.8346148266357176e-05, + "loss": 0.048, + "step": 2380 + }, + { + "epoch": 0.2581309627059844, + "grad_norm": 4.079891204833984, + "learning_rate": 1.8345422892789788e-05, + "loss": 0.0421, + "step": 2381 + }, + { + "epoch": 0.25823937554206416, + "grad_norm": 1.920676589012146, + "learning_rate": 1.83446975192224e-05, + "loss": 0.0153, + "step": 2382 + }, + { + "epoch": 0.258347788378144, + "grad_norm": 3.0104172229766846, + "learning_rate": 1.8343972145655012e-05, + "loss": 0.0306, + "step": 2383 + }, + { + "epoch": 0.25845620121422375, + "grad_norm": 2.348498582839966, + "learning_rate": 1.8343246772087625e-05, + "loss": 0.0275, + "step": 2384 + }, + { + "epoch": 0.25856461405030357, + "grad_norm": 2.515376329421997, + "learning_rate": 1.834252139852024e-05, + "loss": 0.0551, + "step": 2385 + }, + { + "epoch": 0.25867302688638333, + "grad_norm": 5.3374409675598145, + "learning_rate": 1.8341796024952852e-05, + "loss": 0.0569, + "step": 2386 + }, + { + "epoch": 0.25878143972246315, + "grad_norm": 4.105186939239502, + "learning_rate": 1.8341070651385464e-05, + "loss": 0.0442, + "step": 2387 + }, + { + "epoch": 0.2588898525585429, + "grad_norm": 1.930112361907959, + "learning_rate": 1.8340345277818076e-05, + "loss": 0.0259, + "step": 2388 + }, + { + "epoch": 0.25899826539462273, + "grad_norm": 3.265601396560669, + "learning_rate": 1.833961990425069e-05, + "loss": 0.0576, + "step": 2389 + }, + { + "epoch": 0.2591066782307025, + "grad_norm": 2.92000412940979, + "learning_rate": 1.8338894530683304e-05, + "loss": 0.028, + "step": 2390 + }, + { + "epoch": 0.2592150910667823, + "grad_norm": 3.6306397914886475, + "learning_rate": 1.8338169157115916e-05, + "loss": 0.0304, + "step": 2391 + }, + { + "epoch": 0.2593235039028621, + "grad_norm": 3.0283548831939697, + "learning_rate": 1.833744378354853e-05, + "loss": 0.0294, + "step": 2392 + }, + { + "epoch": 0.2594319167389419, + "grad_norm": 7.465962886810303, + "learning_rate": 1.8336718409981144e-05, + "loss": 0.0484, + "step": 2393 + }, + { + "epoch": 0.25954032957502166, + "grad_norm": 6.4476518630981445, + "learning_rate": 1.8335993036413756e-05, + "loss": 0.051, + "step": 2394 + }, + { + "epoch": 0.2596487424111015, + "grad_norm": 2.4857194423675537, + "learning_rate": 1.8335267662846368e-05, + "loss": 0.0474, + "step": 2395 + }, + { + "epoch": 0.2597571552471813, + "grad_norm": 3.4636569023132324, + "learning_rate": 1.833454228927898e-05, + "loss": 0.0455, + "step": 2396 + }, + { + "epoch": 0.25986556808326106, + "grad_norm": 2.211459159851074, + "learning_rate": 1.8333816915711593e-05, + "loss": 0.0181, + "step": 2397 + }, + { + "epoch": 0.2599739809193409, + "grad_norm": 2.1903154850006104, + "learning_rate": 1.8333091542144205e-05, + "loss": 0.0288, + "step": 2398 + }, + { + "epoch": 0.26008239375542064, + "grad_norm": 2.5578134059906006, + "learning_rate": 1.8332366168576817e-05, + "loss": 0.0253, + "step": 2399 + }, + { + "epoch": 0.26019080659150046, + "grad_norm": 3.650827646255493, + "learning_rate": 1.8331640795009432e-05, + "loss": 0.0279, + "step": 2400 + }, + { + "epoch": 0.2602992194275802, + "grad_norm": 2.4672231674194336, + "learning_rate": 1.8330915421442045e-05, + "loss": 0.0561, + "step": 2401 + }, + { + "epoch": 0.26040763226366004, + "grad_norm": 11.803242683410645, + "learning_rate": 1.8330190047874657e-05, + "loss": 0.0599, + "step": 2402 + }, + { + "epoch": 0.2605160450997398, + "grad_norm": 7.532435894012451, + "learning_rate": 1.832946467430727e-05, + "loss": 0.0411, + "step": 2403 + }, + { + "epoch": 0.2606244579358196, + "grad_norm": 1.8741800785064697, + "learning_rate": 1.832873930073988e-05, + "loss": 0.0211, + "step": 2404 + }, + { + "epoch": 0.2607328707718994, + "grad_norm": 1.6385862827301025, + "learning_rate": 1.8328013927172493e-05, + "loss": 0.0139, + "step": 2405 + }, + { + "epoch": 0.2608412836079792, + "grad_norm": 1.3521095514297485, + "learning_rate": 1.832728855360511e-05, + "loss": 0.0185, + "step": 2406 + }, + { + "epoch": 0.26094969644405897, + "grad_norm": 9.927142143249512, + "learning_rate": 1.832656318003772e-05, + "loss": 0.1035, + "step": 2407 + }, + { + "epoch": 0.2610581092801388, + "grad_norm": 0.8037011623382568, + "learning_rate": 1.8325837806470336e-05, + "loss": 0.0191, + "step": 2408 + }, + { + "epoch": 0.26116652211621855, + "grad_norm": 4.367736339569092, + "learning_rate": 1.832511243290295e-05, + "loss": 0.0612, + "step": 2409 + }, + { + "epoch": 0.26127493495229837, + "grad_norm": 2.9399521350860596, + "learning_rate": 1.832438705933556e-05, + "loss": 0.0425, + "step": 2410 + }, + { + "epoch": 0.26138334778837813, + "grad_norm": 2.0078060626983643, + "learning_rate": 1.8323661685768173e-05, + "loss": 0.0357, + "step": 2411 + }, + { + "epoch": 0.26149176062445795, + "grad_norm": 2.8420217037200928, + "learning_rate": 1.8322936312200785e-05, + "loss": 0.0549, + "step": 2412 + }, + { + "epoch": 0.2616001734605377, + "grad_norm": 7.3102521896362305, + "learning_rate": 1.8322210938633397e-05, + "loss": 0.0577, + "step": 2413 + }, + { + "epoch": 0.26170858629661753, + "grad_norm": 1.3140895366668701, + "learning_rate": 1.832148556506601e-05, + "loss": 0.0273, + "step": 2414 + }, + { + "epoch": 0.2618169991326973, + "grad_norm": 4.254485607147217, + "learning_rate": 1.8320760191498625e-05, + "loss": 0.0339, + "step": 2415 + }, + { + "epoch": 0.2619254119687771, + "grad_norm": 0.5496554374694824, + "learning_rate": 1.8320034817931237e-05, + "loss": 0.0057, + "step": 2416 + }, + { + "epoch": 0.2620338248048569, + "grad_norm": 1.3611687421798706, + "learning_rate": 1.831930944436385e-05, + "loss": 0.0121, + "step": 2417 + }, + { + "epoch": 0.2621422376409367, + "grad_norm": 5.218635082244873, + "learning_rate": 1.831858407079646e-05, + "loss": 0.0625, + "step": 2418 + }, + { + "epoch": 0.26225065047701646, + "grad_norm": 6.100493431091309, + "learning_rate": 1.8317858697229073e-05, + "loss": 0.0659, + "step": 2419 + }, + { + "epoch": 0.2623590633130963, + "grad_norm": 7.608099460601807, + "learning_rate": 1.8317133323661685e-05, + "loss": 0.0372, + "step": 2420 + }, + { + "epoch": 0.26246747614917604, + "grad_norm": 2.3556530475616455, + "learning_rate": 1.83164079500943e-05, + "loss": 0.0205, + "step": 2421 + }, + { + "epoch": 0.26257588898525586, + "grad_norm": 8.10224437713623, + "learning_rate": 1.8315682576526913e-05, + "loss": 0.0658, + "step": 2422 + }, + { + "epoch": 0.2626843018213356, + "grad_norm": 5.345652103424072, + "learning_rate": 1.8314957202959525e-05, + "loss": 0.0268, + "step": 2423 + }, + { + "epoch": 0.26279271465741544, + "grad_norm": 5.225749969482422, + "learning_rate": 1.831423182939214e-05, + "loss": 0.0174, + "step": 2424 + }, + { + "epoch": 0.26290112749349526, + "grad_norm": 4.2377848625183105, + "learning_rate": 1.8313506455824753e-05, + "loss": 0.047, + "step": 2425 + }, + { + "epoch": 0.263009540329575, + "grad_norm": 1.9096710681915283, + "learning_rate": 1.8312781082257365e-05, + "loss": 0.0377, + "step": 2426 + }, + { + "epoch": 0.26311795316565484, + "grad_norm": 2.3807713985443115, + "learning_rate": 1.8312055708689977e-05, + "loss": 0.0393, + "step": 2427 + }, + { + "epoch": 0.2632263660017346, + "grad_norm": 1.6620080471038818, + "learning_rate": 1.831133033512259e-05, + "loss": 0.0148, + "step": 2428 + }, + { + "epoch": 0.2633347788378144, + "grad_norm": 5.576217174530029, + "learning_rate": 1.83106049615552e-05, + "loss": 0.0527, + "step": 2429 + }, + { + "epoch": 0.2634431916738942, + "grad_norm": 5.249363899230957, + "learning_rate": 1.8309879587987817e-05, + "loss": 0.0864, + "step": 2430 + }, + { + "epoch": 0.263551604509974, + "grad_norm": 1.6911892890930176, + "learning_rate": 1.830915421442043e-05, + "loss": 0.0218, + "step": 2431 + }, + { + "epoch": 0.26366001734605377, + "grad_norm": 4.698164463043213, + "learning_rate": 1.830842884085304e-05, + "loss": 0.0491, + "step": 2432 + }, + { + "epoch": 0.2637684301821336, + "grad_norm": 6.700082302093506, + "learning_rate": 1.8307703467285653e-05, + "loss": 0.0478, + "step": 2433 + }, + { + "epoch": 0.26387684301821335, + "grad_norm": 4.2184038162231445, + "learning_rate": 1.8306978093718266e-05, + "loss": 0.0217, + "step": 2434 + }, + { + "epoch": 0.26398525585429317, + "grad_norm": 3.951120138168335, + "learning_rate": 1.8306252720150878e-05, + "loss": 0.0507, + "step": 2435 + }, + { + "epoch": 0.26409366869037293, + "grad_norm": 11.703688621520996, + "learning_rate": 1.830552734658349e-05, + "loss": 0.032, + "step": 2436 + }, + { + "epoch": 0.26420208152645275, + "grad_norm": 5.931341648101807, + "learning_rate": 1.8304801973016105e-05, + "loss": 0.05, + "step": 2437 + }, + { + "epoch": 0.2643104943625325, + "grad_norm": 2.192159652709961, + "learning_rate": 1.8304076599448718e-05, + "loss": 0.032, + "step": 2438 + }, + { + "epoch": 0.26441890719861233, + "grad_norm": 2.539797782897949, + "learning_rate": 1.830335122588133e-05, + "loss": 0.0414, + "step": 2439 + }, + { + "epoch": 0.2645273200346921, + "grad_norm": 2.6276795864105225, + "learning_rate": 1.8302625852313942e-05, + "loss": 0.0359, + "step": 2440 + }, + { + "epoch": 0.2646357328707719, + "grad_norm": 6.972290515899658, + "learning_rate": 1.8301900478746557e-05, + "loss": 0.0987, + "step": 2441 + }, + { + "epoch": 0.2647441457068517, + "grad_norm": 6.5112762451171875, + "learning_rate": 1.830117510517917e-05, + "loss": 0.0815, + "step": 2442 + }, + { + "epoch": 0.2648525585429315, + "grad_norm": 5.281688690185547, + "learning_rate": 1.830044973161178e-05, + "loss": 0.1209, + "step": 2443 + }, + { + "epoch": 0.26496097137901126, + "grad_norm": 4.47501802444458, + "learning_rate": 1.8299724358044394e-05, + "loss": 0.0602, + "step": 2444 + }, + { + "epoch": 0.2650693842150911, + "grad_norm": 2.3574020862579346, + "learning_rate": 1.829899898447701e-05, + "loss": 0.0552, + "step": 2445 + }, + { + "epoch": 0.26517779705117084, + "grad_norm": 1.5870801210403442, + "learning_rate": 1.829827361090962e-05, + "loss": 0.0316, + "step": 2446 + }, + { + "epoch": 0.26528620988725066, + "grad_norm": 3.527923583984375, + "learning_rate": 1.8297548237342234e-05, + "loss": 0.0239, + "step": 2447 + }, + { + "epoch": 0.2653946227233304, + "grad_norm": 2.0971691608428955, + "learning_rate": 1.8296822863774846e-05, + "loss": 0.0476, + "step": 2448 + }, + { + "epoch": 0.26550303555941024, + "grad_norm": 3.9033782482147217, + "learning_rate": 1.8296097490207458e-05, + "loss": 0.0643, + "step": 2449 + }, + { + "epoch": 0.26561144839549, + "grad_norm": 1.1275486946105957, + "learning_rate": 1.829537211664007e-05, + "loss": 0.0239, + "step": 2450 + }, + { + "epoch": 0.2657198612315698, + "grad_norm": 3.954902410507202, + "learning_rate": 1.8294646743072682e-05, + "loss": 0.0785, + "step": 2451 + }, + { + "epoch": 0.2658282740676496, + "grad_norm": 3.6526405811309814, + "learning_rate": 1.8293921369505298e-05, + "loss": 0.0601, + "step": 2452 + }, + { + "epoch": 0.2659366869037294, + "grad_norm": 2.264113664627075, + "learning_rate": 1.829319599593791e-05, + "loss": 0.0482, + "step": 2453 + }, + { + "epoch": 0.2660450997398092, + "grad_norm": 1.7409697771072388, + "learning_rate": 1.8292470622370522e-05, + "loss": 0.0214, + "step": 2454 + }, + { + "epoch": 0.266153512575889, + "grad_norm": 3.4613680839538574, + "learning_rate": 1.8291745248803134e-05, + "loss": 0.0549, + "step": 2455 + }, + { + "epoch": 0.2662619254119688, + "grad_norm": 2.0148096084594727, + "learning_rate": 1.8291019875235746e-05, + "loss": 0.0605, + "step": 2456 + }, + { + "epoch": 0.26637033824804857, + "grad_norm": 4.7068705558776855, + "learning_rate": 1.829029450166836e-05, + "loss": 0.0446, + "step": 2457 + }, + { + "epoch": 0.2664787510841284, + "grad_norm": 1.51814866065979, + "learning_rate": 1.8289569128100974e-05, + "loss": 0.0203, + "step": 2458 + }, + { + "epoch": 0.26658716392020815, + "grad_norm": 1.9082322120666504, + "learning_rate": 1.8288843754533586e-05, + "loss": 0.0442, + "step": 2459 + }, + { + "epoch": 0.26669557675628797, + "grad_norm": 6.894189834594727, + "learning_rate": 1.82881183809662e-05, + "loss": 0.0554, + "step": 2460 + }, + { + "epoch": 0.26680398959236773, + "grad_norm": 2.9110593795776367, + "learning_rate": 1.8287393007398814e-05, + "loss": 0.0479, + "step": 2461 + }, + { + "epoch": 0.26691240242844755, + "grad_norm": 0.7610825300216675, + "learning_rate": 1.8286667633831426e-05, + "loss": 0.0159, + "step": 2462 + }, + { + "epoch": 0.2670208152645273, + "grad_norm": 2.206465482711792, + "learning_rate": 1.8285942260264038e-05, + "loss": 0.0362, + "step": 2463 + }, + { + "epoch": 0.26712922810060713, + "grad_norm": 2.622762441635132, + "learning_rate": 1.828521688669665e-05, + "loss": 0.0127, + "step": 2464 + }, + { + "epoch": 0.2672376409366869, + "grad_norm": 1.8428268432617188, + "learning_rate": 1.8284491513129262e-05, + "loss": 0.0231, + "step": 2465 + }, + { + "epoch": 0.2673460537727667, + "grad_norm": 3.18438720703125, + "learning_rate": 1.8283766139561874e-05, + "loss": 0.0565, + "step": 2466 + }, + { + "epoch": 0.2674544666088465, + "grad_norm": 0.6824972033500671, + "learning_rate": 1.828304076599449e-05, + "loss": 0.0061, + "step": 2467 + }, + { + "epoch": 0.2675628794449263, + "grad_norm": 2.690300941467285, + "learning_rate": 1.8282315392427102e-05, + "loss": 0.0446, + "step": 2468 + }, + { + "epoch": 0.26767129228100606, + "grad_norm": 4.273416519165039, + "learning_rate": 1.8281590018859714e-05, + "loss": 0.0724, + "step": 2469 + }, + { + "epoch": 0.2677797051170859, + "grad_norm": 2.885435104370117, + "learning_rate": 1.8280864645292326e-05, + "loss": 0.0511, + "step": 2470 + }, + { + "epoch": 0.26788811795316564, + "grad_norm": 2.1730167865753174, + "learning_rate": 1.828013927172494e-05, + "loss": 0.0363, + "step": 2471 + }, + { + "epoch": 0.26799653078924546, + "grad_norm": 1.8032655715942383, + "learning_rate": 1.827941389815755e-05, + "loss": 0.0415, + "step": 2472 + }, + { + "epoch": 0.2681049436253252, + "grad_norm": 5.78097677230835, + "learning_rate": 1.8278688524590166e-05, + "loss": 0.0501, + "step": 2473 + }, + { + "epoch": 0.26821335646140504, + "grad_norm": 2.08345103263855, + "learning_rate": 1.827796315102278e-05, + "loss": 0.0529, + "step": 2474 + }, + { + "epoch": 0.2683217692974848, + "grad_norm": 2.994398355484009, + "learning_rate": 1.827723777745539e-05, + "loss": 0.067, + "step": 2475 + }, + { + "epoch": 0.2684301821335646, + "grad_norm": 3.151616334915161, + "learning_rate": 1.8276512403888006e-05, + "loss": 0.0486, + "step": 2476 + }, + { + "epoch": 0.2685385949696444, + "grad_norm": 4.712758541107178, + "learning_rate": 1.8275787030320618e-05, + "loss": 0.1225, + "step": 2477 + }, + { + "epoch": 0.2686470078057242, + "grad_norm": 1.8320821523666382, + "learning_rate": 1.827506165675323e-05, + "loss": 0.0415, + "step": 2478 + }, + { + "epoch": 0.26875542064180397, + "grad_norm": 4.964414596557617, + "learning_rate": 1.8274336283185843e-05, + "loss": 0.0302, + "step": 2479 + }, + { + "epoch": 0.2688638334778838, + "grad_norm": 6.01250696182251, + "learning_rate": 1.8273610909618455e-05, + "loss": 0.0415, + "step": 2480 + }, + { + "epoch": 0.26897224631396355, + "grad_norm": 1.9502837657928467, + "learning_rate": 1.8272885536051067e-05, + "loss": 0.0271, + "step": 2481 + }, + { + "epoch": 0.26908065915004337, + "grad_norm": 7.469944477081299, + "learning_rate": 1.8272160162483682e-05, + "loss": 0.066, + "step": 2482 + }, + { + "epoch": 0.26918907198612313, + "grad_norm": 9.201247215270996, + "learning_rate": 1.8271434788916294e-05, + "loss": 0.0381, + "step": 2483 + }, + { + "epoch": 0.26929748482220295, + "grad_norm": 2.621553897857666, + "learning_rate": 1.8270709415348907e-05, + "loss": 0.0498, + "step": 2484 + }, + { + "epoch": 0.26940589765828277, + "grad_norm": 4.831352233886719, + "learning_rate": 1.826998404178152e-05, + "loss": 0.1573, + "step": 2485 + }, + { + "epoch": 0.26951431049436253, + "grad_norm": 4.184063911437988, + "learning_rate": 1.826925866821413e-05, + "loss": 0.0545, + "step": 2486 + }, + { + "epoch": 0.26962272333044235, + "grad_norm": 8.304553031921387, + "learning_rate": 1.8268533294646743e-05, + "loss": 0.0887, + "step": 2487 + }, + { + "epoch": 0.2697311361665221, + "grad_norm": 3.558180809020996, + "learning_rate": 1.8267807921079355e-05, + "loss": 0.0456, + "step": 2488 + }, + { + "epoch": 0.26983954900260193, + "grad_norm": 3.214787483215332, + "learning_rate": 1.826708254751197e-05, + "loss": 0.0428, + "step": 2489 + }, + { + "epoch": 0.2699479618386817, + "grad_norm": 6.089509963989258, + "learning_rate": 1.8266357173944583e-05, + "loss": 0.1026, + "step": 2490 + }, + { + "epoch": 0.2700563746747615, + "grad_norm": 7.302911281585693, + "learning_rate": 1.8265631800377195e-05, + "loss": 0.0566, + "step": 2491 + }, + { + "epoch": 0.2701647875108413, + "grad_norm": 2.042600154876709, + "learning_rate": 1.8264906426809807e-05, + "loss": 0.0397, + "step": 2492 + }, + { + "epoch": 0.2702732003469211, + "grad_norm": 4.619688987731934, + "learning_rate": 1.8264181053242423e-05, + "loss": 0.0892, + "step": 2493 + }, + { + "epoch": 0.27038161318300086, + "grad_norm": 1.2966798543930054, + "learning_rate": 1.8263455679675035e-05, + "loss": 0.0411, + "step": 2494 + }, + { + "epoch": 0.2704900260190807, + "grad_norm": 3.247410535812378, + "learning_rate": 1.8262730306107647e-05, + "loss": 0.0343, + "step": 2495 + }, + { + "epoch": 0.27059843885516044, + "grad_norm": 4.060296058654785, + "learning_rate": 1.826200493254026e-05, + "loss": 0.0637, + "step": 2496 + }, + { + "epoch": 0.27070685169124026, + "grad_norm": 2.347334861755371, + "learning_rate": 1.8261279558972875e-05, + "loss": 0.0556, + "step": 2497 + }, + { + "epoch": 0.27081526452732, + "grad_norm": 7.6914191246032715, + "learning_rate": 1.8260554185405487e-05, + "loss": 0.0477, + "step": 2498 + }, + { + "epoch": 0.27092367736339984, + "grad_norm": 1.6765750646591187, + "learning_rate": 1.82598288118381e-05, + "loss": 0.0348, + "step": 2499 + }, + { + "epoch": 0.2710320901994796, + "grad_norm": 2.7027065753936768, + "learning_rate": 1.825910343827071e-05, + "loss": 0.072, + "step": 2500 + }, + { + "epoch": 0.2711405030355594, + "grad_norm": 1.523539423942566, + "learning_rate": 1.8258378064703323e-05, + "loss": 0.0446, + "step": 2501 + }, + { + "epoch": 0.2712489158716392, + "grad_norm": 2.4230129718780518, + "learning_rate": 1.8257652691135935e-05, + "loss": 0.0435, + "step": 2502 + }, + { + "epoch": 0.271357328707719, + "grad_norm": 4.489567756652832, + "learning_rate": 1.8256927317568548e-05, + "loss": 0.0899, + "step": 2503 + }, + { + "epoch": 0.27146574154379877, + "grad_norm": 5.218160629272461, + "learning_rate": 1.8256201944001163e-05, + "loss": 0.0864, + "step": 2504 + }, + { + "epoch": 0.2715741543798786, + "grad_norm": 4.614497661590576, + "learning_rate": 1.8255476570433775e-05, + "loss": 0.0722, + "step": 2505 + }, + { + "epoch": 0.27168256721595835, + "grad_norm": 1.973706603050232, + "learning_rate": 1.8254751196866387e-05, + "loss": 0.0238, + "step": 2506 + }, + { + "epoch": 0.27179098005203817, + "grad_norm": 3.6512839794158936, + "learning_rate": 1.8254025823299e-05, + "loss": 0.0508, + "step": 2507 + }, + { + "epoch": 0.27189939288811793, + "grad_norm": 3.800687551498413, + "learning_rate": 1.825330044973161e-05, + "loss": 0.0587, + "step": 2508 + }, + { + "epoch": 0.27200780572419775, + "grad_norm": 3.28627610206604, + "learning_rate": 1.8252575076164224e-05, + "loss": 0.0732, + "step": 2509 + }, + { + "epoch": 0.2721162185602775, + "grad_norm": 4.730267524719238, + "learning_rate": 1.825184970259684e-05, + "loss": 0.0348, + "step": 2510 + }, + { + "epoch": 0.27222463139635733, + "grad_norm": 2.6280651092529297, + "learning_rate": 1.825112432902945e-05, + "loss": 0.0647, + "step": 2511 + }, + { + "epoch": 0.2723330442324371, + "grad_norm": 8.280287742614746, + "learning_rate": 1.8250398955462067e-05, + "loss": 0.0445, + "step": 2512 + }, + { + "epoch": 0.2724414570685169, + "grad_norm": 2.4648942947387695, + "learning_rate": 1.824967358189468e-05, + "loss": 0.0483, + "step": 2513 + }, + { + "epoch": 0.27254986990459673, + "grad_norm": 2.8684823513031006, + "learning_rate": 1.824894820832729e-05, + "loss": 0.0547, + "step": 2514 + }, + { + "epoch": 0.2726582827406765, + "grad_norm": 3.0501596927642822, + "learning_rate": 1.8248222834759903e-05, + "loss": 0.0676, + "step": 2515 + }, + { + "epoch": 0.2727666955767563, + "grad_norm": 2.445772171020508, + "learning_rate": 1.8247497461192516e-05, + "loss": 0.0596, + "step": 2516 + }, + { + "epoch": 0.2728751084128361, + "grad_norm": 2.0656120777130127, + "learning_rate": 1.8246772087625128e-05, + "loss": 0.026, + "step": 2517 + }, + { + "epoch": 0.2729835212489159, + "grad_norm": 1.9432532787322998, + "learning_rate": 1.824604671405774e-05, + "loss": 0.0259, + "step": 2518 + }, + { + "epoch": 0.27309193408499566, + "grad_norm": 1.9033833742141724, + "learning_rate": 1.8245321340490355e-05, + "loss": 0.0261, + "step": 2519 + }, + { + "epoch": 0.2732003469210755, + "grad_norm": 1.443785548210144, + "learning_rate": 1.8244595966922968e-05, + "loss": 0.0259, + "step": 2520 + }, + { + "epoch": 0.27330875975715524, + "grad_norm": 2.1901633739471436, + "learning_rate": 1.824387059335558e-05, + "loss": 0.0615, + "step": 2521 + }, + { + "epoch": 0.27341717259323506, + "grad_norm": 2.532289743423462, + "learning_rate": 1.8243145219788192e-05, + "loss": 0.0343, + "step": 2522 + }, + { + "epoch": 0.2735255854293148, + "grad_norm": 3.7761621475219727, + "learning_rate": 1.8242419846220804e-05, + "loss": 0.0382, + "step": 2523 + }, + { + "epoch": 0.27363399826539464, + "grad_norm": 2.293026924133301, + "learning_rate": 1.8241694472653416e-05, + "loss": 0.0236, + "step": 2524 + }, + { + "epoch": 0.2737424111014744, + "grad_norm": 2.3315114974975586, + "learning_rate": 1.824096909908603e-05, + "loss": 0.0265, + "step": 2525 + }, + { + "epoch": 0.2738508239375542, + "grad_norm": 6.289303302764893, + "learning_rate": 1.8240243725518644e-05, + "loss": 0.0607, + "step": 2526 + }, + { + "epoch": 0.273959236773634, + "grad_norm": 1.34465491771698, + "learning_rate": 1.8239518351951256e-05, + "loss": 0.0318, + "step": 2527 + }, + { + "epoch": 0.2740676496097138, + "grad_norm": 5.39076566696167, + "learning_rate": 1.823879297838387e-05, + "loss": 0.0604, + "step": 2528 + }, + { + "epoch": 0.27417606244579357, + "grad_norm": 1.2862411737442017, + "learning_rate": 1.8238067604816484e-05, + "loss": 0.0121, + "step": 2529 + }, + { + "epoch": 0.2742844752818734, + "grad_norm": 2.4105703830718994, + "learning_rate": 1.8237342231249096e-05, + "loss": 0.0502, + "step": 2530 + }, + { + "epoch": 0.27439288811795315, + "grad_norm": 3.0230889320373535, + "learning_rate": 1.8236616857681708e-05, + "loss": 0.0388, + "step": 2531 + }, + { + "epoch": 0.27450130095403297, + "grad_norm": 6.403781890869141, + "learning_rate": 1.823589148411432e-05, + "loss": 0.0656, + "step": 2532 + }, + { + "epoch": 0.27460971379011273, + "grad_norm": 3.843397378921509, + "learning_rate": 1.8235166110546932e-05, + "loss": 0.0231, + "step": 2533 + }, + { + "epoch": 0.27471812662619255, + "grad_norm": 3.716244697570801, + "learning_rate": 1.8234440736979548e-05, + "loss": 0.0609, + "step": 2534 + }, + { + "epoch": 0.2748265394622723, + "grad_norm": 4.07695198059082, + "learning_rate": 1.823371536341216e-05, + "loss": 0.0679, + "step": 2535 + }, + { + "epoch": 0.27493495229835213, + "grad_norm": 9.052422523498535, + "learning_rate": 1.8232989989844772e-05, + "loss": 0.0513, + "step": 2536 + }, + { + "epoch": 0.2750433651344319, + "grad_norm": 4.084612846374512, + "learning_rate": 1.8232264616277384e-05, + "loss": 0.0462, + "step": 2537 + }, + { + "epoch": 0.2751517779705117, + "grad_norm": 1.4497929811477661, + "learning_rate": 1.8231539242709996e-05, + "loss": 0.0685, + "step": 2538 + }, + { + "epoch": 0.2752601908065915, + "grad_norm": 6.921525478363037, + "learning_rate": 1.823081386914261e-05, + "loss": 0.0772, + "step": 2539 + }, + { + "epoch": 0.2753686036426713, + "grad_norm": 6.217686653137207, + "learning_rate": 1.823008849557522e-05, + "loss": 0.0845, + "step": 2540 + }, + { + "epoch": 0.27547701647875106, + "grad_norm": 8.638376235961914, + "learning_rate": 1.8229363122007836e-05, + "loss": 0.0668, + "step": 2541 + }, + { + "epoch": 0.2755854293148309, + "grad_norm": 4.464447975158691, + "learning_rate": 1.8228637748440448e-05, + "loss": 0.1009, + "step": 2542 + }, + { + "epoch": 0.2756938421509107, + "grad_norm": 3.8416812419891357, + "learning_rate": 1.822791237487306e-05, + "loss": 0.044, + "step": 2543 + }, + { + "epoch": 0.27580225498699046, + "grad_norm": 2.607605457305908, + "learning_rate": 1.8227187001305673e-05, + "loss": 0.0365, + "step": 2544 + }, + { + "epoch": 0.2759106678230703, + "grad_norm": 3.600135564804077, + "learning_rate": 1.8226461627738288e-05, + "loss": 0.0312, + "step": 2545 + }, + { + "epoch": 0.27601908065915004, + "grad_norm": 7.398359775543213, + "learning_rate": 1.82257362541709e-05, + "loss": 0.0971, + "step": 2546 + }, + { + "epoch": 0.27612749349522986, + "grad_norm": 2.850698232650757, + "learning_rate": 1.8225010880603512e-05, + "loss": 0.0641, + "step": 2547 + }, + { + "epoch": 0.2762359063313096, + "grad_norm": 2.632554054260254, + "learning_rate": 1.8224285507036124e-05, + "loss": 0.0495, + "step": 2548 + }, + { + "epoch": 0.27634431916738944, + "grad_norm": 6.358157157897949, + "learning_rate": 1.822356013346874e-05, + "loss": 0.1043, + "step": 2549 + }, + { + "epoch": 0.2764527320034692, + "grad_norm": 5.0432538986206055, + "learning_rate": 1.8222834759901352e-05, + "loss": 0.0783, + "step": 2550 + }, + { + "epoch": 0.276561144839549, + "grad_norm": 3.032310962677002, + "learning_rate": 1.8222109386333964e-05, + "loss": 0.0717, + "step": 2551 + }, + { + "epoch": 0.2766695576756288, + "grad_norm": 2.165043830871582, + "learning_rate": 1.8221384012766576e-05, + "loss": 0.0355, + "step": 2552 + }, + { + "epoch": 0.2767779705117086, + "grad_norm": 5.526517868041992, + "learning_rate": 1.822065863919919e-05, + "loss": 0.0603, + "step": 2553 + }, + { + "epoch": 0.27688638334778837, + "grad_norm": 2.638026475906372, + "learning_rate": 1.82199332656318e-05, + "loss": 0.0346, + "step": 2554 + }, + { + "epoch": 0.2769947961838682, + "grad_norm": 2.7358269691467285, + "learning_rate": 1.8219207892064413e-05, + "loss": 0.026, + "step": 2555 + }, + { + "epoch": 0.27710320901994795, + "grad_norm": 2.246046781539917, + "learning_rate": 1.821848251849703e-05, + "loss": 0.0361, + "step": 2556 + }, + { + "epoch": 0.27721162185602777, + "grad_norm": 2.597259759902954, + "learning_rate": 1.821775714492964e-05, + "loss": 0.0755, + "step": 2557 + }, + { + "epoch": 0.27732003469210753, + "grad_norm": 2.4489803314208984, + "learning_rate": 1.8217031771362253e-05, + "loss": 0.0358, + "step": 2558 + }, + { + "epoch": 0.27742844752818735, + "grad_norm": 1.946637749671936, + "learning_rate": 1.8216306397794865e-05, + "loss": 0.0184, + "step": 2559 + }, + { + "epoch": 0.2775368603642671, + "grad_norm": 1.3001161813735962, + "learning_rate": 1.8215581024227477e-05, + "loss": 0.0341, + "step": 2560 + }, + { + "epoch": 0.27764527320034693, + "grad_norm": 2.8983066082000732, + "learning_rate": 1.821485565066009e-05, + "loss": 0.0342, + "step": 2561 + }, + { + "epoch": 0.2777536860364267, + "grad_norm": 5.176815032958984, + "learning_rate": 1.8214130277092705e-05, + "loss": 0.0524, + "step": 2562 + }, + { + "epoch": 0.2778620988725065, + "grad_norm": 2.909076452255249, + "learning_rate": 1.8213404903525317e-05, + "loss": 0.0318, + "step": 2563 + }, + { + "epoch": 0.2779705117085863, + "grad_norm": 1.7005822658538818, + "learning_rate": 1.8212679529957932e-05, + "loss": 0.0625, + "step": 2564 + }, + { + "epoch": 0.2780789245446661, + "grad_norm": 1.4913285970687866, + "learning_rate": 1.8211954156390544e-05, + "loss": 0.0178, + "step": 2565 + }, + { + "epoch": 0.27818733738074586, + "grad_norm": 7.079832553863525, + "learning_rate": 1.8211228782823157e-05, + "loss": 0.0539, + "step": 2566 + }, + { + "epoch": 0.2782957502168257, + "grad_norm": 2.142212152481079, + "learning_rate": 1.821050340925577e-05, + "loss": 0.0404, + "step": 2567 + }, + { + "epoch": 0.27840416305290544, + "grad_norm": 5.311133861541748, + "learning_rate": 1.820977803568838e-05, + "loss": 0.0446, + "step": 2568 + }, + { + "epoch": 0.27851257588898526, + "grad_norm": 2.243713140487671, + "learning_rate": 1.8209052662120993e-05, + "loss": 0.0529, + "step": 2569 + }, + { + "epoch": 0.278620988725065, + "grad_norm": 4.4479827880859375, + "learning_rate": 1.8208327288553605e-05, + "loss": 0.0862, + "step": 2570 + }, + { + "epoch": 0.27872940156114484, + "grad_norm": 3.166003465652466, + "learning_rate": 1.820760191498622e-05, + "loss": 0.0358, + "step": 2571 + }, + { + "epoch": 0.2788378143972246, + "grad_norm": 2.6864774227142334, + "learning_rate": 1.8206876541418833e-05, + "loss": 0.0831, + "step": 2572 + }, + { + "epoch": 0.2789462272333044, + "grad_norm": 2.1044368743896484, + "learning_rate": 1.8206151167851445e-05, + "loss": 0.0613, + "step": 2573 + }, + { + "epoch": 0.27905464006938424, + "grad_norm": 7.553991794586182, + "learning_rate": 1.8205425794284057e-05, + "loss": 0.0966, + "step": 2574 + }, + { + "epoch": 0.279163052905464, + "grad_norm": 3.4234213829040527, + "learning_rate": 1.820470042071667e-05, + "loss": 0.0423, + "step": 2575 + }, + { + "epoch": 0.2792714657415438, + "grad_norm": 3.719761371612549, + "learning_rate": 1.820397504714928e-05, + "loss": 0.0863, + "step": 2576 + }, + { + "epoch": 0.2793798785776236, + "grad_norm": 3.323247194290161, + "learning_rate": 1.8203249673581897e-05, + "loss": 0.0413, + "step": 2577 + }, + { + "epoch": 0.2794882914137034, + "grad_norm": 2.7849137783050537, + "learning_rate": 1.820252430001451e-05, + "loss": 0.0343, + "step": 2578 + }, + { + "epoch": 0.27959670424978317, + "grad_norm": 3.4700779914855957, + "learning_rate": 1.820179892644712e-05, + "loss": 0.0534, + "step": 2579 + }, + { + "epoch": 0.279705117085863, + "grad_norm": 2.676071882247925, + "learning_rate": 1.8201073552879737e-05, + "loss": 0.0576, + "step": 2580 + }, + { + "epoch": 0.27981352992194275, + "grad_norm": 2.093024253845215, + "learning_rate": 1.820034817931235e-05, + "loss": 0.0371, + "step": 2581 + }, + { + "epoch": 0.27992194275802257, + "grad_norm": 3.1305196285247803, + "learning_rate": 1.819962280574496e-05, + "loss": 0.0515, + "step": 2582 + }, + { + "epoch": 0.28003035559410233, + "grad_norm": 2.6267476081848145, + "learning_rate": 1.8198897432177573e-05, + "loss": 0.0511, + "step": 2583 + }, + { + "epoch": 0.28013876843018215, + "grad_norm": 3.8793375492095947, + "learning_rate": 1.8198172058610185e-05, + "loss": 0.0558, + "step": 2584 + }, + { + "epoch": 0.2802471812662619, + "grad_norm": 3.077454090118408, + "learning_rate": 1.8197446685042797e-05, + "loss": 0.0868, + "step": 2585 + }, + { + "epoch": 0.28035559410234173, + "grad_norm": 1.4785008430480957, + "learning_rate": 1.8196721311475413e-05, + "loss": 0.0428, + "step": 2586 + }, + { + "epoch": 0.2804640069384215, + "grad_norm": 1.2952755689620972, + "learning_rate": 1.8195995937908025e-05, + "loss": 0.0279, + "step": 2587 + }, + { + "epoch": 0.2805724197745013, + "grad_norm": 1.7495770454406738, + "learning_rate": 1.8195270564340637e-05, + "loss": 0.0269, + "step": 2588 + }, + { + "epoch": 0.2806808326105811, + "grad_norm": 2.2570130825042725, + "learning_rate": 1.819454519077325e-05, + "loss": 0.0372, + "step": 2589 + }, + { + "epoch": 0.2807892454466609, + "grad_norm": 3.003675699234009, + "learning_rate": 1.819381981720586e-05, + "loss": 0.0421, + "step": 2590 + }, + { + "epoch": 0.28089765828274066, + "grad_norm": 2.5073959827423096, + "learning_rate": 1.8193094443638474e-05, + "loss": 0.0485, + "step": 2591 + }, + { + "epoch": 0.2810060711188205, + "grad_norm": 1.9155495166778564, + "learning_rate": 1.8192369070071086e-05, + "loss": 0.0379, + "step": 2592 + }, + { + "epoch": 0.28111448395490024, + "grad_norm": 4.457486152648926, + "learning_rate": 1.81916436965037e-05, + "loss": 0.089, + "step": 2593 + }, + { + "epoch": 0.28122289679098006, + "grad_norm": 2.9598662853240967, + "learning_rate": 1.8190918322936314e-05, + "loss": 0.0249, + "step": 2594 + }, + { + "epoch": 0.2813313096270598, + "grad_norm": 2.151339054107666, + "learning_rate": 1.8190192949368926e-05, + "loss": 0.0512, + "step": 2595 + }, + { + "epoch": 0.28143972246313964, + "grad_norm": 2.710585832595825, + "learning_rate": 1.8189467575801538e-05, + "loss": 0.0518, + "step": 2596 + }, + { + "epoch": 0.2815481352992194, + "grad_norm": 1.9489927291870117, + "learning_rate": 1.8188742202234153e-05, + "loss": 0.0513, + "step": 2597 + }, + { + "epoch": 0.2816565481352992, + "grad_norm": 1.3347790241241455, + "learning_rate": 1.8188016828666766e-05, + "loss": 0.0299, + "step": 2598 + }, + { + "epoch": 0.281764960971379, + "grad_norm": 0.8661681413650513, + "learning_rate": 1.8187291455099378e-05, + "loss": 0.0137, + "step": 2599 + }, + { + "epoch": 0.2818733738074588, + "grad_norm": 1.7962040901184082, + "learning_rate": 1.818656608153199e-05, + "loss": 0.0319, + "step": 2600 + }, + { + "epoch": 0.28198178664353857, + "grad_norm": 1.2933194637298584, + "learning_rate": 1.8185840707964605e-05, + "loss": 0.0112, + "step": 2601 + }, + { + "epoch": 0.2820901994796184, + "grad_norm": 3.9906606674194336, + "learning_rate": 1.8185115334397217e-05, + "loss": 0.0406, + "step": 2602 + }, + { + "epoch": 0.2821986123156982, + "grad_norm": 2.802093744277954, + "learning_rate": 1.818438996082983e-05, + "loss": 0.0515, + "step": 2603 + }, + { + "epoch": 0.28230702515177797, + "grad_norm": 2.2156643867492676, + "learning_rate": 1.8183664587262442e-05, + "loss": 0.0456, + "step": 2604 + }, + { + "epoch": 0.2824154379878578, + "grad_norm": 1.364597201347351, + "learning_rate": 1.8182939213695054e-05, + "loss": 0.0223, + "step": 2605 + }, + { + "epoch": 0.28252385082393755, + "grad_norm": 1.9302611351013184, + "learning_rate": 1.8182213840127666e-05, + "loss": 0.0291, + "step": 2606 + }, + { + "epoch": 0.28263226366001737, + "grad_norm": 5.851749420166016, + "learning_rate": 1.8181488466560278e-05, + "loss": 0.0322, + "step": 2607 + }, + { + "epoch": 0.28274067649609713, + "grad_norm": 6.527594566345215, + "learning_rate": 1.8180763092992894e-05, + "loss": 0.0752, + "step": 2608 + }, + { + "epoch": 0.28284908933217695, + "grad_norm": 3.9319911003112793, + "learning_rate": 1.8180037719425506e-05, + "loss": 0.0235, + "step": 2609 + }, + { + "epoch": 0.2829575021682567, + "grad_norm": 5.417104721069336, + "learning_rate": 1.8179312345858118e-05, + "loss": 0.0927, + "step": 2610 + }, + { + "epoch": 0.28306591500433653, + "grad_norm": 2.222146511077881, + "learning_rate": 1.817858697229073e-05, + "loss": 0.0157, + "step": 2611 + }, + { + "epoch": 0.2831743278404163, + "grad_norm": 7.776077747344971, + "learning_rate": 1.8177861598723342e-05, + "loss": 0.1066, + "step": 2612 + }, + { + "epoch": 0.2832827406764961, + "grad_norm": 10.464537620544434, + "learning_rate": 1.8177136225155954e-05, + "loss": 0.0719, + "step": 2613 + }, + { + "epoch": 0.2833911535125759, + "grad_norm": 7.0110578536987305, + "learning_rate": 1.817641085158857e-05, + "loss": 0.094, + "step": 2614 + }, + { + "epoch": 0.2834995663486557, + "grad_norm": 6.373886585235596, + "learning_rate": 1.8175685478021182e-05, + "loss": 0.0615, + "step": 2615 + }, + { + "epoch": 0.28360797918473546, + "grad_norm": 8.801752090454102, + "learning_rate": 1.8174960104453798e-05, + "loss": 0.0752, + "step": 2616 + }, + { + "epoch": 0.2837163920208153, + "grad_norm": 9.843016624450684, + "learning_rate": 1.817423473088641e-05, + "loss": 0.1225, + "step": 2617 + }, + { + "epoch": 0.28382480485689504, + "grad_norm": 7.242090225219727, + "learning_rate": 1.8173509357319022e-05, + "loss": 0.0715, + "step": 2618 + }, + { + "epoch": 0.28393321769297486, + "grad_norm": 6.044967174530029, + "learning_rate": 1.8172783983751634e-05, + "loss": 0.123, + "step": 2619 + }, + { + "epoch": 0.2840416305290546, + "grad_norm": 2.315070867538452, + "learning_rate": 1.8172058610184246e-05, + "loss": 0.0381, + "step": 2620 + }, + { + "epoch": 0.28415004336513444, + "grad_norm": 1.1268924474716187, + "learning_rate": 1.817133323661686e-05, + "loss": 0.0088, + "step": 2621 + }, + { + "epoch": 0.2842584562012142, + "grad_norm": 2.0029571056365967, + "learning_rate": 1.817060786304947e-05, + "loss": 0.0126, + "step": 2622 + }, + { + "epoch": 0.284366869037294, + "grad_norm": 3.1291472911834717, + "learning_rate": 1.8169882489482086e-05, + "loss": 0.0518, + "step": 2623 + }, + { + "epoch": 0.2844752818733738, + "grad_norm": 4.854176044464111, + "learning_rate": 1.8169157115914698e-05, + "loss": 0.0551, + "step": 2624 + }, + { + "epoch": 0.2845836947094536, + "grad_norm": 5.597166061401367, + "learning_rate": 1.816843174234731e-05, + "loss": 0.0408, + "step": 2625 + }, + { + "epoch": 0.28469210754553337, + "grad_norm": 1.8558073043823242, + "learning_rate": 1.8167706368779922e-05, + "loss": 0.0349, + "step": 2626 + }, + { + "epoch": 0.2848005203816132, + "grad_norm": 5.829656600952148, + "learning_rate": 1.8166980995212535e-05, + "loss": 0.0917, + "step": 2627 + }, + { + "epoch": 0.28490893321769295, + "grad_norm": 3.5459413528442383, + "learning_rate": 1.8166255621645147e-05, + "loss": 0.0876, + "step": 2628 + }, + { + "epoch": 0.28501734605377277, + "grad_norm": 3.695038318634033, + "learning_rate": 1.8165530248077762e-05, + "loss": 0.0526, + "step": 2629 + }, + { + "epoch": 0.28512575888985253, + "grad_norm": 2.60090970993042, + "learning_rate": 1.8164804874510374e-05, + "loss": 0.0398, + "step": 2630 + }, + { + "epoch": 0.28523417172593235, + "grad_norm": 8.650062561035156, + "learning_rate": 1.8164079500942987e-05, + "loss": 0.0381, + "step": 2631 + }, + { + "epoch": 0.28534258456201217, + "grad_norm": 1.3815643787384033, + "learning_rate": 1.8163354127375602e-05, + "loss": 0.0319, + "step": 2632 + }, + { + "epoch": 0.28545099739809193, + "grad_norm": 4.263142108917236, + "learning_rate": 1.8162628753808214e-05, + "loss": 0.0523, + "step": 2633 + }, + { + "epoch": 0.28555941023417175, + "grad_norm": 6.232356548309326, + "learning_rate": 1.8161903380240826e-05, + "loss": 0.0519, + "step": 2634 + }, + { + "epoch": 0.2856678230702515, + "grad_norm": 1.0521410703659058, + "learning_rate": 1.816117800667344e-05, + "loss": 0.0215, + "step": 2635 + }, + { + "epoch": 0.28577623590633133, + "grad_norm": 5.020655155181885, + "learning_rate": 1.816045263310605e-05, + "loss": 0.0349, + "step": 2636 + }, + { + "epoch": 0.2858846487424111, + "grad_norm": 2.850043773651123, + "learning_rate": 1.8159727259538663e-05, + "loss": 0.0452, + "step": 2637 + }, + { + "epoch": 0.2859930615784909, + "grad_norm": 5.453893184661865, + "learning_rate": 1.815900188597128e-05, + "loss": 0.0835, + "step": 2638 + }, + { + "epoch": 0.2861014744145707, + "grad_norm": 2.4917709827423096, + "learning_rate": 1.815827651240389e-05, + "loss": 0.0537, + "step": 2639 + }, + { + "epoch": 0.2862098872506505, + "grad_norm": 2.436290979385376, + "learning_rate": 1.8157551138836503e-05, + "loss": 0.0309, + "step": 2640 + }, + { + "epoch": 0.28631830008673026, + "grad_norm": 3.3134469985961914, + "learning_rate": 1.8156825765269115e-05, + "loss": 0.0632, + "step": 2641 + }, + { + "epoch": 0.2864267129228101, + "grad_norm": 4.273314952850342, + "learning_rate": 1.8156100391701727e-05, + "loss": 0.0297, + "step": 2642 + }, + { + "epoch": 0.28653512575888984, + "grad_norm": 4.427077293395996, + "learning_rate": 1.815537501813434e-05, + "loss": 0.0563, + "step": 2643 + }, + { + "epoch": 0.28664353859496966, + "grad_norm": 2.2373673915863037, + "learning_rate": 1.815464964456695e-05, + "loss": 0.0311, + "step": 2644 + }, + { + "epoch": 0.2867519514310494, + "grad_norm": 3.5479629039764404, + "learning_rate": 1.8153924270999567e-05, + "loss": 0.0559, + "step": 2645 + }, + { + "epoch": 0.28686036426712924, + "grad_norm": 4.731746196746826, + "learning_rate": 1.815319889743218e-05, + "loss": 0.0602, + "step": 2646 + }, + { + "epoch": 0.286968777103209, + "grad_norm": 2.112499952316284, + "learning_rate": 1.815247352386479e-05, + "loss": 0.056, + "step": 2647 + }, + { + "epoch": 0.2870771899392888, + "grad_norm": 5.3994550704956055, + "learning_rate": 1.8151748150297403e-05, + "loss": 0.0488, + "step": 2648 + }, + { + "epoch": 0.2871856027753686, + "grad_norm": 4.3254899978637695, + "learning_rate": 1.815102277673002e-05, + "loss": 0.0705, + "step": 2649 + }, + { + "epoch": 0.2872940156114484, + "grad_norm": 2.094089984893799, + "learning_rate": 1.815029740316263e-05, + "loss": 0.0367, + "step": 2650 + }, + { + "epoch": 0.28740242844752817, + "grad_norm": 2.012420892715454, + "learning_rate": 1.8149572029595243e-05, + "loss": 0.0252, + "step": 2651 + }, + { + "epoch": 0.287510841283608, + "grad_norm": 4.598001956939697, + "learning_rate": 1.8148846656027855e-05, + "loss": 0.0642, + "step": 2652 + }, + { + "epoch": 0.28761925411968775, + "grad_norm": 3.4042932987213135, + "learning_rate": 1.814812128246047e-05, + "loss": 0.0412, + "step": 2653 + }, + { + "epoch": 0.28772766695576757, + "grad_norm": 1.7689454555511475, + "learning_rate": 1.8147395908893083e-05, + "loss": 0.0302, + "step": 2654 + }, + { + "epoch": 0.28783607979184733, + "grad_norm": 4.267643928527832, + "learning_rate": 1.8146670535325695e-05, + "loss": 0.0831, + "step": 2655 + }, + { + "epoch": 0.28794449262792715, + "grad_norm": 3.088901996612549, + "learning_rate": 1.8145945161758307e-05, + "loss": 0.0352, + "step": 2656 + }, + { + "epoch": 0.2880529054640069, + "grad_norm": 1.2432576417922974, + "learning_rate": 1.814521978819092e-05, + "loss": 0.0186, + "step": 2657 + }, + { + "epoch": 0.28816131830008673, + "grad_norm": 2.798563241958618, + "learning_rate": 1.814449441462353e-05, + "loss": 0.053, + "step": 2658 + }, + { + "epoch": 0.2882697311361665, + "grad_norm": 3.3555924892425537, + "learning_rate": 1.8143769041056144e-05, + "loss": 0.0431, + "step": 2659 + }, + { + "epoch": 0.2883781439722463, + "grad_norm": 3.4196510314941406, + "learning_rate": 1.814304366748876e-05, + "loss": 0.0589, + "step": 2660 + }, + { + "epoch": 0.28848655680832613, + "grad_norm": 3.4523732662200928, + "learning_rate": 1.814231829392137e-05, + "loss": 0.0611, + "step": 2661 + }, + { + "epoch": 0.2885949696444059, + "grad_norm": 2.528409242630005, + "learning_rate": 1.8141592920353983e-05, + "loss": 0.0434, + "step": 2662 + }, + { + "epoch": 0.2887033824804857, + "grad_norm": 3.0723748207092285, + "learning_rate": 1.8140867546786595e-05, + "loss": 0.0267, + "step": 2663 + }, + { + "epoch": 0.2888117953165655, + "grad_norm": 2.836270332336426, + "learning_rate": 1.8140142173219208e-05, + "loss": 0.0468, + "step": 2664 + }, + { + "epoch": 0.2889202081526453, + "grad_norm": 5.884287357330322, + "learning_rate": 1.813941679965182e-05, + "loss": 0.0741, + "step": 2665 + }, + { + "epoch": 0.28902862098872506, + "grad_norm": 2.408987283706665, + "learning_rate": 1.8138691426084435e-05, + "loss": 0.0285, + "step": 2666 + }, + { + "epoch": 0.2891370338248049, + "grad_norm": 2.729339599609375, + "learning_rate": 1.8137966052517047e-05, + "loss": 0.0326, + "step": 2667 + }, + { + "epoch": 0.28924544666088464, + "grad_norm": 3.1774375438690186, + "learning_rate": 1.8137240678949663e-05, + "loss": 0.0441, + "step": 2668 + }, + { + "epoch": 0.28935385949696446, + "grad_norm": 2.313079833984375, + "learning_rate": 1.8136515305382275e-05, + "loss": 0.0386, + "step": 2669 + }, + { + "epoch": 0.2894622723330442, + "grad_norm": 6.641918659210205, + "learning_rate": 1.8135789931814887e-05, + "loss": 0.0943, + "step": 2670 + }, + { + "epoch": 0.28957068516912404, + "grad_norm": 22.02590560913086, + "learning_rate": 1.81350645582475e-05, + "loss": 0.0177, + "step": 2671 + }, + { + "epoch": 0.2896790980052038, + "grad_norm": 2.8001790046691895, + "learning_rate": 1.813433918468011e-05, + "loss": 0.0441, + "step": 2672 + }, + { + "epoch": 0.2897875108412836, + "grad_norm": 1.7620211839675903, + "learning_rate": 1.8133613811112724e-05, + "loss": 0.0157, + "step": 2673 + }, + { + "epoch": 0.2898959236773634, + "grad_norm": 3.975017786026001, + "learning_rate": 1.8132888437545336e-05, + "loss": 0.0789, + "step": 2674 + }, + { + "epoch": 0.2900043365134432, + "grad_norm": 2.518676280975342, + "learning_rate": 1.813216306397795e-05, + "loss": 0.0609, + "step": 2675 + }, + { + "epoch": 0.29011274934952297, + "grad_norm": 3.7687954902648926, + "learning_rate": 1.8131437690410564e-05, + "loss": 0.0639, + "step": 2676 + }, + { + "epoch": 0.2902211621856028, + "grad_norm": 2.3371047973632812, + "learning_rate": 1.8130712316843176e-05, + "loss": 0.0199, + "step": 2677 + }, + { + "epoch": 0.29032957502168255, + "grad_norm": 4.730618953704834, + "learning_rate": 1.8129986943275788e-05, + "loss": 0.1088, + "step": 2678 + }, + { + "epoch": 0.29043798785776237, + "grad_norm": 5.364319801330566, + "learning_rate": 1.81292615697084e-05, + "loss": 0.0589, + "step": 2679 + }, + { + "epoch": 0.29054640069384213, + "grad_norm": 4.251720905303955, + "learning_rate": 1.8128536196141012e-05, + "loss": 0.0371, + "step": 2680 + }, + { + "epoch": 0.29065481352992195, + "grad_norm": 2.4875080585479736, + "learning_rate": 1.8127810822573628e-05, + "loss": 0.075, + "step": 2681 + }, + { + "epoch": 0.2907632263660017, + "grad_norm": 3.9481050968170166, + "learning_rate": 1.812708544900624e-05, + "loss": 0.0554, + "step": 2682 + }, + { + "epoch": 0.29087163920208153, + "grad_norm": 3.325110673904419, + "learning_rate": 1.8126360075438852e-05, + "loss": 0.0241, + "step": 2683 + }, + { + "epoch": 0.2909800520381613, + "grad_norm": 3.6819632053375244, + "learning_rate": 1.8125634701871467e-05, + "loss": 0.0377, + "step": 2684 + }, + { + "epoch": 0.2910884648742411, + "grad_norm": 2.9935309886932373, + "learning_rate": 1.812490932830408e-05, + "loss": 0.0202, + "step": 2685 + }, + { + "epoch": 0.2911968777103209, + "grad_norm": 1.8158044815063477, + "learning_rate": 1.8124183954736692e-05, + "loss": 0.0404, + "step": 2686 + }, + { + "epoch": 0.2913052905464007, + "grad_norm": 2.395672082901001, + "learning_rate": 1.8123458581169304e-05, + "loss": 0.0528, + "step": 2687 + }, + { + "epoch": 0.29141370338248046, + "grad_norm": 2.829627752304077, + "learning_rate": 1.8122733207601916e-05, + "loss": 0.048, + "step": 2688 + }, + { + "epoch": 0.2915221162185603, + "grad_norm": 7.068665504455566, + "learning_rate": 1.8122007834034528e-05, + "loss": 0.065, + "step": 2689 + }, + { + "epoch": 0.29163052905464004, + "grad_norm": 2.8775832653045654, + "learning_rate": 1.8121282460467144e-05, + "loss": 0.0439, + "step": 2690 + }, + { + "epoch": 0.29173894189071986, + "grad_norm": 2.091799259185791, + "learning_rate": 1.8120557086899756e-05, + "loss": 0.0303, + "step": 2691 + }, + { + "epoch": 0.2918473547267997, + "grad_norm": 1.5365169048309326, + "learning_rate": 1.8119831713332368e-05, + "loss": 0.0276, + "step": 2692 + }, + { + "epoch": 0.29195576756287944, + "grad_norm": 1.2155154943466187, + "learning_rate": 1.811910633976498e-05, + "loss": 0.0174, + "step": 2693 + }, + { + "epoch": 0.29206418039895926, + "grad_norm": 2.065688133239746, + "learning_rate": 1.8118380966197592e-05, + "loss": 0.0406, + "step": 2694 + }, + { + "epoch": 0.292172593235039, + "grad_norm": 2.814603328704834, + "learning_rate": 1.8117655592630204e-05, + "loss": 0.0816, + "step": 2695 + }, + { + "epoch": 0.29228100607111884, + "grad_norm": 2.14913272857666, + "learning_rate": 1.8116930219062817e-05, + "loss": 0.028, + "step": 2696 + }, + { + "epoch": 0.2923894189071986, + "grad_norm": 4.172898292541504, + "learning_rate": 1.8116204845495432e-05, + "loss": 0.0319, + "step": 2697 + }, + { + "epoch": 0.2924978317432784, + "grad_norm": 2.4132423400878906, + "learning_rate": 1.8115479471928044e-05, + "loss": 0.0607, + "step": 2698 + }, + { + "epoch": 0.2926062445793582, + "grad_norm": 6.480194568634033, + "learning_rate": 1.8114754098360656e-05, + "loss": 0.0383, + "step": 2699 + }, + { + "epoch": 0.292714657415438, + "grad_norm": 4.149620056152344, + "learning_rate": 1.8114028724793272e-05, + "loss": 0.1151, + "step": 2700 + }, + { + "epoch": 0.29282307025151777, + "grad_norm": 4.324495315551758, + "learning_rate": 1.8113303351225884e-05, + "loss": 0.0678, + "step": 2701 + }, + { + "epoch": 0.2929314830875976, + "grad_norm": 3.2434475421905518, + "learning_rate": 1.8112577977658496e-05, + "loss": 0.036, + "step": 2702 + }, + { + "epoch": 0.29303989592367735, + "grad_norm": 4.054433345794678, + "learning_rate": 1.811185260409111e-05, + "loss": 0.0681, + "step": 2703 + }, + { + "epoch": 0.29314830875975717, + "grad_norm": 1.7715122699737549, + "learning_rate": 1.811112723052372e-05, + "loss": 0.0228, + "step": 2704 + }, + { + "epoch": 0.29325672159583693, + "grad_norm": 1.8995107412338257, + "learning_rate": 1.8110401856956336e-05, + "loss": 0.0403, + "step": 2705 + }, + { + "epoch": 0.29336513443191675, + "grad_norm": 1.7865849733352661, + "learning_rate": 1.8109676483388948e-05, + "loss": 0.036, + "step": 2706 + }, + { + "epoch": 0.2934735472679965, + "grad_norm": 2.513542890548706, + "learning_rate": 1.810895110982156e-05, + "loss": 0.0344, + "step": 2707 + }, + { + "epoch": 0.29358196010407633, + "grad_norm": 3.7069718837738037, + "learning_rate": 1.8108225736254172e-05, + "loss": 0.0781, + "step": 2708 + }, + { + "epoch": 0.2936903729401561, + "grad_norm": 3.13982892036438, + "learning_rate": 1.8107500362686785e-05, + "loss": 0.0484, + "step": 2709 + }, + { + "epoch": 0.2937987857762359, + "grad_norm": 4.787164211273193, + "learning_rate": 1.8106774989119397e-05, + "loss": 0.0558, + "step": 2710 + }, + { + "epoch": 0.2939071986123157, + "grad_norm": 5.934209823608398, + "learning_rate": 1.810604961555201e-05, + "loss": 0.0719, + "step": 2711 + }, + { + "epoch": 0.2940156114483955, + "grad_norm": 6.8322834968566895, + "learning_rate": 1.8105324241984624e-05, + "loss": 0.052, + "step": 2712 + }, + { + "epoch": 0.29412402428447526, + "grad_norm": 2.446206569671631, + "learning_rate": 1.8104598868417237e-05, + "loss": 0.0404, + "step": 2713 + }, + { + "epoch": 0.2942324371205551, + "grad_norm": 2.7047781944274902, + "learning_rate": 1.810387349484985e-05, + "loss": 0.0305, + "step": 2714 + }, + { + "epoch": 0.29434084995663484, + "grad_norm": 1.7110971212387085, + "learning_rate": 1.810314812128246e-05, + "loss": 0.0271, + "step": 2715 + }, + { + "epoch": 0.29444926279271466, + "grad_norm": 6.875162124633789, + "learning_rate": 1.8102422747715073e-05, + "loss": 0.0773, + "step": 2716 + }, + { + "epoch": 0.2945576756287944, + "grad_norm": 2.7776384353637695, + "learning_rate": 1.8101697374147685e-05, + "loss": 0.0275, + "step": 2717 + }, + { + "epoch": 0.29466608846487424, + "grad_norm": 3.3218636512756348, + "learning_rate": 1.81009720005803e-05, + "loss": 0.037, + "step": 2718 + }, + { + "epoch": 0.294774501300954, + "grad_norm": 1.1214587688446045, + "learning_rate": 1.8100246627012913e-05, + "loss": 0.0316, + "step": 2719 + }, + { + "epoch": 0.2948829141370338, + "grad_norm": 2.7029755115509033, + "learning_rate": 1.809952125344553e-05, + "loss": 0.0565, + "step": 2720 + }, + { + "epoch": 0.29499132697311364, + "grad_norm": 4.994269847869873, + "learning_rate": 1.809879587987814e-05, + "loss": 0.0535, + "step": 2721 + }, + { + "epoch": 0.2950997398091934, + "grad_norm": 4.248020648956299, + "learning_rate": 1.8098070506310753e-05, + "loss": 0.0646, + "step": 2722 + }, + { + "epoch": 0.2952081526452732, + "grad_norm": 1.1808912754058838, + "learning_rate": 1.8097345132743365e-05, + "loss": 0.0111, + "step": 2723 + }, + { + "epoch": 0.295316565481353, + "grad_norm": 1.294560194015503, + "learning_rate": 1.8096619759175977e-05, + "loss": 0.0198, + "step": 2724 + }, + { + "epoch": 0.2954249783174328, + "grad_norm": 2.6676747798919678, + "learning_rate": 1.809589438560859e-05, + "loss": 0.0759, + "step": 2725 + }, + { + "epoch": 0.29553339115351257, + "grad_norm": 3.925802230834961, + "learning_rate": 1.80951690120412e-05, + "loss": 0.0589, + "step": 2726 + }, + { + "epoch": 0.2956418039895924, + "grad_norm": 4.14885139465332, + "learning_rate": 1.8094443638473817e-05, + "loss": 0.0596, + "step": 2727 + }, + { + "epoch": 0.29575021682567215, + "grad_norm": 5.222335338592529, + "learning_rate": 1.809371826490643e-05, + "loss": 0.0817, + "step": 2728 + }, + { + "epoch": 0.29585862966175197, + "grad_norm": 1.877341628074646, + "learning_rate": 1.809299289133904e-05, + "loss": 0.0188, + "step": 2729 + }, + { + "epoch": 0.29596704249783173, + "grad_norm": 4.406148910522461, + "learning_rate": 1.8092267517771653e-05, + "loss": 0.0754, + "step": 2730 + }, + { + "epoch": 0.29607545533391155, + "grad_norm": 3.54126238822937, + "learning_rate": 1.8091542144204265e-05, + "loss": 0.0558, + "step": 2731 + }, + { + "epoch": 0.2961838681699913, + "grad_norm": 3.0025851726531982, + "learning_rate": 1.8090816770636877e-05, + "loss": 0.0372, + "step": 2732 + }, + { + "epoch": 0.29629228100607113, + "grad_norm": 3.003028392791748, + "learning_rate": 1.8090091397069493e-05, + "loss": 0.0304, + "step": 2733 + }, + { + "epoch": 0.2964006938421509, + "grad_norm": 3.6703381538391113, + "learning_rate": 1.8089366023502105e-05, + "loss": 0.0557, + "step": 2734 + }, + { + "epoch": 0.2965091066782307, + "grad_norm": 2.179842710494995, + "learning_rate": 1.8088640649934717e-05, + "loss": 0.0387, + "step": 2735 + }, + { + "epoch": 0.2966175195143105, + "grad_norm": 1.7411400079727173, + "learning_rate": 1.8087915276367333e-05, + "loss": 0.0351, + "step": 2736 + }, + { + "epoch": 0.2967259323503903, + "grad_norm": 3.695512533187866, + "learning_rate": 1.8087189902799945e-05, + "loss": 0.0495, + "step": 2737 + }, + { + "epoch": 0.29683434518647006, + "grad_norm": 4.545815944671631, + "learning_rate": 1.8086464529232557e-05, + "loss": 0.0617, + "step": 2738 + }, + { + "epoch": 0.2969427580225499, + "grad_norm": 1.44680655002594, + "learning_rate": 1.808573915566517e-05, + "loss": 0.0352, + "step": 2739 + }, + { + "epoch": 0.29705117085862964, + "grad_norm": 2.3725550174713135, + "learning_rate": 1.808501378209778e-05, + "loss": 0.0472, + "step": 2740 + }, + { + "epoch": 0.29715958369470946, + "grad_norm": 9.24361801147461, + "learning_rate": 1.8084288408530393e-05, + "loss": 0.1267, + "step": 2741 + }, + { + "epoch": 0.2972679965307892, + "grad_norm": 1.5953083038330078, + "learning_rate": 1.808356303496301e-05, + "loss": 0.0315, + "step": 2742 + }, + { + "epoch": 0.29737640936686904, + "grad_norm": 1.9331481456756592, + "learning_rate": 1.808283766139562e-05, + "loss": 0.0224, + "step": 2743 + }, + { + "epoch": 0.2974848222029488, + "grad_norm": 2.972411870956421, + "learning_rate": 1.8082112287828233e-05, + "loss": 0.0469, + "step": 2744 + }, + { + "epoch": 0.2975932350390286, + "grad_norm": 3.1890957355499268, + "learning_rate": 1.8081386914260845e-05, + "loss": 0.0382, + "step": 2745 + }, + { + "epoch": 0.2977016478751084, + "grad_norm": 2.2546956539154053, + "learning_rate": 1.8080661540693458e-05, + "loss": 0.0422, + "step": 2746 + }, + { + "epoch": 0.2978100607111882, + "grad_norm": 0.7488752007484436, + "learning_rate": 1.807993616712607e-05, + "loss": 0.0107, + "step": 2747 + }, + { + "epoch": 0.29791847354726797, + "grad_norm": 6.759978771209717, + "learning_rate": 1.8079210793558682e-05, + "loss": 0.0386, + "step": 2748 + }, + { + "epoch": 0.2980268863833478, + "grad_norm": 3.928243637084961, + "learning_rate": 1.8078485419991297e-05, + "loss": 0.0379, + "step": 2749 + }, + { + "epoch": 0.2981352992194276, + "grad_norm": 2.880223035812378, + "learning_rate": 1.807776004642391e-05, + "loss": 0.0503, + "step": 2750 + }, + { + "epoch": 0.29824371205550737, + "grad_norm": 2.637664556503296, + "learning_rate": 1.8077034672856522e-05, + "loss": 0.0343, + "step": 2751 + }, + { + "epoch": 0.2983521248915872, + "grad_norm": 3.8994617462158203, + "learning_rate": 1.8076309299289137e-05, + "loss": 0.0617, + "step": 2752 + }, + { + "epoch": 0.29846053772766695, + "grad_norm": 2.7850663661956787, + "learning_rate": 1.807558392572175e-05, + "loss": 0.0374, + "step": 2753 + }, + { + "epoch": 0.29856895056374677, + "grad_norm": 4.050201892852783, + "learning_rate": 1.807485855215436e-05, + "loss": 0.0325, + "step": 2754 + }, + { + "epoch": 0.29867736339982653, + "grad_norm": 5.100451946258545, + "learning_rate": 1.8074133178586974e-05, + "loss": 0.021, + "step": 2755 + }, + { + "epoch": 0.29878577623590635, + "grad_norm": 0.598237931728363, + "learning_rate": 1.8073407805019586e-05, + "loss": 0.006, + "step": 2756 + }, + { + "epoch": 0.2988941890719861, + "grad_norm": 5.640381336212158, + "learning_rate": 1.80726824314522e-05, + "loss": 0.0651, + "step": 2757 + }, + { + "epoch": 0.29900260190806593, + "grad_norm": 2.9098281860351562, + "learning_rate": 1.8071957057884813e-05, + "loss": 0.0571, + "step": 2758 + }, + { + "epoch": 0.2991110147441457, + "grad_norm": 5.880359649658203, + "learning_rate": 1.8071231684317426e-05, + "loss": 0.0552, + "step": 2759 + }, + { + "epoch": 0.2992194275802255, + "grad_norm": 2.597628593444824, + "learning_rate": 1.8070506310750038e-05, + "loss": 0.0374, + "step": 2760 + }, + { + "epoch": 0.2993278404163053, + "grad_norm": 4.103765487670898, + "learning_rate": 1.806978093718265e-05, + "loss": 0.0695, + "step": 2761 + }, + { + "epoch": 0.2994362532523851, + "grad_norm": 8.185319900512695, + "learning_rate": 1.8069055563615262e-05, + "loss": 0.0813, + "step": 2762 + }, + { + "epoch": 0.29954466608846486, + "grad_norm": 7.285505294799805, + "learning_rate": 1.8068330190047874e-05, + "loss": 0.0495, + "step": 2763 + }, + { + "epoch": 0.2996530789245447, + "grad_norm": 1.1463921070098877, + "learning_rate": 1.806760481648049e-05, + "loss": 0.0244, + "step": 2764 + }, + { + "epoch": 0.29976149176062444, + "grad_norm": 4.741563320159912, + "learning_rate": 1.8066879442913102e-05, + "loss": 0.0559, + "step": 2765 + }, + { + "epoch": 0.29986990459670426, + "grad_norm": 1.6366832256317139, + "learning_rate": 1.8066154069345714e-05, + "loss": 0.024, + "step": 2766 + }, + { + "epoch": 0.299978317432784, + "grad_norm": 2.6488068103790283, + "learning_rate": 1.8065428695778326e-05, + "loss": 0.0499, + "step": 2767 + }, + { + "epoch": 0.30008673026886384, + "grad_norm": 2.2173948287963867, + "learning_rate": 1.806470332221094e-05, + "loss": 0.0251, + "step": 2768 + }, + { + "epoch": 0.3001951431049436, + "grad_norm": 2.6002278327941895, + "learning_rate": 1.806397794864355e-05, + "loss": 0.0499, + "step": 2769 + }, + { + "epoch": 0.3003035559410234, + "grad_norm": 6.3993330001831055, + "learning_rate": 1.8063252575076166e-05, + "loss": 0.1156, + "step": 2770 + }, + { + "epoch": 0.3004119687771032, + "grad_norm": 5.029718399047852, + "learning_rate": 1.8062527201508778e-05, + "loss": 0.0498, + "step": 2771 + }, + { + "epoch": 0.300520381613183, + "grad_norm": 4.232873916625977, + "learning_rate": 1.8061801827941394e-05, + "loss": 0.0511, + "step": 2772 + }, + { + "epoch": 0.30062879444926277, + "grad_norm": 4.513326168060303, + "learning_rate": 1.8061076454374006e-05, + "loss": 0.0557, + "step": 2773 + }, + { + "epoch": 0.3007372072853426, + "grad_norm": 2.577080011367798, + "learning_rate": 1.8060351080806618e-05, + "loss": 0.0372, + "step": 2774 + }, + { + "epoch": 0.30084562012142235, + "grad_norm": 5.3611674308776855, + "learning_rate": 1.805962570723923e-05, + "loss": 0.1122, + "step": 2775 + }, + { + "epoch": 0.30095403295750217, + "grad_norm": 4.446436405181885, + "learning_rate": 1.8058900333671842e-05, + "loss": 0.0741, + "step": 2776 + }, + { + "epoch": 0.30106244579358193, + "grad_norm": 2.993177652359009, + "learning_rate": 1.8058174960104454e-05, + "loss": 0.0468, + "step": 2777 + }, + { + "epoch": 0.30117085862966175, + "grad_norm": 2.340867757797241, + "learning_rate": 1.8057449586537067e-05, + "loss": 0.0298, + "step": 2778 + }, + { + "epoch": 0.3012792714657415, + "grad_norm": 2.338702440261841, + "learning_rate": 1.8056724212969682e-05, + "loss": 0.0757, + "step": 2779 + }, + { + "epoch": 0.30138768430182133, + "grad_norm": 7.450332164764404, + "learning_rate": 1.8055998839402294e-05, + "loss": 0.0523, + "step": 2780 + }, + { + "epoch": 0.30149609713790115, + "grad_norm": 5.938622951507568, + "learning_rate": 1.8055273465834906e-05, + "loss": 0.0697, + "step": 2781 + }, + { + "epoch": 0.3016045099739809, + "grad_norm": 3.3223111629486084, + "learning_rate": 1.805454809226752e-05, + "loss": 0.0305, + "step": 2782 + }, + { + "epoch": 0.30171292281006074, + "grad_norm": 2.4595565795898438, + "learning_rate": 1.805382271870013e-05, + "loss": 0.024, + "step": 2783 + }, + { + "epoch": 0.3018213356461405, + "grad_norm": 2.384594678878784, + "learning_rate": 1.8053097345132743e-05, + "loss": 0.0557, + "step": 2784 + }, + { + "epoch": 0.3019297484822203, + "grad_norm": 3.3088557720184326, + "learning_rate": 1.8052371971565358e-05, + "loss": 0.0773, + "step": 2785 + }, + { + "epoch": 0.3020381613183001, + "grad_norm": 1.5578701496124268, + "learning_rate": 1.805164659799797e-05, + "loss": 0.0333, + "step": 2786 + }, + { + "epoch": 0.3021465741543799, + "grad_norm": 1.8880977630615234, + "learning_rate": 1.8050921224430583e-05, + "loss": 0.0392, + "step": 2787 + }, + { + "epoch": 0.30225498699045966, + "grad_norm": 2.3590712547302246, + "learning_rate": 1.8050195850863198e-05, + "loss": 0.0531, + "step": 2788 + }, + { + "epoch": 0.3023633998265395, + "grad_norm": 1.8194111585617065, + "learning_rate": 1.804947047729581e-05, + "loss": 0.0324, + "step": 2789 + }, + { + "epoch": 0.30247181266261924, + "grad_norm": 2.6061787605285645, + "learning_rate": 1.8048745103728422e-05, + "loss": 0.0345, + "step": 2790 + }, + { + "epoch": 0.30258022549869906, + "grad_norm": 1.2972450256347656, + "learning_rate": 1.8048019730161035e-05, + "loss": 0.0189, + "step": 2791 + }, + { + "epoch": 0.3026886383347788, + "grad_norm": 1.7948589324951172, + "learning_rate": 1.8047294356593647e-05, + "loss": 0.0264, + "step": 2792 + }, + { + "epoch": 0.30279705117085864, + "grad_norm": 1.7610241174697876, + "learning_rate": 1.804656898302626e-05, + "loss": 0.0264, + "step": 2793 + }, + { + "epoch": 0.3029054640069384, + "grad_norm": 1.8861041069030762, + "learning_rate": 1.8045843609458874e-05, + "loss": 0.0465, + "step": 2794 + }, + { + "epoch": 0.3030138768430182, + "grad_norm": 4.046802520751953, + "learning_rate": 1.8045118235891487e-05, + "loss": 0.0664, + "step": 2795 + }, + { + "epoch": 0.303122289679098, + "grad_norm": 4.846681594848633, + "learning_rate": 1.80443928623241e-05, + "loss": 0.0912, + "step": 2796 + }, + { + "epoch": 0.3032307025151778, + "grad_norm": 2.6214141845703125, + "learning_rate": 1.804366748875671e-05, + "loss": 0.0536, + "step": 2797 + }, + { + "epoch": 0.30333911535125757, + "grad_norm": 2.534022569656372, + "learning_rate": 1.8042942115189323e-05, + "loss": 0.0339, + "step": 2798 + }, + { + "epoch": 0.3034475281873374, + "grad_norm": 7.250705242156982, + "learning_rate": 1.8042216741621935e-05, + "loss": 0.0602, + "step": 2799 + }, + { + "epoch": 0.30355594102341715, + "grad_norm": 4.468280792236328, + "learning_rate": 1.804149136805455e-05, + "loss": 0.0494, + "step": 2800 + }, + { + "epoch": 0.30366435385949697, + "grad_norm": 3.499136209487915, + "learning_rate": 1.8040765994487163e-05, + "loss": 0.0612, + "step": 2801 + }, + { + "epoch": 0.30377276669557673, + "grad_norm": 2.8660120964050293, + "learning_rate": 1.8040040620919775e-05, + "loss": 0.047, + "step": 2802 + }, + { + "epoch": 0.30388117953165655, + "grad_norm": 1.9491503238677979, + "learning_rate": 1.8039315247352387e-05, + "loss": 0.0282, + "step": 2803 + }, + { + "epoch": 0.3039895923677363, + "grad_norm": 1.3784984350204468, + "learning_rate": 1.8038589873785003e-05, + "loss": 0.0153, + "step": 2804 + }, + { + "epoch": 0.30409800520381614, + "grad_norm": 3.1096386909484863, + "learning_rate": 1.8037864500217615e-05, + "loss": 0.0386, + "step": 2805 + }, + { + "epoch": 0.3042064180398959, + "grad_norm": 1.2828596830368042, + "learning_rate": 1.8037139126650227e-05, + "loss": 0.0226, + "step": 2806 + }, + { + "epoch": 0.3043148308759757, + "grad_norm": 1.516862154006958, + "learning_rate": 1.803641375308284e-05, + "loss": 0.0289, + "step": 2807 + }, + { + "epoch": 0.3044232437120555, + "grad_norm": 6.439659595489502, + "learning_rate": 1.803568837951545e-05, + "loss": 0.0791, + "step": 2808 + }, + { + "epoch": 0.3045316565481353, + "grad_norm": 3.961717367172241, + "learning_rate": 1.8034963005948067e-05, + "loss": 0.0844, + "step": 2809 + }, + { + "epoch": 0.3046400693842151, + "grad_norm": 5.044033050537109, + "learning_rate": 1.803423763238068e-05, + "loss": 0.0487, + "step": 2810 + }, + { + "epoch": 0.3047484822202949, + "grad_norm": 8.70350170135498, + "learning_rate": 1.803351225881329e-05, + "loss": 0.0922, + "step": 2811 + }, + { + "epoch": 0.3048568950563747, + "grad_norm": 1.6034809350967407, + "learning_rate": 1.8032786885245903e-05, + "loss": 0.0439, + "step": 2812 + }, + { + "epoch": 0.30496530789245446, + "grad_norm": 0.5838194489479065, + "learning_rate": 1.8032061511678515e-05, + "loss": 0.0079, + "step": 2813 + }, + { + "epoch": 0.3050737207285343, + "grad_norm": 12.586624145507812, + "learning_rate": 1.8031336138111127e-05, + "loss": 0.0494, + "step": 2814 + }, + { + "epoch": 0.30518213356461404, + "grad_norm": 4.010959625244141, + "learning_rate": 1.803061076454374e-05, + "loss": 0.0451, + "step": 2815 + }, + { + "epoch": 0.30529054640069386, + "grad_norm": 5.331350326538086, + "learning_rate": 1.8029885390976355e-05, + "loss": 0.0613, + "step": 2816 + }, + { + "epoch": 0.3053989592367736, + "grad_norm": 2.217766523361206, + "learning_rate": 1.8029160017408967e-05, + "loss": 0.0311, + "step": 2817 + }, + { + "epoch": 0.30550737207285344, + "grad_norm": 3.595163345336914, + "learning_rate": 1.802843464384158e-05, + "loss": 0.0438, + "step": 2818 + }, + { + "epoch": 0.3056157849089332, + "grad_norm": 4.4717817306518555, + "learning_rate": 1.802770927027419e-05, + "loss": 0.0542, + "step": 2819 + }, + { + "epoch": 0.305724197745013, + "grad_norm": 2.318575620651245, + "learning_rate": 1.8026983896706804e-05, + "loss": 0.0311, + "step": 2820 + }, + { + "epoch": 0.3058326105810928, + "grad_norm": 2.2239186763763428, + "learning_rate": 1.802625852313942e-05, + "loss": 0.0188, + "step": 2821 + }, + { + "epoch": 0.3059410234171726, + "grad_norm": 3.8725008964538574, + "learning_rate": 1.802553314957203e-05, + "loss": 0.0415, + "step": 2822 + }, + { + "epoch": 0.30604943625325237, + "grad_norm": 2.7929577827453613, + "learning_rate": 1.8024807776004643e-05, + "loss": 0.0307, + "step": 2823 + }, + { + "epoch": 0.3061578490893322, + "grad_norm": 4.172501087188721, + "learning_rate": 1.802408240243726e-05, + "loss": 0.098, + "step": 2824 + }, + { + "epoch": 0.30626626192541195, + "grad_norm": 5.611998081207275, + "learning_rate": 1.802335702886987e-05, + "loss": 0.0728, + "step": 2825 + }, + { + "epoch": 0.30637467476149177, + "grad_norm": 2.4071743488311768, + "learning_rate": 1.8022631655302483e-05, + "loss": 0.0152, + "step": 2826 + }, + { + "epoch": 0.30648308759757154, + "grad_norm": 4.229058742523193, + "learning_rate": 1.8021906281735095e-05, + "loss": 0.0839, + "step": 2827 + }, + { + "epoch": 0.30659150043365135, + "grad_norm": 3.5584442615509033, + "learning_rate": 1.8021180908167708e-05, + "loss": 0.0919, + "step": 2828 + }, + { + "epoch": 0.3066999132697311, + "grad_norm": 1.6346691846847534, + "learning_rate": 1.802045553460032e-05, + "loss": 0.026, + "step": 2829 + }, + { + "epoch": 0.30680832610581094, + "grad_norm": 2.5905961990356445, + "learning_rate": 1.8019730161032932e-05, + "loss": 0.0325, + "step": 2830 + }, + { + "epoch": 0.3069167389418907, + "grad_norm": 2.6979482173919678, + "learning_rate": 1.8019004787465547e-05, + "loss": 0.0403, + "step": 2831 + }, + { + "epoch": 0.3070251517779705, + "grad_norm": 1.5987675189971924, + "learning_rate": 1.801827941389816e-05, + "loss": 0.0194, + "step": 2832 + }, + { + "epoch": 0.3071335646140503, + "grad_norm": 3.6367714405059814, + "learning_rate": 1.801755404033077e-05, + "loss": 0.0498, + "step": 2833 + }, + { + "epoch": 0.3072419774501301, + "grad_norm": 2.075155258178711, + "learning_rate": 1.8016828666763384e-05, + "loss": 0.0445, + "step": 2834 + }, + { + "epoch": 0.30735039028620986, + "grad_norm": 5.127031326293945, + "learning_rate": 1.8016103293195996e-05, + "loss": 0.0688, + "step": 2835 + }, + { + "epoch": 0.3074588031222897, + "grad_norm": 3.776430130004883, + "learning_rate": 1.8015377919628608e-05, + "loss": 0.0524, + "step": 2836 + }, + { + "epoch": 0.30756721595836944, + "grad_norm": 3.175305128097534, + "learning_rate": 1.8014652546061224e-05, + "loss": 0.0408, + "step": 2837 + }, + { + "epoch": 0.30767562879444926, + "grad_norm": 2.6590120792388916, + "learning_rate": 1.8013927172493836e-05, + "loss": 0.0557, + "step": 2838 + }, + { + "epoch": 0.3077840416305291, + "grad_norm": 3.296558141708374, + "learning_rate": 1.8013201798926448e-05, + "loss": 0.0675, + "step": 2839 + }, + { + "epoch": 0.30789245446660884, + "grad_norm": 1.3074535131454468, + "learning_rate": 1.8012476425359063e-05, + "loss": 0.0223, + "step": 2840 + }, + { + "epoch": 0.30800086730268866, + "grad_norm": 0.9660930037498474, + "learning_rate": 1.8011751051791676e-05, + "loss": 0.02, + "step": 2841 + }, + { + "epoch": 0.3081092801387684, + "grad_norm": 3.3757529258728027, + "learning_rate": 1.8011025678224288e-05, + "loss": 0.0468, + "step": 2842 + }, + { + "epoch": 0.30821769297484825, + "grad_norm": 4.020369052886963, + "learning_rate": 1.80103003046569e-05, + "loss": 0.1299, + "step": 2843 + }, + { + "epoch": 0.308326105810928, + "grad_norm": 1.130661964416504, + "learning_rate": 1.8009574931089512e-05, + "loss": 0.0176, + "step": 2844 + }, + { + "epoch": 0.3084345186470078, + "grad_norm": 1.185036063194275, + "learning_rate": 1.8008849557522124e-05, + "loss": 0.0182, + "step": 2845 + }, + { + "epoch": 0.3085429314830876, + "grad_norm": 1.5978769063949585, + "learning_rate": 1.800812418395474e-05, + "loss": 0.0268, + "step": 2846 + }, + { + "epoch": 0.3086513443191674, + "grad_norm": 3.5574162006378174, + "learning_rate": 1.8007398810387352e-05, + "loss": 0.0782, + "step": 2847 + }, + { + "epoch": 0.30875975715524717, + "grad_norm": 4.238571643829346, + "learning_rate": 1.8006673436819964e-05, + "loss": 0.0178, + "step": 2848 + }, + { + "epoch": 0.308868169991327, + "grad_norm": 1.3754823207855225, + "learning_rate": 1.8005948063252576e-05, + "loss": 0.0349, + "step": 2849 + }, + { + "epoch": 0.30897658282740675, + "grad_norm": 2.6823651790618896, + "learning_rate": 1.8005222689685188e-05, + "loss": 0.032, + "step": 2850 + }, + { + "epoch": 0.3090849956634866, + "grad_norm": 2.534153699874878, + "learning_rate": 1.80044973161178e-05, + "loss": 0.0507, + "step": 2851 + }, + { + "epoch": 0.30919340849956634, + "grad_norm": 3.1575052738189697, + "learning_rate": 1.8003771942550416e-05, + "loss": 0.0232, + "step": 2852 + }, + { + "epoch": 0.30930182133564615, + "grad_norm": 5.102913856506348, + "learning_rate": 1.8003046568983028e-05, + "loss": 0.1209, + "step": 2853 + }, + { + "epoch": 0.3094102341717259, + "grad_norm": 2.097338914871216, + "learning_rate": 1.800232119541564e-05, + "loss": 0.0382, + "step": 2854 + }, + { + "epoch": 0.30951864700780574, + "grad_norm": 1.4713904857635498, + "learning_rate": 1.8001595821848252e-05, + "loss": 0.0139, + "step": 2855 + }, + { + "epoch": 0.3096270598438855, + "grad_norm": 1.040427803993225, + "learning_rate": 1.8000870448280868e-05, + "loss": 0.016, + "step": 2856 + }, + { + "epoch": 0.3097354726799653, + "grad_norm": 1.2041738033294678, + "learning_rate": 1.800014507471348e-05, + "loss": 0.0096, + "step": 2857 + }, + { + "epoch": 0.3098438855160451, + "grad_norm": 3.784162998199463, + "learning_rate": 1.7999419701146092e-05, + "loss": 0.0391, + "step": 2858 + }, + { + "epoch": 0.3099522983521249, + "grad_norm": 4.17588472366333, + "learning_rate": 1.7998694327578704e-05, + "loss": 0.0201, + "step": 2859 + }, + { + "epoch": 0.31006071118820466, + "grad_norm": 5.501913070678711, + "learning_rate": 1.7997968954011316e-05, + "loss": 0.0992, + "step": 2860 + }, + { + "epoch": 0.3101691240242845, + "grad_norm": 2.427274703979492, + "learning_rate": 1.7997243580443932e-05, + "loss": 0.0315, + "step": 2861 + }, + { + "epoch": 0.31027753686036424, + "grad_norm": 3.1575663089752197, + "learning_rate": 1.7996518206876544e-05, + "loss": 0.0367, + "step": 2862 + }, + { + "epoch": 0.31038594969644406, + "grad_norm": 1.5710277557373047, + "learning_rate": 1.7995792833309156e-05, + "loss": 0.0117, + "step": 2863 + }, + { + "epoch": 0.3104943625325238, + "grad_norm": 6.825112342834473, + "learning_rate": 1.799506745974177e-05, + "loss": 0.0424, + "step": 2864 + }, + { + "epoch": 0.31060277536860365, + "grad_norm": 4.163990020751953, + "learning_rate": 1.799434208617438e-05, + "loss": 0.0216, + "step": 2865 + }, + { + "epoch": 0.3107111882046834, + "grad_norm": 9.276862144470215, + "learning_rate": 1.7993616712606993e-05, + "loss": 0.0749, + "step": 2866 + }, + { + "epoch": 0.3108196010407632, + "grad_norm": 8.016409873962402, + "learning_rate": 1.7992891339039605e-05, + "loss": 0.0535, + "step": 2867 + }, + { + "epoch": 0.31092801387684305, + "grad_norm": 11.0970458984375, + "learning_rate": 1.799216596547222e-05, + "loss": 0.087, + "step": 2868 + }, + { + "epoch": 0.3110364267129228, + "grad_norm": 3.824552059173584, + "learning_rate": 1.7991440591904833e-05, + "loss": 0.0517, + "step": 2869 + }, + { + "epoch": 0.3111448395490026, + "grad_norm": 2.086381196975708, + "learning_rate": 1.7990715218337445e-05, + "loss": 0.0163, + "step": 2870 + }, + { + "epoch": 0.3112532523850824, + "grad_norm": 3.7943966388702393, + "learning_rate": 1.7989989844770057e-05, + "loss": 0.0494, + "step": 2871 + }, + { + "epoch": 0.3113616652211622, + "grad_norm": 1.837120532989502, + "learning_rate": 1.798926447120267e-05, + "loss": 0.0214, + "step": 2872 + }, + { + "epoch": 0.311470078057242, + "grad_norm": 1.4752265214920044, + "learning_rate": 1.7988539097635285e-05, + "loss": 0.0229, + "step": 2873 + }, + { + "epoch": 0.3115784908933218, + "grad_norm": 4.1291022300720215, + "learning_rate": 1.7987813724067897e-05, + "loss": 0.0808, + "step": 2874 + }, + { + "epoch": 0.31168690372940155, + "grad_norm": 7.474968433380127, + "learning_rate": 1.798708835050051e-05, + "loss": 0.0445, + "step": 2875 + }, + { + "epoch": 0.3117953165654814, + "grad_norm": 2.7031748294830322, + "learning_rate": 1.7986362976933124e-05, + "loss": 0.0234, + "step": 2876 + }, + { + "epoch": 0.31190372940156114, + "grad_norm": 2.320929765701294, + "learning_rate": 1.7985637603365736e-05, + "loss": 0.0145, + "step": 2877 + }, + { + "epoch": 0.31201214223764095, + "grad_norm": 3.4343624114990234, + "learning_rate": 1.798491222979835e-05, + "loss": 0.0448, + "step": 2878 + }, + { + "epoch": 0.3121205550737207, + "grad_norm": 3.044365644454956, + "learning_rate": 1.798418685623096e-05, + "loss": 0.0253, + "step": 2879 + }, + { + "epoch": 0.31222896790980054, + "grad_norm": 2.8634397983551025, + "learning_rate": 1.7983461482663573e-05, + "loss": 0.0748, + "step": 2880 + }, + { + "epoch": 0.3123373807458803, + "grad_norm": 5.656635761260986, + "learning_rate": 1.7982736109096185e-05, + "loss": 0.0641, + "step": 2881 + }, + { + "epoch": 0.3124457935819601, + "grad_norm": 10.960258483886719, + "learning_rate": 1.7982010735528797e-05, + "loss": 0.0754, + "step": 2882 + }, + { + "epoch": 0.3125542064180399, + "grad_norm": 5.31209659576416, + "learning_rate": 1.7981285361961413e-05, + "loss": 0.1065, + "step": 2883 + }, + { + "epoch": 0.3126626192541197, + "grad_norm": 2.886282205581665, + "learning_rate": 1.7980559988394025e-05, + "loss": 0.0373, + "step": 2884 + }, + { + "epoch": 0.31277103209019946, + "grad_norm": 1.8267766237258911, + "learning_rate": 1.7979834614826637e-05, + "loss": 0.0237, + "step": 2885 + }, + { + "epoch": 0.3128794449262793, + "grad_norm": 3.70416259765625, + "learning_rate": 1.797910924125925e-05, + "loss": 0.0537, + "step": 2886 + }, + { + "epoch": 0.31298785776235905, + "grad_norm": 1.553178310394287, + "learning_rate": 1.797838386769186e-05, + "loss": 0.0304, + "step": 2887 + }, + { + "epoch": 0.31309627059843886, + "grad_norm": 7.119046688079834, + "learning_rate": 1.7977658494124473e-05, + "loss": 0.052, + "step": 2888 + }, + { + "epoch": 0.3132046834345186, + "grad_norm": 1.485065221786499, + "learning_rate": 1.797693312055709e-05, + "loss": 0.0271, + "step": 2889 + }, + { + "epoch": 0.31331309627059845, + "grad_norm": 2.4208316802978516, + "learning_rate": 1.79762077469897e-05, + "loss": 0.0359, + "step": 2890 + }, + { + "epoch": 0.3134215091066782, + "grad_norm": 4.675154685974121, + "learning_rate": 1.7975482373422313e-05, + "loss": 0.0342, + "step": 2891 + }, + { + "epoch": 0.313529921942758, + "grad_norm": 1.9820630550384521, + "learning_rate": 1.797475699985493e-05, + "loss": 0.0465, + "step": 2892 + }, + { + "epoch": 0.3136383347788378, + "grad_norm": 4.973825454711914, + "learning_rate": 1.797403162628754e-05, + "loss": 0.0242, + "step": 2893 + }, + { + "epoch": 0.3137467476149176, + "grad_norm": 4.302704811096191, + "learning_rate": 1.7973306252720153e-05, + "loss": 0.0721, + "step": 2894 + }, + { + "epoch": 0.3138551604509974, + "grad_norm": 2.0394997596740723, + "learning_rate": 1.7972580879152765e-05, + "loss": 0.0217, + "step": 2895 + }, + { + "epoch": 0.3139635732870772, + "grad_norm": 1.7885782718658447, + "learning_rate": 1.7971855505585377e-05, + "loss": 0.0349, + "step": 2896 + }, + { + "epoch": 0.31407198612315695, + "grad_norm": 2.8304598331451416, + "learning_rate": 1.797113013201799e-05, + "loss": 0.023, + "step": 2897 + }, + { + "epoch": 0.3141803989592368, + "grad_norm": 3.475273847579956, + "learning_rate": 1.7970404758450605e-05, + "loss": 0.029, + "step": 2898 + }, + { + "epoch": 0.3142888117953166, + "grad_norm": 3.6099698543548584, + "learning_rate": 1.7969679384883217e-05, + "loss": 0.0715, + "step": 2899 + }, + { + "epoch": 0.31439722463139635, + "grad_norm": 1.6953214406967163, + "learning_rate": 1.796895401131583e-05, + "loss": 0.0328, + "step": 2900 + }, + { + "epoch": 0.3145056374674762, + "grad_norm": 1.3215000629425049, + "learning_rate": 1.796822863774844e-05, + "loss": 0.0507, + "step": 2901 + }, + { + "epoch": 0.31461405030355594, + "grad_norm": 5.262498378753662, + "learning_rate": 1.7967503264181054e-05, + "loss": 0.0578, + "step": 2902 + }, + { + "epoch": 0.31472246313963576, + "grad_norm": 2.591495990753174, + "learning_rate": 1.7966777890613666e-05, + "loss": 0.0631, + "step": 2903 + }, + { + "epoch": 0.3148308759757155, + "grad_norm": 2.321777105331421, + "learning_rate": 1.796605251704628e-05, + "loss": 0.0501, + "step": 2904 + }, + { + "epoch": 0.31493928881179534, + "grad_norm": 2.215830087661743, + "learning_rate": 1.7965327143478893e-05, + "loss": 0.0391, + "step": 2905 + }, + { + "epoch": 0.3150477016478751, + "grad_norm": 3.4126386642456055, + "learning_rate": 1.7964601769911506e-05, + "loss": 0.0762, + "step": 2906 + }, + { + "epoch": 0.3151561144839549, + "grad_norm": 2.3281030654907227, + "learning_rate": 1.7963876396344118e-05, + "loss": 0.0825, + "step": 2907 + }, + { + "epoch": 0.3152645273200347, + "grad_norm": 2.219346046447754, + "learning_rate": 1.7963151022776733e-05, + "loss": 0.033, + "step": 2908 + }, + { + "epoch": 0.3153729401561145, + "grad_norm": 1.1067653894424438, + "learning_rate": 1.7962425649209345e-05, + "loss": 0.0249, + "step": 2909 + }, + { + "epoch": 0.31548135299219426, + "grad_norm": 3.454185724258423, + "learning_rate": 1.7961700275641958e-05, + "loss": 0.083, + "step": 2910 + }, + { + "epoch": 0.3155897658282741, + "grad_norm": 3.2124338150024414, + "learning_rate": 1.796097490207457e-05, + "loss": 0.0728, + "step": 2911 + }, + { + "epoch": 0.31569817866435385, + "grad_norm": 1.6985036134719849, + "learning_rate": 1.7960249528507182e-05, + "loss": 0.0243, + "step": 2912 + }, + { + "epoch": 0.31580659150043366, + "grad_norm": 1.2598493099212646, + "learning_rate": 1.7959524154939797e-05, + "loss": 0.0297, + "step": 2913 + }, + { + "epoch": 0.3159150043365134, + "grad_norm": 3.236572027206421, + "learning_rate": 1.795879878137241e-05, + "loss": 0.0658, + "step": 2914 + }, + { + "epoch": 0.31602341717259325, + "grad_norm": 3.732858180999756, + "learning_rate": 1.795807340780502e-05, + "loss": 0.0633, + "step": 2915 + }, + { + "epoch": 0.316131830008673, + "grad_norm": 2.9318790435791016, + "learning_rate": 1.7957348034237634e-05, + "loss": 0.0618, + "step": 2916 + }, + { + "epoch": 0.31624024284475283, + "grad_norm": 3.1884052753448486, + "learning_rate": 1.7956622660670246e-05, + "loss": 0.0611, + "step": 2917 + }, + { + "epoch": 0.3163486556808326, + "grad_norm": 2.758434772491455, + "learning_rate": 1.7955897287102858e-05, + "loss": 0.0392, + "step": 2918 + }, + { + "epoch": 0.3164570685169124, + "grad_norm": 1.2895241975784302, + "learning_rate": 1.795517191353547e-05, + "loss": 0.0238, + "step": 2919 + }, + { + "epoch": 0.3165654813529922, + "grad_norm": 0.9784451127052307, + "learning_rate": 1.7954446539968086e-05, + "loss": 0.0216, + "step": 2920 + }, + { + "epoch": 0.316673894189072, + "grad_norm": 2.75930118560791, + "learning_rate": 1.7953721166400698e-05, + "loss": 0.0474, + "step": 2921 + }, + { + "epoch": 0.31678230702515175, + "grad_norm": 1.9185564517974854, + "learning_rate": 1.795299579283331e-05, + "loss": 0.0301, + "step": 2922 + }, + { + "epoch": 0.3168907198612316, + "grad_norm": 3.561739444732666, + "learning_rate": 1.7952270419265922e-05, + "loss": 0.0712, + "step": 2923 + }, + { + "epoch": 0.31699913269731134, + "grad_norm": 3.0660581588745117, + "learning_rate": 1.7951545045698534e-05, + "loss": 0.0426, + "step": 2924 + }, + { + "epoch": 0.31710754553339116, + "grad_norm": 4.939038276672363, + "learning_rate": 1.795081967213115e-05, + "loss": 0.0605, + "step": 2925 + }, + { + "epoch": 0.3172159583694709, + "grad_norm": 3.4801037311553955, + "learning_rate": 1.7950094298563762e-05, + "loss": 0.0841, + "step": 2926 + }, + { + "epoch": 0.31732437120555074, + "grad_norm": 2.4079477787017822, + "learning_rate": 1.7949368924996374e-05, + "loss": 0.0125, + "step": 2927 + }, + { + "epoch": 0.31743278404163056, + "grad_norm": 3.0939948558807373, + "learning_rate": 1.794864355142899e-05, + "loss": 0.04, + "step": 2928 + }, + { + "epoch": 0.3175411968777103, + "grad_norm": 4.5976057052612305, + "learning_rate": 1.7947918177861602e-05, + "loss": 0.0233, + "step": 2929 + }, + { + "epoch": 0.31764960971379014, + "grad_norm": 4.016524314880371, + "learning_rate": 1.7947192804294214e-05, + "loss": 0.0519, + "step": 2930 + }, + { + "epoch": 0.3177580225498699, + "grad_norm": 5.907998085021973, + "learning_rate": 1.7946467430726826e-05, + "loss": 0.0405, + "step": 2931 + }, + { + "epoch": 0.3178664353859497, + "grad_norm": 1.4855796098709106, + "learning_rate": 1.7945742057159438e-05, + "loss": 0.0168, + "step": 2932 + }, + { + "epoch": 0.3179748482220295, + "grad_norm": 0.8351123929023743, + "learning_rate": 1.794501668359205e-05, + "loss": 0.0131, + "step": 2933 + }, + { + "epoch": 0.3180832610581093, + "grad_norm": 3.1989355087280273, + "learning_rate": 1.7944291310024663e-05, + "loss": 0.0667, + "step": 2934 + }, + { + "epoch": 0.31819167389418906, + "grad_norm": 3.424467086791992, + "learning_rate": 1.7943565936457278e-05, + "loss": 0.0511, + "step": 2935 + }, + { + "epoch": 0.3183000867302689, + "grad_norm": 2.357269763946533, + "learning_rate": 1.794284056288989e-05, + "loss": 0.0331, + "step": 2936 + }, + { + "epoch": 0.31840849956634865, + "grad_norm": 2.84177827835083, + "learning_rate": 1.7942115189322502e-05, + "loss": 0.0526, + "step": 2937 + }, + { + "epoch": 0.31851691240242846, + "grad_norm": 2.4654855728149414, + "learning_rate": 1.7941389815755114e-05, + "loss": 0.0178, + "step": 2938 + }, + { + "epoch": 0.31862532523850823, + "grad_norm": 7.649999141693115, + "learning_rate": 1.7940664442187727e-05, + "loss": 0.048, + "step": 2939 + }, + { + "epoch": 0.31873373807458805, + "grad_norm": 7.321631908416748, + "learning_rate": 1.793993906862034e-05, + "loss": 0.0714, + "step": 2940 + }, + { + "epoch": 0.3188421509106678, + "grad_norm": 4.072824954986572, + "learning_rate": 1.7939213695052954e-05, + "loss": 0.0409, + "step": 2941 + }, + { + "epoch": 0.31895056374674763, + "grad_norm": 2.3085782527923584, + "learning_rate": 1.7938488321485566e-05, + "loss": 0.0168, + "step": 2942 + }, + { + "epoch": 0.3190589765828274, + "grad_norm": 1.7296218872070312, + "learning_rate": 1.793776294791818e-05, + "loss": 0.0143, + "step": 2943 + }, + { + "epoch": 0.3191673894189072, + "grad_norm": 4.156371593475342, + "learning_rate": 1.7937037574350794e-05, + "loss": 0.0597, + "step": 2944 + }, + { + "epoch": 0.319275802254987, + "grad_norm": 6.03984260559082, + "learning_rate": 1.7936312200783406e-05, + "loss": 0.0936, + "step": 2945 + }, + { + "epoch": 0.3193842150910668, + "grad_norm": 0.6078132390975952, + "learning_rate": 1.793558682721602e-05, + "loss": 0.0079, + "step": 2946 + }, + { + "epoch": 0.31949262792714656, + "grad_norm": 4.271247863769531, + "learning_rate": 1.793486145364863e-05, + "loss": 0.0515, + "step": 2947 + }, + { + "epoch": 0.3196010407632264, + "grad_norm": 2.0998611450195312, + "learning_rate": 1.7934136080081243e-05, + "loss": 0.0153, + "step": 2948 + }, + { + "epoch": 0.31970945359930614, + "grad_norm": 1.237442135810852, + "learning_rate": 1.7933410706513855e-05, + "loss": 0.0074, + "step": 2949 + }, + { + "epoch": 0.31981786643538596, + "grad_norm": 3.3924078941345215, + "learning_rate": 1.793268533294647e-05, + "loss": 0.055, + "step": 2950 + }, + { + "epoch": 0.3199262792714657, + "grad_norm": 5.871874809265137, + "learning_rate": 1.7931959959379083e-05, + "loss": 0.1486, + "step": 2951 + }, + { + "epoch": 0.32003469210754554, + "grad_norm": 12.394488334655762, + "learning_rate": 1.7931234585811695e-05, + "loss": 0.1186, + "step": 2952 + }, + { + "epoch": 0.3201431049436253, + "grad_norm": 0.3033674359321594, + "learning_rate": 1.7930509212244307e-05, + "loss": 0.0027, + "step": 2953 + }, + { + "epoch": 0.3202515177797051, + "grad_norm": 1.6128790378570557, + "learning_rate": 1.792978383867692e-05, + "loss": 0.0198, + "step": 2954 + }, + { + "epoch": 0.3203599306157849, + "grad_norm": 6.039405345916748, + "learning_rate": 1.792905846510953e-05, + "loss": 0.0349, + "step": 2955 + }, + { + "epoch": 0.3204683434518647, + "grad_norm": 2.5146799087524414, + "learning_rate": 1.7928333091542147e-05, + "loss": 0.0108, + "step": 2956 + }, + { + "epoch": 0.3205767562879445, + "grad_norm": 4.036270618438721, + "learning_rate": 1.792760771797476e-05, + "loss": 0.0465, + "step": 2957 + }, + { + "epoch": 0.3206851691240243, + "grad_norm": 2.9852724075317383, + "learning_rate": 1.792688234440737e-05, + "loss": 0.037, + "step": 2958 + }, + { + "epoch": 0.3207935819601041, + "grad_norm": 4.133852005004883, + "learning_rate": 1.7926156970839983e-05, + "loss": 0.0405, + "step": 2959 + }, + { + "epoch": 0.32090199479618386, + "grad_norm": 4.345728874206543, + "learning_rate": 1.79254315972726e-05, + "loss": 0.0609, + "step": 2960 + }, + { + "epoch": 0.3210104076322637, + "grad_norm": 2.4048972129821777, + "learning_rate": 1.792470622370521e-05, + "loss": 0.0299, + "step": 2961 + }, + { + "epoch": 0.32111882046834345, + "grad_norm": 1.8166451454162598, + "learning_rate": 1.7923980850137823e-05, + "loss": 0.0411, + "step": 2962 + }, + { + "epoch": 0.32122723330442327, + "grad_norm": 2.5389673709869385, + "learning_rate": 1.7923255476570435e-05, + "loss": 0.0353, + "step": 2963 + }, + { + "epoch": 0.32133564614050303, + "grad_norm": 5.549444675445557, + "learning_rate": 1.7922530103003047e-05, + "loss": 0.0919, + "step": 2964 + }, + { + "epoch": 0.32144405897658285, + "grad_norm": 1.686328411102295, + "learning_rate": 1.7921804729435663e-05, + "loss": 0.0429, + "step": 2965 + }, + { + "epoch": 0.3215524718126626, + "grad_norm": 5.541773319244385, + "learning_rate": 1.7921079355868275e-05, + "loss": 0.0304, + "step": 2966 + }, + { + "epoch": 0.32166088464874243, + "grad_norm": 4.068087577819824, + "learning_rate": 1.7920353982300887e-05, + "loss": 0.0497, + "step": 2967 + }, + { + "epoch": 0.3217692974848222, + "grad_norm": 3.414395809173584, + "learning_rate": 1.79196286087335e-05, + "loss": 0.0268, + "step": 2968 + }, + { + "epoch": 0.321877710320902, + "grad_norm": 1.5241007804870605, + "learning_rate": 1.791890323516611e-05, + "loss": 0.0201, + "step": 2969 + }, + { + "epoch": 0.3219861231569818, + "grad_norm": 2.5876576900482178, + "learning_rate": 1.7918177861598723e-05, + "loss": 0.0564, + "step": 2970 + }, + { + "epoch": 0.3220945359930616, + "grad_norm": 5.622854709625244, + "learning_rate": 1.7917452488031336e-05, + "loss": 0.0578, + "step": 2971 + }, + { + "epoch": 0.32220294882914136, + "grad_norm": 5.983704566955566, + "learning_rate": 1.791672711446395e-05, + "loss": 0.0985, + "step": 2972 + }, + { + "epoch": 0.3223113616652212, + "grad_norm": 5.366800308227539, + "learning_rate": 1.7916001740896563e-05, + "loss": 0.0534, + "step": 2973 + }, + { + "epoch": 0.32241977450130094, + "grad_norm": 4.08053731918335, + "learning_rate": 1.7915276367329175e-05, + "loss": 0.0582, + "step": 2974 + }, + { + "epoch": 0.32252818733738076, + "grad_norm": 1.0293821096420288, + "learning_rate": 1.7914550993761788e-05, + "loss": 0.0163, + "step": 2975 + }, + { + "epoch": 0.3226366001734605, + "grad_norm": 2.3414723873138428, + "learning_rate": 1.79138256201944e-05, + "loss": 0.0482, + "step": 2976 + }, + { + "epoch": 0.32274501300954034, + "grad_norm": 8.341785430908203, + "learning_rate": 1.7913100246627015e-05, + "loss": 0.0691, + "step": 2977 + }, + { + "epoch": 0.3228534258456201, + "grad_norm": 1.2722586393356323, + "learning_rate": 1.7912374873059627e-05, + "loss": 0.0194, + "step": 2978 + }, + { + "epoch": 0.3229618386816999, + "grad_norm": 1.830697774887085, + "learning_rate": 1.791164949949224e-05, + "loss": 0.0251, + "step": 2979 + }, + { + "epoch": 0.3230702515177797, + "grad_norm": 6.702472686767578, + "learning_rate": 1.7910924125924855e-05, + "loss": 0.0556, + "step": 2980 + }, + { + "epoch": 0.3231786643538595, + "grad_norm": 5.3454484939575195, + "learning_rate": 1.7910198752357467e-05, + "loss": 0.0176, + "step": 2981 + }, + { + "epoch": 0.32328707718993926, + "grad_norm": 3.0587825775146484, + "learning_rate": 1.790947337879008e-05, + "loss": 0.0426, + "step": 2982 + }, + { + "epoch": 0.3233954900260191, + "grad_norm": 1.5504213571548462, + "learning_rate": 1.790874800522269e-05, + "loss": 0.0238, + "step": 2983 + }, + { + "epoch": 0.32350390286209885, + "grad_norm": 3.4017744064331055, + "learning_rate": 1.7908022631655304e-05, + "loss": 0.0612, + "step": 2984 + }, + { + "epoch": 0.32361231569817867, + "grad_norm": 1.5850796699523926, + "learning_rate": 1.7907297258087916e-05, + "loss": 0.0413, + "step": 2985 + }, + { + "epoch": 0.3237207285342585, + "grad_norm": 8.211823463439941, + "learning_rate": 1.7906571884520528e-05, + "loss": 0.0402, + "step": 2986 + }, + { + "epoch": 0.32382914137033825, + "grad_norm": 2.956798553466797, + "learning_rate": 1.7905846510953143e-05, + "loss": 0.0543, + "step": 2987 + }, + { + "epoch": 0.32393755420641807, + "grad_norm": 6.955246925354004, + "learning_rate": 1.7905121137385756e-05, + "loss": 0.0644, + "step": 2988 + }, + { + "epoch": 0.32404596704249783, + "grad_norm": 14.473323822021484, + "learning_rate": 1.7904395763818368e-05, + "loss": 0.0445, + "step": 2989 + }, + { + "epoch": 0.32415437987857765, + "grad_norm": 4.653624057769775, + "learning_rate": 1.790367039025098e-05, + "loss": 0.073, + "step": 2990 + }, + { + "epoch": 0.3242627927146574, + "grad_norm": 1.4972823858261108, + "learning_rate": 1.7902945016683592e-05, + "loss": 0.0326, + "step": 2991 + }, + { + "epoch": 0.32437120555073723, + "grad_norm": 2.9644243717193604, + "learning_rate": 1.7902219643116204e-05, + "loss": 0.0281, + "step": 2992 + }, + { + "epoch": 0.324479618386817, + "grad_norm": 5.25039529800415, + "learning_rate": 1.790149426954882e-05, + "loss": 0.0895, + "step": 2993 + }, + { + "epoch": 0.3245880312228968, + "grad_norm": 5.695943355560303, + "learning_rate": 1.7900768895981432e-05, + "loss": 0.052, + "step": 2994 + }, + { + "epoch": 0.3246964440589766, + "grad_norm": 1.6875602006912231, + "learning_rate": 1.7900043522414047e-05, + "loss": 0.0411, + "step": 2995 + }, + { + "epoch": 0.3248048568950564, + "grad_norm": 3.0386693477630615, + "learning_rate": 1.789931814884666e-05, + "loss": 0.0511, + "step": 2996 + }, + { + "epoch": 0.32491326973113616, + "grad_norm": 2.944190502166748, + "learning_rate": 1.789859277527927e-05, + "loss": 0.0422, + "step": 2997 + }, + { + "epoch": 0.325021682567216, + "grad_norm": 1.0942585468292236, + "learning_rate": 1.7897867401711884e-05, + "loss": 0.0158, + "step": 2998 + }, + { + "epoch": 0.32513009540329574, + "grad_norm": 3.0617268085479736, + "learning_rate": 1.7897142028144496e-05, + "loss": 0.0639, + "step": 2999 + }, + { + "epoch": 0.32523850823937556, + "grad_norm": 1.3362783193588257, + "learning_rate": 1.7896416654577108e-05, + "loss": 0.036, + "step": 3000 + }, + { + "epoch": 0.3253469210754553, + "grad_norm": 1.8693199157714844, + "learning_rate": 1.789569128100972e-05, + "loss": 0.0354, + "step": 3001 + }, + { + "epoch": 0.32545533391153514, + "grad_norm": 19.724964141845703, + "learning_rate": 1.7894965907442336e-05, + "loss": 0.083, + "step": 3002 + }, + { + "epoch": 0.3255637467476149, + "grad_norm": 5.58958625793457, + "learning_rate": 1.7894240533874948e-05, + "loss": 0.0541, + "step": 3003 + }, + { + "epoch": 0.3256721595836947, + "grad_norm": 1.1538715362548828, + "learning_rate": 1.789351516030756e-05, + "loss": 0.0242, + "step": 3004 + }, + { + "epoch": 0.3257805724197745, + "grad_norm": 0.4379907548427582, + "learning_rate": 1.7892789786740172e-05, + "loss": 0.017, + "step": 3005 + }, + { + "epoch": 0.3258889852558543, + "grad_norm": 2.110882043838501, + "learning_rate": 1.7892064413172784e-05, + "loss": 0.0724, + "step": 3006 + }, + { + "epoch": 0.32599739809193407, + "grad_norm": 3.8600170612335205, + "learning_rate": 1.7891339039605396e-05, + "loss": 0.0243, + "step": 3007 + }, + { + "epoch": 0.3261058109280139, + "grad_norm": 1.034112811088562, + "learning_rate": 1.7890613666038012e-05, + "loss": 0.0198, + "step": 3008 + }, + { + "epoch": 0.32621422376409365, + "grad_norm": 1.3082170486450195, + "learning_rate": 1.7889888292470624e-05, + "loss": 0.0204, + "step": 3009 + }, + { + "epoch": 0.32632263660017347, + "grad_norm": 2.2724244594573975, + "learning_rate": 1.7889162918903236e-05, + "loss": 0.0734, + "step": 3010 + }, + { + "epoch": 0.32643104943625323, + "grad_norm": 3.8661844730377197, + "learning_rate": 1.7888437545335852e-05, + "loss": 0.0411, + "step": 3011 + }, + { + "epoch": 0.32653946227233305, + "grad_norm": 2.874342918395996, + "learning_rate": 1.7887712171768464e-05, + "loss": 0.049, + "step": 3012 + }, + { + "epoch": 0.3266478751084128, + "grad_norm": 5.034705638885498, + "learning_rate": 1.7886986798201076e-05, + "loss": 0.0526, + "step": 3013 + }, + { + "epoch": 0.32675628794449263, + "grad_norm": 3.567281723022461, + "learning_rate": 1.7886261424633688e-05, + "loss": 0.0774, + "step": 3014 + }, + { + "epoch": 0.3268647007805724, + "grad_norm": 3.071017265319824, + "learning_rate": 1.78855360510663e-05, + "loss": 0.0357, + "step": 3015 + }, + { + "epoch": 0.3269731136166522, + "grad_norm": 6.198752403259277, + "learning_rate": 1.7884810677498913e-05, + "loss": 0.0691, + "step": 3016 + }, + { + "epoch": 0.32708152645273203, + "grad_norm": 2.7290947437286377, + "learning_rate": 1.7884085303931528e-05, + "loss": 0.1087, + "step": 3017 + }, + { + "epoch": 0.3271899392888118, + "grad_norm": 2.2459237575531006, + "learning_rate": 1.788335993036414e-05, + "loss": 0.0383, + "step": 3018 + }, + { + "epoch": 0.3272983521248916, + "grad_norm": 1.6430726051330566, + "learning_rate": 1.7882634556796752e-05, + "loss": 0.0178, + "step": 3019 + }, + { + "epoch": 0.3274067649609714, + "grad_norm": 4.524199962615967, + "learning_rate": 1.7881909183229364e-05, + "loss": 0.0584, + "step": 3020 + }, + { + "epoch": 0.3275151777970512, + "grad_norm": 1.3625789880752563, + "learning_rate": 1.7881183809661977e-05, + "loss": 0.035, + "step": 3021 + }, + { + "epoch": 0.32762359063313096, + "grad_norm": 3.0459866523742676, + "learning_rate": 1.788045843609459e-05, + "loss": 0.0418, + "step": 3022 + }, + { + "epoch": 0.3277320034692108, + "grad_norm": 6.163491249084473, + "learning_rate": 1.78797330625272e-05, + "loss": 0.0563, + "step": 3023 + }, + { + "epoch": 0.32784041630529054, + "grad_norm": 1.526442050933838, + "learning_rate": 1.7879007688959816e-05, + "loss": 0.0264, + "step": 3024 + }, + { + "epoch": 0.32794882914137036, + "grad_norm": 2.198302984237671, + "learning_rate": 1.787828231539243e-05, + "loss": 0.0362, + "step": 3025 + }, + { + "epoch": 0.3280572419774501, + "grad_norm": 2.2375328540802, + "learning_rate": 1.787755694182504e-05, + "loss": 0.0305, + "step": 3026 + }, + { + "epoch": 0.32816565481352994, + "grad_norm": 7.039582252502441, + "learning_rate": 1.7876831568257653e-05, + "loss": 0.0529, + "step": 3027 + }, + { + "epoch": 0.3282740676496097, + "grad_norm": 1.2228142023086548, + "learning_rate": 1.7876106194690265e-05, + "loss": 0.0126, + "step": 3028 + }, + { + "epoch": 0.3283824804856895, + "grad_norm": 1.7413396835327148, + "learning_rate": 1.787538082112288e-05, + "loss": 0.0356, + "step": 3029 + }, + { + "epoch": 0.3284908933217693, + "grad_norm": 2.082608222961426, + "learning_rate": 1.7874655447555493e-05, + "loss": 0.0208, + "step": 3030 + }, + { + "epoch": 0.3285993061578491, + "grad_norm": 2.637465715408325, + "learning_rate": 1.7873930073988105e-05, + "loss": 0.0141, + "step": 3031 + }, + { + "epoch": 0.32870771899392887, + "grad_norm": 2.1717190742492676, + "learning_rate": 1.787320470042072e-05, + "loss": 0.0224, + "step": 3032 + }, + { + "epoch": 0.3288161318300087, + "grad_norm": 0.8948341012001038, + "learning_rate": 1.7872479326853332e-05, + "loss": 0.026, + "step": 3033 + }, + { + "epoch": 0.32892454466608845, + "grad_norm": 1.432342290878296, + "learning_rate": 1.7871753953285945e-05, + "loss": 0.0198, + "step": 3034 + }, + { + "epoch": 0.32903295750216827, + "grad_norm": 2.3751230239868164, + "learning_rate": 1.7871028579718557e-05, + "loss": 0.0849, + "step": 3035 + }, + { + "epoch": 0.32914137033824803, + "grad_norm": 3.8705592155456543, + "learning_rate": 1.787030320615117e-05, + "loss": 0.0692, + "step": 3036 + }, + { + "epoch": 0.32924978317432785, + "grad_norm": 3.4753670692443848, + "learning_rate": 1.786957783258378e-05, + "loss": 0.0368, + "step": 3037 + }, + { + "epoch": 0.3293581960104076, + "grad_norm": 2.3268585205078125, + "learning_rate": 1.7868852459016393e-05, + "loss": 0.0212, + "step": 3038 + }, + { + "epoch": 0.32946660884648743, + "grad_norm": 4.833503723144531, + "learning_rate": 1.786812708544901e-05, + "loss": 0.0623, + "step": 3039 + }, + { + "epoch": 0.3295750216825672, + "grad_norm": 2.5573155879974365, + "learning_rate": 1.786740171188162e-05, + "loss": 0.0353, + "step": 3040 + }, + { + "epoch": 0.329683434518647, + "grad_norm": 5.28800630569458, + "learning_rate": 1.7866676338314233e-05, + "loss": 0.0524, + "step": 3041 + }, + { + "epoch": 0.3297918473547268, + "grad_norm": 2.911194324493408, + "learning_rate": 1.7865950964746845e-05, + "loss": 0.032, + "step": 3042 + }, + { + "epoch": 0.3299002601908066, + "grad_norm": 2.18282413482666, + "learning_rate": 1.7865225591179457e-05, + "loss": 0.0336, + "step": 3043 + }, + { + "epoch": 0.33000867302688636, + "grad_norm": 2.256711006164551, + "learning_rate": 1.786450021761207e-05, + "loss": 0.0349, + "step": 3044 + }, + { + "epoch": 0.3301170858629662, + "grad_norm": 2.1468615531921387, + "learning_rate": 1.7863774844044685e-05, + "loss": 0.019, + "step": 3045 + }, + { + "epoch": 0.330225498699046, + "grad_norm": 3.429983139038086, + "learning_rate": 1.7863049470477297e-05, + "loss": 0.0389, + "step": 3046 + }, + { + "epoch": 0.33033391153512576, + "grad_norm": 3.21347975730896, + "learning_rate": 1.7862324096909913e-05, + "loss": 0.0208, + "step": 3047 + }, + { + "epoch": 0.3304423243712056, + "grad_norm": 4.732263565063477, + "learning_rate": 1.7861598723342525e-05, + "loss": 0.0171, + "step": 3048 + }, + { + "epoch": 0.33055073720728534, + "grad_norm": 2.859794855117798, + "learning_rate": 1.7860873349775137e-05, + "loss": 0.046, + "step": 3049 + }, + { + "epoch": 0.33065915004336516, + "grad_norm": 1.487738013267517, + "learning_rate": 1.786014797620775e-05, + "loss": 0.0134, + "step": 3050 + }, + { + "epoch": 0.3307675628794449, + "grad_norm": 5.615499496459961, + "learning_rate": 1.785942260264036e-05, + "loss": 0.0393, + "step": 3051 + }, + { + "epoch": 0.33087597571552474, + "grad_norm": 5.049776077270508, + "learning_rate": 1.7858697229072973e-05, + "loss": 0.0482, + "step": 3052 + }, + { + "epoch": 0.3309843885516045, + "grad_norm": 6.925439834594727, + "learning_rate": 1.7857971855505586e-05, + "loss": 0.0696, + "step": 3053 + }, + { + "epoch": 0.3310928013876843, + "grad_norm": 6.5739545822143555, + "learning_rate": 1.78572464819382e-05, + "loss": 0.0653, + "step": 3054 + }, + { + "epoch": 0.3312012142237641, + "grad_norm": 2.1748480796813965, + "learning_rate": 1.7856521108370813e-05, + "loss": 0.0705, + "step": 3055 + }, + { + "epoch": 0.3313096270598439, + "grad_norm": 4.732572555541992, + "learning_rate": 1.7855795734803425e-05, + "loss": 0.0408, + "step": 3056 + }, + { + "epoch": 0.33141803989592367, + "grad_norm": 4.439199924468994, + "learning_rate": 1.7855070361236037e-05, + "loss": 0.026, + "step": 3057 + }, + { + "epoch": 0.3315264527320035, + "grad_norm": 3.552410840988159, + "learning_rate": 1.785434498766865e-05, + "loss": 0.0506, + "step": 3058 + }, + { + "epoch": 0.33163486556808325, + "grad_norm": 4.273713111877441, + "learning_rate": 1.7853619614101262e-05, + "loss": 0.0235, + "step": 3059 + }, + { + "epoch": 0.33174327840416307, + "grad_norm": 1.4780118465423584, + "learning_rate": 1.7852894240533877e-05, + "loss": 0.0105, + "step": 3060 + }, + { + "epoch": 0.33185169124024283, + "grad_norm": 4.019708156585693, + "learning_rate": 1.785216886696649e-05, + "loss": 0.0276, + "step": 3061 + }, + { + "epoch": 0.33196010407632265, + "grad_norm": 4.362269878387451, + "learning_rate": 1.78514434933991e-05, + "loss": 0.0265, + "step": 3062 + }, + { + "epoch": 0.3320685169124024, + "grad_norm": 5.672086715698242, + "learning_rate": 1.7850718119831717e-05, + "loss": 0.0756, + "step": 3063 + }, + { + "epoch": 0.33217692974848223, + "grad_norm": 8.250395774841309, + "learning_rate": 1.784999274626433e-05, + "loss": 0.0225, + "step": 3064 + }, + { + "epoch": 0.332285342584562, + "grad_norm": 8.897716522216797, + "learning_rate": 1.784926737269694e-05, + "loss": 0.0544, + "step": 3065 + }, + { + "epoch": 0.3323937554206418, + "grad_norm": 1.988656759262085, + "learning_rate": 1.7848541999129554e-05, + "loss": 0.0432, + "step": 3066 + }, + { + "epoch": 0.3325021682567216, + "grad_norm": 9.296608924865723, + "learning_rate": 1.7847816625562166e-05, + "loss": 0.0351, + "step": 3067 + }, + { + "epoch": 0.3326105810928014, + "grad_norm": 2.367027997970581, + "learning_rate": 1.7847091251994778e-05, + "loss": 0.0357, + "step": 3068 + }, + { + "epoch": 0.33271899392888116, + "grad_norm": 6.544503688812256, + "learning_rate": 1.7846365878427393e-05, + "loss": 0.0703, + "step": 3069 + }, + { + "epoch": 0.332827406764961, + "grad_norm": 2.8262441158294678, + "learning_rate": 1.7845640504860006e-05, + "loss": 0.0479, + "step": 3070 + }, + { + "epoch": 0.33293581960104074, + "grad_norm": 1.054211974143982, + "learning_rate": 1.7844915131292618e-05, + "loss": 0.0095, + "step": 3071 + }, + { + "epoch": 0.33304423243712056, + "grad_norm": 4.114895343780518, + "learning_rate": 1.784418975772523e-05, + "loss": 0.0452, + "step": 3072 + }, + { + "epoch": 0.3331526452732003, + "grad_norm": 2.123685121536255, + "learning_rate": 1.7843464384157842e-05, + "loss": 0.0303, + "step": 3073 + }, + { + "epoch": 0.33326105810928014, + "grad_norm": 5.491241931915283, + "learning_rate": 1.7842739010590454e-05, + "loss": 0.0226, + "step": 3074 + }, + { + "epoch": 0.33336947094535996, + "grad_norm": 5.697206497192383, + "learning_rate": 1.7842013637023066e-05, + "loss": 0.032, + "step": 3075 + }, + { + "epoch": 0.3334778837814397, + "grad_norm": 4.593652725219727, + "learning_rate": 1.7841288263455682e-05, + "loss": 0.064, + "step": 3076 + }, + { + "epoch": 0.33358629661751954, + "grad_norm": 6.755881309509277, + "learning_rate": 1.7840562889888294e-05, + "loss": 0.0593, + "step": 3077 + }, + { + "epoch": 0.3336947094535993, + "grad_norm": 2.3104655742645264, + "learning_rate": 1.7839837516320906e-05, + "loss": 0.0166, + "step": 3078 + }, + { + "epoch": 0.3338031222896791, + "grad_norm": 3.9641730785369873, + "learning_rate": 1.7839112142753518e-05, + "loss": 0.0212, + "step": 3079 + }, + { + "epoch": 0.3339115351257589, + "grad_norm": 3.9461710453033447, + "learning_rate": 1.783838676918613e-05, + "loss": 0.0494, + "step": 3080 + }, + { + "epoch": 0.3340199479618387, + "grad_norm": 1.9258077144622803, + "learning_rate": 1.7837661395618746e-05, + "loss": 0.0364, + "step": 3081 + }, + { + "epoch": 0.33412836079791847, + "grad_norm": 4.726726055145264, + "learning_rate": 1.7836936022051358e-05, + "loss": 0.0417, + "step": 3082 + }, + { + "epoch": 0.3342367736339983, + "grad_norm": 4.0791802406311035, + "learning_rate": 1.783621064848397e-05, + "loss": 0.0422, + "step": 3083 + }, + { + "epoch": 0.33434518647007805, + "grad_norm": 10.065118789672852, + "learning_rate": 1.7835485274916586e-05, + "loss": 0.0543, + "step": 3084 + }, + { + "epoch": 0.33445359930615787, + "grad_norm": 2.7132105827331543, + "learning_rate": 1.7834759901349198e-05, + "loss": 0.0709, + "step": 3085 + }, + { + "epoch": 0.33456201214223763, + "grad_norm": 5.61018180847168, + "learning_rate": 1.783403452778181e-05, + "loss": 0.0865, + "step": 3086 + }, + { + "epoch": 0.33467042497831745, + "grad_norm": 5.210923194885254, + "learning_rate": 1.7833309154214422e-05, + "loss": 0.0398, + "step": 3087 + }, + { + "epoch": 0.3347788378143972, + "grad_norm": 3.024508476257324, + "learning_rate": 1.7832583780647034e-05, + "loss": 0.06, + "step": 3088 + }, + { + "epoch": 0.33488725065047703, + "grad_norm": 3.671039342880249, + "learning_rate": 1.7831858407079646e-05, + "loss": 0.0269, + "step": 3089 + }, + { + "epoch": 0.3349956634865568, + "grad_norm": 1.4114694595336914, + "learning_rate": 1.783113303351226e-05, + "loss": 0.0253, + "step": 3090 + }, + { + "epoch": 0.3351040763226366, + "grad_norm": 2.7512571811676025, + "learning_rate": 1.7830407659944874e-05, + "loss": 0.0404, + "step": 3091 + }, + { + "epoch": 0.3352124891587164, + "grad_norm": 2.6968629360198975, + "learning_rate": 1.7829682286377486e-05, + "loss": 0.032, + "step": 3092 + }, + { + "epoch": 0.3353209019947962, + "grad_norm": 2.1044909954071045, + "learning_rate": 1.78289569128101e-05, + "loss": 0.0287, + "step": 3093 + }, + { + "epoch": 0.33542931483087596, + "grad_norm": 0.9314191341400146, + "learning_rate": 1.782823153924271e-05, + "loss": 0.0087, + "step": 3094 + }, + { + "epoch": 0.3355377276669558, + "grad_norm": 5.9530415534973145, + "learning_rate": 1.7827506165675323e-05, + "loss": 0.054, + "step": 3095 + }, + { + "epoch": 0.33564614050303554, + "grad_norm": 1.9984571933746338, + "learning_rate": 1.7826780792107935e-05, + "loss": 0.0164, + "step": 3096 + }, + { + "epoch": 0.33575455333911536, + "grad_norm": 1.6742726564407349, + "learning_rate": 1.782605541854055e-05, + "loss": 0.0203, + "step": 3097 + }, + { + "epoch": 0.3358629661751951, + "grad_norm": 6.529160022735596, + "learning_rate": 1.7825330044973162e-05, + "loss": 0.0851, + "step": 3098 + }, + { + "epoch": 0.33597137901127494, + "grad_norm": 5.081089973449707, + "learning_rate": 1.7824604671405778e-05, + "loss": 0.0766, + "step": 3099 + }, + { + "epoch": 0.3360797918473547, + "grad_norm": 3.884354829788208, + "learning_rate": 1.782387929783839e-05, + "loss": 0.0429, + "step": 3100 + }, + { + "epoch": 0.3361882046834345, + "grad_norm": 11.162044525146484, + "learning_rate": 1.7823153924271002e-05, + "loss": 0.0937, + "step": 3101 + }, + { + "epoch": 0.3362966175195143, + "grad_norm": 5.957160472869873, + "learning_rate": 1.7822428550703614e-05, + "loss": 0.0324, + "step": 3102 + }, + { + "epoch": 0.3364050303555941, + "grad_norm": 3.561439037322998, + "learning_rate": 1.7821703177136227e-05, + "loss": 0.0656, + "step": 3103 + }, + { + "epoch": 0.33651344319167387, + "grad_norm": 2.1720190048217773, + "learning_rate": 1.782097780356884e-05, + "loss": 0.029, + "step": 3104 + }, + { + "epoch": 0.3366218560277537, + "grad_norm": 3.309699535369873, + "learning_rate": 1.782025243000145e-05, + "loss": 0.0476, + "step": 3105 + }, + { + "epoch": 0.3367302688638335, + "grad_norm": 7.2465434074401855, + "learning_rate": 1.7819527056434066e-05, + "loss": 0.0727, + "step": 3106 + }, + { + "epoch": 0.33683868169991327, + "grad_norm": 6.298462390899658, + "learning_rate": 1.781880168286668e-05, + "loss": 0.0328, + "step": 3107 + }, + { + "epoch": 0.3369470945359931, + "grad_norm": 3.0622332096099854, + "learning_rate": 1.781807630929929e-05, + "loss": 0.035, + "step": 3108 + }, + { + "epoch": 0.33705550737207285, + "grad_norm": 3.078228235244751, + "learning_rate": 1.7817350935731903e-05, + "loss": 0.07, + "step": 3109 + }, + { + "epoch": 0.33716392020815267, + "grad_norm": 3.9129581451416016, + "learning_rate": 1.7816625562164515e-05, + "loss": 0.0657, + "step": 3110 + }, + { + "epoch": 0.33727233304423243, + "grad_norm": 3.2523720264434814, + "learning_rate": 1.7815900188597127e-05, + "loss": 0.0222, + "step": 3111 + }, + { + "epoch": 0.33738074588031225, + "grad_norm": 3.7525713443756104, + "learning_rate": 1.7815174815029743e-05, + "loss": 0.0315, + "step": 3112 + }, + { + "epoch": 0.337489158716392, + "grad_norm": 2.259420394897461, + "learning_rate": 1.7814449441462355e-05, + "loss": 0.0379, + "step": 3113 + }, + { + "epoch": 0.33759757155247183, + "grad_norm": 3.920499086380005, + "learning_rate": 1.7813724067894967e-05, + "loss": 0.0614, + "step": 3114 + }, + { + "epoch": 0.3377059843885516, + "grad_norm": 1.9361157417297363, + "learning_rate": 1.7812998694327582e-05, + "loss": 0.0293, + "step": 3115 + }, + { + "epoch": 0.3378143972246314, + "grad_norm": 1.8761849403381348, + "learning_rate": 1.7812273320760195e-05, + "loss": 0.0202, + "step": 3116 + }, + { + "epoch": 0.3379228100607112, + "grad_norm": 2.483424663543701, + "learning_rate": 1.7811547947192807e-05, + "loss": 0.0162, + "step": 3117 + }, + { + "epoch": 0.338031222896791, + "grad_norm": 3.675661563873291, + "learning_rate": 1.781082257362542e-05, + "loss": 0.0585, + "step": 3118 + }, + { + "epoch": 0.33813963573287076, + "grad_norm": 1.7692373991012573, + "learning_rate": 1.781009720005803e-05, + "loss": 0.0359, + "step": 3119 + }, + { + "epoch": 0.3382480485689506, + "grad_norm": 1.9644324779510498, + "learning_rate": 1.7809371826490643e-05, + "loss": 0.0359, + "step": 3120 + }, + { + "epoch": 0.33835646140503034, + "grad_norm": 2.6283698081970215, + "learning_rate": 1.780864645292326e-05, + "loss": 0.054, + "step": 3121 + }, + { + "epoch": 0.33846487424111016, + "grad_norm": 3.160480260848999, + "learning_rate": 1.780792107935587e-05, + "loss": 0.0651, + "step": 3122 + }, + { + "epoch": 0.3385732870771899, + "grad_norm": 6.243264198303223, + "learning_rate": 1.7807195705788483e-05, + "loss": 0.0505, + "step": 3123 + }, + { + "epoch": 0.33868169991326974, + "grad_norm": 4.491207122802734, + "learning_rate": 1.7806470332221095e-05, + "loss": 0.0332, + "step": 3124 + }, + { + "epoch": 0.3387901127493495, + "grad_norm": 3.2815864086151123, + "learning_rate": 1.7805744958653707e-05, + "loss": 0.043, + "step": 3125 + }, + { + "epoch": 0.3388985255854293, + "grad_norm": 3.081400156021118, + "learning_rate": 1.780501958508632e-05, + "loss": 0.0314, + "step": 3126 + }, + { + "epoch": 0.3390069384215091, + "grad_norm": 2.724637985229492, + "learning_rate": 1.780429421151893e-05, + "loss": 0.0368, + "step": 3127 + }, + { + "epoch": 0.3391153512575889, + "grad_norm": 3.042382001876831, + "learning_rate": 1.7803568837951547e-05, + "loss": 0.0694, + "step": 3128 + }, + { + "epoch": 0.33922376409366867, + "grad_norm": 1.5437395572662354, + "learning_rate": 1.780284346438416e-05, + "loss": 0.0157, + "step": 3129 + }, + { + "epoch": 0.3393321769297485, + "grad_norm": 4.9632110595703125, + "learning_rate": 1.780211809081677e-05, + "loss": 0.0512, + "step": 3130 + }, + { + "epoch": 0.33944058976582825, + "grad_norm": 2.4552106857299805, + "learning_rate": 1.7801392717249384e-05, + "loss": 0.0496, + "step": 3131 + }, + { + "epoch": 0.33954900260190807, + "grad_norm": 1.3297513723373413, + "learning_rate": 1.7800667343682e-05, + "loss": 0.0223, + "step": 3132 + }, + { + "epoch": 0.33965741543798783, + "grad_norm": 6.3506903648376465, + "learning_rate": 1.779994197011461e-05, + "loss": 0.0874, + "step": 3133 + }, + { + "epoch": 0.33976582827406765, + "grad_norm": 1.2711526155471802, + "learning_rate": 1.7799216596547223e-05, + "loss": 0.0102, + "step": 3134 + }, + { + "epoch": 0.33987424111014747, + "grad_norm": 7.531414031982422, + "learning_rate": 1.7798491222979835e-05, + "loss": 0.0558, + "step": 3135 + }, + { + "epoch": 0.33998265394622723, + "grad_norm": 0.9930984377861023, + "learning_rate": 1.779776584941245e-05, + "loss": 0.0223, + "step": 3136 + }, + { + "epoch": 0.34009106678230705, + "grad_norm": 4.344967842102051, + "learning_rate": 1.7797040475845063e-05, + "loss": 0.0417, + "step": 3137 + }, + { + "epoch": 0.3401994796183868, + "grad_norm": 2.7489306926727295, + "learning_rate": 1.7796315102277675e-05, + "loss": 0.0246, + "step": 3138 + }, + { + "epoch": 0.34030789245446663, + "grad_norm": 2.38954758644104, + "learning_rate": 1.7795589728710287e-05, + "loss": 0.026, + "step": 3139 + }, + { + "epoch": 0.3404163052905464, + "grad_norm": 8.13370418548584, + "learning_rate": 1.77948643551429e-05, + "loss": 0.0731, + "step": 3140 + }, + { + "epoch": 0.3405247181266262, + "grad_norm": 1.8158217668533325, + "learning_rate": 1.7794138981575512e-05, + "loss": 0.0205, + "step": 3141 + }, + { + "epoch": 0.340633130962706, + "grad_norm": 2.3392388820648193, + "learning_rate": 1.7793413608008124e-05, + "loss": 0.0381, + "step": 3142 + }, + { + "epoch": 0.3407415437987858, + "grad_norm": 1.475861668586731, + "learning_rate": 1.779268823444074e-05, + "loss": 0.0175, + "step": 3143 + }, + { + "epoch": 0.34084995663486556, + "grad_norm": 2.990546703338623, + "learning_rate": 1.779196286087335e-05, + "loss": 0.0258, + "step": 3144 + }, + { + "epoch": 0.3409583694709454, + "grad_norm": 2.194432258605957, + "learning_rate": 1.7791237487305964e-05, + "loss": 0.0632, + "step": 3145 + }, + { + "epoch": 0.34106678230702514, + "grad_norm": 3.849672555923462, + "learning_rate": 1.7790512113738576e-05, + "loss": 0.0838, + "step": 3146 + }, + { + "epoch": 0.34117519514310496, + "grad_norm": 5.843372821807861, + "learning_rate": 1.7789786740171188e-05, + "loss": 0.0571, + "step": 3147 + }, + { + "epoch": 0.3412836079791847, + "grad_norm": 3.6928820610046387, + "learning_rate": 1.77890613666038e-05, + "loss": 0.055, + "step": 3148 + }, + { + "epoch": 0.34139202081526454, + "grad_norm": 3.5706541538238525, + "learning_rate": 1.7788335993036416e-05, + "loss": 0.0485, + "step": 3149 + }, + { + "epoch": 0.3415004336513443, + "grad_norm": 3.78281569480896, + "learning_rate": 1.7787610619469028e-05, + "loss": 0.0854, + "step": 3150 + }, + { + "epoch": 0.3416088464874241, + "grad_norm": 11.140869140625, + "learning_rate": 1.7786885245901643e-05, + "loss": 0.0503, + "step": 3151 + }, + { + "epoch": 0.3417172593235039, + "grad_norm": 6.365270614624023, + "learning_rate": 1.7786159872334255e-05, + "loss": 0.0301, + "step": 3152 + }, + { + "epoch": 0.3418256721595837, + "grad_norm": 5.691088676452637, + "learning_rate": 1.7785434498766868e-05, + "loss": 0.0324, + "step": 3153 + }, + { + "epoch": 0.34193408499566347, + "grad_norm": 3.8867361545562744, + "learning_rate": 1.778470912519948e-05, + "loss": 0.1055, + "step": 3154 + }, + { + "epoch": 0.3420424978317433, + "grad_norm": 7.975442409515381, + "learning_rate": 1.7783983751632092e-05, + "loss": 0.0456, + "step": 3155 + }, + { + "epoch": 0.34215091066782305, + "grad_norm": 1.9701186418533325, + "learning_rate": 1.7783258378064704e-05, + "loss": 0.0454, + "step": 3156 + }, + { + "epoch": 0.34225932350390287, + "grad_norm": 4.152474403381348, + "learning_rate": 1.7782533004497316e-05, + "loss": 0.0691, + "step": 3157 + }, + { + "epoch": 0.34236773633998263, + "grad_norm": 1.9457701444625854, + "learning_rate": 1.7781807630929932e-05, + "loss": 0.0312, + "step": 3158 + }, + { + "epoch": 0.34247614917606245, + "grad_norm": 3.0130109786987305, + "learning_rate": 1.7781082257362544e-05, + "loss": 0.0323, + "step": 3159 + }, + { + "epoch": 0.3425845620121422, + "grad_norm": 3.385575771331787, + "learning_rate": 1.7780356883795156e-05, + "loss": 0.024, + "step": 3160 + }, + { + "epoch": 0.34269297484822203, + "grad_norm": 2.791781187057495, + "learning_rate": 1.7779631510227768e-05, + "loss": 0.0619, + "step": 3161 + }, + { + "epoch": 0.3428013876843018, + "grad_norm": 2.345489025115967, + "learning_rate": 1.777890613666038e-05, + "loss": 0.05, + "step": 3162 + }, + { + "epoch": 0.3429098005203816, + "grad_norm": 1.7010232210159302, + "learning_rate": 1.7778180763092992e-05, + "loss": 0.0323, + "step": 3163 + }, + { + "epoch": 0.34301821335646143, + "grad_norm": 2.305123805999756, + "learning_rate": 1.7777455389525608e-05, + "loss": 0.036, + "step": 3164 + }, + { + "epoch": 0.3431266261925412, + "grad_norm": 1.4259427785873413, + "learning_rate": 1.777673001595822e-05, + "loss": 0.0373, + "step": 3165 + }, + { + "epoch": 0.343235039028621, + "grad_norm": 5.8454670906066895, + "learning_rate": 1.7776004642390832e-05, + "loss": 0.0819, + "step": 3166 + }, + { + "epoch": 0.3433434518647008, + "grad_norm": 1.108068585395813, + "learning_rate": 1.7775279268823448e-05, + "loss": 0.0264, + "step": 3167 + }, + { + "epoch": 0.3434518647007806, + "grad_norm": 5.9630584716796875, + "learning_rate": 1.777455389525606e-05, + "loss": 0.0713, + "step": 3168 + }, + { + "epoch": 0.34356027753686036, + "grad_norm": 8.36380386352539, + "learning_rate": 1.7773828521688672e-05, + "loss": 0.0771, + "step": 3169 + }, + { + "epoch": 0.3436686903729402, + "grad_norm": 1.746650218963623, + "learning_rate": 1.7773103148121284e-05, + "loss": 0.0295, + "step": 3170 + }, + { + "epoch": 0.34377710320901994, + "grad_norm": 1.705533504486084, + "learning_rate": 1.7772377774553896e-05, + "loss": 0.0435, + "step": 3171 + }, + { + "epoch": 0.34388551604509976, + "grad_norm": 3.1089186668395996, + "learning_rate": 1.777165240098651e-05, + "loss": 0.0539, + "step": 3172 + }, + { + "epoch": 0.3439939288811795, + "grad_norm": 2.4865341186523438, + "learning_rate": 1.7770927027419124e-05, + "loss": 0.0443, + "step": 3173 + }, + { + "epoch": 0.34410234171725934, + "grad_norm": 3.083296298980713, + "learning_rate": 1.7770201653851736e-05, + "loss": 0.0298, + "step": 3174 + }, + { + "epoch": 0.3442107545533391, + "grad_norm": 4.7879204750061035, + "learning_rate": 1.776947628028435e-05, + "loss": 0.0726, + "step": 3175 + }, + { + "epoch": 0.3443191673894189, + "grad_norm": 3.3916068077087402, + "learning_rate": 1.776875090671696e-05, + "loss": 0.054, + "step": 3176 + }, + { + "epoch": 0.3444275802254987, + "grad_norm": 1.9902054071426392, + "learning_rate": 1.7768025533149573e-05, + "loss": 0.043, + "step": 3177 + }, + { + "epoch": 0.3445359930615785, + "grad_norm": 1.7053680419921875, + "learning_rate": 1.7767300159582185e-05, + "loss": 0.0168, + "step": 3178 + }, + { + "epoch": 0.34464440589765827, + "grad_norm": 1.993080735206604, + "learning_rate": 1.7766574786014797e-05, + "loss": 0.0555, + "step": 3179 + }, + { + "epoch": 0.3447528187337381, + "grad_norm": 2.8052937984466553, + "learning_rate": 1.7765849412447412e-05, + "loss": 0.0662, + "step": 3180 + }, + { + "epoch": 0.34486123156981785, + "grad_norm": 4.889697551727295, + "learning_rate": 1.7765124038880025e-05, + "loss": 0.0457, + "step": 3181 + }, + { + "epoch": 0.34496964440589767, + "grad_norm": 3.16023588180542, + "learning_rate": 1.7764398665312637e-05, + "loss": 0.0407, + "step": 3182 + }, + { + "epoch": 0.34507805724197743, + "grad_norm": 4.352933883666992, + "learning_rate": 1.776367329174525e-05, + "loss": 0.0228, + "step": 3183 + }, + { + "epoch": 0.34518647007805725, + "grad_norm": 4.657009601593018, + "learning_rate": 1.7762947918177864e-05, + "loss": 0.0387, + "step": 3184 + }, + { + "epoch": 0.345294882914137, + "grad_norm": 1.9572018384933472, + "learning_rate": 1.7762222544610477e-05, + "loss": 0.0446, + "step": 3185 + }, + { + "epoch": 0.34540329575021683, + "grad_norm": 3.3632876873016357, + "learning_rate": 1.776149717104309e-05, + "loss": 0.0485, + "step": 3186 + }, + { + "epoch": 0.3455117085862966, + "grad_norm": 1.2131186723709106, + "learning_rate": 1.77607717974757e-05, + "loss": 0.0204, + "step": 3187 + }, + { + "epoch": 0.3456201214223764, + "grad_norm": 2.864773750305176, + "learning_rate": 1.7760046423908316e-05, + "loss": 0.019, + "step": 3188 + }, + { + "epoch": 0.3457285342584562, + "grad_norm": 3.9798216819763184, + "learning_rate": 1.775932105034093e-05, + "loss": 0.0882, + "step": 3189 + }, + { + "epoch": 0.345836947094536, + "grad_norm": 3.5179617404937744, + "learning_rate": 1.775859567677354e-05, + "loss": 0.0353, + "step": 3190 + }, + { + "epoch": 0.34594535993061576, + "grad_norm": 3.334951162338257, + "learning_rate": 1.7757870303206153e-05, + "loss": 0.0576, + "step": 3191 + }, + { + "epoch": 0.3460537727666956, + "grad_norm": 2.456387519836426, + "learning_rate": 1.7757144929638765e-05, + "loss": 0.0347, + "step": 3192 + }, + { + "epoch": 0.3461621856027754, + "grad_norm": 1.3483717441558838, + "learning_rate": 1.7756419556071377e-05, + "loss": 0.0191, + "step": 3193 + }, + { + "epoch": 0.34627059843885516, + "grad_norm": 2.5996756553649902, + "learning_rate": 1.775569418250399e-05, + "loss": 0.0716, + "step": 3194 + }, + { + "epoch": 0.346379011274935, + "grad_norm": 5.064035892486572, + "learning_rate": 1.7754968808936605e-05, + "loss": 0.0779, + "step": 3195 + }, + { + "epoch": 0.34648742411101474, + "grad_norm": 1.0714954137802124, + "learning_rate": 1.7754243435369217e-05, + "loss": 0.0253, + "step": 3196 + }, + { + "epoch": 0.34659583694709456, + "grad_norm": 1.999647855758667, + "learning_rate": 1.775351806180183e-05, + "loss": 0.031, + "step": 3197 + }, + { + "epoch": 0.3467042497831743, + "grad_norm": 3.08722186088562, + "learning_rate": 1.775279268823444e-05, + "loss": 0.0416, + "step": 3198 + }, + { + "epoch": 0.34681266261925414, + "grad_norm": 3.265911102294922, + "learning_rate": 1.7752067314667053e-05, + "loss": 0.028, + "step": 3199 + }, + { + "epoch": 0.3469210754553339, + "grad_norm": 1.6155215501785278, + "learning_rate": 1.7751341941099665e-05, + "loss": 0.0274, + "step": 3200 + }, + { + "epoch": 0.3470294882914137, + "grad_norm": 4.120359420776367, + "learning_rate": 1.775061656753228e-05, + "loss": 0.0646, + "step": 3201 + }, + { + "epoch": 0.3471379011274935, + "grad_norm": 2.6596615314483643, + "learning_rate": 1.7749891193964893e-05, + "loss": 0.057, + "step": 3202 + }, + { + "epoch": 0.3472463139635733, + "grad_norm": 3.7117834091186523, + "learning_rate": 1.774916582039751e-05, + "loss": 0.0454, + "step": 3203 + }, + { + "epoch": 0.34735472679965307, + "grad_norm": 1.418250322341919, + "learning_rate": 1.774844044683012e-05, + "loss": 0.0467, + "step": 3204 + }, + { + "epoch": 0.3474631396357329, + "grad_norm": 5.394033908843994, + "learning_rate": 1.7747715073262733e-05, + "loss": 0.0907, + "step": 3205 + }, + { + "epoch": 0.34757155247181265, + "grad_norm": 1.141340732574463, + "learning_rate": 1.7746989699695345e-05, + "loss": 0.017, + "step": 3206 + }, + { + "epoch": 0.34767996530789247, + "grad_norm": 2.0917482376098633, + "learning_rate": 1.7746264326127957e-05, + "loss": 0.0457, + "step": 3207 + }, + { + "epoch": 0.34778837814397223, + "grad_norm": 0.7696893811225891, + "learning_rate": 1.774553895256057e-05, + "loss": 0.0178, + "step": 3208 + }, + { + "epoch": 0.34789679098005205, + "grad_norm": 2.8461153507232666, + "learning_rate": 1.774481357899318e-05, + "loss": 0.0512, + "step": 3209 + }, + { + "epoch": 0.3480052038161318, + "grad_norm": 1.627835750579834, + "learning_rate": 1.7744088205425797e-05, + "loss": 0.0359, + "step": 3210 + }, + { + "epoch": 0.34811361665221163, + "grad_norm": 3.2644898891448975, + "learning_rate": 1.774336283185841e-05, + "loss": 0.0481, + "step": 3211 + }, + { + "epoch": 0.3482220294882914, + "grad_norm": 2.726097583770752, + "learning_rate": 1.774263745829102e-05, + "loss": 0.0492, + "step": 3212 + }, + { + "epoch": 0.3483304423243712, + "grad_norm": 1.9266866445541382, + "learning_rate": 1.7741912084723634e-05, + "loss": 0.0143, + "step": 3213 + }, + { + "epoch": 0.348438855160451, + "grad_norm": 2.4695065021514893, + "learning_rate": 1.7741186711156246e-05, + "loss": 0.0436, + "step": 3214 + }, + { + "epoch": 0.3485472679965308, + "grad_norm": 2.476797580718994, + "learning_rate": 1.7740461337588858e-05, + "loss": 0.0627, + "step": 3215 + }, + { + "epoch": 0.34865568083261056, + "grad_norm": 0.8637637495994568, + "learning_rate": 1.7739735964021473e-05, + "loss": 0.0285, + "step": 3216 + }, + { + "epoch": 0.3487640936686904, + "grad_norm": 1.6352391242980957, + "learning_rate": 1.7739010590454085e-05, + "loss": 0.0401, + "step": 3217 + }, + { + "epoch": 0.34887250650477014, + "grad_norm": 7.345306396484375, + "learning_rate": 1.7738285216886698e-05, + "loss": 0.0554, + "step": 3218 + }, + { + "epoch": 0.34898091934084996, + "grad_norm": 2.6450014114379883, + "learning_rate": 1.7737559843319313e-05, + "loss": 0.0893, + "step": 3219 + }, + { + "epoch": 0.3490893321769297, + "grad_norm": 1.3778765201568604, + "learning_rate": 1.7736834469751925e-05, + "loss": 0.0328, + "step": 3220 + }, + { + "epoch": 0.34919774501300954, + "grad_norm": 1.0410354137420654, + "learning_rate": 1.7736109096184537e-05, + "loss": 0.0131, + "step": 3221 + }, + { + "epoch": 0.3493061578490893, + "grad_norm": 2.472620964050293, + "learning_rate": 1.773538372261715e-05, + "loss": 0.0291, + "step": 3222 + }, + { + "epoch": 0.3494145706851691, + "grad_norm": 4.320973873138428, + "learning_rate": 1.7734658349049762e-05, + "loss": 0.0259, + "step": 3223 + }, + { + "epoch": 0.34952298352124894, + "grad_norm": 0.9964883327484131, + "learning_rate": 1.7733932975482374e-05, + "loss": 0.0155, + "step": 3224 + }, + { + "epoch": 0.3496313963573287, + "grad_norm": 0.8719242215156555, + "learning_rate": 1.773320760191499e-05, + "loss": 0.0389, + "step": 3225 + }, + { + "epoch": 0.3497398091934085, + "grad_norm": 14.255569458007812, + "learning_rate": 1.77324822283476e-05, + "loss": 0.1021, + "step": 3226 + }, + { + "epoch": 0.3498482220294883, + "grad_norm": 2.695650815963745, + "learning_rate": 1.7731756854780214e-05, + "loss": 0.0448, + "step": 3227 + }, + { + "epoch": 0.3499566348655681, + "grad_norm": 1.9154733419418335, + "learning_rate": 1.7731031481212826e-05, + "loss": 0.0293, + "step": 3228 + }, + { + "epoch": 0.35006504770164787, + "grad_norm": 7.609716892242432, + "learning_rate": 1.7730306107645438e-05, + "loss": 0.0606, + "step": 3229 + }, + { + "epoch": 0.3501734605377277, + "grad_norm": 3.1445839405059814, + "learning_rate": 1.772958073407805e-05, + "loss": 0.0702, + "step": 3230 + }, + { + "epoch": 0.35028187337380745, + "grad_norm": 3.82326340675354, + "learning_rate": 1.7728855360510662e-05, + "loss": 0.0853, + "step": 3231 + }, + { + "epoch": 0.35039028620988727, + "grad_norm": 2.97585391998291, + "learning_rate": 1.7728129986943278e-05, + "loss": 0.0116, + "step": 3232 + }, + { + "epoch": 0.35049869904596703, + "grad_norm": 3.9222376346588135, + "learning_rate": 1.772740461337589e-05, + "loss": 0.0288, + "step": 3233 + }, + { + "epoch": 0.35060711188204685, + "grad_norm": 8.878240585327148, + "learning_rate": 1.7726679239808502e-05, + "loss": 0.0745, + "step": 3234 + }, + { + "epoch": 0.3507155247181266, + "grad_norm": 3.4735796451568604, + "learning_rate": 1.7725953866241114e-05, + "loss": 0.0796, + "step": 3235 + }, + { + "epoch": 0.35082393755420643, + "grad_norm": 4.689833164215088, + "learning_rate": 1.772522849267373e-05, + "loss": 0.0743, + "step": 3236 + }, + { + "epoch": 0.3509323503902862, + "grad_norm": 6.9673752784729, + "learning_rate": 1.7724503119106342e-05, + "loss": 0.0434, + "step": 3237 + }, + { + "epoch": 0.351040763226366, + "grad_norm": 3.199738025665283, + "learning_rate": 1.7723777745538954e-05, + "loss": 0.0302, + "step": 3238 + }, + { + "epoch": 0.3511491760624458, + "grad_norm": 2.9802472591400146, + "learning_rate": 1.7723052371971566e-05, + "loss": 0.0859, + "step": 3239 + }, + { + "epoch": 0.3512575888985256, + "grad_norm": 2.6287336349487305, + "learning_rate": 1.7722326998404182e-05, + "loss": 0.044, + "step": 3240 + }, + { + "epoch": 0.35136600173460536, + "grad_norm": 0.8784615993499756, + "learning_rate": 1.7721601624836794e-05, + "loss": 0.0142, + "step": 3241 + }, + { + "epoch": 0.3514744145706852, + "grad_norm": 4.641603469848633, + "learning_rate": 1.7720876251269406e-05, + "loss": 0.0581, + "step": 3242 + }, + { + "epoch": 0.35158282740676494, + "grad_norm": 5.7267537117004395, + "learning_rate": 1.7720150877702018e-05, + "loss": 0.026, + "step": 3243 + }, + { + "epoch": 0.35169124024284476, + "grad_norm": 2.901841402053833, + "learning_rate": 1.771942550413463e-05, + "loss": 0.0392, + "step": 3244 + }, + { + "epoch": 0.3517996530789245, + "grad_norm": 9.716276168823242, + "learning_rate": 1.7718700130567242e-05, + "loss": 0.0524, + "step": 3245 + }, + { + "epoch": 0.35190806591500434, + "grad_norm": 4.1823410987854, + "learning_rate": 1.7717974756999855e-05, + "loss": 0.0577, + "step": 3246 + }, + { + "epoch": 0.3520164787510841, + "grad_norm": 2.638357162475586, + "learning_rate": 1.771724938343247e-05, + "loss": 0.0316, + "step": 3247 + }, + { + "epoch": 0.3521248915871639, + "grad_norm": 1.5895200967788696, + "learning_rate": 1.7716524009865082e-05, + "loss": 0.0193, + "step": 3248 + }, + { + "epoch": 0.3522333044232437, + "grad_norm": 4.907148361206055, + "learning_rate": 1.7715798636297694e-05, + "loss": 0.102, + "step": 3249 + }, + { + "epoch": 0.3523417172593235, + "grad_norm": 2.8896467685699463, + "learning_rate": 1.7715073262730307e-05, + "loss": 0.0347, + "step": 3250 + }, + { + "epoch": 0.35245013009540327, + "grad_norm": 1.4423991441726685, + "learning_rate": 1.771434788916292e-05, + "loss": 0.015, + "step": 3251 + }, + { + "epoch": 0.3525585429314831, + "grad_norm": 1.1363869905471802, + "learning_rate": 1.771362251559553e-05, + "loss": 0.0099, + "step": 3252 + }, + { + "epoch": 0.3526669557675629, + "grad_norm": 0.8695142269134521, + "learning_rate": 1.7712897142028146e-05, + "loss": 0.0031, + "step": 3253 + }, + { + "epoch": 0.35277536860364267, + "grad_norm": 4.665433883666992, + "learning_rate": 1.771217176846076e-05, + "loss": 0.028, + "step": 3254 + }, + { + "epoch": 0.3528837814397225, + "grad_norm": 6.110390663146973, + "learning_rate": 1.7711446394893374e-05, + "loss": 0.072, + "step": 3255 + }, + { + "epoch": 0.35299219427580225, + "grad_norm": 2.9954001903533936, + "learning_rate": 1.7710721021325986e-05, + "loss": 0.0214, + "step": 3256 + }, + { + "epoch": 0.35310060711188207, + "grad_norm": 1.5481573343276978, + "learning_rate": 1.7709995647758598e-05, + "loss": 0.0066, + "step": 3257 + }, + { + "epoch": 0.35320901994796183, + "grad_norm": 3.623018741607666, + "learning_rate": 1.770927027419121e-05, + "loss": 0.0476, + "step": 3258 + }, + { + "epoch": 0.35331743278404165, + "grad_norm": 0.4959982931613922, + "learning_rate": 1.7708544900623823e-05, + "loss": 0.0063, + "step": 3259 + }, + { + "epoch": 0.3534258456201214, + "grad_norm": 1.935157060623169, + "learning_rate": 1.7707819527056435e-05, + "loss": 0.0318, + "step": 3260 + }, + { + "epoch": 0.35353425845620123, + "grad_norm": 6.133102893829346, + "learning_rate": 1.7707094153489047e-05, + "loss": 0.0215, + "step": 3261 + }, + { + "epoch": 0.353642671292281, + "grad_norm": 2.480440616607666, + "learning_rate": 1.7706368779921662e-05, + "loss": 0.0346, + "step": 3262 + }, + { + "epoch": 0.3537510841283608, + "grad_norm": 4.6252522468566895, + "learning_rate": 1.7705643406354275e-05, + "loss": 0.0627, + "step": 3263 + }, + { + "epoch": 0.3538594969644406, + "grad_norm": 8.276698112487793, + "learning_rate": 1.7704918032786887e-05, + "loss": 0.0446, + "step": 3264 + }, + { + "epoch": 0.3539679098005204, + "grad_norm": 3.5522003173828125, + "learning_rate": 1.77041926592195e-05, + "loss": 0.049, + "step": 3265 + }, + { + "epoch": 0.35407632263660016, + "grad_norm": 1.8767521381378174, + "learning_rate": 1.770346728565211e-05, + "loss": 0.0351, + "step": 3266 + }, + { + "epoch": 0.35418473547268, + "grad_norm": 7.052039623260498, + "learning_rate": 1.7702741912084723e-05, + "loss": 0.1377, + "step": 3267 + }, + { + "epoch": 0.35429314830875974, + "grad_norm": 3.7074553966522217, + "learning_rate": 1.770201653851734e-05, + "loss": 0.0431, + "step": 3268 + }, + { + "epoch": 0.35440156114483956, + "grad_norm": 4.575679302215576, + "learning_rate": 1.770129116494995e-05, + "loss": 0.103, + "step": 3269 + }, + { + "epoch": 0.3545099739809193, + "grad_norm": 7.0246968269348145, + "learning_rate": 1.7700565791382563e-05, + "loss": 0.0135, + "step": 3270 + }, + { + "epoch": 0.35461838681699914, + "grad_norm": 1.3685976266860962, + "learning_rate": 1.769984041781518e-05, + "loss": 0.015, + "step": 3271 + }, + { + "epoch": 0.3547267996530789, + "grad_norm": 1.5214653015136719, + "learning_rate": 1.769911504424779e-05, + "loss": 0.0074, + "step": 3272 + }, + { + "epoch": 0.3548352124891587, + "grad_norm": 2.614352226257324, + "learning_rate": 1.7698389670680403e-05, + "loss": 0.0372, + "step": 3273 + }, + { + "epoch": 0.3549436253252385, + "grad_norm": 3.1107585430145264, + "learning_rate": 1.7697664297113015e-05, + "loss": 0.0433, + "step": 3274 + }, + { + "epoch": 0.3550520381613183, + "grad_norm": 1.2371162176132202, + "learning_rate": 1.7696938923545627e-05, + "loss": 0.0136, + "step": 3275 + }, + { + "epoch": 0.35516045099739807, + "grad_norm": 4.565176486968994, + "learning_rate": 1.769621354997824e-05, + "loss": 0.0642, + "step": 3276 + }, + { + "epoch": 0.3552688638334779, + "grad_norm": 2.8124310970306396, + "learning_rate": 1.7695488176410855e-05, + "loss": 0.0507, + "step": 3277 + }, + { + "epoch": 0.35537727666955765, + "grad_norm": 3.9391891956329346, + "learning_rate": 1.7694762802843467e-05, + "loss": 0.0426, + "step": 3278 + }, + { + "epoch": 0.35548568950563747, + "grad_norm": 2.5216517448425293, + "learning_rate": 1.769403742927608e-05, + "loss": 0.0558, + "step": 3279 + }, + { + "epoch": 0.35559410234171723, + "grad_norm": 8.554863929748535, + "learning_rate": 1.769331205570869e-05, + "loss": 0.1106, + "step": 3280 + }, + { + "epoch": 0.35570251517779705, + "grad_norm": 4.087806224822998, + "learning_rate": 1.7692586682141303e-05, + "loss": 0.024, + "step": 3281 + }, + { + "epoch": 0.35581092801387687, + "grad_norm": 5.625392436981201, + "learning_rate": 1.7691861308573915e-05, + "loss": 0.0775, + "step": 3282 + }, + { + "epoch": 0.35591934084995663, + "grad_norm": 0.5689032077789307, + "learning_rate": 1.7691135935006528e-05, + "loss": 0.0096, + "step": 3283 + }, + { + "epoch": 0.35602775368603645, + "grad_norm": 2.9933950901031494, + "learning_rate": 1.7690410561439143e-05, + "loss": 0.0217, + "step": 3284 + }, + { + "epoch": 0.3561361665221162, + "grad_norm": 2.267369031906128, + "learning_rate": 1.7689685187871755e-05, + "loss": 0.0192, + "step": 3285 + }, + { + "epoch": 0.35624457935819603, + "grad_norm": 2.7942545413970947, + "learning_rate": 1.7688959814304367e-05, + "loss": 0.0435, + "step": 3286 + }, + { + "epoch": 0.3563529921942758, + "grad_norm": 3.6413910388946533, + "learning_rate": 1.768823444073698e-05, + "loss": 0.0841, + "step": 3287 + }, + { + "epoch": 0.3564614050303556, + "grad_norm": 5.335748672485352, + "learning_rate": 1.7687509067169595e-05, + "loss": 0.0501, + "step": 3288 + }, + { + "epoch": 0.3565698178664354, + "grad_norm": 1.670711874961853, + "learning_rate": 1.7686783693602207e-05, + "loss": 0.02, + "step": 3289 + }, + { + "epoch": 0.3566782307025152, + "grad_norm": 3.8573479652404785, + "learning_rate": 1.768605832003482e-05, + "loss": 0.0231, + "step": 3290 + }, + { + "epoch": 0.35678664353859496, + "grad_norm": 3.0647501945495605, + "learning_rate": 1.768533294646743e-05, + "loss": 0.0664, + "step": 3291 + }, + { + "epoch": 0.3568950563746748, + "grad_norm": 1.8566620349884033, + "learning_rate": 1.7684607572900047e-05, + "loss": 0.0216, + "step": 3292 + }, + { + "epoch": 0.35700346921075454, + "grad_norm": 2.576907157897949, + "learning_rate": 1.768388219933266e-05, + "loss": 0.0411, + "step": 3293 + }, + { + "epoch": 0.35711188204683436, + "grad_norm": 1.9695404767990112, + "learning_rate": 1.768315682576527e-05, + "loss": 0.0175, + "step": 3294 + }, + { + "epoch": 0.3572202948829141, + "grad_norm": 4.250535011291504, + "learning_rate": 1.7682431452197883e-05, + "loss": 0.0563, + "step": 3295 + }, + { + "epoch": 0.35732870771899394, + "grad_norm": 3.509188652038574, + "learning_rate": 1.7681706078630496e-05, + "loss": 0.0187, + "step": 3296 + }, + { + "epoch": 0.3574371205550737, + "grad_norm": 1.399837613105774, + "learning_rate": 1.7680980705063108e-05, + "loss": 0.0177, + "step": 3297 + }, + { + "epoch": 0.3575455333911535, + "grad_norm": 3.32088303565979, + "learning_rate": 1.768025533149572e-05, + "loss": 0.0729, + "step": 3298 + }, + { + "epoch": 0.3576539462272333, + "grad_norm": 4.144646167755127, + "learning_rate": 1.7679529957928335e-05, + "loss": 0.0215, + "step": 3299 + }, + { + "epoch": 0.3577623590633131, + "grad_norm": 8.964085578918457, + "learning_rate": 1.7678804584360948e-05, + "loss": 0.0376, + "step": 3300 + }, + { + "epoch": 0.35787077189939287, + "grad_norm": 3.575589418411255, + "learning_rate": 1.767807921079356e-05, + "loss": 0.0412, + "step": 3301 + }, + { + "epoch": 0.3579791847354727, + "grad_norm": 3.999887228012085, + "learning_rate": 1.7677353837226172e-05, + "loss": 0.0684, + "step": 3302 + }, + { + "epoch": 0.35808759757155245, + "grad_norm": 1.1187704801559448, + "learning_rate": 1.7676628463658784e-05, + "loss": 0.0151, + "step": 3303 + }, + { + "epoch": 0.35819601040763227, + "grad_norm": 3.453901767730713, + "learning_rate": 1.7675903090091396e-05, + "loss": 0.0361, + "step": 3304 + }, + { + "epoch": 0.35830442324371203, + "grad_norm": 1.5327162742614746, + "learning_rate": 1.767517771652401e-05, + "loss": 0.0141, + "step": 3305 + }, + { + "epoch": 0.35841283607979185, + "grad_norm": 5.916789531707764, + "learning_rate": 1.7674452342956624e-05, + "loss": 0.061, + "step": 3306 + }, + { + "epoch": 0.3585212489158716, + "grad_norm": 6.69876766204834, + "learning_rate": 1.767372696938924e-05, + "loss": 0.0433, + "step": 3307 + }, + { + "epoch": 0.35862966175195143, + "grad_norm": 2.8612499237060547, + "learning_rate": 1.767300159582185e-05, + "loss": 0.0203, + "step": 3308 + }, + { + "epoch": 0.3587380745880312, + "grad_norm": 3.4510366916656494, + "learning_rate": 1.7672276222254464e-05, + "loss": 0.0901, + "step": 3309 + }, + { + "epoch": 0.358846487424111, + "grad_norm": 3.637334108352661, + "learning_rate": 1.7671550848687076e-05, + "loss": 0.0767, + "step": 3310 + }, + { + "epoch": 0.3589549002601908, + "grad_norm": 4.705819129943848, + "learning_rate": 1.7670825475119688e-05, + "loss": 0.0732, + "step": 3311 + }, + { + "epoch": 0.3590633130962706, + "grad_norm": 3.964461326599121, + "learning_rate": 1.76701001015523e-05, + "loss": 0.0381, + "step": 3312 + }, + { + "epoch": 0.3591717259323504, + "grad_norm": 1.2988874912261963, + "learning_rate": 1.7669374727984912e-05, + "loss": 0.0324, + "step": 3313 + }, + { + "epoch": 0.3592801387684302, + "grad_norm": 4.3727641105651855, + "learning_rate": 1.7668649354417528e-05, + "loss": 0.0396, + "step": 3314 + }, + { + "epoch": 0.35938855160451, + "grad_norm": 5.508593559265137, + "learning_rate": 1.766792398085014e-05, + "loss": 0.05, + "step": 3315 + }, + { + "epoch": 0.35949696444058976, + "grad_norm": 4.379275321960449, + "learning_rate": 1.7667198607282752e-05, + "loss": 0.0841, + "step": 3316 + }, + { + "epoch": 0.3596053772766696, + "grad_norm": 4.121335983276367, + "learning_rate": 1.7666473233715364e-05, + "loss": 0.0509, + "step": 3317 + }, + { + "epoch": 0.35971379011274934, + "grad_norm": 1.8700456619262695, + "learning_rate": 1.7665747860147976e-05, + "loss": 0.0143, + "step": 3318 + }, + { + "epoch": 0.35982220294882916, + "grad_norm": 5.432994365692139, + "learning_rate": 1.766502248658059e-05, + "loss": 0.0345, + "step": 3319 + }, + { + "epoch": 0.3599306157849089, + "grad_norm": 4.550700664520264, + "learning_rate": 1.7664297113013204e-05, + "loss": 0.0819, + "step": 3320 + }, + { + "epoch": 0.36003902862098874, + "grad_norm": 4.712596416473389, + "learning_rate": 1.7663571739445816e-05, + "loss": 0.0785, + "step": 3321 + }, + { + "epoch": 0.3601474414570685, + "grad_norm": 2.7601263523101807, + "learning_rate": 1.7662846365878428e-05, + "loss": 0.0477, + "step": 3322 + }, + { + "epoch": 0.3602558542931483, + "grad_norm": 3.7137842178344727, + "learning_rate": 1.7662120992311044e-05, + "loss": 0.0315, + "step": 3323 + }, + { + "epoch": 0.3603642671292281, + "grad_norm": 5.942421913146973, + "learning_rate": 1.7661395618743656e-05, + "loss": 0.11, + "step": 3324 + }, + { + "epoch": 0.3604726799653079, + "grad_norm": 4.82403564453125, + "learning_rate": 1.7660670245176268e-05, + "loss": 0.0389, + "step": 3325 + }, + { + "epoch": 0.36058109280138767, + "grad_norm": 2.8022239208221436, + "learning_rate": 1.765994487160888e-05, + "loss": 0.0811, + "step": 3326 + }, + { + "epoch": 0.3606895056374675, + "grad_norm": 1.440161108970642, + "learning_rate": 1.7659219498041492e-05, + "loss": 0.0296, + "step": 3327 + }, + { + "epoch": 0.36079791847354725, + "grad_norm": 3.9734458923339844, + "learning_rate": 1.7658494124474105e-05, + "loss": 0.0444, + "step": 3328 + }, + { + "epoch": 0.36090633130962707, + "grad_norm": 2.870957851409912, + "learning_rate": 1.765776875090672e-05, + "loss": 0.0366, + "step": 3329 + }, + { + "epoch": 0.36101474414570683, + "grad_norm": 1.9035803079605103, + "learning_rate": 1.7657043377339332e-05, + "loss": 0.0304, + "step": 3330 + }, + { + "epoch": 0.36112315698178665, + "grad_norm": 3.1779863834381104, + "learning_rate": 1.7656318003771944e-05, + "loss": 0.0535, + "step": 3331 + }, + { + "epoch": 0.3612315698178664, + "grad_norm": 2.008668899536133, + "learning_rate": 1.7655592630204556e-05, + "loss": 0.0207, + "step": 3332 + }, + { + "epoch": 0.36133998265394623, + "grad_norm": 2.9814529418945312, + "learning_rate": 1.765486725663717e-05, + "loss": 0.0653, + "step": 3333 + }, + { + "epoch": 0.361448395490026, + "grad_norm": 5.269363880157471, + "learning_rate": 1.765414188306978e-05, + "loss": 0.0684, + "step": 3334 + }, + { + "epoch": 0.3615568083261058, + "grad_norm": 1.5598108768463135, + "learning_rate": 1.7653416509502393e-05, + "loss": 0.0349, + "step": 3335 + }, + { + "epoch": 0.3616652211621856, + "grad_norm": 2.174180030822754, + "learning_rate": 1.765269113593501e-05, + "loss": 0.035, + "step": 3336 + }, + { + "epoch": 0.3617736339982654, + "grad_norm": 2.0591256618499756, + "learning_rate": 1.765196576236762e-05, + "loss": 0.025, + "step": 3337 + }, + { + "epoch": 0.36188204683434516, + "grad_norm": 2.818218231201172, + "learning_rate": 1.7651240388800233e-05, + "loss": 0.0546, + "step": 3338 + }, + { + "epoch": 0.361990459670425, + "grad_norm": 3.551995038986206, + "learning_rate": 1.7650515015232845e-05, + "loss": 0.0371, + "step": 3339 + }, + { + "epoch": 0.36209887250650474, + "grad_norm": 4.054752349853516, + "learning_rate": 1.764978964166546e-05, + "loss": 0.0454, + "step": 3340 + }, + { + "epoch": 0.36220728534258456, + "grad_norm": 2.0855915546417236, + "learning_rate": 1.7649064268098073e-05, + "loss": 0.0584, + "step": 3341 + }, + { + "epoch": 0.3623156981786644, + "grad_norm": 1.6504932641983032, + "learning_rate": 1.7648338894530685e-05, + "loss": 0.0333, + "step": 3342 + }, + { + "epoch": 0.36242411101474414, + "grad_norm": 1.86553156375885, + "learning_rate": 1.7647613520963297e-05, + "loss": 0.042, + "step": 3343 + }, + { + "epoch": 0.36253252385082396, + "grad_norm": 1.4016045331954956, + "learning_rate": 1.7646888147395912e-05, + "loss": 0.0167, + "step": 3344 + }, + { + "epoch": 0.3626409366869037, + "grad_norm": 2.8828933238983154, + "learning_rate": 1.7646162773828525e-05, + "loss": 0.038, + "step": 3345 + }, + { + "epoch": 0.36274934952298354, + "grad_norm": 2.6272573471069336, + "learning_rate": 1.7645437400261137e-05, + "loss": 0.0225, + "step": 3346 + }, + { + "epoch": 0.3628577623590633, + "grad_norm": 5.604996204376221, + "learning_rate": 1.764471202669375e-05, + "loss": 0.1124, + "step": 3347 + }, + { + "epoch": 0.3629661751951431, + "grad_norm": 2.3619489669799805, + "learning_rate": 1.764398665312636e-05, + "loss": 0.0338, + "step": 3348 + }, + { + "epoch": 0.3630745880312229, + "grad_norm": 3.2770228385925293, + "learning_rate": 1.7643261279558973e-05, + "loss": 0.0364, + "step": 3349 + }, + { + "epoch": 0.3631830008673027, + "grad_norm": 3.36732816696167, + "learning_rate": 1.7642535905991585e-05, + "loss": 0.0807, + "step": 3350 + }, + { + "epoch": 0.36329141370338247, + "grad_norm": 3.3916749954223633, + "learning_rate": 1.76418105324242e-05, + "loss": 0.0415, + "step": 3351 + }, + { + "epoch": 0.3633998265394623, + "grad_norm": 2.574599504470825, + "learning_rate": 1.7641085158856813e-05, + "loss": 0.0508, + "step": 3352 + }, + { + "epoch": 0.36350823937554205, + "grad_norm": 0.9439007043838501, + "learning_rate": 1.7640359785289425e-05, + "loss": 0.0144, + "step": 3353 + }, + { + "epoch": 0.36361665221162187, + "grad_norm": 3.3448939323425293, + "learning_rate": 1.7639634411722037e-05, + "loss": 0.0409, + "step": 3354 + }, + { + "epoch": 0.36372506504770163, + "grad_norm": 5.946524620056152, + "learning_rate": 1.763890903815465e-05, + "loss": 0.0263, + "step": 3355 + }, + { + "epoch": 0.36383347788378145, + "grad_norm": 4.873315334320068, + "learning_rate": 1.763818366458726e-05, + "loss": 0.0672, + "step": 3356 + }, + { + "epoch": 0.3639418907198612, + "grad_norm": 5.725998401641846, + "learning_rate": 1.7637458291019877e-05, + "loss": 0.0648, + "step": 3357 + }, + { + "epoch": 0.36405030355594103, + "grad_norm": 0.410131573677063, + "learning_rate": 1.763673291745249e-05, + "loss": 0.0043, + "step": 3358 + }, + { + "epoch": 0.3641587163920208, + "grad_norm": 3.8514671325683594, + "learning_rate": 1.7636007543885105e-05, + "loss": 0.0199, + "step": 3359 + }, + { + "epoch": 0.3642671292281006, + "grad_norm": 3.556917905807495, + "learning_rate": 1.7635282170317717e-05, + "loss": 0.0209, + "step": 3360 + }, + { + "epoch": 0.3643755420641804, + "grad_norm": 3.524077892303467, + "learning_rate": 1.763455679675033e-05, + "loss": 0.0673, + "step": 3361 + }, + { + "epoch": 0.3644839549002602, + "grad_norm": 3.324150800704956, + "learning_rate": 1.763383142318294e-05, + "loss": 0.0489, + "step": 3362 + }, + { + "epoch": 0.36459236773633996, + "grad_norm": 2.921954393386841, + "learning_rate": 1.7633106049615553e-05, + "loss": 0.0385, + "step": 3363 + }, + { + "epoch": 0.3647007805724198, + "grad_norm": 8.674083709716797, + "learning_rate": 1.7632380676048165e-05, + "loss": 0.0491, + "step": 3364 + }, + { + "epoch": 0.36480919340849954, + "grad_norm": 2.9331343173980713, + "learning_rate": 1.7631655302480778e-05, + "loss": 0.0448, + "step": 3365 + }, + { + "epoch": 0.36491760624457936, + "grad_norm": 1.2665650844573975, + "learning_rate": 1.7630929928913393e-05, + "loss": 0.0116, + "step": 3366 + }, + { + "epoch": 0.3650260190806591, + "grad_norm": 2.8756871223449707, + "learning_rate": 1.7630204555346005e-05, + "loss": 0.034, + "step": 3367 + }, + { + "epoch": 0.36513443191673894, + "grad_norm": 1.7696934938430786, + "learning_rate": 1.7629479181778617e-05, + "loss": 0.0496, + "step": 3368 + }, + { + "epoch": 0.3652428447528187, + "grad_norm": 1.1263160705566406, + "learning_rate": 1.762875380821123e-05, + "loss": 0.0102, + "step": 3369 + }, + { + "epoch": 0.3653512575888985, + "grad_norm": 4.66292667388916, + "learning_rate": 1.762802843464384e-05, + "loss": 0.0703, + "step": 3370 + }, + { + "epoch": 0.36545967042497834, + "grad_norm": 2.950810194015503, + "learning_rate": 1.7627303061076454e-05, + "loss": 0.0587, + "step": 3371 + }, + { + "epoch": 0.3655680832610581, + "grad_norm": 1.144696831703186, + "learning_rate": 1.762657768750907e-05, + "loss": 0.013, + "step": 3372 + }, + { + "epoch": 0.3656764960971379, + "grad_norm": 4.172791957855225, + "learning_rate": 1.762585231394168e-05, + "loss": 0.1121, + "step": 3373 + }, + { + "epoch": 0.3657849089332177, + "grad_norm": 2.1114871501922607, + "learning_rate": 1.7625126940374294e-05, + "loss": 0.0163, + "step": 3374 + }, + { + "epoch": 0.3658933217692975, + "grad_norm": 3.270481586456299, + "learning_rate": 1.762440156680691e-05, + "loss": 0.0544, + "step": 3375 + }, + { + "epoch": 0.36600173460537727, + "grad_norm": 5.040493488311768, + "learning_rate": 1.762367619323952e-05, + "loss": 0.032, + "step": 3376 + }, + { + "epoch": 0.3661101474414571, + "grad_norm": 3.731886625289917, + "learning_rate": 1.7622950819672133e-05, + "loss": 0.0152, + "step": 3377 + }, + { + "epoch": 0.36621856027753685, + "grad_norm": 7.004007339477539, + "learning_rate": 1.7622225446104746e-05, + "loss": 0.0387, + "step": 3378 + }, + { + "epoch": 0.36632697311361667, + "grad_norm": 4.612916946411133, + "learning_rate": 1.7621500072537358e-05, + "loss": 0.0612, + "step": 3379 + }, + { + "epoch": 0.36643538594969643, + "grad_norm": 3.0453789234161377, + "learning_rate": 1.762077469896997e-05, + "loss": 0.0413, + "step": 3380 + }, + { + "epoch": 0.36654379878577625, + "grad_norm": 7.455329895019531, + "learning_rate": 1.7620049325402585e-05, + "loss": 0.092, + "step": 3381 + }, + { + "epoch": 0.366652211621856, + "grad_norm": 5.112606048583984, + "learning_rate": 1.7619323951835198e-05, + "loss": 0.0452, + "step": 3382 + }, + { + "epoch": 0.36676062445793584, + "grad_norm": 2.5106630325317383, + "learning_rate": 1.761859857826781e-05, + "loss": 0.0369, + "step": 3383 + }, + { + "epoch": 0.3668690372940156, + "grad_norm": 1.168839693069458, + "learning_rate": 1.7617873204700422e-05, + "loss": 0.0161, + "step": 3384 + }, + { + "epoch": 0.3669774501300954, + "grad_norm": 1.6202051639556885, + "learning_rate": 1.7617147831133034e-05, + "loss": 0.0378, + "step": 3385 + }, + { + "epoch": 0.3670858629661752, + "grad_norm": 0.32917705178260803, + "learning_rate": 1.7616422457565646e-05, + "loss": 0.0043, + "step": 3386 + }, + { + "epoch": 0.367194275802255, + "grad_norm": 2.9827897548675537, + "learning_rate": 1.7615697083998258e-05, + "loss": 0.0483, + "step": 3387 + }, + { + "epoch": 0.36730268863833476, + "grad_norm": 1.2000956535339355, + "learning_rate": 1.7614971710430874e-05, + "loss": 0.0153, + "step": 3388 + }, + { + "epoch": 0.3674111014744146, + "grad_norm": 5.267902374267578, + "learning_rate": 1.7614246336863486e-05, + "loss": 0.0486, + "step": 3389 + }, + { + "epoch": 0.36751951431049434, + "grad_norm": 2.369145631790161, + "learning_rate": 1.7613520963296098e-05, + "loss": 0.0349, + "step": 3390 + }, + { + "epoch": 0.36762792714657416, + "grad_norm": 3.2062888145446777, + "learning_rate": 1.761279558972871e-05, + "loss": 0.049, + "step": 3391 + }, + { + "epoch": 0.3677363399826539, + "grad_norm": 4.085447311401367, + "learning_rate": 1.7612070216161326e-05, + "loss": 0.0367, + "step": 3392 + }, + { + "epoch": 0.36784475281873374, + "grad_norm": 3.7221548557281494, + "learning_rate": 1.7611344842593938e-05, + "loss": 0.035, + "step": 3393 + }, + { + "epoch": 0.3679531656548135, + "grad_norm": 2.6384365558624268, + "learning_rate": 1.761061946902655e-05, + "loss": 0.0255, + "step": 3394 + }, + { + "epoch": 0.3680615784908933, + "grad_norm": 2.1876041889190674, + "learning_rate": 1.7609894095459162e-05, + "loss": 0.0258, + "step": 3395 + }, + { + "epoch": 0.3681699913269731, + "grad_norm": 4.872441291809082, + "learning_rate": 1.7609168721891778e-05, + "loss": 0.0609, + "step": 3396 + }, + { + "epoch": 0.3682784041630529, + "grad_norm": 1.1177605390548706, + "learning_rate": 1.760844334832439e-05, + "loss": 0.0201, + "step": 3397 + }, + { + "epoch": 0.36838681699913267, + "grad_norm": 15.853787422180176, + "learning_rate": 1.7607717974757002e-05, + "loss": 0.033, + "step": 3398 + }, + { + "epoch": 0.3684952298352125, + "grad_norm": 3.6857316493988037, + "learning_rate": 1.7606992601189614e-05, + "loss": 0.0399, + "step": 3399 + }, + { + "epoch": 0.3686036426712923, + "grad_norm": 8.354720115661621, + "learning_rate": 1.7606267227622226e-05, + "loss": 0.0625, + "step": 3400 + }, + { + "epoch": 0.36871205550737207, + "grad_norm": 8.775534629821777, + "learning_rate": 1.760554185405484e-05, + "loss": 0.0446, + "step": 3401 + }, + { + "epoch": 0.3688204683434519, + "grad_norm": 6.753998756408691, + "learning_rate": 1.760481648048745e-05, + "loss": 0.0236, + "step": 3402 + }, + { + "epoch": 0.36892888117953165, + "grad_norm": 5.216869831085205, + "learning_rate": 1.7604091106920066e-05, + "loss": 0.0475, + "step": 3403 + }, + { + "epoch": 0.3690372940156115, + "grad_norm": 2.698894739151001, + "learning_rate": 1.7603365733352678e-05, + "loss": 0.0155, + "step": 3404 + }, + { + "epoch": 0.36914570685169124, + "grad_norm": 1.859431505203247, + "learning_rate": 1.760264035978529e-05, + "loss": 0.0161, + "step": 3405 + }, + { + "epoch": 0.36925411968777105, + "grad_norm": 0.7965856790542603, + "learning_rate": 1.7601914986217903e-05, + "loss": 0.0163, + "step": 3406 + }, + { + "epoch": 0.3693625325238508, + "grad_norm": 2.9245753288269043, + "learning_rate": 1.7601189612650515e-05, + "loss": 0.0582, + "step": 3407 + }, + { + "epoch": 0.36947094535993064, + "grad_norm": 0.8864175081253052, + "learning_rate": 1.7600464239083127e-05, + "loss": 0.0083, + "step": 3408 + }, + { + "epoch": 0.3695793581960104, + "grad_norm": 5.277891159057617, + "learning_rate": 1.7599738865515742e-05, + "loss": 0.0926, + "step": 3409 + }, + { + "epoch": 0.3696877710320902, + "grad_norm": 12.047258377075195, + "learning_rate": 1.7599013491948354e-05, + "loss": 0.0879, + "step": 3410 + }, + { + "epoch": 0.36979618386817, + "grad_norm": 1.2177846431732178, + "learning_rate": 1.759828811838097e-05, + "loss": 0.0087, + "step": 3411 + }, + { + "epoch": 0.3699045967042498, + "grad_norm": 4.528152942657471, + "learning_rate": 1.7597562744813582e-05, + "loss": 0.0538, + "step": 3412 + }, + { + "epoch": 0.37001300954032956, + "grad_norm": 3.877981185913086, + "learning_rate": 1.7596837371246194e-05, + "loss": 0.0203, + "step": 3413 + }, + { + "epoch": 0.3701214223764094, + "grad_norm": 2.0778448581695557, + "learning_rate": 1.7596111997678806e-05, + "loss": 0.0387, + "step": 3414 + }, + { + "epoch": 0.37022983521248914, + "grad_norm": 2.1995675563812256, + "learning_rate": 1.759538662411142e-05, + "loss": 0.0414, + "step": 3415 + }, + { + "epoch": 0.37033824804856896, + "grad_norm": 5.837597370147705, + "learning_rate": 1.759466125054403e-05, + "loss": 0.0295, + "step": 3416 + }, + { + "epoch": 0.3704466608846487, + "grad_norm": 3.6985795497894287, + "learning_rate": 1.7593935876976643e-05, + "loss": 0.0389, + "step": 3417 + }, + { + "epoch": 0.37055507372072854, + "grad_norm": 6.733206748962402, + "learning_rate": 1.759321050340926e-05, + "loss": 0.0777, + "step": 3418 + }, + { + "epoch": 0.3706634865568083, + "grad_norm": 2.7197952270507812, + "learning_rate": 1.759248512984187e-05, + "loss": 0.0453, + "step": 3419 + }, + { + "epoch": 0.3707718993928881, + "grad_norm": 2.9418582916259766, + "learning_rate": 1.7591759756274483e-05, + "loss": 0.042, + "step": 3420 + }, + { + "epoch": 0.3708803122289679, + "grad_norm": 5.173525810241699, + "learning_rate": 1.7591034382707095e-05, + "loss": 0.047, + "step": 3421 + }, + { + "epoch": 0.3709887250650477, + "grad_norm": 5.482346534729004, + "learning_rate": 1.7590309009139707e-05, + "loss": 0.0452, + "step": 3422 + }, + { + "epoch": 0.37109713790112747, + "grad_norm": 6.173842430114746, + "learning_rate": 1.758958363557232e-05, + "loss": 0.059, + "step": 3423 + }, + { + "epoch": 0.3712055507372073, + "grad_norm": 4.397928714752197, + "learning_rate": 1.7588858262004935e-05, + "loss": 0.0986, + "step": 3424 + }, + { + "epoch": 0.37131396357328705, + "grad_norm": 3.0845065116882324, + "learning_rate": 1.7588132888437547e-05, + "loss": 0.0295, + "step": 3425 + }, + { + "epoch": 0.3714223764093669, + "grad_norm": 3.5119800567626953, + "learning_rate": 1.758740751487016e-05, + "loss": 0.033, + "step": 3426 + }, + { + "epoch": 0.37153078924544664, + "grad_norm": 2.2063965797424316, + "learning_rate": 1.7586682141302774e-05, + "loss": 0.0368, + "step": 3427 + }, + { + "epoch": 0.37163920208152645, + "grad_norm": 5.998653411865234, + "learning_rate": 1.7585956767735387e-05, + "loss": 0.0373, + "step": 3428 + }, + { + "epoch": 0.3717476149176062, + "grad_norm": 8.789198875427246, + "learning_rate": 1.7585231394168e-05, + "loss": 0.115, + "step": 3429 + }, + { + "epoch": 0.37185602775368604, + "grad_norm": 2.87369441986084, + "learning_rate": 1.758450602060061e-05, + "loss": 0.0673, + "step": 3430 + }, + { + "epoch": 0.37196444058976585, + "grad_norm": 3.4387388229370117, + "learning_rate": 1.7583780647033223e-05, + "loss": 0.0562, + "step": 3431 + }, + { + "epoch": 0.3720728534258456, + "grad_norm": 2.462695598602295, + "learning_rate": 1.7583055273465835e-05, + "loss": 0.0478, + "step": 3432 + }, + { + "epoch": 0.37218126626192544, + "grad_norm": 3.227652072906494, + "learning_rate": 1.758232989989845e-05, + "loss": 0.0522, + "step": 3433 + }, + { + "epoch": 0.3722896790980052, + "grad_norm": 2.7471134662628174, + "learning_rate": 1.7581604526331063e-05, + "loss": 0.0323, + "step": 3434 + }, + { + "epoch": 0.372398091934085, + "grad_norm": 2.5121285915374756, + "learning_rate": 1.7580879152763675e-05, + "loss": 0.0507, + "step": 3435 + }, + { + "epoch": 0.3725065047701648, + "grad_norm": 4.905398368835449, + "learning_rate": 1.7580153779196287e-05, + "loss": 0.0857, + "step": 3436 + }, + { + "epoch": 0.3726149176062446, + "grad_norm": 1.8415504693984985, + "learning_rate": 1.75794284056289e-05, + "loss": 0.0241, + "step": 3437 + }, + { + "epoch": 0.37272333044232436, + "grad_norm": 2.52950382232666, + "learning_rate": 1.757870303206151e-05, + "loss": 0.02, + "step": 3438 + }, + { + "epoch": 0.3728317432784042, + "grad_norm": 2.799156904220581, + "learning_rate": 1.7577977658494127e-05, + "loss": 0.0383, + "step": 3439 + }, + { + "epoch": 0.37294015611448394, + "grad_norm": 3.0263752937316895, + "learning_rate": 1.757725228492674e-05, + "loss": 0.0639, + "step": 3440 + }, + { + "epoch": 0.37304856895056376, + "grad_norm": 1.4097235202789307, + "learning_rate": 1.757652691135935e-05, + "loss": 0.0169, + "step": 3441 + }, + { + "epoch": 0.3731569817866435, + "grad_norm": 3.36704421043396, + "learning_rate": 1.7575801537791963e-05, + "loss": 0.1032, + "step": 3442 + }, + { + "epoch": 0.37326539462272335, + "grad_norm": 1.515661597251892, + "learning_rate": 1.757507616422458e-05, + "loss": 0.0309, + "step": 3443 + }, + { + "epoch": 0.3733738074588031, + "grad_norm": 2.274108648300171, + "learning_rate": 1.757435079065719e-05, + "loss": 0.0273, + "step": 3444 + }, + { + "epoch": 0.3734822202948829, + "grad_norm": 2.2597551345825195, + "learning_rate": 1.7573625417089803e-05, + "loss": 0.0332, + "step": 3445 + }, + { + "epoch": 0.3735906331309627, + "grad_norm": 2.508629322052002, + "learning_rate": 1.7572900043522415e-05, + "loss": 0.0345, + "step": 3446 + }, + { + "epoch": 0.3736990459670425, + "grad_norm": 3.657127618789673, + "learning_rate": 1.7572174669955028e-05, + "loss": 0.0638, + "step": 3447 + }, + { + "epoch": 0.3738074588031223, + "grad_norm": 1.2817801237106323, + "learning_rate": 1.7571449296387643e-05, + "loss": 0.0422, + "step": 3448 + }, + { + "epoch": 0.3739158716392021, + "grad_norm": 3.90838885307312, + "learning_rate": 1.7570723922820255e-05, + "loss": 0.0749, + "step": 3449 + }, + { + "epoch": 0.37402428447528185, + "grad_norm": 3.361041784286499, + "learning_rate": 1.7569998549252867e-05, + "loss": 0.0544, + "step": 3450 + }, + { + "epoch": 0.3741326973113617, + "grad_norm": 3.3427915573120117, + "learning_rate": 1.756927317568548e-05, + "loss": 0.0662, + "step": 3451 + }, + { + "epoch": 0.37424111014744144, + "grad_norm": 1.8027594089508057, + "learning_rate": 1.756854780211809e-05, + "loss": 0.0136, + "step": 3452 + }, + { + "epoch": 0.37434952298352125, + "grad_norm": 6.39603328704834, + "learning_rate": 1.7567822428550704e-05, + "loss": 0.0253, + "step": 3453 + }, + { + "epoch": 0.374457935819601, + "grad_norm": 3.9082584381103516, + "learning_rate": 1.7567097054983316e-05, + "loss": 0.0609, + "step": 3454 + }, + { + "epoch": 0.37456634865568084, + "grad_norm": 2.323439836502075, + "learning_rate": 1.756637168141593e-05, + "loss": 0.0243, + "step": 3455 + }, + { + "epoch": 0.3746747614917606, + "grad_norm": 1.3492424488067627, + "learning_rate": 1.7565646307848544e-05, + "loss": 0.0211, + "step": 3456 + }, + { + "epoch": 0.3747831743278404, + "grad_norm": 2.830587148666382, + "learning_rate": 1.7564920934281156e-05, + "loss": 0.0686, + "step": 3457 + }, + { + "epoch": 0.3748915871639202, + "grad_norm": 2.411001205444336, + "learning_rate": 1.7564195560713768e-05, + "loss": 0.0322, + "step": 3458 + }, + { + "epoch": 0.375, + "grad_norm": 2.540555238723755, + "learning_rate": 1.756347018714638e-05, + "loss": 0.0348, + "step": 3459 + }, + { + "epoch": 0.3751084128360798, + "grad_norm": 1.3288023471832275, + "learning_rate": 1.7562744813578992e-05, + "loss": 0.0246, + "step": 3460 + }, + { + "epoch": 0.3752168256721596, + "grad_norm": 1.3542094230651855, + "learning_rate": 1.7562019440011608e-05, + "loss": 0.0247, + "step": 3461 + }, + { + "epoch": 0.3753252385082394, + "grad_norm": 4.728233337402344, + "learning_rate": 1.756129406644422e-05, + "loss": 0.0362, + "step": 3462 + }, + { + "epoch": 0.37543365134431916, + "grad_norm": 1.9383946657180786, + "learning_rate": 1.7560568692876835e-05, + "loss": 0.0249, + "step": 3463 + }, + { + "epoch": 0.375542064180399, + "grad_norm": 1.9391286373138428, + "learning_rate": 1.7559843319309448e-05, + "loss": 0.0233, + "step": 3464 + }, + { + "epoch": 0.37565047701647875, + "grad_norm": 2.9091241359710693, + "learning_rate": 1.755911794574206e-05, + "loss": 0.0472, + "step": 3465 + }, + { + "epoch": 0.37575888985255856, + "grad_norm": 5.320124626159668, + "learning_rate": 1.7558392572174672e-05, + "loss": 0.0269, + "step": 3466 + }, + { + "epoch": 0.3758673026886383, + "grad_norm": 2.0823495388031006, + "learning_rate": 1.7557667198607284e-05, + "loss": 0.0373, + "step": 3467 + }, + { + "epoch": 0.37597571552471815, + "grad_norm": 2.828024387359619, + "learning_rate": 1.7556941825039896e-05, + "loss": 0.0323, + "step": 3468 + }, + { + "epoch": 0.3760841283607979, + "grad_norm": 1.2687617540359497, + "learning_rate": 1.7556216451472508e-05, + "loss": 0.0144, + "step": 3469 + }, + { + "epoch": 0.3761925411968777, + "grad_norm": 3.3887736797332764, + "learning_rate": 1.7555491077905124e-05, + "loss": 0.0434, + "step": 3470 + }, + { + "epoch": 0.3763009540329575, + "grad_norm": 1.1559077501296997, + "learning_rate": 1.7554765704337736e-05, + "loss": 0.0319, + "step": 3471 + }, + { + "epoch": 0.3764093668690373, + "grad_norm": 3.6922216415405273, + "learning_rate": 1.7554040330770348e-05, + "loss": 0.0451, + "step": 3472 + }, + { + "epoch": 0.3765177797051171, + "grad_norm": 7.011097431182861, + "learning_rate": 1.755331495720296e-05, + "loss": 0.0209, + "step": 3473 + }, + { + "epoch": 0.3766261925411969, + "grad_norm": 1.4560670852661133, + "learning_rate": 1.7552589583635572e-05, + "loss": 0.0216, + "step": 3474 + }, + { + "epoch": 0.37673460537727665, + "grad_norm": 3.960345983505249, + "learning_rate": 1.7551864210068184e-05, + "loss": 0.119, + "step": 3475 + }, + { + "epoch": 0.3768430182133565, + "grad_norm": 1.238897442817688, + "learning_rate": 1.75511388365008e-05, + "loss": 0.0257, + "step": 3476 + }, + { + "epoch": 0.37695143104943624, + "grad_norm": 5.224833965301514, + "learning_rate": 1.7550413462933412e-05, + "loss": 0.0473, + "step": 3477 + }, + { + "epoch": 0.37705984388551606, + "grad_norm": 2.590000629425049, + "learning_rate": 1.7549688089366024e-05, + "loss": 0.028, + "step": 3478 + }, + { + "epoch": 0.3771682567215958, + "grad_norm": 2.3540871143341064, + "learning_rate": 1.754896271579864e-05, + "loss": 0.0298, + "step": 3479 + }, + { + "epoch": 0.37727666955767564, + "grad_norm": 2.9856441020965576, + "learning_rate": 1.7548237342231252e-05, + "loss": 0.0473, + "step": 3480 + }, + { + "epoch": 0.3773850823937554, + "grad_norm": 3.207077980041504, + "learning_rate": 1.7547511968663864e-05, + "loss": 0.0506, + "step": 3481 + }, + { + "epoch": 0.3774934952298352, + "grad_norm": 3.615302801132202, + "learning_rate": 1.7546786595096476e-05, + "loss": 0.0657, + "step": 3482 + }, + { + "epoch": 0.377601908065915, + "grad_norm": 1.2822084426879883, + "learning_rate": 1.754606122152909e-05, + "loss": 0.0194, + "step": 3483 + }, + { + "epoch": 0.3777103209019948, + "grad_norm": 1.6056021451950073, + "learning_rate": 1.75453358479617e-05, + "loss": 0.0395, + "step": 3484 + }, + { + "epoch": 0.37781873373807456, + "grad_norm": 3.1505727767944336, + "learning_rate": 1.7544610474394316e-05, + "loss": 0.0149, + "step": 3485 + }, + { + "epoch": 0.3779271465741544, + "grad_norm": 2.681279182434082, + "learning_rate": 1.7543885100826928e-05, + "loss": 0.0232, + "step": 3486 + }, + { + "epoch": 0.37803555941023415, + "grad_norm": 2.6004347801208496, + "learning_rate": 1.754315972725954e-05, + "loss": 0.0424, + "step": 3487 + }, + { + "epoch": 0.37814397224631396, + "grad_norm": 1.0525063276290894, + "learning_rate": 1.7542434353692153e-05, + "loss": 0.0138, + "step": 3488 + }, + { + "epoch": 0.3782523850823938, + "grad_norm": 2.70304536819458, + "learning_rate": 1.7541708980124765e-05, + "loss": 0.0516, + "step": 3489 + }, + { + "epoch": 0.37836079791847355, + "grad_norm": 2.0456717014312744, + "learning_rate": 1.7540983606557377e-05, + "loss": 0.0192, + "step": 3490 + }, + { + "epoch": 0.37846921075455336, + "grad_norm": 4.469120025634766, + "learning_rate": 1.7540258232989992e-05, + "loss": 0.0472, + "step": 3491 + }, + { + "epoch": 0.3785776235906331, + "grad_norm": 4.875046730041504, + "learning_rate": 1.7539532859422604e-05, + "loss": 0.0596, + "step": 3492 + }, + { + "epoch": 0.37868603642671295, + "grad_norm": 2.6574485301971436, + "learning_rate": 1.7538807485855217e-05, + "loss": 0.0457, + "step": 3493 + }, + { + "epoch": 0.3787944492627927, + "grad_norm": 1.4874635934829712, + "learning_rate": 1.753808211228783e-05, + "loss": 0.0325, + "step": 3494 + }, + { + "epoch": 0.37890286209887253, + "grad_norm": 8.628400802612305, + "learning_rate": 1.7537356738720444e-05, + "loss": 0.0316, + "step": 3495 + }, + { + "epoch": 0.3790112749349523, + "grad_norm": 5.53914213180542, + "learning_rate": 1.7536631365153056e-05, + "loss": 0.1005, + "step": 3496 + }, + { + "epoch": 0.3791196877710321, + "grad_norm": 1.4175114631652832, + "learning_rate": 1.753590599158567e-05, + "loss": 0.0211, + "step": 3497 + }, + { + "epoch": 0.3792281006071119, + "grad_norm": 2.7375946044921875, + "learning_rate": 1.753518061801828e-05, + "loss": 0.0336, + "step": 3498 + }, + { + "epoch": 0.3793365134431917, + "grad_norm": 0.7990259528160095, + "learning_rate": 1.7534455244450893e-05, + "loss": 0.0142, + "step": 3499 + }, + { + "epoch": 0.37944492627927146, + "grad_norm": 6.4226908683776855, + "learning_rate": 1.753372987088351e-05, + "loss": 0.0074, + "step": 3500 + }, + { + "epoch": 0.3795533391153513, + "grad_norm": 2.2339565753936768, + "learning_rate": 1.753300449731612e-05, + "loss": 0.011, + "step": 3501 + }, + { + "epoch": 0.37966175195143104, + "grad_norm": 2.8053524494171143, + "learning_rate": 1.7532279123748733e-05, + "loss": 0.0148, + "step": 3502 + }, + { + "epoch": 0.37977016478751086, + "grad_norm": 3.27184796333313, + "learning_rate": 1.7531553750181345e-05, + "loss": 0.041, + "step": 3503 + }, + { + "epoch": 0.3798785776235906, + "grad_norm": 5.466512680053711, + "learning_rate": 1.7530828376613957e-05, + "loss": 0.0285, + "step": 3504 + }, + { + "epoch": 0.37998699045967044, + "grad_norm": 0.9187247157096863, + "learning_rate": 1.753010300304657e-05, + "loss": 0.0145, + "step": 3505 + }, + { + "epoch": 0.3800954032957502, + "grad_norm": 1.7103761434555054, + "learning_rate": 1.752937762947918e-05, + "loss": 0.0241, + "step": 3506 + }, + { + "epoch": 0.38020381613183, + "grad_norm": 8.201961517333984, + "learning_rate": 1.7528652255911797e-05, + "loss": 0.0541, + "step": 3507 + }, + { + "epoch": 0.3803122289679098, + "grad_norm": 1.416109561920166, + "learning_rate": 1.752792688234441e-05, + "loss": 0.0292, + "step": 3508 + }, + { + "epoch": 0.3804206418039896, + "grad_norm": 2.3967936038970947, + "learning_rate": 1.752720150877702e-05, + "loss": 0.0521, + "step": 3509 + }, + { + "epoch": 0.38052905464006936, + "grad_norm": 1.118674635887146, + "learning_rate": 1.7526476135209633e-05, + "loss": 0.0199, + "step": 3510 + }, + { + "epoch": 0.3806374674761492, + "grad_norm": 4.097545146942139, + "learning_rate": 1.7525750761642245e-05, + "loss": 0.0737, + "step": 3511 + }, + { + "epoch": 0.38074588031222895, + "grad_norm": 8.607985496520996, + "learning_rate": 1.752502538807486e-05, + "loss": 0.0959, + "step": 3512 + }, + { + "epoch": 0.38085429314830876, + "grad_norm": 4.204405784606934, + "learning_rate": 1.7524300014507473e-05, + "loss": 0.0647, + "step": 3513 + }, + { + "epoch": 0.3809627059843885, + "grad_norm": 2.9304122924804688, + "learning_rate": 1.7523574640940085e-05, + "loss": 0.0404, + "step": 3514 + }, + { + "epoch": 0.38107111882046835, + "grad_norm": 3.8293850421905518, + "learning_rate": 1.75228492673727e-05, + "loss": 0.0366, + "step": 3515 + }, + { + "epoch": 0.3811795316565481, + "grad_norm": 4.263487339019775, + "learning_rate": 1.7522123893805313e-05, + "loss": 0.0435, + "step": 3516 + }, + { + "epoch": 0.38128794449262793, + "grad_norm": 3.92425799369812, + "learning_rate": 1.7521398520237925e-05, + "loss": 0.0767, + "step": 3517 + }, + { + "epoch": 0.3813963573287077, + "grad_norm": 4.143032550811768, + "learning_rate": 1.7520673146670537e-05, + "loss": 0.0986, + "step": 3518 + }, + { + "epoch": 0.3815047701647875, + "grad_norm": 2.1500444412231445, + "learning_rate": 1.751994777310315e-05, + "loss": 0.0409, + "step": 3519 + }, + { + "epoch": 0.38161318300086733, + "grad_norm": 1.2962251901626587, + "learning_rate": 1.751922239953576e-05, + "loss": 0.0289, + "step": 3520 + }, + { + "epoch": 0.3817215958369471, + "grad_norm": 3.4499192237854004, + "learning_rate": 1.7518497025968374e-05, + "loss": 0.0757, + "step": 3521 + }, + { + "epoch": 0.3818300086730269, + "grad_norm": 4.280137538909912, + "learning_rate": 1.751777165240099e-05, + "loss": 0.0312, + "step": 3522 + }, + { + "epoch": 0.3819384215091067, + "grad_norm": 4.495190143585205, + "learning_rate": 1.75170462788336e-05, + "loss": 0.0395, + "step": 3523 + }, + { + "epoch": 0.3820468343451865, + "grad_norm": 1.826593279838562, + "learning_rate": 1.7516320905266213e-05, + "loss": 0.0329, + "step": 3524 + }, + { + "epoch": 0.38215524718126626, + "grad_norm": 2.593994379043579, + "learning_rate": 1.7515595531698826e-05, + "loss": 0.0146, + "step": 3525 + }, + { + "epoch": 0.3822636600173461, + "grad_norm": 3.2328343391418457, + "learning_rate": 1.7514870158131438e-05, + "loss": 0.0418, + "step": 3526 + }, + { + "epoch": 0.38237207285342584, + "grad_norm": 1.7815901041030884, + "learning_rate": 1.751414478456405e-05, + "loss": 0.0096, + "step": 3527 + }, + { + "epoch": 0.38248048568950566, + "grad_norm": 3.2945947647094727, + "learning_rate": 1.7513419410996665e-05, + "loss": 0.0542, + "step": 3528 + }, + { + "epoch": 0.3825888985255854, + "grad_norm": 3.0913963317871094, + "learning_rate": 1.7512694037429277e-05, + "loss": 0.0192, + "step": 3529 + }, + { + "epoch": 0.38269731136166524, + "grad_norm": 2.3871705532073975, + "learning_rate": 1.751196866386189e-05, + "loss": 0.0651, + "step": 3530 + }, + { + "epoch": 0.382805724197745, + "grad_norm": 1.0771929025650024, + "learning_rate": 1.7511243290294505e-05, + "loss": 0.011, + "step": 3531 + }, + { + "epoch": 0.3829141370338248, + "grad_norm": 8.659120559692383, + "learning_rate": 1.7510517916727117e-05, + "loss": 0.0649, + "step": 3532 + }, + { + "epoch": 0.3830225498699046, + "grad_norm": 1.806313395500183, + "learning_rate": 1.750979254315973e-05, + "loss": 0.0216, + "step": 3533 + }, + { + "epoch": 0.3831309627059844, + "grad_norm": 3.1832308769226074, + "learning_rate": 1.750906716959234e-05, + "loss": 0.0541, + "step": 3534 + }, + { + "epoch": 0.38323937554206416, + "grad_norm": 2.6471760272979736, + "learning_rate": 1.7508341796024954e-05, + "loss": 0.0505, + "step": 3535 + }, + { + "epoch": 0.383347788378144, + "grad_norm": 3.1821141242980957, + "learning_rate": 1.7507616422457566e-05, + "loss": 0.0281, + "step": 3536 + }, + { + "epoch": 0.38345620121422375, + "grad_norm": 4.812459468841553, + "learning_rate": 1.750689104889018e-05, + "loss": 0.0587, + "step": 3537 + }, + { + "epoch": 0.38356461405030357, + "grad_norm": 2.442958354949951, + "learning_rate": 1.7506165675322794e-05, + "loss": 0.0577, + "step": 3538 + }, + { + "epoch": 0.38367302688638333, + "grad_norm": 1.6286933422088623, + "learning_rate": 1.7505440301755406e-05, + "loss": 0.0434, + "step": 3539 + }, + { + "epoch": 0.38378143972246315, + "grad_norm": 2.221299409866333, + "learning_rate": 1.7504714928188018e-05, + "loss": 0.0389, + "step": 3540 + }, + { + "epoch": 0.3838898525585429, + "grad_norm": 1.6440967321395874, + "learning_rate": 1.750398955462063e-05, + "loss": 0.0453, + "step": 3541 + }, + { + "epoch": 0.38399826539462273, + "grad_norm": 2.3033957481384277, + "learning_rate": 1.7503264181053242e-05, + "loss": 0.0371, + "step": 3542 + }, + { + "epoch": 0.3841066782307025, + "grad_norm": 5.438976764678955, + "learning_rate": 1.7502538807485858e-05, + "loss": 0.0345, + "step": 3543 + }, + { + "epoch": 0.3842150910667823, + "grad_norm": 4.934466361999512, + "learning_rate": 1.750181343391847e-05, + "loss": 0.0486, + "step": 3544 + }, + { + "epoch": 0.3843235039028621, + "grad_norm": 4.2398362159729, + "learning_rate": 1.7501088060351082e-05, + "loss": 0.0286, + "step": 3545 + }, + { + "epoch": 0.3844319167389419, + "grad_norm": 3.567718744277954, + "learning_rate": 1.7500362686783694e-05, + "loss": 0.0455, + "step": 3546 + }, + { + "epoch": 0.38454032957502166, + "grad_norm": 3.3244423866271973, + "learning_rate": 1.749963731321631e-05, + "loss": 0.108, + "step": 3547 + }, + { + "epoch": 0.3846487424111015, + "grad_norm": 4.559867858886719, + "learning_rate": 1.7498911939648922e-05, + "loss": 0.0496, + "step": 3548 + }, + { + "epoch": 0.3847571552471813, + "grad_norm": 6.383998394012451, + "learning_rate": 1.7498186566081534e-05, + "loss": 0.0445, + "step": 3549 + }, + { + "epoch": 0.38486556808326106, + "grad_norm": 3.9754512310028076, + "learning_rate": 1.7497461192514146e-05, + "loss": 0.1167, + "step": 3550 + }, + { + "epoch": 0.3849739809193409, + "grad_norm": 1.9006061553955078, + "learning_rate": 1.7496735818946758e-05, + "loss": 0.0189, + "step": 3551 + }, + { + "epoch": 0.38508239375542064, + "grad_norm": 1.1571890115737915, + "learning_rate": 1.7496010445379374e-05, + "loss": 0.0158, + "step": 3552 + }, + { + "epoch": 0.38519080659150046, + "grad_norm": 3.1256446838378906, + "learning_rate": 1.7495285071811986e-05, + "loss": 0.0708, + "step": 3553 + }, + { + "epoch": 0.3852992194275802, + "grad_norm": 6.446224212646484, + "learning_rate": 1.7494559698244598e-05, + "loss": 0.0505, + "step": 3554 + }, + { + "epoch": 0.38540763226366004, + "grad_norm": 2.956758499145508, + "learning_rate": 1.749383432467721e-05, + "loss": 0.0345, + "step": 3555 + }, + { + "epoch": 0.3855160450997398, + "grad_norm": 5.896022796630859, + "learning_rate": 1.7493108951109822e-05, + "loss": 0.1, + "step": 3556 + }, + { + "epoch": 0.3856244579358196, + "grad_norm": 1.3473165035247803, + "learning_rate": 1.7492383577542434e-05, + "loss": 0.0427, + "step": 3557 + }, + { + "epoch": 0.3857328707718994, + "grad_norm": 2.556706666946411, + "learning_rate": 1.7491658203975047e-05, + "loss": 0.0735, + "step": 3558 + }, + { + "epoch": 0.3858412836079792, + "grad_norm": 3.202613353729248, + "learning_rate": 1.7490932830407662e-05, + "loss": 0.0544, + "step": 3559 + }, + { + "epoch": 0.38594969644405897, + "grad_norm": 1.2069042921066284, + "learning_rate": 1.7490207456840274e-05, + "loss": 0.0248, + "step": 3560 + }, + { + "epoch": 0.3860581092801388, + "grad_norm": 3.120499610900879, + "learning_rate": 1.7489482083272886e-05, + "loss": 0.046, + "step": 3561 + }, + { + "epoch": 0.38616652211621855, + "grad_norm": 2.7642204761505127, + "learning_rate": 1.74887567097055e-05, + "loss": 0.0359, + "step": 3562 + }, + { + "epoch": 0.38627493495229837, + "grad_norm": 3.8912525177001953, + "learning_rate": 1.748803133613811e-05, + "loss": 0.047, + "step": 3563 + }, + { + "epoch": 0.38638334778837813, + "grad_norm": 3.1172266006469727, + "learning_rate": 1.7487305962570726e-05, + "loss": 0.0295, + "step": 3564 + }, + { + "epoch": 0.38649176062445795, + "grad_norm": 1.9328868389129639, + "learning_rate": 1.748658058900334e-05, + "loss": 0.0563, + "step": 3565 + }, + { + "epoch": 0.3866001734605377, + "grad_norm": 3.1569995880126953, + "learning_rate": 1.748585521543595e-05, + "loss": 0.0572, + "step": 3566 + }, + { + "epoch": 0.38670858629661753, + "grad_norm": 1.4148321151733398, + "learning_rate": 1.7485129841868566e-05, + "loss": 0.0289, + "step": 3567 + }, + { + "epoch": 0.3868169991326973, + "grad_norm": 4.667187213897705, + "learning_rate": 1.7484404468301178e-05, + "loss": 0.0541, + "step": 3568 + }, + { + "epoch": 0.3869254119687771, + "grad_norm": 3.374298572540283, + "learning_rate": 1.748367909473379e-05, + "loss": 0.0821, + "step": 3569 + }, + { + "epoch": 0.3870338248048569, + "grad_norm": 2.1538517475128174, + "learning_rate": 1.7482953721166402e-05, + "loss": 0.0262, + "step": 3570 + }, + { + "epoch": 0.3871422376409367, + "grad_norm": 2.6918251514434814, + "learning_rate": 1.7482228347599015e-05, + "loss": 0.0533, + "step": 3571 + }, + { + "epoch": 0.38725065047701646, + "grad_norm": 2.748007297515869, + "learning_rate": 1.7481502974031627e-05, + "loss": 0.0705, + "step": 3572 + }, + { + "epoch": 0.3873590633130963, + "grad_norm": 5.3205084800720215, + "learning_rate": 1.748077760046424e-05, + "loss": 0.0506, + "step": 3573 + }, + { + "epoch": 0.38746747614917604, + "grad_norm": 6.639270305633545, + "learning_rate": 1.7480052226896854e-05, + "loss": 0.0612, + "step": 3574 + }, + { + "epoch": 0.38757588898525586, + "grad_norm": 2.9940357208251953, + "learning_rate": 1.7479326853329467e-05, + "loss": 0.0374, + "step": 3575 + }, + { + "epoch": 0.3876843018213356, + "grad_norm": 2.361564874649048, + "learning_rate": 1.747860147976208e-05, + "loss": 0.0599, + "step": 3576 + }, + { + "epoch": 0.38779271465741544, + "grad_norm": 2.370938301086426, + "learning_rate": 1.747787610619469e-05, + "loss": 0.0329, + "step": 3577 + }, + { + "epoch": 0.38790112749349526, + "grad_norm": 2.1685943603515625, + "learning_rate": 1.7477150732627303e-05, + "loss": 0.0146, + "step": 3578 + }, + { + "epoch": 0.388009540329575, + "grad_norm": 2.956777811050415, + "learning_rate": 1.7476425359059915e-05, + "loss": 0.0472, + "step": 3579 + }, + { + "epoch": 0.38811795316565484, + "grad_norm": 0.6061866879463196, + "learning_rate": 1.747569998549253e-05, + "loss": 0.0109, + "step": 3580 + }, + { + "epoch": 0.3882263660017346, + "grad_norm": 3.397963523864746, + "learning_rate": 1.7474974611925143e-05, + "loss": 0.0357, + "step": 3581 + }, + { + "epoch": 0.3883347788378144, + "grad_norm": 3.0566020011901855, + "learning_rate": 1.7474249238357755e-05, + "loss": 0.0436, + "step": 3582 + }, + { + "epoch": 0.3884431916738942, + "grad_norm": 3.915740489959717, + "learning_rate": 1.747352386479037e-05, + "loss": 0.0627, + "step": 3583 + }, + { + "epoch": 0.388551604509974, + "grad_norm": 7.8712158203125, + "learning_rate": 1.7472798491222983e-05, + "loss": 0.1043, + "step": 3584 + }, + { + "epoch": 0.38866001734605377, + "grad_norm": 2.368508815765381, + "learning_rate": 1.7472073117655595e-05, + "loss": 0.0296, + "step": 3585 + }, + { + "epoch": 0.3887684301821336, + "grad_norm": 6.49967622756958, + "learning_rate": 1.7471347744088207e-05, + "loss": 0.0942, + "step": 3586 + }, + { + "epoch": 0.38887684301821335, + "grad_norm": 7.540646076202393, + "learning_rate": 1.747062237052082e-05, + "loss": 0.1051, + "step": 3587 + }, + { + "epoch": 0.38898525585429317, + "grad_norm": 4.0393290519714355, + "learning_rate": 1.746989699695343e-05, + "loss": 0.0425, + "step": 3588 + }, + { + "epoch": 0.38909366869037293, + "grad_norm": 3.6242570877075195, + "learning_rate": 1.7469171623386047e-05, + "loss": 0.0214, + "step": 3589 + }, + { + "epoch": 0.38920208152645275, + "grad_norm": 5.863751411437988, + "learning_rate": 1.746844624981866e-05, + "loss": 0.0611, + "step": 3590 + }, + { + "epoch": 0.3893104943625325, + "grad_norm": 2.074422836303711, + "learning_rate": 1.746772087625127e-05, + "loss": 0.0281, + "step": 3591 + }, + { + "epoch": 0.38941890719861233, + "grad_norm": 2.1599435806274414, + "learning_rate": 1.7466995502683883e-05, + "loss": 0.0167, + "step": 3592 + }, + { + "epoch": 0.3895273200346921, + "grad_norm": 3.4632351398468018, + "learning_rate": 1.7466270129116495e-05, + "loss": 0.0782, + "step": 3593 + }, + { + "epoch": 0.3896357328707719, + "grad_norm": 3.010516405105591, + "learning_rate": 1.7465544755549107e-05, + "loss": 0.0273, + "step": 3594 + }, + { + "epoch": 0.3897441457068517, + "grad_norm": 3.386373996734619, + "learning_rate": 1.7464819381981723e-05, + "loss": 0.0481, + "step": 3595 + }, + { + "epoch": 0.3898525585429315, + "grad_norm": 3.390551805496216, + "learning_rate": 1.7464094008414335e-05, + "loss": 0.0311, + "step": 3596 + }, + { + "epoch": 0.38996097137901126, + "grad_norm": 3.2444489002227783, + "learning_rate": 1.7463368634846947e-05, + "loss": 0.0653, + "step": 3597 + }, + { + "epoch": 0.3900693842150911, + "grad_norm": 2.3810272216796875, + "learning_rate": 1.746264326127956e-05, + "loss": 0.0462, + "step": 3598 + }, + { + "epoch": 0.39017779705117084, + "grad_norm": 2.208030939102173, + "learning_rate": 1.7461917887712175e-05, + "loss": 0.0232, + "step": 3599 + }, + { + "epoch": 0.39028620988725066, + "grad_norm": 2.419773817062378, + "learning_rate": 1.7461192514144787e-05, + "loss": 0.0311, + "step": 3600 + }, + { + "epoch": 0.3903946227233304, + "grad_norm": 3.1322340965270996, + "learning_rate": 1.74604671405774e-05, + "loss": 0.0632, + "step": 3601 + }, + { + "epoch": 0.39050303555941024, + "grad_norm": 5.2309346199035645, + "learning_rate": 1.745974176701001e-05, + "loss": 0.0977, + "step": 3602 + }, + { + "epoch": 0.39061144839549, + "grad_norm": 3.2155652046203613, + "learning_rate": 1.7459016393442624e-05, + "loss": 0.0734, + "step": 3603 + }, + { + "epoch": 0.3907198612315698, + "grad_norm": 2.2261853218078613, + "learning_rate": 1.745829101987524e-05, + "loss": 0.0694, + "step": 3604 + }, + { + "epoch": 0.3908282740676496, + "grad_norm": 1.5750744342803955, + "learning_rate": 1.745756564630785e-05, + "loss": 0.02, + "step": 3605 + }, + { + "epoch": 0.3909366869037294, + "grad_norm": 0.9198883175849915, + "learning_rate": 1.7456840272740463e-05, + "loss": 0.0258, + "step": 3606 + }, + { + "epoch": 0.3910450997398092, + "grad_norm": 3.291961908340454, + "learning_rate": 1.7456114899173075e-05, + "loss": 0.0634, + "step": 3607 + }, + { + "epoch": 0.391153512575889, + "grad_norm": 3.9522500038146973, + "learning_rate": 1.7455389525605688e-05, + "loss": 0.0621, + "step": 3608 + }, + { + "epoch": 0.3912619254119688, + "grad_norm": 1.2695132493972778, + "learning_rate": 1.74546641520383e-05, + "loss": 0.0365, + "step": 3609 + }, + { + "epoch": 0.39137033824804857, + "grad_norm": 4.005901336669922, + "learning_rate": 1.7453938778470912e-05, + "loss": 0.0666, + "step": 3610 + }, + { + "epoch": 0.3914787510841284, + "grad_norm": 1.9340736865997314, + "learning_rate": 1.7453213404903527e-05, + "loss": 0.0293, + "step": 3611 + }, + { + "epoch": 0.39158716392020815, + "grad_norm": 4.19989538192749, + "learning_rate": 1.745248803133614e-05, + "loss": 0.0997, + "step": 3612 + }, + { + "epoch": 0.39169557675628797, + "grad_norm": 1.6811412572860718, + "learning_rate": 1.7451762657768752e-05, + "loss": 0.0165, + "step": 3613 + }, + { + "epoch": 0.39180398959236773, + "grad_norm": 2.197342872619629, + "learning_rate": 1.7451037284201364e-05, + "loss": 0.0278, + "step": 3614 + }, + { + "epoch": 0.39191240242844755, + "grad_norm": 1.2522550821304321, + "learning_rate": 1.7450311910633976e-05, + "loss": 0.014, + "step": 3615 + }, + { + "epoch": 0.3920208152645273, + "grad_norm": 2.294503927230835, + "learning_rate": 1.744958653706659e-05, + "loss": 0.0311, + "step": 3616 + }, + { + "epoch": 0.39212922810060713, + "grad_norm": 1.9050626754760742, + "learning_rate": 1.7448861163499204e-05, + "loss": 0.0566, + "step": 3617 + }, + { + "epoch": 0.3922376409366869, + "grad_norm": 0.8788483142852783, + "learning_rate": 1.7448135789931816e-05, + "loss": 0.0305, + "step": 3618 + }, + { + "epoch": 0.3923460537727667, + "grad_norm": 3.410550832748413, + "learning_rate": 1.744741041636443e-05, + "loss": 0.066, + "step": 3619 + }, + { + "epoch": 0.3924544666088465, + "grad_norm": 1.2300390005111694, + "learning_rate": 1.7446685042797044e-05, + "loss": 0.0196, + "step": 3620 + }, + { + "epoch": 0.3925628794449263, + "grad_norm": 3.9056179523468018, + "learning_rate": 1.7445959669229656e-05, + "loss": 0.0328, + "step": 3621 + }, + { + "epoch": 0.39267129228100606, + "grad_norm": 5.303720474243164, + "learning_rate": 1.7445234295662268e-05, + "loss": 0.0392, + "step": 3622 + }, + { + "epoch": 0.3927797051170859, + "grad_norm": 4.130235195159912, + "learning_rate": 1.744450892209488e-05, + "loss": 0.0347, + "step": 3623 + }, + { + "epoch": 0.39288811795316564, + "grad_norm": 2.7765941619873047, + "learning_rate": 1.7443783548527492e-05, + "loss": 0.0227, + "step": 3624 + }, + { + "epoch": 0.39299653078924546, + "grad_norm": 1.8921394348144531, + "learning_rate": 1.7443058174960104e-05, + "loss": 0.0228, + "step": 3625 + }, + { + "epoch": 0.3931049436253252, + "grad_norm": 3.783036947250366, + "learning_rate": 1.744233280139272e-05, + "loss": 0.0521, + "step": 3626 + }, + { + "epoch": 0.39321335646140504, + "grad_norm": 2.3446898460388184, + "learning_rate": 1.7441607427825332e-05, + "loss": 0.0393, + "step": 3627 + }, + { + "epoch": 0.3933217692974848, + "grad_norm": 7.131758213043213, + "learning_rate": 1.7440882054257944e-05, + "loss": 0.0567, + "step": 3628 + }, + { + "epoch": 0.3934301821335646, + "grad_norm": 2.0515456199645996, + "learning_rate": 1.7440156680690556e-05, + "loss": 0.0274, + "step": 3629 + }, + { + "epoch": 0.3935385949696444, + "grad_norm": 3.519620656967163, + "learning_rate": 1.743943130712317e-05, + "loss": 0.0232, + "step": 3630 + }, + { + "epoch": 0.3936470078057242, + "grad_norm": 1.9861074686050415, + "learning_rate": 1.743870593355578e-05, + "loss": 0.0477, + "step": 3631 + }, + { + "epoch": 0.39375542064180397, + "grad_norm": 4.425195693969727, + "learning_rate": 1.7437980559988396e-05, + "loss": 0.0256, + "step": 3632 + }, + { + "epoch": 0.3938638334778838, + "grad_norm": 4.957149982452393, + "learning_rate": 1.7437255186421008e-05, + "loss": 0.1201, + "step": 3633 + }, + { + "epoch": 0.39397224631396355, + "grad_norm": 4.832335472106934, + "learning_rate": 1.7436529812853624e-05, + "loss": 0.061, + "step": 3634 + }, + { + "epoch": 0.39408065915004337, + "grad_norm": 3.363233804702759, + "learning_rate": 1.7435804439286236e-05, + "loss": 0.0399, + "step": 3635 + }, + { + "epoch": 0.39418907198612313, + "grad_norm": 4.3913373947143555, + "learning_rate": 1.7435079065718848e-05, + "loss": 0.0444, + "step": 3636 + }, + { + "epoch": 0.39429748482220295, + "grad_norm": 4.1709818840026855, + "learning_rate": 1.743435369215146e-05, + "loss": 0.0137, + "step": 3637 + }, + { + "epoch": 0.39440589765828277, + "grad_norm": 0.9823550581932068, + "learning_rate": 1.7433628318584072e-05, + "loss": 0.0113, + "step": 3638 + }, + { + "epoch": 0.39451431049436253, + "grad_norm": 1.7872480154037476, + "learning_rate": 1.7432902945016684e-05, + "loss": 0.0649, + "step": 3639 + }, + { + "epoch": 0.39462272333044235, + "grad_norm": 1.2168655395507812, + "learning_rate": 1.7432177571449297e-05, + "loss": 0.0078, + "step": 3640 + }, + { + "epoch": 0.3947311361665221, + "grad_norm": 1.723241925239563, + "learning_rate": 1.7431452197881912e-05, + "loss": 0.0311, + "step": 3641 + }, + { + "epoch": 0.39483954900260193, + "grad_norm": 4.717487335205078, + "learning_rate": 1.7430726824314524e-05, + "loss": 0.0345, + "step": 3642 + }, + { + "epoch": 0.3949479618386817, + "grad_norm": 3.711412191390991, + "learning_rate": 1.7430001450747136e-05, + "loss": 0.0505, + "step": 3643 + }, + { + "epoch": 0.3950563746747615, + "grad_norm": 2.416734218597412, + "learning_rate": 1.742927607717975e-05, + "loss": 0.0481, + "step": 3644 + }, + { + "epoch": 0.3951647875108413, + "grad_norm": 5.216752052307129, + "learning_rate": 1.742855070361236e-05, + "loss": 0.0664, + "step": 3645 + }, + { + "epoch": 0.3952732003469211, + "grad_norm": 1.3760920763015747, + "learning_rate": 1.7427825330044973e-05, + "loss": 0.046, + "step": 3646 + }, + { + "epoch": 0.39538161318300086, + "grad_norm": 13.181066513061523, + "learning_rate": 1.742709995647759e-05, + "loss": 0.0272, + "step": 3647 + }, + { + "epoch": 0.3954900260190807, + "grad_norm": 4.153227806091309, + "learning_rate": 1.74263745829102e-05, + "loss": 0.0457, + "step": 3648 + }, + { + "epoch": 0.39559843885516044, + "grad_norm": 1.7556642293930054, + "learning_rate": 1.7425649209342813e-05, + "loss": 0.0395, + "step": 3649 + }, + { + "epoch": 0.39570685169124026, + "grad_norm": 1.7085593938827515, + "learning_rate": 1.7424923835775425e-05, + "loss": 0.0297, + "step": 3650 + }, + { + "epoch": 0.39581526452732, + "grad_norm": 2.0188775062561035, + "learning_rate": 1.742419846220804e-05, + "loss": 0.0659, + "step": 3651 + }, + { + "epoch": 0.39592367736339984, + "grad_norm": 2.8999595642089844, + "learning_rate": 1.7423473088640652e-05, + "loss": 0.0664, + "step": 3652 + }, + { + "epoch": 0.3960320901994796, + "grad_norm": 5.199618816375732, + "learning_rate": 1.7422747715073265e-05, + "loss": 0.0719, + "step": 3653 + }, + { + "epoch": 0.3961405030355594, + "grad_norm": 3.586367607116699, + "learning_rate": 1.7422022341505877e-05, + "loss": 0.0468, + "step": 3654 + }, + { + "epoch": 0.3962489158716392, + "grad_norm": 1.4310780763626099, + "learning_rate": 1.742129696793849e-05, + "loss": 0.0215, + "step": 3655 + }, + { + "epoch": 0.396357328707719, + "grad_norm": 1.9108572006225586, + "learning_rate": 1.7420571594371104e-05, + "loss": 0.034, + "step": 3656 + }, + { + "epoch": 0.39646574154379877, + "grad_norm": 3.8685882091522217, + "learning_rate": 1.7419846220803717e-05, + "loss": 0.0531, + "step": 3657 + }, + { + "epoch": 0.3965741543798786, + "grad_norm": 2.5375633239746094, + "learning_rate": 1.741912084723633e-05, + "loss": 0.054, + "step": 3658 + }, + { + "epoch": 0.39668256721595835, + "grad_norm": 1.9328713417053223, + "learning_rate": 1.741839547366894e-05, + "loss": 0.0394, + "step": 3659 + }, + { + "epoch": 0.39679098005203817, + "grad_norm": 3.062664747238159, + "learning_rate": 1.7417670100101553e-05, + "loss": 0.057, + "step": 3660 + }, + { + "epoch": 0.39689939288811793, + "grad_norm": 3.9510629177093506, + "learning_rate": 1.7416944726534165e-05, + "loss": 0.079, + "step": 3661 + }, + { + "epoch": 0.39700780572419775, + "grad_norm": 2.358989715576172, + "learning_rate": 1.7416219352966777e-05, + "loss": 0.026, + "step": 3662 + }, + { + "epoch": 0.3971162185602775, + "grad_norm": 0.9146252274513245, + "learning_rate": 1.7415493979399393e-05, + "loss": 0.0145, + "step": 3663 + }, + { + "epoch": 0.39722463139635733, + "grad_norm": 1.2244298458099365, + "learning_rate": 1.7414768605832005e-05, + "loss": 0.0223, + "step": 3664 + }, + { + "epoch": 0.3973330442324371, + "grad_norm": 1.3425930738449097, + "learning_rate": 1.7414043232264617e-05, + "loss": 0.0247, + "step": 3665 + }, + { + "epoch": 0.3974414570685169, + "grad_norm": 3.257827043533325, + "learning_rate": 1.741331785869723e-05, + "loss": 0.0993, + "step": 3666 + }, + { + "epoch": 0.39754986990459673, + "grad_norm": 5.758270263671875, + "learning_rate": 1.741259248512984e-05, + "loss": 0.0522, + "step": 3667 + }, + { + "epoch": 0.3976582827406765, + "grad_norm": 0.9469918608665466, + "learning_rate": 1.7411867111562457e-05, + "loss": 0.0148, + "step": 3668 + }, + { + "epoch": 0.3977666955767563, + "grad_norm": 3.315096616744995, + "learning_rate": 1.741114173799507e-05, + "loss": 0.0438, + "step": 3669 + }, + { + "epoch": 0.3978751084128361, + "grad_norm": 3.1311726570129395, + "learning_rate": 1.741041636442768e-05, + "loss": 0.0413, + "step": 3670 + }, + { + "epoch": 0.3979835212489159, + "grad_norm": 1.748441457748413, + "learning_rate": 1.7409690990860297e-05, + "loss": 0.0246, + "step": 3671 + }, + { + "epoch": 0.39809193408499566, + "grad_norm": 4.61676549911499, + "learning_rate": 1.740896561729291e-05, + "loss": 0.0435, + "step": 3672 + }, + { + "epoch": 0.3982003469210755, + "grad_norm": 3.3451802730560303, + "learning_rate": 1.740824024372552e-05, + "loss": 0.0335, + "step": 3673 + }, + { + "epoch": 0.39830875975715524, + "grad_norm": 4.1760149002075195, + "learning_rate": 1.7407514870158133e-05, + "loss": 0.0597, + "step": 3674 + }, + { + "epoch": 0.39841717259323506, + "grad_norm": 3.3510944843292236, + "learning_rate": 1.7406789496590745e-05, + "loss": 0.0738, + "step": 3675 + }, + { + "epoch": 0.3985255854293148, + "grad_norm": 4.638632297515869, + "learning_rate": 1.7406064123023357e-05, + "loss": 0.068, + "step": 3676 + }, + { + "epoch": 0.39863399826539464, + "grad_norm": 1.3891814947128296, + "learning_rate": 1.740533874945597e-05, + "loss": 0.0227, + "step": 3677 + }, + { + "epoch": 0.3987424111014744, + "grad_norm": 4.404209613800049, + "learning_rate": 1.7404613375888585e-05, + "loss": 0.0675, + "step": 3678 + }, + { + "epoch": 0.3988508239375542, + "grad_norm": 0.8991419672966003, + "learning_rate": 1.7403888002321197e-05, + "loss": 0.0066, + "step": 3679 + }, + { + "epoch": 0.398959236773634, + "grad_norm": 2.082859754562378, + "learning_rate": 1.740316262875381e-05, + "loss": 0.023, + "step": 3680 + }, + { + "epoch": 0.3990676496097138, + "grad_norm": 1.2896913290023804, + "learning_rate": 1.740243725518642e-05, + "loss": 0.0111, + "step": 3681 + }, + { + "epoch": 0.39917606244579357, + "grad_norm": 1.3013501167297363, + "learning_rate": 1.7401711881619034e-05, + "loss": 0.0108, + "step": 3682 + }, + { + "epoch": 0.3992844752818734, + "grad_norm": 3.013495683670044, + "learning_rate": 1.7400986508051646e-05, + "loss": 0.0554, + "step": 3683 + }, + { + "epoch": 0.39939288811795315, + "grad_norm": 1.388706088066101, + "learning_rate": 1.740026113448426e-05, + "loss": 0.0301, + "step": 3684 + }, + { + "epoch": 0.39950130095403297, + "grad_norm": 2.3985629081726074, + "learning_rate": 1.7399535760916874e-05, + "loss": 0.0186, + "step": 3685 + }, + { + "epoch": 0.39960971379011273, + "grad_norm": 4.183211326599121, + "learning_rate": 1.739881038734949e-05, + "loss": 0.0628, + "step": 3686 + }, + { + "epoch": 0.39971812662619255, + "grad_norm": 14.389114379882812, + "learning_rate": 1.73980850137821e-05, + "loss": 0.0725, + "step": 3687 + }, + { + "epoch": 0.3998265394622723, + "grad_norm": 7.399021148681641, + "learning_rate": 1.7397359640214713e-05, + "loss": 0.0165, + "step": 3688 + }, + { + "epoch": 0.39993495229835213, + "grad_norm": 3.753567934036255, + "learning_rate": 1.7396634266647325e-05, + "loss": 0.057, + "step": 3689 + }, + { + "epoch": 0.4000433651344319, + "grad_norm": 2.195146083831787, + "learning_rate": 1.7395908893079938e-05, + "loss": 0.0165, + "step": 3690 + }, + { + "epoch": 0.4001517779705117, + "grad_norm": 1.754022479057312, + "learning_rate": 1.739518351951255e-05, + "loss": 0.0502, + "step": 3691 + }, + { + "epoch": 0.4002601908065915, + "grad_norm": 0.7958113551139832, + "learning_rate": 1.7394458145945162e-05, + "loss": 0.0118, + "step": 3692 + }, + { + "epoch": 0.4003686036426713, + "grad_norm": 0.8368503451347351, + "learning_rate": 1.7393732772377777e-05, + "loss": 0.0079, + "step": 3693 + }, + { + "epoch": 0.40047701647875106, + "grad_norm": 4.197626113891602, + "learning_rate": 1.739300739881039e-05, + "loss": 0.0269, + "step": 3694 + }, + { + "epoch": 0.4005854293148309, + "grad_norm": 9.281250953674316, + "learning_rate": 1.7392282025243002e-05, + "loss": 0.0261, + "step": 3695 + }, + { + "epoch": 0.4006938421509107, + "grad_norm": 4.678873538970947, + "learning_rate": 1.7391556651675614e-05, + "loss": 0.0361, + "step": 3696 + }, + { + "epoch": 0.40080225498699046, + "grad_norm": 4.2663702964782715, + "learning_rate": 1.7390831278108226e-05, + "loss": 0.0525, + "step": 3697 + }, + { + "epoch": 0.4009106678230703, + "grad_norm": 4.092833042144775, + "learning_rate": 1.7390105904540838e-05, + "loss": 0.0776, + "step": 3698 + }, + { + "epoch": 0.40101908065915004, + "grad_norm": 0.12404310703277588, + "learning_rate": 1.7389380530973454e-05, + "loss": 0.0008, + "step": 3699 + }, + { + "epoch": 0.40112749349522986, + "grad_norm": 0.325162798166275, + "learning_rate": 1.7388655157406066e-05, + "loss": 0.0029, + "step": 3700 + }, + { + "epoch": 0.4012359063313096, + "grad_norm": 0.17947107553482056, + "learning_rate": 1.7387929783838678e-05, + "loss": 0.002, + "step": 3701 + }, + { + "epoch": 0.40134431916738944, + "grad_norm": 20.13673210144043, + "learning_rate": 1.7387204410271293e-05, + "loss": 0.076, + "step": 3702 + }, + { + "epoch": 0.4014527320034692, + "grad_norm": 3.1611392498016357, + "learning_rate": 1.7386479036703906e-05, + "loss": 0.0338, + "step": 3703 + }, + { + "epoch": 0.401561144839549, + "grad_norm": 22.52634620666504, + "learning_rate": 1.7385753663136518e-05, + "loss": 0.0747, + "step": 3704 + }, + { + "epoch": 0.4016695576756288, + "grad_norm": 7.520878314971924, + "learning_rate": 1.738502828956913e-05, + "loss": 0.0864, + "step": 3705 + }, + { + "epoch": 0.4017779705117086, + "grad_norm": 2.058568239212036, + "learning_rate": 1.7384302916001742e-05, + "loss": 0.0309, + "step": 3706 + }, + { + "epoch": 0.40188638334778837, + "grad_norm": 5.60919189453125, + "learning_rate": 1.7383577542434354e-05, + "loss": 0.0263, + "step": 3707 + }, + { + "epoch": 0.4019947961838682, + "grad_norm": 3.038052797317505, + "learning_rate": 1.738285216886697e-05, + "loss": 0.0916, + "step": 3708 + }, + { + "epoch": 0.40210320901994795, + "grad_norm": 5.030117034912109, + "learning_rate": 1.7382126795299582e-05, + "loss": 0.0482, + "step": 3709 + }, + { + "epoch": 0.40221162185602777, + "grad_norm": 0.6418067812919617, + "learning_rate": 1.7381401421732194e-05, + "loss": 0.0085, + "step": 3710 + }, + { + "epoch": 0.40232003469210753, + "grad_norm": 4.938529014587402, + "learning_rate": 1.7380676048164806e-05, + "loss": 0.0852, + "step": 3711 + }, + { + "epoch": 0.40242844752818735, + "grad_norm": 8.223357200622559, + "learning_rate": 1.737995067459742e-05, + "loss": 0.0362, + "step": 3712 + }, + { + "epoch": 0.4025368603642671, + "grad_norm": 10.09626579284668, + "learning_rate": 1.737922530103003e-05, + "loss": 0.0908, + "step": 3713 + }, + { + "epoch": 0.40264527320034693, + "grad_norm": 5.351646423339844, + "learning_rate": 1.7378499927462643e-05, + "loss": 0.0508, + "step": 3714 + }, + { + "epoch": 0.4027536860364267, + "grad_norm": 2.094304084777832, + "learning_rate": 1.7377774553895258e-05, + "loss": 0.0383, + "step": 3715 + }, + { + "epoch": 0.4028620988725065, + "grad_norm": 3.7211830615997314, + "learning_rate": 1.737704918032787e-05, + "loss": 0.0456, + "step": 3716 + }, + { + "epoch": 0.4029705117085863, + "grad_norm": 10.647688865661621, + "learning_rate": 1.7376323806760482e-05, + "loss": 0.102, + "step": 3717 + }, + { + "epoch": 0.4030789245446661, + "grad_norm": 2.900099992752075, + "learning_rate": 1.7375598433193095e-05, + "loss": 0.0506, + "step": 3718 + }, + { + "epoch": 0.40318733738074586, + "grad_norm": 4.381338596343994, + "learning_rate": 1.7374873059625707e-05, + "loss": 0.0541, + "step": 3719 + }, + { + "epoch": 0.4032957502168257, + "grad_norm": 4.953042507171631, + "learning_rate": 1.7374147686058322e-05, + "loss": 0.0617, + "step": 3720 + }, + { + "epoch": 0.40340416305290544, + "grad_norm": 3.4676706790924072, + "learning_rate": 1.7373422312490934e-05, + "loss": 0.0307, + "step": 3721 + }, + { + "epoch": 0.40351257588898526, + "grad_norm": 4.852441310882568, + "learning_rate": 1.7372696938923547e-05, + "loss": 0.1622, + "step": 3722 + }, + { + "epoch": 0.403620988725065, + "grad_norm": 2.209512948989868, + "learning_rate": 1.7371971565356162e-05, + "loss": 0.0519, + "step": 3723 + }, + { + "epoch": 0.40372940156114484, + "grad_norm": 2.024266004562378, + "learning_rate": 1.7371246191788774e-05, + "loss": 0.0543, + "step": 3724 + }, + { + "epoch": 0.4038378143972246, + "grad_norm": 4.156872272491455, + "learning_rate": 1.7370520818221386e-05, + "loss": 0.0779, + "step": 3725 + }, + { + "epoch": 0.4039462272333044, + "grad_norm": 2.680561065673828, + "learning_rate": 1.7369795444654e-05, + "loss": 0.0256, + "step": 3726 + }, + { + "epoch": 0.40405464006938424, + "grad_norm": 2.701892614364624, + "learning_rate": 1.736907007108661e-05, + "loss": 0.0547, + "step": 3727 + }, + { + "epoch": 0.404163052905464, + "grad_norm": 4.020703315734863, + "learning_rate": 1.7368344697519223e-05, + "loss": 0.057, + "step": 3728 + }, + { + "epoch": 0.4042714657415438, + "grad_norm": 1.7381620407104492, + "learning_rate": 1.7367619323951835e-05, + "loss": 0.0214, + "step": 3729 + }, + { + "epoch": 0.4043798785776236, + "grad_norm": 3.538790702819824, + "learning_rate": 1.736689395038445e-05, + "loss": 0.0488, + "step": 3730 + }, + { + "epoch": 0.4044882914137034, + "grad_norm": 2.9368362426757812, + "learning_rate": 1.7366168576817063e-05, + "loss": 0.0733, + "step": 3731 + }, + { + "epoch": 0.40459670424978317, + "grad_norm": 6.420685768127441, + "learning_rate": 1.7365443203249675e-05, + "loss": 0.053, + "step": 3732 + }, + { + "epoch": 0.404705117085863, + "grad_norm": 1.138363242149353, + "learning_rate": 1.7364717829682287e-05, + "loss": 0.0171, + "step": 3733 + }, + { + "epoch": 0.40481352992194275, + "grad_norm": 3.1822471618652344, + "learning_rate": 1.73639924561149e-05, + "loss": 0.0917, + "step": 3734 + }, + { + "epoch": 0.40492194275802257, + "grad_norm": 6.329826831817627, + "learning_rate": 1.736326708254751e-05, + "loss": 0.035, + "step": 3735 + }, + { + "epoch": 0.40503035559410233, + "grad_norm": 1.155211329460144, + "learning_rate": 1.7362541708980127e-05, + "loss": 0.0287, + "step": 3736 + }, + { + "epoch": 0.40513876843018215, + "grad_norm": 5.171082973480225, + "learning_rate": 1.736181633541274e-05, + "loss": 0.0528, + "step": 3737 + }, + { + "epoch": 0.4052471812662619, + "grad_norm": 1.306331992149353, + "learning_rate": 1.7361090961845354e-05, + "loss": 0.0178, + "step": 3738 + }, + { + "epoch": 0.40535559410234173, + "grad_norm": 1.791973352432251, + "learning_rate": 1.7360365588277967e-05, + "loss": 0.0311, + "step": 3739 + }, + { + "epoch": 0.4054640069384215, + "grad_norm": 1.4977869987487793, + "learning_rate": 1.735964021471058e-05, + "loss": 0.0189, + "step": 3740 + }, + { + "epoch": 0.4055724197745013, + "grad_norm": 2.3625447750091553, + "learning_rate": 1.735891484114319e-05, + "loss": 0.0361, + "step": 3741 + }, + { + "epoch": 0.4056808326105811, + "grad_norm": 1.982101321220398, + "learning_rate": 1.7358189467575803e-05, + "loss": 0.0411, + "step": 3742 + }, + { + "epoch": 0.4057892454466609, + "grad_norm": 2.0252325534820557, + "learning_rate": 1.7357464094008415e-05, + "loss": 0.0424, + "step": 3743 + }, + { + "epoch": 0.40589765828274066, + "grad_norm": 3.1369717121124268, + "learning_rate": 1.7356738720441027e-05, + "loss": 0.0703, + "step": 3744 + }, + { + "epoch": 0.4060060711188205, + "grad_norm": 3.909411668777466, + "learning_rate": 1.7356013346873643e-05, + "loss": 0.0286, + "step": 3745 + }, + { + "epoch": 0.40611448395490024, + "grad_norm": 3.0403170585632324, + "learning_rate": 1.7355287973306255e-05, + "loss": 0.0626, + "step": 3746 + }, + { + "epoch": 0.40622289679098006, + "grad_norm": 1.947528600692749, + "learning_rate": 1.7354562599738867e-05, + "loss": 0.0287, + "step": 3747 + }, + { + "epoch": 0.4063313096270598, + "grad_norm": 1.3417608737945557, + "learning_rate": 1.735383722617148e-05, + "loss": 0.0358, + "step": 3748 + }, + { + "epoch": 0.40643972246313964, + "grad_norm": 2.831066131591797, + "learning_rate": 1.735311185260409e-05, + "loss": 0.0608, + "step": 3749 + }, + { + "epoch": 0.4065481352992194, + "grad_norm": 1.8329002857208252, + "learning_rate": 1.7352386479036703e-05, + "loss": 0.0325, + "step": 3750 + }, + { + "epoch": 0.4066565481352992, + "grad_norm": 1.9375160932540894, + "learning_rate": 1.735166110546932e-05, + "loss": 0.0262, + "step": 3751 + }, + { + "epoch": 0.406764960971379, + "grad_norm": 6.18037223815918, + "learning_rate": 1.735093573190193e-05, + "loss": 0.0773, + "step": 3752 + }, + { + "epoch": 0.4068733738074588, + "grad_norm": 3.3368160724639893, + "learning_rate": 1.7350210358334543e-05, + "loss": 0.0413, + "step": 3753 + }, + { + "epoch": 0.40698178664353857, + "grad_norm": 3.751958131790161, + "learning_rate": 1.734948498476716e-05, + "loss": 0.0526, + "step": 3754 + }, + { + "epoch": 0.4070901994796184, + "grad_norm": 4.162976264953613, + "learning_rate": 1.734875961119977e-05, + "loss": 0.0669, + "step": 3755 + }, + { + "epoch": 0.4071986123156982, + "grad_norm": 2.9987235069274902, + "learning_rate": 1.7348034237632383e-05, + "loss": 0.0469, + "step": 3756 + }, + { + "epoch": 0.40730702515177797, + "grad_norm": 1.1671779155731201, + "learning_rate": 1.7347308864064995e-05, + "loss": 0.014, + "step": 3757 + }, + { + "epoch": 0.4074154379878578, + "grad_norm": 7.929171562194824, + "learning_rate": 1.7346583490497607e-05, + "loss": 0.0289, + "step": 3758 + }, + { + "epoch": 0.40752385082393755, + "grad_norm": 1.0527963638305664, + "learning_rate": 1.734585811693022e-05, + "loss": 0.0199, + "step": 3759 + }, + { + "epoch": 0.40763226366001737, + "grad_norm": 1.6982989311218262, + "learning_rate": 1.7345132743362835e-05, + "loss": 0.0264, + "step": 3760 + }, + { + "epoch": 0.40774067649609713, + "grad_norm": 3.554129123687744, + "learning_rate": 1.7344407369795447e-05, + "loss": 0.0555, + "step": 3761 + }, + { + "epoch": 0.40784908933217695, + "grad_norm": 2.734692335128784, + "learning_rate": 1.734368199622806e-05, + "loss": 0.0512, + "step": 3762 + }, + { + "epoch": 0.4079575021682567, + "grad_norm": 3.023193597793579, + "learning_rate": 1.734295662266067e-05, + "loss": 0.0615, + "step": 3763 + }, + { + "epoch": 0.40806591500433653, + "grad_norm": 2.849200487136841, + "learning_rate": 1.7342231249093284e-05, + "loss": 0.0508, + "step": 3764 + }, + { + "epoch": 0.4081743278404163, + "grad_norm": 4.013984203338623, + "learning_rate": 1.7341505875525896e-05, + "loss": 0.0397, + "step": 3765 + }, + { + "epoch": 0.4082827406764961, + "grad_norm": 1.531520962715149, + "learning_rate": 1.7340780501958508e-05, + "loss": 0.0308, + "step": 3766 + }, + { + "epoch": 0.4083911535125759, + "grad_norm": 4.935925483703613, + "learning_rate": 1.7340055128391123e-05, + "loss": 0.008, + "step": 3767 + }, + { + "epoch": 0.4084995663486557, + "grad_norm": 3.3989763259887695, + "learning_rate": 1.7339329754823736e-05, + "loss": 0.0489, + "step": 3768 + }, + { + "epoch": 0.40860797918473546, + "grad_norm": 3.9238638877868652, + "learning_rate": 1.7338604381256348e-05, + "loss": 0.0199, + "step": 3769 + }, + { + "epoch": 0.4087163920208153, + "grad_norm": 2.9301414489746094, + "learning_rate": 1.733787900768896e-05, + "loss": 0.0756, + "step": 3770 + }, + { + "epoch": 0.40882480485689504, + "grad_norm": 1.3993511199951172, + "learning_rate": 1.7337153634121572e-05, + "loss": 0.0162, + "step": 3771 + }, + { + "epoch": 0.40893321769297486, + "grad_norm": 1.632948398590088, + "learning_rate": 1.7336428260554188e-05, + "loss": 0.0154, + "step": 3772 + }, + { + "epoch": 0.4090416305290546, + "grad_norm": 3.01960825920105, + "learning_rate": 1.73357028869868e-05, + "loss": 0.0317, + "step": 3773 + }, + { + "epoch": 0.40915004336513444, + "grad_norm": 6.996201038360596, + "learning_rate": 1.7334977513419412e-05, + "loss": 0.0364, + "step": 3774 + }, + { + "epoch": 0.4092584562012142, + "grad_norm": 3.1948275566101074, + "learning_rate": 1.7334252139852027e-05, + "loss": 0.0324, + "step": 3775 + }, + { + "epoch": 0.409366869037294, + "grad_norm": 1.5997837781906128, + "learning_rate": 1.733352676628464e-05, + "loss": 0.0307, + "step": 3776 + }, + { + "epoch": 0.4094752818733738, + "grad_norm": 4.703310966491699, + "learning_rate": 1.733280139271725e-05, + "loss": 0.0436, + "step": 3777 + }, + { + "epoch": 0.4095836947094536, + "grad_norm": 2.9033896923065186, + "learning_rate": 1.7332076019149864e-05, + "loss": 0.02, + "step": 3778 + }, + { + "epoch": 0.40969210754553337, + "grad_norm": 5.543920516967773, + "learning_rate": 1.7331350645582476e-05, + "loss": 0.0383, + "step": 3779 + }, + { + "epoch": 0.4098005203816132, + "grad_norm": 4.165319442749023, + "learning_rate": 1.7330625272015088e-05, + "loss": 0.0339, + "step": 3780 + }, + { + "epoch": 0.40990893321769295, + "grad_norm": 4.026458263397217, + "learning_rate": 1.73298998984477e-05, + "loss": 0.048, + "step": 3781 + }, + { + "epoch": 0.41001734605377277, + "grad_norm": 1.6461093425750732, + "learning_rate": 1.7329174524880316e-05, + "loss": 0.0151, + "step": 3782 + }, + { + "epoch": 0.41012575888985253, + "grad_norm": 3.045064687728882, + "learning_rate": 1.7328449151312928e-05, + "loss": 0.0443, + "step": 3783 + }, + { + "epoch": 0.41023417172593235, + "grad_norm": 2.158379077911377, + "learning_rate": 1.732772377774554e-05, + "loss": 0.0308, + "step": 3784 + }, + { + "epoch": 0.41034258456201217, + "grad_norm": 2.51021146774292, + "learning_rate": 1.7326998404178152e-05, + "loss": 0.0285, + "step": 3785 + }, + { + "epoch": 0.41045099739809193, + "grad_norm": 3.2544772624969482, + "learning_rate": 1.7326273030610764e-05, + "loss": 0.0329, + "step": 3786 + }, + { + "epoch": 0.41055941023417175, + "grad_norm": 8.222790718078613, + "learning_rate": 1.7325547657043377e-05, + "loss": 0.0895, + "step": 3787 + }, + { + "epoch": 0.4106678230702515, + "grad_norm": 6.837479114532471, + "learning_rate": 1.7324822283475992e-05, + "loss": 0.0741, + "step": 3788 + }, + { + "epoch": 0.41077623590633133, + "grad_norm": 1.8314967155456543, + "learning_rate": 1.7324096909908604e-05, + "loss": 0.0172, + "step": 3789 + }, + { + "epoch": 0.4108846487424111, + "grad_norm": 4.8427886962890625, + "learning_rate": 1.732337153634122e-05, + "loss": 0.0624, + "step": 3790 + }, + { + "epoch": 0.4109930615784909, + "grad_norm": 4.2458696365356445, + "learning_rate": 1.7322646162773832e-05, + "loss": 0.0646, + "step": 3791 + }, + { + "epoch": 0.4111014744145707, + "grad_norm": 3.9180610179901123, + "learning_rate": 1.7321920789206444e-05, + "loss": 0.0888, + "step": 3792 + }, + { + "epoch": 0.4112098872506505, + "grad_norm": 1.626326560974121, + "learning_rate": 1.7321195415639056e-05, + "loss": 0.0207, + "step": 3793 + }, + { + "epoch": 0.41131830008673026, + "grad_norm": 3.8758437633514404, + "learning_rate": 1.7320470042071668e-05, + "loss": 0.0388, + "step": 3794 + }, + { + "epoch": 0.4114267129228101, + "grad_norm": 2.1209895610809326, + "learning_rate": 1.731974466850428e-05, + "loss": 0.0277, + "step": 3795 + }, + { + "epoch": 0.41153512575888984, + "grad_norm": 2.106884241104126, + "learning_rate": 1.7319019294936893e-05, + "loss": 0.0287, + "step": 3796 + }, + { + "epoch": 0.41164353859496966, + "grad_norm": 5.373446941375732, + "learning_rate": 1.7318293921369508e-05, + "loss": 0.0878, + "step": 3797 + }, + { + "epoch": 0.4117519514310494, + "grad_norm": 1.6244065761566162, + "learning_rate": 1.731756854780212e-05, + "loss": 0.0309, + "step": 3798 + }, + { + "epoch": 0.41186036426712924, + "grad_norm": 2.363145112991333, + "learning_rate": 1.7316843174234732e-05, + "loss": 0.0207, + "step": 3799 + }, + { + "epoch": 0.411968777103209, + "grad_norm": 5.626629829406738, + "learning_rate": 1.7316117800667345e-05, + "loss": 0.1085, + "step": 3800 + }, + { + "epoch": 0.4120771899392888, + "grad_norm": 2.6827027797698975, + "learning_rate": 1.7315392427099957e-05, + "loss": 0.0644, + "step": 3801 + }, + { + "epoch": 0.4121856027753686, + "grad_norm": 1.7882983684539795, + "learning_rate": 1.731466705353257e-05, + "loss": 0.0298, + "step": 3802 + }, + { + "epoch": 0.4122940156114484, + "grad_norm": 1.604780673980713, + "learning_rate": 1.7313941679965184e-05, + "loss": 0.0289, + "step": 3803 + }, + { + "epoch": 0.41240242844752817, + "grad_norm": 3.0767226219177246, + "learning_rate": 1.7313216306397796e-05, + "loss": 0.0181, + "step": 3804 + }, + { + "epoch": 0.412510841283608, + "grad_norm": 2.667541980743408, + "learning_rate": 1.731249093283041e-05, + "loss": 0.0211, + "step": 3805 + }, + { + "epoch": 0.41261925411968775, + "grad_norm": 2.2867870330810547, + "learning_rate": 1.7311765559263024e-05, + "loss": 0.0398, + "step": 3806 + }, + { + "epoch": 0.41272766695576757, + "grad_norm": 1.2084273099899292, + "learning_rate": 1.7311040185695636e-05, + "loss": 0.0154, + "step": 3807 + }, + { + "epoch": 0.41283607979184733, + "grad_norm": 2.757852792739868, + "learning_rate": 1.731031481212825e-05, + "loss": 0.0347, + "step": 3808 + }, + { + "epoch": 0.41294449262792715, + "grad_norm": 4.685291290283203, + "learning_rate": 1.730958943856086e-05, + "loss": 0.0635, + "step": 3809 + }, + { + "epoch": 0.4130529054640069, + "grad_norm": 0.5844528675079346, + "learning_rate": 1.7308864064993473e-05, + "loss": 0.0048, + "step": 3810 + }, + { + "epoch": 0.41316131830008673, + "grad_norm": 1.7635506391525269, + "learning_rate": 1.7308138691426085e-05, + "loss": 0.0649, + "step": 3811 + }, + { + "epoch": 0.4132697311361665, + "grad_norm": 5.020039081573486, + "learning_rate": 1.73074133178587e-05, + "loss": 0.0742, + "step": 3812 + }, + { + "epoch": 0.4133781439722463, + "grad_norm": 2.14214825630188, + "learning_rate": 1.7306687944291313e-05, + "loss": 0.0259, + "step": 3813 + }, + { + "epoch": 0.41348655680832613, + "grad_norm": 2.9164538383483887, + "learning_rate": 1.7305962570723925e-05, + "loss": 0.034, + "step": 3814 + }, + { + "epoch": 0.4135949696444059, + "grad_norm": 5.928623199462891, + "learning_rate": 1.7305237197156537e-05, + "loss": 0.0876, + "step": 3815 + }, + { + "epoch": 0.4137033824804857, + "grad_norm": 0.852740466594696, + "learning_rate": 1.730451182358915e-05, + "loss": 0.0117, + "step": 3816 + }, + { + "epoch": 0.4138117953165655, + "grad_norm": 4.853782653808594, + "learning_rate": 1.730378645002176e-05, + "loss": 0.0581, + "step": 3817 + }, + { + "epoch": 0.4139202081526453, + "grad_norm": 1.122643232345581, + "learning_rate": 1.7303061076454373e-05, + "loss": 0.0154, + "step": 3818 + }, + { + "epoch": 0.41402862098872506, + "grad_norm": 2.2813720703125, + "learning_rate": 1.730233570288699e-05, + "loss": 0.0211, + "step": 3819 + }, + { + "epoch": 0.4141370338248049, + "grad_norm": 1.589039921760559, + "learning_rate": 1.73016103293196e-05, + "loss": 0.0286, + "step": 3820 + }, + { + "epoch": 0.41424544666088464, + "grad_norm": 5.3292999267578125, + "learning_rate": 1.7300884955752213e-05, + "loss": 0.0979, + "step": 3821 + }, + { + "epoch": 0.41435385949696446, + "grad_norm": 3.4408435821533203, + "learning_rate": 1.7300159582184825e-05, + "loss": 0.0317, + "step": 3822 + }, + { + "epoch": 0.4144622723330442, + "grad_norm": 6.716139793395996, + "learning_rate": 1.729943420861744e-05, + "loss": 0.0214, + "step": 3823 + }, + { + "epoch": 0.41457068516912404, + "grad_norm": 5.959414958953857, + "learning_rate": 1.7298708835050053e-05, + "loss": 0.0299, + "step": 3824 + }, + { + "epoch": 0.4146790980052038, + "grad_norm": 2.9281129837036133, + "learning_rate": 1.7297983461482665e-05, + "loss": 0.024, + "step": 3825 + }, + { + "epoch": 0.4147875108412836, + "grad_norm": 5.390382766723633, + "learning_rate": 1.7297258087915277e-05, + "loss": 0.0828, + "step": 3826 + }, + { + "epoch": 0.4148959236773634, + "grad_norm": 3.0025393962860107, + "learning_rate": 1.7296532714347893e-05, + "loss": 0.0536, + "step": 3827 + }, + { + "epoch": 0.4150043365134432, + "grad_norm": 3.3234801292419434, + "learning_rate": 1.7295807340780505e-05, + "loss": 0.064, + "step": 3828 + }, + { + "epoch": 0.41511274934952297, + "grad_norm": 1.7491503953933716, + "learning_rate": 1.7295081967213117e-05, + "loss": 0.0222, + "step": 3829 + }, + { + "epoch": 0.4152211621856028, + "grad_norm": 2.213405132293701, + "learning_rate": 1.729435659364573e-05, + "loss": 0.0513, + "step": 3830 + }, + { + "epoch": 0.41532957502168255, + "grad_norm": 3.2103970050811768, + "learning_rate": 1.729363122007834e-05, + "loss": 0.0519, + "step": 3831 + }, + { + "epoch": 0.41543798785776237, + "grad_norm": 3.1570723056793213, + "learning_rate": 1.7292905846510953e-05, + "loss": 0.0405, + "step": 3832 + }, + { + "epoch": 0.41554640069384213, + "grad_norm": 4.665624618530273, + "learning_rate": 1.7292180472943566e-05, + "loss": 0.0449, + "step": 3833 + }, + { + "epoch": 0.41565481352992195, + "grad_norm": 1.682875156402588, + "learning_rate": 1.729145509937618e-05, + "loss": 0.0267, + "step": 3834 + }, + { + "epoch": 0.4157632263660017, + "grad_norm": 4.0482683181762695, + "learning_rate": 1.7290729725808793e-05, + "loss": 0.0417, + "step": 3835 + }, + { + "epoch": 0.41587163920208153, + "grad_norm": 2.9353504180908203, + "learning_rate": 1.7290004352241405e-05, + "loss": 0.0471, + "step": 3836 + }, + { + "epoch": 0.4159800520381613, + "grad_norm": 2.57015061378479, + "learning_rate": 1.7289278978674018e-05, + "loss": 0.0481, + "step": 3837 + }, + { + "epoch": 0.4160884648742411, + "grad_norm": 4.671820640563965, + "learning_rate": 1.728855360510663e-05, + "loss": 0.0217, + "step": 3838 + }, + { + "epoch": 0.4161968777103209, + "grad_norm": 1.7382936477661133, + "learning_rate": 1.7287828231539242e-05, + "loss": 0.0273, + "step": 3839 + }, + { + "epoch": 0.4163052905464007, + "grad_norm": 1.4660052061080933, + "learning_rate": 1.7287102857971857e-05, + "loss": 0.0245, + "step": 3840 + }, + { + "epoch": 0.41641370338248046, + "grad_norm": 7.912234306335449, + "learning_rate": 1.728637748440447e-05, + "loss": 0.0565, + "step": 3841 + }, + { + "epoch": 0.4165221162185603, + "grad_norm": 2.197736978530884, + "learning_rate": 1.7285652110837085e-05, + "loss": 0.0213, + "step": 3842 + }, + { + "epoch": 0.41663052905464004, + "grad_norm": 1.7319984436035156, + "learning_rate": 1.7284926737269697e-05, + "loss": 0.0348, + "step": 3843 + }, + { + "epoch": 0.41673894189071986, + "grad_norm": 5.190864086151123, + "learning_rate": 1.728420136370231e-05, + "loss": 0.0426, + "step": 3844 + }, + { + "epoch": 0.4168473547267997, + "grad_norm": 1.4912748336791992, + "learning_rate": 1.728347599013492e-05, + "loss": 0.0174, + "step": 3845 + }, + { + "epoch": 0.41695576756287944, + "grad_norm": 2.529711961746216, + "learning_rate": 1.7282750616567534e-05, + "loss": 0.0588, + "step": 3846 + }, + { + "epoch": 0.41706418039895926, + "grad_norm": 1.3808424472808838, + "learning_rate": 1.7282025243000146e-05, + "loss": 0.0181, + "step": 3847 + }, + { + "epoch": 0.417172593235039, + "grad_norm": 2.9327609539031982, + "learning_rate": 1.7281299869432758e-05, + "loss": 0.0417, + "step": 3848 + }, + { + "epoch": 0.41728100607111884, + "grad_norm": 2.615145683288574, + "learning_rate": 1.7280574495865373e-05, + "loss": 0.0305, + "step": 3849 + }, + { + "epoch": 0.4173894189071986, + "grad_norm": 2.4119443893432617, + "learning_rate": 1.7279849122297986e-05, + "loss": 0.0352, + "step": 3850 + }, + { + "epoch": 0.4174978317432784, + "grad_norm": 2.728598117828369, + "learning_rate": 1.7279123748730598e-05, + "loss": 0.0328, + "step": 3851 + }, + { + "epoch": 0.4176062445793582, + "grad_norm": 3.611501932144165, + "learning_rate": 1.727839837516321e-05, + "loss": 0.0334, + "step": 3852 + }, + { + "epoch": 0.417714657415438, + "grad_norm": 3.354278326034546, + "learning_rate": 1.7277673001595822e-05, + "loss": 0.0167, + "step": 3853 + }, + { + "epoch": 0.41782307025151777, + "grad_norm": 1.171744465827942, + "learning_rate": 1.7276947628028434e-05, + "loss": 0.0199, + "step": 3854 + }, + { + "epoch": 0.4179314830875976, + "grad_norm": 1.0272464752197266, + "learning_rate": 1.727622225446105e-05, + "loss": 0.0095, + "step": 3855 + }, + { + "epoch": 0.41803989592367735, + "grad_norm": 6.346652507781982, + "learning_rate": 1.7275496880893662e-05, + "loss": 0.0802, + "step": 3856 + }, + { + "epoch": 0.41814830875975717, + "grad_norm": 3.47702693939209, + "learning_rate": 1.7274771507326274e-05, + "loss": 0.0603, + "step": 3857 + }, + { + "epoch": 0.41825672159583693, + "grad_norm": 7.574621200561523, + "learning_rate": 1.727404613375889e-05, + "loss": 0.0677, + "step": 3858 + }, + { + "epoch": 0.41836513443191675, + "grad_norm": 4.571320056915283, + "learning_rate": 1.72733207601915e-05, + "loss": 0.0377, + "step": 3859 + }, + { + "epoch": 0.4184735472679965, + "grad_norm": 5.219276428222656, + "learning_rate": 1.7272595386624114e-05, + "loss": 0.0815, + "step": 3860 + }, + { + "epoch": 0.41858196010407633, + "grad_norm": 1.3933613300323486, + "learning_rate": 1.7271870013056726e-05, + "loss": 0.0155, + "step": 3861 + }, + { + "epoch": 0.4186903729401561, + "grad_norm": 2.2271804809570312, + "learning_rate": 1.7271144639489338e-05, + "loss": 0.0373, + "step": 3862 + }, + { + "epoch": 0.4187987857762359, + "grad_norm": 4.780783653259277, + "learning_rate": 1.727041926592195e-05, + "loss": 0.0289, + "step": 3863 + }, + { + "epoch": 0.4189071986123157, + "grad_norm": 1.9710886478424072, + "learning_rate": 1.7269693892354566e-05, + "loss": 0.0263, + "step": 3864 + }, + { + "epoch": 0.4190156114483955, + "grad_norm": 7.625316143035889, + "learning_rate": 1.7268968518787178e-05, + "loss": 0.1043, + "step": 3865 + }, + { + "epoch": 0.41912402428447526, + "grad_norm": 1.6319456100463867, + "learning_rate": 1.726824314521979e-05, + "loss": 0.0172, + "step": 3866 + }, + { + "epoch": 0.4192324371205551, + "grad_norm": 4.6306867599487305, + "learning_rate": 1.7267517771652402e-05, + "loss": 0.0385, + "step": 3867 + }, + { + "epoch": 0.41934084995663484, + "grad_norm": 4.910335540771484, + "learning_rate": 1.7266792398085014e-05, + "loss": 0.0323, + "step": 3868 + }, + { + "epoch": 0.41944926279271466, + "grad_norm": 4.330243110656738, + "learning_rate": 1.7266067024517626e-05, + "loss": 0.0786, + "step": 3869 + }, + { + "epoch": 0.4195576756287944, + "grad_norm": 6.0695271492004395, + "learning_rate": 1.726534165095024e-05, + "loss": 0.0266, + "step": 3870 + }, + { + "epoch": 0.41966608846487424, + "grad_norm": 5.438863277435303, + "learning_rate": 1.7264616277382854e-05, + "loss": 0.0212, + "step": 3871 + }, + { + "epoch": 0.419774501300954, + "grad_norm": 3.963033437728882, + "learning_rate": 1.7263890903815466e-05, + "loss": 0.0512, + "step": 3872 + }, + { + "epoch": 0.4198829141370338, + "grad_norm": 5.541579246520996, + "learning_rate": 1.726316553024808e-05, + "loss": 0.0796, + "step": 3873 + }, + { + "epoch": 0.41999132697311364, + "grad_norm": 3.0511972904205322, + "learning_rate": 1.726244015668069e-05, + "loss": 0.0375, + "step": 3874 + }, + { + "epoch": 0.4200997398091934, + "grad_norm": 2.706510543823242, + "learning_rate": 1.7261714783113306e-05, + "loss": 0.0225, + "step": 3875 + }, + { + "epoch": 0.4202081526452732, + "grad_norm": 3.274970054626465, + "learning_rate": 1.7260989409545918e-05, + "loss": 0.054, + "step": 3876 + }, + { + "epoch": 0.420316565481353, + "grad_norm": 6.107010841369629, + "learning_rate": 1.726026403597853e-05, + "loss": 0.0732, + "step": 3877 + }, + { + "epoch": 0.4204249783174328, + "grad_norm": 1.7489497661590576, + "learning_rate": 1.7259538662411143e-05, + "loss": 0.0192, + "step": 3878 + }, + { + "epoch": 0.42053339115351257, + "grad_norm": 2.6175765991210938, + "learning_rate": 1.7258813288843758e-05, + "loss": 0.0396, + "step": 3879 + }, + { + "epoch": 0.4206418039895924, + "grad_norm": 6.743903636932373, + "learning_rate": 1.725808791527637e-05, + "loss": 0.0322, + "step": 3880 + }, + { + "epoch": 0.42075021682567215, + "grad_norm": 3.9170210361480713, + "learning_rate": 1.7257362541708982e-05, + "loss": 0.0352, + "step": 3881 + }, + { + "epoch": 0.42085862966175197, + "grad_norm": 4.458588600158691, + "learning_rate": 1.7256637168141594e-05, + "loss": 0.0955, + "step": 3882 + }, + { + "epoch": 0.42096704249783173, + "grad_norm": 4.245110034942627, + "learning_rate": 1.7255911794574207e-05, + "loss": 0.0771, + "step": 3883 + }, + { + "epoch": 0.42107545533391155, + "grad_norm": 1.961476445198059, + "learning_rate": 1.725518642100682e-05, + "loss": 0.0373, + "step": 3884 + }, + { + "epoch": 0.4211838681699913, + "grad_norm": 1.9694408178329468, + "learning_rate": 1.725446104743943e-05, + "loss": 0.0257, + "step": 3885 + }, + { + "epoch": 0.42129228100607113, + "grad_norm": 2.133056163787842, + "learning_rate": 1.7253735673872046e-05, + "loss": 0.0129, + "step": 3886 + }, + { + "epoch": 0.4214006938421509, + "grad_norm": 1.1910011768341064, + "learning_rate": 1.725301030030466e-05, + "loss": 0.0098, + "step": 3887 + }, + { + "epoch": 0.4215091066782307, + "grad_norm": 4.3945393562316895, + "learning_rate": 1.725228492673727e-05, + "loss": 0.0471, + "step": 3888 + }, + { + "epoch": 0.4216175195143105, + "grad_norm": 2.491676092147827, + "learning_rate": 1.7251559553169883e-05, + "loss": 0.0544, + "step": 3889 + }, + { + "epoch": 0.4217259323503903, + "grad_norm": 3.3743927478790283, + "learning_rate": 1.7250834179602495e-05, + "loss": 0.0617, + "step": 3890 + }, + { + "epoch": 0.42183434518647006, + "grad_norm": 3.1561601161956787, + "learning_rate": 1.7250108806035107e-05, + "loss": 0.0212, + "step": 3891 + }, + { + "epoch": 0.4219427580225499, + "grad_norm": 3.2812328338623047, + "learning_rate": 1.7249383432467723e-05, + "loss": 0.0588, + "step": 3892 + }, + { + "epoch": 0.42205117085862964, + "grad_norm": 2.929633855819702, + "learning_rate": 1.7248658058900335e-05, + "loss": 0.0595, + "step": 3893 + }, + { + "epoch": 0.42215958369470946, + "grad_norm": 3.8091955184936523, + "learning_rate": 1.724793268533295e-05, + "loss": 0.0811, + "step": 3894 + }, + { + "epoch": 0.4222679965307892, + "grad_norm": 2.2327311038970947, + "learning_rate": 1.7247207311765563e-05, + "loss": 0.0328, + "step": 3895 + }, + { + "epoch": 0.42237640936686904, + "grad_norm": 3.4462404251098633, + "learning_rate": 1.7246481938198175e-05, + "loss": 0.0781, + "step": 3896 + }, + { + "epoch": 0.4224848222029488, + "grad_norm": 1.518433690071106, + "learning_rate": 1.7245756564630787e-05, + "loss": 0.0209, + "step": 3897 + }, + { + "epoch": 0.4225932350390286, + "grad_norm": 2.74560284614563, + "learning_rate": 1.72450311910634e-05, + "loss": 0.0411, + "step": 3898 + }, + { + "epoch": 0.4227016478751084, + "grad_norm": 6.089324474334717, + "learning_rate": 1.724430581749601e-05, + "loss": 0.0483, + "step": 3899 + }, + { + "epoch": 0.4228100607111882, + "grad_norm": 4.9456048011779785, + "learning_rate": 1.7243580443928623e-05, + "loss": 0.0503, + "step": 3900 + }, + { + "epoch": 0.42291847354726797, + "grad_norm": 1.7479881048202515, + "learning_rate": 1.724285507036124e-05, + "loss": 0.0287, + "step": 3901 + }, + { + "epoch": 0.4230268863833478, + "grad_norm": 0.9589519500732422, + "learning_rate": 1.724212969679385e-05, + "loss": 0.0218, + "step": 3902 + }, + { + "epoch": 0.4231352992194276, + "grad_norm": 1.8024864196777344, + "learning_rate": 1.7241404323226463e-05, + "loss": 0.0345, + "step": 3903 + }, + { + "epoch": 0.42324371205550737, + "grad_norm": 6.262121200561523, + "learning_rate": 1.7240678949659075e-05, + "loss": 0.0259, + "step": 3904 + }, + { + "epoch": 0.4233521248915872, + "grad_norm": 1.099487543106079, + "learning_rate": 1.7239953576091687e-05, + "loss": 0.0162, + "step": 3905 + }, + { + "epoch": 0.42346053772766695, + "grad_norm": 1.5703623294830322, + "learning_rate": 1.72392282025243e-05, + "loss": 0.0187, + "step": 3906 + }, + { + "epoch": 0.42356895056374677, + "grad_norm": 5.3576555252075195, + "learning_rate": 1.7238502828956915e-05, + "loss": 0.0853, + "step": 3907 + }, + { + "epoch": 0.42367736339982653, + "grad_norm": 6.097900390625, + "learning_rate": 1.7237777455389527e-05, + "loss": 0.0476, + "step": 3908 + }, + { + "epoch": 0.42378577623590635, + "grad_norm": 5.570598602294922, + "learning_rate": 1.723705208182214e-05, + "loss": 0.0546, + "step": 3909 + }, + { + "epoch": 0.4238941890719861, + "grad_norm": 1.2237321138381958, + "learning_rate": 1.7236326708254755e-05, + "loss": 0.0136, + "step": 3910 + }, + { + "epoch": 0.42400260190806593, + "grad_norm": 2.279118537902832, + "learning_rate": 1.7235601334687367e-05, + "loss": 0.0399, + "step": 3911 + }, + { + "epoch": 0.4241110147441457, + "grad_norm": 6.516868591308594, + "learning_rate": 1.723487596111998e-05, + "loss": 0.0324, + "step": 3912 + }, + { + "epoch": 0.4242194275802255, + "grad_norm": 1.361261010169983, + "learning_rate": 1.723415058755259e-05, + "loss": 0.0252, + "step": 3913 + }, + { + "epoch": 0.4243278404163053, + "grad_norm": 2.454376459121704, + "learning_rate": 1.7233425213985203e-05, + "loss": 0.0237, + "step": 3914 + }, + { + "epoch": 0.4244362532523851, + "grad_norm": 2.635852098464966, + "learning_rate": 1.7232699840417816e-05, + "loss": 0.0536, + "step": 3915 + }, + { + "epoch": 0.42454466608846486, + "grad_norm": 3.5632400512695312, + "learning_rate": 1.723197446685043e-05, + "loss": 0.0802, + "step": 3916 + }, + { + "epoch": 0.4246530789245447, + "grad_norm": 5.984455108642578, + "learning_rate": 1.7231249093283043e-05, + "loss": 0.1045, + "step": 3917 + }, + { + "epoch": 0.42476149176062444, + "grad_norm": 1.0292905569076538, + "learning_rate": 1.7230523719715655e-05, + "loss": 0.0073, + "step": 3918 + }, + { + "epoch": 0.42486990459670426, + "grad_norm": 5.5086283683776855, + "learning_rate": 1.7229798346148268e-05, + "loss": 0.0244, + "step": 3919 + }, + { + "epoch": 0.424978317432784, + "grad_norm": 3.7804973125457764, + "learning_rate": 1.722907297258088e-05, + "loss": 0.0567, + "step": 3920 + }, + { + "epoch": 0.42508673026886384, + "grad_norm": 1.2519727945327759, + "learning_rate": 1.7228347599013492e-05, + "loss": 0.0132, + "step": 3921 + }, + { + "epoch": 0.4251951431049436, + "grad_norm": 5.346220970153809, + "learning_rate": 1.7227622225446104e-05, + "loss": 0.0281, + "step": 3922 + }, + { + "epoch": 0.4253035559410234, + "grad_norm": 9.4461669921875, + "learning_rate": 1.722689685187872e-05, + "loss": 0.0508, + "step": 3923 + }, + { + "epoch": 0.4254119687771032, + "grad_norm": 2.014263868331909, + "learning_rate": 1.722617147831133e-05, + "loss": 0.0189, + "step": 3924 + }, + { + "epoch": 0.425520381613183, + "grad_norm": 0.8980073928833008, + "learning_rate": 1.7225446104743944e-05, + "loss": 0.0062, + "step": 3925 + }, + { + "epoch": 0.42562879444926277, + "grad_norm": 1.9875025749206543, + "learning_rate": 1.7224720731176556e-05, + "loss": 0.0441, + "step": 3926 + }, + { + "epoch": 0.4257372072853426, + "grad_norm": 1.7529393434524536, + "learning_rate": 1.722399535760917e-05, + "loss": 0.0273, + "step": 3927 + }, + { + "epoch": 0.42584562012142235, + "grad_norm": 4.472391128540039, + "learning_rate": 1.7223269984041784e-05, + "loss": 0.0551, + "step": 3928 + }, + { + "epoch": 0.42595403295750217, + "grad_norm": 4.332386016845703, + "learning_rate": 1.7222544610474396e-05, + "loss": 0.0383, + "step": 3929 + }, + { + "epoch": 0.42606244579358193, + "grad_norm": 3.593315601348877, + "learning_rate": 1.7221819236907008e-05, + "loss": 0.0318, + "step": 3930 + }, + { + "epoch": 0.42617085862966175, + "grad_norm": 2.1655614376068115, + "learning_rate": 1.7221093863339623e-05, + "loss": 0.0095, + "step": 3931 + }, + { + "epoch": 0.4262792714657415, + "grad_norm": 5.803516864776611, + "learning_rate": 1.7220368489772236e-05, + "loss": 0.056, + "step": 3932 + }, + { + "epoch": 0.42638768430182133, + "grad_norm": 1.856900691986084, + "learning_rate": 1.7219643116204848e-05, + "loss": 0.0275, + "step": 3933 + }, + { + "epoch": 0.42649609713790115, + "grad_norm": 2.2702293395996094, + "learning_rate": 1.721891774263746e-05, + "loss": 0.01, + "step": 3934 + }, + { + "epoch": 0.4266045099739809, + "grad_norm": 4.247564315795898, + "learning_rate": 1.7218192369070072e-05, + "loss": 0.0782, + "step": 3935 + }, + { + "epoch": 0.42671292281006074, + "grad_norm": 3.303804636001587, + "learning_rate": 1.7217466995502684e-05, + "loss": 0.0475, + "step": 3936 + }, + { + "epoch": 0.4268213356461405, + "grad_norm": 2.192288398742676, + "learning_rate": 1.7216741621935296e-05, + "loss": 0.0344, + "step": 3937 + }, + { + "epoch": 0.4269297484822203, + "grad_norm": 6.687643051147461, + "learning_rate": 1.7216016248367912e-05, + "loss": 0.0811, + "step": 3938 + }, + { + "epoch": 0.4270381613183001, + "grad_norm": 2.227369546890259, + "learning_rate": 1.7215290874800524e-05, + "loss": 0.032, + "step": 3939 + }, + { + "epoch": 0.4271465741543799, + "grad_norm": 5.508747100830078, + "learning_rate": 1.7214565501233136e-05, + "loss": 0.0453, + "step": 3940 + }, + { + "epoch": 0.42725498699045966, + "grad_norm": 3.8427979946136475, + "learning_rate": 1.7213840127665748e-05, + "loss": 0.0573, + "step": 3941 + }, + { + "epoch": 0.4273633998265395, + "grad_norm": 1.9508293867111206, + "learning_rate": 1.721311475409836e-05, + "loss": 0.0307, + "step": 3942 + }, + { + "epoch": 0.42747181266261924, + "grad_norm": 5.260270595550537, + "learning_rate": 1.7212389380530973e-05, + "loss": 0.0643, + "step": 3943 + }, + { + "epoch": 0.42758022549869906, + "grad_norm": 5.312746047973633, + "learning_rate": 1.7211664006963588e-05, + "loss": 0.0433, + "step": 3944 + }, + { + "epoch": 0.4276886383347788, + "grad_norm": 0.27812182903289795, + "learning_rate": 1.72109386333962e-05, + "loss": 0.0014, + "step": 3945 + }, + { + "epoch": 0.42779705117085864, + "grad_norm": 2.4416663646698, + "learning_rate": 1.7210213259828816e-05, + "loss": 0.0594, + "step": 3946 + }, + { + "epoch": 0.4279054640069384, + "grad_norm": 2.8234620094299316, + "learning_rate": 1.7209487886261428e-05, + "loss": 0.0405, + "step": 3947 + }, + { + "epoch": 0.4280138768430182, + "grad_norm": 5.630441665649414, + "learning_rate": 1.720876251269404e-05, + "loss": 0.0712, + "step": 3948 + }, + { + "epoch": 0.428122289679098, + "grad_norm": 3.101438522338867, + "learning_rate": 1.7208037139126652e-05, + "loss": 0.0278, + "step": 3949 + }, + { + "epoch": 0.4282307025151778, + "grad_norm": 1.2593703269958496, + "learning_rate": 1.7207311765559264e-05, + "loss": 0.016, + "step": 3950 + }, + { + "epoch": 0.42833911535125757, + "grad_norm": 6.51328706741333, + "learning_rate": 1.7206586391991876e-05, + "loss": 0.044, + "step": 3951 + }, + { + "epoch": 0.4284475281873374, + "grad_norm": 1.0506713390350342, + "learning_rate": 1.720586101842449e-05, + "loss": 0.0288, + "step": 3952 + }, + { + "epoch": 0.42855594102341715, + "grad_norm": 6.087812900543213, + "learning_rate": 1.7205135644857104e-05, + "loss": 0.047, + "step": 3953 + }, + { + "epoch": 0.42866435385949697, + "grad_norm": 4.548233509063721, + "learning_rate": 1.7204410271289716e-05, + "loss": 0.0342, + "step": 3954 + }, + { + "epoch": 0.42877276669557673, + "grad_norm": 6.830498218536377, + "learning_rate": 1.720368489772233e-05, + "loss": 0.0158, + "step": 3955 + }, + { + "epoch": 0.42888117953165655, + "grad_norm": 7.1962456703186035, + "learning_rate": 1.720295952415494e-05, + "loss": 0.0784, + "step": 3956 + }, + { + "epoch": 0.4289895923677363, + "grad_norm": 1.7248456478118896, + "learning_rate": 1.7202234150587553e-05, + "loss": 0.0217, + "step": 3957 + }, + { + "epoch": 0.42909800520381614, + "grad_norm": 4.62910795211792, + "learning_rate": 1.7201508777020165e-05, + "loss": 0.0653, + "step": 3958 + }, + { + "epoch": 0.4292064180398959, + "grad_norm": 3.349670648574829, + "learning_rate": 1.720078340345278e-05, + "loss": 0.0491, + "step": 3959 + }, + { + "epoch": 0.4293148308759757, + "grad_norm": 2.0577611923217773, + "learning_rate": 1.7200058029885393e-05, + "loss": 0.0487, + "step": 3960 + }, + { + "epoch": 0.4294232437120555, + "grad_norm": 3.4171059131622314, + "learning_rate": 1.7199332656318005e-05, + "loss": 0.0967, + "step": 3961 + }, + { + "epoch": 0.4295316565481353, + "grad_norm": 5.129594326019287, + "learning_rate": 1.719860728275062e-05, + "loss": 0.0598, + "step": 3962 + }, + { + "epoch": 0.4296400693842151, + "grad_norm": 2.4003663063049316, + "learning_rate": 1.7197881909183232e-05, + "loss": 0.0227, + "step": 3963 + }, + { + "epoch": 0.4297484822202949, + "grad_norm": 2.157400131225586, + "learning_rate": 1.7197156535615844e-05, + "loss": 0.0188, + "step": 3964 + }, + { + "epoch": 0.4298568950563747, + "grad_norm": 2.4586844444274902, + "learning_rate": 1.7196431162048457e-05, + "loss": 0.0285, + "step": 3965 + }, + { + "epoch": 0.42996530789245446, + "grad_norm": 1.5432583093643188, + "learning_rate": 1.719570578848107e-05, + "loss": 0.0195, + "step": 3966 + }, + { + "epoch": 0.4300737207285343, + "grad_norm": 1.56804621219635, + "learning_rate": 1.719498041491368e-05, + "loss": 0.072, + "step": 3967 + }, + { + "epoch": 0.43018213356461404, + "grad_norm": 1.632080316543579, + "learning_rate": 1.7194255041346296e-05, + "loss": 0.0231, + "step": 3968 + }, + { + "epoch": 0.43029054640069386, + "grad_norm": 2.1756751537323, + "learning_rate": 1.719352966777891e-05, + "loss": 0.0307, + "step": 3969 + }, + { + "epoch": 0.4303989592367736, + "grad_norm": 0.7257576584815979, + "learning_rate": 1.719280429421152e-05, + "loss": 0.0071, + "step": 3970 + }, + { + "epoch": 0.43050737207285344, + "grad_norm": 2.6193976402282715, + "learning_rate": 1.7192078920644133e-05, + "loss": 0.0425, + "step": 3971 + }, + { + "epoch": 0.4306157849089332, + "grad_norm": 2.7592225074768066, + "learning_rate": 1.7191353547076745e-05, + "loss": 0.0542, + "step": 3972 + }, + { + "epoch": 0.430724197745013, + "grad_norm": 14.760022163391113, + "learning_rate": 1.7190628173509357e-05, + "loss": 0.0738, + "step": 3973 + }, + { + "epoch": 0.4308326105810928, + "grad_norm": 2.8993723392486572, + "learning_rate": 1.718990279994197e-05, + "loss": 0.0346, + "step": 3974 + }, + { + "epoch": 0.4309410234171726, + "grad_norm": 5.224318981170654, + "learning_rate": 1.7189177426374585e-05, + "loss": 0.0676, + "step": 3975 + }, + { + "epoch": 0.43104943625325237, + "grad_norm": 8.48237133026123, + "learning_rate": 1.7188452052807197e-05, + "loss": 0.0459, + "step": 3976 + }, + { + "epoch": 0.4311578490893322, + "grad_norm": 6.7012553215026855, + "learning_rate": 1.718772667923981e-05, + "loss": 0.0705, + "step": 3977 + }, + { + "epoch": 0.43126626192541195, + "grad_norm": 5.930606842041016, + "learning_rate": 1.718700130567242e-05, + "loss": 0.077, + "step": 3978 + }, + { + "epoch": 0.43137467476149177, + "grad_norm": 1.6489601135253906, + "learning_rate": 1.7186275932105037e-05, + "loss": 0.0136, + "step": 3979 + }, + { + "epoch": 0.43148308759757154, + "grad_norm": 1.9829089641571045, + "learning_rate": 1.718555055853765e-05, + "loss": 0.043, + "step": 3980 + }, + { + "epoch": 0.43159150043365135, + "grad_norm": 1.935173749923706, + "learning_rate": 1.718482518497026e-05, + "loss": 0.0229, + "step": 3981 + }, + { + "epoch": 0.4316999132697311, + "grad_norm": 1.4062464237213135, + "learning_rate": 1.7184099811402873e-05, + "loss": 0.0203, + "step": 3982 + }, + { + "epoch": 0.43180832610581094, + "grad_norm": 3.9179115295410156, + "learning_rate": 1.718337443783549e-05, + "loss": 0.0564, + "step": 3983 + }, + { + "epoch": 0.4319167389418907, + "grad_norm": 2.9408771991729736, + "learning_rate": 1.71826490642681e-05, + "loss": 0.0546, + "step": 3984 + }, + { + "epoch": 0.4320251517779705, + "grad_norm": 2.4111533164978027, + "learning_rate": 1.7181923690700713e-05, + "loss": 0.0423, + "step": 3985 + }, + { + "epoch": 0.4321335646140503, + "grad_norm": 2.140747308731079, + "learning_rate": 1.7181198317133325e-05, + "loss": 0.0101, + "step": 3986 + }, + { + "epoch": 0.4322419774501301, + "grad_norm": 4.842681407928467, + "learning_rate": 1.7180472943565937e-05, + "loss": 0.0663, + "step": 3987 + }, + { + "epoch": 0.43235039028620986, + "grad_norm": 6.708953380584717, + "learning_rate": 1.717974756999855e-05, + "loss": 0.0692, + "step": 3988 + }, + { + "epoch": 0.4324588031222897, + "grad_norm": 0.8065577149391174, + "learning_rate": 1.717902219643116e-05, + "loss": 0.0187, + "step": 3989 + }, + { + "epoch": 0.43256721595836944, + "grad_norm": 1.6582664251327515, + "learning_rate": 1.7178296822863777e-05, + "loss": 0.0269, + "step": 3990 + }, + { + "epoch": 0.43267562879444926, + "grad_norm": 3.6304752826690674, + "learning_rate": 1.717757144929639e-05, + "loss": 0.1005, + "step": 3991 + }, + { + "epoch": 0.4327840416305291, + "grad_norm": 2.535177707672119, + "learning_rate": 1.7176846075729e-05, + "loss": 0.011, + "step": 3992 + }, + { + "epoch": 0.43289245446660884, + "grad_norm": 2.7173783779144287, + "learning_rate": 1.7176120702161614e-05, + "loss": 0.0208, + "step": 3993 + }, + { + "epoch": 0.43300086730268866, + "grad_norm": 4.691150188446045, + "learning_rate": 1.7175395328594226e-05, + "loss": 0.0771, + "step": 3994 + }, + { + "epoch": 0.4331092801387684, + "grad_norm": 2.1372272968292236, + "learning_rate": 1.7174669955026838e-05, + "loss": 0.0347, + "step": 3995 + }, + { + "epoch": 0.43321769297484825, + "grad_norm": 1.7221155166625977, + "learning_rate": 1.7173944581459453e-05, + "loss": 0.0132, + "step": 3996 + }, + { + "epoch": 0.433326105810928, + "grad_norm": 2.921945571899414, + "learning_rate": 1.7173219207892066e-05, + "loss": 0.0141, + "step": 3997 + }, + { + "epoch": 0.4334345186470078, + "grad_norm": 1.6402426958084106, + "learning_rate": 1.717249383432468e-05, + "loss": 0.037, + "step": 3998 + }, + { + "epoch": 0.4335429314830876, + "grad_norm": 1.7428878545761108, + "learning_rate": 1.7171768460757293e-05, + "loss": 0.0225, + "step": 3999 + }, + { + "epoch": 0.4336513443191674, + "grad_norm": 7.764048099517822, + "learning_rate": 1.7171043087189905e-05, + "loss": 0.0734, + "step": 4000 + }, + { + "epoch": 0.43375975715524717, + "grad_norm": 4.523402690887451, + "learning_rate": 1.7170317713622517e-05, + "loss": 0.0665, + "step": 4001 + }, + { + "epoch": 0.433868169991327, + "grad_norm": 3.017719030380249, + "learning_rate": 1.716959234005513e-05, + "loss": 0.0311, + "step": 4002 + }, + { + "epoch": 0.43397658282740675, + "grad_norm": 7.367895603179932, + "learning_rate": 1.7168866966487742e-05, + "loss": 0.0765, + "step": 4003 + }, + { + "epoch": 0.4340849956634866, + "grad_norm": 0.7926501035690308, + "learning_rate": 1.7168141592920354e-05, + "loss": 0.0063, + "step": 4004 + }, + { + "epoch": 0.43419340849956634, + "grad_norm": 2.282872438430786, + "learning_rate": 1.716741621935297e-05, + "loss": 0.0445, + "step": 4005 + }, + { + "epoch": 0.43430182133564615, + "grad_norm": 4.900802135467529, + "learning_rate": 1.716669084578558e-05, + "loss": 0.0135, + "step": 4006 + }, + { + "epoch": 0.4344102341717259, + "grad_norm": 1.6696202754974365, + "learning_rate": 1.7165965472218194e-05, + "loss": 0.0142, + "step": 4007 + }, + { + "epoch": 0.43451864700780574, + "grad_norm": 1.628526210784912, + "learning_rate": 1.7165240098650806e-05, + "loss": 0.0148, + "step": 4008 + }, + { + "epoch": 0.4346270598438855, + "grad_norm": 2.6679718494415283, + "learning_rate": 1.7164514725083418e-05, + "loss": 0.065, + "step": 4009 + }, + { + "epoch": 0.4347354726799653, + "grad_norm": 2.287215232849121, + "learning_rate": 1.716378935151603e-05, + "loss": 0.0457, + "step": 4010 + }, + { + "epoch": 0.4348438855160451, + "grad_norm": 6.189607620239258, + "learning_rate": 1.7163063977948646e-05, + "loss": 0.0294, + "step": 4011 + }, + { + "epoch": 0.4349522983521249, + "grad_norm": 2.9572746753692627, + "learning_rate": 1.7162338604381258e-05, + "loss": 0.0998, + "step": 4012 + }, + { + "epoch": 0.43506071118820466, + "grad_norm": 6.6702070236206055, + "learning_rate": 1.716161323081387e-05, + "loss": 0.0741, + "step": 4013 + }, + { + "epoch": 0.4351691240242845, + "grad_norm": 3.3663015365600586, + "learning_rate": 1.7160887857246486e-05, + "loss": 0.0406, + "step": 4014 + }, + { + "epoch": 0.43527753686036424, + "grad_norm": 1.3046094179153442, + "learning_rate": 1.7160162483679098e-05, + "loss": 0.0093, + "step": 4015 + }, + { + "epoch": 0.43538594969644406, + "grad_norm": 6.666154861450195, + "learning_rate": 1.715943711011171e-05, + "loss": 0.0631, + "step": 4016 + }, + { + "epoch": 0.4354943625325238, + "grad_norm": 1.8599637746810913, + "learning_rate": 1.7158711736544322e-05, + "loss": 0.0192, + "step": 4017 + }, + { + "epoch": 0.43560277536860365, + "grad_norm": 2.8197970390319824, + "learning_rate": 1.7157986362976934e-05, + "loss": 0.0416, + "step": 4018 + }, + { + "epoch": 0.4357111882046834, + "grad_norm": 1.1321306228637695, + "learning_rate": 1.7157260989409546e-05, + "loss": 0.0139, + "step": 4019 + }, + { + "epoch": 0.4358196010407632, + "grad_norm": 2.130330801010132, + "learning_rate": 1.7156535615842162e-05, + "loss": 0.0317, + "step": 4020 + }, + { + "epoch": 0.43592801387684305, + "grad_norm": 2.050781726837158, + "learning_rate": 1.7155810242274774e-05, + "loss": 0.0226, + "step": 4021 + }, + { + "epoch": 0.4360364267129228, + "grad_norm": 2.038970708847046, + "learning_rate": 1.7155084868707386e-05, + "loss": 0.0245, + "step": 4022 + }, + { + "epoch": 0.4361448395490026, + "grad_norm": 4.722787380218506, + "learning_rate": 1.7154359495139998e-05, + "loss": 0.0814, + "step": 4023 + }, + { + "epoch": 0.4362532523850824, + "grad_norm": 3.7514889240264893, + "learning_rate": 1.715363412157261e-05, + "loss": 0.0679, + "step": 4024 + }, + { + "epoch": 0.4363616652211622, + "grad_norm": 5.43030309677124, + "learning_rate": 1.7152908748005222e-05, + "loss": 0.0964, + "step": 4025 + }, + { + "epoch": 0.436470078057242, + "grad_norm": 2.0928566455841064, + "learning_rate": 1.7152183374437838e-05, + "loss": 0.0135, + "step": 4026 + }, + { + "epoch": 0.4365784908933218, + "grad_norm": 5.324298858642578, + "learning_rate": 1.715145800087045e-05, + "loss": 0.0933, + "step": 4027 + }, + { + "epoch": 0.43668690372940155, + "grad_norm": 2.894519329071045, + "learning_rate": 1.7150732627303062e-05, + "loss": 0.0727, + "step": 4028 + }, + { + "epoch": 0.4367953165654814, + "grad_norm": 4.452707290649414, + "learning_rate": 1.7150007253735674e-05, + "loss": 0.0484, + "step": 4029 + }, + { + "epoch": 0.43690372940156114, + "grad_norm": 4.838139533996582, + "learning_rate": 1.7149281880168287e-05, + "loss": 0.0622, + "step": 4030 + }, + { + "epoch": 0.43701214223764095, + "grad_norm": 7.660518169403076, + "learning_rate": 1.7148556506600902e-05, + "loss": 0.0358, + "step": 4031 + }, + { + "epoch": 0.4371205550737207, + "grad_norm": 4.992181777954102, + "learning_rate": 1.7147831133033514e-05, + "loss": 0.0798, + "step": 4032 + }, + { + "epoch": 0.43722896790980054, + "grad_norm": 4.0710062980651855, + "learning_rate": 1.7147105759466126e-05, + "loss": 0.062, + "step": 4033 + }, + { + "epoch": 0.4373373807458803, + "grad_norm": 2.1791558265686035, + "learning_rate": 1.714638038589874e-05, + "loss": 0.0443, + "step": 4034 + }, + { + "epoch": 0.4374457935819601, + "grad_norm": 2.3420119285583496, + "learning_rate": 1.7145655012331354e-05, + "loss": 0.0481, + "step": 4035 + }, + { + "epoch": 0.4375542064180399, + "grad_norm": 4.400576114654541, + "learning_rate": 1.7144929638763966e-05, + "loss": 0.0405, + "step": 4036 + }, + { + "epoch": 0.4376626192541197, + "grad_norm": 2.0066473484039307, + "learning_rate": 1.714420426519658e-05, + "loss": 0.0328, + "step": 4037 + }, + { + "epoch": 0.43777103209019946, + "grad_norm": 3.519486904144287, + "learning_rate": 1.714347889162919e-05, + "loss": 0.0687, + "step": 4038 + }, + { + "epoch": 0.4378794449262793, + "grad_norm": 3.904247283935547, + "learning_rate": 1.7142753518061803e-05, + "loss": 0.0147, + "step": 4039 + }, + { + "epoch": 0.43798785776235905, + "grad_norm": 1.942013144493103, + "learning_rate": 1.7142028144494415e-05, + "loss": 0.0498, + "step": 4040 + }, + { + "epoch": 0.43809627059843886, + "grad_norm": 1.3660271167755127, + "learning_rate": 1.7141302770927027e-05, + "loss": 0.0377, + "step": 4041 + }, + { + "epoch": 0.4382046834345186, + "grad_norm": 4.128398895263672, + "learning_rate": 1.7140577397359642e-05, + "loss": 0.0558, + "step": 4042 + }, + { + "epoch": 0.43831309627059845, + "grad_norm": 4.142878532409668, + "learning_rate": 1.7139852023792255e-05, + "loss": 0.0703, + "step": 4043 + }, + { + "epoch": 0.4384215091066782, + "grad_norm": 2.8046698570251465, + "learning_rate": 1.7139126650224867e-05, + "loss": 0.0336, + "step": 4044 + }, + { + "epoch": 0.438529921942758, + "grad_norm": 3.757976531982422, + "learning_rate": 1.713840127665748e-05, + "loss": 0.0413, + "step": 4045 + }, + { + "epoch": 0.4386383347788378, + "grad_norm": 2.27228045463562, + "learning_rate": 1.713767590309009e-05, + "loss": 0.0417, + "step": 4046 + }, + { + "epoch": 0.4387467476149176, + "grad_norm": 3.9056236743927, + "learning_rate": 1.7136950529522703e-05, + "loss": 0.0283, + "step": 4047 + }, + { + "epoch": 0.4388551604509974, + "grad_norm": 4.00669527053833, + "learning_rate": 1.713622515595532e-05, + "loss": 0.0286, + "step": 4048 + }, + { + "epoch": 0.4389635732870772, + "grad_norm": 6.9183125495910645, + "learning_rate": 1.713549978238793e-05, + "loss": 0.0825, + "step": 4049 + }, + { + "epoch": 0.43907198612315695, + "grad_norm": 2.1205718517303467, + "learning_rate": 1.7134774408820546e-05, + "loss": 0.0527, + "step": 4050 + }, + { + "epoch": 0.4391803989592368, + "grad_norm": 3.788012742996216, + "learning_rate": 1.713404903525316e-05, + "loss": 0.0681, + "step": 4051 + }, + { + "epoch": 0.4392888117953166, + "grad_norm": 0.9929600358009338, + "learning_rate": 1.713332366168577e-05, + "loss": 0.0139, + "step": 4052 + }, + { + "epoch": 0.43939722463139635, + "grad_norm": 2.820844888687134, + "learning_rate": 1.7132598288118383e-05, + "loss": 0.0211, + "step": 4053 + }, + { + "epoch": 0.4395056374674762, + "grad_norm": 2.8745718002319336, + "learning_rate": 1.7131872914550995e-05, + "loss": 0.0429, + "step": 4054 + }, + { + "epoch": 0.43961405030355594, + "grad_norm": 2.144663095474243, + "learning_rate": 1.7131147540983607e-05, + "loss": 0.0288, + "step": 4055 + }, + { + "epoch": 0.43972246313963576, + "grad_norm": 2.406298875808716, + "learning_rate": 1.713042216741622e-05, + "loss": 0.0586, + "step": 4056 + }, + { + "epoch": 0.4398308759757155, + "grad_norm": 0.79182368516922, + "learning_rate": 1.7129696793848835e-05, + "loss": 0.0114, + "step": 4057 + }, + { + "epoch": 0.43993928881179534, + "grad_norm": 1.2549716234207153, + "learning_rate": 1.7128971420281447e-05, + "loss": 0.0165, + "step": 4058 + }, + { + "epoch": 0.4400477016478751, + "grad_norm": 1.1523886919021606, + "learning_rate": 1.712824604671406e-05, + "loss": 0.0156, + "step": 4059 + }, + { + "epoch": 0.4401561144839549, + "grad_norm": 5.132842063903809, + "learning_rate": 1.712752067314667e-05, + "loss": 0.0477, + "step": 4060 + }, + { + "epoch": 0.4402645273200347, + "grad_norm": 2.9402153491973877, + "learning_rate": 1.7126795299579283e-05, + "loss": 0.0375, + "step": 4061 + }, + { + "epoch": 0.4403729401561145, + "grad_norm": 2.926300525665283, + "learning_rate": 1.7126069926011896e-05, + "loss": 0.0257, + "step": 4062 + }, + { + "epoch": 0.44048135299219426, + "grad_norm": 0.309114009141922, + "learning_rate": 1.712534455244451e-05, + "loss": 0.0044, + "step": 4063 + }, + { + "epoch": 0.4405897658282741, + "grad_norm": 2.9061427116394043, + "learning_rate": 1.7124619178877123e-05, + "loss": 0.0343, + "step": 4064 + }, + { + "epoch": 0.44069817866435385, + "grad_norm": 0.8946327567100525, + "learning_rate": 1.7123893805309735e-05, + "loss": 0.0065, + "step": 4065 + }, + { + "epoch": 0.44080659150043366, + "grad_norm": 2.0459909439086914, + "learning_rate": 1.712316843174235e-05, + "loss": 0.0382, + "step": 4066 + }, + { + "epoch": 0.4409150043365134, + "grad_norm": 2.330225706100464, + "learning_rate": 1.7122443058174963e-05, + "loss": 0.0779, + "step": 4067 + }, + { + "epoch": 0.44102341717259325, + "grad_norm": 2.796213388442993, + "learning_rate": 1.7121717684607575e-05, + "loss": 0.0373, + "step": 4068 + }, + { + "epoch": 0.441131830008673, + "grad_norm": 4.302934646606445, + "learning_rate": 1.7120992311040187e-05, + "loss": 0.0549, + "step": 4069 + }, + { + "epoch": 0.44124024284475283, + "grad_norm": 2.725295066833496, + "learning_rate": 1.71202669374728e-05, + "loss": 0.0346, + "step": 4070 + }, + { + "epoch": 0.4413486556808326, + "grad_norm": 1.3011687994003296, + "learning_rate": 1.711954156390541e-05, + "loss": 0.0174, + "step": 4071 + }, + { + "epoch": 0.4414570685169124, + "grad_norm": 1.998639702796936, + "learning_rate": 1.7118816190338027e-05, + "loss": 0.0317, + "step": 4072 + }, + { + "epoch": 0.4415654813529922, + "grad_norm": 1.4589167833328247, + "learning_rate": 1.711809081677064e-05, + "loss": 0.0094, + "step": 4073 + }, + { + "epoch": 0.441673894189072, + "grad_norm": 2.125462770462036, + "learning_rate": 1.711736544320325e-05, + "loss": 0.0178, + "step": 4074 + }, + { + "epoch": 0.44178230702515175, + "grad_norm": 5.033949851989746, + "learning_rate": 1.7116640069635864e-05, + "loss": 0.115, + "step": 4075 + }, + { + "epoch": 0.4418907198612316, + "grad_norm": 2.8041253089904785, + "learning_rate": 1.7115914696068476e-05, + "loss": 0.0483, + "step": 4076 + }, + { + "epoch": 0.44199913269731134, + "grad_norm": 3.980555772781372, + "learning_rate": 1.7115189322501088e-05, + "loss": 0.0306, + "step": 4077 + }, + { + "epoch": 0.44210754553339116, + "grad_norm": 4.059553623199463, + "learning_rate": 1.7114463948933703e-05, + "loss": 0.0658, + "step": 4078 + }, + { + "epoch": 0.4422159583694709, + "grad_norm": 1.2469793558120728, + "learning_rate": 1.7113738575366315e-05, + "loss": 0.0341, + "step": 4079 + }, + { + "epoch": 0.44232437120555074, + "grad_norm": 0.3095446527004242, + "learning_rate": 1.7113013201798928e-05, + "loss": 0.0033, + "step": 4080 + }, + { + "epoch": 0.44243278404163056, + "grad_norm": 3.15041184425354, + "learning_rate": 1.711228782823154e-05, + "loss": 0.0183, + "step": 4081 + }, + { + "epoch": 0.4425411968777103, + "grad_norm": 1.4772467613220215, + "learning_rate": 1.7111562454664152e-05, + "loss": 0.0436, + "step": 4082 + }, + { + "epoch": 0.44264960971379014, + "grad_norm": 2.112473726272583, + "learning_rate": 1.7110837081096767e-05, + "loss": 0.0161, + "step": 4083 + }, + { + "epoch": 0.4427580225498699, + "grad_norm": 3.7642276287078857, + "learning_rate": 1.711011170752938e-05, + "loss": 0.0317, + "step": 4084 + }, + { + "epoch": 0.4428664353859497, + "grad_norm": 8.500578880310059, + "learning_rate": 1.7109386333961992e-05, + "loss": 0.0311, + "step": 4085 + }, + { + "epoch": 0.4429748482220295, + "grad_norm": 11.295109748840332, + "learning_rate": 1.7108660960394604e-05, + "loss": 0.0234, + "step": 4086 + }, + { + "epoch": 0.4430832610581093, + "grad_norm": 3.3787424564361572, + "learning_rate": 1.710793558682722e-05, + "loss": 0.0494, + "step": 4087 + }, + { + "epoch": 0.44319167389418906, + "grad_norm": 1.9079936742782593, + "learning_rate": 1.710721021325983e-05, + "loss": 0.0237, + "step": 4088 + }, + { + "epoch": 0.4433000867302689, + "grad_norm": 2.3395471572875977, + "learning_rate": 1.7106484839692444e-05, + "loss": 0.0229, + "step": 4089 + }, + { + "epoch": 0.44340849956634865, + "grad_norm": 2.2901787757873535, + "learning_rate": 1.7105759466125056e-05, + "loss": 0.0385, + "step": 4090 + }, + { + "epoch": 0.44351691240242846, + "grad_norm": 1.0459911823272705, + "learning_rate": 1.7105034092557668e-05, + "loss": 0.0129, + "step": 4091 + }, + { + "epoch": 0.44362532523850823, + "grad_norm": 4.234433174133301, + "learning_rate": 1.710430871899028e-05, + "loss": 0.0951, + "step": 4092 + }, + { + "epoch": 0.44373373807458805, + "grad_norm": 4.8916215896606445, + "learning_rate": 1.7103583345422892e-05, + "loss": 0.1032, + "step": 4093 + }, + { + "epoch": 0.4438421509106678, + "grad_norm": 5.477846622467041, + "learning_rate": 1.7102857971855508e-05, + "loss": 0.0396, + "step": 4094 + }, + { + "epoch": 0.44395056374674763, + "grad_norm": 3.52354097366333, + "learning_rate": 1.710213259828812e-05, + "loss": 0.025, + "step": 4095 + }, + { + "epoch": 0.4440589765828274, + "grad_norm": 2.4781086444854736, + "learning_rate": 1.7101407224720732e-05, + "loss": 0.0492, + "step": 4096 + }, + { + "epoch": 0.4441673894189072, + "grad_norm": 2.183182716369629, + "learning_rate": 1.7100681851153344e-05, + "loss": 0.0172, + "step": 4097 + }, + { + "epoch": 0.444275802254987, + "grad_norm": 6.065431118011475, + "learning_rate": 1.7099956477585956e-05, + "loss": 0.0373, + "step": 4098 + }, + { + "epoch": 0.4443842150910668, + "grad_norm": 6.141956806182861, + "learning_rate": 1.709923110401857e-05, + "loss": 0.1058, + "step": 4099 + }, + { + "epoch": 0.44449262792714656, + "grad_norm": 1.3509268760681152, + "learning_rate": 1.7098505730451184e-05, + "loss": 0.0228, + "step": 4100 + }, + { + "epoch": 0.4446010407632264, + "grad_norm": 2.2263994216918945, + "learning_rate": 1.7097780356883796e-05, + "loss": 0.0282, + "step": 4101 + }, + { + "epoch": 0.44470945359930614, + "grad_norm": 2.5614731311798096, + "learning_rate": 1.7097054983316412e-05, + "loss": 0.0426, + "step": 4102 + }, + { + "epoch": 0.44481786643538596, + "grad_norm": 1.486000657081604, + "learning_rate": 1.7096329609749024e-05, + "loss": 0.0228, + "step": 4103 + }, + { + "epoch": 0.4449262792714657, + "grad_norm": 1.367527723312378, + "learning_rate": 1.7095604236181636e-05, + "loss": 0.0178, + "step": 4104 + }, + { + "epoch": 0.44503469210754554, + "grad_norm": 3.20393443107605, + "learning_rate": 1.7094878862614248e-05, + "loss": 0.0339, + "step": 4105 + }, + { + "epoch": 0.4451431049436253, + "grad_norm": 2.069373846054077, + "learning_rate": 1.709415348904686e-05, + "loss": 0.0152, + "step": 4106 + }, + { + "epoch": 0.4452515177797051, + "grad_norm": 2.2355308532714844, + "learning_rate": 1.7093428115479472e-05, + "loss": 0.0146, + "step": 4107 + }, + { + "epoch": 0.4453599306157849, + "grad_norm": 4.307426929473877, + "learning_rate": 1.7092702741912085e-05, + "loss": 0.0473, + "step": 4108 + }, + { + "epoch": 0.4454683434518647, + "grad_norm": 5.278342247009277, + "learning_rate": 1.70919773683447e-05, + "loss": 0.0508, + "step": 4109 + }, + { + "epoch": 0.4455767562879445, + "grad_norm": 3.0360524654388428, + "learning_rate": 1.7091251994777312e-05, + "loss": 0.0416, + "step": 4110 + }, + { + "epoch": 0.4456851691240243, + "grad_norm": 1.8327243328094482, + "learning_rate": 1.7090526621209924e-05, + "loss": 0.0268, + "step": 4111 + }, + { + "epoch": 0.4457935819601041, + "grad_norm": 2.656586170196533, + "learning_rate": 1.7089801247642537e-05, + "loss": 0.0171, + "step": 4112 + }, + { + "epoch": 0.44590199479618386, + "grad_norm": 3.820396661758423, + "learning_rate": 1.708907587407515e-05, + "loss": 0.0336, + "step": 4113 + }, + { + "epoch": 0.4460104076322637, + "grad_norm": 4.658166408538818, + "learning_rate": 1.708835050050776e-05, + "loss": 0.0222, + "step": 4114 + }, + { + "epoch": 0.44611882046834345, + "grad_norm": 1.6628166437149048, + "learning_rate": 1.7087625126940376e-05, + "loss": 0.0273, + "step": 4115 + }, + { + "epoch": 0.44622723330442327, + "grad_norm": 3.5934507846832275, + "learning_rate": 1.708689975337299e-05, + "loss": 0.0448, + "step": 4116 + }, + { + "epoch": 0.44633564614050303, + "grad_norm": 6.429623126983643, + "learning_rate": 1.70861743798056e-05, + "loss": 0.0826, + "step": 4117 + }, + { + "epoch": 0.44644405897658285, + "grad_norm": 1.6874083280563354, + "learning_rate": 1.7085449006238216e-05, + "loss": 0.0131, + "step": 4118 + }, + { + "epoch": 0.4465524718126626, + "grad_norm": 2.5012245178222656, + "learning_rate": 1.708472363267083e-05, + "loss": 0.0437, + "step": 4119 + }, + { + "epoch": 0.44666088464874243, + "grad_norm": 1.643362283706665, + "learning_rate": 1.708399825910344e-05, + "loss": 0.0148, + "step": 4120 + }, + { + "epoch": 0.4467692974848222, + "grad_norm": 4.095820426940918, + "learning_rate": 1.7083272885536053e-05, + "loss": 0.0674, + "step": 4121 + }, + { + "epoch": 0.446877710320902, + "grad_norm": 0.3252134919166565, + "learning_rate": 1.7082547511968665e-05, + "loss": 0.0043, + "step": 4122 + }, + { + "epoch": 0.4469861231569818, + "grad_norm": 1.405486822128296, + "learning_rate": 1.7081822138401277e-05, + "loss": 0.0117, + "step": 4123 + }, + { + "epoch": 0.4470945359930616, + "grad_norm": 3.500778913497925, + "learning_rate": 1.7081096764833892e-05, + "loss": 0.0282, + "step": 4124 + }, + { + "epoch": 0.44720294882914136, + "grad_norm": 3.551691770553589, + "learning_rate": 1.7080371391266505e-05, + "loss": 0.0347, + "step": 4125 + }, + { + "epoch": 0.4473113616652212, + "grad_norm": 2.0705015659332275, + "learning_rate": 1.7079646017699117e-05, + "loss": 0.0462, + "step": 4126 + }, + { + "epoch": 0.44741977450130094, + "grad_norm": 3.393246650695801, + "learning_rate": 1.707892064413173e-05, + "loss": 0.0455, + "step": 4127 + }, + { + "epoch": 0.44752818733738076, + "grad_norm": 2.690186023712158, + "learning_rate": 1.707819527056434e-05, + "loss": 0.0347, + "step": 4128 + }, + { + "epoch": 0.4476366001734605, + "grad_norm": 2.5017433166503906, + "learning_rate": 1.7077469896996953e-05, + "loss": 0.0602, + "step": 4129 + }, + { + "epoch": 0.44774501300954034, + "grad_norm": 3.783139228820801, + "learning_rate": 1.707674452342957e-05, + "loss": 0.0688, + "step": 4130 + }, + { + "epoch": 0.4478534258456201, + "grad_norm": 2.7775120735168457, + "learning_rate": 1.707601914986218e-05, + "loss": 0.0241, + "step": 4131 + }, + { + "epoch": 0.4479618386816999, + "grad_norm": 3.8614766597747803, + "learning_rate": 1.7075293776294793e-05, + "loss": 0.0396, + "step": 4132 + }, + { + "epoch": 0.4480702515177797, + "grad_norm": 2.147796869277954, + "learning_rate": 1.7074568402727405e-05, + "loss": 0.0218, + "step": 4133 + }, + { + "epoch": 0.4481786643538595, + "grad_norm": 5.816850662231445, + "learning_rate": 1.707384302916002e-05, + "loss": 0.0512, + "step": 4134 + }, + { + "epoch": 0.44828707718993926, + "grad_norm": 2.3005552291870117, + "learning_rate": 1.7073117655592633e-05, + "loss": 0.0435, + "step": 4135 + }, + { + "epoch": 0.4483954900260191, + "grad_norm": 3.48996639251709, + "learning_rate": 1.7072392282025245e-05, + "loss": 0.069, + "step": 4136 + }, + { + "epoch": 0.44850390286209885, + "grad_norm": 5.485073566436768, + "learning_rate": 1.7071666908457857e-05, + "loss": 0.1115, + "step": 4137 + }, + { + "epoch": 0.44861231569817867, + "grad_norm": 3.08377742767334, + "learning_rate": 1.707094153489047e-05, + "loss": 0.0342, + "step": 4138 + }, + { + "epoch": 0.4487207285342585, + "grad_norm": 1.6700599193572998, + "learning_rate": 1.7070216161323085e-05, + "loss": 0.0291, + "step": 4139 + }, + { + "epoch": 0.44882914137033825, + "grad_norm": 1.9040182828903198, + "learning_rate": 1.7069490787755697e-05, + "loss": 0.031, + "step": 4140 + }, + { + "epoch": 0.44893755420641807, + "grad_norm": 1.8848278522491455, + "learning_rate": 1.706876541418831e-05, + "loss": 0.0337, + "step": 4141 + }, + { + "epoch": 0.44904596704249783, + "grad_norm": 5.297695159912109, + "learning_rate": 1.706804004062092e-05, + "loss": 0.1251, + "step": 4142 + }, + { + "epoch": 0.44915437987857765, + "grad_norm": 3.175151824951172, + "learning_rate": 1.7067314667053533e-05, + "loss": 0.0444, + "step": 4143 + }, + { + "epoch": 0.4492627927146574, + "grad_norm": 5.082479953765869, + "learning_rate": 1.7066589293486145e-05, + "loss": 0.0754, + "step": 4144 + }, + { + "epoch": 0.44937120555073723, + "grad_norm": 1.9357634782791138, + "learning_rate": 1.7065863919918758e-05, + "loss": 0.0464, + "step": 4145 + }, + { + "epoch": 0.449479618386817, + "grad_norm": 7.137246131896973, + "learning_rate": 1.7065138546351373e-05, + "loss": 0.109, + "step": 4146 + }, + { + "epoch": 0.4495880312228968, + "grad_norm": 1.827762246131897, + "learning_rate": 1.7064413172783985e-05, + "loss": 0.0354, + "step": 4147 + }, + { + "epoch": 0.4496964440589766, + "grad_norm": 2.085258960723877, + "learning_rate": 1.7063687799216597e-05, + "loss": 0.0248, + "step": 4148 + }, + { + "epoch": 0.4498048568950564, + "grad_norm": 2.7500765323638916, + "learning_rate": 1.706296242564921e-05, + "loss": 0.0501, + "step": 4149 + }, + { + "epoch": 0.44991326973113616, + "grad_norm": 2.4326207637786865, + "learning_rate": 1.7062237052081822e-05, + "loss": 0.0291, + "step": 4150 + }, + { + "epoch": 0.450021682567216, + "grad_norm": 2.9958221912384033, + "learning_rate": 1.7061511678514434e-05, + "loss": 0.0583, + "step": 4151 + }, + { + "epoch": 0.45013009540329574, + "grad_norm": 2.0536322593688965, + "learning_rate": 1.706078630494705e-05, + "loss": 0.025, + "step": 4152 + }, + { + "epoch": 0.45023850823937556, + "grad_norm": 4.320369720458984, + "learning_rate": 1.706006093137966e-05, + "loss": 0.0513, + "step": 4153 + }, + { + "epoch": 0.4503469210754553, + "grad_norm": 2.2392830848693848, + "learning_rate": 1.7059335557812277e-05, + "loss": 0.0348, + "step": 4154 + }, + { + "epoch": 0.45045533391153514, + "grad_norm": 2.296297311782837, + "learning_rate": 1.705861018424489e-05, + "loss": 0.0626, + "step": 4155 + }, + { + "epoch": 0.4505637467476149, + "grad_norm": 2.345141887664795, + "learning_rate": 1.70578848106775e-05, + "loss": 0.046, + "step": 4156 + }, + { + "epoch": 0.4506721595836947, + "grad_norm": 2.2939016819000244, + "learning_rate": 1.7057159437110114e-05, + "loss": 0.0739, + "step": 4157 + }, + { + "epoch": 0.4507805724197745, + "grad_norm": 0.7529460787773132, + "learning_rate": 1.7056434063542726e-05, + "loss": 0.0136, + "step": 4158 + }, + { + "epoch": 0.4508889852558543, + "grad_norm": 1.7343189716339111, + "learning_rate": 1.7055708689975338e-05, + "loss": 0.0394, + "step": 4159 + }, + { + "epoch": 0.45099739809193407, + "grad_norm": 0.9408934116363525, + "learning_rate": 1.705498331640795e-05, + "loss": 0.0174, + "step": 4160 + }, + { + "epoch": 0.4511058109280139, + "grad_norm": 5.72635555267334, + "learning_rate": 1.7054257942840565e-05, + "loss": 0.058, + "step": 4161 + }, + { + "epoch": 0.45121422376409365, + "grad_norm": 1.0900076627731323, + "learning_rate": 1.7053532569273178e-05, + "loss": 0.0148, + "step": 4162 + }, + { + "epoch": 0.45132263660017347, + "grad_norm": 5.7204413414001465, + "learning_rate": 1.705280719570579e-05, + "loss": 0.0387, + "step": 4163 + }, + { + "epoch": 0.45143104943625323, + "grad_norm": 1.82813560962677, + "learning_rate": 1.7052081822138402e-05, + "loss": 0.0359, + "step": 4164 + }, + { + "epoch": 0.45153946227233305, + "grad_norm": 4.866833209991455, + "learning_rate": 1.7051356448571014e-05, + "loss": 0.0535, + "step": 4165 + }, + { + "epoch": 0.4516478751084128, + "grad_norm": 2.5442721843719482, + "learning_rate": 1.7050631075003626e-05, + "loss": 0.0673, + "step": 4166 + }, + { + "epoch": 0.45175628794449263, + "grad_norm": 2.2466700077056885, + "learning_rate": 1.7049905701436242e-05, + "loss": 0.026, + "step": 4167 + }, + { + "epoch": 0.4518647007805724, + "grad_norm": 2.6875600814819336, + "learning_rate": 1.7049180327868854e-05, + "loss": 0.0738, + "step": 4168 + }, + { + "epoch": 0.4519731136166522, + "grad_norm": 2.986294984817505, + "learning_rate": 1.7048454954301466e-05, + "loss": 0.0388, + "step": 4169 + }, + { + "epoch": 0.45208152645273203, + "grad_norm": 2.150193929672241, + "learning_rate": 1.704772958073408e-05, + "loss": 0.0406, + "step": 4170 + }, + { + "epoch": 0.4521899392888118, + "grad_norm": 7.072264671325684, + "learning_rate": 1.7047004207166694e-05, + "loss": 0.078, + "step": 4171 + }, + { + "epoch": 0.4522983521248916, + "grad_norm": 1.321550965309143, + "learning_rate": 1.7046278833599306e-05, + "loss": 0.0264, + "step": 4172 + }, + { + "epoch": 0.4524067649609714, + "grad_norm": 3.487621307373047, + "learning_rate": 1.7045553460031918e-05, + "loss": 0.058, + "step": 4173 + }, + { + "epoch": 0.4525151777970512, + "grad_norm": 4.023947238922119, + "learning_rate": 1.704482808646453e-05, + "loss": 0.1072, + "step": 4174 + }, + { + "epoch": 0.45262359063313096, + "grad_norm": 2.0360524654388428, + "learning_rate": 1.7044102712897142e-05, + "loss": 0.0315, + "step": 4175 + }, + { + "epoch": 0.4527320034692108, + "grad_norm": 2.742703914642334, + "learning_rate": 1.7043377339329758e-05, + "loss": 0.0606, + "step": 4176 + }, + { + "epoch": 0.45284041630529054, + "grad_norm": 4.003826141357422, + "learning_rate": 1.704265196576237e-05, + "loss": 0.0657, + "step": 4177 + }, + { + "epoch": 0.45294882914137036, + "grad_norm": 1.827617883682251, + "learning_rate": 1.7041926592194982e-05, + "loss": 0.0157, + "step": 4178 + }, + { + "epoch": 0.4530572419774501, + "grad_norm": 2.1661269664764404, + "learning_rate": 1.7041201218627594e-05, + "loss": 0.0804, + "step": 4179 + }, + { + "epoch": 0.45316565481352994, + "grad_norm": 4.462652683258057, + "learning_rate": 1.7040475845060206e-05, + "loss": 0.1031, + "step": 4180 + }, + { + "epoch": 0.4532740676496097, + "grad_norm": 1.6856728792190552, + "learning_rate": 1.703975047149282e-05, + "loss": 0.0377, + "step": 4181 + }, + { + "epoch": 0.4533824804856895, + "grad_norm": 2.5578339099884033, + "learning_rate": 1.7039025097925434e-05, + "loss": 0.0363, + "step": 4182 + }, + { + "epoch": 0.4534908933217693, + "grad_norm": 0.3631773889064789, + "learning_rate": 1.7038299724358046e-05, + "loss": 0.0044, + "step": 4183 + }, + { + "epoch": 0.4535993061578491, + "grad_norm": 3.4300084114074707, + "learning_rate": 1.703757435079066e-05, + "loss": 0.0209, + "step": 4184 + }, + { + "epoch": 0.45370771899392887, + "grad_norm": 1.5256630182266235, + "learning_rate": 1.703684897722327e-05, + "loss": 0.0207, + "step": 4185 + }, + { + "epoch": 0.4538161318300087, + "grad_norm": 4.01906156539917, + "learning_rate": 1.7036123603655886e-05, + "loss": 0.0412, + "step": 4186 + }, + { + "epoch": 0.45392454466608845, + "grad_norm": 0.9316742420196533, + "learning_rate": 1.7035398230088498e-05, + "loss": 0.016, + "step": 4187 + }, + { + "epoch": 0.45403295750216827, + "grad_norm": 2.5232484340667725, + "learning_rate": 1.703467285652111e-05, + "loss": 0.033, + "step": 4188 + }, + { + "epoch": 0.45414137033824803, + "grad_norm": 2.43583345413208, + "learning_rate": 1.7033947482953722e-05, + "loss": 0.0196, + "step": 4189 + }, + { + "epoch": 0.45424978317432785, + "grad_norm": 3.9479167461395264, + "learning_rate": 1.7033222109386335e-05, + "loss": 0.0497, + "step": 4190 + }, + { + "epoch": 0.4543581960104076, + "grad_norm": 8.859834671020508, + "learning_rate": 1.703249673581895e-05, + "loss": 0.0618, + "step": 4191 + }, + { + "epoch": 0.45446660884648743, + "grad_norm": 3.0675103664398193, + "learning_rate": 1.7031771362251562e-05, + "loss": 0.0365, + "step": 4192 + }, + { + "epoch": 0.4545750216825672, + "grad_norm": 0.43305832147598267, + "learning_rate": 1.7031045988684174e-05, + "loss": 0.0037, + "step": 4193 + }, + { + "epoch": 0.454683434518647, + "grad_norm": 6.573017597198486, + "learning_rate": 1.7030320615116787e-05, + "loss": 0.0666, + "step": 4194 + }, + { + "epoch": 0.4547918473547268, + "grad_norm": 3.570399045944214, + "learning_rate": 1.70295952415494e-05, + "loss": 0.0814, + "step": 4195 + }, + { + "epoch": 0.4549002601908066, + "grad_norm": 3.552053451538086, + "learning_rate": 1.702886986798201e-05, + "loss": 0.0493, + "step": 4196 + }, + { + "epoch": 0.45500867302688636, + "grad_norm": 4.5995025634765625, + "learning_rate": 1.7028144494414623e-05, + "loss": 0.0588, + "step": 4197 + }, + { + "epoch": 0.4551170858629662, + "grad_norm": 1.043447732925415, + "learning_rate": 1.702741912084724e-05, + "loss": 0.0139, + "step": 4198 + }, + { + "epoch": 0.455225498699046, + "grad_norm": 1.8671075105667114, + "learning_rate": 1.702669374727985e-05, + "loss": 0.0241, + "step": 4199 + }, + { + "epoch": 0.45533391153512576, + "grad_norm": 2.5313706398010254, + "learning_rate": 1.7025968373712463e-05, + "loss": 0.0079, + "step": 4200 + }, + { + "epoch": 0.4554423243712056, + "grad_norm": 6.32307767868042, + "learning_rate": 1.7025243000145075e-05, + "loss": 0.0453, + "step": 4201 + }, + { + "epoch": 0.45555073720728534, + "grad_norm": 4.445122718811035, + "learning_rate": 1.7024517626577687e-05, + "loss": 0.0527, + "step": 4202 + }, + { + "epoch": 0.45565915004336516, + "grad_norm": 0.8731651902198792, + "learning_rate": 1.7023792253010303e-05, + "loss": 0.0109, + "step": 4203 + }, + { + "epoch": 0.4557675628794449, + "grad_norm": 8.746415138244629, + "learning_rate": 1.7023066879442915e-05, + "loss": 0.1182, + "step": 4204 + }, + { + "epoch": 0.45587597571552474, + "grad_norm": 1.589820384979248, + "learning_rate": 1.7022341505875527e-05, + "loss": 0.021, + "step": 4205 + }, + { + "epoch": 0.4559843885516045, + "grad_norm": 2.8358445167541504, + "learning_rate": 1.7021616132308142e-05, + "loss": 0.0242, + "step": 4206 + }, + { + "epoch": 0.4560928013876843, + "grad_norm": 1.8188822269439697, + "learning_rate": 1.7020890758740755e-05, + "loss": 0.0284, + "step": 4207 + }, + { + "epoch": 0.4562012142237641, + "grad_norm": 5.237939834594727, + "learning_rate": 1.7020165385173367e-05, + "loss": 0.0379, + "step": 4208 + }, + { + "epoch": 0.4563096270598439, + "grad_norm": 6.165538787841797, + "learning_rate": 1.701944001160598e-05, + "loss": 0.0502, + "step": 4209 + }, + { + "epoch": 0.45641803989592367, + "grad_norm": 2.3427300453186035, + "learning_rate": 1.701871463803859e-05, + "loss": 0.0156, + "step": 4210 + }, + { + "epoch": 0.4565264527320035, + "grad_norm": 1.5270684957504272, + "learning_rate": 1.7017989264471203e-05, + "loss": 0.043, + "step": 4211 + }, + { + "epoch": 0.45663486556808325, + "grad_norm": 0.6696369051933289, + "learning_rate": 1.7017263890903815e-05, + "loss": 0.0071, + "step": 4212 + }, + { + "epoch": 0.45674327840416307, + "grad_norm": 3.9871599674224854, + "learning_rate": 1.701653851733643e-05, + "loss": 0.1002, + "step": 4213 + }, + { + "epoch": 0.45685169124024283, + "grad_norm": 4.949384689331055, + "learning_rate": 1.7015813143769043e-05, + "loss": 0.0193, + "step": 4214 + }, + { + "epoch": 0.45696010407632265, + "grad_norm": 5.260954856872559, + "learning_rate": 1.7015087770201655e-05, + "loss": 0.0304, + "step": 4215 + }, + { + "epoch": 0.4570685169124024, + "grad_norm": 3.0198910236358643, + "learning_rate": 1.7014362396634267e-05, + "loss": 0.0519, + "step": 4216 + }, + { + "epoch": 0.45717692974848223, + "grad_norm": 3.0372071266174316, + "learning_rate": 1.701363702306688e-05, + "loss": 0.0297, + "step": 4217 + }, + { + "epoch": 0.457285342584562, + "grad_norm": 3.5217249393463135, + "learning_rate": 1.701291164949949e-05, + "loss": 0.088, + "step": 4218 + }, + { + "epoch": 0.4573937554206418, + "grad_norm": 2.9213027954101562, + "learning_rate": 1.7012186275932107e-05, + "loss": 0.0207, + "step": 4219 + }, + { + "epoch": 0.4575021682567216, + "grad_norm": 4.657233715057373, + "learning_rate": 1.701146090236472e-05, + "loss": 0.0711, + "step": 4220 + }, + { + "epoch": 0.4576105810928014, + "grad_norm": 4.901986598968506, + "learning_rate": 1.7010735528797335e-05, + "loss": 0.0818, + "step": 4221 + }, + { + "epoch": 0.45771899392888116, + "grad_norm": 3.8745734691619873, + "learning_rate": 1.7010010155229947e-05, + "loss": 0.0443, + "step": 4222 + }, + { + "epoch": 0.457827406764961, + "grad_norm": 4.092204570770264, + "learning_rate": 1.700928478166256e-05, + "loss": 0.026, + "step": 4223 + }, + { + "epoch": 0.45793581960104074, + "grad_norm": 5.274410247802734, + "learning_rate": 1.700855940809517e-05, + "loss": 0.0669, + "step": 4224 + }, + { + "epoch": 0.45804423243712056, + "grad_norm": 1.8599529266357422, + "learning_rate": 1.7007834034527783e-05, + "loss": 0.0354, + "step": 4225 + }, + { + "epoch": 0.4581526452732003, + "grad_norm": 3.6568005084991455, + "learning_rate": 1.7007108660960395e-05, + "loss": 0.0571, + "step": 4226 + }, + { + "epoch": 0.45826105810928014, + "grad_norm": 2.085369348526001, + "learning_rate": 1.7006383287393008e-05, + "loss": 0.0247, + "step": 4227 + }, + { + "epoch": 0.45836947094535996, + "grad_norm": 1.9224334955215454, + "learning_rate": 1.7005657913825623e-05, + "loss": 0.02, + "step": 4228 + }, + { + "epoch": 0.4584778837814397, + "grad_norm": 1.4509849548339844, + "learning_rate": 1.7004932540258235e-05, + "loss": 0.0233, + "step": 4229 + }, + { + "epoch": 0.45858629661751954, + "grad_norm": 2.1786930561065674, + "learning_rate": 1.7004207166690847e-05, + "loss": 0.038, + "step": 4230 + }, + { + "epoch": 0.4586947094535993, + "grad_norm": 3.707845449447632, + "learning_rate": 1.700348179312346e-05, + "loss": 0.0455, + "step": 4231 + }, + { + "epoch": 0.4588031222896791, + "grad_norm": 2.064483642578125, + "learning_rate": 1.700275641955607e-05, + "loss": 0.0288, + "step": 4232 + }, + { + "epoch": 0.4589115351257589, + "grad_norm": 1.4372726678848267, + "learning_rate": 1.7002031045988684e-05, + "loss": 0.019, + "step": 4233 + }, + { + "epoch": 0.4590199479618387, + "grad_norm": 4.039114952087402, + "learning_rate": 1.70013056724213e-05, + "loss": 0.0531, + "step": 4234 + }, + { + "epoch": 0.45912836079791847, + "grad_norm": 2.110121965408325, + "learning_rate": 1.700058029885391e-05, + "loss": 0.0366, + "step": 4235 + }, + { + "epoch": 0.4592367736339983, + "grad_norm": 3.400256633758545, + "learning_rate": 1.6999854925286524e-05, + "loss": 0.0671, + "step": 4236 + }, + { + "epoch": 0.45934518647007805, + "grad_norm": 2.028193235397339, + "learning_rate": 1.6999129551719136e-05, + "loss": 0.0314, + "step": 4237 + }, + { + "epoch": 0.45945359930615787, + "grad_norm": 2.8120813369750977, + "learning_rate": 1.699840417815175e-05, + "loss": 0.061, + "step": 4238 + }, + { + "epoch": 0.45956201214223763, + "grad_norm": 3.731675386428833, + "learning_rate": 1.6997678804584363e-05, + "loss": 0.0945, + "step": 4239 + }, + { + "epoch": 0.45967042497831745, + "grad_norm": 2.6047523021698, + "learning_rate": 1.6996953431016976e-05, + "loss": 0.0309, + "step": 4240 + }, + { + "epoch": 0.4597788378143972, + "grad_norm": 4.941725730895996, + "learning_rate": 1.6996228057449588e-05, + "loss": 0.0341, + "step": 4241 + }, + { + "epoch": 0.45988725065047703, + "grad_norm": 0.6802063584327698, + "learning_rate": 1.69955026838822e-05, + "loss": 0.0062, + "step": 4242 + }, + { + "epoch": 0.4599956634865568, + "grad_norm": 3.769732713699341, + "learning_rate": 1.6994777310314815e-05, + "loss": 0.0632, + "step": 4243 + }, + { + "epoch": 0.4601040763226366, + "grad_norm": 2.8162527084350586, + "learning_rate": 1.6994051936747428e-05, + "loss": 0.0426, + "step": 4244 + }, + { + "epoch": 0.4602124891587164, + "grad_norm": 4.675906181335449, + "learning_rate": 1.699332656318004e-05, + "loss": 0.0731, + "step": 4245 + }, + { + "epoch": 0.4603209019947962, + "grad_norm": 1.2596189975738525, + "learning_rate": 1.6992601189612652e-05, + "loss": 0.0175, + "step": 4246 + }, + { + "epoch": 0.46042931483087596, + "grad_norm": 2.325641393661499, + "learning_rate": 1.6991875816045264e-05, + "loss": 0.0481, + "step": 4247 + }, + { + "epoch": 0.4605377276669558, + "grad_norm": 3.488723039627075, + "learning_rate": 1.6991150442477876e-05, + "loss": 0.0258, + "step": 4248 + }, + { + "epoch": 0.46064614050303554, + "grad_norm": 0.8574606776237488, + "learning_rate": 1.6990425068910488e-05, + "loss": 0.0101, + "step": 4249 + }, + { + "epoch": 0.46075455333911536, + "grad_norm": 2.2129905223846436, + "learning_rate": 1.6989699695343104e-05, + "loss": 0.0299, + "step": 4250 + }, + { + "epoch": 0.4608629661751951, + "grad_norm": 3.7632710933685303, + "learning_rate": 1.6988974321775716e-05, + "loss": 0.0947, + "step": 4251 + }, + { + "epoch": 0.46097137901127494, + "grad_norm": 5.356408596038818, + "learning_rate": 1.6988248948208328e-05, + "loss": 0.025, + "step": 4252 + }, + { + "epoch": 0.4610797918473547, + "grad_norm": 5.321102142333984, + "learning_rate": 1.698752357464094e-05, + "loss": 0.0353, + "step": 4253 + }, + { + "epoch": 0.4611882046834345, + "grad_norm": 3.663944959640503, + "learning_rate": 1.6986798201073552e-05, + "loss": 0.0798, + "step": 4254 + }, + { + "epoch": 0.4612966175195143, + "grad_norm": 2.9526703357696533, + "learning_rate": 1.6986072827506168e-05, + "loss": 0.0483, + "step": 4255 + }, + { + "epoch": 0.4614050303555941, + "grad_norm": 0.641257643699646, + "learning_rate": 1.698534745393878e-05, + "loss": 0.0076, + "step": 4256 + }, + { + "epoch": 0.46151344319167387, + "grad_norm": 1.7786264419555664, + "learning_rate": 1.6984622080371392e-05, + "loss": 0.0218, + "step": 4257 + }, + { + "epoch": 0.4616218560277537, + "grad_norm": 1.5984435081481934, + "learning_rate": 1.6983896706804008e-05, + "loss": 0.0117, + "step": 4258 + }, + { + "epoch": 0.4617302688638335, + "grad_norm": 2.55964994430542, + "learning_rate": 1.698317133323662e-05, + "loss": 0.0525, + "step": 4259 + }, + { + "epoch": 0.46183868169991327, + "grad_norm": 3.4328248500823975, + "learning_rate": 1.6982445959669232e-05, + "loss": 0.0704, + "step": 4260 + }, + { + "epoch": 0.4619470945359931, + "grad_norm": 2.5485005378723145, + "learning_rate": 1.6981720586101844e-05, + "loss": 0.0696, + "step": 4261 + }, + { + "epoch": 0.46205550737207285, + "grad_norm": 3.9510080814361572, + "learning_rate": 1.6980995212534456e-05, + "loss": 0.1191, + "step": 4262 + }, + { + "epoch": 0.46216392020815267, + "grad_norm": 3.2836523056030273, + "learning_rate": 1.698026983896707e-05, + "loss": 0.0526, + "step": 4263 + }, + { + "epoch": 0.46227233304423243, + "grad_norm": 2.3197896480560303, + "learning_rate": 1.697954446539968e-05, + "loss": 0.0317, + "step": 4264 + }, + { + "epoch": 0.46238074588031225, + "grad_norm": 3.896162986755371, + "learning_rate": 1.6978819091832296e-05, + "loss": 0.0851, + "step": 4265 + }, + { + "epoch": 0.462489158716392, + "grad_norm": 3.039846181869507, + "learning_rate": 1.6978093718264908e-05, + "loss": 0.0571, + "step": 4266 + }, + { + "epoch": 0.46259757155247183, + "grad_norm": 2.3864035606384277, + "learning_rate": 1.697736834469752e-05, + "loss": 0.0274, + "step": 4267 + }, + { + "epoch": 0.4627059843885516, + "grad_norm": 5.344044208526611, + "learning_rate": 1.6976642971130133e-05, + "loss": 0.0469, + "step": 4268 + }, + { + "epoch": 0.4628143972246314, + "grad_norm": 2.485077381134033, + "learning_rate": 1.6975917597562745e-05, + "loss": 0.0508, + "step": 4269 + }, + { + "epoch": 0.4629228100607112, + "grad_norm": 4.03607702255249, + "learning_rate": 1.6975192223995357e-05, + "loss": 0.0796, + "step": 4270 + }, + { + "epoch": 0.463031222896791, + "grad_norm": 6.275140285491943, + "learning_rate": 1.6974466850427972e-05, + "loss": 0.0289, + "step": 4271 + }, + { + "epoch": 0.46313963573287076, + "grad_norm": 2.5866713523864746, + "learning_rate": 1.6973741476860585e-05, + "loss": 0.0261, + "step": 4272 + }, + { + "epoch": 0.4632480485689506, + "grad_norm": 2.6919376850128174, + "learning_rate": 1.69730161032932e-05, + "loss": 0.0347, + "step": 4273 + }, + { + "epoch": 0.46335646140503034, + "grad_norm": 3.0526154041290283, + "learning_rate": 1.6972290729725812e-05, + "loss": 0.0252, + "step": 4274 + }, + { + "epoch": 0.46346487424111016, + "grad_norm": 1.028113842010498, + "learning_rate": 1.6971565356158424e-05, + "loss": 0.0132, + "step": 4275 + }, + { + "epoch": 0.4635732870771899, + "grad_norm": 0.6612816452980042, + "learning_rate": 1.6970839982591036e-05, + "loss": 0.0072, + "step": 4276 + }, + { + "epoch": 0.46368169991326974, + "grad_norm": 3.869230270385742, + "learning_rate": 1.697011460902365e-05, + "loss": 0.1041, + "step": 4277 + }, + { + "epoch": 0.4637901127493495, + "grad_norm": 3.093059778213501, + "learning_rate": 1.696938923545626e-05, + "loss": 0.0357, + "step": 4278 + }, + { + "epoch": 0.4638985255854293, + "grad_norm": 3.523890256881714, + "learning_rate": 1.6968663861888873e-05, + "loss": 0.0373, + "step": 4279 + }, + { + "epoch": 0.4640069384215091, + "grad_norm": 3.1529040336608887, + "learning_rate": 1.696793848832149e-05, + "loss": 0.0232, + "step": 4280 + }, + { + "epoch": 0.4641153512575889, + "grad_norm": 2.9701437950134277, + "learning_rate": 1.69672131147541e-05, + "loss": 0.0458, + "step": 4281 + }, + { + "epoch": 0.46422376409366867, + "grad_norm": 1.6359665393829346, + "learning_rate": 1.6966487741186713e-05, + "loss": 0.0149, + "step": 4282 + }, + { + "epoch": 0.4643321769297485, + "grad_norm": 0.9231381416320801, + "learning_rate": 1.6965762367619325e-05, + "loss": 0.0176, + "step": 4283 + }, + { + "epoch": 0.46444058976582825, + "grad_norm": 3.7438957691192627, + "learning_rate": 1.6965036994051937e-05, + "loss": 0.0567, + "step": 4284 + }, + { + "epoch": 0.46454900260190807, + "grad_norm": 2.53254771232605, + "learning_rate": 1.696431162048455e-05, + "loss": 0.0527, + "step": 4285 + }, + { + "epoch": 0.46465741543798783, + "grad_norm": 1.5640249252319336, + "learning_rate": 1.6963586246917165e-05, + "loss": 0.0255, + "step": 4286 + }, + { + "epoch": 0.46476582827406765, + "grad_norm": 5.1468939781188965, + "learning_rate": 1.6962860873349777e-05, + "loss": 0.0706, + "step": 4287 + }, + { + "epoch": 0.46487424111014747, + "grad_norm": 5.210297107696533, + "learning_rate": 1.696213549978239e-05, + "loss": 0.0942, + "step": 4288 + }, + { + "epoch": 0.46498265394622723, + "grad_norm": 6.507889747619629, + "learning_rate": 1.6961410126215e-05, + "loss": 0.0634, + "step": 4289 + }, + { + "epoch": 0.46509106678230705, + "grad_norm": 5.694176197052002, + "learning_rate": 1.6960684752647617e-05, + "loss": 0.0507, + "step": 4290 + }, + { + "epoch": 0.4651994796183868, + "grad_norm": 2.893846035003662, + "learning_rate": 1.695995937908023e-05, + "loss": 0.0289, + "step": 4291 + }, + { + "epoch": 0.46530789245446663, + "grad_norm": 4.201303958892822, + "learning_rate": 1.695923400551284e-05, + "loss": 0.0416, + "step": 4292 + }, + { + "epoch": 0.4654163052905464, + "grad_norm": 2.517230272293091, + "learning_rate": 1.6958508631945453e-05, + "loss": 0.0512, + "step": 4293 + }, + { + "epoch": 0.4655247181266262, + "grad_norm": 5.096835613250732, + "learning_rate": 1.6957783258378065e-05, + "loss": 0.0253, + "step": 4294 + }, + { + "epoch": 0.465633130962706, + "grad_norm": 3.3652334213256836, + "learning_rate": 1.695705788481068e-05, + "loss": 0.0467, + "step": 4295 + }, + { + "epoch": 0.4657415437987858, + "grad_norm": 1.717220425605774, + "learning_rate": 1.6956332511243293e-05, + "loss": 0.0344, + "step": 4296 + }, + { + "epoch": 0.46584995663486556, + "grad_norm": 3.2674574851989746, + "learning_rate": 1.6955607137675905e-05, + "loss": 0.0312, + "step": 4297 + }, + { + "epoch": 0.4659583694709454, + "grad_norm": 2.174971342086792, + "learning_rate": 1.6954881764108517e-05, + "loss": 0.018, + "step": 4298 + }, + { + "epoch": 0.46606678230702514, + "grad_norm": 3.230926990509033, + "learning_rate": 1.695415639054113e-05, + "loss": 0.0763, + "step": 4299 + }, + { + "epoch": 0.46617519514310496, + "grad_norm": 1.1368712186813354, + "learning_rate": 1.695343101697374e-05, + "loss": 0.0143, + "step": 4300 + }, + { + "epoch": 0.4662836079791847, + "grad_norm": 3.134610176086426, + "learning_rate": 1.6952705643406354e-05, + "loss": 0.0299, + "step": 4301 + }, + { + "epoch": 0.46639202081526454, + "grad_norm": 3.0689244270324707, + "learning_rate": 1.695198026983897e-05, + "loss": 0.0509, + "step": 4302 + }, + { + "epoch": 0.4665004336513443, + "grad_norm": 2.8242480754852295, + "learning_rate": 1.695125489627158e-05, + "loss": 0.0501, + "step": 4303 + }, + { + "epoch": 0.4666088464874241, + "grad_norm": 3.766636371612549, + "learning_rate": 1.6950529522704193e-05, + "loss": 0.0719, + "step": 4304 + }, + { + "epoch": 0.4667172593235039, + "grad_norm": 4.1143975257873535, + "learning_rate": 1.6949804149136806e-05, + "loss": 0.0532, + "step": 4305 + }, + { + "epoch": 0.4668256721595837, + "grad_norm": 2.5796053409576416, + "learning_rate": 1.6949078775569418e-05, + "loss": 0.0292, + "step": 4306 + }, + { + "epoch": 0.46693408499566347, + "grad_norm": 5.162386894226074, + "learning_rate": 1.6948353402002033e-05, + "loss": 0.0477, + "step": 4307 + }, + { + "epoch": 0.4670424978317433, + "grad_norm": 2.8503715991973877, + "learning_rate": 1.6947628028434645e-05, + "loss": 0.0696, + "step": 4308 + }, + { + "epoch": 0.46715091066782305, + "grad_norm": 4.893138885498047, + "learning_rate": 1.6946902654867258e-05, + "loss": 0.0369, + "step": 4309 + }, + { + "epoch": 0.46725932350390287, + "grad_norm": 2.1872975826263428, + "learning_rate": 1.6946177281299873e-05, + "loss": 0.0255, + "step": 4310 + }, + { + "epoch": 0.46736773633998263, + "grad_norm": 1.3545210361480713, + "learning_rate": 1.6945451907732485e-05, + "loss": 0.0289, + "step": 4311 + }, + { + "epoch": 0.46747614917606245, + "grad_norm": 1.6032614707946777, + "learning_rate": 1.6944726534165097e-05, + "loss": 0.0138, + "step": 4312 + }, + { + "epoch": 0.4675845620121422, + "grad_norm": 2.98032283782959, + "learning_rate": 1.694400116059771e-05, + "loss": 0.0578, + "step": 4313 + }, + { + "epoch": 0.46769297484822203, + "grad_norm": 2.5393154621124268, + "learning_rate": 1.694327578703032e-05, + "loss": 0.0284, + "step": 4314 + }, + { + "epoch": 0.4678013876843018, + "grad_norm": 1.2768982648849487, + "learning_rate": 1.6942550413462934e-05, + "loss": 0.0191, + "step": 4315 + }, + { + "epoch": 0.4679098005203816, + "grad_norm": 1.7075440883636475, + "learning_rate": 1.6941825039895546e-05, + "loss": 0.0447, + "step": 4316 + }, + { + "epoch": 0.46801821335646143, + "grad_norm": 3.4647579193115234, + "learning_rate": 1.694109966632816e-05, + "loss": 0.0279, + "step": 4317 + }, + { + "epoch": 0.4681266261925412, + "grad_norm": 4.16401481628418, + "learning_rate": 1.6940374292760774e-05, + "loss": 0.0352, + "step": 4318 + }, + { + "epoch": 0.468235039028621, + "grad_norm": 1.1437814235687256, + "learning_rate": 1.6939648919193386e-05, + "loss": 0.012, + "step": 4319 + }, + { + "epoch": 0.4683434518647008, + "grad_norm": 2.633981227874756, + "learning_rate": 1.6938923545625998e-05, + "loss": 0.0291, + "step": 4320 + }, + { + "epoch": 0.4684518647007806, + "grad_norm": 3.4663288593292236, + "learning_rate": 1.693819817205861e-05, + "loss": 0.0643, + "step": 4321 + }, + { + "epoch": 0.46856027753686036, + "grad_norm": 5.343357086181641, + "learning_rate": 1.6937472798491222e-05, + "loss": 0.0924, + "step": 4322 + }, + { + "epoch": 0.4686686903729402, + "grad_norm": 2.834704637527466, + "learning_rate": 1.6936747424923838e-05, + "loss": 0.0349, + "step": 4323 + }, + { + "epoch": 0.46877710320901994, + "grad_norm": 2.988981008529663, + "learning_rate": 1.693602205135645e-05, + "loss": 0.044, + "step": 4324 + }, + { + "epoch": 0.46888551604509976, + "grad_norm": 3.4611849784851074, + "learning_rate": 1.6935296677789065e-05, + "loss": 0.0204, + "step": 4325 + }, + { + "epoch": 0.4689939288811795, + "grad_norm": 0.9032403230667114, + "learning_rate": 1.6934571304221678e-05, + "loss": 0.0079, + "step": 4326 + }, + { + "epoch": 0.46910234171725934, + "grad_norm": 2.26774525642395, + "learning_rate": 1.693384593065429e-05, + "loss": 0.0314, + "step": 4327 + }, + { + "epoch": 0.4692107545533391, + "grad_norm": 3.8860373497009277, + "learning_rate": 1.6933120557086902e-05, + "loss": 0.0866, + "step": 4328 + }, + { + "epoch": 0.4693191673894189, + "grad_norm": 1.44521164894104, + "learning_rate": 1.6932395183519514e-05, + "loss": 0.0159, + "step": 4329 + }, + { + "epoch": 0.4694275802254987, + "grad_norm": 3.1960883140563965, + "learning_rate": 1.6931669809952126e-05, + "loss": 0.0406, + "step": 4330 + }, + { + "epoch": 0.4695359930615785, + "grad_norm": 1.8500912189483643, + "learning_rate": 1.6930944436384738e-05, + "loss": 0.0447, + "step": 4331 + }, + { + "epoch": 0.46964440589765827, + "grad_norm": 4.276816368103027, + "learning_rate": 1.6930219062817354e-05, + "loss": 0.0547, + "step": 4332 + }, + { + "epoch": 0.4697528187337381, + "grad_norm": 4.602271556854248, + "learning_rate": 1.6929493689249966e-05, + "loss": 0.0693, + "step": 4333 + }, + { + "epoch": 0.46986123156981785, + "grad_norm": 2.145765781402588, + "learning_rate": 1.6928768315682578e-05, + "loss": 0.0258, + "step": 4334 + }, + { + "epoch": 0.46996964440589767, + "grad_norm": 2.767078399658203, + "learning_rate": 1.692804294211519e-05, + "loss": 0.0406, + "step": 4335 + }, + { + "epoch": 0.47007805724197743, + "grad_norm": 5.693793296813965, + "learning_rate": 1.6927317568547802e-05, + "loss": 0.0515, + "step": 4336 + }, + { + "epoch": 0.47018647007805725, + "grad_norm": 3.6417131423950195, + "learning_rate": 1.6926592194980415e-05, + "loss": 0.0461, + "step": 4337 + }, + { + "epoch": 0.470294882914137, + "grad_norm": 1.7193666696548462, + "learning_rate": 1.692586682141303e-05, + "loss": 0.042, + "step": 4338 + }, + { + "epoch": 0.47040329575021683, + "grad_norm": 5.18550443649292, + "learning_rate": 1.6925141447845642e-05, + "loss": 0.079, + "step": 4339 + }, + { + "epoch": 0.4705117085862966, + "grad_norm": 3.1794028282165527, + "learning_rate": 1.6924416074278254e-05, + "loss": 0.0732, + "step": 4340 + }, + { + "epoch": 0.4706201214223764, + "grad_norm": 2.905557155609131, + "learning_rate": 1.6923690700710866e-05, + "loss": 0.0436, + "step": 4341 + }, + { + "epoch": 0.4707285342584562, + "grad_norm": 3.512287139892578, + "learning_rate": 1.6922965327143482e-05, + "loss": 0.028, + "step": 4342 + }, + { + "epoch": 0.470836947094536, + "grad_norm": 0.946077287197113, + "learning_rate": 1.6922239953576094e-05, + "loss": 0.0171, + "step": 4343 + }, + { + "epoch": 0.47094535993061576, + "grad_norm": 4.840882778167725, + "learning_rate": 1.6921514580008706e-05, + "loss": 0.05, + "step": 4344 + }, + { + "epoch": 0.4710537727666956, + "grad_norm": 0.6406776905059814, + "learning_rate": 1.692078920644132e-05, + "loss": 0.0097, + "step": 4345 + }, + { + "epoch": 0.4711621856027754, + "grad_norm": 1.8568412065505981, + "learning_rate": 1.692006383287393e-05, + "loss": 0.0176, + "step": 4346 + }, + { + "epoch": 0.47127059843885516, + "grad_norm": 2.3042805194854736, + "learning_rate": 1.6919338459306546e-05, + "loss": 0.0406, + "step": 4347 + }, + { + "epoch": 0.471379011274935, + "grad_norm": 1.526668906211853, + "learning_rate": 1.6918613085739158e-05, + "loss": 0.032, + "step": 4348 + }, + { + "epoch": 0.47148742411101474, + "grad_norm": 3.161378860473633, + "learning_rate": 1.691788771217177e-05, + "loss": 0.0774, + "step": 4349 + }, + { + "epoch": 0.47159583694709456, + "grad_norm": 7.393195629119873, + "learning_rate": 1.6917162338604383e-05, + "loss": 0.0599, + "step": 4350 + }, + { + "epoch": 0.4717042497831743, + "grad_norm": 7.562840461730957, + "learning_rate": 1.6916436965036995e-05, + "loss": 0.0362, + "step": 4351 + }, + { + "epoch": 0.47181266261925414, + "grad_norm": 4.406378269195557, + "learning_rate": 1.6915711591469607e-05, + "loss": 0.0397, + "step": 4352 + }, + { + "epoch": 0.4719210754553339, + "grad_norm": 2.8092668056488037, + "learning_rate": 1.691498621790222e-05, + "loss": 0.0422, + "step": 4353 + }, + { + "epoch": 0.4720294882914137, + "grad_norm": 7.082157611846924, + "learning_rate": 1.6914260844334834e-05, + "loss": 0.0238, + "step": 4354 + }, + { + "epoch": 0.4721379011274935, + "grad_norm": 3.9569506645202637, + "learning_rate": 1.6913535470767447e-05, + "loss": 0.0481, + "step": 4355 + }, + { + "epoch": 0.4722463139635733, + "grad_norm": 3.2734909057617188, + "learning_rate": 1.691281009720006e-05, + "loss": 0.026, + "step": 4356 + }, + { + "epoch": 0.47235472679965307, + "grad_norm": 2.043815851211548, + "learning_rate": 1.691208472363267e-05, + "loss": 0.0216, + "step": 4357 + }, + { + "epoch": 0.4724631396357329, + "grad_norm": 3.939946413040161, + "learning_rate": 1.6911359350065283e-05, + "loss": 0.0614, + "step": 4358 + }, + { + "epoch": 0.47257155247181265, + "grad_norm": 3.787721872329712, + "learning_rate": 1.69106339764979e-05, + "loss": 0.0514, + "step": 4359 + }, + { + "epoch": 0.47267996530789247, + "grad_norm": 3.1683568954467773, + "learning_rate": 1.690990860293051e-05, + "loss": 0.0456, + "step": 4360 + }, + { + "epoch": 0.47278837814397223, + "grad_norm": 2.5365517139434814, + "learning_rate": 1.6909183229363123e-05, + "loss": 0.0528, + "step": 4361 + }, + { + "epoch": 0.47289679098005205, + "grad_norm": 3.41381573677063, + "learning_rate": 1.690845785579574e-05, + "loss": 0.0694, + "step": 4362 + }, + { + "epoch": 0.4730052038161318, + "grad_norm": 4.555696487426758, + "learning_rate": 1.690773248222835e-05, + "loss": 0.0395, + "step": 4363 + }, + { + "epoch": 0.47311361665221163, + "grad_norm": 2.1256892681121826, + "learning_rate": 1.6907007108660963e-05, + "loss": 0.0119, + "step": 4364 + }, + { + "epoch": 0.4732220294882914, + "grad_norm": 3.6081607341766357, + "learning_rate": 1.6906281735093575e-05, + "loss": 0.0234, + "step": 4365 + }, + { + "epoch": 0.4733304423243712, + "grad_norm": 3.1012721061706543, + "learning_rate": 1.6905556361526187e-05, + "loss": 0.0386, + "step": 4366 + }, + { + "epoch": 0.473438855160451, + "grad_norm": 1.7716882228851318, + "learning_rate": 1.69048309879588e-05, + "loss": 0.0328, + "step": 4367 + }, + { + "epoch": 0.4735472679965308, + "grad_norm": 4.203100681304932, + "learning_rate": 1.690410561439141e-05, + "loss": 0.0534, + "step": 4368 + }, + { + "epoch": 0.47365568083261056, + "grad_norm": 2.308743715286255, + "learning_rate": 1.6903380240824027e-05, + "loss": 0.0469, + "step": 4369 + }, + { + "epoch": 0.4737640936686904, + "grad_norm": 1.7806602716445923, + "learning_rate": 1.690265486725664e-05, + "loss": 0.0216, + "step": 4370 + }, + { + "epoch": 0.47387250650477014, + "grad_norm": 4.655505657196045, + "learning_rate": 1.690192949368925e-05, + "loss": 0.039, + "step": 4371 + }, + { + "epoch": 0.47398091934084996, + "grad_norm": 4.780037879943848, + "learning_rate": 1.6901204120121863e-05, + "loss": 0.0402, + "step": 4372 + }, + { + "epoch": 0.4740893321769297, + "grad_norm": 3.6377928256988525, + "learning_rate": 1.6900478746554475e-05, + "loss": 0.0771, + "step": 4373 + }, + { + "epoch": 0.47419774501300954, + "grad_norm": 2.2029178142547607, + "learning_rate": 1.6899753372987088e-05, + "loss": 0.0485, + "step": 4374 + }, + { + "epoch": 0.4743061578490893, + "grad_norm": 4.358334064483643, + "learning_rate": 1.6899027999419703e-05, + "loss": 0.0822, + "step": 4375 + }, + { + "epoch": 0.4744145706851691, + "grad_norm": 5.129088401794434, + "learning_rate": 1.6898302625852315e-05, + "loss": 0.0541, + "step": 4376 + }, + { + "epoch": 0.47452298352124894, + "grad_norm": 1.4390177726745605, + "learning_rate": 1.689757725228493e-05, + "loss": 0.0195, + "step": 4377 + }, + { + "epoch": 0.4746313963573287, + "grad_norm": 0.8997501134872437, + "learning_rate": 1.6896851878717543e-05, + "loss": 0.0155, + "step": 4378 + }, + { + "epoch": 0.4747398091934085, + "grad_norm": 4.585682392120361, + "learning_rate": 1.6896126505150155e-05, + "loss": 0.0665, + "step": 4379 + }, + { + "epoch": 0.4748482220294883, + "grad_norm": 0.9135122895240784, + "learning_rate": 1.6895401131582767e-05, + "loss": 0.0137, + "step": 4380 + }, + { + "epoch": 0.4749566348655681, + "grad_norm": 2.9928243160247803, + "learning_rate": 1.689467575801538e-05, + "loss": 0.0351, + "step": 4381 + }, + { + "epoch": 0.47506504770164787, + "grad_norm": 6.992832660675049, + "learning_rate": 1.689395038444799e-05, + "loss": 0.0551, + "step": 4382 + }, + { + "epoch": 0.4751734605377277, + "grad_norm": 1.428786039352417, + "learning_rate": 1.6893225010880604e-05, + "loss": 0.0344, + "step": 4383 + }, + { + "epoch": 0.47528187337380745, + "grad_norm": 4.04830265045166, + "learning_rate": 1.689249963731322e-05, + "loss": 0.1041, + "step": 4384 + }, + { + "epoch": 0.47539028620988727, + "grad_norm": 3.179222583770752, + "learning_rate": 1.689177426374583e-05, + "loss": 0.0529, + "step": 4385 + }, + { + "epoch": 0.47549869904596703, + "grad_norm": 3.321143388748169, + "learning_rate": 1.6891048890178443e-05, + "loss": 0.0851, + "step": 4386 + }, + { + "epoch": 0.47560711188204685, + "grad_norm": 1.118914246559143, + "learning_rate": 1.6890323516611056e-05, + "loss": 0.0176, + "step": 4387 + }, + { + "epoch": 0.4757155247181266, + "grad_norm": 3.074643611907959, + "learning_rate": 1.6889598143043668e-05, + "loss": 0.0419, + "step": 4388 + }, + { + "epoch": 0.47582393755420643, + "grad_norm": 2.2187561988830566, + "learning_rate": 1.688887276947628e-05, + "loss": 0.0598, + "step": 4389 + }, + { + "epoch": 0.4759323503902862, + "grad_norm": 4.51185941696167, + "learning_rate": 1.6888147395908895e-05, + "loss": 0.044, + "step": 4390 + }, + { + "epoch": 0.476040763226366, + "grad_norm": 2.6908552646636963, + "learning_rate": 1.6887422022341508e-05, + "loss": 0.0741, + "step": 4391 + }, + { + "epoch": 0.4761491760624458, + "grad_norm": 2.9782776832580566, + "learning_rate": 1.688669664877412e-05, + "loss": 0.0427, + "step": 4392 + }, + { + "epoch": 0.4762575888985256, + "grad_norm": 5.926055908203125, + "learning_rate": 1.6885971275206732e-05, + "loss": 0.0733, + "step": 4393 + }, + { + "epoch": 0.47636600173460536, + "grad_norm": 0.752280056476593, + "learning_rate": 1.6885245901639347e-05, + "loss": 0.0122, + "step": 4394 + }, + { + "epoch": 0.4764744145706852, + "grad_norm": 4.047726631164551, + "learning_rate": 1.688452052807196e-05, + "loss": 0.0345, + "step": 4395 + }, + { + "epoch": 0.47658282740676494, + "grad_norm": 2.6570746898651123, + "learning_rate": 1.688379515450457e-05, + "loss": 0.0909, + "step": 4396 + }, + { + "epoch": 0.47669124024284476, + "grad_norm": 1.7361584901809692, + "learning_rate": 1.6883069780937184e-05, + "loss": 0.0182, + "step": 4397 + }, + { + "epoch": 0.4767996530789245, + "grad_norm": 0.7551249265670776, + "learning_rate": 1.6882344407369796e-05, + "loss": 0.0079, + "step": 4398 + }, + { + "epoch": 0.47690806591500434, + "grad_norm": 4.213406085968018, + "learning_rate": 1.688161903380241e-05, + "loss": 0.0418, + "step": 4399 + }, + { + "epoch": 0.4770164787510841, + "grad_norm": 1.9855026006698608, + "learning_rate": 1.6880893660235024e-05, + "loss": 0.0764, + "step": 4400 + }, + { + "epoch": 0.4771248915871639, + "grad_norm": 1.9997233152389526, + "learning_rate": 1.6880168286667636e-05, + "loss": 0.0299, + "step": 4401 + }, + { + "epoch": 0.4772333044232437, + "grad_norm": 2.8076279163360596, + "learning_rate": 1.6879442913100248e-05, + "loss": 0.0601, + "step": 4402 + }, + { + "epoch": 0.4773417172593235, + "grad_norm": 2.249474048614502, + "learning_rate": 1.687871753953286e-05, + "loss": 0.0399, + "step": 4403 + }, + { + "epoch": 0.47745013009540327, + "grad_norm": 1.8566690683364868, + "learning_rate": 1.6877992165965472e-05, + "loss": 0.0491, + "step": 4404 + }, + { + "epoch": 0.4775585429314831, + "grad_norm": 6.1595330238342285, + "learning_rate": 1.6877266792398084e-05, + "loss": 0.0363, + "step": 4405 + }, + { + "epoch": 0.4776669557675629, + "grad_norm": 3.092238426208496, + "learning_rate": 1.68765414188307e-05, + "loss": 0.1188, + "step": 4406 + }, + { + "epoch": 0.47777536860364267, + "grad_norm": 3.4351868629455566, + "learning_rate": 1.6875816045263312e-05, + "loss": 0.0316, + "step": 4407 + }, + { + "epoch": 0.4778837814397225, + "grad_norm": 1.3832749128341675, + "learning_rate": 1.6875090671695924e-05, + "loss": 0.0482, + "step": 4408 + }, + { + "epoch": 0.47799219427580225, + "grad_norm": 2.437690496444702, + "learning_rate": 1.6874365298128536e-05, + "loss": 0.0346, + "step": 4409 + }, + { + "epoch": 0.47810060711188207, + "grad_norm": 1.040818691253662, + "learning_rate": 1.687363992456115e-05, + "loss": 0.0177, + "step": 4410 + }, + { + "epoch": 0.47820901994796183, + "grad_norm": 2.3809053897857666, + "learning_rate": 1.6872914550993764e-05, + "loss": 0.0516, + "step": 4411 + }, + { + "epoch": 0.47831743278404165, + "grad_norm": 1.655861258506775, + "learning_rate": 1.6872189177426376e-05, + "loss": 0.0283, + "step": 4412 + }, + { + "epoch": 0.4784258456201214, + "grad_norm": 5.5497636795043945, + "learning_rate": 1.6871463803858988e-05, + "loss": 0.1201, + "step": 4413 + }, + { + "epoch": 0.47853425845620123, + "grad_norm": 2.7910964488983154, + "learning_rate": 1.6870738430291604e-05, + "loss": 0.0525, + "step": 4414 + }, + { + "epoch": 0.478642671292281, + "grad_norm": 2.9252564907073975, + "learning_rate": 1.6870013056724216e-05, + "loss": 0.0542, + "step": 4415 + }, + { + "epoch": 0.4787510841283608, + "grad_norm": 5.566764831542969, + "learning_rate": 1.6869287683156828e-05, + "loss": 0.0324, + "step": 4416 + }, + { + "epoch": 0.4788594969644406, + "grad_norm": 1.7783591747283936, + "learning_rate": 1.686856230958944e-05, + "loss": 0.0201, + "step": 4417 + }, + { + "epoch": 0.4789679098005204, + "grad_norm": 2.508617877960205, + "learning_rate": 1.6867836936022052e-05, + "loss": 0.0541, + "step": 4418 + }, + { + "epoch": 0.47907632263660016, + "grad_norm": 3.427119255065918, + "learning_rate": 1.6867111562454664e-05, + "loss": 0.0687, + "step": 4419 + }, + { + "epoch": 0.47918473547268, + "grad_norm": 2.627427816390991, + "learning_rate": 1.6866386188887277e-05, + "loss": 0.0349, + "step": 4420 + }, + { + "epoch": 0.47929314830875974, + "grad_norm": 2.7571868896484375, + "learning_rate": 1.6865660815319892e-05, + "loss": 0.0426, + "step": 4421 + }, + { + "epoch": 0.47940156114483956, + "grad_norm": 0.5685505867004395, + "learning_rate": 1.6864935441752504e-05, + "loss": 0.0125, + "step": 4422 + }, + { + "epoch": 0.4795099739809193, + "grad_norm": 6.45375919342041, + "learning_rate": 1.6864210068185116e-05, + "loss": 0.0571, + "step": 4423 + }, + { + "epoch": 0.47961838681699914, + "grad_norm": 0.8808764815330505, + "learning_rate": 1.686348469461773e-05, + "loss": 0.0187, + "step": 4424 + }, + { + "epoch": 0.4797267996530789, + "grad_norm": 2.0193419456481934, + "learning_rate": 1.686275932105034e-05, + "loss": 0.0455, + "step": 4425 + }, + { + "epoch": 0.4798352124891587, + "grad_norm": 4.0348801612854, + "learning_rate": 1.6862033947482953e-05, + "loss": 0.056, + "step": 4426 + }, + { + "epoch": 0.4799436253252385, + "grad_norm": 0.9012851715087891, + "learning_rate": 1.686130857391557e-05, + "loss": 0.0197, + "step": 4427 + }, + { + "epoch": 0.4800520381613183, + "grad_norm": 2.876640796661377, + "learning_rate": 1.686058320034818e-05, + "loss": 0.0289, + "step": 4428 + }, + { + "epoch": 0.48016045099739807, + "grad_norm": 3.470029830932617, + "learning_rate": 1.6859857826780796e-05, + "loss": 0.0805, + "step": 4429 + }, + { + "epoch": 0.4802688638334779, + "grad_norm": 0.9233266115188599, + "learning_rate": 1.6859132453213408e-05, + "loss": 0.0111, + "step": 4430 + }, + { + "epoch": 0.48037727666955765, + "grad_norm": 2.4102365970611572, + "learning_rate": 1.685840707964602e-05, + "loss": 0.041, + "step": 4431 + }, + { + "epoch": 0.48048568950563747, + "grad_norm": 1.408799409866333, + "learning_rate": 1.6857681706078633e-05, + "loss": 0.0133, + "step": 4432 + }, + { + "epoch": 0.48059410234171723, + "grad_norm": 0.522782027721405, + "learning_rate": 1.6856956332511245e-05, + "loss": 0.005, + "step": 4433 + }, + { + "epoch": 0.48070251517779705, + "grad_norm": 3.339466094970703, + "learning_rate": 1.6856230958943857e-05, + "loss": 0.0877, + "step": 4434 + }, + { + "epoch": 0.48081092801387687, + "grad_norm": 2.181450128555298, + "learning_rate": 1.685550558537647e-05, + "loss": 0.0228, + "step": 4435 + }, + { + "epoch": 0.48091934084995663, + "grad_norm": 4.476340293884277, + "learning_rate": 1.6854780211809084e-05, + "loss": 0.0685, + "step": 4436 + }, + { + "epoch": 0.48102775368603645, + "grad_norm": 2.7658097743988037, + "learning_rate": 1.6854054838241697e-05, + "loss": 0.0396, + "step": 4437 + }, + { + "epoch": 0.4811361665221162, + "grad_norm": 4.6996331214904785, + "learning_rate": 1.685332946467431e-05, + "loss": 0.0473, + "step": 4438 + }, + { + "epoch": 0.48124457935819603, + "grad_norm": 2.1858644485473633, + "learning_rate": 1.685260409110692e-05, + "loss": 0.0248, + "step": 4439 + }, + { + "epoch": 0.4813529921942758, + "grad_norm": 6.704414367675781, + "learning_rate": 1.6851878717539533e-05, + "loss": 0.0307, + "step": 4440 + }, + { + "epoch": 0.4814614050303556, + "grad_norm": 7.310563087463379, + "learning_rate": 1.6851153343972145e-05, + "loss": 0.0121, + "step": 4441 + }, + { + "epoch": 0.4815698178664354, + "grad_norm": 5.294905662536621, + "learning_rate": 1.685042797040476e-05, + "loss": 0.0801, + "step": 4442 + }, + { + "epoch": 0.4816782307025152, + "grad_norm": 6.455778121948242, + "learning_rate": 1.6849702596837373e-05, + "loss": 0.0938, + "step": 4443 + }, + { + "epoch": 0.48178664353859496, + "grad_norm": 2.9388465881347656, + "learning_rate": 1.6848977223269985e-05, + "loss": 0.0256, + "step": 4444 + }, + { + "epoch": 0.4818950563746748, + "grad_norm": 1.4668891429901123, + "learning_rate": 1.68482518497026e-05, + "loss": 0.0169, + "step": 4445 + }, + { + "epoch": 0.48200346921075454, + "grad_norm": 2.2386043071746826, + "learning_rate": 1.6847526476135213e-05, + "loss": 0.0101, + "step": 4446 + }, + { + "epoch": 0.48211188204683436, + "grad_norm": 8.174972534179688, + "learning_rate": 1.6846801102567825e-05, + "loss": 0.0691, + "step": 4447 + }, + { + "epoch": 0.4822202948829141, + "grad_norm": 4.8256306648254395, + "learning_rate": 1.6846075729000437e-05, + "loss": 0.0672, + "step": 4448 + }, + { + "epoch": 0.48232870771899394, + "grad_norm": 0.9600998759269714, + "learning_rate": 1.684535035543305e-05, + "loss": 0.0148, + "step": 4449 + }, + { + "epoch": 0.4824371205550737, + "grad_norm": 2.376845598220825, + "learning_rate": 1.684462498186566e-05, + "loss": 0.0243, + "step": 4450 + }, + { + "epoch": 0.4825455333911535, + "grad_norm": 4.904407501220703, + "learning_rate": 1.6843899608298277e-05, + "loss": 0.065, + "step": 4451 + }, + { + "epoch": 0.4826539462272333, + "grad_norm": 5.709595203399658, + "learning_rate": 1.684317423473089e-05, + "loss": 0.0452, + "step": 4452 + }, + { + "epoch": 0.4827623590633131, + "grad_norm": 3.0947165489196777, + "learning_rate": 1.68424488611635e-05, + "loss": 0.0543, + "step": 4453 + }, + { + "epoch": 0.48287077189939287, + "grad_norm": 6.084900379180908, + "learning_rate": 1.6841723487596113e-05, + "loss": 0.0475, + "step": 4454 + }, + { + "epoch": 0.4829791847354727, + "grad_norm": 3.257803201675415, + "learning_rate": 1.6840998114028725e-05, + "loss": 0.0265, + "step": 4455 + }, + { + "epoch": 0.48308759757155245, + "grad_norm": 6.342535972595215, + "learning_rate": 1.6840272740461337e-05, + "loss": 0.0978, + "step": 4456 + }, + { + "epoch": 0.48319601040763227, + "grad_norm": 2.6996445655822754, + "learning_rate": 1.683954736689395e-05, + "loss": 0.0354, + "step": 4457 + }, + { + "epoch": 0.48330442324371203, + "grad_norm": 3.609541654586792, + "learning_rate": 1.6838821993326565e-05, + "loss": 0.0275, + "step": 4458 + }, + { + "epoch": 0.48341283607979185, + "grad_norm": 6.662419319152832, + "learning_rate": 1.6838096619759177e-05, + "loss": 0.0657, + "step": 4459 + }, + { + "epoch": 0.4835212489158716, + "grad_norm": 6.526035308837891, + "learning_rate": 1.683737124619179e-05, + "loss": 0.094, + "step": 4460 + }, + { + "epoch": 0.48362966175195143, + "grad_norm": 2.8946073055267334, + "learning_rate": 1.68366458726244e-05, + "loss": 0.046, + "step": 4461 + }, + { + "epoch": 0.4837380745880312, + "grad_norm": 2.6609954833984375, + "learning_rate": 1.6835920499057014e-05, + "loss": 0.07, + "step": 4462 + }, + { + "epoch": 0.483846487424111, + "grad_norm": 2.3733906745910645, + "learning_rate": 1.683519512548963e-05, + "loss": 0.0753, + "step": 4463 + }, + { + "epoch": 0.4839549002601908, + "grad_norm": 3.526045560836792, + "learning_rate": 1.683446975192224e-05, + "loss": 0.0605, + "step": 4464 + }, + { + "epoch": 0.4840633130962706, + "grad_norm": 3.5730795860290527, + "learning_rate": 1.6833744378354854e-05, + "loss": 0.0124, + "step": 4465 + }, + { + "epoch": 0.4841717259323504, + "grad_norm": 1.945346474647522, + "learning_rate": 1.683301900478747e-05, + "loss": 0.037, + "step": 4466 + }, + { + "epoch": 0.4842801387684302, + "grad_norm": 1.3449627161026, + "learning_rate": 1.683229363122008e-05, + "loss": 0.04, + "step": 4467 + }, + { + "epoch": 0.48438855160451, + "grad_norm": 4.514261722564697, + "learning_rate": 1.6831568257652693e-05, + "loss": 0.0589, + "step": 4468 + }, + { + "epoch": 0.48449696444058976, + "grad_norm": 1.4813706874847412, + "learning_rate": 1.6830842884085306e-05, + "loss": 0.0305, + "step": 4469 + }, + { + "epoch": 0.4846053772766696, + "grad_norm": 2.5010759830474854, + "learning_rate": 1.6830117510517918e-05, + "loss": 0.0519, + "step": 4470 + }, + { + "epoch": 0.48471379011274934, + "grad_norm": 2.8209714889526367, + "learning_rate": 1.682939213695053e-05, + "loss": 0.0212, + "step": 4471 + }, + { + "epoch": 0.48482220294882916, + "grad_norm": 1.2538570165634155, + "learning_rate": 1.6828666763383142e-05, + "loss": 0.0511, + "step": 4472 + }, + { + "epoch": 0.4849306157849089, + "grad_norm": 2.064746379852295, + "learning_rate": 1.6827941389815757e-05, + "loss": 0.0415, + "step": 4473 + }, + { + "epoch": 0.48503902862098874, + "grad_norm": 2.55513858795166, + "learning_rate": 1.682721601624837e-05, + "loss": 0.0222, + "step": 4474 + }, + { + "epoch": 0.4851474414570685, + "grad_norm": 3.6458499431610107, + "learning_rate": 1.6826490642680982e-05, + "loss": 0.0657, + "step": 4475 + }, + { + "epoch": 0.4852558542931483, + "grad_norm": 1.732184886932373, + "learning_rate": 1.6825765269113594e-05, + "loss": 0.0353, + "step": 4476 + }, + { + "epoch": 0.4853642671292281, + "grad_norm": 4.038102626800537, + "learning_rate": 1.6825039895546206e-05, + "loss": 0.0432, + "step": 4477 + }, + { + "epoch": 0.4854726799653079, + "grad_norm": 2.7947146892547607, + "learning_rate": 1.6824314521978818e-05, + "loss": 0.0366, + "step": 4478 + }, + { + "epoch": 0.48558109280138767, + "grad_norm": 0.8075164556503296, + "learning_rate": 1.6823589148411434e-05, + "loss": 0.0081, + "step": 4479 + }, + { + "epoch": 0.4856895056374675, + "grad_norm": 1.4122402667999268, + "learning_rate": 1.6822863774844046e-05, + "loss": 0.033, + "step": 4480 + }, + { + "epoch": 0.48579791847354725, + "grad_norm": 1.3936089277267456, + "learning_rate": 1.682213840127666e-05, + "loss": 0.0341, + "step": 4481 + }, + { + "epoch": 0.48590633130962707, + "grad_norm": 1.9714431762695312, + "learning_rate": 1.6821413027709274e-05, + "loss": 0.0316, + "step": 4482 + }, + { + "epoch": 0.48601474414570683, + "grad_norm": 4.034191131591797, + "learning_rate": 1.6820687654141886e-05, + "loss": 0.0388, + "step": 4483 + }, + { + "epoch": 0.48612315698178665, + "grad_norm": 1.0849988460540771, + "learning_rate": 1.6819962280574498e-05, + "loss": 0.0201, + "step": 4484 + }, + { + "epoch": 0.4862315698178664, + "grad_norm": 1.2838038206100464, + "learning_rate": 1.681923690700711e-05, + "loss": 0.0406, + "step": 4485 + }, + { + "epoch": 0.48633998265394623, + "grad_norm": 4.689419746398926, + "learning_rate": 1.6818511533439722e-05, + "loss": 0.0355, + "step": 4486 + }, + { + "epoch": 0.486448395490026, + "grad_norm": 4.339396953582764, + "learning_rate": 1.6817786159872334e-05, + "loss": 0.0434, + "step": 4487 + }, + { + "epoch": 0.4865568083261058, + "grad_norm": 1.1133184432983398, + "learning_rate": 1.681706078630495e-05, + "loss": 0.0091, + "step": 4488 + }, + { + "epoch": 0.4866652211621856, + "grad_norm": 3.5586040019989014, + "learning_rate": 1.6816335412737562e-05, + "loss": 0.06, + "step": 4489 + }, + { + "epoch": 0.4867736339982654, + "grad_norm": 8.971960067749023, + "learning_rate": 1.6815610039170174e-05, + "loss": 0.109, + "step": 4490 + }, + { + "epoch": 0.48688204683434516, + "grad_norm": 2.0785017013549805, + "learning_rate": 1.6814884665602786e-05, + "loss": 0.0357, + "step": 4491 + }, + { + "epoch": 0.486990459670425, + "grad_norm": 1.5941087007522583, + "learning_rate": 1.68141592920354e-05, + "loss": 0.0304, + "step": 4492 + }, + { + "epoch": 0.48709887250650474, + "grad_norm": 2.7535719871520996, + "learning_rate": 1.681343391846801e-05, + "loss": 0.0521, + "step": 4493 + }, + { + "epoch": 0.48720728534258456, + "grad_norm": 2.3074357509613037, + "learning_rate": 1.6812708544900626e-05, + "loss": 0.0517, + "step": 4494 + }, + { + "epoch": 0.4873156981786644, + "grad_norm": 1.9662879705429077, + "learning_rate": 1.6811983171333238e-05, + "loss": 0.034, + "step": 4495 + }, + { + "epoch": 0.48742411101474414, + "grad_norm": 4.274440288543701, + "learning_rate": 1.681125779776585e-05, + "loss": 0.0691, + "step": 4496 + }, + { + "epoch": 0.48753252385082396, + "grad_norm": 2.828563928604126, + "learning_rate": 1.6810532424198466e-05, + "loss": 0.0781, + "step": 4497 + }, + { + "epoch": 0.4876409366869037, + "grad_norm": 2.7441251277923584, + "learning_rate": 1.6809807050631078e-05, + "loss": 0.0639, + "step": 4498 + }, + { + "epoch": 0.48774934952298354, + "grad_norm": 1.2826205492019653, + "learning_rate": 1.680908167706369e-05, + "loss": 0.0324, + "step": 4499 + }, + { + "epoch": 0.4878577623590633, + "grad_norm": 4.290728569030762, + "learning_rate": 1.6808356303496302e-05, + "loss": 0.0722, + "step": 4500 + }, + { + "epoch": 0.4879661751951431, + "grad_norm": 3.0183584690093994, + "learning_rate": 1.6807630929928914e-05, + "loss": 0.0458, + "step": 4501 + }, + { + "epoch": 0.4880745880312229, + "grad_norm": 2.6427054405212402, + "learning_rate": 1.6806905556361527e-05, + "loss": 0.0418, + "step": 4502 + }, + { + "epoch": 0.4881830008673027, + "grad_norm": 2.827984571456909, + "learning_rate": 1.6806180182794142e-05, + "loss": 0.0474, + "step": 4503 + }, + { + "epoch": 0.48829141370338247, + "grad_norm": 1.6433979272842407, + "learning_rate": 1.6805454809226754e-05, + "loss": 0.0261, + "step": 4504 + }, + { + "epoch": 0.4883998265394623, + "grad_norm": 4.3226823806762695, + "learning_rate": 1.6804729435659366e-05, + "loss": 0.0324, + "step": 4505 + }, + { + "epoch": 0.48850823937554205, + "grad_norm": 4.602495193481445, + "learning_rate": 1.680400406209198e-05, + "loss": 0.0456, + "step": 4506 + }, + { + "epoch": 0.48861665221162187, + "grad_norm": 2.632462739944458, + "learning_rate": 1.680327868852459e-05, + "loss": 0.024, + "step": 4507 + }, + { + "epoch": 0.48872506504770163, + "grad_norm": 1.283164381980896, + "learning_rate": 1.6802553314957203e-05, + "loss": 0.0241, + "step": 4508 + }, + { + "epoch": 0.48883347788378145, + "grad_norm": 1.9488508701324463, + "learning_rate": 1.6801827941389815e-05, + "loss": 0.0379, + "step": 4509 + }, + { + "epoch": 0.4889418907198612, + "grad_norm": 0.7933711409568787, + "learning_rate": 1.680110256782243e-05, + "loss": 0.0153, + "step": 4510 + }, + { + "epoch": 0.48905030355594103, + "grad_norm": 2.4360342025756836, + "learning_rate": 1.6800377194255043e-05, + "loss": 0.0291, + "step": 4511 + }, + { + "epoch": 0.4891587163920208, + "grad_norm": 1.0682505369186401, + "learning_rate": 1.6799651820687655e-05, + "loss": 0.0246, + "step": 4512 + }, + { + "epoch": 0.4892671292281006, + "grad_norm": 2.6520144939422607, + "learning_rate": 1.6798926447120267e-05, + "loss": 0.0435, + "step": 4513 + }, + { + "epoch": 0.4893755420641804, + "grad_norm": 11.998244285583496, + "learning_rate": 1.6798201073552882e-05, + "loss": 0.063, + "step": 4514 + }, + { + "epoch": 0.4894839549002602, + "grad_norm": 1.9189276695251465, + "learning_rate": 1.6797475699985495e-05, + "loss": 0.0355, + "step": 4515 + }, + { + "epoch": 0.48959236773633996, + "grad_norm": 2.0832431316375732, + "learning_rate": 1.6796750326418107e-05, + "loss": 0.0198, + "step": 4516 + }, + { + "epoch": 0.4897007805724198, + "grad_norm": 0.6181147694587708, + "learning_rate": 1.679602495285072e-05, + "loss": 0.0076, + "step": 4517 + }, + { + "epoch": 0.48980919340849954, + "grad_norm": 7.425376892089844, + "learning_rate": 1.6795299579283334e-05, + "loss": 0.0477, + "step": 4518 + }, + { + "epoch": 0.48991760624457936, + "grad_norm": 3.3778059482574463, + "learning_rate": 1.6794574205715947e-05, + "loss": 0.0312, + "step": 4519 + }, + { + "epoch": 0.4900260190806591, + "grad_norm": 1.9690208435058594, + "learning_rate": 1.679384883214856e-05, + "loss": 0.0318, + "step": 4520 + }, + { + "epoch": 0.49013443191673894, + "grad_norm": 2.7819225788116455, + "learning_rate": 1.679312345858117e-05, + "loss": 0.0461, + "step": 4521 + }, + { + "epoch": 0.4902428447528187, + "grad_norm": 3.425234794616699, + "learning_rate": 1.6792398085013783e-05, + "loss": 0.081, + "step": 4522 + }, + { + "epoch": 0.4903512575888985, + "grad_norm": 3.616643190383911, + "learning_rate": 1.6791672711446395e-05, + "loss": 0.029, + "step": 4523 + }, + { + "epoch": 0.49045967042497834, + "grad_norm": 2.8658037185668945, + "learning_rate": 1.6790947337879007e-05, + "loss": 0.0201, + "step": 4524 + }, + { + "epoch": 0.4905680832610581, + "grad_norm": 1.7912286520004272, + "learning_rate": 1.6790221964311623e-05, + "loss": 0.0333, + "step": 4525 + }, + { + "epoch": 0.4906764960971379, + "grad_norm": 2.3147308826446533, + "learning_rate": 1.6789496590744235e-05, + "loss": 0.0356, + "step": 4526 + }, + { + "epoch": 0.4907849089332177, + "grad_norm": 1.4324216842651367, + "learning_rate": 1.6788771217176847e-05, + "loss": 0.0262, + "step": 4527 + }, + { + "epoch": 0.4908933217692975, + "grad_norm": 7.997109413146973, + "learning_rate": 1.678804584360946e-05, + "loss": 0.079, + "step": 4528 + }, + { + "epoch": 0.49100173460537727, + "grad_norm": 2.4405882358551025, + "learning_rate": 1.678732047004207e-05, + "loss": 0.0337, + "step": 4529 + }, + { + "epoch": 0.4911101474414571, + "grad_norm": 9.712556838989258, + "learning_rate": 1.6786595096474684e-05, + "loss": 0.0911, + "step": 4530 + }, + { + "epoch": 0.49121856027753685, + "grad_norm": 3.906190872192383, + "learning_rate": 1.67858697229073e-05, + "loss": 0.0476, + "step": 4531 + }, + { + "epoch": 0.49132697311361667, + "grad_norm": 4.389848232269287, + "learning_rate": 1.678514434933991e-05, + "loss": 0.0393, + "step": 4532 + }, + { + "epoch": 0.49143538594969643, + "grad_norm": 0.9036026000976562, + "learning_rate": 1.6784418975772527e-05, + "loss": 0.0526, + "step": 4533 + }, + { + "epoch": 0.49154379878577625, + "grad_norm": 3.351301670074463, + "learning_rate": 1.678369360220514e-05, + "loss": 0.0408, + "step": 4534 + }, + { + "epoch": 0.491652211621856, + "grad_norm": 1.844503402709961, + "learning_rate": 1.678296822863775e-05, + "loss": 0.0111, + "step": 4535 + }, + { + "epoch": 0.49176062445793584, + "grad_norm": 2.921917676925659, + "learning_rate": 1.6782242855070363e-05, + "loss": 0.0509, + "step": 4536 + }, + { + "epoch": 0.4918690372940156, + "grad_norm": 2.7674529552459717, + "learning_rate": 1.6781517481502975e-05, + "loss": 0.0323, + "step": 4537 + }, + { + "epoch": 0.4919774501300954, + "grad_norm": 2.75809645652771, + "learning_rate": 1.6780792107935587e-05, + "loss": 0.0548, + "step": 4538 + }, + { + "epoch": 0.4920858629661752, + "grad_norm": 1.705200433731079, + "learning_rate": 1.67800667343682e-05, + "loss": 0.0221, + "step": 4539 + }, + { + "epoch": 0.492194275802255, + "grad_norm": 5.323071002960205, + "learning_rate": 1.6779341360800815e-05, + "loss": 0.074, + "step": 4540 + }, + { + "epoch": 0.49230268863833476, + "grad_norm": 7.562272548675537, + "learning_rate": 1.6778615987233427e-05, + "loss": 0.0711, + "step": 4541 + }, + { + "epoch": 0.4924111014744146, + "grad_norm": 5.888721942901611, + "learning_rate": 1.677789061366604e-05, + "loss": 0.0336, + "step": 4542 + }, + { + "epoch": 0.49251951431049434, + "grad_norm": 3.186516046524048, + "learning_rate": 1.677716524009865e-05, + "loss": 0.0724, + "step": 4543 + }, + { + "epoch": 0.49262792714657416, + "grad_norm": 3.1818954944610596, + "learning_rate": 1.6776439866531264e-05, + "loss": 0.0482, + "step": 4544 + }, + { + "epoch": 0.4927363399826539, + "grad_norm": 1.701079249382019, + "learning_rate": 1.6775714492963876e-05, + "loss": 0.0307, + "step": 4545 + }, + { + "epoch": 0.49284475281873374, + "grad_norm": 1.0019489526748657, + "learning_rate": 1.677498911939649e-05, + "loss": 0.0102, + "step": 4546 + }, + { + "epoch": 0.4929531656548135, + "grad_norm": 2.5249545574188232, + "learning_rate": 1.6774263745829104e-05, + "loss": 0.042, + "step": 4547 + }, + { + "epoch": 0.4930615784908933, + "grad_norm": 4.693789482116699, + "learning_rate": 1.6773538372261716e-05, + "loss": 0.1151, + "step": 4548 + }, + { + "epoch": 0.4931699913269731, + "grad_norm": 2.5821964740753174, + "learning_rate": 1.677281299869433e-05, + "loss": 0.0733, + "step": 4549 + }, + { + "epoch": 0.4932784041630529, + "grad_norm": 0.9890873432159424, + "learning_rate": 1.6772087625126943e-05, + "loss": 0.0207, + "step": 4550 + }, + { + "epoch": 0.49338681699913267, + "grad_norm": 3.6186792850494385, + "learning_rate": 1.6771362251559555e-05, + "loss": 0.0251, + "step": 4551 + }, + { + "epoch": 0.4934952298352125, + "grad_norm": 6.5059638023376465, + "learning_rate": 1.6770636877992168e-05, + "loss": 0.0955, + "step": 4552 + }, + { + "epoch": 0.4936036426712923, + "grad_norm": 3.1353049278259277, + "learning_rate": 1.676991150442478e-05, + "loss": 0.0395, + "step": 4553 + }, + { + "epoch": 0.49371205550737207, + "grad_norm": 4.318317890167236, + "learning_rate": 1.6769186130857392e-05, + "loss": 0.0189, + "step": 4554 + }, + { + "epoch": 0.4938204683434519, + "grad_norm": 2.147599935531616, + "learning_rate": 1.6768460757290007e-05, + "loss": 0.0315, + "step": 4555 + }, + { + "epoch": 0.49392888117953165, + "grad_norm": 1.472783088684082, + "learning_rate": 1.676773538372262e-05, + "loss": 0.0423, + "step": 4556 + }, + { + "epoch": 0.4940372940156115, + "grad_norm": 3.091566562652588, + "learning_rate": 1.6767010010155232e-05, + "loss": 0.045, + "step": 4557 + }, + { + "epoch": 0.49414570685169124, + "grad_norm": 4.742379665374756, + "learning_rate": 1.6766284636587844e-05, + "loss": 0.0723, + "step": 4558 + }, + { + "epoch": 0.49425411968777105, + "grad_norm": 1.1229125261306763, + "learning_rate": 1.6765559263020456e-05, + "loss": 0.0152, + "step": 4559 + }, + { + "epoch": 0.4943625325238508, + "grad_norm": 3.4464213848114014, + "learning_rate": 1.6764833889453068e-05, + "loss": 0.0325, + "step": 4560 + }, + { + "epoch": 0.49447094535993064, + "grad_norm": 2.812307596206665, + "learning_rate": 1.676410851588568e-05, + "loss": 0.0243, + "step": 4561 + }, + { + "epoch": 0.4945793581960104, + "grad_norm": 2.095472812652588, + "learning_rate": 1.6763383142318296e-05, + "loss": 0.0315, + "step": 4562 + }, + { + "epoch": 0.4946877710320902, + "grad_norm": 2.4433045387268066, + "learning_rate": 1.6762657768750908e-05, + "loss": 0.0338, + "step": 4563 + }, + { + "epoch": 0.49479618386817, + "grad_norm": 1.4315459728240967, + "learning_rate": 1.676193239518352e-05, + "loss": 0.0275, + "step": 4564 + }, + { + "epoch": 0.4949045967042498, + "grad_norm": 6.765658378601074, + "learning_rate": 1.6761207021616132e-05, + "loss": 0.1091, + "step": 4565 + }, + { + "epoch": 0.49501300954032956, + "grad_norm": 3.278973340988159, + "learning_rate": 1.6760481648048748e-05, + "loss": 0.0193, + "step": 4566 + }, + { + "epoch": 0.4951214223764094, + "grad_norm": 1.1620160341262817, + "learning_rate": 1.675975627448136e-05, + "loss": 0.0143, + "step": 4567 + }, + { + "epoch": 0.49522983521248914, + "grad_norm": 4.907799243927002, + "learning_rate": 1.6759030900913972e-05, + "loss": 0.0681, + "step": 4568 + }, + { + "epoch": 0.49533824804856896, + "grad_norm": 5.515619277954102, + "learning_rate": 1.6758305527346584e-05, + "loss": 0.049, + "step": 4569 + }, + { + "epoch": 0.4954466608846487, + "grad_norm": 4.604776859283447, + "learning_rate": 1.67575801537792e-05, + "loss": 0.0268, + "step": 4570 + }, + { + "epoch": 0.49555507372072854, + "grad_norm": 6.31388521194458, + "learning_rate": 1.6756854780211812e-05, + "loss": 0.0296, + "step": 4571 + }, + { + "epoch": 0.4956634865568083, + "grad_norm": 4.310097694396973, + "learning_rate": 1.6756129406644424e-05, + "loss": 0.0722, + "step": 4572 + }, + { + "epoch": 0.4957718993928881, + "grad_norm": 1.3118258714675903, + "learning_rate": 1.6755404033077036e-05, + "loss": 0.0162, + "step": 4573 + }, + { + "epoch": 0.4958803122289679, + "grad_norm": 1.617424726486206, + "learning_rate": 1.675467865950965e-05, + "loss": 0.0533, + "step": 4574 + }, + { + "epoch": 0.4959887250650477, + "grad_norm": 2.4193973541259766, + "learning_rate": 1.675395328594226e-05, + "loss": 0.0505, + "step": 4575 + }, + { + "epoch": 0.49609713790112747, + "grad_norm": 3.13797664642334, + "learning_rate": 1.6753227912374873e-05, + "loss": 0.0412, + "step": 4576 + }, + { + "epoch": 0.4962055507372073, + "grad_norm": 3.422022819519043, + "learning_rate": 1.6752502538807488e-05, + "loss": 0.0293, + "step": 4577 + }, + { + "epoch": 0.49631396357328705, + "grad_norm": 3.9658429622650146, + "learning_rate": 1.67517771652401e-05, + "loss": 0.0897, + "step": 4578 + }, + { + "epoch": 0.4964223764093669, + "grad_norm": 2.0007050037384033, + "learning_rate": 1.6751051791672712e-05, + "loss": 0.0415, + "step": 4579 + }, + { + "epoch": 0.49653078924544664, + "grad_norm": 4.091763496398926, + "learning_rate": 1.6750326418105325e-05, + "loss": 0.0314, + "step": 4580 + }, + { + "epoch": 0.49663920208152645, + "grad_norm": 5.0524773597717285, + "learning_rate": 1.6749601044537937e-05, + "loss": 0.0521, + "step": 4581 + }, + { + "epoch": 0.4967476149176062, + "grad_norm": 2.237384080886841, + "learning_rate": 1.674887567097055e-05, + "loss": 0.0489, + "step": 4582 + }, + { + "epoch": 0.49685602775368604, + "grad_norm": 1.056670904159546, + "learning_rate": 1.6748150297403164e-05, + "loss": 0.022, + "step": 4583 + }, + { + "epoch": 0.49696444058976585, + "grad_norm": 2.4624545574188232, + "learning_rate": 1.6747424923835777e-05, + "loss": 0.0175, + "step": 4584 + }, + { + "epoch": 0.4970728534258456, + "grad_norm": 3.040264368057251, + "learning_rate": 1.6746699550268392e-05, + "loss": 0.0376, + "step": 4585 + }, + { + "epoch": 0.49718126626192544, + "grad_norm": 2.9210593700408936, + "learning_rate": 1.6745974176701004e-05, + "loss": 0.0326, + "step": 4586 + }, + { + "epoch": 0.4972896790980052, + "grad_norm": 2.997708559036255, + "learning_rate": 1.6745248803133616e-05, + "loss": 0.0301, + "step": 4587 + }, + { + "epoch": 0.497398091934085, + "grad_norm": 3.9931743144989014, + "learning_rate": 1.674452342956623e-05, + "loss": 0.0692, + "step": 4588 + }, + { + "epoch": 0.4975065047701648, + "grad_norm": 1.1051782369613647, + "learning_rate": 1.674379805599884e-05, + "loss": 0.014, + "step": 4589 + }, + { + "epoch": 0.4976149176062446, + "grad_norm": 4.5225348472595215, + "learning_rate": 1.6743072682431453e-05, + "loss": 0.0541, + "step": 4590 + }, + { + "epoch": 0.49772333044232436, + "grad_norm": 1.4611256122589111, + "learning_rate": 1.6742347308864065e-05, + "loss": 0.0276, + "step": 4591 + }, + { + "epoch": 0.4978317432784042, + "grad_norm": 1.9181493520736694, + "learning_rate": 1.674162193529668e-05, + "loss": 0.0566, + "step": 4592 + }, + { + "epoch": 0.49794015611448394, + "grad_norm": 1.2424193620681763, + "learning_rate": 1.6740896561729293e-05, + "loss": 0.0185, + "step": 4593 + }, + { + "epoch": 0.49804856895056376, + "grad_norm": 2.614839792251587, + "learning_rate": 1.6740171188161905e-05, + "loss": 0.0407, + "step": 4594 + }, + { + "epoch": 0.4981569817866435, + "grad_norm": 1.0881662368774414, + "learning_rate": 1.6739445814594517e-05, + "loss": 0.0173, + "step": 4595 + }, + { + "epoch": 0.49826539462272335, + "grad_norm": 3.4651081562042236, + "learning_rate": 1.673872044102713e-05, + "loss": 0.0842, + "step": 4596 + }, + { + "epoch": 0.4983738074588031, + "grad_norm": 1.652126669883728, + "learning_rate": 1.673799506745974e-05, + "loss": 0.0252, + "step": 4597 + }, + { + "epoch": 0.4984822202948829, + "grad_norm": 1.9482512474060059, + "learning_rate": 1.6737269693892357e-05, + "loss": 0.024, + "step": 4598 + }, + { + "epoch": 0.4985906331309627, + "grad_norm": 0.5176246166229248, + "learning_rate": 1.673654432032497e-05, + "loss": 0.0084, + "step": 4599 + }, + { + "epoch": 0.4986990459670425, + "grad_norm": 2.146324634552002, + "learning_rate": 1.673581894675758e-05, + "loss": 0.0232, + "step": 4600 + }, + { + "epoch": 0.4988074588031223, + "grad_norm": 4.4464592933654785, + "learning_rate": 1.6735093573190197e-05, + "loss": 0.0508, + "step": 4601 + }, + { + "epoch": 0.4989158716392021, + "grad_norm": 2.602210283279419, + "learning_rate": 1.673436819962281e-05, + "loss": 0.0399, + "step": 4602 + }, + { + "epoch": 0.49902428447528185, + "grad_norm": 6.699312210083008, + "learning_rate": 1.673364282605542e-05, + "loss": 0.0475, + "step": 4603 + }, + { + "epoch": 0.4991326973113617, + "grad_norm": 0.7775063514709473, + "learning_rate": 1.6732917452488033e-05, + "loss": 0.0098, + "step": 4604 + }, + { + "epoch": 0.49924111014744144, + "grad_norm": 2.9465348720550537, + "learning_rate": 1.6732192078920645e-05, + "loss": 0.023, + "step": 4605 + }, + { + "epoch": 0.49934952298352125, + "grad_norm": 0.8424277901649475, + "learning_rate": 1.6731466705353257e-05, + "loss": 0.011, + "step": 4606 + }, + { + "epoch": 0.499457935819601, + "grad_norm": 8.405474662780762, + "learning_rate": 1.6730741331785873e-05, + "loss": 0.047, + "step": 4607 + }, + { + "epoch": 0.49956634865568084, + "grad_norm": 2.870436906814575, + "learning_rate": 1.6730015958218485e-05, + "loss": 0.0285, + "step": 4608 + }, + { + "epoch": 0.4996747614917606, + "grad_norm": 2.5877156257629395, + "learning_rate": 1.6729290584651097e-05, + "loss": 0.0299, + "step": 4609 + }, + { + "epoch": 0.4997831743278404, + "grad_norm": 0.7271683216094971, + "learning_rate": 1.672856521108371e-05, + "loss": 0.0105, + "step": 4610 + }, + { + "epoch": 0.4998915871639202, + "grad_norm": 1.8379112482070923, + "learning_rate": 1.672783983751632e-05, + "loss": 0.0097, + "step": 4611 + }, + { + "epoch": 0.5, + "grad_norm": 2.175999402999878, + "learning_rate": 1.6727114463948934e-05, + "loss": 0.0285, + "step": 4612 + }, + { + "epoch": 0.5001084128360798, + "grad_norm": 0.20611460506916046, + "learning_rate": 1.6726389090381546e-05, + "loss": 0.0022, + "step": 4613 + }, + { + "epoch": 0.5002168256721596, + "grad_norm": 4.077826976776123, + "learning_rate": 1.672566371681416e-05, + "loss": 0.0976, + "step": 4614 + }, + { + "epoch": 0.5003252385082394, + "grad_norm": 3.778219699859619, + "learning_rate": 1.6724938343246773e-05, + "loss": 0.0412, + "step": 4615 + }, + { + "epoch": 0.5004336513443192, + "grad_norm": 2.5174942016601562, + "learning_rate": 1.6724212969679385e-05, + "loss": 0.0337, + "step": 4616 + }, + { + "epoch": 0.5005420641803989, + "grad_norm": 0.33191901445388794, + "learning_rate": 1.6723487596111998e-05, + "loss": 0.004, + "step": 4617 + }, + { + "epoch": 0.5006504770164788, + "grad_norm": 2.7791788578033447, + "learning_rate": 1.6722762222544613e-05, + "loss": 0.0557, + "step": 4618 + }, + { + "epoch": 0.5007588898525586, + "grad_norm": 1.070298433303833, + "learning_rate": 1.6722036848977225e-05, + "loss": 0.0039, + "step": 4619 + }, + { + "epoch": 0.5008673026886383, + "grad_norm": 4.333750247955322, + "learning_rate": 1.6721311475409837e-05, + "loss": 0.0637, + "step": 4620 + }, + { + "epoch": 0.5009757155247181, + "grad_norm": 1.4283169507980347, + "learning_rate": 1.672058610184245e-05, + "loss": 0.0401, + "step": 4621 + }, + { + "epoch": 0.501084128360798, + "grad_norm": 3.54632306098938, + "learning_rate": 1.6719860728275065e-05, + "loss": 0.0156, + "step": 4622 + }, + { + "epoch": 0.5011925411968777, + "grad_norm": 2.873049259185791, + "learning_rate": 1.6719135354707677e-05, + "loss": 0.0276, + "step": 4623 + }, + { + "epoch": 0.5013009540329575, + "grad_norm": 3.796480655670166, + "learning_rate": 1.671840998114029e-05, + "loss": 0.0543, + "step": 4624 + }, + { + "epoch": 0.5014093668690373, + "grad_norm": 2.759819984436035, + "learning_rate": 1.67176846075729e-05, + "loss": 0.0418, + "step": 4625 + }, + { + "epoch": 0.5015177797051171, + "grad_norm": 3.8994030952453613, + "learning_rate": 1.6716959234005514e-05, + "loss": 0.058, + "step": 4626 + }, + { + "epoch": 0.5016261925411969, + "grad_norm": 4.655514240264893, + "learning_rate": 1.6716233860438126e-05, + "loss": 0.023, + "step": 4627 + }, + { + "epoch": 0.5017346053772767, + "grad_norm": 2.625568151473999, + "learning_rate": 1.6715508486870738e-05, + "loss": 0.0238, + "step": 4628 + }, + { + "epoch": 0.5018430182133564, + "grad_norm": 0.36555543541908264, + "learning_rate": 1.6714783113303354e-05, + "loss": 0.0042, + "step": 4629 + }, + { + "epoch": 0.5019514310494363, + "grad_norm": 9.924644470214844, + "learning_rate": 1.6714057739735966e-05, + "loss": 0.014, + "step": 4630 + }, + { + "epoch": 0.502059843885516, + "grad_norm": 0.8911523222923279, + "learning_rate": 1.6713332366168578e-05, + "loss": 0.0178, + "step": 4631 + }, + { + "epoch": 0.5021682567215958, + "grad_norm": 3.597550630569458, + "learning_rate": 1.671260699260119e-05, + "loss": 0.0209, + "step": 4632 + }, + { + "epoch": 0.5022766695576756, + "grad_norm": 0.9766392111778259, + "learning_rate": 1.6711881619033802e-05, + "loss": 0.0073, + "step": 4633 + }, + { + "epoch": 0.5023850823937555, + "grad_norm": 5.534913063049316, + "learning_rate": 1.6711156245466414e-05, + "loss": 0.0262, + "step": 4634 + }, + { + "epoch": 0.5024934952298352, + "grad_norm": 1.848031997680664, + "learning_rate": 1.671043087189903e-05, + "loss": 0.0193, + "step": 4635 + }, + { + "epoch": 0.502601908065915, + "grad_norm": 4.073891639709473, + "learning_rate": 1.6709705498331642e-05, + "loss": 0.0144, + "step": 4636 + }, + { + "epoch": 0.5027103209019947, + "grad_norm": 4.641273021697998, + "learning_rate": 1.6708980124764257e-05, + "loss": 0.0383, + "step": 4637 + }, + { + "epoch": 0.5028187337380746, + "grad_norm": 8.031255722045898, + "learning_rate": 1.670825475119687e-05, + "loss": 0.0648, + "step": 4638 + }, + { + "epoch": 0.5029271465741544, + "grad_norm": 6.414639472961426, + "learning_rate": 1.6707529377629482e-05, + "loss": 0.0572, + "step": 4639 + }, + { + "epoch": 0.5030355594102341, + "grad_norm": 3.7662272453308105, + "learning_rate": 1.6706804004062094e-05, + "loss": 0.0398, + "step": 4640 + }, + { + "epoch": 0.5031439722463139, + "grad_norm": 4.054555892944336, + "learning_rate": 1.6706078630494706e-05, + "loss": 0.0474, + "step": 4641 + }, + { + "epoch": 0.5032523850823938, + "grad_norm": 1.7317343950271606, + "learning_rate": 1.6705353256927318e-05, + "loss": 0.0158, + "step": 4642 + }, + { + "epoch": 0.5033607979184735, + "grad_norm": 5.884375095367432, + "learning_rate": 1.670462788335993e-05, + "loss": 0.0351, + "step": 4643 + }, + { + "epoch": 0.5034692107545533, + "grad_norm": 5.413352966308594, + "learning_rate": 1.6703902509792546e-05, + "loss": 0.0977, + "step": 4644 + }, + { + "epoch": 0.5035776235906332, + "grad_norm": 3.296931028366089, + "learning_rate": 1.6703177136225158e-05, + "loss": 0.0712, + "step": 4645 + }, + { + "epoch": 0.503686036426713, + "grad_norm": 6.990636348724365, + "learning_rate": 1.670245176265777e-05, + "loss": 0.0347, + "step": 4646 + }, + { + "epoch": 0.5037944492627927, + "grad_norm": 4.8439412117004395, + "learning_rate": 1.6701726389090382e-05, + "loss": 0.0614, + "step": 4647 + }, + { + "epoch": 0.5039028620988725, + "grad_norm": 4.515338897705078, + "learning_rate": 1.6701001015522994e-05, + "loss": 0.1014, + "step": 4648 + }, + { + "epoch": 0.5040112749349523, + "grad_norm": 3.734900951385498, + "learning_rate": 1.6700275641955607e-05, + "loss": 0.0221, + "step": 4649 + }, + { + "epoch": 0.5041196877710321, + "grad_norm": 3.38956618309021, + "learning_rate": 1.6699550268388222e-05, + "loss": 0.0265, + "step": 4650 + }, + { + "epoch": 0.5042281006071119, + "grad_norm": 2.7469401359558105, + "learning_rate": 1.6698824894820834e-05, + "loss": 0.0463, + "step": 4651 + }, + { + "epoch": 0.5043365134431916, + "grad_norm": 3.1678378582000732, + "learning_rate": 1.6698099521253446e-05, + "loss": 0.0654, + "step": 4652 + }, + { + "epoch": 0.5044449262792715, + "grad_norm": 1.15753972530365, + "learning_rate": 1.6697374147686062e-05, + "loss": 0.0296, + "step": 4653 + }, + { + "epoch": 0.5045533391153513, + "grad_norm": 1.8827993869781494, + "learning_rate": 1.6696648774118674e-05, + "loss": 0.0597, + "step": 4654 + }, + { + "epoch": 0.504661751951431, + "grad_norm": 3.2394301891326904, + "learning_rate": 1.6695923400551286e-05, + "loss": 0.0488, + "step": 4655 + }, + { + "epoch": 0.5047701647875108, + "grad_norm": 2.1407811641693115, + "learning_rate": 1.66951980269839e-05, + "loss": 0.0209, + "step": 4656 + }, + { + "epoch": 0.5048785776235907, + "grad_norm": 1.1472710371017456, + "learning_rate": 1.669447265341651e-05, + "loss": 0.0086, + "step": 4657 + }, + { + "epoch": 0.5049869904596704, + "grad_norm": 4.074243068695068, + "learning_rate": 1.6693747279849123e-05, + "loss": 0.0604, + "step": 4658 + }, + { + "epoch": 0.5050954032957502, + "grad_norm": 1.9217363595962524, + "learning_rate": 1.6693021906281738e-05, + "loss": 0.0453, + "step": 4659 + }, + { + "epoch": 0.50520381613183, + "grad_norm": 1.6691957712173462, + "learning_rate": 1.669229653271435e-05, + "loss": 0.032, + "step": 4660 + }, + { + "epoch": 0.5053122289679098, + "grad_norm": 2.7036633491516113, + "learning_rate": 1.6691571159146962e-05, + "loss": 0.061, + "step": 4661 + }, + { + "epoch": 0.5054206418039896, + "grad_norm": 0.7166115045547485, + "learning_rate": 1.6690845785579575e-05, + "loss": 0.0199, + "step": 4662 + }, + { + "epoch": 0.5055290546400694, + "grad_norm": 1.6899985074996948, + "learning_rate": 1.6690120412012187e-05, + "loss": 0.0279, + "step": 4663 + }, + { + "epoch": 0.5056374674761491, + "grad_norm": 5.321112155914307, + "learning_rate": 1.66893950384448e-05, + "loss": 0.0726, + "step": 4664 + }, + { + "epoch": 0.505745880312229, + "grad_norm": 2.033956527709961, + "learning_rate": 1.6688669664877414e-05, + "loss": 0.0415, + "step": 4665 + }, + { + "epoch": 0.5058542931483088, + "grad_norm": 3.892167806625366, + "learning_rate": 1.6687944291310027e-05, + "loss": 0.0963, + "step": 4666 + }, + { + "epoch": 0.5059627059843885, + "grad_norm": 1.737677812576294, + "learning_rate": 1.668721891774264e-05, + "loss": 0.0315, + "step": 4667 + }, + { + "epoch": 0.5060711188204683, + "grad_norm": 3.51914381980896, + "learning_rate": 1.668649354417525e-05, + "loss": 0.0329, + "step": 4668 + }, + { + "epoch": 0.5061795316565482, + "grad_norm": 3.5866010189056396, + "learning_rate": 1.6685768170607863e-05, + "loss": 0.0181, + "step": 4669 + }, + { + "epoch": 0.5062879444926279, + "grad_norm": 3.1307361125946045, + "learning_rate": 1.668504279704048e-05, + "loss": 0.0559, + "step": 4670 + }, + { + "epoch": 0.5063963573287077, + "grad_norm": 3.5944058895111084, + "learning_rate": 1.668431742347309e-05, + "loss": 0.0495, + "step": 4671 + }, + { + "epoch": 0.5065047701647876, + "grad_norm": 1.3279756307601929, + "learning_rate": 1.6683592049905703e-05, + "loss": 0.0212, + "step": 4672 + }, + { + "epoch": 0.5066131830008673, + "grad_norm": 3.2348906993865967, + "learning_rate": 1.6682866676338315e-05, + "loss": 0.0732, + "step": 4673 + }, + { + "epoch": 0.5067215958369471, + "grad_norm": 1.8767125606536865, + "learning_rate": 1.668214130277093e-05, + "loss": 0.0475, + "step": 4674 + }, + { + "epoch": 0.5068300086730269, + "grad_norm": 5.9938483238220215, + "learning_rate": 1.6681415929203543e-05, + "loss": 0.033, + "step": 4675 + }, + { + "epoch": 0.5069384215091067, + "grad_norm": 1.7726587057113647, + "learning_rate": 1.6680690555636155e-05, + "loss": 0.0221, + "step": 4676 + }, + { + "epoch": 0.5070468343451865, + "grad_norm": 5.263266563415527, + "learning_rate": 1.6679965182068767e-05, + "loss": 0.0677, + "step": 4677 + }, + { + "epoch": 0.5071552471812663, + "grad_norm": 4.554039478302002, + "learning_rate": 1.667923980850138e-05, + "loss": 0.0459, + "step": 4678 + }, + { + "epoch": 0.507263660017346, + "grad_norm": 0.6805658936500549, + "learning_rate": 1.667851443493399e-05, + "loss": 0.0083, + "step": 4679 + }, + { + "epoch": 0.5073720728534259, + "grad_norm": 1.2438431978225708, + "learning_rate": 1.6677789061366603e-05, + "loss": 0.0276, + "step": 4680 + }, + { + "epoch": 0.5074804856895057, + "grad_norm": 2.0310423374176025, + "learning_rate": 1.667706368779922e-05, + "loss": 0.0435, + "step": 4681 + }, + { + "epoch": 0.5075888985255854, + "grad_norm": 2.7589683532714844, + "learning_rate": 1.667633831423183e-05, + "loss": 0.048, + "step": 4682 + }, + { + "epoch": 0.5076973113616652, + "grad_norm": 6.341707706451416, + "learning_rate": 1.6675612940664443e-05, + "loss": 0.0679, + "step": 4683 + }, + { + "epoch": 0.5078057241977451, + "grad_norm": 2.014653205871582, + "learning_rate": 1.6674887567097055e-05, + "loss": 0.0476, + "step": 4684 + }, + { + "epoch": 0.5079141370338248, + "grad_norm": 2.265484094619751, + "learning_rate": 1.6674162193529667e-05, + "loss": 0.0455, + "step": 4685 + }, + { + "epoch": 0.5080225498699046, + "grad_norm": 2.2072243690490723, + "learning_rate": 1.667343681996228e-05, + "loss": 0.0388, + "step": 4686 + }, + { + "epoch": 0.5081309627059843, + "grad_norm": 7.1816511154174805, + "learning_rate": 1.6672711446394895e-05, + "loss": 0.0854, + "step": 4687 + }, + { + "epoch": 0.5082393755420642, + "grad_norm": 4.0957560539245605, + "learning_rate": 1.6671986072827507e-05, + "loss": 0.0813, + "step": 4688 + }, + { + "epoch": 0.508347788378144, + "grad_norm": 3.038170576095581, + "learning_rate": 1.6671260699260123e-05, + "loss": 0.0563, + "step": 4689 + }, + { + "epoch": 0.5084562012142237, + "grad_norm": 2.494467258453369, + "learning_rate": 1.6670535325692735e-05, + "loss": 0.0721, + "step": 4690 + }, + { + "epoch": 0.5085646140503035, + "grad_norm": 3.1041977405548096, + "learning_rate": 1.6669809952125347e-05, + "loss": 0.0273, + "step": 4691 + }, + { + "epoch": 0.5086730268863834, + "grad_norm": 4.601886749267578, + "learning_rate": 1.666908457855796e-05, + "loss": 0.0634, + "step": 4692 + }, + { + "epoch": 0.5087814397224631, + "grad_norm": 12.953119277954102, + "learning_rate": 1.666835920499057e-05, + "loss": 0.0431, + "step": 4693 + }, + { + "epoch": 0.5088898525585429, + "grad_norm": 2.480243444442749, + "learning_rate": 1.6667633831423183e-05, + "loss": 0.04, + "step": 4694 + }, + { + "epoch": 0.5089982653946227, + "grad_norm": 2.0744192600250244, + "learning_rate": 1.6666908457855796e-05, + "loss": 0.0168, + "step": 4695 + }, + { + "epoch": 0.5091066782307025, + "grad_norm": 1.3341600894927979, + "learning_rate": 1.666618308428841e-05, + "loss": 0.0453, + "step": 4696 + }, + { + "epoch": 0.5092150910667823, + "grad_norm": 2.0609395503997803, + "learning_rate": 1.6665457710721023e-05, + "loss": 0.0202, + "step": 4697 + }, + { + "epoch": 0.5093235039028621, + "grad_norm": 2.9086005687713623, + "learning_rate": 1.6664732337153635e-05, + "loss": 0.0522, + "step": 4698 + }, + { + "epoch": 0.5094319167389418, + "grad_norm": 1.0303409099578857, + "learning_rate": 1.6664006963586248e-05, + "loss": 0.0178, + "step": 4699 + }, + { + "epoch": 0.5095403295750217, + "grad_norm": 3.143831968307495, + "learning_rate": 1.666328159001886e-05, + "loss": 0.0481, + "step": 4700 + }, + { + "epoch": 0.5096487424111015, + "grad_norm": 2.154165744781494, + "learning_rate": 1.6662556216451472e-05, + "loss": 0.0169, + "step": 4701 + }, + { + "epoch": 0.5097571552471812, + "grad_norm": 4.345488548278809, + "learning_rate": 1.6661830842884087e-05, + "loss": 0.0176, + "step": 4702 + }, + { + "epoch": 0.5098655680832611, + "grad_norm": 3.371687889099121, + "learning_rate": 1.66611054693167e-05, + "loss": 0.0948, + "step": 4703 + }, + { + "epoch": 0.5099739809193409, + "grad_norm": 1.5746986865997314, + "learning_rate": 1.666038009574931e-05, + "loss": 0.0225, + "step": 4704 + }, + { + "epoch": 0.5100823937554206, + "grad_norm": 2.2266054153442383, + "learning_rate": 1.6659654722181927e-05, + "loss": 0.0301, + "step": 4705 + }, + { + "epoch": 0.5101908065915004, + "grad_norm": 2.2544515132904053, + "learning_rate": 1.665892934861454e-05, + "loss": 0.0308, + "step": 4706 + }, + { + "epoch": 0.5102992194275803, + "grad_norm": 0.9143131971359253, + "learning_rate": 1.665820397504715e-05, + "loss": 0.0163, + "step": 4707 + }, + { + "epoch": 0.51040763226366, + "grad_norm": 1.5078938007354736, + "learning_rate": 1.6657478601479764e-05, + "loss": 0.0292, + "step": 4708 + }, + { + "epoch": 0.5105160450997398, + "grad_norm": 2.3790063858032227, + "learning_rate": 1.6656753227912376e-05, + "loss": 0.0368, + "step": 4709 + }, + { + "epoch": 0.5106244579358196, + "grad_norm": 3.8245768547058105, + "learning_rate": 1.6656027854344988e-05, + "loss": 0.0646, + "step": 4710 + }, + { + "epoch": 0.5107328707718994, + "grad_norm": 0.37288784980773926, + "learning_rate": 1.6655302480777603e-05, + "loss": 0.0047, + "step": 4711 + }, + { + "epoch": 0.5108412836079792, + "grad_norm": 2.579779863357544, + "learning_rate": 1.6654577107210216e-05, + "loss": 0.0442, + "step": 4712 + }, + { + "epoch": 0.510949696444059, + "grad_norm": 2.4482245445251465, + "learning_rate": 1.6653851733642828e-05, + "loss": 0.0435, + "step": 4713 + }, + { + "epoch": 0.5110581092801387, + "grad_norm": 2.5762133598327637, + "learning_rate": 1.665312636007544e-05, + "loss": 0.0579, + "step": 4714 + }, + { + "epoch": 0.5111665221162186, + "grad_norm": 2.1364846229553223, + "learning_rate": 1.6652400986508052e-05, + "loss": 0.041, + "step": 4715 + }, + { + "epoch": 0.5112749349522984, + "grad_norm": 1.5597339868545532, + "learning_rate": 1.6651675612940664e-05, + "loss": 0.0351, + "step": 4716 + }, + { + "epoch": 0.5113833477883781, + "grad_norm": 3.950113296508789, + "learning_rate": 1.665095023937328e-05, + "loss": 0.0796, + "step": 4717 + }, + { + "epoch": 0.5114917606244579, + "grad_norm": 4.649710655212402, + "learning_rate": 1.6650224865805892e-05, + "loss": 0.0869, + "step": 4718 + }, + { + "epoch": 0.5116001734605378, + "grad_norm": 0.6812368631362915, + "learning_rate": 1.6649499492238504e-05, + "loss": 0.0063, + "step": 4719 + }, + { + "epoch": 0.5117085862966175, + "grad_norm": 3.957498550415039, + "learning_rate": 1.6648774118671116e-05, + "loss": 0.0608, + "step": 4720 + }, + { + "epoch": 0.5118169991326973, + "grad_norm": 1.0983271598815918, + "learning_rate": 1.6648048745103728e-05, + "loss": 0.0287, + "step": 4721 + }, + { + "epoch": 0.511925411968777, + "grad_norm": 4.676031112670898, + "learning_rate": 1.6647323371536344e-05, + "loss": 0.0742, + "step": 4722 + }, + { + "epoch": 0.5120338248048569, + "grad_norm": 0.9623878002166748, + "learning_rate": 1.6646597997968956e-05, + "loss": 0.0137, + "step": 4723 + }, + { + "epoch": 0.5121422376409367, + "grad_norm": 3.0985870361328125, + "learning_rate": 1.6645872624401568e-05, + "loss": 0.0444, + "step": 4724 + }, + { + "epoch": 0.5122506504770165, + "grad_norm": 7.786823749542236, + "learning_rate": 1.664514725083418e-05, + "loss": 0.0559, + "step": 4725 + }, + { + "epoch": 0.5123590633130962, + "grad_norm": 3.033903121948242, + "learning_rate": 1.6644421877266796e-05, + "loss": 0.0141, + "step": 4726 + }, + { + "epoch": 0.5124674761491761, + "grad_norm": 3.479531764984131, + "learning_rate": 1.6643696503699408e-05, + "loss": 0.0434, + "step": 4727 + }, + { + "epoch": 0.5125758889852559, + "grad_norm": 1.502650499343872, + "learning_rate": 1.664297113013202e-05, + "loss": 0.0307, + "step": 4728 + }, + { + "epoch": 0.5126843018213356, + "grad_norm": 7.388299942016602, + "learning_rate": 1.6642245756564632e-05, + "loss": 0.0737, + "step": 4729 + }, + { + "epoch": 0.5127927146574154, + "grad_norm": 0.9599138498306274, + "learning_rate": 1.6641520382997244e-05, + "loss": 0.0087, + "step": 4730 + }, + { + "epoch": 0.5129011274934953, + "grad_norm": 3.565629720687866, + "learning_rate": 1.6640795009429857e-05, + "loss": 0.0185, + "step": 4731 + }, + { + "epoch": 0.513009540329575, + "grad_norm": 1.9976650476455688, + "learning_rate": 1.664006963586247e-05, + "loss": 0.0068, + "step": 4732 + }, + { + "epoch": 0.5131179531656548, + "grad_norm": 5.811744689941406, + "learning_rate": 1.6639344262295084e-05, + "loss": 0.0529, + "step": 4733 + }, + { + "epoch": 0.5132263660017347, + "grad_norm": 3.949899673461914, + "learning_rate": 1.6638618888727696e-05, + "loss": 0.0291, + "step": 4734 + }, + { + "epoch": 0.5133347788378144, + "grad_norm": 4.1686015129089355, + "learning_rate": 1.663789351516031e-05, + "loss": 0.041, + "step": 4735 + }, + { + "epoch": 0.5134431916738942, + "grad_norm": 1.3135366439819336, + "learning_rate": 1.663716814159292e-05, + "loss": 0.0051, + "step": 4736 + }, + { + "epoch": 0.513551604509974, + "grad_norm": 5.971502780914307, + "learning_rate": 1.6636442768025533e-05, + "loss": 0.0528, + "step": 4737 + }, + { + "epoch": 0.5136600173460538, + "grad_norm": 2.802671194076538, + "learning_rate": 1.6635717394458145e-05, + "loss": 0.0525, + "step": 4738 + }, + { + "epoch": 0.5137684301821336, + "grad_norm": 4.617991924285889, + "learning_rate": 1.663499202089076e-05, + "loss": 0.0762, + "step": 4739 + }, + { + "epoch": 0.5138768430182133, + "grad_norm": 4.436395168304443, + "learning_rate": 1.6634266647323373e-05, + "loss": 0.0185, + "step": 4740 + }, + { + "epoch": 0.5139852558542931, + "grad_norm": 1.2243462800979614, + "learning_rate": 1.6633541273755988e-05, + "loss": 0.0174, + "step": 4741 + }, + { + "epoch": 0.514093668690373, + "grad_norm": 3.6901071071624756, + "learning_rate": 1.66328159001886e-05, + "loss": 0.0629, + "step": 4742 + }, + { + "epoch": 0.5142020815264527, + "grad_norm": 4.504311561584473, + "learning_rate": 1.6632090526621212e-05, + "loss": 0.014, + "step": 4743 + }, + { + "epoch": 0.5143104943625325, + "grad_norm": 2.3932814598083496, + "learning_rate": 1.6631365153053825e-05, + "loss": 0.0357, + "step": 4744 + }, + { + "epoch": 0.5144189071986123, + "grad_norm": 2.010185718536377, + "learning_rate": 1.6630639779486437e-05, + "loss": 0.0348, + "step": 4745 + }, + { + "epoch": 0.5145273200346921, + "grad_norm": 0.8728238940238953, + "learning_rate": 1.662991440591905e-05, + "loss": 0.0057, + "step": 4746 + }, + { + "epoch": 0.5146357328707719, + "grad_norm": 5.466737270355225, + "learning_rate": 1.662918903235166e-05, + "loss": 0.0324, + "step": 4747 + }, + { + "epoch": 0.5147441457068517, + "grad_norm": 1.8274003267288208, + "learning_rate": 1.6628463658784276e-05, + "loss": 0.026, + "step": 4748 + }, + { + "epoch": 0.5148525585429314, + "grad_norm": 5.596933364868164, + "learning_rate": 1.662773828521689e-05, + "loss": 0.0878, + "step": 4749 + }, + { + "epoch": 0.5149609713790113, + "grad_norm": 2.792661666870117, + "learning_rate": 1.66270129116495e-05, + "loss": 0.0422, + "step": 4750 + }, + { + "epoch": 0.5150693842150911, + "grad_norm": 2.2570595741271973, + "learning_rate": 1.6626287538082113e-05, + "loss": 0.0313, + "step": 4751 + }, + { + "epoch": 0.5151777970511708, + "grad_norm": 1.780289888381958, + "learning_rate": 1.6625562164514725e-05, + "loss": 0.0154, + "step": 4752 + }, + { + "epoch": 0.5152862098872506, + "grad_norm": 5.006690502166748, + "learning_rate": 1.6624836790947337e-05, + "loss": 0.0462, + "step": 4753 + }, + { + "epoch": 0.5153946227233305, + "grad_norm": 3.1289968490600586, + "learning_rate": 1.6624111417379953e-05, + "loss": 0.0597, + "step": 4754 + }, + { + "epoch": 0.5155030355594102, + "grad_norm": 6.27437162399292, + "learning_rate": 1.6623386043812565e-05, + "loss": 0.0947, + "step": 4755 + }, + { + "epoch": 0.51561144839549, + "grad_norm": 0.8605690598487854, + "learning_rate": 1.6622660670245177e-05, + "loss": 0.017, + "step": 4756 + }, + { + "epoch": 0.5157198612315698, + "grad_norm": 4.666478157043457, + "learning_rate": 1.6621935296677793e-05, + "loss": 0.0416, + "step": 4757 + }, + { + "epoch": 0.5158282740676496, + "grad_norm": 3.4767143726348877, + "learning_rate": 1.6621209923110405e-05, + "loss": 0.0249, + "step": 4758 + }, + { + "epoch": 0.5159366869037294, + "grad_norm": 4.01909875869751, + "learning_rate": 1.6620484549543017e-05, + "loss": 0.0277, + "step": 4759 + }, + { + "epoch": 0.5160450997398092, + "grad_norm": 4.524288654327393, + "learning_rate": 1.661975917597563e-05, + "loss": 0.0378, + "step": 4760 + }, + { + "epoch": 0.516153512575889, + "grad_norm": 8.015003204345703, + "learning_rate": 1.661903380240824e-05, + "loss": 0.0751, + "step": 4761 + }, + { + "epoch": 0.5162619254119688, + "grad_norm": 4.094924449920654, + "learning_rate": 1.6618308428840853e-05, + "loss": 0.0385, + "step": 4762 + }, + { + "epoch": 0.5163703382480486, + "grad_norm": 1.3274699449539185, + "learning_rate": 1.661758305527347e-05, + "loss": 0.0277, + "step": 4763 + }, + { + "epoch": 0.5164787510841283, + "grad_norm": 1.9175084829330444, + "learning_rate": 1.661685768170608e-05, + "loss": 0.015, + "step": 4764 + }, + { + "epoch": 0.5165871639202082, + "grad_norm": 1.9472368955612183, + "learning_rate": 1.6616132308138693e-05, + "loss": 0.0116, + "step": 4765 + }, + { + "epoch": 0.516695576756288, + "grad_norm": 7.013864517211914, + "learning_rate": 1.6615406934571305e-05, + "loss": 0.0398, + "step": 4766 + }, + { + "epoch": 0.5168039895923677, + "grad_norm": 8.406414031982422, + "learning_rate": 1.6614681561003917e-05, + "loss": 0.0382, + "step": 4767 + }, + { + "epoch": 0.5169124024284475, + "grad_norm": 1.4405426979064941, + "learning_rate": 1.661395618743653e-05, + "loss": 0.0188, + "step": 4768 + }, + { + "epoch": 0.5170208152645274, + "grad_norm": 1.6282219886779785, + "learning_rate": 1.6613230813869145e-05, + "loss": 0.0228, + "step": 4769 + }, + { + "epoch": 0.5171292281006071, + "grad_norm": 2.4605345726013184, + "learning_rate": 1.6612505440301757e-05, + "loss": 0.0283, + "step": 4770 + }, + { + "epoch": 0.5172376409366869, + "grad_norm": 2.2691755294799805, + "learning_rate": 1.661178006673437e-05, + "loss": 0.0182, + "step": 4771 + }, + { + "epoch": 0.5173460537727667, + "grad_norm": 2.882096767425537, + "learning_rate": 1.661105469316698e-05, + "loss": 0.0492, + "step": 4772 + }, + { + "epoch": 0.5174544666088465, + "grad_norm": 3.506002426147461, + "learning_rate": 1.6610329319599594e-05, + "loss": 0.0412, + "step": 4773 + }, + { + "epoch": 0.5175628794449263, + "grad_norm": 0.9027864933013916, + "learning_rate": 1.660960394603221e-05, + "loss": 0.0225, + "step": 4774 + }, + { + "epoch": 0.5176712922810061, + "grad_norm": 12.09327507019043, + "learning_rate": 1.660887857246482e-05, + "loss": 0.0849, + "step": 4775 + }, + { + "epoch": 0.5177797051170858, + "grad_norm": 0.4588984549045563, + "learning_rate": 1.6608153198897433e-05, + "loss": 0.0137, + "step": 4776 + }, + { + "epoch": 0.5178881179531657, + "grad_norm": 1.96686589717865, + "learning_rate": 1.6607427825330046e-05, + "loss": 0.0247, + "step": 4777 + }, + { + "epoch": 0.5179965307892455, + "grad_norm": 2.013380765914917, + "learning_rate": 1.660670245176266e-05, + "loss": 0.0685, + "step": 4778 + }, + { + "epoch": 0.5181049436253252, + "grad_norm": 4.478318214416504, + "learning_rate": 1.6605977078195273e-05, + "loss": 0.0673, + "step": 4779 + }, + { + "epoch": 0.518213356461405, + "grad_norm": 2.170880079269409, + "learning_rate": 1.6605251704627885e-05, + "loss": 0.0592, + "step": 4780 + }, + { + "epoch": 0.5183217692974849, + "grad_norm": 2.6902265548706055, + "learning_rate": 1.6604526331060498e-05, + "loss": 0.0584, + "step": 4781 + }, + { + "epoch": 0.5184301821335646, + "grad_norm": 1.8979532718658447, + "learning_rate": 1.660380095749311e-05, + "loss": 0.0233, + "step": 4782 + }, + { + "epoch": 0.5185385949696444, + "grad_norm": 0.8573009967803955, + "learning_rate": 1.6603075583925722e-05, + "loss": 0.0143, + "step": 4783 + }, + { + "epoch": 0.5186470078057241, + "grad_norm": 0.5299654603004456, + "learning_rate": 1.6602350210358334e-05, + "loss": 0.006, + "step": 4784 + }, + { + "epoch": 0.518755420641804, + "grad_norm": 3.2769744396209717, + "learning_rate": 1.660162483679095e-05, + "loss": 0.0435, + "step": 4785 + }, + { + "epoch": 0.5188638334778838, + "grad_norm": 3.4057021141052246, + "learning_rate": 1.660089946322356e-05, + "loss": 0.0487, + "step": 4786 + }, + { + "epoch": 0.5189722463139635, + "grad_norm": 3.6929261684417725, + "learning_rate": 1.6600174089656174e-05, + "loss": 0.0882, + "step": 4787 + }, + { + "epoch": 0.5190806591500433, + "grad_norm": 3.460705518722534, + "learning_rate": 1.6599448716088786e-05, + "loss": 0.0459, + "step": 4788 + }, + { + "epoch": 0.5191890719861232, + "grad_norm": 3.8103108406066895, + "learning_rate": 1.6598723342521398e-05, + "loss": 0.0474, + "step": 4789 + }, + { + "epoch": 0.519297484822203, + "grad_norm": 0.968204140663147, + "learning_rate": 1.659799796895401e-05, + "loss": 0.0145, + "step": 4790 + }, + { + "epoch": 0.5194058976582827, + "grad_norm": 2.8407886028289795, + "learning_rate": 1.6597272595386626e-05, + "loss": 0.057, + "step": 4791 + }, + { + "epoch": 0.5195143104943626, + "grad_norm": 3.8856728076934814, + "learning_rate": 1.6596547221819238e-05, + "loss": 0.0252, + "step": 4792 + }, + { + "epoch": 0.5196227233304423, + "grad_norm": 4.609353065490723, + "learning_rate": 1.6595821848251853e-05, + "loss": 0.0432, + "step": 4793 + }, + { + "epoch": 0.5197311361665221, + "grad_norm": 3.5084729194641113, + "learning_rate": 1.6595096474684466e-05, + "loss": 0.0481, + "step": 4794 + }, + { + "epoch": 0.5198395490026019, + "grad_norm": 2.0183358192443848, + "learning_rate": 1.6594371101117078e-05, + "loss": 0.0357, + "step": 4795 + }, + { + "epoch": 0.5199479618386817, + "grad_norm": 2.951747417449951, + "learning_rate": 1.659364572754969e-05, + "loss": 0.028, + "step": 4796 + }, + { + "epoch": 0.5200563746747615, + "grad_norm": 3.5161375999450684, + "learning_rate": 1.6592920353982302e-05, + "loss": 0.0309, + "step": 4797 + }, + { + "epoch": 0.5201647875108413, + "grad_norm": 8.435470581054688, + "learning_rate": 1.6592194980414914e-05, + "loss": 0.0718, + "step": 4798 + }, + { + "epoch": 0.520273200346921, + "grad_norm": 3.2962028980255127, + "learning_rate": 1.6591469606847526e-05, + "loss": 0.0446, + "step": 4799 + }, + { + "epoch": 0.5203816131830009, + "grad_norm": 10.13066291809082, + "learning_rate": 1.6590744233280142e-05, + "loss": 0.024, + "step": 4800 + }, + { + "epoch": 0.5204900260190807, + "grad_norm": 3.4040472507476807, + "learning_rate": 1.6590018859712754e-05, + "loss": 0.0402, + "step": 4801 + }, + { + "epoch": 0.5205984388551604, + "grad_norm": 1.4319756031036377, + "learning_rate": 1.6589293486145366e-05, + "loss": 0.0162, + "step": 4802 + }, + { + "epoch": 0.5207068516912402, + "grad_norm": 2.4625439643859863, + "learning_rate": 1.6588568112577978e-05, + "loss": 0.0249, + "step": 4803 + }, + { + "epoch": 0.5208152645273201, + "grad_norm": 4.685250282287598, + "learning_rate": 1.658784273901059e-05, + "loss": 0.0433, + "step": 4804 + }, + { + "epoch": 0.5209236773633998, + "grad_norm": 3.0452678203582764, + "learning_rate": 1.6587117365443203e-05, + "loss": 0.0629, + "step": 4805 + }, + { + "epoch": 0.5210320901994796, + "grad_norm": 1.7407636642456055, + "learning_rate": 1.6586391991875818e-05, + "loss": 0.0515, + "step": 4806 + }, + { + "epoch": 0.5211405030355594, + "grad_norm": 4.385132312774658, + "learning_rate": 1.658566661830843e-05, + "loss": 0.0552, + "step": 4807 + }, + { + "epoch": 0.5212489158716392, + "grad_norm": 4.198019027709961, + "learning_rate": 1.6584941244741042e-05, + "loss": 0.0182, + "step": 4808 + }, + { + "epoch": 0.521357328707719, + "grad_norm": 0.3213529884815216, + "learning_rate": 1.6584215871173658e-05, + "loss": 0.0041, + "step": 4809 + }, + { + "epoch": 0.5214657415437988, + "grad_norm": 1.8104233741760254, + "learning_rate": 1.658349049760627e-05, + "loss": 0.0342, + "step": 4810 + }, + { + "epoch": 0.5215741543798785, + "grad_norm": 2.2957065105438232, + "learning_rate": 1.6582765124038882e-05, + "loss": 0.03, + "step": 4811 + }, + { + "epoch": 0.5216825672159584, + "grad_norm": 2.1644163131713867, + "learning_rate": 1.6582039750471494e-05, + "loss": 0.0624, + "step": 4812 + }, + { + "epoch": 0.5217909800520382, + "grad_norm": 3.887728452682495, + "learning_rate": 1.6581314376904106e-05, + "loss": 0.0627, + "step": 4813 + }, + { + "epoch": 0.5218993928881179, + "grad_norm": 2.1037867069244385, + "learning_rate": 1.658058900333672e-05, + "loss": 0.0481, + "step": 4814 + }, + { + "epoch": 0.5220078057241977, + "grad_norm": 5.508378505706787, + "learning_rate": 1.6579863629769334e-05, + "loss": 0.1029, + "step": 4815 + }, + { + "epoch": 0.5221162185602776, + "grad_norm": 2.1386215686798096, + "learning_rate": 1.6579138256201946e-05, + "loss": 0.0485, + "step": 4816 + }, + { + "epoch": 0.5222246313963573, + "grad_norm": 2.593317747116089, + "learning_rate": 1.657841288263456e-05, + "loss": 0.0587, + "step": 4817 + }, + { + "epoch": 0.5223330442324371, + "grad_norm": 2.295620918273926, + "learning_rate": 1.657768750906717e-05, + "loss": 0.0223, + "step": 4818 + }, + { + "epoch": 0.522441457068517, + "grad_norm": 2.5367565155029297, + "learning_rate": 1.6576962135499783e-05, + "loss": 0.0472, + "step": 4819 + }, + { + "epoch": 0.5225498699045967, + "grad_norm": 1.8952913284301758, + "learning_rate": 1.6576236761932395e-05, + "loss": 0.0514, + "step": 4820 + }, + { + "epoch": 0.5226582827406765, + "grad_norm": 1.685128927230835, + "learning_rate": 1.657551138836501e-05, + "loss": 0.0335, + "step": 4821 + }, + { + "epoch": 0.5227666955767563, + "grad_norm": 1.1597479581832886, + "learning_rate": 1.6574786014797623e-05, + "loss": 0.0191, + "step": 4822 + }, + { + "epoch": 0.5228751084128361, + "grad_norm": 2.367415428161621, + "learning_rate": 1.6574060641230235e-05, + "loss": 0.0417, + "step": 4823 + }, + { + "epoch": 0.5229835212489159, + "grad_norm": 2.489769697189331, + "learning_rate": 1.6573335267662847e-05, + "loss": 0.0213, + "step": 4824 + }, + { + "epoch": 0.5230919340849957, + "grad_norm": 3.1484158039093018, + "learning_rate": 1.6572609894095462e-05, + "loss": 0.0794, + "step": 4825 + }, + { + "epoch": 0.5232003469210754, + "grad_norm": 3.9753544330596924, + "learning_rate": 1.6571884520528074e-05, + "loss": 0.0787, + "step": 4826 + }, + { + "epoch": 0.5233087597571553, + "grad_norm": 0.8699100613594055, + "learning_rate": 1.6571159146960687e-05, + "loss": 0.0151, + "step": 4827 + }, + { + "epoch": 0.5234171725932351, + "grad_norm": 3.6216604709625244, + "learning_rate": 1.65704337733933e-05, + "loss": 0.0354, + "step": 4828 + }, + { + "epoch": 0.5235255854293148, + "grad_norm": 7.55878210067749, + "learning_rate": 1.656970839982591e-05, + "loss": 0.0465, + "step": 4829 + }, + { + "epoch": 0.5236339982653946, + "grad_norm": 1.896962285041809, + "learning_rate": 1.6568983026258526e-05, + "loss": 0.0362, + "step": 4830 + }, + { + "epoch": 0.5237424111014745, + "grad_norm": 1.6966056823730469, + "learning_rate": 1.656825765269114e-05, + "loss": 0.0158, + "step": 4831 + }, + { + "epoch": 0.5238508239375542, + "grad_norm": 1.2342612743377686, + "learning_rate": 1.656753227912375e-05, + "loss": 0.0201, + "step": 4832 + }, + { + "epoch": 0.523959236773634, + "grad_norm": 1.7716922760009766, + "learning_rate": 1.6566806905556363e-05, + "loss": 0.0334, + "step": 4833 + }, + { + "epoch": 0.5240676496097137, + "grad_norm": 2.185584306716919, + "learning_rate": 1.6566081531988975e-05, + "loss": 0.0707, + "step": 4834 + }, + { + "epoch": 0.5241760624457936, + "grad_norm": 0.7794736623764038, + "learning_rate": 1.6565356158421587e-05, + "loss": 0.0104, + "step": 4835 + }, + { + "epoch": 0.5242844752818734, + "grad_norm": 0.9973711371421814, + "learning_rate": 1.65646307848542e-05, + "loss": 0.026, + "step": 4836 + }, + { + "epoch": 0.5243928881179531, + "grad_norm": 4.400254726409912, + "learning_rate": 1.6563905411286815e-05, + "loss": 0.041, + "step": 4837 + }, + { + "epoch": 0.5245013009540329, + "grad_norm": 2.6201436519622803, + "learning_rate": 1.6563180037719427e-05, + "loss": 0.0325, + "step": 4838 + }, + { + "epoch": 0.5246097137901128, + "grad_norm": 1.913550615310669, + "learning_rate": 1.656245466415204e-05, + "loss": 0.0273, + "step": 4839 + }, + { + "epoch": 0.5247181266261925, + "grad_norm": 5.119163513183594, + "learning_rate": 1.656172929058465e-05, + "loss": 0.0373, + "step": 4840 + }, + { + "epoch": 0.5248265394622723, + "grad_norm": 1.8012847900390625, + "learning_rate": 1.6561003917017263e-05, + "loss": 0.0372, + "step": 4841 + }, + { + "epoch": 0.5249349522983521, + "grad_norm": 0.41003304719924927, + "learning_rate": 1.6560278543449876e-05, + "loss": 0.0048, + "step": 4842 + }, + { + "epoch": 0.525043365134432, + "grad_norm": 4.8848042488098145, + "learning_rate": 1.655955316988249e-05, + "loss": 0.0401, + "step": 4843 + }, + { + "epoch": 0.5251517779705117, + "grad_norm": 3.4188504219055176, + "learning_rate": 1.6558827796315103e-05, + "loss": 0.0288, + "step": 4844 + }, + { + "epoch": 0.5252601908065915, + "grad_norm": 3.8730359077453613, + "learning_rate": 1.655810242274772e-05, + "loss": 0.0494, + "step": 4845 + }, + { + "epoch": 0.5253686036426712, + "grad_norm": 1.5227220058441162, + "learning_rate": 1.655737704918033e-05, + "loss": 0.0125, + "step": 4846 + }, + { + "epoch": 0.5254770164787511, + "grad_norm": 1.2049561738967896, + "learning_rate": 1.6556651675612943e-05, + "loss": 0.0196, + "step": 4847 + }, + { + "epoch": 0.5255854293148309, + "grad_norm": 5.011872291564941, + "learning_rate": 1.6555926302045555e-05, + "loss": 0.0484, + "step": 4848 + }, + { + "epoch": 0.5256938421509106, + "grad_norm": 5.916911602020264, + "learning_rate": 1.6555200928478167e-05, + "loss": 0.0278, + "step": 4849 + }, + { + "epoch": 0.5258022549869905, + "grad_norm": 3.7479052543640137, + "learning_rate": 1.655447555491078e-05, + "loss": 0.0449, + "step": 4850 + }, + { + "epoch": 0.5259106678230703, + "grad_norm": 5.909242153167725, + "learning_rate": 1.655375018134339e-05, + "loss": 0.0872, + "step": 4851 + }, + { + "epoch": 0.52601908065915, + "grad_norm": 0.8439807295799255, + "learning_rate": 1.6553024807776007e-05, + "loss": 0.0088, + "step": 4852 + }, + { + "epoch": 0.5261274934952298, + "grad_norm": 4.674795150756836, + "learning_rate": 1.655229943420862e-05, + "loss": 0.0912, + "step": 4853 + }, + { + "epoch": 0.5262359063313097, + "grad_norm": 6.55457067489624, + "learning_rate": 1.655157406064123e-05, + "loss": 0.0558, + "step": 4854 + }, + { + "epoch": 0.5263443191673894, + "grad_norm": 1.5918248891830444, + "learning_rate": 1.6550848687073844e-05, + "loss": 0.0263, + "step": 4855 + }, + { + "epoch": 0.5264527320034692, + "grad_norm": 1.0178965330123901, + "learning_rate": 1.6550123313506456e-05, + "loss": 0.0128, + "step": 4856 + }, + { + "epoch": 0.526561144839549, + "grad_norm": 4.272635459899902, + "learning_rate": 1.6549397939939068e-05, + "loss": 0.0684, + "step": 4857 + }, + { + "epoch": 0.5266695576756288, + "grad_norm": 7.55299186706543, + "learning_rate": 1.6548672566371683e-05, + "loss": 0.1162, + "step": 4858 + }, + { + "epoch": 0.5267779705117086, + "grad_norm": 2.3208258152008057, + "learning_rate": 1.6547947192804296e-05, + "loss": 0.044, + "step": 4859 + }, + { + "epoch": 0.5268863833477884, + "grad_norm": 1.4327540397644043, + "learning_rate": 1.654722181923691e-05, + "loss": 0.0185, + "step": 4860 + }, + { + "epoch": 0.5269947961838681, + "grad_norm": 3.840554714202881, + "learning_rate": 1.6546496445669523e-05, + "loss": 0.0471, + "step": 4861 + }, + { + "epoch": 0.527103209019948, + "grad_norm": 2.5064499378204346, + "learning_rate": 1.6545771072102135e-05, + "loss": 0.0265, + "step": 4862 + }, + { + "epoch": 0.5272116218560278, + "grad_norm": 1.5307093858718872, + "learning_rate": 1.6545045698534748e-05, + "loss": 0.0124, + "step": 4863 + }, + { + "epoch": 0.5273200346921075, + "grad_norm": 3.0895495414733887, + "learning_rate": 1.654432032496736e-05, + "loss": 0.0583, + "step": 4864 + }, + { + "epoch": 0.5274284475281873, + "grad_norm": 5.07768440246582, + "learning_rate": 1.6543594951399972e-05, + "loss": 0.0255, + "step": 4865 + }, + { + "epoch": 0.5275368603642672, + "grad_norm": 3.6311533451080322, + "learning_rate": 1.6542869577832584e-05, + "loss": 0.0213, + "step": 4866 + }, + { + "epoch": 0.5276452732003469, + "grad_norm": 4.378561496734619, + "learning_rate": 1.65421442042652e-05, + "loss": 0.0682, + "step": 4867 + }, + { + "epoch": 0.5277536860364267, + "grad_norm": 4.437617778778076, + "learning_rate": 1.654141883069781e-05, + "loss": 0.0588, + "step": 4868 + }, + { + "epoch": 0.5278620988725065, + "grad_norm": 1.5902681350708008, + "learning_rate": 1.6540693457130424e-05, + "loss": 0.0307, + "step": 4869 + }, + { + "epoch": 0.5279705117085863, + "grad_norm": 1.4627171754837036, + "learning_rate": 1.6539968083563036e-05, + "loss": 0.0079, + "step": 4870 + }, + { + "epoch": 0.5280789245446661, + "grad_norm": 3.507539749145508, + "learning_rate": 1.6539242709995648e-05, + "loss": 0.0628, + "step": 4871 + }, + { + "epoch": 0.5281873373807459, + "grad_norm": 8.701287269592285, + "learning_rate": 1.653851733642826e-05, + "loss": 0.0618, + "step": 4872 + }, + { + "epoch": 0.5282957502168256, + "grad_norm": 2.3879051208496094, + "learning_rate": 1.6537791962860876e-05, + "loss": 0.0383, + "step": 4873 + }, + { + "epoch": 0.5284041630529055, + "grad_norm": 2.8745789527893066, + "learning_rate": 1.6537066589293488e-05, + "loss": 0.0397, + "step": 4874 + }, + { + "epoch": 0.5285125758889853, + "grad_norm": 2.97296142578125, + "learning_rate": 1.65363412157261e-05, + "loss": 0.0442, + "step": 4875 + }, + { + "epoch": 0.528620988725065, + "grad_norm": 8.687459945678711, + "learning_rate": 1.6535615842158712e-05, + "loss": 0.0306, + "step": 4876 + }, + { + "epoch": 0.5287294015611448, + "grad_norm": 4.324615001678467, + "learning_rate": 1.6534890468591328e-05, + "loss": 0.0666, + "step": 4877 + }, + { + "epoch": 0.5288378143972247, + "grad_norm": 1.7031513452529907, + "learning_rate": 1.653416509502394e-05, + "loss": 0.0525, + "step": 4878 + }, + { + "epoch": 0.5289462272333044, + "grad_norm": 4.269677639007568, + "learning_rate": 1.6533439721456552e-05, + "loss": 0.0369, + "step": 4879 + }, + { + "epoch": 0.5290546400693842, + "grad_norm": 3.375741720199585, + "learning_rate": 1.6532714347889164e-05, + "loss": 0.0918, + "step": 4880 + }, + { + "epoch": 0.5291630529054641, + "grad_norm": 5.261469841003418, + "learning_rate": 1.6531988974321776e-05, + "loss": 0.0466, + "step": 4881 + }, + { + "epoch": 0.5292714657415438, + "grad_norm": 4.017150402069092, + "learning_rate": 1.6531263600754392e-05, + "loss": 0.0833, + "step": 4882 + }, + { + "epoch": 0.5293798785776236, + "grad_norm": 3.329902172088623, + "learning_rate": 1.6530538227187004e-05, + "loss": 0.0439, + "step": 4883 + }, + { + "epoch": 0.5294882914137033, + "grad_norm": 0.8147914409637451, + "learning_rate": 1.6529812853619616e-05, + "loss": 0.0202, + "step": 4884 + }, + { + "epoch": 0.5295967042497832, + "grad_norm": 1.7565065622329712, + "learning_rate": 1.6529087480052228e-05, + "loss": 0.078, + "step": 4885 + }, + { + "epoch": 0.529705117085863, + "grad_norm": 2.653200149536133, + "learning_rate": 1.652836210648484e-05, + "loss": 0.0265, + "step": 4886 + }, + { + "epoch": 0.5298135299219427, + "grad_norm": 0.47308990359306335, + "learning_rate": 1.6527636732917453e-05, + "loss": 0.0215, + "step": 4887 + }, + { + "epoch": 0.5299219427580225, + "grad_norm": 0.6556981801986694, + "learning_rate": 1.6526911359350065e-05, + "loss": 0.0086, + "step": 4888 + }, + { + "epoch": 0.5300303555941024, + "grad_norm": 1.0130906105041504, + "learning_rate": 1.652618598578268e-05, + "loss": 0.0116, + "step": 4889 + }, + { + "epoch": 0.5301387684301822, + "grad_norm": 1.0279003381729126, + "learning_rate": 1.6525460612215292e-05, + "loss": 0.0212, + "step": 4890 + }, + { + "epoch": 0.5302471812662619, + "grad_norm": 4.704721927642822, + "learning_rate": 1.6524735238647904e-05, + "loss": 0.0347, + "step": 4891 + }, + { + "epoch": 0.5303555941023417, + "grad_norm": 1.8154219388961792, + "learning_rate": 1.6524009865080517e-05, + "loss": 0.0298, + "step": 4892 + }, + { + "epoch": 0.5304640069384216, + "grad_norm": 3.2522366046905518, + "learning_rate": 1.652328449151313e-05, + "loss": 0.0445, + "step": 4893 + }, + { + "epoch": 0.5305724197745013, + "grad_norm": 0.7674040198326111, + "learning_rate": 1.652255911794574e-05, + "loss": 0.0154, + "step": 4894 + }, + { + "epoch": 0.5306808326105811, + "grad_norm": 1.559651255607605, + "learning_rate": 1.6521833744378356e-05, + "loss": 0.0233, + "step": 4895 + }, + { + "epoch": 0.5307892454466608, + "grad_norm": 1.9650150537490845, + "learning_rate": 1.652110837081097e-05, + "loss": 0.0385, + "step": 4896 + }, + { + "epoch": 0.5308976582827407, + "grad_norm": 4.374350070953369, + "learning_rate": 1.6520382997243584e-05, + "loss": 0.0116, + "step": 4897 + }, + { + "epoch": 0.5310060711188205, + "grad_norm": 2.8857901096343994, + "learning_rate": 1.6519657623676196e-05, + "loss": 0.0146, + "step": 4898 + }, + { + "epoch": 0.5311144839549002, + "grad_norm": 2.1296892166137695, + "learning_rate": 1.651893225010881e-05, + "loss": 0.0242, + "step": 4899 + }, + { + "epoch": 0.53122289679098, + "grad_norm": 3.420822858810425, + "learning_rate": 1.651820687654142e-05, + "loss": 0.0598, + "step": 4900 + }, + { + "epoch": 0.5313313096270599, + "grad_norm": 1.7484444379806519, + "learning_rate": 1.6517481502974033e-05, + "loss": 0.0177, + "step": 4901 + }, + { + "epoch": 0.5314397224631396, + "grad_norm": 1.7222148180007935, + "learning_rate": 1.6516756129406645e-05, + "loss": 0.0226, + "step": 4902 + }, + { + "epoch": 0.5315481352992194, + "grad_norm": 0.9290526509284973, + "learning_rate": 1.6516030755839257e-05, + "loss": 0.0193, + "step": 4903 + }, + { + "epoch": 0.5316565481352992, + "grad_norm": 3.174954652786255, + "learning_rate": 1.6515305382271873e-05, + "loss": 0.061, + "step": 4904 + }, + { + "epoch": 0.531764960971379, + "grad_norm": 1.749436616897583, + "learning_rate": 1.6514580008704485e-05, + "loss": 0.0491, + "step": 4905 + }, + { + "epoch": 0.5318733738074588, + "grad_norm": 7.4100189208984375, + "learning_rate": 1.6513854635137097e-05, + "loss": 0.0817, + "step": 4906 + }, + { + "epoch": 0.5319817866435386, + "grad_norm": 3.782010078430176, + "learning_rate": 1.651312926156971e-05, + "loss": 0.0322, + "step": 4907 + }, + { + "epoch": 0.5320901994796184, + "grad_norm": 2.3180363178253174, + "learning_rate": 1.651240388800232e-05, + "loss": 0.0463, + "step": 4908 + }, + { + "epoch": 0.5321986123156982, + "grad_norm": 2.9956722259521484, + "learning_rate": 1.6511678514434933e-05, + "loss": 0.0729, + "step": 4909 + }, + { + "epoch": 0.532307025151778, + "grad_norm": 1.3829094171524048, + "learning_rate": 1.651095314086755e-05, + "loss": 0.0424, + "step": 4910 + }, + { + "epoch": 0.5324154379878577, + "grad_norm": 4.629758834838867, + "learning_rate": 1.651022776730016e-05, + "loss": 0.0973, + "step": 4911 + }, + { + "epoch": 0.5325238508239376, + "grad_norm": 2.725048780441284, + "learning_rate": 1.6509502393732776e-05, + "loss": 0.0223, + "step": 4912 + }, + { + "epoch": 0.5326322636600174, + "grad_norm": 6.323185920715332, + "learning_rate": 1.650877702016539e-05, + "loss": 0.0431, + "step": 4913 + }, + { + "epoch": 0.5327406764960971, + "grad_norm": 7.450569152832031, + "learning_rate": 1.6508051646598e-05, + "loss": 0.0469, + "step": 4914 + }, + { + "epoch": 0.5328490893321769, + "grad_norm": 2.2239279747009277, + "learning_rate": 1.6507326273030613e-05, + "loss": 0.0332, + "step": 4915 + }, + { + "epoch": 0.5329575021682568, + "grad_norm": 2.8055977821350098, + "learning_rate": 1.6506600899463225e-05, + "loss": 0.0423, + "step": 4916 + }, + { + "epoch": 0.5330659150043365, + "grad_norm": 2.1815640926361084, + "learning_rate": 1.6505875525895837e-05, + "loss": 0.0378, + "step": 4917 + }, + { + "epoch": 0.5331743278404163, + "grad_norm": 4.577796459197998, + "learning_rate": 1.650515015232845e-05, + "loss": 0.0877, + "step": 4918 + }, + { + "epoch": 0.5332827406764961, + "grad_norm": 2.6450932025909424, + "learning_rate": 1.6504424778761065e-05, + "loss": 0.0417, + "step": 4919 + }, + { + "epoch": 0.5333911535125759, + "grad_norm": 4.25867223739624, + "learning_rate": 1.6503699405193677e-05, + "loss": 0.0386, + "step": 4920 + }, + { + "epoch": 0.5334995663486557, + "grad_norm": 1.3546167612075806, + "learning_rate": 1.650297403162629e-05, + "loss": 0.0441, + "step": 4921 + }, + { + "epoch": 0.5336079791847355, + "grad_norm": 3.309025764465332, + "learning_rate": 1.65022486580589e-05, + "loss": 0.0435, + "step": 4922 + }, + { + "epoch": 0.5337163920208152, + "grad_norm": 2.0384202003479004, + "learning_rate": 1.6501523284491513e-05, + "loss": 0.0499, + "step": 4923 + }, + { + "epoch": 0.5338248048568951, + "grad_norm": 3.816265821456909, + "learning_rate": 1.6500797910924126e-05, + "loss": 0.0918, + "step": 4924 + }, + { + "epoch": 0.5339332176929749, + "grad_norm": 0.636897087097168, + "learning_rate": 1.650007253735674e-05, + "loss": 0.0153, + "step": 4925 + }, + { + "epoch": 0.5340416305290546, + "grad_norm": 4.85003137588501, + "learning_rate": 1.6499347163789353e-05, + "loss": 0.0271, + "step": 4926 + }, + { + "epoch": 0.5341500433651344, + "grad_norm": 4.396016597747803, + "learning_rate": 1.6498621790221965e-05, + "loss": 0.0644, + "step": 4927 + }, + { + "epoch": 0.5342584562012143, + "grad_norm": 4.354980945587158, + "learning_rate": 1.6497896416654577e-05, + "loss": 0.0423, + "step": 4928 + }, + { + "epoch": 0.534366869037294, + "grad_norm": 1.3497886657714844, + "learning_rate": 1.6497171043087193e-05, + "loss": 0.032, + "step": 4929 + }, + { + "epoch": 0.5344752818733738, + "grad_norm": 5.850977897644043, + "learning_rate": 1.6496445669519805e-05, + "loss": 0.0732, + "step": 4930 + }, + { + "epoch": 0.5345836947094535, + "grad_norm": 1.2122230529785156, + "learning_rate": 1.6495720295952417e-05, + "loss": 0.0135, + "step": 4931 + }, + { + "epoch": 0.5346921075455334, + "grad_norm": 1.757696270942688, + "learning_rate": 1.649499492238503e-05, + "loss": 0.031, + "step": 4932 + }, + { + "epoch": 0.5348005203816132, + "grad_norm": 4.614891052246094, + "learning_rate": 1.649426954881764e-05, + "loss": 0.0735, + "step": 4933 + }, + { + "epoch": 0.534908933217693, + "grad_norm": 1.5283719301223755, + "learning_rate": 1.6493544175250257e-05, + "loss": 0.0211, + "step": 4934 + }, + { + "epoch": 0.5350173460537727, + "grad_norm": 7.806076526641846, + "learning_rate": 1.649281880168287e-05, + "loss": 0.0844, + "step": 4935 + }, + { + "epoch": 0.5351257588898526, + "grad_norm": 2.341137647628784, + "learning_rate": 1.649209342811548e-05, + "loss": 0.094, + "step": 4936 + }, + { + "epoch": 0.5352341717259324, + "grad_norm": 4.965423107147217, + "learning_rate": 1.6491368054548094e-05, + "loss": 0.0301, + "step": 4937 + }, + { + "epoch": 0.5353425845620121, + "grad_norm": 0.7510169744491577, + "learning_rate": 1.6490642680980706e-05, + "loss": 0.0178, + "step": 4938 + }, + { + "epoch": 0.535450997398092, + "grad_norm": 2.666245460510254, + "learning_rate": 1.6489917307413318e-05, + "loss": 0.018, + "step": 4939 + }, + { + "epoch": 0.5355594102341718, + "grad_norm": 6.333392143249512, + "learning_rate": 1.648919193384593e-05, + "loss": 0.048, + "step": 4940 + }, + { + "epoch": 0.5356678230702515, + "grad_norm": 1.240892767906189, + "learning_rate": 1.6488466560278546e-05, + "loss": 0.0146, + "step": 4941 + }, + { + "epoch": 0.5357762359063313, + "grad_norm": 1.194549798965454, + "learning_rate": 1.6487741186711158e-05, + "loss": 0.0267, + "step": 4942 + }, + { + "epoch": 0.5358846487424112, + "grad_norm": 0.8996971845626831, + "learning_rate": 1.648701581314377e-05, + "loss": 0.013, + "step": 4943 + }, + { + "epoch": 0.5359930615784909, + "grad_norm": 3.257582902908325, + "learning_rate": 1.6486290439576382e-05, + "loss": 0.0491, + "step": 4944 + }, + { + "epoch": 0.5361014744145707, + "grad_norm": 2.887737989425659, + "learning_rate": 1.6485565066008994e-05, + "loss": 0.0252, + "step": 4945 + }, + { + "epoch": 0.5362098872506504, + "grad_norm": 1.4927306175231934, + "learning_rate": 1.648483969244161e-05, + "loss": 0.0179, + "step": 4946 + }, + { + "epoch": 0.5363183000867303, + "grad_norm": 1.6840498447418213, + "learning_rate": 1.6484114318874222e-05, + "loss": 0.0299, + "step": 4947 + }, + { + "epoch": 0.5364267129228101, + "grad_norm": 3.166288137435913, + "learning_rate": 1.6483388945306834e-05, + "loss": 0.0559, + "step": 4948 + }, + { + "epoch": 0.5365351257588898, + "grad_norm": 1.9922114610671997, + "learning_rate": 1.648266357173945e-05, + "loss": 0.0354, + "step": 4949 + }, + { + "epoch": 0.5366435385949696, + "grad_norm": 3.4100501537323, + "learning_rate": 1.648193819817206e-05, + "loss": 0.051, + "step": 4950 + }, + { + "epoch": 0.5367519514310495, + "grad_norm": 2.925539970397949, + "learning_rate": 1.6481212824604674e-05, + "loss": 0.0593, + "step": 4951 + }, + { + "epoch": 0.5368603642671292, + "grad_norm": 4.829415321350098, + "learning_rate": 1.6480487451037286e-05, + "loss": 0.0519, + "step": 4952 + }, + { + "epoch": 0.536968777103209, + "grad_norm": 6.913331508636475, + "learning_rate": 1.6479762077469898e-05, + "loss": 0.0246, + "step": 4953 + }, + { + "epoch": 0.5370771899392888, + "grad_norm": 0.17233186960220337, + "learning_rate": 1.647903670390251e-05, + "loss": 0.0026, + "step": 4954 + }, + { + "epoch": 0.5371856027753686, + "grad_norm": 1.8297572135925293, + "learning_rate": 1.6478311330335122e-05, + "loss": 0.0347, + "step": 4955 + }, + { + "epoch": 0.5372940156114484, + "grad_norm": 2.2803401947021484, + "learning_rate": 1.6477585956767738e-05, + "loss": 0.0382, + "step": 4956 + }, + { + "epoch": 0.5374024284475282, + "grad_norm": 2.147195339202881, + "learning_rate": 1.647686058320035e-05, + "loss": 0.0272, + "step": 4957 + }, + { + "epoch": 0.5375108412836079, + "grad_norm": 1.813038945198059, + "learning_rate": 1.6476135209632962e-05, + "loss": 0.0412, + "step": 4958 + }, + { + "epoch": 0.5376192541196878, + "grad_norm": 1.9224393367767334, + "learning_rate": 1.6475409836065574e-05, + "loss": 0.0265, + "step": 4959 + }, + { + "epoch": 0.5377276669557676, + "grad_norm": 2.2844111919403076, + "learning_rate": 1.6474684462498186e-05, + "loss": 0.024, + "step": 4960 + }, + { + "epoch": 0.5378360797918473, + "grad_norm": 0.6510205864906311, + "learning_rate": 1.64739590889308e-05, + "loss": 0.0064, + "step": 4961 + }, + { + "epoch": 0.5379444926279271, + "grad_norm": 4.003460884094238, + "learning_rate": 1.6473233715363414e-05, + "loss": 0.0265, + "step": 4962 + }, + { + "epoch": 0.538052905464007, + "grad_norm": 2.8357064723968506, + "learning_rate": 1.6472508341796026e-05, + "loss": 0.0359, + "step": 4963 + }, + { + "epoch": 0.5381613183000867, + "grad_norm": 1.6654431819915771, + "learning_rate": 1.6471782968228642e-05, + "loss": 0.0236, + "step": 4964 + }, + { + "epoch": 0.5382697311361665, + "grad_norm": 1.5441248416900635, + "learning_rate": 1.6471057594661254e-05, + "loss": 0.0186, + "step": 4965 + }, + { + "epoch": 0.5383781439722463, + "grad_norm": 2.712049722671509, + "learning_rate": 1.6470332221093866e-05, + "loss": 0.0356, + "step": 4966 + }, + { + "epoch": 0.5384865568083261, + "grad_norm": 1.805452823638916, + "learning_rate": 1.6469606847526478e-05, + "loss": 0.0177, + "step": 4967 + }, + { + "epoch": 0.5385949696444059, + "grad_norm": 0.8680498600006104, + "learning_rate": 1.646888147395909e-05, + "loss": 0.0074, + "step": 4968 + }, + { + "epoch": 0.5387033824804857, + "grad_norm": 3.8047611713409424, + "learning_rate": 1.6468156100391702e-05, + "loss": 0.0766, + "step": 4969 + }, + { + "epoch": 0.5388117953165655, + "grad_norm": 5.279074192047119, + "learning_rate": 1.6467430726824315e-05, + "loss": 0.0524, + "step": 4970 + }, + { + "epoch": 0.5389202081526453, + "grad_norm": 2.386446714401245, + "learning_rate": 1.646670535325693e-05, + "loss": 0.0384, + "step": 4971 + }, + { + "epoch": 0.5390286209887251, + "grad_norm": 0.45868346095085144, + "learning_rate": 1.6465979979689542e-05, + "loss": 0.0072, + "step": 4972 + }, + { + "epoch": 0.5391370338248048, + "grad_norm": 3.755002737045288, + "learning_rate": 1.6465254606122154e-05, + "loss": 0.0339, + "step": 4973 + }, + { + "epoch": 0.5392454466608847, + "grad_norm": 2.7561533451080322, + "learning_rate": 1.6464529232554767e-05, + "loss": 0.0575, + "step": 4974 + }, + { + "epoch": 0.5393538594969645, + "grad_norm": 2.7678146362304688, + "learning_rate": 1.646380385898738e-05, + "loss": 0.0363, + "step": 4975 + }, + { + "epoch": 0.5394622723330442, + "grad_norm": 2.550419569015503, + "learning_rate": 1.646307848541999e-05, + "loss": 0.0735, + "step": 4976 + }, + { + "epoch": 0.539570685169124, + "grad_norm": 2.1886565685272217, + "learning_rate": 1.6462353111852606e-05, + "loss": 0.0236, + "step": 4977 + }, + { + "epoch": 0.5396790980052039, + "grad_norm": 2.961038112640381, + "learning_rate": 1.646162773828522e-05, + "loss": 0.0195, + "step": 4978 + }, + { + "epoch": 0.5397875108412836, + "grad_norm": 3.841688632965088, + "learning_rate": 1.646090236471783e-05, + "loss": 0.0794, + "step": 4979 + }, + { + "epoch": 0.5398959236773634, + "grad_norm": 1.3093112707138062, + "learning_rate": 1.6460176991150443e-05, + "loss": 0.0067, + "step": 4980 + }, + { + "epoch": 0.5400043365134432, + "grad_norm": 2.379230499267578, + "learning_rate": 1.645945161758306e-05, + "loss": 0.0619, + "step": 4981 + }, + { + "epoch": 0.540112749349523, + "grad_norm": 3.0533928871154785, + "learning_rate": 1.645872624401567e-05, + "loss": 0.068, + "step": 4982 + }, + { + "epoch": 0.5402211621856028, + "grad_norm": 1.6894444227218628, + "learning_rate": 1.6458000870448283e-05, + "loss": 0.0364, + "step": 4983 + }, + { + "epoch": 0.5403295750216826, + "grad_norm": 4.364635467529297, + "learning_rate": 1.6457275496880895e-05, + "loss": 0.0303, + "step": 4984 + }, + { + "epoch": 0.5404379878577623, + "grad_norm": 1.1810102462768555, + "learning_rate": 1.6456550123313507e-05, + "loss": 0.0226, + "step": 4985 + }, + { + "epoch": 0.5405464006938422, + "grad_norm": 3.7352449893951416, + "learning_rate": 1.6455824749746122e-05, + "loss": 0.0187, + "step": 4986 + }, + { + "epoch": 0.540654813529922, + "grad_norm": 1.8726212978363037, + "learning_rate": 1.6455099376178735e-05, + "loss": 0.0142, + "step": 4987 + }, + { + "epoch": 0.5407632263660017, + "grad_norm": 3.409898281097412, + "learning_rate": 1.6454374002611347e-05, + "loss": 0.042, + "step": 4988 + }, + { + "epoch": 0.5408716392020815, + "grad_norm": 0.3390190005302429, + "learning_rate": 1.645364862904396e-05, + "loss": 0.0063, + "step": 4989 + }, + { + "epoch": 0.5409800520381614, + "grad_norm": 0.4659437835216522, + "learning_rate": 1.645292325547657e-05, + "loss": 0.002, + "step": 4990 + }, + { + "epoch": 0.5410884648742411, + "grad_norm": 3.462942123413086, + "learning_rate": 1.6452197881909183e-05, + "loss": 0.0438, + "step": 4991 + }, + { + "epoch": 0.5411968777103209, + "grad_norm": 2.7886011600494385, + "learning_rate": 1.6451472508341795e-05, + "loss": 0.0327, + "step": 4992 + }, + { + "epoch": 0.5413052905464006, + "grad_norm": 3.864123821258545, + "learning_rate": 1.645074713477441e-05, + "loss": 0.0579, + "step": 4993 + }, + { + "epoch": 0.5414137033824805, + "grad_norm": 4.667160511016846, + "learning_rate": 1.6450021761207023e-05, + "loss": 0.0497, + "step": 4994 + }, + { + "epoch": 0.5415221162185603, + "grad_norm": 3.53991436958313, + "learning_rate": 1.6449296387639635e-05, + "loss": 0.0212, + "step": 4995 + }, + { + "epoch": 0.54163052905464, + "grad_norm": 2.8866500854492188, + "learning_rate": 1.6448571014072247e-05, + "loss": 0.0452, + "step": 4996 + }, + { + "epoch": 0.5417389418907199, + "grad_norm": 2.9916017055511475, + "learning_rate": 1.644784564050486e-05, + "loss": 0.0594, + "step": 4997 + }, + { + "epoch": 0.5418473547267997, + "grad_norm": 2.620950937271118, + "learning_rate": 1.6447120266937475e-05, + "loss": 0.0585, + "step": 4998 + }, + { + "epoch": 0.5419557675628794, + "grad_norm": 3.2498793601989746, + "learning_rate": 1.6446394893370087e-05, + "loss": 0.0376, + "step": 4999 + }, + { + "epoch": 0.5420641803989592, + "grad_norm": 3.2775585651397705, + "learning_rate": 1.64456695198027e-05, + "loss": 0.0294, + "step": 5000 + }, + { + "epoch": 0.5421725932350391, + "grad_norm": 2.2462852001190186, + "learning_rate": 1.6444944146235315e-05, + "loss": 0.0439, + "step": 5001 + }, + { + "epoch": 0.5422810060711188, + "grad_norm": 4.067817687988281, + "learning_rate": 1.6444218772667927e-05, + "loss": 0.0483, + "step": 5002 + }, + { + "epoch": 0.5423894189071986, + "grad_norm": 6.40342903137207, + "learning_rate": 1.644349339910054e-05, + "loss": 0.0304, + "step": 5003 + }, + { + "epoch": 0.5424978317432784, + "grad_norm": 1.6075190305709839, + "learning_rate": 1.644276802553315e-05, + "loss": 0.0304, + "step": 5004 + }, + { + "epoch": 0.5426062445793582, + "grad_norm": 2.3802013397216797, + "learning_rate": 1.6442042651965763e-05, + "loss": 0.0438, + "step": 5005 + }, + { + "epoch": 0.542714657415438, + "grad_norm": 7.647125244140625, + "learning_rate": 1.6441317278398376e-05, + "loss": 0.0779, + "step": 5006 + }, + { + "epoch": 0.5428230702515178, + "grad_norm": 3.0481510162353516, + "learning_rate": 1.6440591904830988e-05, + "loss": 0.0229, + "step": 5007 + }, + { + "epoch": 0.5429314830875975, + "grad_norm": 3.5447096824645996, + "learning_rate": 1.6439866531263603e-05, + "loss": 0.0345, + "step": 5008 + }, + { + "epoch": 0.5430398959236774, + "grad_norm": 1.9252361059188843, + "learning_rate": 1.6439141157696215e-05, + "loss": 0.0472, + "step": 5009 + }, + { + "epoch": 0.5431483087597572, + "grad_norm": 2.441843271255493, + "learning_rate": 1.6438415784128827e-05, + "loss": 0.0485, + "step": 5010 + }, + { + "epoch": 0.5432567215958369, + "grad_norm": 2.466804265975952, + "learning_rate": 1.643769041056144e-05, + "loss": 0.066, + "step": 5011 + }, + { + "epoch": 0.5433651344319167, + "grad_norm": 8.425455093383789, + "learning_rate": 1.6436965036994052e-05, + "loss": 0.0561, + "step": 5012 + }, + { + "epoch": 0.5434735472679966, + "grad_norm": 1.8130046129226685, + "learning_rate": 1.6436239663426664e-05, + "loss": 0.0145, + "step": 5013 + }, + { + "epoch": 0.5435819601040763, + "grad_norm": 2.4745473861694336, + "learning_rate": 1.643551428985928e-05, + "loss": 0.0217, + "step": 5014 + }, + { + "epoch": 0.5436903729401561, + "grad_norm": 2.1929643154144287, + "learning_rate": 1.643478891629189e-05, + "loss": 0.0425, + "step": 5015 + }, + { + "epoch": 0.5437987857762359, + "grad_norm": 3.8272812366485596, + "learning_rate": 1.6434063542724507e-05, + "loss": 0.0675, + "step": 5016 + }, + { + "epoch": 0.5439071986123157, + "grad_norm": 2.082502841949463, + "learning_rate": 1.643333816915712e-05, + "loss": 0.0785, + "step": 5017 + }, + { + "epoch": 0.5440156114483955, + "grad_norm": 0.7153574824333191, + "learning_rate": 1.643261279558973e-05, + "loss": 0.0072, + "step": 5018 + }, + { + "epoch": 0.5441240242844753, + "grad_norm": 3.827810525894165, + "learning_rate": 1.6431887422022344e-05, + "loss": 0.0883, + "step": 5019 + }, + { + "epoch": 0.544232437120555, + "grad_norm": 0.6846656203269958, + "learning_rate": 1.6431162048454956e-05, + "loss": 0.0157, + "step": 5020 + }, + { + "epoch": 0.5443408499566349, + "grad_norm": 0.9487529397010803, + "learning_rate": 1.6430436674887568e-05, + "loss": 0.0183, + "step": 5021 + }, + { + "epoch": 0.5444492627927147, + "grad_norm": 1.7223552465438843, + "learning_rate": 1.642971130132018e-05, + "loss": 0.034, + "step": 5022 + }, + { + "epoch": 0.5445576756287944, + "grad_norm": 2.745596408843994, + "learning_rate": 1.6428985927752795e-05, + "loss": 0.0568, + "step": 5023 + }, + { + "epoch": 0.5446660884648742, + "grad_norm": 1.1889370679855347, + "learning_rate": 1.6428260554185408e-05, + "loss": 0.0236, + "step": 5024 + }, + { + "epoch": 0.5447745013009541, + "grad_norm": 2.710216760635376, + "learning_rate": 1.642753518061802e-05, + "loss": 0.0546, + "step": 5025 + }, + { + "epoch": 0.5448829141370338, + "grad_norm": 2.552889823913574, + "learning_rate": 1.6426809807050632e-05, + "loss": 0.0515, + "step": 5026 + }, + { + "epoch": 0.5449913269731136, + "grad_norm": 1.4534006118774414, + "learning_rate": 1.6426084433483244e-05, + "loss": 0.0224, + "step": 5027 + }, + { + "epoch": 0.5450997398091935, + "grad_norm": 0.8451440930366516, + "learning_rate": 1.6425359059915856e-05, + "loss": 0.0097, + "step": 5028 + }, + { + "epoch": 0.5452081526452732, + "grad_norm": 1.6322193145751953, + "learning_rate": 1.6424633686348472e-05, + "loss": 0.0346, + "step": 5029 + }, + { + "epoch": 0.545316565481353, + "grad_norm": 4.293122291564941, + "learning_rate": 1.6423908312781084e-05, + "loss": 0.0639, + "step": 5030 + }, + { + "epoch": 0.5454249783174328, + "grad_norm": 1.5230896472930908, + "learning_rate": 1.6423182939213696e-05, + "loss": 0.0272, + "step": 5031 + }, + { + "epoch": 0.5455333911535126, + "grad_norm": 2.8208494186401367, + "learning_rate": 1.6422457565646308e-05, + "loss": 0.0539, + "step": 5032 + }, + { + "epoch": 0.5456418039895924, + "grad_norm": 4.510250091552734, + "learning_rate": 1.6421732192078924e-05, + "loss": 0.0825, + "step": 5033 + }, + { + "epoch": 0.5457502168256722, + "grad_norm": 2.585862159729004, + "learning_rate": 1.6421006818511536e-05, + "loss": 0.0462, + "step": 5034 + }, + { + "epoch": 0.5458586296617519, + "grad_norm": 6.211108207702637, + "learning_rate": 1.6420281444944148e-05, + "loss": 0.079, + "step": 5035 + }, + { + "epoch": 0.5459670424978318, + "grad_norm": 2.2544026374816895, + "learning_rate": 1.641955607137676e-05, + "loss": 0.0342, + "step": 5036 + }, + { + "epoch": 0.5460754553339116, + "grad_norm": 1.320810079574585, + "learning_rate": 1.6418830697809372e-05, + "loss": 0.0283, + "step": 5037 + }, + { + "epoch": 0.5461838681699913, + "grad_norm": 2.111820936203003, + "learning_rate": 1.6418105324241988e-05, + "loss": 0.0474, + "step": 5038 + }, + { + "epoch": 0.5462922810060711, + "grad_norm": 2.8861303329467773, + "learning_rate": 1.64173799506746e-05, + "loss": 0.0559, + "step": 5039 + }, + { + "epoch": 0.546400693842151, + "grad_norm": 10.565464973449707, + "learning_rate": 1.6416654577107212e-05, + "loss": 0.1333, + "step": 5040 + }, + { + "epoch": 0.5465091066782307, + "grad_norm": 3.418355941772461, + "learning_rate": 1.6415929203539824e-05, + "loss": 0.0421, + "step": 5041 + }, + { + "epoch": 0.5466175195143105, + "grad_norm": 1.4750019311904907, + "learning_rate": 1.6415203829972436e-05, + "loss": 0.0312, + "step": 5042 + }, + { + "epoch": 0.5467259323503902, + "grad_norm": 2.2409534454345703, + "learning_rate": 1.641447845640505e-05, + "loss": 0.0384, + "step": 5043 + }, + { + "epoch": 0.5468343451864701, + "grad_norm": 1.763737678527832, + "learning_rate": 1.641375308283766e-05, + "loss": 0.0358, + "step": 5044 + }, + { + "epoch": 0.5469427580225499, + "grad_norm": 3.259178638458252, + "learning_rate": 1.6413027709270276e-05, + "loss": 0.0497, + "step": 5045 + }, + { + "epoch": 0.5470511708586296, + "grad_norm": 2.2667486667633057, + "learning_rate": 1.641230233570289e-05, + "loss": 0.0369, + "step": 5046 + }, + { + "epoch": 0.5471595836947094, + "grad_norm": 3.6723971366882324, + "learning_rate": 1.64115769621355e-05, + "loss": 0.0648, + "step": 5047 + }, + { + "epoch": 0.5472679965307893, + "grad_norm": 1.6277600526809692, + "learning_rate": 1.6410851588568113e-05, + "loss": 0.033, + "step": 5048 + }, + { + "epoch": 0.547376409366869, + "grad_norm": 2.485616683959961, + "learning_rate": 1.6410126215000725e-05, + "loss": 0.0664, + "step": 5049 + }, + { + "epoch": 0.5474848222029488, + "grad_norm": 2.6926419734954834, + "learning_rate": 1.640940084143334e-05, + "loss": 0.0418, + "step": 5050 + }, + { + "epoch": 0.5475932350390286, + "grad_norm": 3.6088998317718506, + "learning_rate": 1.6408675467865952e-05, + "loss": 0.0565, + "step": 5051 + }, + { + "epoch": 0.5477016478751084, + "grad_norm": 3.3768372535705566, + "learning_rate": 1.6407950094298565e-05, + "loss": 0.0787, + "step": 5052 + }, + { + "epoch": 0.5478100607111882, + "grad_norm": 2.0946993827819824, + "learning_rate": 1.640722472073118e-05, + "loss": 0.0316, + "step": 5053 + }, + { + "epoch": 0.547918473547268, + "grad_norm": 1.554484248161316, + "learning_rate": 1.6406499347163792e-05, + "loss": 0.0372, + "step": 5054 + }, + { + "epoch": 0.5480268863833477, + "grad_norm": 3.4499869346618652, + "learning_rate": 1.6405773973596404e-05, + "loss": 0.0552, + "step": 5055 + }, + { + "epoch": 0.5481352992194276, + "grad_norm": 2.273040771484375, + "learning_rate": 1.6405048600029017e-05, + "loss": 0.036, + "step": 5056 + }, + { + "epoch": 0.5482437120555074, + "grad_norm": 0.8583455681800842, + "learning_rate": 1.640432322646163e-05, + "loss": 0.0142, + "step": 5057 + }, + { + "epoch": 0.5483521248915871, + "grad_norm": 1.9078741073608398, + "learning_rate": 1.640359785289424e-05, + "loss": 0.0518, + "step": 5058 + }, + { + "epoch": 0.548460537727667, + "grad_norm": 2.5157339572906494, + "learning_rate": 1.6402872479326853e-05, + "loss": 0.0173, + "step": 5059 + }, + { + "epoch": 0.5485689505637468, + "grad_norm": 3.4767160415649414, + "learning_rate": 1.640214710575947e-05, + "loss": 0.0809, + "step": 5060 + }, + { + "epoch": 0.5486773633998265, + "grad_norm": 2.1386306285858154, + "learning_rate": 1.640142173219208e-05, + "loss": 0.0216, + "step": 5061 + }, + { + "epoch": 0.5487857762359063, + "grad_norm": 1.0456316471099854, + "learning_rate": 1.6400696358624693e-05, + "loss": 0.016, + "step": 5062 + }, + { + "epoch": 0.5488941890719862, + "grad_norm": 1.7603209018707275, + "learning_rate": 1.6399970985057305e-05, + "loss": 0.0254, + "step": 5063 + }, + { + "epoch": 0.5490026019080659, + "grad_norm": 3.1958119869232178, + "learning_rate": 1.6399245611489917e-05, + "loss": 0.0838, + "step": 5064 + }, + { + "epoch": 0.5491110147441457, + "grad_norm": 3.643941640853882, + "learning_rate": 1.639852023792253e-05, + "loss": 0.0929, + "step": 5065 + }, + { + "epoch": 0.5492194275802255, + "grad_norm": 2.3742401599884033, + "learning_rate": 1.6397794864355145e-05, + "loss": 0.0357, + "step": 5066 + }, + { + "epoch": 0.5493278404163053, + "grad_norm": 1.3566210269927979, + "learning_rate": 1.6397069490787757e-05, + "loss": 0.0212, + "step": 5067 + }, + { + "epoch": 0.5494362532523851, + "grad_norm": 1.5444239377975464, + "learning_rate": 1.6396344117220372e-05, + "loss": 0.0453, + "step": 5068 + }, + { + "epoch": 0.5495446660884649, + "grad_norm": 5.284215927124023, + "learning_rate": 1.6395618743652985e-05, + "loss": 0.0482, + "step": 5069 + }, + { + "epoch": 0.5496530789245446, + "grad_norm": 0.7587088346481323, + "learning_rate": 1.6394893370085597e-05, + "loss": 0.0048, + "step": 5070 + }, + { + "epoch": 0.5497614917606245, + "grad_norm": 3.3856282234191895, + "learning_rate": 1.639416799651821e-05, + "loss": 0.0377, + "step": 5071 + }, + { + "epoch": 0.5498699045967043, + "grad_norm": 2.63958740234375, + "learning_rate": 1.639344262295082e-05, + "loss": 0.0276, + "step": 5072 + }, + { + "epoch": 0.549978317432784, + "grad_norm": 1.6280609369277954, + "learning_rate": 1.6392717249383433e-05, + "loss": 0.03, + "step": 5073 + }, + { + "epoch": 0.5500867302688638, + "grad_norm": 3.5562329292297363, + "learning_rate": 1.6391991875816045e-05, + "loss": 0.0523, + "step": 5074 + }, + { + "epoch": 0.5501951431049437, + "grad_norm": 1.128608226776123, + "learning_rate": 1.639126650224866e-05, + "loss": 0.0217, + "step": 5075 + }, + { + "epoch": 0.5503035559410234, + "grad_norm": 3.8859357833862305, + "learning_rate": 1.6390541128681273e-05, + "loss": 0.0672, + "step": 5076 + }, + { + "epoch": 0.5504119687771032, + "grad_norm": 1.8327172994613647, + "learning_rate": 1.6389815755113885e-05, + "loss": 0.0413, + "step": 5077 + }, + { + "epoch": 0.550520381613183, + "grad_norm": 2.69262957572937, + "learning_rate": 1.6389090381546497e-05, + "loss": 0.0347, + "step": 5078 + }, + { + "epoch": 0.5506287944492628, + "grad_norm": 1.7447338104248047, + "learning_rate": 1.638836500797911e-05, + "loss": 0.0241, + "step": 5079 + }, + { + "epoch": 0.5507372072853426, + "grad_norm": 1.538925290107727, + "learning_rate": 1.638763963441172e-05, + "loss": 0.0175, + "step": 5080 + }, + { + "epoch": 0.5508456201214224, + "grad_norm": 1.7914605140686035, + "learning_rate": 1.6386914260844337e-05, + "loss": 0.014, + "step": 5081 + }, + { + "epoch": 0.5509540329575021, + "grad_norm": 5.365731716156006, + "learning_rate": 1.638618888727695e-05, + "loss": 0.0769, + "step": 5082 + }, + { + "epoch": 0.551062445793582, + "grad_norm": 3.5401172637939453, + "learning_rate": 1.638546351370956e-05, + "loss": 0.0446, + "step": 5083 + }, + { + "epoch": 0.5511708586296618, + "grad_norm": 3.704366445541382, + "learning_rate": 1.6384738140142174e-05, + "loss": 0.0925, + "step": 5084 + }, + { + "epoch": 0.5512792714657415, + "grad_norm": 2.6563239097595215, + "learning_rate": 1.638401276657479e-05, + "loss": 0.0734, + "step": 5085 + }, + { + "epoch": 0.5513876843018214, + "grad_norm": 1.6279473304748535, + "learning_rate": 1.63832873930074e-05, + "loss": 0.027, + "step": 5086 + }, + { + "epoch": 0.5514960971379012, + "grad_norm": 8.203385353088379, + "learning_rate": 1.6382562019440013e-05, + "loss": 0.0487, + "step": 5087 + }, + { + "epoch": 0.5516045099739809, + "grad_norm": 3.023491382598877, + "learning_rate": 1.6381836645872625e-05, + "loss": 0.1068, + "step": 5088 + }, + { + "epoch": 0.5517129228100607, + "grad_norm": 3.383204460144043, + "learning_rate": 1.6381111272305238e-05, + "loss": 0.0534, + "step": 5089 + }, + { + "epoch": 0.5518213356461406, + "grad_norm": 4.864977836608887, + "learning_rate": 1.6380385898737853e-05, + "loss": 0.0674, + "step": 5090 + }, + { + "epoch": 0.5519297484822203, + "grad_norm": 2.6292500495910645, + "learning_rate": 1.6379660525170465e-05, + "loss": 0.0804, + "step": 5091 + }, + { + "epoch": 0.5520381613183001, + "grad_norm": 1.3484959602355957, + "learning_rate": 1.6378935151603077e-05, + "loss": 0.0245, + "step": 5092 + }, + { + "epoch": 0.5521465741543798, + "grad_norm": 1.693988561630249, + "learning_rate": 1.637820977803569e-05, + "loss": 0.028, + "step": 5093 + }, + { + "epoch": 0.5522549869904597, + "grad_norm": 2.291281223297119, + "learning_rate": 1.6377484404468302e-05, + "loss": 0.0347, + "step": 5094 + }, + { + "epoch": 0.5523633998265395, + "grad_norm": 3.6909115314483643, + "learning_rate": 1.6376759030900914e-05, + "loss": 0.0401, + "step": 5095 + }, + { + "epoch": 0.5524718126626192, + "grad_norm": 1.3233542442321777, + "learning_rate": 1.6376033657333526e-05, + "loss": 0.0253, + "step": 5096 + }, + { + "epoch": 0.552580225498699, + "grad_norm": 2.9225211143493652, + "learning_rate": 1.637530828376614e-05, + "loss": 0.0726, + "step": 5097 + }, + { + "epoch": 0.5526886383347789, + "grad_norm": 4.903811454772949, + "learning_rate": 1.6374582910198754e-05, + "loss": 0.095, + "step": 5098 + }, + { + "epoch": 0.5527970511708586, + "grad_norm": 4.626526355743408, + "learning_rate": 1.6373857536631366e-05, + "loss": 0.0294, + "step": 5099 + }, + { + "epoch": 0.5529054640069384, + "grad_norm": 3.2061917781829834, + "learning_rate": 1.6373132163063978e-05, + "loss": 0.0866, + "step": 5100 + }, + { + "epoch": 0.5530138768430182, + "grad_norm": 4.195862293243408, + "learning_rate": 1.637240678949659e-05, + "loss": 0.023, + "step": 5101 + }, + { + "epoch": 0.553122289679098, + "grad_norm": 1.3478672504425049, + "learning_rate": 1.6371681415929206e-05, + "loss": 0.0512, + "step": 5102 + }, + { + "epoch": 0.5532307025151778, + "grad_norm": 2.9004650115966797, + "learning_rate": 1.6370956042361818e-05, + "loss": 0.0768, + "step": 5103 + }, + { + "epoch": 0.5533391153512576, + "grad_norm": 1.502955436706543, + "learning_rate": 1.637023066879443e-05, + "loss": 0.0423, + "step": 5104 + }, + { + "epoch": 0.5534475281873373, + "grad_norm": 2.309342622756958, + "learning_rate": 1.6369505295227045e-05, + "loss": 0.0365, + "step": 5105 + }, + { + "epoch": 0.5535559410234172, + "grad_norm": 3.084690809249878, + "learning_rate": 1.6368779921659658e-05, + "loss": 0.0235, + "step": 5106 + }, + { + "epoch": 0.553664353859497, + "grad_norm": 3.2023980617523193, + "learning_rate": 1.636805454809227e-05, + "loss": 0.0804, + "step": 5107 + }, + { + "epoch": 0.5537727666955767, + "grad_norm": 1.6171826124191284, + "learning_rate": 1.6367329174524882e-05, + "loss": 0.0319, + "step": 5108 + }, + { + "epoch": 0.5538811795316565, + "grad_norm": 4.179808139801025, + "learning_rate": 1.6366603800957494e-05, + "loss": 0.0512, + "step": 5109 + }, + { + "epoch": 0.5539895923677364, + "grad_norm": 2.067439556121826, + "learning_rate": 1.6365878427390106e-05, + "loss": 0.0415, + "step": 5110 + }, + { + "epoch": 0.5540980052038161, + "grad_norm": 1.126861572265625, + "learning_rate": 1.636515305382272e-05, + "loss": 0.0324, + "step": 5111 + }, + { + "epoch": 0.5542064180398959, + "grad_norm": 0.6396703124046326, + "learning_rate": 1.6364427680255334e-05, + "loss": 0.0117, + "step": 5112 + }, + { + "epoch": 0.5543148308759757, + "grad_norm": 2.0417380332946777, + "learning_rate": 1.6363702306687946e-05, + "loss": 0.0634, + "step": 5113 + }, + { + "epoch": 0.5544232437120555, + "grad_norm": 1.9772326946258545, + "learning_rate": 1.6362976933120558e-05, + "loss": 0.033, + "step": 5114 + }, + { + "epoch": 0.5545316565481353, + "grad_norm": 3.6371052265167236, + "learning_rate": 1.636225155955317e-05, + "loss": 0.0544, + "step": 5115 + }, + { + "epoch": 0.5546400693842151, + "grad_norm": 3.9697365760803223, + "learning_rate": 1.6361526185985782e-05, + "loss": 0.095, + "step": 5116 + }, + { + "epoch": 0.5547484822202949, + "grad_norm": 2.3036892414093018, + "learning_rate": 1.6360800812418395e-05, + "loss": 0.0469, + "step": 5117 + }, + { + "epoch": 0.5548568950563747, + "grad_norm": 1.8933926820755005, + "learning_rate": 1.636007543885101e-05, + "loss": 0.0151, + "step": 5118 + }, + { + "epoch": 0.5549653078924545, + "grad_norm": 2.0600759983062744, + "learning_rate": 1.6359350065283622e-05, + "loss": 0.0363, + "step": 5119 + }, + { + "epoch": 0.5550737207285342, + "grad_norm": 2.1259381771087646, + "learning_rate": 1.6358624691716238e-05, + "loss": 0.0252, + "step": 5120 + }, + { + "epoch": 0.5551821335646141, + "grad_norm": 2.1480941772460938, + "learning_rate": 1.635789931814885e-05, + "loss": 0.0392, + "step": 5121 + }, + { + "epoch": 0.5552905464006939, + "grad_norm": 4.215796947479248, + "learning_rate": 1.6357173944581462e-05, + "loss": 0.084, + "step": 5122 + }, + { + "epoch": 0.5553989592367736, + "grad_norm": 1.6010388135910034, + "learning_rate": 1.6356448571014074e-05, + "loss": 0.026, + "step": 5123 + }, + { + "epoch": 0.5555073720728534, + "grad_norm": 1.0921252965927124, + "learning_rate": 1.6355723197446686e-05, + "loss": 0.0306, + "step": 5124 + }, + { + "epoch": 0.5556157849089333, + "grad_norm": 4.439383029937744, + "learning_rate": 1.63549978238793e-05, + "loss": 0.028, + "step": 5125 + }, + { + "epoch": 0.555724197745013, + "grad_norm": 1.8050683736801147, + "learning_rate": 1.635427245031191e-05, + "loss": 0.0343, + "step": 5126 + }, + { + "epoch": 0.5558326105810928, + "grad_norm": 1.48624849319458, + "learning_rate": 1.6353547076744526e-05, + "loss": 0.0286, + "step": 5127 + }, + { + "epoch": 0.5559410234171726, + "grad_norm": 1.5549880266189575, + "learning_rate": 1.635282170317714e-05, + "loss": 0.0366, + "step": 5128 + }, + { + "epoch": 0.5560494362532524, + "grad_norm": 2.3564155101776123, + "learning_rate": 1.635209632960975e-05, + "loss": 0.0511, + "step": 5129 + }, + { + "epoch": 0.5561578490893322, + "grad_norm": 0.9540880918502808, + "learning_rate": 1.6351370956042363e-05, + "loss": 0.0181, + "step": 5130 + }, + { + "epoch": 0.556266261925412, + "grad_norm": 1.4149880409240723, + "learning_rate": 1.6350645582474975e-05, + "loss": 0.0274, + "step": 5131 + }, + { + "epoch": 0.5563746747614917, + "grad_norm": 0.7735758423805237, + "learning_rate": 1.6349920208907587e-05, + "loss": 0.0172, + "step": 5132 + }, + { + "epoch": 0.5564830875975716, + "grad_norm": 2.3908753395080566, + "learning_rate": 1.6349194835340202e-05, + "loss": 0.0597, + "step": 5133 + }, + { + "epoch": 0.5565915004336514, + "grad_norm": 4.839807987213135, + "learning_rate": 1.6348469461772815e-05, + "loss": 0.0219, + "step": 5134 + }, + { + "epoch": 0.5566999132697311, + "grad_norm": 1.5110752582550049, + "learning_rate": 1.6347744088205427e-05, + "loss": 0.0221, + "step": 5135 + }, + { + "epoch": 0.5568083261058109, + "grad_norm": 2.570100784301758, + "learning_rate": 1.6347018714638042e-05, + "loss": 0.0194, + "step": 5136 + }, + { + "epoch": 0.5569167389418908, + "grad_norm": 3.9641733169555664, + "learning_rate": 1.6346293341070654e-05, + "loss": 0.0499, + "step": 5137 + }, + { + "epoch": 0.5570251517779705, + "grad_norm": 6.364387035369873, + "learning_rate": 1.6345567967503267e-05, + "loss": 0.0382, + "step": 5138 + }, + { + "epoch": 0.5571335646140503, + "grad_norm": 5.152579307556152, + "learning_rate": 1.634484259393588e-05, + "loss": 0.0211, + "step": 5139 + }, + { + "epoch": 0.55724197745013, + "grad_norm": 2.032949209213257, + "learning_rate": 1.634411722036849e-05, + "loss": 0.0295, + "step": 5140 + }, + { + "epoch": 0.5573503902862099, + "grad_norm": 0.7414957880973816, + "learning_rate": 1.6343391846801103e-05, + "loss": 0.0108, + "step": 5141 + }, + { + "epoch": 0.5574588031222897, + "grad_norm": 3.2642416954040527, + "learning_rate": 1.634266647323372e-05, + "loss": 0.0567, + "step": 5142 + }, + { + "epoch": 0.5575672159583694, + "grad_norm": 6.758565902709961, + "learning_rate": 1.634194109966633e-05, + "loss": 0.0522, + "step": 5143 + }, + { + "epoch": 0.5576756287944492, + "grad_norm": 3.263155937194824, + "learning_rate": 1.6341215726098943e-05, + "loss": 0.0733, + "step": 5144 + }, + { + "epoch": 0.5577840416305291, + "grad_norm": 4.840835094451904, + "learning_rate": 1.6340490352531555e-05, + "loss": 0.0185, + "step": 5145 + }, + { + "epoch": 0.5578924544666088, + "grad_norm": 2.7096002101898193, + "learning_rate": 1.6339764978964167e-05, + "loss": 0.0407, + "step": 5146 + }, + { + "epoch": 0.5580008673026886, + "grad_norm": 4.237142086029053, + "learning_rate": 1.633903960539678e-05, + "loss": 0.035, + "step": 5147 + }, + { + "epoch": 0.5581092801387685, + "grad_norm": 1.901175856590271, + "learning_rate": 1.633831423182939e-05, + "loss": 0.0218, + "step": 5148 + }, + { + "epoch": 0.5582176929748482, + "grad_norm": 4.94233512878418, + "learning_rate": 1.6337588858262007e-05, + "loss": 0.044, + "step": 5149 + }, + { + "epoch": 0.558326105810928, + "grad_norm": 1.802440881729126, + "learning_rate": 1.633686348469462e-05, + "loss": 0.0366, + "step": 5150 + }, + { + "epoch": 0.5584345186470078, + "grad_norm": 2.206983804702759, + "learning_rate": 1.633613811112723e-05, + "loss": 0.0277, + "step": 5151 + }, + { + "epoch": 0.5585429314830876, + "grad_norm": 1.0741589069366455, + "learning_rate": 1.6335412737559843e-05, + "loss": 0.0099, + "step": 5152 + }, + { + "epoch": 0.5586513443191674, + "grad_norm": 5.373044013977051, + "learning_rate": 1.6334687363992455e-05, + "loss": 0.0635, + "step": 5153 + }, + { + "epoch": 0.5587597571552472, + "grad_norm": 4.206640243530273, + "learning_rate": 1.633396199042507e-05, + "loss": 0.0896, + "step": 5154 + }, + { + "epoch": 0.5588681699913269, + "grad_norm": 5.567643165588379, + "learning_rate": 1.6333236616857683e-05, + "loss": 0.0673, + "step": 5155 + }, + { + "epoch": 0.5589765828274068, + "grad_norm": 4.056031227111816, + "learning_rate": 1.6332511243290295e-05, + "loss": 0.0349, + "step": 5156 + }, + { + "epoch": 0.5590849956634866, + "grad_norm": 2.1687140464782715, + "learning_rate": 1.633178586972291e-05, + "loss": 0.028, + "step": 5157 + }, + { + "epoch": 0.5591934084995663, + "grad_norm": 3.019845724105835, + "learning_rate": 1.6331060496155523e-05, + "loss": 0.0746, + "step": 5158 + }, + { + "epoch": 0.5593018213356461, + "grad_norm": 11.753154754638672, + "learning_rate": 1.6330335122588135e-05, + "loss": 0.1122, + "step": 5159 + }, + { + "epoch": 0.559410234171726, + "grad_norm": 2.171184539794922, + "learning_rate": 1.6329609749020747e-05, + "loss": 0.014, + "step": 5160 + }, + { + "epoch": 0.5595186470078057, + "grad_norm": 3.2709248065948486, + "learning_rate": 1.632888437545336e-05, + "loss": 0.0213, + "step": 5161 + }, + { + "epoch": 0.5596270598438855, + "grad_norm": 2.9245755672454834, + "learning_rate": 1.632815900188597e-05, + "loss": 0.0932, + "step": 5162 + }, + { + "epoch": 0.5597354726799653, + "grad_norm": 4.73089599609375, + "learning_rate": 1.6327433628318584e-05, + "loss": 0.0661, + "step": 5163 + }, + { + "epoch": 0.5598438855160451, + "grad_norm": 3.706360340118408, + "learning_rate": 1.63267082547512e-05, + "loss": 0.1017, + "step": 5164 + }, + { + "epoch": 0.5599522983521249, + "grad_norm": 2.1361641883850098, + "learning_rate": 1.632598288118381e-05, + "loss": 0.0297, + "step": 5165 + }, + { + "epoch": 0.5600607111882047, + "grad_norm": 2.5614614486694336, + "learning_rate": 1.6325257507616423e-05, + "loss": 0.0238, + "step": 5166 + }, + { + "epoch": 0.5601691240242844, + "grad_norm": 2.999551296234131, + "learning_rate": 1.6324532134049036e-05, + "loss": 0.0197, + "step": 5167 + }, + { + "epoch": 0.5602775368603643, + "grad_norm": 4.107407093048096, + "learning_rate": 1.6323806760481648e-05, + "loss": 0.0593, + "step": 5168 + }, + { + "epoch": 0.5603859496964441, + "grad_norm": 5.543707370758057, + "learning_rate": 1.632308138691426e-05, + "loss": 0.0641, + "step": 5169 + }, + { + "epoch": 0.5604943625325238, + "grad_norm": 1.2094587087631226, + "learning_rate": 1.6322356013346875e-05, + "loss": 0.0063, + "step": 5170 + }, + { + "epoch": 0.5606027753686036, + "grad_norm": 1.5998605489730835, + "learning_rate": 1.6321630639779488e-05, + "loss": 0.0343, + "step": 5171 + }, + { + "epoch": 0.5607111882046835, + "grad_norm": 3.9151864051818848, + "learning_rate": 1.6320905266212103e-05, + "loss": 0.0632, + "step": 5172 + }, + { + "epoch": 0.5608196010407632, + "grad_norm": 0.6869471073150635, + "learning_rate": 1.6320179892644715e-05, + "loss": 0.0045, + "step": 5173 + }, + { + "epoch": 0.560928013876843, + "grad_norm": 1.3526936769485474, + "learning_rate": 1.6319454519077327e-05, + "loss": 0.0235, + "step": 5174 + }, + { + "epoch": 0.5610364267129229, + "grad_norm": 2.7098913192749023, + "learning_rate": 1.631872914550994e-05, + "loss": 0.0439, + "step": 5175 + }, + { + "epoch": 0.5611448395490026, + "grad_norm": 1.8382817506790161, + "learning_rate": 1.631800377194255e-05, + "loss": 0.0322, + "step": 5176 + }, + { + "epoch": 0.5612532523850824, + "grad_norm": 2.497119426727295, + "learning_rate": 1.6317278398375164e-05, + "loss": 0.0733, + "step": 5177 + }, + { + "epoch": 0.5613616652211622, + "grad_norm": 3.0273826122283936, + "learning_rate": 1.6316553024807776e-05, + "loss": 0.0428, + "step": 5178 + }, + { + "epoch": 0.561470078057242, + "grad_norm": 0.5653853416442871, + "learning_rate": 1.631582765124039e-05, + "loss": 0.0045, + "step": 5179 + }, + { + "epoch": 0.5615784908933218, + "grad_norm": 4.786931037902832, + "learning_rate": 1.6315102277673004e-05, + "loss": 0.0476, + "step": 5180 + }, + { + "epoch": 0.5616869037294016, + "grad_norm": 4.063449382781982, + "learning_rate": 1.6314376904105616e-05, + "loss": 0.0721, + "step": 5181 + }, + { + "epoch": 0.5617953165654813, + "grad_norm": 1.374902606010437, + "learning_rate": 1.6313651530538228e-05, + "loss": 0.025, + "step": 5182 + }, + { + "epoch": 0.5619037294015612, + "grad_norm": 1.4928038120269775, + "learning_rate": 1.631292615697084e-05, + "loss": 0.0123, + "step": 5183 + }, + { + "epoch": 0.562012142237641, + "grad_norm": 3.4076461791992188, + "learning_rate": 1.6312200783403452e-05, + "loss": 0.0265, + "step": 5184 + }, + { + "epoch": 0.5621205550737207, + "grad_norm": 2.495006799697876, + "learning_rate": 1.6311475409836068e-05, + "loss": 0.0175, + "step": 5185 + }, + { + "epoch": 0.5622289679098005, + "grad_norm": 3.2925543785095215, + "learning_rate": 1.631075003626868e-05, + "loss": 0.0475, + "step": 5186 + }, + { + "epoch": 0.5623373807458804, + "grad_norm": 4.196090221405029, + "learning_rate": 1.6310024662701292e-05, + "loss": 0.0394, + "step": 5187 + }, + { + "epoch": 0.5624457935819601, + "grad_norm": 2.0340631008148193, + "learning_rate": 1.6309299289133908e-05, + "loss": 0.0429, + "step": 5188 + }, + { + "epoch": 0.5625542064180399, + "grad_norm": 3.486473321914673, + "learning_rate": 1.630857391556652e-05, + "loss": 0.0467, + "step": 5189 + }, + { + "epoch": 0.5626626192541196, + "grad_norm": 3.2000608444213867, + "learning_rate": 1.6307848541999132e-05, + "loss": 0.0247, + "step": 5190 + }, + { + "epoch": 0.5627710320901995, + "grad_norm": 3.377206802368164, + "learning_rate": 1.6307123168431744e-05, + "loss": 0.0312, + "step": 5191 + }, + { + "epoch": 0.5628794449262793, + "grad_norm": 4.260329723358154, + "learning_rate": 1.6306397794864356e-05, + "loss": 0.0284, + "step": 5192 + }, + { + "epoch": 0.562987857762359, + "grad_norm": 5.612096309661865, + "learning_rate": 1.6305672421296968e-05, + "loss": 0.0758, + "step": 5193 + }, + { + "epoch": 0.5630962705984388, + "grad_norm": 5.179226398468018, + "learning_rate": 1.6304947047729584e-05, + "loss": 0.0949, + "step": 5194 + }, + { + "epoch": 0.5632046834345187, + "grad_norm": 3.790357828140259, + "learning_rate": 1.6304221674162196e-05, + "loss": 0.0353, + "step": 5195 + }, + { + "epoch": 0.5633130962705984, + "grad_norm": 1.7909817695617676, + "learning_rate": 1.6303496300594808e-05, + "loss": 0.0245, + "step": 5196 + }, + { + "epoch": 0.5634215091066782, + "grad_norm": 1.39255690574646, + "learning_rate": 1.630277092702742e-05, + "loss": 0.0272, + "step": 5197 + }, + { + "epoch": 0.563529921942758, + "grad_norm": 2.7219812870025635, + "learning_rate": 1.6302045553460032e-05, + "loss": 0.054, + "step": 5198 + }, + { + "epoch": 0.5636383347788378, + "grad_norm": 1.8366403579711914, + "learning_rate": 1.6301320179892645e-05, + "loss": 0.0421, + "step": 5199 + }, + { + "epoch": 0.5637467476149176, + "grad_norm": 2.781665086746216, + "learning_rate": 1.6300594806325257e-05, + "loss": 0.0539, + "step": 5200 + }, + { + "epoch": 0.5638551604509974, + "grad_norm": 2.8814520835876465, + "learning_rate": 1.6299869432757872e-05, + "loss": 0.0254, + "step": 5201 + }, + { + "epoch": 0.5639635732870771, + "grad_norm": 3.2619359493255615, + "learning_rate": 1.6299144059190484e-05, + "loss": 0.0823, + "step": 5202 + }, + { + "epoch": 0.564071986123157, + "grad_norm": 2.2410380840301514, + "learning_rate": 1.6298418685623097e-05, + "loss": 0.0195, + "step": 5203 + }, + { + "epoch": 0.5641803989592368, + "grad_norm": 1.6254780292510986, + "learning_rate": 1.629769331205571e-05, + "loss": 0.018, + "step": 5204 + }, + { + "epoch": 0.5642888117953165, + "grad_norm": 3.3830881118774414, + "learning_rate": 1.6296967938488324e-05, + "loss": 0.0704, + "step": 5205 + }, + { + "epoch": 0.5643972246313964, + "grad_norm": 2.9379169940948486, + "learning_rate": 1.6296242564920936e-05, + "loss": 0.0459, + "step": 5206 + }, + { + "epoch": 0.5645056374674762, + "grad_norm": 8.189810752868652, + "learning_rate": 1.629551719135355e-05, + "loss": 0.053, + "step": 5207 + }, + { + "epoch": 0.5646140503035559, + "grad_norm": 5.665319442749023, + "learning_rate": 1.629479181778616e-05, + "loss": 0.0446, + "step": 5208 + }, + { + "epoch": 0.5647224631396357, + "grad_norm": 0.9014101624488831, + "learning_rate": 1.6294066444218776e-05, + "loss": 0.0123, + "step": 5209 + }, + { + "epoch": 0.5648308759757156, + "grad_norm": 4.98649787902832, + "learning_rate": 1.6293341070651388e-05, + "loss": 0.0841, + "step": 5210 + }, + { + "epoch": 0.5649392888117953, + "grad_norm": 1.8716496229171753, + "learning_rate": 1.6292615697084e-05, + "loss": 0.0248, + "step": 5211 + }, + { + "epoch": 0.5650477016478751, + "grad_norm": 2.0477099418640137, + "learning_rate": 1.6291890323516613e-05, + "loss": 0.0363, + "step": 5212 + }, + { + "epoch": 0.5651561144839549, + "grad_norm": 1.0944163799285889, + "learning_rate": 1.6291164949949225e-05, + "loss": 0.0162, + "step": 5213 + }, + { + "epoch": 0.5652645273200347, + "grad_norm": 1.12092125415802, + "learning_rate": 1.6290439576381837e-05, + "loss": 0.0143, + "step": 5214 + }, + { + "epoch": 0.5653729401561145, + "grad_norm": 2.5396766662597656, + "learning_rate": 1.628971420281445e-05, + "loss": 0.058, + "step": 5215 + }, + { + "epoch": 0.5654813529921943, + "grad_norm": 0.8147549033164978, + "learning_rate": 1.6288988829247065e-05, + "loss": 0.0047, + "step": 5216 + }, + { + "epoch": 0.565589765828274, + "grad_norm": 1.940107822418213, + "learning_rate": 1.6288263455679677e-05, + "loss": 0.0224, + "step": 5217 + }, + { + "epoch": 0.5656981786643539, + "grad_norm": 1.748453974723816, + "learning_rate": 1.628753808211229e-05, + "loss": 0.054, + "step": 5218 + }, + { + "epoch": 0.5658065915004337, + "grad_norm": 1.0618023872375488, + "learning_rate": 1.62868127085449e-05, + "loss": 0.0186, + "step": 5219 + }, + { + "epoch": 0.5659150043365134, + "grad_norm": 2.6352462768554688, + "learning_rate": 1.6286087334977513e-05, + "loss": 0.0162, + "step": 5220 + }, + { + "epoch": 0.5660234171725932, + "grad_norm": 2.846100330352783, + "learning_rate": 1.6285361961410125e-05, + "loss": 0.0465, + "step": 5221 + }, + { + "epoch": 0.5661318300086731, + "grad_norm": 2.3492531776428223, + "learning_rate": 1.628463658784274e-05, + "loss": 0.042, + "step": 5222 + }, + { + "epoch": 0.5662402428447528, + "grad_norm": 3.823028326034546, + "learning_rate": 1.6283911214275353e-05, + "loss": 0.0301, + "step": 5223 + }, + { + "epoch": 0.5663486556808326, + "grad_norm": 3.8096680641174316, + "learning_rate": 1.628318584070797e-05, + "loss": 0.0608, + "step": 5224 + }, + { + "epoch": 0.5664570685169124, + "grad_norm": 1.7454932928085327, + "learning_rate": 1.628246046714058e-05, + "loss": 0.0203, + "step": 5225 + }, + { + "epoch": 0.5665654813529922, + "grad_norm": 2.2374086380004883, + "learning_rate": 1.6281735093573193e-05, + "loss": 0.0222, + "step": 5226 + }, + { + "epoch": 0.566673894189072, + "grad_norm": 2.8628079891204834, + "learning_rate": 1.6281009720005805e-05, + "loss": 0.05, + "step": 5227 + }, + { + "epoch": 0.5667823070251518, + "grad_norm": 2.0448055267333984, + "learning_rate": 1.6280284346438417e-05, + "loss": 0.0116, + "step": 5228 + }, + { + "epoch": 0.5668907198612315, + "grad_norm": 3.252000093460083, + "learning_rate": 1.627955897287103e-05, + "loss": 0.0607, + "step": 5229 + }, + { + "epoch": 0.5669991326973114, + "grad_norm": 3.247689723968506, + "learning_rate": 1.627883359930364e-05, + "loss": 0.0415, + "step": 5230 + }, + { + "epoch": 0.5671075455333912, + "grad_norm": 4.647761821746826, + "learning_rate": 1.6278108225736257e-05, + "loss": 0.0558, + "step": 5231 + }, + { + "epoch": 0.5672159583694709, + "grad_norm": 5.589905738830566, + "learning_rate": 1.627738285216887e-05, + "loss": 0.0831, + "step": 5232 + }, + { + "epoch": 0.5673243712055508, + "grad_norm": 3.773484468460083, + "learning_rate": 1.627665747860148e-05, + "loss": 0.0281, + "step": 5233 + }, + { + "epoch": 0.5674327840416306, + "grad_norm": 1.9897761344909668, + "learning_rate": 1.6275932105034093e-05, + "loss": 0.0346, + "step": 5234 + }, + { + "epoch": 0.5675411968777103, + "grad_norm": 6.798101425170898, + "learning_rate": 1.6275206731466705e-05, + "loss": 0.0625, + "step": 5235 + }, + { + "epoch": 0.5676496097137901, + "grad_norm": 3.1878373622894287, + "learning_rate": 1.6274481357899318e-05, + "loss": 0.046, + "step": 5236 + }, + { + "epoch": 0.56775802254987, + "grad_norm": 4.483451843261719, + "learning_rate": 1.6273755984331933e-05, + "loss": 0.0244, + "step": 5237 + }, + { + "epoch": 0.5678664353859497, + "grad_norm": 4.078670978546143, + "learning_rate": 1.6273030610764545e-05, + "loss": 0.0311, + "step": 5238 + }, + { + "epoch": 0.5679748482220295, + "grad_norm": 0.9147211313247681, + "learning_rate": 1.6272305237197157e-05, + "loss": 0.0221, + "step": 5239 + }, + { + "epoch": 0.5680832610581092, + "grad_norm": 4.162057876586914, + "learning_rate": 1.6271579863629773e-05, + "loss": 0.0512, + "step": 5240 + }, + { + "epoch": 0.5681916738941891, + "grad_norm": 1.3945006132125854, + "learning_rate": 1.6270854490062385e-05, + "loss": 0.0114, + "step": 5241 + }, + { + "epoch": 0.5683000867302689, + "grad_norm": 1.7980490922927856, + "learning_rate": 1.6270129116494997e-05, + "loss": 0.025, + "step": 5242 + }, + { + "epoch": 0.5684084995663486, + "grad_norm": 4.583281993865967, + "learning_rate": 1.626940374292761e-05, + "loss": 0.0907, + "step": 5243 + }, + { + "epoch": 0.5685169124024284, + "grad_norm": 1.6427212953567505, + "learning_rate": 1.626867836936022e-05, + "loss": 0.0442, + "step": 5244 + }, + { + "epoch": 0.5686253252385083, + "grad_norm": 3.1747376918792725, + "learning_rate": 1.6267952995792834e-05, + "loss": 0.018, + "step": 5245 + }, + { + "epoch": 0.568733738074588, + "grad_norm": 3.9883570671081543, + "learning_rate": 1.626722762222545e-05, + "loss": 0.0409, + "step": 5246 + }, + { + "epoch": 0.5688421509106678, + "grad_norm": 1.3352001905441284, + "learning_rate": 1.626650224865806e-05, + "loss": 0.0226, + "step": 5247 + }, + { + "epoch": 0.5689505637467476, + "grad_norm": 2.2240371704101562, + "learning_rate": 1.6265776875090673e-05, + "loss": 0.0507, + "step": 5248 + }, + { + "epoch": 0.5690589765828274, + "grad_norm": 0.9287235736846924, + "learning_rate": 1.6265051501523286e-05, + "loss": 0.0176, + "step": 5249 + }, + { + "epoch": 0.5691673894189072, + "grad_norm": 5.328152656555176, + "learning_rate": 1.6264326127955898e-05, + "loss": 0.0322, + "step": 5250 + }, + { + "epoch": 0.569275802254987, + "grad_norm": 2.1947877407073975, + "learning_rate": 1.626360075438851e-05, + "loss": 0.022, + "step": 5251 + }, + { + "epoch": 0.5693842150910667, + "grad_norm": 3.5893802642822266, + "learning_rate": 1.6262875380821125e-05, + "loss": 0.1011, + "step": 5252 + }, + { + "epoch": 0.5694926279271466, + "grad_norm": 2.4303719997406006, + "learning_rate": 1.6262150007253738e-05, + "loss": 0.0291, + "step": 5253 + }, + { + "epoch": 0.5696010407632264, + "grad_norm": 2.904127359390259, + "learning_rate": 1.626142463368635e-05, + "loss": 0.0353, + "step": 5254 + }, + { + "epoch": 0.5697094535993061, + "grad_norm": 0.48067018389701843, + "learning_rate": 1.6260699260118962e-05, + "loss": 0.0152, + "step": 5255 + }, + { + "epoch": 0.5698178664353859, + "grad_norm": 5.9353532791137695, + "learning_rate": 1.6259973886551574e-05, + "loss": 0.0269, + "step": 5256 + }, + { + "epoch": 0.5699262792714658, + "grad_norm": 2.704540491104126, + "learning_rate": 1.625924851298419e-05, + "loss": 0.0352, + "step": 5257 + }, + { + "epoch": 0.5700346921075455, + "grad_norm": 5.468922138214111, + "learning_rate": 1.62585231394168e-05, + "loss": 0.0611, + "step": 5258 + }, + { + "epoch": 0.5701431049436253, + "grad_norm": 3.2617902755737305, + "learning_rate": 1.6257797765849414e-05, + "loss": 0.0518, + "step": 5259 + }, + { + "epoch": 0.5702515177797051, + "grad_norm": 2.033036708831787, + "learning_rate": 1.6257072392282026e-05, + "loss": 0.0116, + "step": 5260 + }, + { + "epoch": 0.5703599306157849, + "grad_norm": 5.193704605102539, + "learning_rate": 1.625634701871464e-05, + "loss": 0.0537, + "step": 5261 + }, + { + "epoch": 0.5704683434518647, + "grad_norm": 4.4904890060424805, + "learning_rate": 1.6255621645147254e-05, + "loss": 0.0405, + "step": 5262 + }, + { + "epoch": 0.5705767562879445, + "grad_norm": 2.334906816482544, + "learning_rate": 1.6254896271579866e-05, + "loss": 0.0248, + "step": 5263 + }, + { + "epoch": 0.5706851691240243, + "grad_norm": 2.1067264080047607, + "learning_rate": 1.6254170898012478e-05, + "loss": 0.0412, + "step": 5264 + }, + { + "epoch": 0.5707935819601041, + "grad_norm": 3.0895817279815674, + "learning_rate": 1.625344552444509e-05, + "loss": 0.0503, + "step": 5265 + }, + { + "epoch": 0.5709019947961839, + "grad_norm": 1.227602243423462, + "learning_rate": 1.6252720150877702e-05, + "loss": 0.0102, + "step": 5266 + }, + { + "epoch": 0.5710104076322636, + "grad_norm": 3.820214033126831, + "learning_rate": 1.6251994777310314e-05, + "loss": 0.0448, + "step": 5267 + }, + { + "epoch": 0.5711188204683435, + "grad_norm": 1.5461868047714233, + "learning_rate": 1.625126940374293e-05, + "loss": 0.0335, + "step": 5268 + }, + { + "epoch": 0.5712272333044233, + "grad_norm": 2.355576992034912, + "learning_rate": 1.6250544030175542e-05, + "loss": 0.0077, + "step": 5269 + }, + { + "epoch": 0.571335646140503, + "grad_norm": 1.4700407981872559, + "learning_rate": 1.6249818656608154e-05, + "loss": 0.0229, + "step": 5270 + }, + { + "epoch": 0.5714440589765828, + "grad_norm": 2.1841485500335693, + "learning_rate": 1.6249093283040766e-05, + "loss": 0.0176, + "step": 5271 + }, + { + "epoch": 0.5715524718126627, + "grad_norm": 2.8015308380126953, + "learning_rate": 1.624836790947338e-05, + "loss": 0.0344, + "step": 5272 + }, + { + "epoch": 0.5716608846487424, + "grad_norm": 2.7629618644714355, + "learning_rate": 1.624764253590599e-05, + "loss": 0.0505, + "step": 5273 + }, + { + "epoch": 0.5717692974848222, + "grad_norm": 1.6309070587158203, + "learning_rate": 1.6246917162338606e-05, + "loss": 0.0105, + "step": 5274 + }, + { + "epoch": 0.571877710320902, + "grad_norm": 1.9800699949264526, + "learning_rate": 1.6246191788771218e-05, + "loss": 0.0258, + "step": 5275 + }, + { + "epoch": 0.5719861231569818, + "grad_norm": 2.7174620628356934, + "learning_rate": 1.6245466415203834e-05, + "loss": 0.0365, + "step": 5276 + }, + { + "epoch": 0.5720945359930616, + "grad_norm": 3.4483487606048584, + "learning_rate": 1.6244741041636446e-05, + "loss": 0.0206, + "step": 5277 + }, + { + "epoch": 0.5722029488291414, + "grad_norm": 4.399353981018066, + "learning_rate": 1.6244015668069058e-05, + "loss": 0.0526, + "step": 5278 + }, + { + "epoch": 0.5723113616652211, + "grad_norm": 0.715123176574707, + "learning_rate": 1.624329029450167e-05, + "loss": 0.0086, + "step": 5279 + }, + { + "epoch": 0.572419774501301, + "grad_norm": 4.103656768798828, + "learning_rate": 1.6242564920934282e-05, + "loss": 0.0787, + "step": 5280 + }, + { + "epoch": 0.5725281873373808, + "grad_norm": 3.109549045562744, + "learning_rate": 1.6241839547366895e-05, + "loss": 0.0389, + "step": 5281 + }, + { + "epoch": 0.5726366001734605, + "grad_norm": 1.188319444656372, + "learning_rate": 1.6241114173799507e-05, + "loss": 0.0088, + "step": 5282 + }, + { + "epoch": 0.5727450130095403, + "grad_norm": 2.2685437202453613, + "learning_rate": 1.6240388800232122e-05, + "loss": 0.02, + "step": 5283 + }, + { + "epoch": 0.5728534258456202, + "grad_norm": 3.9989447593688965, + "learning_rate": 1.6239663426664734e-05, + "loss": 0.0724, + "step": 5284 + }, + { + "epoch": 0.5729618386816999, + "grad_norm": 4.955438137054443, + "learning_rate": 1.6238938053097346e-05, + "loss": 0.029, + "step": 5285 + }, + { + "epoch": 0.5730702515177797, + "grad_norm": 3.218674898147583, + "learning_rate": 1.623821267952996e-05, + "loss": 0.0334, + "step": 5286 + }, + { + "epoch": 0.5731786643538594, + "grad_norm": 3.229050874710083, + "learning_rate": 1.623748730596257e-05, + "loss": 0.0601, + "step": 5287 + }, + { + "epoch": 0.5732870771899393, + "grad_norm": 3.6192026138305664, + "learning_rate": 1.6236761932395183e-05, + "loss": 0.0432, + "step": 5288 + }, + { + "epoch": 0.5733954900260191, + "grad_norm": 1.9559262990951538, + "learning_rate": 1.62360365588278e-05, + "loss": 0.0273, + "step": 5289 + }, + { + "epoch": 0.5735039028620988, + "grad_norm": 2.726465940475464, + "learning_rate": 1.623531118526041e-05, + "loss": 0.0456, + "step": 5290 + }, + { + "epoch": 0.5736123156981786, + "grad_norm": 6.670962333679199, + "learning_rate": 1.6234585811693023e-05, + "loss": 0.0913, + "step": 5291 + }, + { + "epoch": 0.5737207285342585, + "grad_norm": 5.1301374435424805, + "learning_rate": 1.6233860438125638e-05, + "loss": 0.0588, + "step": 5292 + }, + { + "epoch": 0.5738291413703382, + "grad_norm": 3.2523117065429688, + "learning_rate": 1.623313506455825e-05, + "loss": 0.038, + "step": 5293 + }, + { + "epoch": 0.573937554206418, + "grad_norm": 3.2597174644470215, + "learning_rate": 1.6232409690990863e-05, + "loss": 0.0717, + "step": 5294 + }, + { + "epoch": 0.5740459670424979, + "grad_norm": 2.234738349914551, + "learning_rate": 1.6231684317423475e-05, + "loss": 0.0547, + "step": 5295 + }, + { + "epoch": 0.5741543798785776, + "grad_norm": 14.439750671386719, + "learning_rate": 1.6230958943856087e-05, + "loss": 0.0636, + "step": 5296 + }, + { + "epoch": 0.5742627927146574, + "grad_norm": 1.014384388923645, + "learning_rate": 1.62302335702887e-05, + "loss": 0.017, + "step": 5297 + }, + { + "epoch": 0.5743712055507372, + "grad_norm": 3.3372912406921387, + "learning_rate": 1.6229508196721314e-05, + "loss": 0.028, + "step": 5298 + }, + { + "epoch": 0.574479618386817, + "grad_norm": 2.4049293994903564, + "learning_rate": 1.6228782823153927e-05, + "loss": 0.0181, + "step": 5299 + }, + { + "epoch": 0.5745880312228968, + "grad_norm": 1.522692322731018, + "learning_rate": 1.622805744958654e-05, + "loss": 0.0392, + "step": 5300 + }, + { + "epoch": 0.5746964440589766, + "grad_norm": 4.380326747894287, + "learning_rate": 1.622733207601915e-05, + "loss": 0.0275, + "step": 5301 + }, + { + "epoch": 0.5748048568950563, + "grad_norm": 4.22769021987915, + "learning_rate": 1.6226606702451763e-05, + "loss": 0.0697, + "step": 5302 + }, + { + "epoch": 0.5749132697311362, + "grad_norm": 2.278947114944458, + "learning_rate": 1.6225881328884375e-05, + "loss": 0.0402, + "step": 5303 + }, + { + "epoch": 0.575021682567216, + "grad_norm": 1.7913002967834473, + "learning_rate": 1.622515595531699e-05, + "loss": 0.036, + "step": 5304 + }, + { + "epoch": 0.5751300954032957, + "grad_norm": 1.6766023635864258, + "learning_rate": 1.6224430581749603e-05, + "loss": 0.0195, + "step": 5305 + }, + { + "epoch": 0.5752385082393755, + "grad_norm": 3.334897994995117, + "learning_rate": 1.6223705208182215e-05, + "loss": 0.028, + "step": 5306 + }, + { + "epoch": 0.5753469210754554, + "grad_norm": 1.6486204862594604, + "learning_rate": 1.6222979834614827e-05, + "loss": 0.0254, + "step": 5307 + }, + { + "epoch": 0.5754553339115351, + "grad_norm": 6.886995792388916, + "learning_rate": 1.622225446104744e-05, + "loss": 0.0763, + "step": 5308 + }, + { + "epoch": 0.5755637467476149, + "grad_norm": 2.50095796585083, + "learning_rate": 1.6221529087480055e-05, + "loss": 0.0472, + "step": 5309 + }, + { + "epoch": 0.5756721595836947, + "grad_norm": 2.944582939147949, + "learning_rate": 1.6220803713912667e-05, + "loss": 0.0207, + "step": 5310 + }, + { + "epoch": 0.5757805724197745, + "grad_norm": 2.4111709594726562, + "learning_rate": 1.622007834034528e-05, + "loss": 0.0448, + "step": 5311 + }, + { + "epoch": 0.5758889852558543, + "grad_norm": 2.602546453475952, + "learning_rate": 1.621935296677789e-05, + "loss": 0.0516, + "step": 5312 + }, + { + "epoch": 0.5759973980919341, + "grad_norm": 3.177354335784912, + "learning_rate": 1.6218627593210507e-05, + "loss": 0.0617, + "step": 5313 + }, + { + "epoch": 0.5761058109280138, + "grad_norm": 3.6424150466918945, + "learning_rate": 1.621790221964312e-05, + "loss": 0.1021, + "step": 5314 + }, + { + "epoch": 0.5762142237640937, + "grad_norm": 2.7174103260040283, + "learning_rate": 1.621717684607573e-05, + "loss": 0.0376, + "step": 5315 + }, + { + "epoch": 0.5763226366001735, + "grad_norm": 3.8231732845306396, + "learning_rate": 1.6216451472508343e-05, + "loss": 0.0433, + "step": 5316 + }, + { + "epoch": 0.5764310494362532, + "grad_norm": 0.5600517392158508, + "learning_rate": 1.6215726098940955e-05, + "loss": 0.0072, + "step": 5317 + }, + { + "epoch": 0.576539462272333, + "grad_norm": 2.9329888820648193, + "learning_rate": 1.6215000725373568e-05, + "loss": 0.0514, + "step": 5318 + }, + { + "epoch": 0.5766478751084129, + "grad_norm": 1.846600890159607, + "learning_rate": 1.621427535180618e-05, + "loss": 0.0244, + "step": 5319 + }, + { + "epoch": 0.5767562879444926, + "grad_norm": 0.5850388407707214, + "learning_rate": 1.6213549978238795e-05, + "loss": 0.0086, + "step": 5320 + }, + { + "epoch": 0.5768647007805724, + "grad_norm": 2.853548049926758, + "learning_rate": 1.6212824604671407e-05, + "loss": 0.0293, + "step": 5321 + }, + { + "epoch": 0.5769731136166523, + "grad_norm": 2.518280506134033, + "learning_rate": 1.621209923110402e-05, + "loss": 0.0307, + "step": 5322 + }, + { + "epoch": 0.577081526452732, + "grad_norm": 3.002363443374634, + "learning_rate": 1.621137385753663e-05, + "loss": 0.055, + "step": 5323 + }, + { + "epoch": 0.5771899392888118, + "grad_norm": 2.2105679512023926, + "learning_rate": 1.6210648483969244e-05, + "loss": 0.0265, + "step": 5324 + }, + { + "epoch": 0.5772983521248916, + "grad_norm": 2.442345142364502, + "learning_rate": 1.6209923110401856e-05, + "loss": 0.0656, + "step": 5325 + }, + { + "epoch": 0.5774067649609714, + "grad_norm": 6.248246669769287, + "learning_rate": 1.620919773683447e-05, + "loss": 0.0856, + "step": 5326 + }, + { + "epoch": 0.5775151777970512, + "grad_norm": 1.2285504341125488, + "learning_rate": 1.6208472363267084e-05, + "loss": 0.033, + "step": 5327 + }, + { + "epoch": 0.577623590633131, + "grad_norm": 2.0769546031951904, + "learning_rate": 1.62077469896997e-05, + "loss": 0.0258, + "step": 5328 + }, + { + "epoch": 0.5777320034692107, + "grad_norm": 1.2320173978805542, + "learning_rate": 1.620702161613231e-05, + "loss": 0.0197, + "step": 5329 + }, + { + "epoch": 0.5778404163052906, + "grad_norm": 4.034855365753174, + "learning_rate": 1.6206296242564923e-05, + "loss": 0.0369, + "step": 5330 + }, + { + "epoch": 0.5779488291413704, + "grad_norm": 2.258179187774658, + "learning_rate": 1.6205570868997536e-05, + "loss": 0.013, + "step": 5331 + }, + { + "epoch": 0.5780572419774501, + "grad_norm": 3.637549877166748, + "learning_rate": 1.6204845495430148e-05, + "loss": 0.0611, + "step": 5332 + }, + { + "epoch": 0.5781656548135299, + "grad_norm": 2.726059675216675, + "learning_rate": 1.620412012186276e-05, + "loss": 0.041, + "step": 5333 + }, + { + "epoch": 0.5782740676496098, + "grad_norm": 2.3789901733398438, + "learning_rate": 1.6203394748295372e-05, + "loss": 0.0157, + "step": 5334 + }, + { + "epoch": 0.5783824804856895, + "grad_norm": 1.178820013999939, + "learning_rate": 1.6202669374727988e-05, + "loss": 0.0112, + "step": 5335 + }, + { + "epoch": 0.5784908933217693, + "grad_norm": 1.065276026725769, + "learning_rate": 1.62019440011606e-05, + "loss": 0.0242, + "step": 5336 + }, + { + "epoch": 0.578599306157849, + "grad_norm": 3.695542097091675, + "learning_rate": 1.6201218627593212e-05, + "loss": 0.0161, + "step": 5337 + }, + { + "epoch": 0.5787077189939289, + "grad_norm": 2.123595952987671, + "learning_rate": 1.6200493254025824e-05, + "loss": 0.0142, + "step": 5338 + }, + { + "epoch": 0.5788161318300087, + "grad_norm": 0.49144870042800903, + "learning_rate": 1.6199767880458436e-05, + "loss": 0.0067, + "step": 5339 + }, + { + "epoch": 0.5789245446660884, + "grad_norm": 4.718293190002441, + "learning_rate": 1.6199042506891048e-05, + "loss": 0.0852, + "step": 5340 + }, + { + "epoch": 0.5790329575021682, + "grad_norm": 5.653567790985107, + "learning_rate": 1.6198317133323664e-05, + "loss": 0.075, + "step": 5341 + }, + { + "epoch": 0.5791413703382481, + "grad_norm": 0.34213587641716003, + "learning_rate": 1.6197591759756276e-05, + "loss": 0.0031, + "step": 5342 + }, + { + "epoch": 0.5792497831743278, + "grad_norm": 4.179341793060303, + "learning_rate": 1.6196866386188888e-05, + "loss": 0.0474, + "step": 5343 + }, + { + "epoch": 0.5793581960104076, + "grad_norm": 1.1002577543258667, + "learning_rate": 1.6196141012621504e-05, + "loss": 0.0192, + "step": 5344 + }, + { + "epoch": 0.5794666088464874, + "grad_norm": 2.6020164489746094, + "learning_rate": 1.6195415639054116e-05, + "loss": 0.0392, + "step": 5345 + }, + { + "epoch": 0.5795750216825672, + "grad_norm": 3.174375295639038, + "learning_rate": 1.6194690265486728e-05, + "loss": 0.0598, + "step": 5346 + }, + { + "epoch": 0.579683434518647, + "grad_norm": 3.8556020259857178, + "learning_rate": 1.619396489191934e-05, + "loss": 0.0509, + "step": 5347 + }, + { + "epoch": 0.5797918473547268, + "grad_norm": 6.11973762512207, + "learning_rate": 1.6193239518351952e-05, + "loss": 0.0756, + "step": 5348 + }, + { + "epoch": 0.5799002601908065, + "grad_norm": 6.044345855712891, + "learning_rate": 1.6192514144784564e-05, + "loss": 0.0618, + "step": 5349 + }, + { + "epoch": 0.5800086730268864, + "grad_norm": 5.896756172180176, + "learning_rate": 1.619178877121718e-05, + "loss": 0.0366, + "step": 5350 + }, + { + "epoch": 0.5801170858629662, + "grad_norm": 4.93554162979126, + "learning_rate": 1.6191063397649792e-05, + "loss": 0.0352, + "step": 5351 + }, + { + "epoch": 0.5802254986990459, + "grad_norm": 0.4907979369163513, + "learning_rate": 1.6190338024082404e-05, + "loss": 0.0088, + "step": 5352 + }, + { + "epoch": 0.5803339115351258, + "grad_norm": 2.71132493019104, + "learning_rate": 1.6189612650515016e-05, + "loss": 0.0655, + "step": 5353 + }, + { + "epoch": 0.5804423243712056, + "grad_norm": 2.652209997177124, + "learning_rate": 1.618888727694763e-05, + "loss": 0.0439, + "step": 5354 + }, + { + "epoch": 0.5805507372072853, + "grad_norm": 3.2848522663116455, + "learning_rate": 1.618816190338024e-05, + "loss": 0.0416, + "step": 5355 + }, + { + "epoch": 0.5806591500433651, + "grad_norm": 7.182081699371338, + "learning_rate": 1.6187436529812856e-05, + "loss": 0.0343, + "step": 5356 + }, + { + "epoch": 0.580767562879445, + "grad_norm": 3.0986521244049072, + "learning_rate": 1.6186711156245468e-05, + "loss": 0.0477, + "step": 5357 + }, + { + "epoch": 0.5808759757155247, + "grad_norm": 4.645793437957764, + "learning_rate": 1.618598578267808e-05, + "loss": 0.0817, + "step": 5358 + }, + { + "epoch": 0.5809843885516045, + "grad_norm": 2.810641050338745, + "learning_rate": 1.6185260409110693e-05, + "loss": 0.025, + "step": 5359 + }, + { + "epoch": 0.5810928013876843, + "grad_norm": 3.6191558837890625, + "learning_rate": 1.6184535035543305e-05, + "loss": 0.037, + "step": 5360 + }, + { + "epoch": 0.5812012142237641, + "grad_norm": 2.0025830268859863, + "learning_rate": 1.618380966197592e-05, + "loss": 0.033, + "step": 5361 + }, + { + "epoch": 0.5813096270598439, + "grad_norm": 1.4936468601226807, + "learning_rate": 1.6183084288408532e-05, + "loss": 0.0213, + "step": 5362 + }, + { + "epoch": 0.5814180398959237, + "grad_norm": 2.9745516777038574, + "learning_rate": 1.6182358914841144e-05, + "loss": 0.0807, + "step": 5363 + }, + { + "epoch": 0.5815264527320034, + "grad_norm": 2.603543281555176, + "learning_rate": 1.6181633541273757e-05, + "loss": 0.0573, + "step": 5364 + }, + { + "epoch": 0.5816348655680833, + "grad_norm": 3.5900559425354004, + "learning_rate": 1.6180908167706372e-05, + "loss": 0.0415, + "step": 5365 + }, + { + "epoch": 0.5817432784041631, + "grad_norm": 2.273425817489624, + "learning_rate": 1.6180182794138984e-05, + "loss": 0.0692, + "step": 5366 + }, + { + "epoch": 0.5818516912402428, + "grad_norm": 2.994779586791992, + "learning_rate": 1.6179457420571596e-05, + "loss": 0.0583, + "step": 5367 + }, + { + "epoch": 0.5819601040763226, + "grad_norm": 2.0471532344818115, + "learning_rate": 1.617873204700421e-05, + "loss": 0.0338, + "step": 5368 + }, + { + "epoch": 0.5820685169124025, + "grad_norm": 6.256675720214844, + "learning_rate": 1.617800667343682e-05, + "loss": 0.0299, + "step": 5369 + }, + { + "epoch": 0.5821769297484822, + "grad_norm": 1.7450956106185913, + "learning_rate": 1.6177281299869433e-05, + "loss": 0.0342, + "step": 5370 + }, + { + "epoch": 0.582285342584562, + "grad_norm": 1.0642763376235962, + "learning_rate": 1.6176555926302045e-05, + "loss": 0.0224, + "step": 5371 + }, + { + "epoch": 0.5823937554206418, + "grad_norm": 3.357840061187744, + "learning_rate": 1.617583055273466e-05, + "loss": 0.0299, + "step": 5372 + }, + { + "epoch": 0.5825021682567216, + "grad_norm": 5.563510894775391, + "learning_rate": 1.6175105179167273e-05, + "loss": 0.0618, + "step": 5373 + }, + { + "epoch": 0.5826105810928014, + "grad_norm": 1.9910746812820435, + "learning_rate": 1.6174379805599885e-05, + "loss": 0.0165, + "step": 5374 + }, + { + "epoch": 0.5827189939288812, + "grad_norm": 2.951505422592163, + "learning_rate": 1.6173654432032497e-05, + "loss": 0.0707, + "step": 5375 + }, + { + "epoch": 0.5828274067649609, + "grad_norm": 1.778327226638794, + "learning_rate": 1.617292905846511e-05, + "loss": 0.0461, + "step": 5376 + }, + { + "epoch": 0.5829358196010408, + "grad_norm": 3.877195358276367, + "learning_rate": 1.617220368489772e-05, + "loss": 0.0243, + "step": 5377 + }, + { + "epoch": 0.5830442324371206, + "grad_norm": 3.878209352493286, + "learning_rate": 1.6171478311330337e-05, + "loss": 0.0208, + "step": 5378 + }, + { + "epoch": 0.5831526452732003, + "grad_norm": 1.7767775058746338, + "learning_rate": 1.617075293776295e-05, + "loss": 0.0271, + "step": 5379 + }, + { + "epoch": 0.5832610581092801, + "grad_norm": 2.1039979457855225, + "learning_rate": 1.6170027564195564e-05, + "loss": 0.0411, + "step": 5380 + }, + { + "epoch": 0.58336947094536, + "grad_norm": 2.823613166809082, + "learning_rate": 1.6169302190628177e-05, + "loss": 0.0456, + "step": 5381 + }, + { + "epoch": 0.5834778837814397, + "grad_norm": 1.5867104530334473, + "learning_rate": 1.616857681706079e-05, + "loss": 0.0326, + "step": 5382 + }, + { + "epoch": 0.5835862966175195, + "grad_norm": 8.342266082763672, + "learning_rate": 1.61678514434934e-05, + "loss": 0.0704, + "step": 5383 + }, + { + "epoch": 0.5836947094535994, + "grad_norm": 4.352209091186523, + "learning_rate": 1.6167126069926013e-05, + "loss": 0.0338, + "step": 5384 + }, + { + "epoch": 0.5838031222896791, + "grad_norm": 4.895063877105713, + "learning_rate": 1.6166400696358625e-05, + "loss": 0.0537, + "step": 5385 + }, + { + "epoch": 0.5839115351257589, + "grad_norm": 3.069074869155884, + "learning_rate": 1.6165675322791237e-05, + "loss": 0.0461, + "step": 5386 + }, + { + "epoch": 0.5840199479618386, + "grad_norm": 2.967855930328369, + "learning_rate": 1.6164949949223853e-05, + "loss": 0.0469, + "step": 5387 + }, + { + "epoch": 0.5841283607979185, + "grad_norm": 3.8026938438415527, + "learning_rate": 1.6164224575656465e-05, + "loss": 0.0369, + "step": 5388 + }, + { + "epoch": 0.5842367736339983, + "grad_norm": 2.1685736179351807, + "learning_rate": 1.6163499202089077e-05, + "loss": 0.0563, + "step": 5389 + }, + { + "epoch": 0.584345186470078, + "grad_norm": 2.1127939224243164, + "learning_rate": 1.616277382852169e-05, + "loss": 0.0746, + "step": 5390 + }, + { + "epoch": 0.5844535993061578, + "grad_norm": 3.5730085372924805, + "learning_rate": 1.61620484549543e-05, + "loss": 0.1107, + "step": 5391 + }, + { + "epoch": 0.5845620121422377, + "grad_norm": 1.074644923210144, + "learning_rate": 1.6161323081386914e-05, + "loss": 0.0449, + "step": 5392 + }, + { + "epoch": 0.5846704249783174, + "grad_norm": 1.052332878112793, + "learning_rate": 1.616059770781953e-05, + "loss": 0.0369, + "step": 5393 + }, + { + "epoch": 0.5847788378143972, + "grad_norm": 1.6059283018112183, + "learning_rate": 1.615987233425214e-05, + "loss": 0.0317, + "step": 5394 + }, + { + "epoch": 0.584887250650477, + "grad_norm": 2.2416210174560547, + "learning_rate": 1.6159146960684753e-05, + "loss": 0.0309, + "step": 5395 + }, + { + "epoch": 0.5849956634865568, + "grad_norm": 2.178144931793213, + "learning_rate": 1.615842158711737e-05, + "loss": 0.0521, + "step": 5396 + }, + { + "epoch": 0.5851040763226366, + "grad_norm": 1.0739048719406128, + "learning_rate": 1.615769621354998e-05, + "loss": 0.0284, + "step": 5397 + }, + { + "epoch": 0.5852124891587164, + "grad_norm": 1.0824028253555298, + "learning_rate": 1.6156970839982593e-05, + "loss": 0.0136, + "step": 5398 + }, + { + "epoch": 0.5853209019947961, + "grad_norm": 3.469701051712036, + "learning_rate": 1.6156245466415205e-05, + "loss": 0.0622, + "step": 5399 + }, + { + "epoch": 0.585429314830876, + "grad_norm": 1.545242190361023, + "learning_rate": 1.6155520092847817e-05, + "loss": 0.0301, + "step": 5400 + }, + { + "epoch": 0.5855377276669558, + "grad_norm": 1.826278567314148, + "learning_rate": 1.615479471928043e-05, + "loss": 0.0224, + "step": 5401 + }, + { + "epoch": 0.5856461405030355, + "grad_norm": 2.311178207397461, + "learning_rate": 1.6154069345713045e-05, + "loss": 0.0325, + "step": 5402 + }, + { + "epoch": 0.5857545533391153, + "grad_norm": 2.9956846237182617, + "learning_rate": 1.6153343972145657e-05, + "loss": 0.0458, + "step": 5403 + }, + { + "epoch": 0.5858629661751952, + "grad_norm": 1.922014832496643, + "learning_rate": 1.615261859857827e-05, + "loss": 0.0433, + "step": 5404 + }, + { + "epoch": 0.5859713790112749, + "grad_norm": 2.6191813945770264, + "learning_rate": 1.615189322501088e-05, + "loss": 0.0446, + "step": 5405 + }, + { + "epoch": 0.5860797918473547, + "grad_norm": 1.4238884449005127, + "learning_rate": 1.6151167851443494e-05, + "loss": 0.0444, + "step": 5406 + }, + { + "epoch": 0.5861882046834345, + "grad_norm": 1.3501840829849243, + "learning_rate": 1.6150442477876106e-05, + "loss": 0.0294, + "step": 5407 + }, + { + "epoch": 0.5862966175195143, + "grad_norm": 1.0426664352416992, + "learning_rate": 1.614971710430872e-05, + "loss": 0.0335, + "step": 5408 + }, + { + "epoch": 0.5864050303555941, + "grad_norm": 4.139912128448486, + "learning_rate": 1.6148991730741334e-05, + "loss": 0.0315, + "step": 5409 + }, + { + "epoch": 0.5865134431916739, + "grad_norm": 0.8868893980979919, + "learning_rate": 1.6148266357173946e-05, + "loss": 0.0188, + "step": 5410 + }, + { + "epoch": 0.5866218560277537, + "grad_norm": 2.196030616760254, + "learning_rate": 1.6147540983606558e-05, + "loss": 0.027, + "step": 5411 + }, + { + "epoch": 0.5867302688638335, + "grad_norm": 2.8365790843963623, + "learning_rate": 1.614681561003917e-05, + "loss": 0.0647, + "step": 5412 + }, + { + "epoch": 0.5868386816999133, + "grad_norm": 1.1067583560943604, + "learning_rate": 1.6146090236471786e-05, + "loss": 0.0339, + "step": 5413 + }, + { + "epoch": 0.586947094535993, + "grad_norm": 1.4704557657241821, + "learning_rate": 1.6145364862904398e-05, + "loss": 0.0222, + "step": 5414 + }, + { + "epoch": 0.5870555073720729, + "grad_norm": 3.528404951095581, + "learning_rate": 1.614463948933701e-05, + "loss": 0.0618, + "step": 5415 + }, + { + "epoch": 0.5871639202081527, + "grad_norm": 3.4525275230407715, + "learning_rate": 1.6143914115769622e-05, + "loss": 0.0705, + "step": 5416 + }, + { + "epoch": 0.5872723330442324, + "grad_norm": 1.474904179573059, + "learning_rate": 1.6143188742202237e-05, + "loss": 0.0134, + "step": 5417 + }, + { + "epoch": 0.5873807458803122, + "grad_norm": 3.363647937774658, + "learning_rate": 1.614246336863485e-05, + "loss": 0.0358, + "step": 5418 + }, + { + "epoch": 0.5874891587163921, + "grad_norm": 3.5775012969970703, + "learning_rate": 1.6141737995067462e-05, + "loss": 0.076, + "step": 5419 + }, + { + "epoch": 0.5875975715524718, + "grad_norm": 1.6805771589279175, + "learning_rate": 1.6141012621500074e-05, + "loss": 0.0344, + "step": 5420 + }, + { + "epoch": 0.5877059843885516, + "grad_norm": 1.679589033126831, + "learning_rate": 1.6140287247932686e-05, + "loss": 0.0267, + "step": 5421 + }, + { + "epoch": 0.5878143972246314, + "grad_norm": 2.1762514114379883, + "learning_rate": 1.6139561874365298e-05, + "loss": 0.0192, + "step": 5422 + }, + { + "epoch": 0.5879228100607112, + "grad_norm": 4.45921516418457, + "learning_rate": 1.613883650079791e-05, + "loss": 0.0805, + "step": 5423 + }, + { + "epoch": 0.588031222896791, + "grad_norm": 1.8251148462295532, + "learning_rate": 1.6138111127230526e-05, + "loss": 0.0322, + "step": 5424 + }, + { + "epoch": 0.5881396357328708, + "grad_norm": 1.3517975807189941, + "learning_rate": 1.6137385753663138e-05, + "loss": 0.013, + "step": 5425 + }, + { + "epoch": 0.5882480485689505, + "grad_norm": 8.502799987792969, + "learning_rate": 1.613666038009575e-05, + "loss": 0.0581, + "step": 5426 + }, + { + "epoch": 0.5883564614050304, + "grad_norm": 1.4082425832748413, + "learning_rate": 1.6135935006528362e-05, + "loss": 0.0046, + "step": 5427 + }, + { + "epoch": 0.5884648742411102, + "grad_norm": 5.181066513061523, + "learning_rate": 1.6135209632960974e-05, + "loss": 0.0742, + "step": 5428 + }, + { + "epoch": 0.5885732870771899, + "grad_norm": 1.8884817361831665, + "learning_rate": 1.6134484259393587e-05, + "loss": 0.0232, + "step": 5429 + }, + { + "epoch": 0.5886816999132697, + "grad_norm": 1.6525664329528809, + "learning_rate": 1.6133758885826202e-05, + "loss": 0.0469, + "step": 5430 + }, + { + "epoch": 0.5887901127493496, + "grad_norm": 5.909686088562012, + "learning_rate": 1.6133033512258814e-05, + "loss": 0.0624, + "step": 5431 + }, + { + "epoch": 0.5888985255854293, + "grad_norm": 4.705449104309082, + "learning_rate": 1.613230813869143e-05, + "loss": 0.0439, + "step": 5432 + }, + { + "epoch": 0.5890069384215091, + "grad_norm": 3.389177083969116, + "learning_rate": 1.6131582765124042e-05, + "loss": 0.0327, + "step": 5433 + }, + { + "epoch": 0.5891153512575888, + "grad_norm": 2.40138840675354, + "learning_rate": 1.6130857391556654e-05, + "loss": 0.037, + "step": 5434 + }, + { + "epoch": 0.5892237640936687, + "grad_norm": 1.881539225578308, + "learning_rate": 1.6130132017989266e-05, + "loss": 0.0312, + "step": 5435 + }, + { + "epoch": 0.5893321769297485, + "grad_norm": 2.94512939453125, + "learning_rate": 1.612940664442188e-05, + "loss": 0.0222, + "step": 5436 + }, + { + "epoch": 0.5894405897658282, + "grad_norm": 3.7645649909973145, + "learning_rate": 1.612868127085449e-05, + "loss": 0.0718, + "step": 5437 + }, + { + "epoch": 0.589549002601908, + "grad_norm": 6.9969282150268555, + "learning_rate": 1.6127955897287103e-05, + "loss": 0.0444, + "step": 5438 + }, + { + "epoch": 0.5896574154379879, + "grad_norm": 4.887207984924316, + "learning_rate": 1.6127230523719718e-05, + "loss": 0.0808, + "step": 5439 + }, + { + "epoch": 0.5897658282740676, + "grad_norm": 3.8772802352905273, + "learning_rate": 1.612650515015233e-05, + "loss": 0.0144, + "step": 5440 + }, + { + "epoch": 0.5898742411101474, + "grad_norm": 4.717006206512451, + "learning_rate": 1.6125779776584942e-05, + "loss": 0.112, + "step": 5441 + }, + { + "epoch": 0.5899826539462273, + "grad_norm": 2.408407211303711, + "learning_rate": 1.6125054403017555e-05, + "loss": 0.0197, + "step": 5442 + }, + { + "epoch": 0.590091066782307, + "grad_norm": 2.2943432331085205, + "learning_rate": 1.6124329029450167e-05, + "loss": 0.0234, + "step": 5443 + }, + { + "epoch": 0.5901994796183868, + "grad_norm": 3.2456440925598145, + "learning_rate": 1.612360365588278e-05, + "loss": 0.0566, + "step": 5444 + }, + { + "epoch": 0.5903078924544666, + "grad_norm": 1.8319733142852783, + "learning_rate": 1.6122878282315394e-05, + "loss": 0.0323, + "step": 5445 + }, + { + "epoch": 0.5904163052905465, + "grad_norm": 2.9747495651245117, + "learning_rate": 1.6122152908748007e-05, + "loss": 0.0664, + "step": 5446 + }, + { + "epoch": 0.5905247181266262, + "grad_norm": 1.3865898847579956, + "learning_rate": 1.6121427535180622e-05, + "loss": 0.0237, + "step": 5447 + }, + { + "epoch": 0.590633130962706, + "grad_norm": 0.9991604089736938, + "learning_rate": 1.6120702161613234e-05, + "loss": 0.0333, + "step": 5448 + }, + { + "epoch": 0.5907415437987857, + "grad_norm": 5.400475025177002, + "learning_rate": 1.6119976788045846e-05, + "loss": 0.0858, + "step": 5449 + }, + { + "epoch": 0.5908499566348656, + "grad_norm": 2.1881496906280518, + "learning_rate": 1.611925141447846e-05, + "loss": 0.0459, + "step": 5450 + }, + { + "epoch": 0.5909583694709454, + "grad_norm": 2.393329620361328, + "learning_rate": 1.611852604091107e-05, + "loss": 0.0182, + "step": 5451 + }, + { + "epoch": 0.5910667823070251, + "grad_norm": 1.372984528541565, + "learning_rate": 1.6117800667343683e-05, + "loss": 0.0148, + "step": 5452 + }, + { + "epoch": 0.5911751951431049, + "grad_norm": 1.3181257247924805, + "learning_rate": 1.6117075293776295e-05, + "loss": 0.0549, + "step": 5453 + }, + { + "epoch": 0.5912836079791848, + "grad_norm": 0.6383550763130188, + "learning_rate": 1.611634992020891e-05, + "loss": 0.0078, + "step": 5454 + }, + { + "epoch": 0.5913920208152645, + "grad_norm": 2.8114888668060303, + "learning_rate": 1.6115624546641523e-05, + "loss": 0.0462, + "step": 5455 + }, + { + "epoch": 0.5915004336513443, + "grad_norm": 1.5287222862243652, + "learning_rate": 1.6114899173074135e-05, + "loss": 0.0137, + "step": 5456 + }, + { + "epoch": 0.5916088464874241, + "grad_norm": 1.2191150188446045, + "learning_rate": 1.6114173799506747e-05, + "loss": 0.0334, + "step": 5457 + }, + { + "epoch": 0.5917172593235039, + "grad_norm": 1.4875860214233398, + "learning_rate": 1.611344842593936e-05, + "loss": 0.0131, + "step": 5458 + }, + { + "epoch": 0.5918256721595837, + "grad_norm": 1.440416932106018, + "learning_rate": 1.611272305237197e-05, + "loss": 0.044, + "step": 5459 + }, + { + "epoch": 0.5919340849956635, + "grad_norm": 3.7150816917419434, + "learning_rate": 1.6111997678804587e-05, + "loss": 0.0139, + "step": 5460 + }, + { + "epoch": 0.5920424978317432, + "grad_norm": 3.063979148864746, + "learning_rate": 1.61112723052372e-05, + "loss": 0.0524, + "step": 5461 + }, + { + "epoch": 0.5921509106678231, + "grad_norm": 1.5694106817245483, + "learning_rate": 1.611054693166981e-05, + "loss": 0.0248, + "step": 5462 + }, + { + "epoch": 0.5922593235039029, + "grad_norm": 1.416095495223999, + "learning_rate": 1.6109821558102423e-05, + "loss": 0.0142, + "step": 5463 + }, + { + "epoch": 0.5923677363399826, + "grad_norm": 3.6799275875091553, + "learning_rate": 1.6109096184535035e-05, + "loss": 0.0374, + "step": 5464 + }, + { + "epoch": 0.5924761491760624, + "grad_norm": 1.6527565717697144, + "learning_rate": 1.610837081096765e-05, + "loss": 0.0415, + "step": 5465 + }, + { + "epoch": 0.5925845620121423, + "grad_norm": 4.068950653076172, + "learning_rate": 1.6107645437400263e-05, + "loss": 0.0719, + "step": 5466 + }, + { + "epoch": 0.592692974848222, + "grad_norm": 0.9695519804954529, + "learning_rate": 1.6106920063832875e-05, + "loss": 0.0195, + "step": 5467 + }, + { + "epoch": 0.5928013876843018, + "grad_norm": 5.134016036987305, + "learning_rate": 1.6106194690265487e-05, + "loss": 0.0628, + "step": 5468 + }, + { + "epoch": 0.5929098005203816, + "grad_norm": 0.37964069843292236, + "learning_rate": 1.6105469316698103e-05, + "loss": 0.0044, + "step": 5469 + }, + { + "epoch": 0.5930182133564614, + "grad_norm": 3.4860873222351074, + "learning_rate": 1.6104743943130715e-05, + "loss": 0.0701, + "step": 5470 + }, + { + "epoch": 0.5931266261925412, + "grad_norm": 1.8861538171768188, + "learning_rate": 1.6104018569563327e-05, + "loss": 0.0208, + "step": 5471 + }, + { + "epoch": 0.593235039028621, + "grad_norm": 1.8395347595214844, + "learning_rate": 1.610329319599594e-05, + "loss": 0.0348, + "step": 5472 + }, + { + "epoch": 0.5933434518647008, + "grad_norm": 8.386988639831543, + "learning_rate": 1.610256782242855e-05, + "loss": 0.0803, + "step": 5473 + }, + { + "epoch": 0.5934518647007806, + "grad_norm": 3.8780159950256348, + "learning_rate": 1.6101842448861164e-05, + "loss": 0.033, + "step": 5474 + }, + { + "epoch": 0.5935602775368604, + "grad_norm": 0.37065839767456055, + "learning_rate": 1.6101117075293776e-05, + "loss": 0.0112, + "step": 5475 + }, + { + "epoch": 0.5936686903729401, + "grad_norm": 3.6761200428009033, + "learning_rate": 1.610039170172639e-05, + "loss": 0.0558, + "step": 5476 + }, + { + "epoch": 0.59377710320902, + "grad_norm": 4.924239635467529, + "learning_rate": 1.6099666328159003e-05, + "loss": 0.032, + "step": 5477 + }, + { + "epoch": 0.5938855160450998, + "grad_norm": 7.297539234161377, + "learning_rate": 1.6098940954591616e-05, + "loss": 0.057, + "step": 5478 + }, + { + "epoch": 0.5939939288811795, + "grad_norm": 2.1984176635742188, + "learning_rate": 1.6098215581024228e-05, + "loss": 0.0352, + "step": 5479 + }, + { + "epoch": 0.5941023417172593, + "grad_norm": 1.0203698873519897, + "learning_rate": 1.609749020745684e-05, + "loss": 0.0146, + "step": 5480 + }, + { + "epoch": 0.5942107545533392, + "grad_norm": 3.9794259071350098, + "learning_rate": 1.6096764833889452e-05, + "loss": 0.0379, + "step": 5481 + }, + { + "epoch": 0.5943191673894189, + "grad_norm": 1.9499434232711792, + "learning_rate": 1.6096039460322067e-05, + "loss": 0.0312, + "step": 5482 + }, + { + "epoch": 0.5944275802254987, + "grad_norm": 3.792969226837158, + "learning_rate": 1.609531408675468e-05, + "loss": 0.0265, + "step": 5483 + }, + { + "epoch": 0.5945359930615784, + "grad_norm": 2.592094898223877, + "learning_rate": 1.6094588713187295e-05, + "loss": 0.0289, + "step": 5484 + }, + { + "epoch": 0.5946444058976583, + "grad_norm": 1.4240034818649292, + "learning_rate": 1.6093863339619907e-05, + "loss": 0.0221, + "step": 5485 + }, + { + "epoch": 0.5947528187337381, + "grad_norm": 4.997046947479248, + "learning_rate": 1.609313796605252e-05, + "loss": 0.0446, + "step": 5486 + }, + { + "epoch": 0.5948612315698178, + "grad_norm": 1.3244051933288574, + "learning_rate": 1.609241259248513e-05, + "loss": 0.0175, + "step": 5487 + }, + { + "epoch": 0.5949696444058976, + "grad_norm": 4.31507682800293, + "learning_rate": 1.6091687218917744e-05, + "loss": 0.0508, + "step": 5488 + }, + { + "epoch": 0.5950780572419775, + "grad_norm": 3.3385086059570312, + "learning_rate": 1.6090961845350356e-05, + "loss": 0.0354, + "step": 5489 + }, + { + "epoch": 0.5951864700780573, + "grad_norm": 3.022871732711792, + "learning_rate": 1.6090236471782968e-05, + "loss": 0.0238, + "step": 5490 + }, + { + "epoch": 0.595294882914137, + "grad_norm": 1.071752905845642, + "learning_rate": 1.6089511098215584e-05, + "loss": 0.0111, + "step": 5491 + }, + { + "epoch": 0.5954032957502168, + "grad_norm": 0.9409400820732117, + "learning_rate": 1.6088785724648196e-05, + "loss": 0.0046, + "step": 5492 + }, + { + "epoch": 0.5955117085862967, + "grad_norm": 1.4608052968978882, + "learning_rate": 1.6088060351080808e-05, + "loss": 0.0368, + "step": 5493 + }, + { + "epoch": 0.5956201214223764, + "grad_norm": 2.121882915496826, + "learning_rate": 1.608733497751342e-05, + "loss": 0.0196, + "step": 5494 + }, + { + "epoch": 0.5957285342584562, + "grad_norm": 5.46293306350708, + "learning_rate": 1.6086609603946032e-05, + "loss": 0.0697, + "step": 5495 + }, + { + "epoch": 0.5958369470945359, + "grad_norm": 0.8905203938484192, + "learning_rate": 1.6085884230378644e-05, + "loss": 0.0133, + "step": 5496 + }, + { + "epoch": 0.5959453599306158, + "grad_norm": 1.409909963607788, + "learning_rate": 1.608515885681126e-05, + "loss": 0.0172, + "step": 5497 + }, + { + "epoch": 0.5960537727666956, + "grad_norm": 3.155043125152588, + "learning_rate": 1.6084433483243872e-05, + "loss": 0.0693, + "step": 5498 + }, + { + "epoch": 0.5961621856027753, + "grad_norm": 3.9829611778259277, + "learning_rate": 1.6083708109676487e-05, + "loss": 0.0285, + "step": 5499 + }, + { + "epoch": 0.5962705984388552, + "grad_norm": 4.293179988861084, + "learning_rate": 1.60829827361091e-05, + "loss": 0.0463, + "step": 5500 + }, + { + "epoch": 0.596379011274935, + "grad_norm": 0.5905930995941162, + "learning_rate": 1.6082257362541712e-05, + "loss": 0.0148, + "step": 5501 + }, + { + "epoch": 0.5964874241110147, + "grad_norm": 2.9253456592559814, + "learning_rate": 1.6081531988974324e-05, + "loss": 0.0256, + "step": 5502 + }, + { + "epoch": 0.5965958369470945, + "grad_norm": 0.5221245288848877, + "learning_rate": 1.6080806615406936e-05, + "loss": 0.0042, + "step": 5503 + }, + { + "epoch": 0.5967042497831744, + "grad_norm": 2.22037935256958, + "learning_rate": 1.6080081241839548e-05, + "loss": 0.0276, + "step": 5504 + }, + { + "epoch": 0.5968126626192541, + "grad_norm": 5.352254867553711, + "learning_rate": 1.607935586827216e-05, + "loss": 0.0619, + "step": 5505 + }, + { + "epoch": 0.5969210754553339, + "grad_norm": 3.040862798690796, + "learning_rate": 1.6078630494704776e-05, + "loss": 0.0056, + "step": 5506 + }, + { + "epoch": 0.5970294882914137, + "grad_norm": 2.2158594131469727, + "learning_rate": 1.6077905121137388e-05, + "loss": 0.0312, + "step": 5507 + }, + { + "epoch": 0.5971379011274935, + "grad_norm": 3.734488010406494, + "learning_rate": 1.607717974757e-05, + "loss": 0.0411, + "step": 5508 + }, + { + "epoch": 0.5972463139635733, + "grad_norm": 3.0335586071014404, + "learning_rate": 1.6076454374002612e-05, + "loss": 0.05, + "step": 5509 + }, + { + "epoch": 0.5973547267996531, + "grad_norm": 2.5104243755340576, + "learning_rate": 1.6075729000435224e-05, + "loss": 0.0619, + "step": 5510 + }, + { + "epoch": 0.5974631396357328, + "grad_norm": 3.5064167976379395, + "learning_rate": 1.6075003626867837e-05, + "loss": 0.0407, + "step": 5511 + }, + { + "epoch": 0.5975715524718127, + "grad_norm": 1.6664915084838867, + "learning_rate": 1.6074278253300452e-05, + "loss": 0.0097, + "step": 5512 + }, + { + "epoch": 0.5976799653078925, + "grad_norm": 7.996669292449951, + "learning_rate": 1.6073552879733064e-05, + "loss": 0.0679, + "step": 5513 + }, + { + "epoch": 0.5977883781439722, + "grad_norm": 5.458017349243164, + "learning_rate": 1.6072827506165676e-05, + "loss": 0.0624, + "step": 5514 + }, + { + "epoch": 0.597896790980052, + "grad_norm": 2.7159337997436523, + "learning_rate": 1.607210213259829e-05, + "loss": 0.0173, + "step": 5515 + }, + { + "epoch": 0.5980052038161319, + "grad_norm": 4.1752848625183105, + "learning_rate": 1.6071376759030904e-05, + "loss": 0.0263, + "step": 5516 + }, + { + "epoch": 0.5981136166522116, + "grad_norm": 6.924830913543701, + "learning_rate": 1.6070651385463516e-05, + "loss": 0.05, + "step": 5517 + }, + { + "epoch": 0.5982220294882914, + "grad_norm": 1.7068867683410645, + "learning_rate": 1.606992601189613e-05, + "loss": 0.0256, + "step": 5518 + }, + { + "epoch": 0.5983304423243712, + "grad_norm": 2.9448487758636475, + "learning_rate": 1.606920063832874e-05, + "loss": 0.0437, + "step": 5519 + }, + { + "epoch": 0.598438855160451, + "grad_norm": 1.363142490386963, + "learning_rate": 1.6068475264761353e-05, + "loss": 0.0308, + "step": 5520 + }, + { + "epoch": 0.5985472679965308, + "grad_norm": 2.6197402477264404, + "learning_rate": 1.6067749891193968e-05, + "loss": 0.0245, + "step": 5521 + }, + { + "epoch": 0.5986556808326106, + "grad_norm": 2.7393171787261963, + "learning_rate": 1.606702451762658e-05, + "loss": 0.0598, + "step": 5522 + }, + { + "epoch": 0.5987640936686903, + "grad_norm": 4.0074052810668945, + "learning_rate": 1.6066299144059192e-05, + "loss": 0.0755, + "step": 5523 + }, + { + "epoch": 0.5988725065047702, + "grad_norm": 1.8751511573791504, + "learning_rate": 1.6065573770491805e-05, + "loss": 0.0377, + "step": 5524 + }, + { + "epoch": 0.59898091934085, + "grad_norm": 1.8070085048675537, + "learning_rate": 1.6064848396924417e-05, + "loss": 0.0273, + "step": 5525 + }, + { + "epoch": 0.5990893321769297, + "grad_norm": 1.5537409782409668, + "learning_rate": 1.606412302335703e-05, + "loss": 0.0089, + "step": 5526 + }, + { + "epoch": 0.5991977450130095, + "grad_norm": 5.896933555603027, + "learning_rate": 1.606339764978964e-05, + "loss": 0.0701, + "step": 5527 + }, + { + "epoch": 0.5993061578490894, + "grad_norm": 0.5789577960968018, + "learning_rate": 1.6062672276222257e-05, + "loss": 0.0049, + "step": 5528 + }, + { + "epoch": 0.5994145706851691, + "grad_norm": 2.7624335289001465, + "learning_rate": 1.606194690265487e-05, + "loss": 0.0471, + "step": 5529 + }, + { + "epoch": 0.5995229835212489, + "grad_norm": 3.2075788974761963, + "learning_rate": 1.606122152908748e-05, + "loss": 0.0795, + "step": 5530 + }, + { + "epoch": 0.5996313963573288, + "grad_norm": 0.2294722944498062, + "learning_rate": 1.6060496155520093e-05, + "loss": 0.0041, + "step": 5531 + }, + { + "epoch": 0.5997398091934085, + "grad_norm": 1.7948626279830933, + "learning_rate": 1.6059770781952705e-05, + "loss": 0.0357, + "step": 5532 + }, + { + "epoch": 0.5998482220294883, + "grad_norm": 1.4938404560089111, + "learning_rate": 1.6059045408385317e-05, + "loss": 0.0154, + "step": 5533 + }, + { + "epoch": 0.599956634865568, + "grad_norm": 4.4550580978393555, + "learning_rate": 1.6058320034817933e-05, + "loss": 0.0656, + "step": 5534 + }, + { + "epoch": 0.6000650477016479, + "grad_norm": 3.89803147315979, + "learning_rate": 1.6057594661250545e-05, + "loss": 0.1231, + "step": 5535 + }, + { + "epoch": 0.6001734605377277, + "grad_norm": 2.2290308475494385, + "learning_rate": 1.605686928768316e-05, + "loss": 0.0341, + "step": 5536 + }, + { + "epoch": 0.6002818733738075, + "grad_norm": 1.0782015323638916, + "learning_rate": 1.6056143914115773e-05, + "loss": 0.0158, + "step": 5537 + }, + { + "epoch": 0.6003902862098872, + "grad_norm": 5.886460781097412, + "learning_rate": 1.6055418540548385e-05, + "loss": 0.042, + "step": 5538 + }, + { + "epoch": 0.6004986990459671, + "grad_norm": 6.854816436767578, + "learning_rate": 1.6054693166980997e-05, + "loss": 0.0542, + "step": 5539 + }, + { + "epoch": 0.6006071118820469, + "grad_norm": 3.8362793922424316, + "learning_rate": 1.605396779341361e-05, + "loss": 0.0406, + "step": 5540 + }, + { + "epoch": 0.6007155247181266, + "grad_norm": 4.445630073547363, + "learning_rate": 1.605324241984622e-05, + "loss": 0.0544, + "step": 5541 + }, + { + "epoch": 0.6008239375542064, + "grad_norm": 4.786797523498535, + "learning_rate": 1.6052517046278833e-05, + "loss": 0.0866, + "step": 5542 + }, + { + "epoch": 0.6009323503902863, + "grad_norm": 3.1810646057128906, + "learning_rate": 1.605179167271145e-05, + "loss": 0.0404, + "step": 5543 + }, + { + "epoch": 0.601040763226366, + "grad_norm": 2.5283329486846924, + "learning_rate": 1.605106629914406e-05, + "loss": 0.0645, + "step": 5544 + }, + { + "epoch": 0.6011491760624458, + "grad_norm": 3.2936856746673584, + "learning_rate": 1.6050340925576673e-05, + "loss": 0.046, + "step": 5545 + }, + { + "epoch": 0.6012575888985255, + "grad_norm": 5.274336814880371, + "learning_rate": 1.6049615552009285e-05, + "loss": 0.0627, + "step": 5546 + }, + { + "epoch": 0.6013660017346054, + "grad_norm": 1.8217947483062744, + "learning_rate": 1.6048890178441897e-05, + "loss": 0.0326, + "step": 5547 + }, + { + "epoch": 0.6014744145706852, + "grad_norm": 2.9483911991119385, + "learning_rate": 1.604816480487451e-05, + "loss": 0.0259, + "step": 5548 + }, + { + "epoch": 0.6015828274067649, + "grad_norm": 4.079671382904053, + "learning_rate": 1.6047439431307125e-05, + "loss": 0.0262, + "step": 5549 + }, + { + "epoch": 0.6016912402428447, + "grad_norm": 2.7005186080932617, + "learning_rate": 1.6046714057739737e-05, + "loss": 0.0283, + "step": 5550 + }, + { + "epoch": 0.6017996530789246, + "grad_norm": 2.047776937484741, + "learning_rate": 1.6045988684172353e-05, + "loss": 0.0387, + "step": 5551 + }, + { + "epoch": 0.6019080659150043, + "grad_norm": 2.206900119781494, + "learning_rate": 1.6045263310604965e-05, + "loss": 0.0376, + "step": 5552 + }, + { + "epoch": 0.6020164787510841, + "grad_norm": 6.4269561767578125, + "learning_rate": 1.6044537937037577e-05, + "loss": 0.0323, + "step": 5553 + }, + { + "epoch": 0.6021248915871639, + "grad_norm": 1.4265491962432861, + "learning_rate": 1.604381256347019e-05, + "loss": 0.0359, + "step": 5554 + }, + { + "epoch": 0.6022333044232437, + "grad_norm": 3.508159637451172, + "learning_rate": 1.60430871899028e-05, + "loss": 0.064, + "step": 5555 + }, + { + "epoch": 0.6023417172593235, + "grad_norm": 3.9541900157928467, + "learning_rate": 1.6042361816335414e-05, + "loss": 0.0564, + "step": 5556 + }, + { + "epoch": 0.6024501300954033, + "grad_norm": 2.6671180725097656, + "learning_rate": 1.6041636442768026e-05, + "loss": 0.0578, + "step": 5557 + }, + { + "epoch": 0.602558542931483, + "grad_norm": 3.690711259841919, + "learning_rate": 1.604091106920064e-05, + "loss": 0.0542, + "step": 5558 + }, + { + "epoch": 0.6026669557675629, + "grad_norm": 5.9566874504089355, + "learning_rate": 1.6040185695633253e-05, + "loss": 0.0869, + "step": 5559 + }, + { + "epoch": 0.6027753686036427, + "grad_norm": 3.417764663696289, + "learning_rate": 1.6039460322065865e-05, + "loss": 0.0527, + "step": 5560 + }, + { + "epoch": 0.6028837814397224, + "grad_norm": 1.870475172996521, + "learning_rate": 1.6038734948498478e-05, + "loss": 0.0571, + "step": 5561 + }, + { + "epoch": 0.6029921942758023, + "grad_norm": 2.4465696811676025, + "learning_rate": 1.603800957493109e-05, + "loss": 0.0534, + "step": 5562 + }, + { + "epoch": 0.6031006071118821, + "grad_norm": 2.075838804244995, + "learning_rate": 1.6037284201363702e-05, + "loss": 0.032, + "step": 5563 + }, + { + "epoch": 0.6032090199479618, + "grad_norm": 4.773810863494873, + "learning_rate": 1.6036558827796317e-05, + "loss": 0.0578, + "step": 5564 + }, + { + "epoch": 0.6033174327840416, + "grad_norm": 4.347052574157715, + "learning_rate": 1.603583345422893e-05, + "loss": 0.0317, + "step": 5565 + }, + { + "epoch": 0.6034258456201215, + "grad_norm": 0.952893853187561, + "learning_rate": 1.6035108080661542e-05, + "loss": 0.0283, + "step": 5566 + }, + { + "epoch": 0.6035342584562012, + "grad_norm": 2.3940603733062744, + "learning_rate": 1.6034382707094154e-05, + "loss": 0.0488, + "step": 5567 + }, + { + "epoch": 0.603642671292281, + "grad_norm": 2.669703960418701, + "learning_rate": 1.603365733352677e-05, + "loss": 0.0209, + "step": 5568 + }, + { + "epoch": 0.6037510841283608, + "grad_norm": 1.9302526712417603, + "learning_rate": 1.603293195995938e-05, + "loss": 0.0402, + "step": 5569 + }, + { + "epoch": 0.6038594969644406, + "grad_norm": 2.666539192199707, + "learning_rate": 1.6032206586391994e-05, + "loss": 0.0427, + "step": 5570 + }, + { + "epoch": 0.6039679098005204, + "grad_norm": 3.5776500701904297, + "learning_rate": 1.6031481212824606e-05, + "loss": 0.0692, + "step": 5571 + }, + { + "epoch": 0.6040763226366002, + "grad_norm": 1.890588641166687, + "learning_rate": 1.6030755839257218e-05, + "loss": 0.0353, + "step": 5572 + }, + { + "epoch": 0.6041847354726799, + "grad_norm": 1.7375061511993408, + "learning_rate": 1.6030030465689834e-05, + "loss": 0.0328, + "step": 5573 + }, + { + "epoch": 0.6042931483087598, + "grad_norm": 2.648423671722412, + "learning_rate": 1.6029305092122446e-05, + "loss": 0.032, + "step": 5574 + }, + { + "epoch": 0.6044015611448396, + "grad_norm": 3.1148452758789062, + "learning_rate": 1.6028579718555058e-05, + "loss": 0.0468, + "step": 5575 + }, + { + "epoch": 0.6045099739809193, + "grad_norm": 1.4078164100646973, + "learning_rate": 1.602785434498767e-05, + "loss": 0.0333, + "step": 5576 + }, + { + "epoch": 0.6046183868169991, + "grad_norm": 1.877724051475525, + "learning_rate": 1.6027128971420282e-05, + "loss": 0.0251, + "step": 5577 + }, + { + "epoch": 0.604726799653079, + "grad_norm": 3.4628679752349854, + "learning_rate": 1.6026403597852894e-05, + "loss": 0.1092, + "step": 5578 + }, + { + "epoch": 0.6048352124891587, + "grad_norm": 1.9903602600097656, + "learning_rate": 1.6025678224285506e-05, + "loss": 0.0364, + "step": 5579 + }, + { + "epoch": 0.6049436253252385, + "grad_norm": 0.9694215059280396, + "learning_rate": 1.6024952850718122e-05, + "loss": 0.0115, + "step": 5580 + }, + { + "epoch": 0.6050520381613183, + "grad_norm": 8.493644714355469, + "learning_rate": 1.6024227477150734e-05, + "loss": 0.0728, + "step": 5581 + }, + { + "epoch": 0.6051604509973981, + "grad_norm": 1.6446504592895508, + "learning_rate": 1.6023502103583346e-05, + "loss": 0.0364, + "step": 5582 + }, + { + "epoch": 0.6052688638334779, + "grad_norm": 46.89434051513672, + "learning_rate": 1.602277673001596e-05, + "loss": 0.0658, + "step": 5583 + }, + { + "epoch": 0.6053772766695577, + "grad_norm": 8.315254211425781, + "learning_rate": 1.602205135644857e-05, + "loss": 0.047, + "step": 5584 + }, + { + "epoch": 0.6054856895056374, + "grad_norm": 4.183222770690918, + "learning_rate": 1.6021325982881183e-05, + "loss": 0.0646, + "step": 5585 + }, + { + "epoch": 0.6055941023417173, + "grad_norm": 3.5290606021881104, + "learning_rate": 1.6020600609313798e-05, + "loss": 0.0432, + "step": 5586 + }, + { + "epoch": 0.605702515177797, + "grad_norm": 3.2335543632507324, + "learning_rate": 1.601987523574641e-05, + "loss": 0.0606, + "step": 5587 + }, + { + "epoch": 0.6058109280138768, + "grad_norm": 4.547922611236572, + "learning_rate": 1.6019149862179026e-05, + "loss": 0.1166, + "step": 5588 + }, + { + "epoch": 0.6059193408499567, + "grad_norm": 2.9397132396698, + "learning_rate": 1.6018424488611638e-05, + "loss": 0.0659, + "step": 5589 + }, + { + "epoch": 0.6060277536860365, + "grad_norm": 1.5313597917556763, + "learning_rate": 1.601769911504425e-05, + "loss": 0.0254, + "step": 5590 + }, + { + "epoch": 0.6061361665221162, + "grad_norm": 3.2841057777404785, + "learning_rate": 1.6016973741476862e-05, + "loss": 0.0418, + "step": 5591 + }, + { + "epoch": 0.606244579358196, + "grad_norm": 3.0590991973876953, + "learning_rate": 1.6016248367909474e-05, + "loss": 0.0149, + "step": 5592 + }, + { + "epoch": 0.6063529921942759, + "grad_norm": 2.283186435699463, + "learning_rate": 1.6015522994342087e-05, + "loss": 0.0365, + "step": 5593 + }, + { + "epoch": 0.6064614050303556, + "grad_norm": 2.731889247894287, + "learning_rate": 1.60147976207747e-05, + "loss": 0.0367, + "step": 5594 + }, + { + "epoch": 0.6065698178664354, + "grad_norm": 4.040363311767578, + "learning_rate": 1.6014072247207314e-05, + "loss": 0.0485, + "step": 5595 + }, + { + "epoch": 0.6066782307025151, + "grad_norm": 1.3000307083129883, + "learning_rate": 1.6013346873639926e-05, + "loss": 0.0132, + "step": 5596 + }, + { + "epoch": 0.606786643538595, + "grad_norm": 3.6399693489074707, + "learning_rate": 1.601262150007254e-05, + "loss": 0.0786, + "step": 5597 + }, + { + "epoch": 0.6068950563746748, + "grad_norm": 1.3030973672866821, + "learning_rate": 1.601189612650515e-05, + "loss": 0.0179, + "step": 5598 + }, + { + "epoch": 0.6070034692107545, + "grad_norm": 0.7235309481620789, + "learning_rate": 1.6011170752937763e-05, + "loss": 0.009, + "step": 5599 + }, + { + "epoch": 0.6071118820468343, + "grad_norm": 5.926695823669434, + "learning_rate": 1.6010445379370375e-05, + "loss": 0.0418, + "step": 5600 + }, + { + "epoch": 0.6072202948829142, + "grad_norm": 1.4028840065002441, + "learning_rate": 1.600972000580299e-05, + "loss": 0.0119, + "step": 5601 + }, + { + "epoch": 0.6073287077189939, + "grad_norm": 1.7720000743865967, + "learning_rate": 1.6008994632235603e-05, + "loss": 0.0273, + "step": 5602 + }, + { + "epoch": 0.6074371205550737, + "grad_norm": 0.32815784215927124, + "learning_rate": 1.6008269258668218e-05, + "loss": 0.003, + "step": 5603 + }, + { + "epoch": 0.6075455333911535, + "grad_norm": 0.9144414067268372, + "learning_rate": 1.600754388510083e-05, + "loss": 0.0059, + "step": 5604 + }, + { + "epoch": 0.6076539462272333, + "grad_norm": 3.8624486923217773, + "learning_rate": 1.6006818511533442e-05, + "loss": 0.034, + "step": 5605 + }, + { + "epoch": 0.6077623590633131, + "grad_norm": 0.9615346789360046, + "learning_rate": 1.6006093137966055e-05, + "loss": 0.0118, + "step": 5606 + }, + { + "epoch": 0.6078707718993929, + "grad_norm": 1.1123818159103394, + "learning_rate": 1.6005367764398667e-05, + "loss": 0.0213, + "step": 5607 + }, + { + "epoch": 0.6079791847354726, + "grad_norm": 4.551454067230225, + "learning_rate": 1.600464239083128e-05, + "loss": 0.0223, + "step": 5608 + }, + { + "epoch": 0.6080875975715525, + "grad_norm": 8.329451560974121, + "learning_rate": 1.600391701726389e-05, + "loss": 0.1042, + "step": 5609 + }, + { + "epoch": 0.6081960104076323, + "grad_norm": 2.075676679611206, + "learning_rate": 1.6003191643696507e-05, + "loss": 0.036, + "step": 5610 + }, + { + "epoch": 0.608304423243712, + "grad_norm": 6.67831563949585, + "learning_rate": 1.600246627012912e-05, + "loss": 0.0409, + "step": 5611 + }, + { + "epoch": 0.6084128360797918, + "grad_norm": 3.1996757984161377, + "learning_rate": 1.600174089656173e-05, + "loss": 0.0449, + "step": 5612 + }, + { + "epoch": 0.6085212489158717, + "grad_norm": 0.3235401213169098, + "learning_rate": 1.6001015522994343e-05, + "loss": 0.0062, + "step": 5613 + }, + { + "epoch": 0.6086296617519514, + "grad_norm": 4.181478977203369, + "learning_rate": 1.6000290149426955e-05, + "loss": 0.0337, + "step": 5614 + }, + { + "epoch": 0.6087380745880312, + "grad_norm": 2.8165628910064697, + "learning_rate": 1.5999564775859567e-05, + "loss": 0.0257, + "step": 5615 + }, + { + "epoch": 0.608846487424111, + "grad_norm": 3.559357166290283, + "learning_rate": 1.5998839402292183e-05, + "loss": 0.0444, + "step": 5616 + }, + { + "epoch": 0.6089549002601908, + "grad_norm": 7.1681036949157715, + "learning_rate": 1.5998114028724795e-05, + "loss": 0.0294, + "step": 5617 + }, + { + "epoch": 0.6090633130962706, + "grad_norm": 0.8474098443984985, + "learning_rate": 1.5997388655157407e-05, + "loss": 0.0083, + "step": 5618 + }, + { + "epoch": 0.6091717259323504, + "grad_norm": 3.0288801193237305, + "learning_rate": 1.599666328159002e-05, + "loss": 0.0146, + "step": 5619 + }, + { + "epoch": 0.6092801387684302, + "grad_norm": 5.024599552154541, + "learning_rate": 1.5995937908022635e-05, + "loss": 0.0488, + "step": 5620 + }, + { + "epoch": 0.60938855160451, + "grad_norm": 1.0166808366775513, + "learning_rate": 1.5995212534455247e-05, + "loss": 0.0086, + "step": 5621 + }, + { + "epoch": 0.6094969644405898, + "grad_norm": 8.754921913146973, + "learning_rate": 1.599448716088786e-05, + "loss": 0.0181, + "step": 5622 + }, + { + "epoch": 0.6096053772766695, + "grad_norm": 2.7122554779052734, + "learning_rate": 1.599376178732047e-05, + "loss": 0.0314, + "step": 5623 + }, + { + "epoch": 0.6097137901127494, + "grad_norm": 10.923108100891113, + "learning_rate": 1.5993036413753083e-05, + "loss": 0.0758, + "step": 5624 + }, + { + "epoch": 0.6098222029488292, + "grad_norm": 8.180861473083496, + "learning_rate": 1.59923110401857e-05, + "loss": 0.0722, + "step": 5625 + }, + { + "epoch": 0.6099306157849089, + "grad_norm": 5.134505748748779, + "learning_rate": 1.599158566661831e-05, + "loss": 0.0939, + "step": 5626 + }, + { + "epoch": 0.6100390286209887, + "grad_norm": 7.1586012840271, + "learning_rate": 1.5990860293050923e-05, + "loss": 0.0496, + "step": 5627 + }, + { + "epoch": 0.6101474414570686, + "grad_norm": 4.744986534118652, + "learning_rate": 1.5990134919483535e-05, + "loss": 0.0514, + "step": 5628 + }, + { + "epoch": 0.6102558542931483, + "grad_norm": 2.588390350341797, + "learning_rate": 1.5989409545916147e-05, + "loss": 0.0241, + "step": 5629 + }, + { + "epoch": 0.6103642671292281, + "grad_norm": 3.383916139602661, + "learning_rate": 1.598868417234876e-05, + "loss": 0.0616, + "step": 5630 + }, + { + "epoch": 0.6104726799653079, + "grad_norm": 5.127309799194336, + "learning_rate": 1.5987958798781372e-05, + "loss": 0.0197, + "step": 5631 + }, + { + "epoch": 0.6105810928013877, + "grad_norm": 2.8846449851989746, + "learning_rate": 1.5987233425213987e-05, + "loss": 0.0346, + "step": 5632 + }, + { + "epoch": 0.6106895056374675, + "grad_norm": 4.983236312866211, + "learning_rate": 1.59865080516466e-05, + "loss": 0.0977, + "step": 5633 + }, + { + "epoch": 0.6107979184735473, + "grad_norm": 9.091633796691895, + "learning_rate": 1.598578267807921e-05, + "loss": 0.0758, + "step": 5634 + }, + { + "epoch": 0.610906331309627, + "grad_norm": 7.223941802978516, + "learning_rate": 1.5985057304511824e-05, + "loss": 0.0773, + "step": 5635 + }, + { + "epoch": 0.6110147441457069, + "grad_norm": 4.281701564788818, + "learning_rate": 1.5984331930944436e-05, + "loss": 0.0799, + "step": 5636 + }, + { + "epoch": 0.6111231569817867, + "grad_norm": 5.932700157165527, + "learning_rate": 1.598360655737705e-05, + "loss": 0.0399, + "step": 5637 + }, + { + "epoch": 0.6112315698178664, + "grad_norm": 2.2812612056732178, + "learning_rate": 1.5982881183809663e-05, + "loss": 0.032, + "step": 5638 + }, + { + "epoch": 0.6113399826539462, + "grad_norm": 0.7409887313842773, + "learning_rate": 1.5982155810242276e-05, + "loss": 0.0196, + "step": 5639 + }, + { + "epoch": 0.611448395490026, + "grad_norm": 2.354330062866211, + "learning_rate": 1.598143043667489e-05, + "loss": 0.0118, + "step": 5640 + }, + { + "epoch": 0.6115568083261058, + "grad_norm": 2.489583969116211, + "learning_rate": 1.5980705063107503e-05, + "loss": 0.0302, + "step": 5641 + }, + { + "epoch": 0.6116652211621856, + "grad_norm": 3.5780155658721924, + "learning_rate": 1.5979979689540115e-05, + "loss": 0.0316, + "step": 5642 + }, + { + "epoch": 0.6117736339982653, + "grad_norm": 2.431504487991333, + "learning_rate": 1.5979254315972728e-05, + "loss": 0.0256, + "step": 5643 + }, + { + "epoch": 0.6118820468343452, + "grad_norm": 2.077819347381592, + "learning_rate": 1.597852894240534e-05, + "loss": 0.0308, + "step": 5644 + }, + { + "epoch": 0.611990459670425, + "grad_norm": 2.0055713653564453, + "learning_rate": 1.5977803568837952e-05, + "loss": 0.0286, + "step": 5645 + }, + { + "epoch": 0.6120988725065047, + "grad_norm": 3.1463193893432617, + "learning_rate": 1.5977078195270564e-05, + "loss": 0.0486, + "step": 5646 + }, + { + "epoch": 0.6122072853425846, + "grad_norm": 1.5401939153671265, + "learning_rate": 1.597635282170318e-05, + "loss": 0.0237, + "step": 5647 + }, + { + "epoch": 0.6123156981786644, + "grad_norm": 2.8308041095733643, + "learning_rate": 1.597562744813579e-05, + "loss": 0.0242, + "step": 5648 + }, + { + "epoch": 0.6124241110147441, + "grad_norm": 2.1836111545562744, + "learning_rate": 1.5974902074568404e-05, + "loss": 0.0246, + "step": 5649 + }, + { + "epoch": 0.6125325238508239, + "grad_norm": 2.7547242641448975, + "learning_rate": 1.5974176701001016e-05, + "loss": 0.0291, + "step": 5650 + }, + { + "epoch": 0.6126409366869038, + "grad_norm": 2.7419164180755615, + "learning_rate": 1.5973451327433628e-05, + "loss": 0.0362, + "step": 5651 + }, + { + "epoch": 0.6127493495229835, + "grad_norm": 3.9885151386260986, + "learning_rate": 1.597272595386624e-05, + "loss": 0.031, + "step": 5652 + }, + { + "epoch": 0.6128577623590633, + "grad_norm": 3.326470375061035, + "learning_rate": 1.5972000580298856e-05, + "loss": 0.032, + "step": 5653 + }, + { + "epoch": 0.6129661751951431, + "grad_norm": 1.6433684825897217, + "learning_rate": 1.5971275206731468e-05, + "loss": 0.045, + "step": 5654 + }, + { + "epoch": 0.613074588031223, + "grad_norm": 2.592639684677124, + "learning_rate": 1.5970549833164083e-05, + "loss": 0.0447, + "step": 5655 + }, + { + "epoch": 0.6131830008673027, + "grad_norm": 3.7349276542663574, + "learning_rate": 1.5969824459596696e-05, + "loss": 0.0497, + "step": 5656 + }, + { + "epoch": 0.6132914137033825, + "grad_norm": 2.306278944015503, + "learning_rate": 1.5969099086029308e-05, + "loss": 0.0156, + "step": 5657 + }, + { + "epoch": 0.6133998265394622, + "grad_norm": 3.738063097000122, + "learning_rate": 1.596837371246192e-05, + "loss": 0.0305, + "step": 5658 + }, + { + "epoch": 0.6135082393755421, + "grad_norm": 3.6415836811065674, + "learning_rate": 1.5967648338894532e-05, + "loss": 0.0416, + "step": 5659 + }, + { + "epoch": 0.6136166522116219, + "grad_norm": 2.0080714225769043, + "learning_rate": 1.5966922965327144e-05, + "loss": 0.0224, + "step": 5660 + }, + { + "epoch": 0.6137250650477016, + "grad_norm": 2.1092474460601807, + "learning_rate": 1.5966197591759756e-05, + "loss": 0.0432, + "step": 5661 + }, + { + "epoch": 0.6138334778837814, + "grad_norm": 4.475510597229004, + "learning_rate": 1.5965472218192372e-05, + "loss": 0.0553, + "step": 5662 + }, + { + "epoch": 0.6139418907198613, + "grad_norm": 2.6626875400543213, + "learning_rate": 1.5964746844624984e-05, + "loss": 0.058, + "step": 5663 + }, + { + "epoch": 0.614050303555941, + "grad_norm": 3.23738431930542, + "learning_rate": 1.5964021471057596e-05, + "loss": 0.0508, + "step": 5664 + }, + { + "epoch": 0.6141587163920208, + "grad_norm": 1.6686177253723145, + "learning_rate": 1.5963296097490208e-05, + "loss": 0.019, + "step": 5665 + }, + { + "epoch": 0.6142671292281006, + "grad_norm": 3.1179699897766113, + "learning_rate": 1.596257072392282e-05, + "loss": 0.022, + "step": 5666 + }, + { + "epoch": 0.6143755420641804, + "grad_norm": 1.2797859907150269, + "learning_rate": 1.5961845350355433e-05, + "loss": 0.0314, + "step": 5667 + }, + { + "epoch": 0.6144839549002602, + "grad_norm": 1.8133658170700073, + "learning_rate": 1.5961119976788048e-05, + "loss": 0.0281, + "step": 5668 + }, + { + "epoch": 0.61459236773634, + "grad_norm": 2.1993730068206787, + "learning_rate": 1.596039460322066e-05, + "loss": 0.0425, + "step": 5669 + }, + { + "epoch": 0.6147007805724197, + "grad_norm": 3.4146852493286133, + "learning_rate": 1.5959669229653272e-05, + "loss": 0.0523, + "step": 5670 + }, + { + "epoch": 0.6148091934084996, + "grad_norm": 1.0634955167770386, + "learning_rate": 1.5958943856085885e-05, + "loss": 0.01, + "step": 5671 + }, + { + "epoch": 0.6149176062445794, + "grad_norm": 2.433866024017334, + "learning_rate": 1.59582184825185e-05, + "loss": 0.04, + "step": 5672 + }, + { + "epoch": 0.6150260190806591, + "grad_norm": 2.3761980533599854, + "learning_rate": 1.5957493108951112e-05, + "loss": 0.0217, + "step": 5673 + }, + { + "epoch": 0.6151344319167389, + "grad_norm": 3.6268064975738525, + "learning_rate": 1.5956767735383724e-05, + "loss": 0.0482, + "step": 5674 + }, + { + "epoch": 0.6152428447528188, + "grad_norm": 2.997424602508545, + "learning_rate": 1.5956042361816337e-05, + "loss": 0.0591, + "step": 5675 + }, + { + "epoch": 0.6153512575888985, + "grad_norm": 3.0527312755584717, + "learning_rate": 1.595531698824895e-05, + "loss": 0.0198, + "step": 5676 + }, + { + "epoch": 0.6154596704249783, + "grad_norm": 2.648484230041504, + "learning_rate": 1.5954591614681564e-05, + "loss": 0.0447, + "step": 5677 + }, + { + "epoch": 0.6155680832610582, + "grad_norm": 2.683408498764038, + "learning_rate": 1.5953866241114176e-05, + "loss": 0.0666, + "step": 5678 + }, + { + "epoch": 0.6156764960971379, + "grad_norm": 3.560833215713501, + "learning_rate": 1.595314086754679e-05, + "loss": 0.0489, + "step": 5679 + }, + { + "epoch": 0.6157849089332177, + "grad_norm": 4.0490522384643555, + "learning_rate": 1.59524154939794e-05, + "loss": 0.0433, + "step": 5680 + }, + { + "epoch": 0.6158933217692975, + "grad_norm": 1.1773563623428345, + "learning_rate": 1.5951690120412013e-05, + "loss": 0.0115, + "step": 5681 + }, + { + "epoch": 0.6160017346053773, + "grad_norm": 1.4403700828552246, + "learning_rate": 1.5950964746844625e-05, + "loss": 0.0209, + "step": 5682 + }, + { + "epoch": 0.6161101474414571, + "grad_norm": 1.2584564685821533, + "learning_rate": 1.5950239373277237e-05, + "loss": 0.0081, + "step": 5683 + }, + { + "epoch": 0.6162185602775369, + "grad_norm": 3.092336416244507, + "learning_rate": 1.5949513999709853e-05, + "loss": 0.0481, + "step": 5684 + }, + { + "epoch": 0.6163269731136166, + "grad_norm": 1.539516568183899, + "learning_rate": 1.5948788626142465e-05, + "loss": 0.0111, + "step": 5685 + }, + { + "epoch": 0.6164353859496965, + "grad_norm": 1.8672658205032349, + "learning_rate": 1.5948063252575077e-05, + "loss": 0.0192, + "step": 5686 + }, + { + "epoch": 0.6165437987857763, + "grad_norm": 4.050486087799072, + "learning_rate": 1.594733787900769e-05, + "loss": 0.058, + "step": 5687 + }, + { + "epoch": 0.616652211621856, + "grad_norm": 3.8122446537017822, + "learning_rate": 1.59466125054403e-05, + "loss": 0.061, + "step": 5688 + }, + { + "epoch": 0.6167606244579358, + "grad_norm": 1.4287903308868408, + "learning_rate": 1.5945887131872917e-05, + "loss": 0.0153, + "step": 5689 + }, + { + "epoch": 0.6168690372940157, + "grad_norm": 2.1055760383605957, + "learning_rate": 1.594516175830553e-05, + "loss": 0.0175, + "step": 5690 + }, + { + "epoch": 0.6169774501300954, + "grad_norm": 1.454035758972168, + "learning_rate": 1.594443638473814e-05, + "loss": 0.0228, + "step": 5691 + }, + { + "epoch": 0.6170858629661752, + "grad_norm": 2.061886787414551, + "learning_rate": 1.5943711011170756e-05, + "loss": 0.0122, + "step": 5692 + }, + { + "epoch": 0.6171942758022549, + "grad_norm": 3.4886696338653564, + "learning_rate": 1.594298563760337e-05, + "loss": 0.032, + "step": 5693 + }, + { + "epoch": 0.6173026886383348, + "grad_norm": 4.328814506530762, + "learning_rate": 1.594226026403598e-05, + "loss": 0.078, + "step": 5694 + }, + { + "epoch": 0.6174111014744146, + "grad_norm": 1.288404107093811, + "learning_rate": 1.5941534890468593e-05, + "loss": 0.0254, + "step": 5695 + }, + { + "epoch": 0.6175195143104943, + "grad_norm": 0.5263507962226868, + "learning_rate": 1.5940809516901205e-05, + "loss": 0.0124, + "step": 5696 + }, + { + "epoch": 0.6176279271465741, + "grad_norm": 6.17104959487915, + "learning_rate": 1.5940084143333817e-05, + "loss": 0.0679, + "step": 5697 + }, + { + "epoch": 0.617736339982654, + "grad_norm": 3.6659860610961914, + "learning_rate": 1.593935876976643e-05, + "loss": 0.0248, + "step": 5698 + }, + { + "epoch": 0.6178447528187337, + "grad_norm": 3.076314687728882, + "learning_rate": 1.5938633396199045e-05, + "loss": 0.077, + "step": 5699 + }, + { + "epoch": 0.6179531656548135, + "grad_norm": 2.727647542953491, + "learning_rate": 1.5937908022631657e-05, + "loss": 0.0365, + "step": 5700 + }, + { + "epoch": 0.6180615784908933, + "grad_norm": 2.8620948791503906, + "learning_rate": 1.593718264906427e-05, + "loss": 0.044, + "step": 5701 + }, + { + "epoch": 0.6181699913269731, + "grad_norm": 3.0510241985321045, + "learning_rate": 1.593645727549688e-05, + "loss": 0.0609, + "step": 5702 + }, + { + "epoch": 0.6182784041630529, + "grad_norm": 5.084969997406006, + "learning_rate": 1.5935731901929493e-05, + "loss": 0.0358, + "step": 5703 + }, + { + "epoch": 0.6183868169991327, + "grad_norm": 5.333993911743164, + "learning_rate": 1.5935006528362106e-05, + "loss": 0.0532, + "step": 5704 + }, + { + "epoch": 0.6184952298352124, + "grad_norm": 4.383157253265381, + "learning_rate": 1.593428115479472e-05, + "loss": 0.0709, + "step": 5705 + }, + { + "epoch": 0.6186036426712923, + "grad_norm": 3.4271316528320312, + "learning_rate": 1.5933555781227333e-05, + "loss": 0.0454, + "step": 5706 + }, + { + "epoch": 0.6187120555073721, + "grad_norm": 2.5688529014587402, + "learning_rate": 1.593283040765995e-05, + "loss": 0.0416, + "step": 5707 + }, + { + "epoch": 0.6188204683434518, + "grad_norm": 2.966243028640747, + "learning_rate": 1.593210503409256e-05, + "loss": 0.0766, + "step": 5708 + }, + { + "epoch": 0.6189288811795317, + "grad_norm": 1.024421215057373, + "learning_rate": 1.5931379660525173e-05, + "loss": 0.0154, + "step": 5709 + }, + { + "epoch": 0.6190372940156115, + "grad_norm": 4.1446099281311035, + "learning_rate": 1.5930654286957785e-05, + "loss": 0.0436, + "step": 5710 + }, + { + "epoch": 0.6191457068516912, + "grad_norm": 2.503577470779419, + "learning_rate": 1.5929928913390397e-05, + "loss": 0.034, + "step": 5711 + }, + { + "epoch": 0.619254119687771, + "grad_norm": 4.223203659057617, + "learning_rate": 1.592920353982301e-05, + "loss": 0.1262, + "step": 5712 + }, + { + "epoch": 0.6193625325238509, + "grad_norm": 6.111632823944092, + "learning_rate": 1.592847816625562e-05, + "loss": 0.0409, + "step": 5713 + }, + { + "epoch": 0.6194709453599306, + "grad_norm": 0.9352089762687683, + "learning_rate": 1.5927752792688237e-05, + "loss": 0.0227, + "step": 5714 + }, + { + "epoch": 0.6195793581960104, + "grad_norm": 3.1631674766540527, + "learning_rate": 1.592702741912085e-05, + "loss": 0.0497, + "step": 5715 + }, + { + "epoch": 0.6196877710320902, + "grad_norm": 2.313464879989624, + "learning_rate": 1.592630204555346e-05, + "loss": 0.0296, + "step": 5716 + }, + { + "epoch": 0.61979618386817, + "grad_norm": 1.4713339805603027, + "learning_rate": 1.5925576671986074e-05, + "loss": 0.0066, + "step": 5717 + }, + { + "epoch": 0.6199045967042498, + "grad_norm": 5.162541389465332, + "learning_rate": 1.5924851298418686e-05, + "loss": 0.084, + "step": 5718 + }, + { + "epoch": 0.6200130095403296, + "grad_norm": 1.4646458625793457, + "learning_rate": 1.5924125924851298e-05, + "loss": 0.0312, + "step": 5719 + }, + { + "epoch": 0.6201214223764093, + "grad_norm": 1.9309842586517334, + "learning_rate": 1.5923400551283913e-05, + "loss": 0.031, + "step": 5720 + }, + { + "epoch": 0.6202298352124892, + "grad_norm": 0.7016682028770447, + "learning_rate": 1.5922675177716526e-05, + "loss": 0.0069, + "step": 5721 + }, + { + "epoch": 0.620338248048569, + "grad_norm": 0.5337362289428711, + "learning_rate": 1.5921949804149138e-05, + "loss": 0.0051, + "step": 5722 + }, + { + "epoch": 0.6204466608846487, + "grad_norm": 7.10960578918457, + "learning_rate": 1.592122443058175e-05, + "loss": 0.1138, + "step": 5723 + }, + { + "epoch": 0.6205550737207285, + "grad_norm": 4.216928958892822, + "learning_rate": 1.5920499057014365e-05, + "loss": 0.0291, + "step": 5724 + }, + { + "epoch": 0.6206634865568084, + "grad_norm": 6.44874382019043, + "learning_rate": 1.5919773683446978e-05, + "loss": 0.0589, + "step": 5725 + }, + { + "epoch": 0.6207718993928881, + "grad_norm": 2.4875073432922363, + "learning_rate": 1.591904830987959e-05, + "loss": 0.0297, + "step": 5726 + }, + { + "epoch": 0.6208803122289679, + "grad_norm": 1.7672141790390015, + "learning_rate": 1.5918322936312202e-05, + "loss": 0.0343, + "step": 5727 + }, + { + "epoch": 0.6209887250650477, + "grad_norm": 1.5587648153305054, + "learning_rate": 1.5917597562744814e-05, + "loss": 0.0485, + "step": 5728 + }, + { + "epoch": 0.6210971379011275, + "grad_norm": 3.9360899925231934, + "learning_rate": 1.591687218917743e-05, + "loss": 0.0486, + "step": 5729 + }, + { + "epoch": 0.6212055507372073, + "grad_norm": 5.4682297706604, + "learning_rate": 1.591614681561004e-05, + "loss": 0.0294, + "step": 5730 + }, + { + "epoch": 0.621313963573287, + "grad_norm": 2.5889718532562256, + "learning_rate": 1.5915421442042654e-05, + "loss": 0.0475, + "step": 5731 + }, + { + "epoch": 0.6214223764093668, + "grad_norm": 4.461302757263184, + "learning_rate": 1.5914696068475266e-05, + "loss": 0.0925, + "step": 5732 + }, + { + "epoch": 0.6215307892454467, + "grad_norm": 1.9094436168670654, + "learning_rate": 1.5913970694907878e-05, + "loss": 0.0596, + "step": 5733 + }, + { + "epoch": 0.6216392020815265, + "grad_norm": 0.7811403870582581, + "learning_rate": 1.591324532134049e-05, + "loss": 0.012, + "step": 5734 + }, + { + "epoch": 0.6217476149176062, + "grad_norm": 1.22239089012146, + "learning_rate": 1.5912519947773102e-05, + "loss": 0.0291, + "step": 5735 + }, + { + "epoch": 0.6218560277536861, + "grad_norm": 3.10666823387146, + "learning_rate": 1.5911794574205718e-05, + "loss": 0.1, + "step": 5736 + }, + { + "epoch": 0.6219644405897659, + "grad_norm": 1.549207091331482, + "learning_rate": 1.591106920063833e-05, + "loss": 0.0384, + "step": 5737 + }, + { + "epoch": 0.6220728534258456, + "grad_norm": 0.96235591173172, + "learning_rate": 1.5910343827070942e-05, + "loss": 0.0212, + "step": 5738 + }, + { + "epoch": 0.6221812662619254, + "grad_norm": 1.6279913187026978, + "learning_rate": 1.5909618453503554e-05, + "loss": 0.0326, + "step": 5739 + }, + { + "epoch": 0.6222896790980053, + "grad_norm": 0.5788346529006958, + "learning_rate": 1.5908893079936166e-05, + "loss": 0.0086, + "step": 5740 + }, + { + "epoch": 0.622398091934085, + "grad_norm": 1.932071566581726, + "learning_rate": 1.5908167706368782e-05, + "loss": 0.0244, + "step": 5741 + }, + { + "epoch": 0.6225065047701648, + "grad_norm": 1.7956942319869995, + "learning_rate": 1.5907442332801394e-05, + "loss": 0.0381, + "step": 5742 + }, + { + "epoch": 0.6226149176062445, + "grad_norm": 3.4651756286621094, + "learning_rate": 1.5906716959234006e-05, + "loss": 0.0686, + "step": 5743 + }, + { + "epoch": 0.6227233304423244, + "grad_norm": 2.2130253314971924, + "learning_rate": 1.5905991585666622e-05, + "loss": 0.0591, + "step": 5744 + }, + { + "epoch": 0.6228317432784042, + "grad_norm": 1.24077570438385, + "learning_rate": 1.5905266212099234e-05, + "loss": 0.0296, + "step": 5745 + }, + { + "epoch": 0.622940156114484, + "grad_norm": 11.056808471679688, + "learning_rate": 1.5904540838531846e-05, + "loss": 0.0427, + "step": 5746 + }, + { + "epoch": 0.6230485689505637, + "grad_norm": 1.5498580932617188, + "learning_rate": 1.5903815464964458e-05, + "loss": 0.0331, + "step": 5747 + }, + { + "epoch": 0.6231569817866436, + "grad_norm": 0.7282893657684326, + "learning_rate": 1.590309009139707e-05, + "loss": 0.0151, + "step": 5748 + }, + { + "epoch": 0.6232653946227233, + "grad_norm": 3.883387327194214, + "learning_rate": 1.5902364717829683e-05, + "loss": 0.0387, + "step": 5749 + }, + { + "epoch": 0.6233738074588031, + "grad_norm": 1.762410044670105, + "learning_rate": 1.5901639344262295e-05, + "loss": 0.0278, + "step": 5750 + }, + { + "epoch": 0.6234822202948829, + "grad_norm": 1.7229281663894653, + "learning_rate": 1.590091397069491e-05, + "loss": 0.034, + "step": 5751 + }, + { + "epoch": 0.6235906331309627, + "grad_norm": 2.832817554473877, + "learning_rate": 1.5900188597127522e-05, + "loss": 0.0192, + "step": 5752 + }, + { + "epoch": 0.6236990459670425, + "grad_norm": 1.6606274843215942, + "learning_rate": 1.5899463223560135e-05, + "loss": 0.017, + "step": 5753 + }, + { + "epoch": 0.6238074588031223, + "grad_norm": 3.459502696990967, + "learning_rate": 1.5898737849992747e-05, + "loss": 0.0608, + "step": 5754 + }, + { + "epoch": 0.623915871639202, + "grad_norm": 1.8488917350769043, + "learning_rate": 1.589801247642536e-05, + "loss": 0.0159, + "step": 5755 + }, + { + "epoch": 0.6240242844752819, + "grad_norm": 0.9637010097503662, + "learning_rate": 1.589728710285797e-05, + "loss": 0.0123, + "step": 5756 + }, + { + "epoch": 0.6241326973113617, + "grad_norm": 3.7802300453186035, + "learning_rate": 1.5896561729290586e-05, + "loss": 0.0804, + "step": 5757 + }, + { + "epoch": 0.6242411101474414, + "grad_norm": 8.036775588989258, + "learning_rate": 1.58958363557232e-05, + "loss": 0.0366, + "step": 5758 + }, + { + "epoch": 0.6243495229835212, + "grad_norm": 0.9356223940849304, + "learning_rate": 1.5895110982155814e-05, + "loss": 0.0134, + "step": 5759 + }, + { + "epoch": 0.6244579358196011, + "grad_norm": 1.5995466709136963, + "learning_rate": 1.5894385608588426e-05, + "loss": 0.0182, + "step": 5760 + }, + { + "epoch": 0.6245663486556808, + "grad_norm": 2.52344012260437, + "learning_rate": 1.589366023502104e-05, + "loss": 0.0172, + "step": 5761 + }, + { + "epoch": 0.6246747614917606, + "grad_norm": 8.097987174987793, + "learning_rate": 1.589293486145365e-05, + "loss": 0.0669, + "step": 5762 + }, + { + "epoch": 0.6247831743278404, + "grad_norm": 5.528139114379883, + "learning_rate": 1.5892209487886263e-05, + "loss": 0.0242, + "step": 5763 + }, + { + "epoch": 0.6248915871639202, + "grad_norm": 3.2933223247528076, + "learning_rate": 1.5891484114318875e-05, + "loss": 0.0483, + "step": 5764 + }, + { + "epoch": 0.625, + "grad_norm": 2.3048994541168213, + "learning_rate": 1.5890758740751487e-05, + "loss": 0.0306, + "step": 5765 + }, + { + "epoch": 0.6251084128360798, + "grad_norm": 5.890291690826416, + "learning_rate": 1.5890033367184103e-05, + "loss": 0.0539, + "step": 5766 + }, + { + "epoch": 0.6252168256721596, + "grad_norm": 4.750545501708984, + "learning_rate": 1.5889307993616715e-05, + "loss": 0.0526, + "step": 5767 + }, + { + "epoch": 0.6253252385082394, + "grad_norm": 3.462501049041748, + "learning_rate": 1.5888582620049327e-05, + "loss": 0.0381, + "step": 5768 + }, + { + "epoch": 0.6254336513443192, + "grad_norm": 4.0876784324646, + "learning_rate": 1.588785724648194e-05, + "loss": 0.0882, + "step": 5769 + }, + { + "epoch": 0.6255420641803989, + "grad_norm": 4.081306457519531, + "learning_rate": 1.588713187291455e-05, + "loss": 0.0125, + "step": 5770 + }, + { + "epoch": 0.6256504770164788, + "grad_norm": 10.811928749084473, + "learning_rate": 1.5886406499347163e-05, + "loss": 0.0212, + "step": 5771 + }, + { + "epoch": 0.6257588898525586, + "grad_norm": 2.502551555633545, + "learning_rate": 1.588568112577978e-05, + "loss": 0.0429, + "step": 5772 + }, + { + "epoch": 0.6258673026886383, + "grad_norm": 4.290502071380615, + "learning_rate": 1.588495575221239e-05, + "loss": 0.0392, + "step": 5773 + }, + { + "epoch": 0.6259757155247181, + "grad_norm": 2.2252800464630127, + "learning_rate": 1.5884230378645003e-05, + "loss": 0.0354, + "step": 5774 + }, + { + "epoch": 0.626084128360798, + "grad_norm": 9.090497016906738, + "learning_rate": 1.5883505005077615e-05, + "loss": 0.0728, + "step": 5775 + }, + { + "epoch": 0.6261925411968777, + "grad_norm": 3.2901554107666016, + "learning_rate": 1.588277963151023e-05, + "loss": 0.0928, + "step": 5776 + }, + { + "epoch": 0.6263009540329575, + "grad_norm": 6.351983070373535, + "learning_rate": 1.5882054257942843e-05, + "loss": 0.1172, + "step": 5777 + }, + { + "epoch": 0.6264093668690373, + "grad_norm": 2.108264923095703, + "learning_rate": 1.5881328884375455e-05, + "loss": 0.0084, + "step": 5778 + }, + { + "epoch": 0.6265177797051171, + "grad_norm": 3.3596413135528564, + "learning_rate": 1.5880603510808067e-05, + "loss": 0.0567, + "step": 5779 + }, + { + "epoch": 0.6266261925411969, + "grad_norm": 0.6609101295471191, + "learning_rate": 1.587987813724068e-05, + "loss": 0.0063, + "step": 5780 + }, + { + "epoch": 0.6267346053772767, + "grad_norm": 2.500351905822754, + "learning_rate": 1.5879152763673295e-05, + "loss": 0.0075, + "step": 5781 + }, + { + "epoch": 0.6268430182133564, + "grad_norm": 4.408476829528809, + "learning_rate": 1.5878427390105907e-05, + "loss": 0.0919, + "step": 5782 + }, + { + "epoch": 0.6269514310494363, + "grad_norm": 2.67195987701416, + "learning_rate": 1.587770201653852e-05, + "loss": 0.0404, + "step": 5783 + }, + { + "epoch": 0.627059843885516, + "grad_norm": 0.5921860933303833, + "learning_rate": 1.587697664297113e-05, + "loss": 0.0059, + "step": 5784 + }, + { + "epoch": 0.6271682567215958, + "grad_norm": 3.7680609226226807, + "learning_rate": 1.5876251269403743e-05, + "loss": 0.0256, + "step": 5785 + }, + { + "epoch": 0.6272766695576756, + "grad_norm": 2.7482106685638428, + "learning_rate": 1.5875525895836356e-05, + "loss": 0.023, + "step": 5786 + }, + { + "epoch": 0.6273850823937555, + "grad_norm": 5.454686641693115, + "learning_rate": 1.5874800522268968e-05, + "loss": 0.0763, + "step": 5787 + }, + { + "epoch": 0.6274934952298352, + "grad_norm": 3.32127046585083, + "learning_rate": 1.5874075148701583e-05, + "loss": 0.0514, + "step": 5788 + }, + { + "epoch": 0.627601908065915, + "grad_norm": 2.429912567138672, + "learning_rate": 1.5873349775134195e-05, + "loss": 0.0181, + "step": 5789 + }, + { + "epoch": 0.6277103209019947, + "grad_norm": 4.8106160163879395, + "learning_rate": 1.5872624401566808e-05, + "loss": 0.0406, + "step": 5790 + }, + { + "epoch": 0.6278187337380746, + "grad_norm": 1.634602665901184, + "learning_rate": 1.587189902799942e-05, + "loss": 0.0468, + "step": 5791 + }, + { + "epoch": 0.6279271465741544, + "grad_norm": 1.1746282577514648, + "learning_rate": 1.5871173654432032e-05, + "loss": 0.0176, + "step": 5792 + }, + { + "epoch": 0.6280355594102341, + "grad_norm": 2.347416877746582, + "learning_rate": 1.5870448280864647e-05, + "loss": 0.032, + "step": 5793 + }, + { + "epoch": 0.6281439722463139, + "grad_norm": 3.1871583461761475, + "learning_rate": 1.586972290729726e-05, + "loss": 0.0274, + "step": 5794 + }, + { + "epoch": 0.6282523850823938, + "grad_norm": 2.3297476768493652, + "learning_rate": 1.586899753372987e-05, + "loss": 0.0351, + "step": 5795 + }, + { + "epoch": 0.6283607979184735, + "grad_norm": 1.106356143951416, + "learning_rate": 1.5868272160162487e-05, + "loss": 0.0112, + "step": 5796 + }, + { + "epoch": 0.6284692107545533, + "grad_norm": 0.9606380462646484, + "learning_rate": 1.58675467865951e-05, + "loss": 0.0108, + "step": 5797 + }, + { + "epoch": 0.6285776235906332, + "grad_norm": 4.897000789642334, + "learning_rate": 1.586682141302771e-05, + "loss": 0.0341, + "step": 5798 + }, + { + "epoch": 0.628686036426713, + "grad_norm": 1.9177788496017456, + "learning_rate": 1.5866096039460324e-05, + "loss": 0.0576, + "step": 5799 + }, + { + "epoch": 0.6287944492627927, + "grad_norm": 1.6183440685272217, + "learning_rate": 1.5865370665892936e-05, + "loss": 0.0208, + "step": 5800 + }, + { + "epoch": 0.6289028620988725, + "grad_norm": 5.164540767669678, + "learning_rate": 1.5864645292325548e-05, + "loss": 0.0791, + "step": 5801 + }, + { + "epoch": 0.6290112749349523, + "grad_norm": 3.591484546661377, + "learning_rate": 1.586391991875816e-05, + "loss": 0.0578, + "step": 5802 + }, + { + "epoch": 0.6291196877710321, + "grad_norm": 2.4207723140716553, + "learning_rate": 1.5863194545190776e-05, + "loss": 0.0582, + "step": 5803 + }, + { + "epoch": 0.6292281006071119, + "grad_norm": 1.5631037950515747, + "learning_rate": 1.5862469171623388e-05, + "loss": 0.0201, + "step": 5804 + }, + { + "epoch": 0.6293365134431916, + "grad_norm": 1.1739063262939453, + "learning_rate": 1.5861743798056e-05, + "loss": 0.0288, + "step": 5805 + }, + { + "epoch": 0.6294449262792715, + "grad_norm": 1.0041959285736084, + "learning_rate": 1.5861018424488612e-05, + "loss": 0.0265, + "step": 5806 + }, + { + "epoch": 0.6295533391153513, + "grad_norm": 2.450319528579712, + "learning_rate": 1.5860293050921224e-05, + "loss": 0.0575, + "step": 5807 + }, + { + "epoch": 0.629661751951431, + "grad_norm": 4.911057949066162, + "learning_rate": 1.5859567677353836e-05, + "loss": 0.073, + "step": 5808 + }, + { + "epoch": 0.6297701647875108, + "grad_norm": 5.394469261169434, + "learning_rate": 1.5858842303786452e-05, + "loss": 0.034, + "step": 5809 + }, + { + "epoch": 0.6298785776235907, + "grad_norm": 4.858827590942383, + "learning_rate": 1.5858116930219064e-05, + "loss": 0.0769, + "step": 5810 + }, + { + "epoch": 0.6299869904596704, + "grad_norm": 4.547558307647705, + "learning_rate": 1.585739155665168e-05, + "loss": 0.0374, + "step": 5811 + }, + { + "epoch": 0.6300954032957502, + "grad_norm": 5.22677755355835, + "learning_rate": 1.585666618308429e-05, + "loss": 0.1035, + "step": 5812 + }, + { + "epoch": 0.63020381613183, + "grad_norm": 2.6964967250823975, + "learning_rate": 1.5855940809516904e-05, + "loss": 0.0361, + "step": 5813 + }, + { + "epoch": 0.6303122289679098, + "grad_norm": 1.631641149520874, + "learning_rate": 1.5855215435949516e-05, + "loss": 0.0279, + "step": 5814 + }, + { + "epoch": 0.6304206418039896, + "grad_norm": 4.946029186248779, + "learning_rate": 1.5854490062382128e-05, + "loss": 0.0767, + "step": 5815 + }, + { + "epoch": 0.6305290546400694, + "grad_norm": 2.6606037616729736, + "learning_rate": 1.585376468881474e-05, + "loss": 0.0443, + "step": 5816 + }, + { + "epoch": 0.6306374674761491, + "grad_norm": 4.351911544799805, + "learning_rate": 1.5853039315247352e-05, + "loss": 0.1017, + "step": 5817 + }, + { + "epoch": 0.630745880312229, + "grad_norm": 3.595640182495117, + "learning_rate": 1.5852313941679968e-05, + "loss": 0.1057, + "step": 5818 + }, + { + "epoch": 0.6308542931483088, + "grad_norm": 1.133139729499817, + "learning_rate": 1.585158856811258e-05, + "loss": 0.0126, + "step": 5819 + }, + { + "epoch": 0.6309627059843885, + "grad_norm": 4.2340216636657715, + "learning_rate": 1.5850863194545192e-05, + "loss": 0.0496, + "step": 5820 + }, + { + "epoch": 0.6310711188204683, + "grad_norm": 2.068573474884033, + "learning_rate": 1.5850137820977804e-05, + "loss": 0.0383, + "step": 5821 + }, + { + "epoch": 0.6311795316565482, + "grad_norm": 2.2171361446380615, + "learning_rate": 1.5849412447410416e-05, + "loss": 0.0196, + "step": 5822 + }, + { + "epoch": 0.6312879444926279, + "grad_norm": 2.961148500442505, + "learning_rate": 1.584868707384303e-05, + "loss": 0.0798, + "step": 5823 + }, + { + "epoch": 0.6313963573287077, + "grad_norm": 1.1823581457138062, + "learning_rate": 1.5847961700275644e-05, + "loss": 0.0092, + "step": 5824 + }, + { + "epoch": 0.6315047701647876, + "grad_norm": 2.365006685256958, + "learning_rate": 1.5847236326708256e-05, + "loss": 0.0233, + "step": 5825 + }, + { + "epoch": 0.6316131830008673, + "grad_norm": 2.1101503372192383, + "learning_rate": 1.584651095314087e-05, + "loss": 0.0279, + "step": 5826 + }, + { + "epoch": 0.6317215958369471, + "grad_norm": 1.8537282943725586, + "learning_rate": 1.5845785579573484e-05, + "loss": 0.0395, + "step": 5827 + }, + { + "epoch": 0.6318300086730269, + "grad_norm": 1.2134774923324585, + "learning_rate": 1.5845060206006096e-05, + "loss": 0.0235, + "step": 5828 + }, + { + "epoch": 0.6319384215091067, + "grad_norm": 1.2431936264038086, + "learning_rate": 1.5844334832438708e-05, + "loss": 0.0205, + "step": 5829 + }, + { + "epoch": 0.6320468343451865, + "grad_norm": 9.342301368713379, + "learning_rate": 1.584360945887132e-05, + "loss": 0.0588, + "step": 5830 + }, + { + "epoch": 0.6321552471812663, + "grad_norm": 4.03784704208374, + "learning_rate": 1.5842884085303933e-05, + "loss": 0.0239, + "step": 5831 + }, + { + "epoch": 0.632263660017346, + "grad_norm": 0.7125971913337708, + "learning_rate": 1.5842158711736545e-05, + "loss": 0.0183, + "step": 5832 + }, + { + "epoch": 0.6323720728534259, + "grad_norm": 4.538700103759766, + "learning_rate": 1.584143333816916e-05, + "loss": 0.0561, + "step": 5833 + }, + { + "epoch": 0.6324804856895057, + "grad_norm": 0.2669282555580139, + "learning_rate": 1.5840707964601772e-05, + "loss": 0.0078, + "step": 5834 + }, + { + "epoch": 0.6325888985255854, + "grad_norm": 5.8653564453125, + "learning_rate": 1.5839982591034384e-05, + "loss": 0.0782, + "step": 5835 + }, + { + "epoch": 0.6326973113616652, + "grad_norm": 1.737546443939209, + "learning_rate": 1.5839257217466997e-05, + "loss": 0.0202, + "step": 5836 + }, + { + "epoch": 0.6328057241977451, + "grad_norm": 6.246695041656494, + "learning_rate": 1.583853184389961e-05, + "loss": 0.0153, + "step": 5837 + }, + { + "epoch": 0.6329141370338248, + "grad_norm": 1.196395993232727, + "learning_rate": 1.583780647033222e-05, + "loss": 0.0106, + "step": 5838 + }, + { + "epoch": 0.6330225498699046, + "grad_norm": 4.196409225463867, + "learning_rate": 1.5837081096764833e-05, + "loss": 0.0754, + "step": 5839 + }, + { + "epoch": 0.6331309627059843, + "grad_norm": 1.36492919921875, + "learning_rate": 1.583635572319745e-05, + "loss": 0.016, + "step": 5840 + }, + { + "epoch": 0.6332393755420642, + "grad_norm": 4.929582595825195, + "learning_rate": 1.583563034963006e-05, + "loss": 0.0302, + "step": 5841 + }, + { + "epoch": 0.633347788378144, + "grad_norm": 4.2604475021362305, + "learning_rate": 1.5834904976062673e-05, + "loss": 0.0465, + "step": 5842 + }, + { + "epoch": 0.6334562012142237, + "grad_norm": 3.342315435409546, + "learning_rate": 1.5834179602495285e-05, + "loss": 0.0475, + "step": 5843 + }, + { + "epoch": 0.6335646140503035, + "grad_norm": 3.4119763374328613, + "learning_rate": 1.5833454228927897e-05, + "loss": 0.0682, + "step": 5844 + }, + { + "epoch": 0.6336730268863834, + "grad_norm": 2.6978392601013184, + "learning_rate": 1.5832728855360513e-05, + "loss": 0.0569, + "step": 5845 + }, + { + "epoch": 0.6337814397224631, + "grad_norm": 3.1197571754455566, + "learning_rate": 1.5832003481793125e-05, + "loss": 0.0349, + "step": 5846 + }, + { + "epoch": 0.6338898525585429, + "grad_norm": 4.5220232009887695, + "learning_rate": 1.5831278108225737e-05, + "loss": 0.0675, + "step": 5847 + }, + { + "epoch": 0.6339982653946227, + "grad_norm": 4.4708685874938965, + "learning_rate": 1.5830552734658353e-05, + "loss": 0.0545, + "step": 5848 + }, + { + "epoch": 0.6341066782307025, + "grad_norm": 2.630438804626465, + "learning_rate": 1.5829827361090965e-05, + "loss": 0.041, + "step": 5849 + }, + { + "epoch": 0.6342150910667823, + "grad_norm": 6.4234185218811035, + "learning_rate": 1.5829101987523577e-05, + "loss": 0.0789, + "step": 5850 + }, + { + "epoch": 0.6343235039028621, + "grad_norm": 4.206418037414551, + "learning_rate": 1.582837661395619e-05, + "loss": 0.0644, + "step": 5851 + }, + { + "epoch": 0.6344319167389418, + "grad_norm": 4.351095199584961, + "learning_rate": 1.58276512403888e-05, + "loss": 0.0481, + "step": 5852 + }, + { + "epoch": 0.6345403295750217, + "grad_norm": 1.5502653121948242, + "learning_rate": 1.5826925866821413e-05, + "loss": 0.0239, + "step": 5853 + }, + { + "epoch": 0.6346487424111015, + "grad_norm": 2.4883241653442383, + "learning_rate": 1.5826200493254025e-05, + "loss": 0.0148, + "step": 5854 + }, + { + "epoch": 0.6347571552471812, + "grad_norm": 5.222198486328125, + "learning_rate": 1.582547511968664e-05, + "loss": 0.0198, + "step": 5855 + }, + { + "epoch": 0.6348655680832611, + "grad_norm": 1.234687328338623, + "learning_rate": 1.5824749746119253e-05, + "loss": 0.0317, + "step": 5856 + }, + { + "epoch": 0.6349739809193409, + "grad_norm": 0.9139000773429871, + "learning_rate": 1.5824024372551865e-05, + "loss": 0.0169, + "step": 5857 + }, + { + "epoch": 0.6350823937554206, + "grad_norm": 2.742765188217163, + "learning_rate": 1.5823298998984477e-05, + "loss": 0.0542, + "step": 5858 + }, + { + "epoch": 0.6351908065915004, + "grad_norm": 1.7154579162597656, + "learning_rate": 1.582257362541709e-05, + "loss": 0.0163, + "step": 5859 + }, + { + "epoch": 0.6352992194275803, + "grad_norm": 2.269481658935547, + "learning_rate": 1.58218482518497e-05, + "loss": 0.0569, + "step": 5860 + }, + { + "epoch": 0.63540763226366, + "grad_norm": 9.719342231750488, + "learning_rate": 1.5821122878282317e-05, + "loss": 0.0959, + "step": 5861 + }, + { + "epoch": 0.6355160450997398, + "grad_norm": 2.46616268157959, + "learning_rate": 1.582039750471493e-05, + "loss": 0.0243, + "step": 5862 + }, + { + "epoch": 0.6356244579358196, + "grad_norm": 2.381730556488037, + "learning_rate": 1.5819672131147545e-05, + "loss": 0.0297, + "step": 5863 + }, + { + "epoch": 0.6357328707718994, + "grad_norm": 4.948078632354736, + "learning_rate": 1.5818946757580157e-05, + "loss": 0.088, + "step": 5864 + }, + { + "epoch": 0.6358412836079792, + "grad_norm": 2.1242856979370117, + "learning_rate": 1.581822138401277e-05, + "loss": 0.0517, + "step": 5865 + }, + { + "epoch": 0.635949696444059, + "grad_norm": 2.5315074920654297, + "learning_rate": 1.581749601044538e-05, + "loss": 0.0291, + "step": 5866 + }, + { + "epoch": 0.6360581092801387, + "grad_norm": 1.6262251138687134, + "learning_rate": 1.5816770636877993e-05, + "loss": 0.0124, + "step": 5867 + }, + { + "epoch": 0.6361665221162186, + "grad_norm": 4.666186332702637, + "learning_rate": 1.5816045263310606e-05, + "loss": 0.0991, + "step": 5868 + }, + { + "epoch": 0.6362749349522984, + "grad_norm": 1.7301256656646729, + "learning_rate": 1.5815319889743218e-05, + "loss": 0.0354, + "step": 5869 + }, + { + "epoch": 0.6363833477883781, + "grad_norm": 2.2247049808502197, + "learning_rate": 1.5814594516175833e-05, + "loss": 0.0155, + "step": 5870 + }, + { + "epoch": 0.6364917606244579, + "grad_norm": 0.9827172160148621, + "learning_rate": 1.5813869142608445e-05, + "loss": 0.0112, + "step": 5871 + }, + { + "epoch": 0.6366001734605378, + "grad_norm": 4.863332271575928, + "learning_rate": 1.5813143769041057e-05, + "loss": 0.0219, + "step": 5872 + }, + { + "epoch": 0.6367085862966175, + "grad_norm": 3.23160457611084, + "learning_rate": 1.581241839547367e-05, + "loss": 0.0864, + "step": 5873 + }, + { + "epoch": 0.6368169991326973, + "grad_norm": 1.801859974861145, + "learning_rate": 1.5811693021906282e-05, + "loss": 0.0332, + "step": 5874 + }, + { + "epoch": 0.636925411968777, + "grad_norm": 3.5554208755493164, + "learning_rate": 1.5810967648338894e-05, + "loss": 0.0326, + "step": 5875 + }, + { + "epoch": 0.6370338248048569, + "grad_norm": 1.3324824571609497, + "learning_rate": 1.581024227477151e-05, + "loss": 0.0234, + "step": 5876 + }, + { + "epoch": 0.6371422376409367, + "grad_norm": 1.8908493518829346, + "learning_rate": 1.580951690120412e-05, + "loss": 0.0337, + "step": 5877 + }, + { + "epoch": 0.6372506504770165, + "grad_norm": 2.0551698207855225, + "learning_rate": 1.5808791527636734e-05, + "loss": 0.0494, + "step": 5878 + }, + { + "epoch": 0.6373590633130962, + "grad_norm": 0.6267329454421997, + "learning_rate": 1.580806615406935e-05, + "loss": 0.0173, + "step": 5879 + }, + { + "epoch": 0.6374674761491761, + "grad_norm": 2.203730344772339, + "learning_rate": 1.580734078050196e-05, + "loss": 0.0338, + "step": 5880 + }, + { + "epoch": 0.6375758889852559, + "grad_norm": 1.3507128953933716, + "learning_rate": 1.5806615406934574e-05, + "loss": 0.0116, + "step": 5881 + }, + { + "epoch": 0.6376843018213356, + "grad_norm": 0.835730254650116, + "learning_rate": 1.5805890033367186e-05, + "loss": 0.0242, + "step": 5882 + }, + { + "epoch": 0.6377927146574154, + "grad_norm": 1.1781694889068604, + "learning_rate": 1.5805164659799798e-05, + "loss": 0.0205, + "step": 5883 + }, + { + "epoch": 0.6379011274934953, + "grad_norm": 1.4262423515319824, + "learning_rate": 1.580443928623241e-05, + "loss": 0.0159, + "step": 5884 + }, + { + "epoch": 0.638009540329575, + "grad_norm": 1.675188660621643, + "learning_rate": 1.5803713912665026e-05, + "loss": 0.0198, + "step": 5885 + }, + { + "epoch": 0.6381179531656548, + "grad_norm": 4.43829345703125, + "learning_rate": 1.5802988539097638e-05, + "loss": 0.0363, + "step": 5886 + }, + { + "epoch": 0.6382263660017347, + "grad_norm": 3.017733097076416, + "learning_rate": 1.580226316553025e-05, + "loss": 0.0743, + "step": 5887 + }, + { + "epoch": 0.6383347788378144, + "grad_norm": 1.0693166255950928, + "learning_rate": 1.5801537791962862e-05, + "loss": 0.0165, + "step": 5888 + }, + { + "epoch": 0.6384431916738942, + "grad_norm": 0.7605170011520386, + "learning_rate": 1.5800812418395474e-05, + "loss": 0.0059, + "step": 5889 + }, + { + "epoch": 0.638551604509974, + "grad_norm": 3.18094539642334, + "learning_rate": 1.5800087044828086e-05, + "loss": 0.0217, + "step": 5890 + }, + { + "epoch": 0.6386600173460538, + "grad_norm": 1.0634971857070923, + "learning_rate": 1.5799361671260702e-05, + "loss": 0.0409, + "step": 5891 + }, + { + "epoch": 0.6387684301821336, + "grad_norm": 5.885505676269531, + "learning_rate": 1.5798636297693314e-05, + "loss": 0.0581, + "step": 5892 + }, + { + "epoch": 0.6388768430182133, + "grad_norm": 2.171786069869995, + "learning_rate": 1.5797910924125926e-05, + "loss": 0.0645, + "step": 5893 + }, + { + "epoch": 0.6389852558542931, + "grad_norm": 3.3582170009613037, + "learning_rate": 1.5797185550558538e-05, + "loss": 0.0377, + "step": 5894 + }, + { + "epoch": 0.639093668690373, + "grad_norm": 3.033994197845459, + "learning_rate": 1.579646017699115e-05, + "loss": 0.0378, + "step": 5895 + }, + { + "epoch": 0.6392020815264527, + "grad_norm": 1.8543941974639893, + "learning_rate": 1.5795734803423766e-05, + "loss": 0.0338, + "step": 5896 + }, + { + "epoch": 0.6393104943625325, + "grad_norm": 4.873364448547363, + "learning_rate": 1.5795009429856378e-05, + "loss": 0.1778, + "step": 5897 + }, + { + "epoch": 0.6394189071986123, + "grad_norm": 3.1637370586395264, + "learning_rate": 1.579428405628899e-05, + "loss": 0.0392, + "step": 5898 + }, + { + "epoch": 0.6395273200346921, + "grad_norm": 4.882558345794678, + "learning_rate": 1.5793558682721602e-05, + "loss": 0.0661, + "step": 5899 + }, + { + "epoch": 0.6396357328707719, + "grad_norm": 1.2988344430923462, + "learning_rate": 1.5792833309154218e-05, + "loss": 0.0114, + "step": 5900 + }, + { + "epoch": 0.6397441457068517, + "grad_norm": 2.797952890396118, + "learning_rate": 1.579210793558683e-05, + "loss": 0.031, + "step": 5901 + }, + { + "epoch": 0.6398525585429314, + "grad_norm": 3.1237287521362305, + "learning_rate": 1.5791382562019442e-05, + "loss": 0.0415, + "step": 5902 + }, + { + "epoch": 0.6399609713790113, + "grad_norm": 0.32277846336364746, + "learning_rate": 1.5790657188452054e-05, + "loss": 0.0016, + "step": 5903 + }, + { + "epoch": 0.6400693842150911, + "grad_norm": 6.099995136260986, + "learning_rate": 1.5789931814884666e-05, + "loss": 0.0883, + "step": 5904 + }, + { + "epoch": 0.6401777970511708, + "grad_norm": 5.577706813812256, + "learning_rate": 1.578920644131728e-05, + "loss": 0.0299, + "step": 5905 + }, + { + "epoch": 0.6402862098872506, + "grad_norm": 3.934678077697754, + "learning_rate": 1.578848106774989e-05, + "loss": 0.084, + "step": 5906 + }, + { + "epoch": 0.6403946227233305, + "grad_norm": 3.627925157546997, + "learning_rate": 1.5787755694182506e-05, + "loss": 0.0683, + "step": 5907 + }, + { + "epoch": 0.6405030355594102, + "grad_norm": 4.347616195678711, + "learning_rate": 1.578703032061512e-05, + "loss": 0.0722, + "step": 5908 + }, + { + "epoch": 0.64061144839549, + "grad_norm": 1.3486884832382202, + "learning_rate": 1.578630494704773e-05, + "loss": 0.0323, + "step": 5909 + }, + { + "epoch": 0.6407198612315698, + "grad_norm": 4.873210906982422, + "learning_rate": 1.5785579573480343e-05, + "loss": 0.0492, + "step": 5910 + }, + { + "epoch": 0.6408282740676496, + "grad_norm": 3.7314608097076416, + "learning_rate": 1.5784854199912955e-05, + "loss": 0.0455, + "step": 5911 + }, + { + "epoch": 0.6409366869037294, + "grad_norm": 4.148571491241455, + "learning_rate": 1.5784128826345567e-05, + "loss": 0.101, + "step": 5912 + }, + { + "epoch": 0.6410450997398092, + "grad_norm": 1.5739988088607788, + "learning_rate": 1.5783403452778182e-05, + "loss": 0.0187, + "step": 5913 + }, + { + "epoch": 0.641153512575889, + "grad_norm": 1.6318539381027222, + "learning_rate": 1.5782678079210795e-05, + "loss": 0.0318, + "step": 5914 + }, + { + "epoch": 0.6412619254119688, + "grad_norm": 5.682806491851807, + "learning_rate": 1.578195270564341e-05, + "loss": 0.0566, + "step": 5915 + }, + { + "epoch": 0.6413703382480486, + "grad_norm": 2.5831894874572754, + "learning_rate": 1.5781227332076022e-05, + "loss": 0.0387, + "step": 5916 + }, + { + "epoch": 0.6414787510841283, + "grad_norm": 3.3840982913970947, + "learning_rate": 1.5780501958508634e-05, + "loss": 0.0804, + "step": 5917 + }, + { + "epoch": 0.6415871639202082, + "grad_norm": 3.222959041595459, + "learning_rate": 1.5779776584941247e-05, + "loss": 0.0407, + "step": 5918 + }, + { + "epoch": 0.641695576756288, + "grad_norm": 2.1061320304870605, + "learning_rate": 1.577905121137386e-05, + "loss": 0.0235, + "step": 5919 + }, + { + "epoch": 0.6418039895923677, + "grad_norm": 1.0517983436584473, + "learning_rate": 1.577832583780647e-05, + "loss": 0.0231, + "step": 5920 + }, + { + "epoch": 0.6419124024284475, + "grad_norm": 1.8173335790634155, + "learning_rate": 1.5777600464239083e-05, + "loss": 0.0425, + "step": 5921 + }, + { + "epoch": 0.6420208152645274, + "grad_norm": 4.699488639831543, + "learning_rate": 1.57768750906717e-05, + "loss": 0.0312, + "step": 5922 + }, + { + "epoch": 0.6421292281006071, + "grad_norm": 4.112161159515381, + "learning_rate": 1.577614971710431e-05, + "loss": 0.0432, + "step": 5923 + }, + { + "epoch": 0.6422376409366869, + "grad_norm": 2.2054460048675537, + "learning_rate": 1.5775424343536923e-05, + "loss": 0.0443, + "step": 5924 + }, + { + "epoch": 0.6423460537727667, + "grad_norm": 3.4159164428710938, + "learning_rate": 1.5774698969969535e-05, + "loss": 0.0348, + "step": 5925 + }, + { + "epoch": 0.6424544666088465, + "grad_norm": 2.3345718383789062, + "learning_rate": 1.5773973596402147e-05, + "loss": 0.0405, + "step": 5926 + }, + { + "epoch": 0.6425628794449263, + "grad_norm": 1.3651635646820068, + "learning_rate": 1.577324822283476e-05, + "loss": 0.0146, + "step": 5927 + }, + { + "epoch": 0.6426712922810061, + "grad_norm": 1.2855435609817505, + "learning_rate": 1.5772522849267375e-05, + "loss": 0.0248, + "step": 5928 + }, + { + "epoch": 0.6427797051170858, + "grad_norm": 1.2450826168060303, + "learning_rate": 1.5771797475699987e-05, + "loss": 0.0243, + "step": 5929 + }, + { + "epoch": 0.6428881179531657, + "grad_norm": 5.086470127105713, + "learning_rate": 1.57710721021326e-05, + "loss": 0.0412, + "step": 5930 + }, + { + "epoch": 0.6429965307892455, + "grad_norm": 3.0532336235046387, + "learning_rate": 1.5770346728565215e-05, + "loss": 0.0388, + "step": 5931 + }, + { + "epoch": 0.6431049436253252, + "grad_norm": 4.6232686042785645, + "learning_rate": 1.5769621354997827e-05, + "loss": 0.0745, + "step": 5932 + }, + { + "epoch": 0.643213356461405, + "grad_norm": 2.4849541187286377, + "learning_rate": 1.576889598143044e-05, + "loss": 0.077, + "step": 5933 + }, + { + "epoch": 0.6433217692974849, + "grad_norm": 3.114569902420044, + "learning_rate": 1.576817060786305e-05, + "loss": 0.0822, + "step": 5934 + }, + { + "epoch": 0.6434301821335646, + "grad_norm": 3.233168601989746, + "learning_rate": 1.5767445234295663e-05, + "loss": 0.0709, + "step": 5935 + }, + { + "epoch": 0.6435385949696444, + "grad_norm": 2.729200839996338, + "learning_rate": 1.5766719860728275e-05, + "loss": 0.0543, + "step": 5936 + }, + { + "epoch": 0.6436470078057241, + "grad_norm": 0.8103414177894592, + "learning_rate": 1.576599448716089e-05, + "loss": 0.0113, + "step": 5937 + }, + { + "epoch": 0.643755420641804, + "grad_norm": 1.1873246431350708, + "learning_rate": 1.5765269113593503e-05, + "loss": 0.0191, + "step": 5938 + }, + { + "epoch": 0.6438638334778838, + "grad_norm": 2.925098180770874, + "learning_rate": 1.5764543740026115e-05, + "loss": 0.1047, + "step": 5939 + }, + { + "epoch": 0.6439722463139635, + "grad_norm": 1.1666646003723145, + "learning_rate": 1.5763818366458727e-05, + "loss": 0.023, + "step": 5940 + }, + { + "epoch": 0.6440806591500433, + "grad_norm": 2.600857973098755, + "learning_rate": 1.576309299289134e-05, + "loss": 0.0498, + "step": 5941 + }, + { + "epoch": 0.6441890719861232, + "grad_norm": 2.054593086242676, + "learning_rate": 1.576236761932395e-05, + "loss": 0.0548, + "step": 5942 + }, + { + "epoch": 0.644297484822203, + "grad_norm": 3.2029435634613037, + "learning_rate": 1.5761642245756567e-05, + "loss": 0.0562, + "step": 5943 + }, + { + "epoch": 0.6444058976582827, + "grad_norm": 2.0844316482543945, + "learning_rate": 1.576091687218918e-05, + "loss": 0.0374, + "step": 5944 + }, + { + "epoch": 0.6445143104943626, + "grad_norm": 1.491193175315857, + "learning_rate": 1.576019149862179e-05, + "loss": 0.0292, + "step": 5945 + }, + { + "epoch": 0.6446227233304423, + "grad_norm": 5.09755802154541, + "learning_rate": 1.5759466125054404e-05, + "loss": 0.0963, + "step": 5946 + }, + { + "epoch": 0.6447311361665221, + "grad_norm": 2.26692795753479, + "learning_rate": 1.5758740751487016e-05, + "loss": 0.0597, + "step": 5947 + }, + { + "epoch": 0.6448395490026019, + "grad_norm": 2.510316848754883, + "learning_rate": 1.575801537791963e-05, + "loss": 0.044, + "step": 5948 + }, + { + "epoch": 0.6449479618386817, + "grad_norm": 0.8638304471969604, + "learning_rate": 1.5757290004352243e-05, + "loss": 0.0107, + "step": 5949 + }, + { + "epoch": 0.6450563746747615, + "grad_norm": 3.917912244796753, + "learning_rate": 1.5756564630784856e-05, + "loss": 0.0295, + "step": 5950 + }, + { + "epoch": 0.6451647875108413, + "grad_norm": 2.1936702728271484, + "learning_rate": 1.5755839257217468e-05, + "loss": 0.03, + "step": 5951 + }, + { + "epoch": 0.645273200346921, + "grad_norm": 3.5807340145111084, + "learning_rate": 1.5755113883650083e-05, + "loss": 0.0622, + "step": 5952 + }, + { + "epoch": 0.6453816131830009, + "grad_norm": 1.830095648765564, + "learning_rate": 1.5754388510082695e-05, + "loss": 0.0458, + "step": 5953 + }, + { + "epoch": 0.6454900260190807, + "grad_norm": 4.343463897705078, + "learning_rate": 1.5753663136515307e-05, + "loss": 0.0673, + "step": 5954 + }, + { + "epoch": 0.6455984388551604, + "grad_norm": 3.0547542572021484, + "learning_rate": 1.575293776294792e-05, + "loss": 0.0682, + "step": 5955 + }, + { + "epoch": 0.6457068516912402, + "grad_norm": 5.877383708953857, + "learning_rate": 1.5752212389380532e-05, + "loss": 0.0343, + "step": 5956 + }, + { + "epoch": 0.6458152645273201, + "grad_norm": 2.373415470123291, + "learning_rate": 1.5751487015813144e-05, + "loss": 0.0124, + "step": 5957 + }, + { + "epoch": 0.6459236773633998, + "grad_norm": 3.011296033859253, + "learning_rate": 1.5750761642245756e-05, + "loss": 0.064, + "step": 5958 + }, + { + "epoch": 0.6460320901994796, + "grad_norm": 1.628387451171875, + "learning_rate": 1.575003626867837e-05, + "loss": 0.0218, + "step": 5959 + }, + { + "epoch": 0.6461405030355594, + "grad_norm": 1.069906234741211, + "learning_rate": 1.5749310895110984e-05, + "loss": 0.0229, + "step": 5960 + }, + { + "epoch": 0.6462489158716392, + "grad_norm": 1.7115447521209717, + "learning_rate": 1.5748585521543596e-05, + "loss": 0.0288, + "step": 5961 + }, + { + "epoch": 0.646357328707719, + "grad_norm": 2.1800224781036377, + "learning_rate": 1.5747860147976208e-05, + "loss": 0.0216, + "step": 5962 + }, + { + "epoch": 0.6464657415437988, + "grad_norm": 3.442474126815796, + "learning_rate": 1.574713477440882e-05, + "loss": 0.0472, + "step": 5963 + }, + { + "epoch": 0.6465741543798785, + "grad_norm": 2.8208632469177246, + "learning_rate": 1.5746409400841432e-05, + "loss": 0.0275, + "step": 5964 + }, + { + "epoch": 0.6466825672159584, + "grad_norm": 3.467439651489258, + "learning_rate": 1.5745684027274048e-05, + "loss": 0.0572, + "step": 5965 + }, + { + "epoch": 0.6467909800520382, + "grad_norm": 3.024960994720459, + "learning_rate": 1.574495865370666e-05, + "loss": 0.0258, + "step": 5966 + }, + { + "epoch": 0.6468993928881179, + "grad_norm": 0.577421247959137, + "learning_rate": 1.5744233280139275e-05, + "loss": 0.0101, + "step": 5967 + }, + { + "epoch": 0.6470078057241977, + "grad_norm": 2.7114310264587402, + "learning_rate": 1.5743507906571888e-05, + "loss": 0.0221, + "step": 5968 + }, + { + "epoch": 0.6471162185602776, + "grad_norm": 3.389761447906494, + "learning_rate": 1.57427825330045e-05, + "loss": 0.0448, + "step": 5969 + }, + { + "epoch": 0.6472246313963573, + "grad_norm": 3.737804651260376, + "learning_rate": 1.5742057159437112e-05, + "loss": 0.0181, + "step": 5970 + }, + { + "epoch": 0.6473330442324371, + "grad_norm": 4.034669876098633, + "learning_rate": 1.5741331785869724e-05, + "loss": 0.0532, + "step": 5971 + }, + { + "epoch": 0.647441457068517, + "grad_norm": 0.7361354231834412, + "learning_rate": 1.5740606412302336e-05, + "loss": 0.0175, + "step": 5972 + }, + { + "epoch": 0.6475498699045967, + "grad_norm": 2.3543338775634766, + "learning_rate": 1.573988103873495e-05, + "loss": 0.0223, + "step": 5973 + }, + { + "epoch": 0.6476582827406765, + "grad_norm": 6.550306797027588, + "learning_rate": 1.5739155665167564e-05, + "loss": 0.0679, + "step": 5974 + }, + { + "epoch": 0.6477666955767563, + "grad_norm": 2.3107590675354004, + "learning_rate": 1.5738430291600176e-05, + "loss": 0.0212, + "step": 5975 + }, + { + "epoch": 0.6478751084128361, + "grad_norm": 3.264915943145752, + "learning_rate": 1.5737704918032788e-05, + "loss": 0.0266, + "step": 5976 + }, + { + "epoch": 0.6479835212489159, + "grad_norm": 1.9525445699691772, + "learning_rate": 1.57369795444654e-05, + "loss": 0.0253, + "step": 5977 + }, + { + "epoch": 0.6480919340849957, + "grad_norm": 4.137040615081787, + "learning_rate": 1.5736254170898012e-05, + "loss": 0.0603, + "step": 5978 + }, + { + "epoch": 0.6482003469210754, + "grad_norm": 0.8034107685089111, + "learning_rate": 1.5735528797330625e-05, + "loss": 0.0048, + "step": 5979 + }, + { + "epoch": 0.6483087597571553, + "grad_norm": 3.737785816192627, + "learning_rate": 1.573480342376324e-05, + "loss": 0.026, + "step": 5980 + }, + { + "epoch": 0.6484171725932351, + "grad_norm": 2.4024529457092285, + "learning_rate": 1.5734078050195852e-05, + "loss": 0.0183, + "step": 5981 + }, + { + "epoch": 0.6485255854293148, + "grad_norm": 4.782386779785156, + "learning_rate": 1.5733352676628464e-05, + "loss": 0.0525, + "step": 5982 + }, + { + "epoch": 0.6486339982653946, + "grad_norm": 0.40419718623161316, + "learning_rate": 1.573262730306108e-05, + "loss": 0.0044, + "step": 5983 + }, + { + "epoch": 0.6487424111014745, + "grad_norm": 6.4539055824279785, + "learning_rate": 1.5731901929493692e-05, + "loss": 0.0852, + "step": 5984 + }, + { + "epoch": 0.6488508239375542, + "grad_norm": 6.283330917358398, + "learning_rate": 1.5731176555926304e-05, + "loss": 0.0515, + "step": 5985 + }, + { + "epoch": 0.648959236773634, + "grad_norm": 2.1018173694610596, + "learning_rate": 1.5730451182358916e-05, + "loss": 0.0273, + "step": 5986 + }, + { + "epoch": 0.6490676496097137, + "grad_norm": 2.394272565841675, + "learning_rate": 1.572972580879153e-05, + "loss": 0.0404, + "step": 5987 + }, + { + "epoch": 0.6491760624457936, + "grad_norm": 4.54078483581543, + "learning_rate": 1.572900043522414e-05, + "loss": 0.0708, + "step": 5988 + }, + { + "epoch": 0.6492844752818734, + "grad_norm": 2.0265488624572754, + "learning_rate": 1.5728275061656756e-05, + "loss": 0.0467, + "step": 5989 + }, + { + "epoch": 0.6493928881179531, + "grad_norm": 6.585023880004883, + "learning_rate": 1.572754968808937e-05, + "loss": 0.0445, + "step": 5990 + }, + { + "epoch": 0.6495013009540329, + "grad_norm": 2.095050096511841, + "learning_rate": 1.572682431452198e-05, + "loss": 0.0293, + "step": 5991 + }, + { + "epoch": 0.6496097137901128, + "grad_norm": 4.009614944458008, + "learning_rate": 1.5726098940954593e-05, + "loss": 0.0737, + "step": 5992 + }, + { + "epoch": 0.6497181266261925, + "grad_norm": 1.697692632675171, + "learning_rate": 1.5725373567387205e-05, + "loss": 0.0129, + "step": 5993 + }, + { + "epoch": 0.6498265394622723, + "grad_norm": 3.014399290084839, + "learning_rate": 1.5724648193819817e-05, + "loss": 0.041, + "step": 5994 + }, + { + "epoch": 0.6499349522983521, + "grad_norm": 2.4514048099517822, + "learning_rate": 1.5723922820252432e-05, + "loss": 0.0391, + "step": 5995 + }, + { + "epoch": 0.650043365134432, + "grad_norm": 1.4040817022323608, + "learning_rate": 1.5723197446685045e-05, + "loss": 0.0351, + "step": 5996 + }, + { + "epoch": 0.6501517779705117, + "grad_norm": 2.7305514812469482, + "learning_rate": 1.5722472073117657e-05, + "loss": 0.0587, + "step": 5997 + }, + { + "epoch": 0.6502601908065915, + "grad_norm": 0.9917778372764587, + "learning_rate": 1.572174669955027e-05, + "loss": 0.0103, + "step": 5998 + }, + { + "epoch": 0.6503686036426712, + "grad_norm": 2.8797714710235596, + "learning_rate": 1.572102132598288e-05, + "loss": 0.0402, + "step": 5999 + }, + { + "epoch": 0.6504770164787511, + "grad_norm": 5.2924485206604, + "learning_rate": 1.5720295952415497e-05, + "loss": 0.0643, + "step": 6000 + }, + { + "epoch": 0.6505854293148309, + "grad_norm": 1.731285572052002, + "learning_rate": 1.571957057884811e-05, + "loss": 0.0155, + "step": 6001 + }, + { + "epoch": 0.6506938421509106, + "grad_norm": 3.807220697402954, + "learning_rate": 1.571884520528072e-05, + "loss": 0.0343, + "step": 6002 + }, + { + "epoch": 0.6508022549869905, + "grad_norm": 1.9228607416152954, + "learning_rate": 1.5718119831713333e-05, + "loss": 0.0299, + "step": 6003 + }, + { + "epoch": 0.6509106678230703, + "grad_norm": 3.0847458839416504, + "learning_rate": 1.571739445814595e-05, + "loss": 0.0376, + "step": 6004 + }, + { + "epoch": 0.65101908065915, + "grad_norm": 3.824052333831787, + "learning_rate": 1.571666908457856e-05, + "loss": 0.035, + "step": 6005 + }, + { + "epoch": 0.6511274934952298, + "grad_norm": 3.4792075157165527, + "learning_rate": 1.5715943711011173e-05, + "loss": 0.0308, + "step": 6006 + }, + { + "epoch": 0.6512359063313097, + "grad_norm": 3.722078561782837, + "learning_rate": 1.5715218337443785e-05, + "loss": 0.0282, + "step": 6007 + }, + { + "epoch": 0.6513443191673894, + "grad_norm": 3.3622372150421143, + "learning_rate": 1.5714492963876397e-05, + "loss": 0.0134, + "step": 6008 + }, + { + "epoch": 0.6514527320034692, + "grad_norm": 2.9185879230499268, + "learning_rate": 1.571376759030901e-05, + "loss": 0.0263, + "step": 6009 + }, + { + "epoch": 0.651561144839549, + "grad_norm": 0.5979353785514832, + "learning_rate": 1.571304221674162e-05, + "loss": 0.0063, + "step": 6010 + }, + { + "epoch": 0.6516695576756288, + "grad_norm": 2.9874866008758545, + "learning_rate": 1.5712316843174237e-05, + "loss": 0.0422, + "step": 6011 + }, + { + "epoch": 0.6517779705117086, + "grad_norm": 6.469560146331787, + "learning_rate": 1.571159146960685e-05, + "loss": 0.1091, + "step": 6012 + }, + { + "epoch": 0.6518863833477884, + "grad_norm": 2.7014167308807373, + "learning_rate": 1.571086609603946e-05, + "loss": 0.0424, + "step": 6013 + }, + { + "epoch": 0.6519947961838681, + "grad_norm": 5.36643648147583, + "learning_rate": 1.5710140722472073e-05, + "loss": 0.0228, + "step": 6014 + }, + { + "epoch": 0.652103209019948, + "grad_norm": 1.899145483970642, + "learning_rate": 1.5709415348904685e-05, + "loss": 0.03, + "step": 6015 + }, + { + "epoch": 0.6522116218560278, + "grad_norm": 2.642974376678467, + "learning_rate": 1.5708689975337298e-05, + "loss": 0.058, + "step": 6016 + }, + { + "epoch": 0.6523200346921075, + "grad_norm": 0.6861225366592407, + "learning_rate": 1.5707964601769913e-05, + "loss": 0.0147, + "step": 6017 + }, + { + "epoch": 0.6524284475281873, + "grad_norm": 3.051898717880249, + "learning_rate": 1.5707239228202525e-05, + "loss": 0.0656, + "step": 6018 + }, + { + "epoch": 0.6525368603642672, + "grad_norm": 4.734655857086182, + "learning_rate": 1.570651385463514e-05, + "loss": 0.0279, + "step": 6019 + }, + { + "epoch": 0.6526452732003469, + "grad_norm": 1.1693589687347412, + "learning_rate": 1.5705788481067753e-05, + "loss": 0.0312, + "step": 6020 + }, + { + "epoch": 0.6527536860364267, + "grad_norm": 1.7636234760284424, + "learning_rate": 1.5705063107500365e-05, + "loss": 0.0403, + "step": 6021 + }, + { + "epoch": 0.6528620988725065, + "grad_norm": 4.163998603820801, + "learning_rate": 1.5704337733932977e-05, + "loss": 0.0725, + "step": 6022 + }, + { + "epoch": 0.6529705117085863, + "grad_norm": 4.791207790374756, + "learning_rate": 1.570361236036559e-05, + "loss": 0.0464, + "step": 6023 + }, + { + "epoch": 0.6530789245446661, + "grad_norm": 4.3857197761535645, + "learning_rate": 1.57028869867982e-05, + "loss": 0.0687, + "step": 6024 + }, + { + "epoch": 0.6531873373807459, + "grad_norm": 2.8604891300201416, + "learning_rate": 1.5702161613230814e-05, + "loss": 0.0803, + "step": 6025 + }, + { + "epoch": 0.6532957502168256, + "grad_norm": 3.4092414379119873, + "learning_rate": 1.570143623966343e-05, + "loss": 0.0498, + "step": 6026 + }, + { + "epoch": 0.6534041630529055, + "grad_norm": 2.20501971244812, + "learning_rate": 1.570071086609604e-05, + "loss": 0.0469, + "step": 6027 + }, + { + "epoch": 0.6535125758889853, + "grad_norm": 2.0721938610076904, + "learning_rate": 1.5699985492528654e-05, + "loss": 0.0387, + "step": 6028 + }, + { + "epoch": 0.653620988725065, + "grad_norm": 2.105156660079956, + "learning_rate": 1.5699260118961266e-05, + "loss": 0.0159, + "step": 6029 + }, + { + "epoch": 0.6537294015611448, + "grad_norm": 1.217538595199585, + "learning_rate": 1.5698534745393878e-05, + "loss": 0.0159, + "step": 6030 + }, + { + "epoch": 0.6538378143972247, + "grad_norm": 2.2326064109802246, + "learning_rate": 1.569780937182649e-05, + "loss": 0.0333, + "step": 6031 + }, + { + "epoch": 0.6539462272333044, + "grad_norm": 2.1184380054473877, + "learning_rate": 1.5697083998259105e-05, + "loss": 0.0758, + "step": 6032 + }, + { + "epoch": 0.6540546400693842, + "grad_norm": 1.625888705253601, + "learning_rate": 1.5696358624691718e-05, + "loss": 0.0308, + "step": 6033 + }, + { + "epoch": 0.6541630529054641, + "grad_norm": 3.494072914123535, + "learning_rate": 1.569563325112433e-05, + "loss": 0.0475, + "step": 6034 + }, + { + "epoch": 0.6542714657415438, + "grad_norm": 3.62935209274292, + "learning_rate": 1.5694907877556945e-05, + "loss": 0.063, + "step": 6035 + }, + { + "epoch": 0.6543798785776236, + "grad_norm": 0.9775118827819824, + "learning_rate": 1.5694182503989557e-05, + "loss": 0.015, + "step": 6036 + }, + { + "epoch": 0.6544882914137033, + "grad_norm": 5.434805870056152, + "learning_rate": 1.569345713042217e-05, + "loss": 0.0337, + "step": 6037 + }, + { + "epoch": 0.6545967042497832, + "grad_norm": 2.5667848587036133, + "learning_rate": 1.5692731756854782e-05, + "loss": 0.0395, + "step": 6038 + }, + { + "epoch": 0.654705117085863, + "grad_norm": 4.548742771148682, + "learning_rate": 1.5692006383287394e-05, + "loss": 0.0337, + "step": 6039 + }, + { + "epoch": 0.6548135299219427, + "grad_norm": 2.7279274463653564, + "learning_rate": 1.5691281009720006e-05, + "loss": 0.0517, + "step": 6040 + }, + { + "epoch": 0.6549219427580225, + "grad_norm": 1.5701199769973755, + "learning_rate": 1.569055563615262e-05, + "loss": 0.0184, + "step": 6041 + }, + { + "epoch": 0.6550303555941024, + "grad_norm": 1.9172840118408203, + "learning_rate": 1.5689830262585234e-05, + "loss": 0.0434, + "step": 6042 + }, + { + "epoch": 0.6551387684301822, + "grad_norm": 3.6194541454315186, + "learning_rate": 1.5689104889017846e-05, + "loss": 0.0932, + "step": 6043 + }, + { + "epoch": 0.6552471812662619, + "grad_norm": 2.70809268951416, + "learning_rate": 1.5688379515450458e-05, + "loss": 0.0421, + "step": 6044 + }, + { + "epoch": 0.6553555941023417, + "grad_norm": 0.7303289175033569, + "learning_rate": 1.568765414188307e-05, + "loss": 0.0131, + "step": 6045 + }, + { + "epoch": 0.6554640069384216, + "grad_norm": 1.0463991165161133, + "learning_rate": 1.5686928768315682e-05, + "loss": 0.016, + "step": 6046 + }, + { + "epoch": 0.6555724197745013, + "grad_norm": 1.9368054866790771, + "learning_rate": 1.5686203394748298e-05, + "loss": 0.0233, + "step": 6047 + }, + { + "epoch": 0.6556808326105811, + "grad_norm": 0.951501727104187, + "learning_rate": 1.568547802118091e-05, + "loss": 0.0229, + "step": 6048 + }, + { + "epoch": 0.6557892454466608, + "grad_norm": 1.7780228853225708, + "learning_rate": 1.5684752647613522e-05, + "loss": 0.0212, + "step": 6049 + }, + { + "epoch": 0.6558976582827407, + "grad_norm": 0.9707716107368469, + "learning_rate": 1.5684027274046134e-05, + "loss": 0.0233, + "step": 6050 + }, + { + "epoch": 0.6560060711188205, + "grad_norm": 3.2181994915008545, + "learning_rate": 1.5683301900478746e-05, + "loss": 0.0376, + "step": 6051 + }, + { + "epoch": 0.6561144839549002, + "grad_norm": 1.734553337097168, + "learning_rate": 1.5682576526911362e-05, + "loss": 0.0108, + "step": 6052 + }, + { + "epoch": 0.65622289679098, + "grad_norm": 0.6119619607925415, + "learning_rate": 1.5681851153343974e-05, + "loss": 0.0098, + "step": 6053 + }, + { + "epoch": 0.6563313096270599, + "grad_norm": 0.311501681804657, + "learning_rate": 1.5681125779776586e-05, + "loss": 0.0023, + "step": 6054 + }, + { + "epoch": 0.6564397224631396, + "grad_norm": 0.5923037528991699, + "learning_rate": 1.56804004062092e-05, + "loss": 0.008, + "step": 6055 + }, + { + "epoch": 0.6565481352992194, + "grad_norm": 8.04394817352295, + "learning_rate": 1.5679675032641814e-05, + "loss": 0.0473, + "step": 6056 + }, + { + "epoch": 0.6566565481352992, + "grad_norm": 9.459967613220215, + "learning_rate": 1.5678949659074426e-05, + "loss": 0.0438, + "step": 6057 + }, + { + "epoch": 0.656764960971379, + "grad_norm": 2.914273500442505, + "learning_rate": 1.5678224285507038e-05, + "loss": 0.0279, + "step": 6058 + }, + { + "epoch": 0.6568733738074588, + "grad_norm": 1.051672339439392, + "learning_rate": 1.567749891193965e-05, + "loss": 0.0118, + "step": 6059 + }, + { + "epoch": 0.6569817866435386, + "grad_norm": 3.635741710662842, + "learning_rate": 1.5676773538372262e-05, + "loss": 0.0211, + "step": 6060 + }, + { + "epoch": 0.6570901994796184, + "grad_norm": 2.2568812370300293, + "learning_rate": 1.5676048164804875e-05, + "loss": 0.045, + "step": 6061 + }, + { + "epoch": 0.6571986123156982, + "grad_norm": 3.7245211601257324, + "learning_rate": 1.5675322791237487e-05, + "loss": 0.0356, + "step": 6062 + }, + { + "epoch": 0.657307025151778, + "grad_norm": 1.1968512535095215, + "learning_rate": 1.5674597417670102e-05, + "loss": 0.0184, + "step": 6063 + }, + { + "epoch": 0.6574154379878577, + "grad_norm": 1.9390817880630493, + "learning_rate": 1.5673872044102714e-05, + "loss": 0.0226, + "step": 6064 + }, + { + "epoch": 0.6575238508239376, + "grad_norm": 3.457296371459961, + "learning_rate": 1.5673146670535327e-05, + "loss": 0.0303, + "step": 6065 + }, + { + "epoch": 0.6576322636600174, + "grad_norm": 4.3234543800354, + "learning_rate": 1.567242129696794e-05, + "loss": 0.1187, + "step": 6066 + }, + { + "epoch": 0.6577406764960971, + "grad_norm": 1.8454574346542358, + "learning_rate": 1.567169592340055e-05, + "loss": 0.0176, + "step": 6067 + }, + { + "epoch": 0.6578490893321769, + "grad_norm": 3.4383530616760254, + "learning_rate": 1.5670970549833163e-05, + "loss": 0.0526, + "step": 6068 + }, + { + "epoch": 0.6579575021682568, + "grad_norm": 5.234161376953125, + "learning_rate": 1.567024517626578e-05, + "loss": 0.0284, + "step": 6069 + }, + { + "epoch": 0.6580659150043365, + "grad_norm": 1.5036468505859375, + "learning_rate": 1.566951980269839e-05, + "loss": 0.0164, + "step": 6070 + }, + { + "epoch": 0.6581743278404163, + "grad_norm": 3.970134735107422, + "learning_rate": 1.5668794429131006e-05, + "loss": 0.0737, + "step": 6071 + }, + { + "epoch": 0.6582827406764961, + "grad_norm": 3.900237798690796, + "learning_rate": 1.566806905556362e-05, + "loss": 0.0157, + "step": 6072 + }, + { + "epoch": 0.6583911535125759, + "grad_norm": 0.4512295424938202, + "learning_rate": 1.566734368199623e-05, + "loss": 0.0045, + "step": 6073 + }, + { + "epoch": 0.6584995663486557, + "grad_norm": 2.8641555309295654, + "learning_rate": 1.5666618308428843e-05, + "loss": 0.0139, + "step": 6074 + }, + { + "epoch": 0.6586079791847355, + "grad_norm": 1.1507844924926758, + "learning_rate": 1.5665892934861455e-05, + "loss": 0.0164, + "step": 6075 + }, + { + "epoch": 0.6587163920208152, + "grad_norm": 1.9622400999069214, + "learning_rate": 1.5665167561294067e-05, + "loss": 0.0337, + "step": 6076 + }, + { + "epoch": 0.6588248048568951, + "grad_norm": 3.0438570976257324, + "learning_rate": 1.566444218772668e-05, + "loss": 0.0717, + "step": 6077 + }, + { + "epoch": 0.6589332176929749, + "grad_norm": 6.141500949859619, + "learning_rate": 1.5663716814159295e-05, + "loss": 0.0804, + "step": 6078 + }, + { + "epoch": 0.6590416305290546, + "grad_norm": 2.7919483184814453, + "learning_rate": 1.5662991440591907e-05, + "loss": 0.0167, + "step": 6079 + }, + { + "epoch": 0.6591500433651344, + "grad_norm": 2.178694248199463, + "learning_rate": 1.566226606702452e-05, + "loss": 0.0473, + "step": 6080 + }, + { + "epoch": 0.6592584562012143, + "grad_norm": 3.883009672164917, + "learning_rate": 1.566154069345713e-05, + "loss": 0.0472, + "step": 6081 + }, + { + "epoch": 0.659366869037294, + "grad_norm": 3.608062744140625, + "learning_rate": 1.5660815319889743e-05, + "loss": 0.0177, + "step": 6082 + }, + { + "epoch": 0.6594752818733738, + "grad_norm": 3.6912848949432373, + "learning_rate": 1.5660089946322355e-05, + "loss": 0.0436, + "step": 6083 + }, + { + "epoch": 0.6595836947094535, + "grad_norm": 4.473208904266357, + "learning_rate": 1.565936457275497e-05, + "loss": 0.0888, + "step": 6084 + }, + { + "epoch": 0.6596921075455334, + "grad_norm": 4.199167251586914, + "learning_rate": 1.5658639199187583e-05, + "loss": 0.041, + "step": 6085 + }, + { + "epoch": 0.6598005203816132, + "grad_norm": 3.3453097343444824, + "learning_rate": 1.5657913825620195e-05, + "loss": 0.0411, + "step": 6086 + }, + { + "epoch": 0.659908933217693, + "grad_norm": 27.58759117126465, + "learning_rate": 1.565718845205281e-05, + "loss": 0.0633, + "step": 6087 + }, + { + "epoch": 0.6600173460537727, + "grad_norm": 8.871274948120117, + "learning_rate": 1.5656463078485423e-05, + "loss": 0.1019, + "step": 6088 + }, + { + "epoch": 0.6601257588898526, + "grad_norm": 6.056930065155029, + "learning_rate": 1.5655737704918035e-05, + "loss": 0.0073, + "step": 6089 + }, + { + "epoch": 0.6602341717259324, + "grad_norm": 1.6269667148590088, + "learning_rate": 1.5655012331350647e-05, + "loss": 0.012, + "step": 6090 + }, + { + "epoch": 0.6603425845620121, + "grad_norm": 4.550601005554199, + "learning_rate": 1.565428695778326e-05, + "loss": 0.0484, + "step": 6091 + }, + { + "epoch": 0.660450997398092, + "grad_norm": 4.133365631103516, + "learning_rate": 1.565356158421587e-05, + "loss": 0.0372, + "step": 6092 + }, + { + "epoch": 0.6605594102341718, + "grad_norm": 1.2225006818771362, + "learning_rate": 1.5652836210648487e-05, + "loss": 0.0153, + "step": 6093 + }, + { + "epoch": 0.6606678230702515, + "grad_norm": 3.3535866737365723, + "learning_rate": 1.56521108370811e-05, + "loss": 0.0388, + "step": 6094 + }, + { + "epoch": 0.6607762359063313, + "grad_norm": 4.160021781921387, + "learning_rate": 1.565138546351371e-05, + "loss": 0.0533, + "step": 6095 + }, + { + "epoch": 0.6608846487424112, + "grad_norm": 1.494924545288086, + "learning_rate": 1.5650660089946323e-05, + "loss": 0.0358, + "step": 6096 + }, + { + "epoch": 0.6609930615784909, + "grad_norm": 2.8687148094177246, + "learning_rate": 1.5649934716378935e-05, + "loss": 0.0587, + "step": 6097 + }, + { + "epoch": 0.6611014744145707, + "grad_norm": 4.794931411743164, + "learning_rate": 1.5649209342811548e-05, + "loss": 0.0267, + "step": 6098 + }, + { + "epoch": 0.6612098872506504, + "grad_norm": 3.6789774894714355, + "learning_rate": 1.5648483969244163e-05, + "loss": 0.059, + "step": 6099 + }, + { + "epoch": 0.6613183000867303, + "grad_norm": 3.6765995025634766, + "learning_rate": 1.5647758595676775e-05, + "loss": 0.0665, + "step": 6100 + }, + { + "epoch": 0.6614267129228101, + "grad_norm": 4.366245269775391, + "learning_rate": 1.5647033222109387e-05, + "loss": 0.0257, + "step": 6101 + }, + { + "epoch": 0.6615351257588898, + "grad_norm": 4.678951263427734, + "learning_rate": 1.5646307848542e-05, + "loss": 0.0359, + "step": 6102 + }, + { + "epoch": 0.6616435385949696, + "grad_norm": 2.1530020236968994, + "learning_rate": 1.5645582474974612e-05, + "loss": 0.0252, + "step": 6103 + }, + { + "epoch": 0.6617519514310495, + "grad_norm": 1.3069207668304443, + "learning_rate": 1.5644857101407227e-05, + "loss": 0.0262, + "step": 6104 + }, + { + "epoch": 0.6618603642671292, + "grad_norm": 1.2811626195907593, + "learning_rate": 1.564413172783984e-05, + "loss": 0.0216, + "step": 6105 + }, + { + "epoch": 0.661968777103209, + "grad_norm": 3.5309860706329346, + "learning_rate": 1.564340635427245e-05, + "loss": 0.0485, + "step": 6106 + }, + { + "epoch": 0.6620771899392888, + "grad_norm": 4.4887309074401855, + "learning_rate": 1.5642680980705064e-05, + "loss": 0.0388, + "step": 6107 + }, + { + "epoch": 0.6621856027753686, + "grad_norm": 3.9008283615112305, + "learning_rate": 1.564195560713768e-05, + "loss": 0.067, + "step": 6108 + }, + { + "epoch": 0.6622940156114484, + "grad_norm": 2.6525514125823975, + "learning_rate": 1.564123023357029e-05, + "loss": 0.0185, + "step": 6109 + }, + { + "epoch": 0.6624024284475282, + "grad_norm": 2.9529268741607666, + "learning_rate": 1.5640504860002903e-05, + "loss": 0.0377, + "step": 6110 + }, + { + "epoch": 0.6625108412836079, + "grad_norm": 5.719120025634766, + "learning_rate": 1.5639779486435516e-05, + "loss": 0.0545, + "step": 6111 + }, + { + "epoch": 0.6626192541196878, + "grad_norm": 2.504098415374756, + "learning_rate": 1.5639054112868128e-05, + "loss": 0.036, + "step": 6112 + }, + { + "epoch": 0.6627276669557676, + "grad_norm": 3.9211220741271973, + "learning_rate": 1.563832873930074e-05, + "loss": 0.0093, + "step": 6113 + }, + { + "epoch": 0.6628360797918473, + "grad_norm": 1.7901114225387573, + "learning_rate": 1.5637603365733352e-05, + "loss": 0.0248, + "step": 6114 + }, + { + "epoch": 0.6629444926279271, + "grad_norm": 3.1673073768615723, + "learning_rate": 1.5636877992165968e-05, + "loss": 0.0471, + "step": 6115 + }, + { + "epoch": 0.663052905464007, + "grad_norm": 2.267972469329834, + "learning_rate": 1.563615261859858e-05, + "loss": 0.0277, + "step": 6116 + }, + { + "epoch": 0.6631613183000867, + "grad_norm": 1.5589734315872192, + "learning_rate": 1.5635427245031192e-05, + "loss": 0.0246, + "step": 6117 + }, + { + "epoch": 0.6632697311361665, + "grad_norm": 1.48273766040802, + "learning_rate": 1.5634701871463804e-05, + "loss": 0.0174, + "step": 6118 + }, + { + "epoch": 0.6633781439722463, + "grad_norm": 1.2062281370162964, + "learning_rate": 1.5633976497896416e-05, + "loss": 0.0299, + "step": 6119 + }, + { + "epoch": 0.6634865568083261, + "grad_norm": 3.5435116291046143, + "learning_rate": 1.5633251124329028e-05, + "loss": 0.0471, + "step": 6120 + }, + { + "epoch": 0.6635949696444059, + "grad_norm": 2.7264084815979004, + "learning_rate": 1.5632525750761644e-05, + "loss": 0.0744, + "step": 6121 + }, + { + "epoch": 0.6637033824804857, + "grad_norm": 1.5609949827194214, + "learning_rate": 1.5631800377194256e-05, + "loss": 0.0141, + "step": 6122 + }, + { + "epoch": 0.6638117953165655, + "grad_norm": 2.3799211978912354, + "learning_rate": 1.563107500362687e-05, + "loss": 0.042, + "step": 6123 + }, + { + "epoch": 0.6639202081526453, + "grad_norm": 1.2398124933242798, + "learning_rate": 1.5630349630059484e-05, + "loss": 0.0293, + "step": 6124 + }, + { + "epoch": 0.6640286209887251, + "grad_norm": 2.901505947113037, + "learning_rate": 1.5629624256492096e-05, + "loss": 0.0341, + "step": 6125 + }, + { + "epoch": 0.6641370338248048, + "grad_norm": 0.7565484046936035, + "learning_rate": 1.5628898882924708e-05, + "loss": 0.0042, + "step": 6126 + }, + { + "epoch": 0.6642454466608847, + "grad_norm": 5.01582145690918, + "learning_rate": 1.562817350935732e-05, + "loss": 0.0309, + "step": 6127 + }, + { + "epoch": 0.6643538594969645, + "grad_norm": 3.4315876960754395, + "learning_rate": 1.5627448135789932e-05, + "loss": 0.0429, + "step": 6128 + }, + { + "epoch": 0.6644622723330442, + "grad_norm": 1.6922279596328735, + "learning_rate": 1.5626722762222544e-05, + "loss": 0.0553, + "step": 6129 + }, + { + "epoch": 0.664570685169124, + "grad_norm": 5.319772243499756, + "learning_rate": 1.562599738865516e-05, + "loss": 0.0874, + "step": 6130 + }, + { + "epoch": 0.6646790980052039, + "grad_norm": 3.682748317718506, + "learning_rate": 1.5625272015087772e-05, + "loss": 0.0585, + "step": 6131 + }, + { + "epoch": 0.6647875108412836, + "grad_norm": 2.1884639263153076, + "learning_rate": 1.5624546641520384e-05, + "loss": 0.0239, + "step": 6132 + }, + { + "epoch": 0.6648959236773634, + "grad_norm": 3.747525691986084, + "learning_rate": 1.5623821267952996e-05, + "loss": 0.0618, + "step": 6133 + }, + { + "epoch": 0.6650043365134432, + "grad_norm": 4.667129993438721, + "learning_rate": 1.562309589438561e-05, + "loss": 0.0194, + "step": 6134 + }, + { + "epoch": 0.665112749349523, + "grad_norm": 1.6379183530807495, + "learning_rate": 1.562237052081822e-05, + "loss": 0.0251, + "step": 6135 + }, + { + "epoch": 0.6652211621856028, + "grad_norm": 2.9800195693969727, + "learning_rate": 1.5621645147250836e-05, + "loss": 0.0667, + "step": 6136 + }, + { + "epoch": 0.6653295750216826, + "grad_norm": 2.0309832096099854, + "learning_rate": 1.5620919773683448e-05, + "loss": 0.0267, + "step": 6137 + }, + { + "epoch": 0.6654379878577623, + "grad_norm": 2.000577688217163, + "learning_rate": 1.5620194400116064e-05, + "loss": 0.057, + "step": 6138 + }, + { + "epoch": 0.6655464006938422, + "grad_norm": 2.193079948425293, + "learning_rate": 1.5619469026548676e-05, + "loss": 0.0336, + "step": 6139 + }, + { + "epoch": 0.665654813529922, + "grad_norm": 3.7942521572113037, + "learning_rate": 1.5618743652981288e-05, + "loss": 0.0489, + "step": 6140 + }, + { + "epoch": 0.6657632263660017, + "grad_norm": 1.2286449670791626, + "learning_rate": 1.56180182794139e-05, + "loss": 0.0129, + "step": 6141 + }, + { + "epoch": 0.6658716392020815, + "grad_norm": 3.810485363006592, + "learning_rate": 1.5617292905846512e-05, + "loss": 0.0725, + "step": 6142 + }, + { + "epoch": 0.6659800520381614, + "grad_norm": 1.701354742050171, + "learning_rate": 1.5616567532279125e-05, + "loss": 0.0337, + "step": 6143 + }, + { + "epoch": 0.6660884648742411, + "grad_norm": 3.8042140007019043, + "learning_rate": 1.5615842158711737e-05, + "loss": 0.0255, + "step": 6144 + }, + { + "epoch": 0.6661968777103209, + "grad_norm": 2.8371083736419678, + "learning_rate": 1.5615116785144352e-05, + "loss": 0.0304, + "step": 6145 + }, + { + "epoch": 0.6663052905464006, + "grad_norm": 3.1027045249938965, + "learning_rate": 1.5614391411576964e-05, + "loss": 0.0684, + "step": 6146 + }, + { + "epoch": 0.6664137033824805, + "grad_norm": 2.4105985164642334, + "learning_rate": 1.5613666038009577e-05, + "loss": 0.0358, + "step": 6147 + }, + { + "epoch": 0.6665221162185603, + "grad_norm": 2.3883845806121826, + "learning_rate": 1.561294066444219e-05, + "loss": 0.0542, + "step": 6148 + }, + { + "epoch": 0.66663052905464, + "grad_norm": 2.4431707859039307, + "learning_rate": 1.56122152908748e-05, + "loss": 0.0287, + "step": 6149 + }, + { + "epoch": 0.6667389418907199, + "grad_norm": 4.030784606933594, + "learning_rate": 1.5611489917307413e-05, + "loss": 0.0319, + "step": 6150 + }, + { + "epoch": 0.6668473547267997, + "grad_norm": 0.9959144592285156, + "learning_rate": 1.561076454374003e-05, + "loss": 0.0327, + "step": 6151 + }, + { + "epoch": 0.6669557675628794, + "grad_norm": 0.8357388973236084, + "learning_rate": 1.561003917017264e-05, + "loss": 0.014, + "step": 6152 + }, + { + "epoch": 0.6670641803989592, + "grad_norm": 2.9587509632110596, + "learning_rate": 1.5609313796605253e-05, + "loss": 0.0306, + "step": 6153 + }, + { + "epoch": 0.6671725932350391, + "grad_norm": 5.776952743530273, + "learning_rate": 1.5608588423037865e-05, + "loss": 0.0712, + "step": 6154 + }, + { + "epoch": 0.6672810060711188, + "grad_norm": 1.3078559637069702, + "learning_rate": 1.5607863049470477e-05, + "loss": 0.0309, + "step": 6155 + }, + { + "epoch": 0.6673894189071986, + "grad_norm": 0.9026651978492737, + "learning_rate": 1.5607137675903093e-05, + "loss": 0.011, + "step": 6156 + }, + { + "epoch": 0.6674978317432784, + "grad_norm": 0.730683445930481, + "learning_rate": 1.5606412302335705e-05, + "loss": 0.0247, + "step": 6157 + }, + { + "epoch": 0.6676062445793582, + "grad_norm": 4.692252159118652, + "learning_rate": 1.5605686928768317e-05, + "loss": 0.0812, + "step": 6158 + }, + { + "epoch": 0.667714657415438, + "grad_norm": 1.9655832052230835, + "learning_rate": 1.560496155520093e-05, + "loss": 0.0222, + "step": 6159 + }, + { + "epoch": 0.6678230702515178, + "grad_norm": 2.8439667224884033, + "learning_rate": 1.5604236181633545e-05, + "loss": 0.0449, + "step": 6160 + }, + { + "epoch": 0.6679314830875975, + "grad_norm": 4.980482578277588, + "learning_rate": 1.5603510808066157e-05, + "loss": 0.023, + "step": 6161 + }, + { + "epoch": 0.6680398959236774, + "grad_norm": 2.2662017345428467, + "learning_rate": 1.560278543449877e-05, + "loss": 0.044, + "step": 6162 + }, + { + "epoch": 0.6681483087597572, + "grad_norm": 2.9352095127105713, + "learning_rate": 1.560206006093138e-05, + "loss": 0.0498, + "step": 6163 + }, + { + "epoch": 0.6682567215958369, + "grad_norm": 2.377857208251953, + "learning_rate": 1.5601334687363993e-05, + "loss": 0.0388, + "step": 6164 + }, + { + "epoch": 0.6683651344319167, + "grad_norm": 3.0115177631378174, + "learning_rate": 1.5600609313796605e-05, + "loss": 0.0634, + "step": 6165 + }, + { + "epoch": 0.6684735472679966, + "grad_norm": 1.9908192157745361, + "learning_rate": 1.5599883940229217e-05, + "loss": 0.0381, + "step": 6166 + }, + { + "epoch": 0.6685819601040763, + "grad_norm": 0.8815810680389404, + "learning_rate": 1.5599158566661833e-05, + "loss": 0.0091, + "step": 6167 + }, + { + "epoch": 0.6686903729401561, + "grad_norm": 3.403059482574463, + "learning_rate": 1.5598433193094445e-05, + "loss": 0.0457, + "step": 6168 + }, + { + "epoch": 0.6687987857762359, + "grad_norm": 2.7381350994110107, + "learning_rate": 1.5597707819527057e-05, + "loss": 0.0298, + "step": 6169 + }, + { + "epoch": 0.6689071986123157, + "grad_norm": 4.743663311004639, + "learning_rate": 1.559698244595967e-05, + "loss": 0.0665, + "step": 6170 + }, + { + "epoch": 0.6690156114483955, + "grad_norm": 8.478625297546387, + "learning_rate": 1.559625707239228e-05, + "loss": 0.0433, + "step": 6171 + }, + { + "epoch": 0.6691240242844753, + "grad_norm": 3.6883766651153564, + "learning_rate": 1.5595531698824894e-05, + "loss": 0.0585, + "step": 6172 + }, + { + "epoch": 0.669232437120555, + "grad_norm": 4.698849201202393, + "learning_rate": 1.559480632525751e-05, + "loss": 0.0439, + "step": 6173 + }, + { + "epoch": 0.6693408499566349, + "grad_norm": 5.1746954917907715, + "learning_rate": 1.559408095169012e-05, + "loss": 0.0755, + "step": 6174 + }, + { + "epoch": 0.6694492627927147, + "grad_norm": 5.4259233474731445, + "learning_rate": 1.5593355578122737e-05, + "loss": 0.0468, + "step": 6175 + }, + { + "epoch": 0.6695576756287944, + "grad_norm": 3.7726659774780273, + "learning_rate": 1.559263020455535e-05, + "loss": 0.0505, + "step": 6176 + }, + { + "epoch": 0.6696660884648742, + "grad_norm": 2.202348232269287, + "learning_rate": 1.559190483098796e-05, + "loss": 0.0221, + "step": 6177 + }, + { + "epoch": 0.6697745013009541, + "grad_norm": 1.4004722833633423, + "learning_rate": 1.5591179457420573e-05, + "loss": 0.0318, + "step": 6178 + }, + { + "epoch": 0.6698829141370338, + "grad_norm": 2.8907978534698486, + "learning_rate": 1.5590454083853185e-05, + "loss": 0.022, + "step": 6179 + }, + { + "epoch": 0.6699913269731136, + "grad_norm": 6.277833938598633, + "learning_rate": 1.5589728710285798e-05, + "loss": 0.0574, + "step": 6180 + }, + { + "epoch": 0.6700997398091935, + "grad_norm": 3.0341475009918213, + "learning_rate": 1.558900333671841e-05, + "loss": 0.0259, + "step": 6181 + }, + { + "epoch": 0.6702081526452732, + "grad_norm": 3.2224369049072266, + "learning_rate": 1.5588277963151025e-05, + "loss": 0.0193, + "step": 6182 + }, + { + "epoch": 0.670316565481353, + "grad_norm": 3.846320867538452, + "learning_rate": 1.5587552589583637e-05, + "loss": 0.093, + "step": 6183 + }, + { + "epoch": 0.6704249783174328, + "grad_norm": 2.4920156002044678, + "learning_rate": 1.558682721601625e-05, + "loss": 0.0565, + "step": 6184 + }, + { + "epoch": 0.6705333911535126, + "grad_norm": 6.484341621398926, + "learning_rate": 1.558610184244886e-05, + "loss": 0.0639, + "step": 6185 + }, + { + "epoch": 0.6706418039895924, + "grad_norm": 1.0425043106079102, + "learning_rate": 1.5585376468881474e-05, + "loss": 0.0237, + "step": 6186 + }, + { + "epoch": 0.6707502168256722, + "grad_norm": 1.9596869945526123, + "learning_rate": 1.5584651095314086e-05, + "loss": 0.0485, + "step": 6187 + }, + { + "epoch": 0.6708586296617519, + "grad_norm": 3.6188700199127197, + "learning_rate": 1.55839257217467e-05, + "loss": 0.0432, + "step": 6188 + }, + { + "epoch": 0.6709670424978318, + "grad_norm": 4.513874530792236, + "learning_rate": 1.5583200348179314e-05, + "loss": 0.0599, + "step": 6189 + }, + { + "epoch": 0.6710754553339116, + "grad_norm": 2.5682384967803955, + "learning_rate": 1.558247497461193e-05, + "loss": 0.0366, + "step": 6190 + }, + { + "epoch": 0.6711838681699913, + "grad_norm": 2.396463394165039, + "learning_rate": 1.558174960104454e-05, + "loss": 0.041, + "step": 6191 + }, + { + "epoch": 0.6712922810060711, + "grad_norm": 5.055145263671875, + "learning_rate": 1.5581024227477153e-05, + "loss": 0.0342, + "step": 6192 + }, + { + "epoch": 0.671400693842151, + "grad_norm": 0.9915660619735718, + "learning_rate": 1.5580298853909766e-05, + "loss": 0.0213, + "step": 6193 + }, + { + "epoch": 0.6715091066782307, + "grad_norm": 5.77639627456665, + "learning_rate": 1.5579573480342378e-05, + "loss": 0.0615, + "step": 6194 + }, + { + "epoch": 0.6716175195143105, + "grad_norm": 1.482914924621582, + "learning_rate": 1.557884810677499e-05, + "loss": 0.0244, + "step": 6195 + }, + { + "epoch": 0.6717259323503902, + "grad_norm": 1.2972661256790161, + "learning_rate": 1.5578122733207602e-05, + "loss": 0.0166, + "step": 6196 + }, + { + "epoch": 0.6718343451864701, + "grad_norm": 4.574367523193359, + "learning_rate": 1.5577397359640218e-05, + "loss": 0.0508, + "step": 6197 + }, + { + "epoch": 0.6719427580225499, + "grad_norm": 3.5747056007385254, + "learning_rate": 1.557667198607283e-05, + "loss": 0.0503, + "step": 6198 + }, + { + "epoch": 0.6720511708586296, + "grad_norm": 1.4010496139526367, + "learning_rate": 1.5575946612505442e-05, + "loss": 0.0159, + "step": 6199 + }, + { + "epoch": 0.6721595836947094, + "grad_norm": 2.9647276401519775, + "learning_rate": 1.5575221238938054e-05, + "loss": 0.0495, + "step": 6200 + }, + { + "epoch": 0.6722679965307893, + "grad_norm": 0.3339548707008362, + "learning_rate": 1.5574495865370666e-05, + "loss": 0.005, + "step": 6201 + }, + { + "epoch": 0.672376409366869, + "grad_norm": 2.9420390129089355, + "learning_rate": 1.5573770491803278e-05, + "loss": 0.0516, + "step": 6202 + }, + { + "epoch": 0.6724848222029488, + "grad_norm": 2.448759078979492, + "learning_rate": 1.5573045118235894e-05, + "loss": 0.0333, + "step": 6203 + }, + { + "epoch": 0.6725932350390286, + "grad_norm": 4.54311990737915, + "learning_rate": 1.5572319744668506e-05, + "loss": 0.0399, + "step": 6204 + }, + { + "epoch": 0.6727016478751084, + "grad_norm": 0.626893937587738, + "learning_rate": 1.5571594371101118e-05, + "loss": 0.0122, + "step": 6205 + }, + { + "epoch": 0.6728100607111882, + "grad_norm": 4.048682689666748, + "learning_rate": 1.557086899753373e-05, + "loss": 0.0727, + "step": 6206 + }, + { + "epoch": 0.672918473547268, + "grad_norm": 3.529951572418213, + "learning_rate": 1.5570143623966346e-05, + "loss": 0.0685, + "step": 6207 + }, + { + "epoch": 0.6730268863833477, + "grad_norm": 4.29930305480957, + "learning_rate": 1.5569418250398958e-05, + "loss": 0.0615, + "step": 6208 + }, + { + "epoch": 0.6731352992194276, + "grad_norm": 3.7008750438690186, + "learning_rate": 1.556869287683157e-05, + "loss": 0.0451, + "step": 6209 + }, + { + "epoch": 0.6732437120555074, + "grad_norm": 5.161681652069092, + "learning_rate": 1.5567967503264182e-05, + "loss": 0.0404, + "step": 6210 + }, + { + "epoch": 0.6733521248915871, + "grad_norm": 1.9257899522781372, + "learning_rate": 1.5567242129696794e-05, + "loss": 0.0398, + "step": 6211 + }, + { + "epoch": 0.673460537727667, + "grad_norm": 4.078176498413086, + "learning_rate": 1.556651675612941e-05, + "loss": 0.0247, + "step": 6212 + }, + { + "epoch": 0.6735689505637468, + "grad_norm": 6.4333977699279785, + "learning_rate": 1.5565791382562022e-05, + "loss": 0.0962, + "step": 6213 + }, + { + "epoch": 0.6736773633998265, + "grad_norm": 0.5778555274009705, + "learning_rate": 1.5565066008994634e-05, + "loss": 0.006, + "step": 6214 + }, + { + "epoch": 0.6737857762359063, + "grad_norm": 4.385668754577637, + "learning_rate": 1.5564340635427246e-05, + "loss": 0.0643, + "step": 6215 + }, + { + "epoch": 0.6738941890719862, + "grad_norm": 1.604062795639038, + "learning_rate": 1.556361526185986e-05, + "loss": 0.0137, + "step": 6216 + }, + { + "epoch": 0.6740026019080659, + "grad_norm": 2.7424094676971436, + "learning_rate": 1.556288988829247e-05, + "loss": 0.051, + "step": 6217 + }, + { + "epoch": 0.6741110147441457, + "grad_norm": 2.000102996826172, + "learning_rate": 1.5562164514725083e-05, + "loss": 0.0248, + "step": 6218 + }, + { + "epoch": 0.6742194275802255, + "grad_norm": 1.5428247451782227, + "learning_rate": 1.5561439141157698e-05, + "loss": 0.0196, + "step": 6219 + }, + { + "epoch": 0.6743278404163053, + "grad_norm": 3.1198983192443848, + "learning_rate": 1.556071376759031e-05, + "loss": 0.0395, + "step": 6220 + }, + { + "epoch": 0.6744362532523851, + "grad_norm": 3.8989596366882324, + "learning_rate": 1.5559988394022923e-05, + "loss": 0.0488, + "step": 6221 + }, + { + "epoch": 0.6745446660884649, + "grad_norm": 1.0541476011276245, + "learning_rate": 1.5559263020455535e-05, + "loss": 0.0165, + "step": 6222 + }, + { + "epoch": 0.6746530789245446, + "grad_norm": 4.6822991371154785, + "learning_rate": 1.5558537646888147e-05, + "loss": 0.0594, + "step": 6223 + }, + { + "epoch": 0.6747614917606245, + "grad_norm": 4.26092529296875, + "learning_rate": 1.555781227332076e-05, + "loss": 0.0307, + "step": 6224 + }, + { + "epoch": 0.6748699045967043, + "grad_norm": 4.605710506439209, + "learning_rate": 1.5557086899753375e-05, + "loss": 0.0792, + "step": 6225 + }, + { + "epoch": 0.674978317432784, + "grad_norm": 0.3015051782131195, + "learning_rate": 1.5556361526185987e-05, + "loss": 0.0047, + "step": 6226 + }, + { + "epoch": 0.6750867302688638, + "grad_norm": 4.103333950042725, + "learning_rate": 1.5555636152618602e-05, + "loss": 0.0703, + "step": 6227 + }, + { + "epoch": 0.6751951431049437, + "grad_norm": 3.077077865600586, + "learning_rate": 1.5554910779051214e-05, + "loss": 0.0277, + "step": 6228 + }, + { + "epoch": 0.6753035559410234, + "grad_norm": 2.3526065349578857, + "learning_rate": 1.5554185405483826e-05, + "loss": 0.0454, + "step": 6229 + }, + { + "epoch": 0.6754119687771032, + "grad_norm": 3.9989469051361084, + "learning_rate": 1.555346003191644e-05, + "loss": 0.0519, + "step": 6230 + }, + { + "epoch": 0.675520381613183, + "grad_norm": 5.455064296722412, + "learning_rate": 1.555273465834905e-05, + "loss": 0.0656, + "step": 6231 + }, + { + "epoch": 0.6756287944492628, + "grad_norm": 3.071298122406006, + "learning_rate": 1.5552009284781663e-05, + "loss": 0.0249, + "step": 6232 + }, + { + "epoch": 0.6757372072853426, + "grad_norm": 5.231883525848389, + "learning_rate": 1.5551283911214275e-05, + "loss": 0.063, + "step": 6233 + }, + { + "epoch": 0.6758456201214224, + "grad_norm": 7.210513114929199, + "learning_rate": 1.555055853764689e-05, + "loss": 0.0466, + "step": 6234 + }, + { + "epoch": 0.6759540329575021, + "grad_norm": 0.510771632194519, + "learning_rate": 1.5549833164079503e-05, + "loss": 0.0058, + "step": 6235 + }, + { + "epoch": 0.676062445793582, + "grad_norm": 3.310340404510498, + "learning_rate": 1.5549107790512115e-05, + "loss": 0.056, + "step": 6236 + }, + { + "epoch": 0.6761708586296618, + "grad_norm": 0.9947631359100342, + "learning_rate": 1.5548382416944727e-05, + "loss": 0.0093, + "step": 6237 + }, + { + "epoch": 0.6762792714657415, + "grad_norm": 1.9869073629379272, + "learning_rate": 1.554765704337734e-05, + "loss": 0.047, + "step": 6238 + }, + { + "epoch": 0.6763876843018214, + "grad_norm": 1.8737897872924805, + "learning_rate": 1.554693166980995e-05, + "loss": 0.0385, + "step": 6239 + }, + { + "epoch": 0.6764960971379012, + "grad_norm": 0.9022863507270813, + "learning_rate": 1.5546206296242567e-05, + "loss": 0.0089, + "step": 6240 + }, + { + "epoch": 0.6766045099739809, + "grad_norm": 3.3809752464294434, + "learning_rate": 1.554548092267518e-05, + "loss": 0.0557, + "step": 6241 + }, + { + "epoch": 0.6767129228100607, + "grad_norm": 6.936369895935059, + "learning_rate": 1.5544755549107794e-05, + "loss": 0.0424, + "step": 6242 + }, + { + "epoch": 0.6768213356461406, + "grad_norm": 3.703446388244629, + "learning_rate": 1.5544030175540407e-05, + "loss": 0.0245, + "step": 6243 + }, + { + "epoch": 0.6769297484822203, + "grad_norm": 4.240456581115723, + "learning_rate": 1.554330480197302e-05, + "loss": 0.0998, + "step": 6244 + }, + { + "epoch": 0.6770381613183001, + "grad_norm": 2.138967275619507, + "learning_rate": 1.554257942840563e-05, + "loss": 0.0197, + "step": 6245 + }, + { + "epoch": 0.6771465741543798, + "grad_norm": 2.0277092456817627, + "learning_rate": 1.5541854054838243e-05, + "loss": 0.0363, + "step": 6246 + }, + { + "epoch": 0.6772549869904597, + "grad_norm": 0.8539581894874573, + "learning_rate": 1.5541128681270855e-05, + "loss": 0.0235, + "step": 6247 + }, + { + "epoch": 0.6773633998265395, + "grad_norm": 1.3732014894485474, + "learning_rate": 1.5540403307703467e-05, + "loss": 0.0323, + "step": 6248 + }, + { + "epoch": 0.6774718126626192, + "grad_norm": 3.5867600440979004, + "learning_rate": 1.5539677934136083e-05, + "loss": 0.0888, + "step": 6249 + }, + { + "epoch": 0.677580225498699, + "grad_norm": 2.293099880218506, + "learning_rate": 1.5538952560568695e-05, + "loss": 0.069, + "step": 6250 + }, + { + "epoch": 0.6776886383347789, + "grad_norm": 3.2712230682373047, + "learning_rate": 1.5538227187001307e-05, + "loss": 0.073, + "step": 6251 + }, + { + "epoch": 0.6777970511708586, + "grad_norm": 1.2272636890411377, + "learning_rate": 1.553750181343392e-05, + "loss": 0.0485, + "step": 6252 + }, + { + "epoch": 0.6779054640069384, + "grad_norm": 1.4934968948364258, + "learning_rate": 1.553677643986653e-05, + "loss": 0.0401, + "step": 6253 + }, + { + "epoch": 0.6780138768430182, + "grad_norm": 2.123610019683838, + "learning_rate": 1.5536051066299144e-05, + "loss": 0.027, + "step": 6254 + }, + { + "epoch": 0.678122289679098, + "grad_norm": 2.648784637451172, + "learning_rate": 1.553532569273176e-05, + "loss": 0.0285, + "step": 6255 + }, + { + "epoch": 0.6782307025151778, + "grad_norm": 2.144819736480713, + "learning_rate": 1.553460031916437e-05, + "loss": 0.0201, + "step": 6256 + }, + { + "epoch": 0.6783391153512576, + "grad_norm": 4.209848403930664, + "learning_rate": 1.5533874945596983e-05, + "loss": 0.0665, + "step": 6257 + }, + { + "epoch": 0.6784475281873373, + "grad_norm": 3.2353641986846924, + "learning_rate": 1.5533149572029596e-05, + "loss": 0.0261, + "step": 6258 + }, + { + "epoch": 0.6785559410234172, + "grad_norm": 4.03487491607666, + "learning_rate": 1.553242419846221e-05, + "loss": 0.0933, + "step": 6259 + }, + { + "epoch": 0.678664353859497, + "grad_norm": 3.600639820098877, + "learning_rate": 1.5531698824894823e-05, + "loss": 0.0441, + "step": 6260 + }, + { + "epoch": 0.6787727666955767, + "grad_norm": 1.1568022966384888, + "learning_rate": 1.5530973451327435e-05, + "loss": 0.0176, + "step": 6261 + }, + { + "epoch": 0.6788811795316565, + "grad_norm": 3.79087495803833, + "learning_rate": 1.5530248077760048e-05, + "loss": 0.0872, + "step": 6262 + }, + { + "epoch": 0.6789895923677364, + "grad_norm": 1.744141697883606, + "learning_rate": 1.552952270419266e-05, + "loss": 0.0274, + "step": 6263 + }, + { + "epoch": 0.6790980052038161, + "grad_norm": 2.891627550125122, + "learning_rate": 1.5528797330625275e-05, + "loss": 0.0265, + "step": 6264 + }, + { + "epoch": 0.6792064180398959, + "grad_norm": 1.8148671388626099, + "learning_rate": 1.5528071957057887e-05, + "loss": 0.054, + "step": 6265 + }, + { + "epoch": 0.6793148308759757, + "grad_norm": 3.6383426189422607, + "learning_rate": 1.55273465834905e-05, + "loss": 0.0555, + "step": 6266 + }, + { + "epoch": 0.6794232437120555, + "grad_norm": 3.90104603767395, + "learning_rate": 1.552662120992311e-05, + "loss": 0.0393, + "step": 6267 + }, + { + "epoch": 0.6795316565481353, + "grad_norm": 2.061767101287842, + "learning_rate": 1.5525895836355724e-05, + "loss": 0.05, + "step": 6268 + }, + { + "epoch": 0.6796400693842151, + "grad_norm": 1.449669361114502, + "learning_rate": 1.5525170462788336e-05, + "loss": 0.028, + "step": 6269 + }, + { + "epoch": 0.6797484822202949, + "grad_norm": 0.9406036138534546, + "learning_rate": 1.5524445089220948e-05, + "loss": 0.0174, + "step": 6270 + }, + { + "epoch": 0.6798568950563747, + "grad_norm": 1.50026535987854, + "learning_rate": 1.5523719715653564e-05, + "loss": 0.0282, + "step": 6271 + }, + { + "epoch": 0.6799653078924545, + "grad_norm": 2.808455467224121, + "learning_rate": 1.5522994342086176e-05, + "loss": 0.0503, + "step": 6272 + }, + { + "epoch": 0.6800737207285342, + "grad_norm": 3.8675408363342285, + "learning_rate": 1.5522268968518788e-05, + "loss": 0.0491, + "step": 6273 + }, + { + "epoch": 0.6801821335646141, + "grad_norm": 2.8310635089874268, + "learning_rate": 1.55215435949514e-05, + "loss": 0.0429, + "step": 6274 + }, + { + "epoch": 0.6802905464006939, + "grad_norm": 2.528406858444214, + "learning_rate": 1.5520818221384012e-05, + "loss": 0.0485, + "step": 6275 + }, + { + "epoch": 0.6803989592367736, + "grad_norm": 0.3308431804180145, + "learning_rate": 1.5520092847816624e-05, + "loss": 0.0078, + "step": 6276 + }, + { + "epoch": 0.6805073720728534, + "grad_norm": 1.5018551349639893, + "learning_rate": 1.551936747424924e-05, + "loss": 0.0305, + "step": 6277 + }, + { + "epoch": 0.6806157849089333, + "grad_norm": 2.6546154022216797, + "learning_rate": 1.5518642100681852e-05, + "loss": 0.0688, + "step": 6278 + }, + { + "epoch": 0.680724197745013, + "grad_norm": 6.277023792266846, + "learning_rate": 1.5517916727114468e-05, + "loss": 0.0449, + "step": 6279 + }, + { + "epoch": 0.6808326105810928, + "grad_norm": 4.501676559448242, + "learning_rate": 1.551719135354708e-05, + "loss": 0.0381, + "step": 6280 + }, + { + "epoch": 0.6809410234171726, + "grad_norm": 1.3825165033340454, + "learning_rate": 1.5516465979979692e-05, + "loss": 0.0311, + "step": 6281 + }, + { + "epoch": 0.6810494362532524, + "grad_norm": 0.34504690766334534, + "learning_rate": 1.5515740606412304e-05, + "loss": 0.0065, + "step": 6282 + }, + { + "epoch": 0.6811578490893322, + "grad_norm": 3.9699230194091797, + "learning_rate": 1.5515015232844916e-05, + "loss": 0.0446, + "step": 6283 + }, + { + "epoch": 0.681266261925412, + "grad_norm": 2.3699915409088135, + "learning_rate": 1.5514289859277528e-05, + "loss": 0.0311, + "step": 6284 + }, + { + "epoch": 0.6813746747614917, + "grad_norm": 4.722332954406738, + "learning_rate": 1.551356448571014e-05, + "loss": 0.0521, + "step": 6285 + }, + { + "epoch": 0.6814830875975716, + "grad_norm": 4.43822717666626, + "learning_rate": 1.5512839112142756e-05, + "loss": 0.0878, + "step": 6286 + }, + { + "epoch": 0.6815915004336514, + "grad_norm": 3.699186086654663, + "learning_rate": 1.5512113738575368e-05, + "loss": 0.0434, + "step": 6287 + }, + { + "epoch": 0.6816999132697311, + "grad_norm": 1.5235775709152222, + "learning_rate": 1.551138836500798e-05, + "loss": 0.0363, + "step": 6288 + }, + { + "epoch": 0.6818083261058109, + "grad_norm": 5.086569786071777, + "learning_rate": 1.5510662991440592e-05, + "loss": 0.0351, + "step": 6289 + }, + { + "epoch": 0.6819167389418908, + "grad_norm": 1.7950913906097412, + "learning_rate": 1.5509937617873204e-05, + "loss": 0.0462, + "step": 6290 + }, + { + "epoch": 0.6820251517779705, + "grad_norm": 5.680915832519531, + "learning_rate": 1.5509212244305817e-05, + "loss": 0.0369, + "step": 6291 + }, + { + "epoch": 0.6821335646140503, + "grad_norm": 0.882776141166687, + "learning_rate": 1.5508486870738432e-05, + "loss": 0.0084, + "step": 6292 + }, + { + "epoch": 0.68224197745013, + "grad_norm": 1.895481824874878, + "learning_rate": 1.5507761497171044e-05, + "loss": 0.0187, + "step": 6293 + }, + { + "epoch": 0.6823503902862099, + "grad_norm": 3.0490944385528564, + "learning_rate": 1.550703612360366e-05, + "loss": 0.0608, + "step": 6294 + }, + { + "epoch": 0.6824588031222897, + "grad_norm": 2.5523619651794434, + "learning_rate": 1.5506310750036272e-05, + "loss": 0.0279, + "step": 6295 + }, + { + "epoch": 0.6825672159583694, + "grad_norm": 1.408063530921936, + "learning_rate": 1.5505585376468884e-05, + "loss": 0.0356, + "step": 6296 + }, + { + "epoch": 0.6826756287944492, + "grad_norm": 3.2262003421783447, + "learning_rate": 1.5504860002901496e-05, + "loss": 0.0394, + "step": 6297 + }, + { + "epoch": 0.6827840416305291, + "grad_norm": 3.6378114223480225, + "learning_rate": 1.550413462933411e-05, + "loss": 0.0626, + "step": 6298 + }, + { + "epoch": 0.6828924544666088, + "grad_norm": 0.5671706795692444, + "learning_rate": 1.550340925576672e-05, + "loss": 0.01, + "step": 6299 + }, + { + "epoch": 0.6830008673026886, + "grad_norm": 3.3542752265930176, + "learning_rate": 1.5502683882199333e-05, + "loss": 0.0516, + "step": 6300 + }, + { + "epoch": 0.6831092801387685, + "grad_norm": 6.36350679397583, + "learning_rate": 1.5501958508631948e-05, + "loss": 0.0786, + "step": 6301 + }, + { + "epoch": 0.6832176929748482, + "grad_norm": 4.636519432067871, + "learning_rate": 1.550123313506456e-05, + "loss": 0.0264, + "step": 6302 + }, + { + "epoch": 0.683326105810928, + "grad_norm": 2.4686813354492188, + "learning_rate": 1.5500507761497173e-05, + "loss": 0.0259, + "step": 6303 + }, + { + "epoch": 0.6834345186470078, + "grad_norm": 2.8188350200653076, + "learning_rate": 1.5499782387929785e-05, + "loss": 0.0426, + "step": 6304 + }, + { + "epoch": 0.6835429314830876, + "grad_norm": 4.059082508087158, + "learning_rate": 1.5499057014362397e-05, + "loss": 0.0507, + "step": 6305 + }, + { + "epoch": 0.6836513443191674, + "grad_norm": 2.437134265899658, + "learning_rate": 1.549833164079501e-05, + "loss": 0.0442, + "step": 6306 + }, + { + "epoch": 0.6837597571552472, + "grad_norm": 1.4342031478881836, + "learning_rate": 1.5497606267227624e-05, + "loss": 0.0093, + "step": 6307 + }, + { + "epoch": 0.6838681699913269, + "grad_norm": 6.611500263214111, + "learning_rate": 1.5496880893660237e-05, + "loss": 0.0316, + "step": 6308 + }, + { + "epoch": 0.6839765828274068, + "grad_norm": 2.8593430519104004, + "learning_rate": 1.549615552009285e-05, + "loss": 0.0505, + "step": 6309 + }, + { + "epoch": 0.6840849956634866, + "grad_norm": 4.426180362701416, + "learning_rate": 1.549543014652546e-05, + "loss": 0.0653, + "step": 6310 + }, + { + "epoch": 0.6841934084995663, + "grad_norm": 2.260258436203003, + "learning_rate": 1.5494704772958076e-05, + "loss": 0.0184, + "step": 6311 + }, + { + "epoch": 0.6843018213356461, + "grad_norm": 3.2711284160614014, + "learning_rate": 1.549397939939069e-05, + "loss": 0.0444, + "step": 6312 + }, + { + "epoch": 0.684410234171726, + "grad_norm": 1.9177112579345703, + "learning_rate": 1.54932540258233e-05, + "loss": 0.0095, + "step": 6313 + }, + { + "epoch": 0.6845186470078057, + "grad_norm": 5.299351692199707, + "learning_rate": 1.5492528652255913e-05, + "loss": 0.0319, + "step": 6314 + }, + { + "epoch": 0.6846270598438855, + "grad_norm": 12.781699180603027, + "learning_rate": 1.5491803278688525e-05, + "loss": 0.0735, + "step": 6315 + }, + { + "epoch": 0.6847354726799653, + "grad_norm": 1.5912996530532837, + "learning_rate": 1.549107790512114e-05, + "loss": 0.0374, + "step": 6316 + }, + { + "epoch": 0.6848438855160451, + "grad_norm": 2.942258596420288, + "learning_rate": 1.5490352531553753e-05, + "loss": 0.0337, + "step": 6317 + }, + { + "epoch": 0.6849522983521249, + "grad_norm": 1.2857569456100464, + "learning_rate": 1.5489627157986365e-05, + "loss": 0.0244, + "step": 6318 + }, + { + "epoch": 0.6850607111882047, + "grad_norm": 2.4029650688171387, + "learning_rate": 1.5488901784418977e-05, + "loss": 0.0327, + "step": 6319 + }, + { + "epoch": 0.6851691240242844, + "grad_norm": 4.897619247436523, + "learning_rate": 1.548817641085159e-05, + "loss": 0.0788, + "step": 6320 + }, + { + "epoch": 0.6852775368603643, + "grad_norm": 4.552427291870117, + "learning_rate": 1.54874510372842e-05, + "loss": 0.0601, + "step": 6321 + }, + { + "epoch": 0.6853859496964441, + "grad_norm": 2.76228666305542, + "learning_rate": 1.5486725663716813e-05, + "loss": 0.0397, + "step": 6322 + }, + { + "epoch": 0.6854943625325238, + "grad_norm": 1.9663809537887573, + "learning_rate": 1.548600029014943e-05, + "loss": 0.0293, + "step": 6323 + }, + { + "epoch": 0.6856027753686036, + "grad_norm": 2.187729597091675, + "learning_rate": 1.548527491658204e-05, + "loss": 0.0223, + "step": 6324 + }, + { + "epoch": 0.6857111882046835, + "grad_norm": 3.0012378692626953, + "learning_rate": 1.5484549543014653e-05, + "loss": 0.0418, + "step": 6325 + }, + { + "epoch": 0.6858196010407632, + "grad_norm": 2.427954912185669, + "learning_rate": 1.5483824169447265e-05, + "loss": 0.0361, + "step": 6326 + }, + { + "epoch": 0.685928013876843, + "grad_norm": 5.481561660766602, + "learning_rate": 1.5483098795879878e-05, + "loss": 0.0165, + "step": 6327 + }, + { + "epoch": 0.6860364267129229, + "grad_norm": 1.3604907989501953, + "learning_rate": 1.5482373422312493e-05, + "loss": 0.0118, + "step": 6328 + }, + { + "epoch": 0.6861448395490026, + "grad_norm": 2.0901262760162354, + "learning_rate": 1.5481648048745105e-05, + "loss": 0.0319, + "step": 6329 + }, + { + "epoch": 0.6862532523850824, + "grad_norm": 3.2930727005004883, + "learning_rate": 1.5480922675177717e-05, + "loss": 0.0656, + "step": 6330 + }, + { + "epoch": 0.6863616652211622, + "grad_norm": 2.0328338146209717, + "learning_rate": 1.5480197301610333e-05, + "loss": 0.0367, + "step": 6331 + }, + { + "epoch": 0.686470078057242, + "grad_norm": 1.6574598550796509, + "learning_rate": 1.5479471928042945e-05, + "loss": 0.0402, + "step": 6332 + }, + { + "epoch": 0.6865784908933218, + "grad_norm": 2.8468878269195557, + "learning_rate": 1.5478746554475557e-05, + "loss": 0.059, + "step": 6333 + }, + { + "epoch": 0.6866869037294016, + "grad_norm": 2.739346742630005, + "learning_rate": 1.547802118090817e-05, + "loss": 0.0619, + "step": 6334 + }, + { + "epoch": 0.6867953165654813, + "grad_norm": 2.1571006774902344, + "learning_rate": 1.547729580734078e-05, + "loss": 0.0251, + "step": 6335 + }, + { + "epoch": 0.6869037294015612, + "grad_norm": 3.5410611629486084, + "learning_rate": 1.5476570433773394e-05, + "loss": 0.0464, + "step": 6336 + }, + { + "epoch": 0.687012142237641, + "grad_norm": 1.8108218908309937, + "learning_rate": 1.5475845060206006e-05, + "loss": 0.0166, + "step": 6337 + }, + { + "epoch": 0.6871205550737207, + "grad_norm": 4.0381622314453125, + "learning_rate": 1.547511968663862e-05, + "loss": 0.0447, + "step": 6338 + }, + { + "epoch": 0.6872289679098005, + "grad_norm": 2.038372278213501, + "learning_rate": 1.5474394313071233e-05, + "loss": 0.0483, + "step": 6339 + }, + { + "epoch": 0.6873373807458804, + "grad_norm": 0.32052645087242126, + "learning_rate": 1.5473668939503846e-05, + "loss": 0.0029, + "step": 6340 + }, + { + "epoch": 0.6874457935819601, + "grad_norm": 0.38007333874702454, + "learning_rate": 1.5472943565936458e-05, + "loss": 0.0051, + "step": 6341 + }, + { + "epoch": 0.6875542064180399, + "grad_norm": 2.0959231853485107, + "learning_rate": 1.547221819236907e-05, + "loss": 0.0212, + "step": 6342 + }, + { + "epoch": 0.6876626192541196, + "grad_norm": 1.3118488788604736, + "learning_rate": 1.5471492818801682e-05, + "loss": 0.0278, + "step": 6343 + }, + { + "epoch": 0.6877710320901995, + "grad_norm": 2.9994208812713623, + "learning_rate": 1.5470767445234297e-05, + "loss": 0.0496, + "step": 6344 + }, + { + "epoch": 0.6878794449262793, + "grad_norm": 4.186224937438965, + "learning_rate": 1.547004207166691e-05, + "loss": 0.0426, + "step": 6345 + }, + { + "epoch": 0.687987857762359, + "grad_norm": 4.421842098236084, + "learning_rate": 1.5469316698099525e-05, + "loss": 0.0397, + "step": 6346 + }, + { + "epoch": 0.6880962705984388, + "grad_norm": 1.5111708641052246, + "learning_rate": 1.5468591324532137e-05, + "loss": 0.0273, + "step": 6347 + }, + { + "epoch": 0.6882046834345187, + "grad_norm": 3.2548367977142334, + "learning_rate": 1.546786595096475e-05, + "loss": 0.0594, + "step": 6348 + }, + { + "epoch": 0.6883130962705984, + "grad_norm": 1.6973379850387573, + "learning_rate": 1.546714057739736e-05, + "loss": 0.0225, + "step": 6349 + }, + { + "epoch": 0.6884215091066782, + "grad_norm": 2.483708143234253, + "learning_rate": 1.5466415203829974e-05, + "loss": 0.0503, + "step": 6350 + }, + { + "epoch": 0.688529921942758, + "grad_norm": 2.470940113067627, + "learning_rate": 1.5465689830262586e-05, + "loss": 0.0347, + "step": 6351 + }, + { + "epoch": 0.6886383347788378, + "grad_norm": 2.7770192623138428, + "learning_rate": 1.5464964456695198e-05, + "loss": 0.0254, + "step": 6352 + }, + { + "epoch": 0.6887467476149176, + "grad_norm": 0.7913181185722351, + "learning_rate": 1.5464239083127814e-05, + "loss": 0.0082, + "step": 6353 + }, + { + "epoch": 0.6888551604509974, + "grad_norm": 2.776520013809204, + "learning_rate": 1.5463513709560426e-05, + "loss": 0.0474, + "step": 6354 + }, + { + "epoch": 0.6889635732870771, + "grad_norm": 7.464134216308594, + "learning_rate": 1.5462788335993038e-05, + "loss": 0.0596, + "step": 6355 + }, + { + "epoch": 0.689071986123157, + "grad_norm": 4.341506004333496, + "learning_rate": 1.546206296242565e-05, + "loss": 0.0402, + "step": 6356 + }, + { + "epoch": 0.6891803989592368, + "grad_norm": 4.8221540451049805, + "learning_rate": 1.5461337588858262e-05, + "loss": 0.079, + "step": 6357 + }, + { + "epoch": 0.6892888117953165, + "grad_norm": 4.284844875335693, + "learning_rate": 1.5460612215290874e-05, + "loss": 0.07, + "step": 6358 + }, + { + "epoch": 0.6893972246313964, + "grad_norm": 2.5573158264160156, + "learning_rate": 1.545988684172349e-05, + "loss": 0.0496, + "step": 6359 + }, + { + "epoch": 0.6895056374674762, + "grad_norm": 4.753836154937744, + "learning_rate": 1.5459161468156102e-05, + "loss": 0.0354, + "step": 6360 + }, + { + "epoch": 0.6896140503035559, + "grad_norm": 2.698944330215454, + "learning_rate": 1.5458436094588714e-05, + "loss": 0.0277, + "step": 6361 + }, + { + "epoch": 0.6897224631396357, + "grad_norm": 2.003829002380371, + "learning_rate": 1.5457710721021326e-05, + "loss": 0.0241, + "step": 6362 + }, + { + "epoch": 0.6898308759757156, + "grad_norm": 2.485530138015747, + "learning_rate": 1.5456985347453942e-05, + "loss": 0.024, + "step": 6363 + }, + { + "epoch": 0.6899392888117953, + "grad_norm": 3.4626221656799316, + "learning_rate": 1.5456259973886554e-05, + "loss": 0.0309, + "step": 6364 + }, + { + "epoch": 0.6900477016478751, + "grad_norm": 2.898923873901367, + "learning_rate": 1.5455534600319166e-05, + "loss": 0.0611, + "step": 6365 + }, + { + "epoch": 0.6901561144839549, + "grad_norm": 1.3456519842147827, + "learning_rate": 1.5454809226751778e-05, + "loss": 0.0214, + "step": 6366 + }, + { + "epoch": 0.6902645273200347, + "grad_norm": 0.5316813588142395, + "learning_rate": 1.545408385318439e-05, + "loss": 0.007, + "step": 6367 + }, + { + "epoch": 0.6903729401561145, + "grad_norm": 7.076413631439209, + "learning_rate": 1.5453358479617006e-05, + "loss": 0.0742, + "step": 6368 + }, + { + "epoch": 0.6904813529921943, + "grad_norm": 1.1756319999694824, + "learning_rate": 1.5452633106049618e-05, + "loss": 0.0169, + "step": 6369 + }, + { + "epoch": 0.690589765828274, + "grad_norm": 2.203547477722168, + "learning_rate": 1.545190773248223e-05, + "loss": 0.0212, + "step": 6370 + }, + { + "epoch": 0.6906981786643539, + "grad_norm": 2.1387460231781006, + "learning_rate": 1.5451182358914842e-05, + "loss": 0.0305, + "step": 6371 + }, + { + "epoch": 0.6908065915004337, + "grad_norm": 3.637702226638794, + "learning_rate": 1.5450456985347454e-05, + "loss": 0.0534, + "step": 6372 + }, + { + "epoch": 0.6909150043365134, + "grad_norm": 2.9834089279174805, + "learning_rate": 1.5449731611780067e-05, + "loss": 0.0435, + "step": 6373 + }, + { + "epoch": 0.6910234171725932, + "grad_norm": 1.9951735734939575, + "learning_rate": 1.544900623821268e-05, + "loss": 0.0115, + "step": 6374 + }, + { + "epoch": 0.6911318300086731, + "grad_norm": 3.177149772644043, + "learning_rate": 1.5448280864645294e-05, + "loss": 0.0669, + "step": 6375 + }, + { + "epoch": 0.6912402428447528, + "grad_norm": 2.5218162536621094, + "learning_rate": 1.5447555491077906e-05, + "loss": 0.0177, + "step": 6376 + }, + { + "epoch": 0.6913486556808326, + "grad_norm": 5.896263122558594, + "learning_rate": 1.544683011751052e-05, + "loss": 0.0697, + "step": 6377 + }, + { + "epoch": 0.6914570685169124, + "grad_norm": 2.6857728958129883, + "learning_rate": 1.544610474394313e-05, + "loss": 0.0348, + "step": 6378 + }, + { + "epoch": 0.6915654813529922, + "grad_norm": 2.7657148838043213, + "learning_rate": 1.5445379370375743e-05, + "loss": 0.0411, + "step": 6379 + }, + { + "epoch": 0.691673894189072, + "grad_norm": 3.126093626022339, + "learning_rate": 1.544465399680836e-05, + "loss": 0.0465, + "step": 6380 + }, + { + "epoch": 0.6917823070251518, + "grad_norm": 1.6224839687347412, + "learning_rate": 1.544392862324097e-05, + "loss": 0.0155, + "step": 6381 + }, + { + "epoch": 0.6918907198612315, + "grad_norm": 3.7230312824249268, + "learning_rate": 1.5443203249673583e-05, + "loss": 0.0557, + "step": 6382 + }, + { + "epoch": 0.6919991326973114, + "grad_norm": 0.9597880244255066, + "learning_rate": 1.5442477876106198e-05, + "loss": 0.0146, + "step": 6383 + }, + { + "epoch": 0.6921075455333912, + "grad_norm": 2.724897861480713, + "learning_rate": 1.544175250253881e-05, + "loss": 0.0447, + "step": 6384 + }, + { + "epoch": 0.6922159583694709, + "grad_norm": 11.25571060180664, + "learning_rate": 1.5441027128971422e-05, + "loss": 0.0643, + "step": 6385 + }, + { + "epoch": 0.6923243712055508, + "grad_norm": 2.020810604095459, + "learning_rate": 1.5440301755404035e-05, + "loss": 0.0115, + "step": 6386 + }, + { + "epoch": 0.6924327840416306, + "grad_norm": 1.0651839971542358, + "learning_rate": 1.5439576381836647e-05, + "loss": 0.0124, + "step": 6387 + }, + { + "epoch": 0.6925411968777103, + "grad_norm": 3.160055637359619, + "learning_rate": 1.543885100826926e-05, + "loss": 0.0479, + "step": 6388 + }, + { + "epoch": 0.6926496097137901, + "grad_norm": 3.8262016773223877, + "learning_rate": 1.543812563470187e-05, + "loss": 0.018, + "step": 6389 + }, + { + "epoch": 0.69275802254987, + "grad_norm": 5.114297866821289, + "learning_rate": 1.5437400261134487e-05, + "loss": 0.0746, + "step": 6390 + }, + { + "epoch": 0.6928664353859497, + "grad_norm": 3.541667938232422, + "learning_rate": 1.54366748875671e-05, + "loss": 0.0366, + "step": 6391 + }, + { + "epoch": 0.6929748482220295, + "grad_norm": 5.478541851043701, + "learning_rate": 1.543594951399971e-05, + "loss": 0.0564, + "step": 6392 + }, + { + "epoch": 0.6930832610581092, + "grad_norm": 5.102950572967529, + "learning_rate": 1.5435224140432323e-05, + "loss": 0.0402, + "step": 6393 + }, + { + "epoch": 0.6931916738941891, + "grad_norm": 5.97018575668335, + "learning_rate": 1.5434498766864935e-05, + "loss": 0.0308, + "step": 6394 + }, + { + "epoch": 0.6933000867302689, + "grad_norm": 3.3901612758636475, + "learning_rate": 1.5433773393297547e-05, + "loss": 0.0208, + "step": 6395 + }, + { + "epoch": 0.6934084995663486, + "grad_norm": 2.0662736892700195, + "learning_rate": 1.5433048019730163e-05, + "loss": 0.0453, + "step": 6396 + }, + { + "epoch": 0.6935169124024284, + "grad_norm": 2.4566352367401123, + "learning_rate": 1.5432322646162775e-05, + "loss": 0.0543, + "step": 6397 + }, + { + "epoch": 0.6936253252385083, + "grad_norm": 0.4403768479824066, + "learning_rate": 1.543159727259539e-05, + "loss": 0.0036, + "step": 6398 + }, + { + "epoch": 0.693733738074588, + "grad_norm": 2.65407133102417, + "learning_rate": 1.5430871899028003e-05, + "loss": 0.0193, + "step": 6399 + }, + { + "epoch": 0.6938421509106678, + "grad_norm": 2.3509576320648193, + "learning_rate": 1.5430146525460615e-05, + "loss": 0.021, + "step": 6400 + }, + { + "epoch": 0.6939505637467476, + "grad_norm": 1.4500610828399658, + "learning_rate": 1.5429421151893227e-05, + "loss": 0.0144, + "step": 6401 + }, + { + "epoch": 0.6940589765828274, + "grad_norm": 5.836851119995117, + "learning_rate": 1.542869577832584e-05, + "loss": 0.0469, + "step": 6402 + }, + { + "epoch": 0.6941673894189072, + "grad_norm": 7.295343399047852, + "learning_rate": 1.542797040475845e-05, + "loss": 0.1035, + "step": 6403 + }, + { + "epoch": 0.694275802254987, + "grad_norm": 4.12368106842041, + "learning_rate": 1.5427245031191063e-05, + "loss": 0.0913, + "step": 6404 + }, + { + "epoch": 0.6943842150910667, + "grad_norm": 6.602219104766846, + "learning_rate": 1.542651965762368e-05, + "loss": 0.0519, + "step": 6405 + }, + { + "epoch": 0.6944926279271466, + "grad_norm": 2.260152816772461, + "learning_rate": 1.542579428405629e-05, + "loss": 0.0286, + "step": 6406 + }, + { + "epoch": 0.6946010407632264, + "grad_norm": 5.169588565826416, + "learning_rate": 1.5425068910488903e-05, + "loss": 0.0686, + "step": 6407 + }, + { + "epoch": 0.6947094535993061, + "grad_norm": 1.9117414951324463, + "learning_rate": 1.5424343536921515e-05, + "loss": 0.0307, + "step": 6408 + }, + { + "epoch": 0.6948178664353859, + "grad_norm": 2.0360512733459473, + "learning_rate": 1.5423618163354127e-05, + "loss": 0.025, + "step": 6409 + }, + { + "epoch": 0.6949262792714658, + "grad_norm": 5.7014641761779785, + "learning_rate": 1.542289278978674e-05, + "loss": 0.051, + "step": 6410 + }, + { + "epoch": 0.6950346921075455, + "grad_norm": 7.049473762512207, + "learning_rate": 1.5422167416219355e-05, + "loss": 0.0508, + "step": 6411 + }, + { + "epoch": 0.6951431049436253, + "grad_norm": 2.755202531814575, + "learning_rate": 1.5421442042651967e-05, + "loss": 0.0527, + "step": 6412 + }, + { + "epoch": 0.6952515177797051, + "grad_norm": 3.889899253845215, + "learning_rate": 1.542071666908458e-05, + "loss": 0.0823, + "step": 6413 + }, + { + "epoch": 0.6953599306157849, + "grad_norm": 3.0134754180908203, + "learning_rate": 1.541999129551719e-05, + "loss": 0.0613, + "step": 6414 + }, + { + "epoch": 0.6954683434518647, + "grad_norm": 2.4231948852539062, + "learning_rate": 1.5419265921949807e-05, + "loss": 0.0253, + "step": 6415 + }, + { + "epoch": 0.6955767562879445, + "grad_norm": 3.124208688735962, + "learning_rate": 1.541854054838242e-05, + "loss": 0.0233, + "step": 6416 + }, + { + "epoch": 0.6956851691240243, + "grad_norm": 4.697427749633789, + "learning_rate": 1.541781517481503e-05, + "loss": 0.0418, + "step": 6417 + }, + { + "epoch": 0.6957935819601041, + "grad_norm": 4.476807594299316, + "learning_rate": 1.5417089801247644e-05, + "loss": 0.0582, + "step": 6418 + }, + { + "epoch": 0.6959019947961839, + "grad_norm": 1.807517647743225, + "learning_rate": 1.5416364427680256e-05, + "loss": 0.0189, + "step": 6419 + }, + { + "epoch": 0.6960104076322636, + "grad_norm": 4.702230453491211, + "learning_rate": 1.541563905411287e-05, + "loss": 0.0418, + "step": 6420 + }, + { + "epoch": 0.6961188204683435, + "grad_norm": 1.617247223854065, + "learning_rate": 1.5414913680545483e-05, + "loss": 0.0291, + "step": 6421 + }, + { + "epoch": 0.6962272333044233, + "grad_norm": 4.01574182510376, + "learning_rate": 1.5414188306978096e-05, + "loss": 0.0228, + "step": 6422 + }, + { + "epoch": 0.696335646140503, + "grad_norm": 3.1310336589813232, + "learning_rate": 1.5413462933410708e-05, + "loss": 0.0508, + "step": 6423 + }, + { + "epoch": 0.6964440589765828, + "grad_norm": 0.8311132192611694, + "learning_rate": 1.541273755984332e-05, + "loss": 0.0161, + "step": 6424 + }, + { + "epoch": 0.6965524718126627, + "grad_norm": 2.056442975997925, + "learning_rate": 1.5412012186275932e-05, + "loss": 0.0402, + "step": 6425 + }, + { + "epoch": 0.6966608846487424, + "grad_norm": 2.8880255222320557, + "learning_rate": 1.5411286812708544e-05, + "loss": 0.0508, + "step": 6426 + }, + { + "epoch": 0.6967692974848222, + "grad_norm": 0.8567746877670288, + "learning_rate": 1.541056143914116e-05, + "loss": 0.0047, + "step": 6427 + }, + { + "epoch": 0.696877710320902, + "grad_norm": 1.6281254291534424, + "learning_rate": 1.5409836065573772e-05, + "loss": 0.0278, + "step": 6428 + }, + { + "epoch": 0.6969861231569818, + "grad_norm": 1.4675229787826538, + "learning_rate": 1.5409110692006384e-05, + "loss": 0.0364, + "step": 6429 + }, + { + "epoch": 0.6970945359930616, + "grad_norm": 4.170100688934326, + "learning_rate": 1.5408385318438996e-05, + "loss": 0.0214, + "step": 6430 + }, + { + "epoch": 0.6972029488291414, + "grad_norm": 1.2465028762817383, + "learning_rate": 1.5407659944871608e-05, + "loss": 0.0122, + "step": 6431 + }, + { + "epoch": 0.6973113616652211, + "grad_norm": 1.2393341064453125, + "learning_rate": 1.5406934571304224e-05, + "loss": 0.0106, + "step": 6432 + }, + { + "epoch": 0.697419774501301, + "grad_norm": 2.796245813369751, + "learning_rate": 1.5406209197736836e-05, + "loss": 0.035, + "step": 6433 + }, + { + "epoch": 0.6975281873373808, + "grad_norm": 2.0033295154571533, + "learning_rate": 1.5405483824169448e-05, + "loss": 0.0214, + "step": 6434 + }, + { + "epoch": 0.6976366001734605, + "grad_norm": 2.208268880844116, + "learning_rate": 1.5404758450602064e-05, + "loss": 0.0235, + "step": 6435 + }, + { + "epoch": 0.6977450130095403, + "grad_norm": 3.134169340133667, + "learning_rate": 1.5404033077034676e-05, + "loss": 0.0509, + "step": 6436 + }, + { + "epoch": 0.6978534258456202, + "grad_norm": 7.11651611328125, + "learning_rate": 1.5403307703467288e-05, + "loss": 0.0749, + "step": 6437 + }, + { + "epoch": 0.6979618386816999, + "grad_norm": 3.9388270378112793, + "learning_rate": 1.54025823298999e-05, + "loss": 0.0377, + "step": 6438 + }, + { + "epoch": 0.6980702515177797, + "grad_norm": 3.3120436668395996, + "learning_rate": 1.5401856956332512e-05, + "loss": 0.0212, + "step": 6439 + }, + { + "epoch": 0.6981786643538594, + "grad_norm": 1.0302274227142334, + "learning_rate": 1.5401131582765124e-05, + "loss": 0.0263, + "step": 6440 + }, + { + "epoch": 0.6982870771899393, + "grad_norm": 1.3705387115478516, + "learning_rate": 1.5400406209197736e-05, + "loss": 0.0368, + "step": 6441 + }, + { + "epoch": 0.6983954900260191, + "grad_norm": 2.053124189376831, + "learning_rate": 1.5399680835630352e-05, + "loss": 0.0294, + "step": 6442 + }, + { + "epoch": 0.6985039028620988, + "grad_norm": 3.497880458831787, + "learning_rate": 1.5398955462062964e-05, + "loss": 0.1196, + "step": 6443 + }, + { + "epoch": 0.6986123156981786, + "grad_norm": 0.7757773399353027, + "learning_rate": 1.5398230088495576e-05, + "loss": 0.0077, + "step": 6444 + }, + { + "epoch": 0.6987207285342585, + "grad_norm": 2.2241458892822266, + "learning_rate": 1.539750471492819e-05, + "loss": 0.0471, + "step": 6445 + }, + { + "epoch": 0.6988291413703382, + "grad_norm": 3.074766159057617, + "learning_rate": 1.53967793413608e-05, + "loss": 0.0346, + "step": 6446 + }, + { + "epoch": 0.698937554206418, + "grad_norm": 6.311899185180664, + "learning_rate": 1.5396053967793413e-05, + "loss": 0.1064, + "step": 6447 + }, + { + "epoch": 0.6990459670424979, + "grad_norm": 0.7884364128112793, + "learning_rate": 1.5395328594226028e-05, + "loss": 0.0159, + "step": 6448 + }, + { + "epoch": 0.6991543798785776, + "grad_norm": 6.1330742835998535, + "learning_rate": 1.539460322065864e-05, + "loss": 0.0616, + "step": 6449 + }, + { + "epoch": 0.6992627927146574, + "grad_norm": 4.996412754058838, + "learning_rate": 1.5393877847091256e-05, + "loss": 0.0739, + "step": 6450 + }, + { + "epoch": 0.6993712055507372, + "grad_norm": 2.131394386291504, + "learning_rate": 1.5393152473523868e-05, + "loss": 0.0187, + "step": 6451 + }, + { + "epoch": 0.699479618386817, + "grad_norm": 2.089310884475708, + "learning_rate": 1.539242709995648e-05, + "loss": 0.0405, + "step": 6452 + }, + { + "epoch": 0.6995880312228968, + "grad_norm": 4.987019062042236, + "learning_rate": 1.5391701726389092e-05, + "loss": 0.0735, + "step": 6453 + }, + { + "epoch": 0.6996964440589766, + "grad_norm": 1.8639267683029175, + "learning_rate": 1.5390976352821704e-05, + "loss": 0.0184, + "step": 6454 + }, + { + "epoch": 0.6998048568950563, + "grad_norm": 3.687720537185669, + "learning_rate": 1.5390250979254317e-05, + "loss": 0.0472, + "step": 6455 + }, + { + "epoch": 0.6999132697311362, + "grad_norm": 2.8101744651794434, + "learning_rate": 1.538952560568693e-05, + "loss": 0.0423, + "step": 6456 + }, + { + "epoch": 0.700021682567216, + "grad_norm": 3.79937481880188, + "learning_rate": 1.5388800232119544e-05, + "loss": 0.044, + "step": 6457 + }, + { + "epoch": 0.7001300954032957, + "grad_norm": 2.4258780479431152, + "learning_rate": 1.5388074858552156e-05, + "loss": 0.0627, + "step": 6458 + }, + { + "epoch": 0.7002385082393755, + "grad_norm": 6.351016998291016, + "learning_rate": 1.538734948498477e-05, + "loss": 0.063, + "step": 6459 + }, + { + "epoch": 0.7003469210754554, + "grad_norm": 2.1328554153442383, + "learning_rate": 1.538662411141738e-05, + "loss": 0.0329, + "step": 6460 + }, + { + "epoch": 0.7004553339115351, + "grad_norm": 1.827115535736084, + "learning_rate": 1.5385898737849993e-05, + "loss": 0.037, + "step": 6461 + }, + { + "epoch": 0.7005637467476149, + "grad_norm": 5.968873500823975, + "learning_rate": 1.5385173364282605e-05, + "loss": 0.0763, + "step": 6462 + }, + { + "epoch": 0.7006721595836947, + "grad_norm": 2.5282320976257324, + "learning_rate": 1.538444799071522e-05, + "loss": 0.0399, + "step": 6463 + }, + { + "epoch": 0.7007805724197745, + "grad_norm": 2.2455461025238037, + "learning_rate": 1.5383722617147833e-05, + "loss": 0.0639, + "step": 6464 + }, + { + "epoch": 0.7008889852558543, + "grad_norm": 2.53737735748291, + "learning_rate": 1.5382997243580445e-05, + "loss": 0.0505, + "step": 6465 + }, + { + "epoch": 0.7009973980919341, + "grad_norm": 1.1773567199707031, + "learning_rate": 1.5382271870013057e-05, + "loss": 0.0172, + "step": 6466 + }, + { + "epoch": 0.7011058109280138, + "grad_norm": 3.1349217891693115, + "learning_rate": 1.5381546496445672e-05, + "loss": 0.0242, + "step": 6467 + }, + { + "epoch": 0.7012142237640937, + "grad_norm": 4.189875602722168, + "learning_rate": 1.5380821122878285e-05, + "loss": 0.0346, + "step": 6468 + }, + { + "epoch": 0.7013226366001735, + "grad_norm": 0.9441370368003845, + "learning_rate": 1.5380095749310897e-05, + "loss": 0.023, + "step": 6469 + }, + { + "epoch": 0.7014310494362532, + "grad_norm": 3.730008602142334, + "learning_rate": 1.537937037574351e-05, + "loss": 0.0374, + "step": 6470 + }, + { + "epoch": 0.701539462272333, + "grad_norm": 0.8124465346336365, + "learning_rate": 1.537864500217612e-05, + "loss": 0.0118, + "step": 6471 + }, + { + "epoch": 0.7016478751084129, + "grad_norm": 8.319415092468262, + "learning_rate": 1.5377919628608737e-05, + "loss": 0.0192, + "step": 6472 + }, + { + "epoch": 0.7017562879444926, + "grad_norm": 2.7829153537750244, + "learning_rate": 1.537719425504135e-05, + "loss": 0.0564, + "step": 6473 + }, + { + "epoch": 0.7018647007805724, + "grad_norm": 2.6659302711486816, + "learning_rate": 1.537646888147396e-05, + "loss": 0.0538, + "step": 6474 + }, + { + "epoch": 0.7019731136166523, + "grad_norm": 4.468470573425293, + "learning_rate": 1.5375743507906573e-05, + "loss": 0.0306, + "step": 6475 + }, + { + "epoch": 0.702081526452732, + "grad_norm": 1.6269813776016235, + "learning_rate": 1.5375018134339185e-05, + "loss": 0.0241, + "step": 6476 + }, + { + "epoch": 0.7021899392888118, + "grad_norm": 1.0911351442337036, + "learning_rate": 1.5374292760771797e-05, + "loss": 0.024, + "step": 6477 + }, + { + "epoch": 0.7022983521248916, + "grad_norm": 2.3917829990386963, + "learning_rate": 1.537356738720441e-05, + "loss": 0.0334, + "step": 6478 + }, + { + "epoch": 0.7024067649609714, + "grad_norm": 1.942299246788025, + "learning_rate": 1.5372842013637025e-05, + "loss": 0.0218, + "step": 6479 + }, + { + "epoch": 0.7025151777970512, + "grad_norm": 3.9034597873687744, + "learning_rate": 1.5372116640069637e-05, + "loss": 0.0438, + "step": 6480 + }, + { + "epoch": 0.702623590633131, + "grad_norm": 2.052725315093994, + "learning_rate": 1.537139126650225e-05, + "loss": 0.0329, + "step": 6481 + }, + { + "epoch": 0.7027320034692107, + "grad_norm": 3.3101577758789062, + "learning_rate": 1.537066589293486e-05, + "loss": 0.0342, + "step": 6482 + }, + { + "epoch": 0.7028404163052906, + "grad_norm": 2.954514265060425, + "learning_rate": 1.5369940519367474e-05, + "loss": 0.0191, + "step": 6483 + }, + { + "epoch": 0.7029488291413704, + "grad_norm": 2.7095117568969727, + "learning_rate": 1.536921514580009e-05, + "loss": 0.0226, + "step": 6484 + }, + { + "epoch": 0.7030572419774501, + "grad_norm": 0.6893970370292664, + "learning_rate": 1.53684897722327e-05, + "loss": 0.0066, + "step": 6485 + }, + { + "epoch": 0.7031656548135299, + "grad_norm": 7.830722808837891, + "learning_rate": 1.5367764398665313e-05, + "loss": 0.0534, + "step": 6486 + }, + { + "epoch": 0.7032740676496098, + "grad_norm": 2.228497266769409, + "learning_rate": 1.536703902509793e-05, + "loss": 0.0411, + "step": 6487 + }, + { + "epoch": 0.7033824804856895, + "grad_norm": 1.879757285118103, + "learning_rate": 1.536631365153054e-05, + "loss": 0.0294, + "step": 6488 + }, + { + "epoch": 0.7034908933217693, + "grad_norm": 4.440798282623291, + "learning_rate": 1.5365588277963153e-05, + "loss": 0.0896, + "step": 6489 + }, + { + "epoch": 0.703599306157849, + "grad_norm": 1.0484944581985474, + "learning_rate": 1.5364862904395765e-05, + "loss": 0.0194, + "step": 6490 + }, + { + "epoch": 0.7037077189939289, + "grad_norm": 1.9232429265975952, + "learning_rate": 1.5364137530828377e-05, + "loss": 0.0456, + "step": 6491 + }, + { + "epoch": 0.7038161318300087, + "grad_norm": 3.750790596008301, + "learning_rate": 1.536341215726099e-05, + "loss": 0.0439, + "step": 6492 + }, + { + "epoch": 0.7039245446660884, + "grad_norm": 2.125175952911377, + "learning_rate": 1.5362686783693602e-05, + "loss": 0.0176, + "step": 6493 + }, + { + "epoch": 0.7040329575021682, + "grad_norm": 0.7633838653564453, + "learning_rate": 1.5361961410126217e-05, + "loss": 0.0294, + "step": 6494 + }, + { + "epoch": 0.7041413703382481, + "grad_norm": 1.5804599523544312, + "learning_rate": 1.536123603655883e-05, + "loss": 0.0217, + "step": 6495 + }, + { + "epoch": 0.7042497831743278, + "grad_norm": 6.11769437789917, + "learning_rate": 1.536051066299144e-05, + "loss": 0.0152, + "step": 6496 + }, + { + "epoch": 0.7043581960104076, + "grad_norm": 5.638584136962891, + "learning_rate": 1.5359785289424054e-05, + "loss": 0.0385, + "step": 6497 + }, + { + "epoch": 0.7044666088464874, + "grad_norm": 0.7246368527412415, + "learning_rate": 1.5359059915856666e-05, + "loss": 0.0053, + "step": 6498 + }, + { + "epoch": 0.7045750216825672, + "grad_norm": 4.545156002044678, + "learning_rate": 1.5358334542289278e-05, + "loss": 0.0622, + "step": 6499 + }, + { + "epoch": 0.704683434518647, + "grad_norm": 4.518915176391602, + "learning_rate": 1.5357609168721894e-05, + "loss": 0.0262, + "step": 6500 + }, + { + "epoch": 0.7047918473547268, + "grad_norm": 2.6193573474884033, + "learning_rate": 1.5356883795154506e-05, + "loss": 0.0302, + "step": 6501 + }, + { + "epoch": 0.7049002601908065, + "grad_norm": 4.596084117889404, + "learning_rate": 1.535615842158712e-05, + "loss": 0.0624, + "step": 6502 + }, + { + "epoch": 0.7050086730268864, + "grad_norm": 1.8275245428085327, + "learning_rate": 1.5355433048019733e-05, + "loss": 0.0093, + "step": 6503 + }, + { + "epoch": 0.7051170858629662, + "grad_norm": 3.278695583343506, + "learning_rate": 1.5354707674452345e-05, + "loss": 0.0665, + "step": 6504 + }, + { + "epoch": 0.7052254986990459, + "grad_norm": 4.21748161315918, + "learning_rate": 1.5353982300884958e-05, + "loss": 0.0421, + "step": 6505 + }, + { + "epoch": 0.7053339115351258, + "grad_norm": 2.0527658462524414, + "learning_rate": 1.535325692731757e-05, + "loss": 0.0361, + "step": 6506 + }, + { + "epoch": 0.7054423243712056, + "grad_norm": 1.9034656286239624, + "learning_rate": 1.5352531553750182e-05, + "loss": 0.0481, + "step": 6507 + }, + { + "epoch": 0.7055507372072853, + "grad_norm": 4.518468856811523, + "learning_rate": 1.5351806180182794e-05, + "loss": 0.0639, + "step": 6508 + }, + { + "epoch": 0.7056591500433651, + "grad_norm": 1.5792386531829834, + "learning_rate": 1.535108080661541e-05, + "loss": 0.0383, + "step": 6509 + }, + { + "epoch": 0.705767562879445, + "grad_norm": 5.025452613830566, + "learning_rate": 1.5350355433048022e-05, + "loss": 0.0504, + "step": 6510 + }, + { + "epoch": 0.7058759757155247, + "grad_norm": 2.738581657409668, + "learning_rate": 1.5349630059480634e-05, + "loss": 0.0249, + "step": 6511 + }, + { + "epoch": 0.7059843885516045, + "grad_norm": 3.1015207767486572, + "learning_rate": 1.5348904685913246e-05, + "loss": 0.0291, + "step": 6512 + }, + { + "epoch": 0.7060928013876843, + "grad_norm": 4.954853534698486, + "learning_rate": 1.5348179312345858e-05, + "loss": 0.0375, + "step": 6513 + }, + { + "epoch": 0.7062012142237641, + "grad_norm": 5.310886859893799, + "learning_rate": 1.534745393877847e-05, + "loss": 0.0803, + "step": 6514 + }, + { + "epoch": 0.7063096270598439, + "grad_norm": 0.8987859487533569, + "learning_rate": 1.5346728565211086e-05, + "loss": 0.0153, + "step": 6515 + }, + { + "epoch": 0.7064180398959237, + "grad_norm": 2.5392556190490723, + "learning_rate": 1.5346003191643698e-05, + "loss": 0.0185, + "step": 6516 + }, + { + "epoch": 0.7065264527320034, + "grad_norm": 2.3338711261749268, + "learning_rate": 1.534527781807631e-05, + "loss": 0.0569, + "step": 6517 + }, + { + "epoch": 0.7066348655680833, + "grad_norm": 2.1921908855438232, + "learning_rate": 1.5344552444508926e-05, + "loss": 0.048, + "step": 6518 + }, + { + "epoch": 0.7067432784041631, + "grad_norm": 6.1394453048706055, + "learning_rate": 1.5343827070941538e-05, + "loss": 0.0735, + "step": 6519 + }, + { + "epoch": 0.7068516912402428, + "grad_norm": 3.6298296451568604, + "learning_rate": 1.534310169737415e-05, + "loss": 0.1054, + "step": 6520 + }, + { + "epoch": 0.7069601040763226, + "grad_norm": 1.960733413696289, + "learning_rate": 1.5342376323806762e-05, + "loss": 0.0708, + "step": 6521 + }, + { + "epoch": 0.7070685169124025, + "grad_norm": 2.186167001724243, + "learning_rate": 1.5341650950239374e-05, + "loss": 0.0442, + "step": 6522 + }, + { + "epoch": 0.7071769297484822, + "grad_norm": 4.310095310211182, + "learning_rate": 1.5340925576671986e-05, + "loss": 0.0412, + "step": 6523 + }, + { + "epoch": 0.707285342584562, + "grad_norm": 3.024230718612671, + "learning_rate": 1.5340200203104602e-05, + "loss": 0.0536, + "step": 6524 + }, + { + "epoch": 0.7073937554206418, + "grad_norm": 3.5753350257873535, + "learning_rate": 1.5339474829537214e-05, + "loss": 0.0351, + "step": 6525 + }, + { + "epoch": 0.7075021682567216, + "grad_norm": 1.8419512510299683, + "learning_rate": 1.5338749455969826e-05, + "loss": 0.0386, + "step": 6526 + }, + { + "epoch": 0.7076105810928014, + "grad_norm": 0.9958664178848267, + "learning_rate": 1.533802408240244e-05, + "loss": 0.015, + "step": 6527 + }, + { + "epoch": 0.7077189939288812, + "grad_norm": 1.8120057582855225, + "learning_rate": 1.533729870883505e-05, + "loss": 0.0584, + "step": 6528 + }, + { + "epoch": 0.7078274067649609, + "grad_norm": 1.0673595666885376, + "learning_rate": 1.5336573335267663e-05, + "loss": 0.0188, + "step": 6529 + }, + { + "epoch": 0.7079358196010408, + "grad_norm": 2.013046979904175, + "learning_rate": 1.5335847961700278e-05, + "loss": 0.0307, + "step": 6530 + }, + { + "epoch": 0.7080442324371206, + "grad_norm": 0.7654993534088135, + "learning_rate": 1.533512258813289e-05, + "loss": 0.0141, + "step": 6531 + }, + { + "epoch": 0.7081526452732003, + "grad_norm": 2.8744914531707764, + "learning_rate": 1.5334397214565502e-05, + "loss": 0.0414, + "step": 6532 + }, + { + "epoch": 0.7082610581092801, + "grad_norm": 2.7884371280670166, + "learning_rate": 1.5333671840998115e-05, + "loss": 0.0416, + "step": 6533 + }, + { + "epoch": 0.70836947094536, + "grad_norm": 3.2250030040740967, + "learning_rate": 1.5332946467430727e-05, + "loss": 0.0389, + "step": 6534 + }, + { + "epoch": 0.7084778837814397, + "grad_norm": 2.2920947074890137, + "learning_rate": 1.533222109386334e-05, + "loss": 0.0436, + "step": 6535 + }, + { + "epoch": 0.7085862966175195, + "grad_norm": 3.3901572227478027, + "learning_rate": 1.5331495720295954e-05, + "loss": 0.045, + "step": 6536 + }, + { + "epoch": 0.7086947094535994, + "grad_norm": 3.339574098587036, + "learning_rate": 1.5330770346728567e-05, + "loss": 0.0629, + "step": 6537 + }, + { + "epoch": 0.7088031222896791, + "grad_norm": 2.050985097885132, + "learning_rate": 1.533004497316118e-05, + "loss": 0.0283, + "step": 6538 + }, + { + "epoch": 0.7089115351257589, + "grad_norm": 1.8399168252944946, + "learning_rate": 1.5329319599593794e-05, + "loss": 0.0168, + "step": 6539 + }, + { + "epoch": 0.7090199479618386, + "grad_norm": 7.136024475097656, + "learning_rate": 1.5328594226026406e-05, + "loss": 0.0454, + "step": 6540 + }, + { + "epoch": 0.7091283607979185, + "grad_norm": 1.7318239212036133, + "learning_rate": 1.532786885245902e-05, + "loss": 0.0441, + "step": 6541 + }, + { + "epoch": 0.7092367736339983, + "grad_norm": 2.606956720352173, + "learning_rate": 1.532714347889163e-05, + "loss": 0.0543, + "step": 6542 + }, + { + "epoch": 0.709345186470078, + "grad_norm": 1.711799144744873, + "learning_rate": 1.5326418105324243e-05, + "loss": 0.0263, + "step": 6543 + }, + { + "epoch": 0.7094535993061578, + "grad_norm": 3.62369966506958, + "learning_rate": 1.5325692731756855e-05, + "loss": 0.0334, + "step": 6544 + }, + { + "epoch": 0.7095620121422377, + "grad_norm": 2.2541587352752686, + "learning_rate": 1.5324967358189467e-05, + "loss": 0.0266, + "step": 6545 + }, + { + "epoch": 0.7096704249783174, + "grad_norm": 0.28627774119377136, + "learning_rate": 1.5324241984622083e-05, + "loss": 0.0041, + "step": 6546 + }, + { + "epoch": 0.7097788378143972, + "grad_norm": 2.212575912475586, + "learning_rate": 1.5323516611054695e-05, + "loss": 0.0135, + "step": 6547 + }, + { + "epoch": 0.709887250650477, + "grad_norm": 2.869673013687134, + "learning_rate": 1.5322791237487307e-05, + "loss": 0.045, + "step": 6548 + }, + { + "epoch": 0.7099956634865568, + "grad_norm": 1.4192330837249756, + "learning_rate": 1.532206586391992e-05, + "loss": 0.0235, + "step": 6549 + }, + { + "epoch": 0.7101040763226366, + "grad_norm": 6.468691349029541, + "learning_rate": 1.532134049035253e-05, + "loss": 0.1045, + "step": 6550 + }, + { + "epoch": 0.7102124891587164, + "grad_norm": 2.712613582611084, + "learning_rate": 1.5320615116785143e-05, + "loss": 0.0685, + "step": 6551 + }, + { + "epoch": 0.7103209019947961, + "grad_norm": 0.49722713232040405, + "learning_rate": 1.531988974321776e-05, + "loss": 0.0083, + "step": 6552 + }, + { + "epoch": 0.710429314830876, + "grad_norm": 2.40541410446167, + "learning_rate": 1.531916436965037e-05, + "loss": 0.0362, + "step": 6553 + }, + { + "epoch": 0.7105377276669558, + "grad_norm": 2.512495279312134, + "learning_rate": 1.5318438996082987e-05, + "loss": 0.0385, + "step": 6554 + }, + { + "epoch": 0.7106461405030355, + "grad_norm": 1.942123532295227, + "learning_rate": 1.53177136225156e-05, + "loss": 0.0339, + "step": 6555 + }, + { + "epoch": 0.7107545533391153, + "grad_norm": 4.213188648223877, + "learning_rate": 1.531698824894821e-05, + "loss": 0.0449, + "step": 6556 + }, + { + "epoch": 0.7108629661751952, + "grad_norm": 5.0038161277771, + "learning_rate": 1.5316262875380823e-05, + "loss": 0.0311, + "step": 6557 + }, + { + "epoch": 0.7109713790112749, + "grad_norm": 1.811711072921753, + "learning_rate": 1.5315537501813435e-05, + "loss": 0.0292, + "step": 6558 + }, + { + "epoch": 0.7110797918473547, + "grad_norm": 5.146531581878662, + "learning_rate": 1.5314812128246047e-05, + "loss": 0.0788, + "step": 6559 + }, + { + "epoch": 0.7111882046834345, + "grad_norm": 3.489025115966797, + "learning_rate": 1.531408675467866e-05, + "loss": 0.0895, + "step": 6560 + }, + { + "epoch": 0.7112966175195143, + "grad_norm": 6.969557762145996, + "learning_rate": 1.5313361381111275e-05, + "loss": 0.0358, + "step": 6561 + }, + { + "epoch": 0.7114050303555941, + "grad_norm": 2.1688997745513916, + "learning_rate": 1.5312636007543887e-05, + "loss": 0.0205, + "step": 6562 + }, + { + "epoch": 0.7115134431916739, + "grad_norm": 3.5247714519500732, + "learning_rate": 1.53119106339765e-05, + "loss": 0.037, + "step": 6563 + }, + { + "epoch": 0.7116218560277537, + "grad_norm": 7.748881816864014, + "learning_rate": 1.531118526040911e-05, + "loss": 0.075, + "step": 6564 + }, + { + "epoch": 0.7117302688638335, + "grad_norm": 3.5531084537506104, + "learning_rate": 1.5310459886841723e-05, + "loss": 0.0181, + "step": 6565 + }, + { + "epoch": 0.7118386816999133, + "grad_norm": 1.0733762979507446, + "learning_rate": 1.5309734513274336e-05, + "loss": 0.0278, + "step": 6566 + }, + { + "epoch": 0.711947094535993, + "grad_norm": 3.756740093231201, + "learning_rate": 1.530900913970695e-05, + "loss": 0.0243, + "step": 6567 + }, + { + "epoch": 0.7120555073720729, + "grad_norm": 5.065311908721924, + "learning_rate": 1.5308283766139563e-05, + "loss": 0.066, + "step": 6568 + }, + { + "epoch": 0.7121639202081527, + "grad_norm": 2.290308713912964, + "learning_rate": 1.5307558392572175e-05, + "loss": 0.0317, + "step": 6569 + }, + { + "epoch": 0.7122723330442324, + "grad_norm": 2.8807599544525146, + "learning_rate": 1.530683301900479e-05, + "loss": 0.0335, + "step": 6570 + }, + { + "epoch": 0.7123807458803122, + "grad_norm": 3.1195383071899414, + "learning_rate": 1.5306107645437403e-05, + "loss": 0.0613, + "step": 6571 + }, + { + "epoch": 0.7124891587163921, + "grad_norm": 3.1050148010253906, + "learning_rate": 1.5305382271870015e-05, + "loss": 0.045, + "step": 6572 + }, + { + "epoch": 0.7125975715524718, + "grad_norm": 3.314127206802368, + "learning_rate": 1.5304656898302627e-05, + "loss": 0.0743, + "step": 6573 + }, + { + "epoch": 0.7127059843885516, + "grad_norm": 5.15139627456665, + "learning_rate": 1.530393152473524e-05, + "loss": 0.0383, + "step": 6574 + }, + { + "epoch": 0.7128143972246314, + "grad_norm": 1.5839916467666626, + "learning_rate": 1.5303206151167852e-05, + "loss": 0.0251, + "step": 6575 + }, + { + "epoch": 0.7129228100607112, + "grad_norm": 2.866661548614502, + "learning_rate": 1.5302480777600467e-05, + "loss": 0.0261, + "step": 6576 + }, + { + "epoch": 0.713031222896791, + "grad_norm": 2.7178537845611572, + "learning_rate": 1.530175540403308e-05, + "loss": 0.0459, + "step": 6577 + }, + { + "epoch": 0.7131396357328708, + "grad_norm": 3.765813112258911, + "learning_rate": 1.530103003046569e-05, + "loss": 0.0579, + "step": 6578 + }, + { + "epoch": 0.7132480485689505, + "grad_norm": 1.7827199697494507, + "learning_rate": 1.5300304656898304e-05, + "loss": 0.0406, + "step": 6579 + }, + { + "epoch": 0.7133564614050304, + "grad_norm": 2.691354274749756, + "learning_rate": 1.5299579283330916e-05, + "loss": 0.0182, + "step": 6580 + }, + { + "epoch": 0.7134648742411102, + "grad_norm": 3.3224995136260986, + "learning_rate": 1.5298853909763528e-05, + "loss": 0.0258, + "step": 6581 + }, + { + "epoch": 0.7135732870771899, + "grad_norm": 1.0497149229049683, + "learning_rate": 1.5298128536196143e-05, + "loss": 0.0134, + "step": 6582 + }, + { + "epoch": 0.7136816999132697, + "grad_norm": 4.133944034576416, + "learning_rate": 1.5297403162628756e-05, + "loss": 0.0235, + "step": 6583 + }, + { + "epoch": 0.7137901127493496, + "grad_norm": 2.7720491886138916, + "learning_rate": 1.5296677789061368e-05, + "loss": 0.06, + "step": 6584 + }, + { + "epoch": 0.7138985255854293, + "grad_norm": 2.382075786590576, + "learning_rate": 1.529595241549398e-05, + "loss": 0.033, + "step": 6585 + }, + { + "epoch": 0.7140069384215091, + "grad_norm": 4.77400541305542, + "learning_rate": 1.5295227041926592e-05, + "loss": 0.1012, + "step": 6586 + }, + { + "epoch": 0.7141153512575888, + "grad_norm": 2.9210312366485596, + "learning_rate": 1.5294501668359204e-05, + "loss": 0.0296, + "step": 6587 + }, + { + "epoch": 0.7142237640936687, + "grad_norm": 2.6526269912719727, + "learning_rate": 1.529377629479182e-05, + "loss": 0.0402, + "step": 6588 + }, + { + "epoch": 0.7143321769297485, + "grad_norm": 1.1245639324188232, + "learning_rate": 1.5293050921224432e-05, + "loss": 0.0193, + "step": 6589 + }, + { + "epoch": 0.7144405897658282, + "grad_norm": 4.372034072875977, + "learning_rate": 1.5292325547657044e-05, + "loss": 0.0238, + "step": 6590 + }, + { + "epoch": 0.714549002601908, + "grad_norm": 3.5859596729278564, + "learning_rate": 1.529160017408966e-05, + "loss": 0.0391, + "step": 6591 + }, + { + "epoch": 0.7146574154379879, + "grad_norm": 0.9146093726158142, + "learning_rate": 1.529087480052227e-05, + "loss": 0.0173, + "step": 6592 + }, + { + "epoch": 0.7147658282740676, + "grad_norm": 2.310546398162842, + "learning_rate": 1.5290149426954884e-05, + "loss": 0.0344, + "step": 6593 + }, + { + "epoch": 0.7148742411101474, + "grad_norm": 4.20704984664917, + "learning_rate": 1.5289424053387496e-05, + "loss": 0.0252, + "step": 6594 + }, + { + "epoch": 0.7149826539462273, + "grad_norm": 1.23106050491333, + "learning_rate": 1.5288698679820108e-05, + "loss": 0.0332, + "step": 6595 + }, + { + "epoch": 0.715091066782307, + "grad_norm": 1.6909065246582031, + "learning_rate": 1.528797330625272e-05, + "loss": 0.0256, + "step": 6596 + }, + { + "epoch": 0.7151994796183868, + "grad_norm": 2.420949935913086, + "learning_rate": 1.5287247932685332e-05, + "loss": 0.0236, + "step": 6597 + }, + { + "epoch": 0.7153078924544666, + "grad_norm": 3.31394100189209, + "learning_rate": 1.5286522559117948e-05, + "loss": 0.0396, + "step": 6598 + }, + { + "epoch": 0.7154163052905465, + "grad_norm": 1.9488754272460938, + "learning_rate": 1.528579718555056e-05, + "loss": 0.032, + "step": 6599 + }, + { + "epoch": 0.7155247181266262, + "grad_norm": 4.518686771392822, + "learning_rate": 1.5285071811983172e-05, + "loss": 0.0454, + "step": 6600 + }, + { + "epoch": 0.715633130962706, + "grad_norm": 2.3300724029541016, + "learning_rate": 1.5284346438415784e-05, + "loss": 0.0419, + "step": 6601 + }, + { + "epoch": 0.7157415437987857, + "grad_norm": 5.869266033172607, + "learning_rate": 1.5283621064848397e-05, + "loss": 0.073, + "step": 6602 + }, + { + "epoch": 0.7158499566348656, + "grad_norm": 1.5411179065704346, + "learning_rate": 1.528289569128101e-05, + "loss": 0.0403, + "step": 6603 + }, + { + "epoch": 0.7159583694709454, + "grad_norm": 0.5202935338020325, + "learning_rate": 1.5282170317713624e-05, + "loss": 0.0088, + "step": 6604 + }, + { + "epoch": 0.7160667823070251, + "grad_norm": 2.5672123432159424, + "learning_rate": 1.5281444944146236e-05, + "loss": 0.0407, + "step": 6605 + }, + { + "epoch": 0.7161751951431049, + "grad_norm": 3.2341957092285156, + "learning_rate": 1.5280719570578852e-05, + "loss": 0.0258, + "step": 6606 + }, + { + "epoch": 0.7162836079791848, + "grad_norm": 4.022128582000732, + "learning_rate": 1.5279994197011464e-05, + "loss": 0.05, + "step": 6607 + }, + { + "epoch": 0.7163920208152645, + "grad_norm": 1.011397123336792, + "learning_rate": 1.5279268823444076e-05, + "loss": 0.0105, + "step": 6608 + }, + { + "epoch": 0.7165004336513443, + "grad_norm": 2.971468448638916, + "learning_rate": 1.5278543449876688e-05, + "loss": 0.0418, + "step": 6609 + }, + { + "epoch": 0.7166088464874241, + "grad_norm": 2.9132182598114014, + "learning_rate": 1.52778180763093e-05, + "loss": 0.0321, + "step": 6610 + }, + { + "epoch": 0.7167172593235039, + "grad_norm": 5.13993501663208, + "learning_rate": 1.5277092702741913e-05, + "loss": 0.0683, + "step": 6611 + }, + { + "epoch": 0.7168256721595837, + "grad_norm": 0.6078640818595886, + "learning_rate": 1.5276367329174525e-05, + "loss": 0.0061, + "step": 6612 + }, + { + "epoch": 0.7169340849956635, + "grad_norm": 1.8235881328582764, + "learning_rate": 1.527564195560714e-05, + "loss": 0.0278, + "step": 6613 + }, + { + "epoch": 0.7170424978317432, + "grad_norm": 2.8410558700561523, + "learning_rate": 1.5274916582039752e-05, + "loss": 0.0437, + "step": 6614 + }, + { + "epoch": 0.7171509106678231, + "grad_norm": 2.138317346572876, + "learning_rate": 1.5274191208472365e-05, + "loss": 0.0783, + "step": 6615 + }, + { + "epoch": 0.7172593235039029, + "grad_norm": 0.5706311464309692, + "learning_rate": 1.5273465834904977e-05, + "loss": 0.0058, + "step": 6616 + }, + { + "epoch": 0.7173677363399826, + "grad_norm": 2.419768810272217, + "learning_rate": 1.527274046133759e-05, + "loss": 0.0157, + "step": 6617 + }, + { + "epoch": 0.7174761491760624, + "grad_norm": 1.3200689554214478, + "learning_rate": 1.52720150877702e-05, + "loss": 0.0331, + "step": 6618 + }, + { + "epoch": 0.7175845620121423, + "grad_norm": 0.2722494304180145, + "learning_rate": 1.5271289714202817e-05, + "loss": 0.0028, + "step": 6619 + }, + { + "epoch": 0.717692974848222, + "grad_norm": 4.000482559204102, + "learning_rate": 1.527056434063543e-05, + "loss": 0.0652, + "step": 6620 + }, + { + "epoch": 0.7178013876843018, + "grad_norm": 4.4130120277404785, + "learning_rate": 1.526983896706804e-05, + "loss": 0.0156, + "step": 6621 + }, + { + "epoch": 0.7179098005203816, + "grad_norm": 1.5127230882644653, + "learning_rate": 1.5269113593500656e-05, + "loss": 0.0102, + "step": 6622 + }, + { + "epoch": 0.7180182133564614, + "grad_norm": 0.7135778665542603, + "learning_rate": 1.526838821993327e-05, + "loss": 0.0062, + "step": 6623 + }, + { + "epoch": 0.7181266261925412, + "grad_norm": 0.6288947463035583, + "learning_rate": 1.526766284636588e-05, + "loss": 0.0059, + "step": 6624 + }, + { + "epoch": 0.718235039028621, + "grad_norm": 2.9630324840545654, + "learning_rate": 1.5266937472798493e-05, + "loss": 0.0451, + "step": 6625 + }, + { + "epoch": 0.7183434518647008, + "grad_norm": 4.614346504211426, + "learning_rate": 1.5266212099231105e-05, + "loss": 0.084, + "step": 6626 + }, + { + "epoch": 0.7184518647007806, + "grad_norm": 2.8087332248687744, + "learning_rate": 1.5265486725663717e-05, + "loss": 0.0346, + "step": 6627 + }, + { + "epoch": 0.7185602775368604, + "grad_norm": 1.5182592868804932, + "learning_rate": 1.5264761352096333e-05, + "loss": 0.0135, + "step": 6628 + }, + { + "epoch": 0.7186686903729401, + "grad_norm": 5.40833854675293, + "learning_rate": 1.5264035978528945e-05, + "loss": 0.0856, + "step": 6629 + }, + { + "epoch": 0.71877710320902, + "grad_norm": 2.7131247520446777, + "learning_rate": 1.5263310604961557e-05, + "loss": 0.0213, + "step": 6630 + }, + { + "epoch": 0.7188855160450998, + "grad_norm": 3.732726573944092, + "learning_rate": 1.526258523139417e-05, + "loss": 0.037, + "step": 6631 + }, + { + "epoch": 0.7189939288811795, + "grad_norm": 3.6104631423950195, + "learning_rate": 1.526185985782678e-05, + "loss": 0.027, + "step": 6632 + }, + { + "epoch": 0.7191023417172593, + "grad_norm": 4.967593669891357, + "learning_rate": 1.5261134484259393e-05, + "loss": 0.0283, + "step": 6633 + }, + { + "epoch": 0.7192107545533392, + "grad_norm": 3.216369152069092, + "learning_rate": 1.526040911069201e-05, + "loss": 0.0424, + "step": 6634 + }, + { + "epoch": 0.7193191673894189, + "grad_norm": 1.368926763534546, + "learning_rate": 1.525968373712462e-05, + "loss": 0.0156, + "step": 6635 + }, + { + "epoch": 0.7194275802254987, + "grad_norm": 2.2568695545196533, + "learning_rate": 1.5258958363557233e-05, + "loss": 0.0256, + "step": 6636 + }, + { + "epoch": 0.7195359930615784, + "grad_norm": 2.947425127029419, + "learning_rate": 1.5258232989989845e-05, + "loss": 0.0144, + "step": 6637 + }, + { + "epoch": 0.7196444058976583, + "grad_norm": 3.422844409942627, + "learning_rate": 1.5257507616422457e-05, + "loss": 0.0591, + "step": 6638 + }, + { + "epoch": 0.7197528187337381, + "grad_norm": 10.219019889831543, + "learning_rate": 1.5256782242855073e-05, + "loss": 0.0494, + "step": 6639 + }, + { + "epoch": 0.7198612315698178, + "grad_norm": 3.2007226943969727, + "learning_rate": 1.5256056869287685e-05, + "loss": 0.0614, + "step": 6640 + }, + { + "epoch": 0.7199696444058976, + "grad_norm": 3.362277030944824, + "learning_rate": 1.5255331495720297e-05, + "loss": 0.0534, + "step": 6641 + }, + { + "epoch": 0.7200780572419775, + "grad_norm": 5.688760280609131, + "learning_rate": 1.5254606122152911e-05, + "loss": 0.0898, + "step": 6642 + }, + { + "epoch": 0.7201864700780573, + "grad_norm": 8.449666976928711, + "learning_rate": 1.5253880748585523e-05, + "loss": 0.1074, + "step": 6643 + }, + { + "epoch": 0.720294882914137, + "grad_norm": 1.8506392240524292, + "learning_rate": 1.5253155375018135e-05, + "loss": 0.0675, + "step": 6644 + }, + { + "epoch": 0.7204032957502168, + "grad_norm": 4.166075706481934, + "learning_rate": 1.525243000145075e-05, + "loss": 0.0912, + "step": 6645 + }, + { + "epoch": 0.7205117085862967, + "grad_norm": 0.638714611530304, + "learning_rate": 1.5251704627883361e-05, + "loss": 0.0088, + "step": 6646 + }, + { + "epoch": 0.7206201214223764, + "grad_norm": 2.4657673835754395, + "learning_rate": 1.5250979254315973e-05, + "loss": 0.0193, + "step": 6647 + }, + { + "epoch": 0.7207285342584562, + "grad_norm": 4.347072124481201, + "learning_rate": 1.5250253880748587e-05, + "loss": 0.0577, + "step": 6648 + }, + { + "epoch": 0.7208369470945359, + "grad_norm": 3.6821365356445312, + "learning_rate": 1.52495285071812e-05, + "loss": 0.0313, + "step": 6649 + }, + { + "epoch": 0.7209453599306158, + "grad_norm": 5.554458141326904, + "learning_rate": 1.5248803133613812e-05, + "loss": 0.0529, + "step": 6650 + }, + { + "epoch": 0.7210537727666956, + "grad_norm": 2.285510540008545, + "learning_rate": 1.5248077760046425e-05, + "loss": 0.0315, + "step": 6651 + }, + { + "epoch": 0.7211621856027753, + "grad_norm": 3.6372478008270264, + "learning_rate": 1.5247352386479038e-05, + "loss": 0.0435, + "step": 6652 + }, + { + "epoch": 0.7212705984388552, + "grad_norm": 3.1118807792663574, + "learning_rate": 1.524662701291165e-05, + "loss": 0.0582, + "step": 6653 + }, + { + "epoch": 0.721379011274935, + "grad_norm": 8.1882963180542, + "learning_rate": 1.5245901639344264e-05, + "loss": 0.0536, + "step": 6654 + }, + { + "epoch": 0.7214874241110147, + "grad_norm": 0.7709168195724487, + "learning_rate": 1.5245176265776876e-05, + "loss": 0.0107, + "step": 6655 + }, + { + "epoch": 0.7215958369470945, + "grad_norm": 2.413869619369507, + "learning_rate": 1.5244450892209488e-05, + "loss": 0.0511, + "step": 6656 + }, + { + "epoch": 0.7217042497831744, + "grad_norm": 6.421358108520508, + "learning_rate": 1.5243725518642103e-05, + "loss": 0.0735, + "step": 6657 + }, + { + "epoch": 0.7218126626192541, + "grad_norm": 3.1919705867767334, + "learning_rate": 1.5243000145074716e-05, + "loss": 0.0611, + "step": 6658 + }, + { + "epoch": 0.7219210754553339, + "grad_norm": 3.5819692611694336, + "learning_rate": 1.5242274771507328e-05, + "loss": 0.0473, + "step": 6659 + }, + { + "epoch": 0.7220294882914137, + "grad_norm": 1.5005654096603394, + "learning_rate": 1.5241549397939941e-05, + "loss": 0.0378, + "step": 6660 + }, + { + "epoch": 0.7221379011274935, + "grad_norm": 3.4607126712799072, + "learning_rate": 1.5240824024372554e-05, + "loss": 0.0498, + "step": 6661 + }, + { + "epoch": 0.7222463139635733, + "grad_norm": 2.0922131538391113, + "learning_rate": 1.5240098650805166e-05, + "loss": 0.0495, + "step": 6662 + }, + { + "epoch": 0.7223547267996531, + "grad_norm": 1.3317757844924927, + "learning_rate": 1.523937327723778e-05, + "loss": 0.017, + "step": 6663 + }, + { + "epoch": 0.7224631396357328, + "grad_norm": 1.7123146057128906, + "learning_rate": 1.5238647903670392e-05, + "loss": 0.0281, + "step": 6664 + }, + { + "epoch": 0.7225715524718127, + "grad_norm": 1.3026227951049805, + "learning_rate": 1.5237922530103004e-05, + "loss": 0.0258, + "step": 6665 + }, + { + "epoch": 0.7226799653078925, + "grad_norm": 3.502211093902588, + "learning_rate": 1.5237197156535618e-05, + "loss": 0.0317, + "step": 6666 + }, + { + "epoch": 0.7227883781439722, + "grad_norm": 2.660996198654175, + "learning_rate": 1.523647178296823e-05, + "loss": 0.066, + "step": 6667 + }, + { + "epoch": 0.722896790980052, + "grad_norm": 0.9136035442352295, + "learning_rate": 1.5235746409400842e-05, + "loss": 0.0161, + "step": 6668 + }, + { + "epoch": 0.7230052038161319, + "grad_norm": 2.901873826980591, + "learning_rate": 1.5235021035833454e-05, + "loss": 0.0465, + "step": 6669 + }, + { + "epoch": 0.7231136166522116, + "grad_norm": 4.895410537719727, + "learning_rate": 1.5234295662266068e-05, + "loss": 0.0195, + "step": 6670 + }, + { + "epoch": 0.7232220294882914, + "grad_norm": 2.692314624786377, + "learning_rate": 1.523357028869868e-05, + "loss": 0.0669, + "step": 6671 + }, + { + "epoch": 0.7233304423243712, + "grad_norm": 4.6260786056518555, + "learning_rate": 1.5232844915131292e-05, + "loss": 0.0388, + "step": 6672 + }, + { + "epoch": 0.723438855160451, + "grad_norm": 1.860863208770752, + "learning_rate": 1.5232119541563906e-05, + "loss": 0.0467, + "step": 6673 + }, + { + "epoch": 0.7235472679965308, + "grad_norm": 4.753927707672119, + "learning_rate": 1.523139416799652e-05, + "loss": 0.0206, + "step": 6674 + }, + { + "epoch": 0.7236556808326106, + "grad_norm": 0.6477971076965332, + "learning_rate": 1.5230668794429134e-05, + "loss": 0.0078, + "step": 6675 + }, + { + "epoch": 0.7237640936686903, + "grad_norm": 1.537627935409546, + "learning_rate": 1.5229943420861746e-05, + "loss": 0.0223, + "step": 6676 + }, + { + "epoch": 0.7238725065047702, + "grad_norm": 1.8125231266021729, + "learning_rate": 1.5229218047294358e-05, + "loss": 0.0254, + "step": 6677 + }, + { + "epoch": 0.72398091934085, + "grad_norm": 2.0656301975250244, + "learning_rate": 1.5228492673726972e-05, + "loss": 0.0294, + "step": 6678 + }, + { + "epoch": 0.7240893321769297, + "grad_norm": 1.347060203552246, + "learning_rate": 1.5227767300159584e-05, + "loss": 0.05, + "step": 6679 + }, + { + "epoch": 0.7241977450130095, + "grad_norm": 3.5160231590270996, + "learning_rate": 1.5227041926592196e-05, + "loss": 0.02, + "step": 6680 + }, + { + "epoch": 0.7243061578490894, + "grad_norm": 3.427682876586914, + "learning_rate": 1.5226316553024808e-05, + "loss": 0.0434, + "step": 6681 + }, + { + "epoch": 0.7244145706851691, + "grad_norm": 6.193972587585449, + "learning_rate": 1.5225591179457422e-05, + "loss": 0.0307, + "step": 6682 + }, + { + "epoch": 0.7245229835212489, + "grad_norm": 7.0545220375061035, + "learning_rate": 1.5224865805890034e-05, + "loss": 0.0675, + "step": 6683 + }, + { + "epoch": 0.7246313963573288, + "grad_norm": 3.155596971511841, + "learning_rate": 1.5224140432322646e-05, + "loss": 0.0503, + "step": 6684 + }, + { + "epoch": 0.7247398091934085, + "grad_norm": 3.0611698627471924, + "learning_rate": 1.522341505875526e-05, + "loss": 0.0464, + "step": 6685 + }, + { + "epoch": 0.7248482220294883, + "grad_norm": 2.517714262008667, + "learning_rate": 1.5222689685187872e-05, + "loss": 0.0404, + "step": 6686 + }, + { + "epoch": 0.724956634865568, + "grad_norm": 2.3941450119018555, + "learning_rate": 1.5221964311620485e-05, + "loss": 0.056, + "step": 6687 + }, + { + "epoch": 0.7250650477016479, + "grad_norm": 2.9098541736602783, + "learning_rate": 1.5221238938053098e-05, + "loss": 0.0336, + "step": 6688 + }, + { + "epoch": 0.7251734605377277, + "grad_norm": 1.5946043729782104, + "learning_rate": 1.522051356448571e-05, + "loss": 0.0364, + "step": 6689 + }, + { + "epoch": 0.7252818733738075, + "grad_norm": 1.9495558738708496, + "learning_rate": 1.5219788190918323e-05, + "loss": 0.0226, + "step": 6690 + }, + { + "epoch": 0.7253902862098872, + "grad_norm": 3.259265422821045, + "learning_rate": 1.5219062817350938e-05, + "loss": 0.0751, + "step": 6691 + }, + { + "epoch": 0.7254986990459671, + "grad_norm": 2.833233594894409, + "learning_rate": 1.521833744378355e-05, + "loss": 0.0407, + "step": 6692 + }, + { + "epoch": 0.7256071118820469, + "grad_norm": 2.068044662475586, + "learning_rate": 1.5217612070216163e-05, + "loss": 0.026, + "step": 6693 + }, + { + "epoch": 0.7257155247181266, + "grad_norm": 3.545839786529541, + "learning_rate": 1.5216886696648776e-05, + "loss": 0.0587, + "step": 6694 + }, + { + "epoch": 0.7258239375542064, + "grad_norm": 1.8023391962051392, + "learning_rate": 1.5216161323081389e-05, + "loss": 0.0196, + "step": 6695 + }, + { + "epoch": 0.7259323503902863, + "grad_norm": 1.1242752075195312, + "learning_rate": 1.5215435949514e-05, + "loss": 0.0107, + "step": 6696 + }, + { + "epoch": 0.726040763226366, + "grad_norm": 2.33671236038208, + "learning_rate": 1.5214710575946615e-05, + "loss": 0.0279, + "step": 6697 + }, + { + "epoch": 0.7261491760624458, + "grad_norm": 2.1080873012542725, + "learning_rate": 1.5213985202379227e-05, + "loss": 0.0807, + "step": 6698 + }, + { + "epoch": 0.7262575888985255, + "grad_norm": 1.8061412572860718, + "learning_rate": 1.5213259828811839e-05, + "loss": 0.0223, + "step": 6699 + }, + { + "epoch": 0.7263660017346054, + "grad_norm": 1.4812734127044678, + "learning_rate": 1.5212534455244453e-05, + "loss": 0.0298, + "step": 6700 + }, + { + "epoch": 0.7264744145706852, + "grad_norm": 0.7065081000328064, + "learning_rate": 1.5211809081677065e-05, + "loss": 0.0114, + "step": 6701 + }, + { + "epoch": 0.7265828274067649, + "grad_norm": 2.4082183837890625, + "learning_rate": 1.5211083708109677e-05, + "loss": 0.0247, + "step": 6702 + }, + { + "epoch": 0.7266912402428447, + "grad_norm": 3.2607455253601074, + "learning_rate": 1.521035833454229e-05, + "loss": 0.0865, + "step": 6703 + }, + { + "epoch": 0.7267996530789246, + "grad_norm": 3.5232326984405518, + "learning_rate": 1.5209632960974903e-05, + "loss": 0.0234, + "step": 6704 + }, + { + "epoch": 0.7269080659150043, + "grad_norm": 4.017436504364014, + "learning_rate": 1.5208907587407515e-05, + "loss": 0.0456, + "step": 6705 + }, + { + "epoch": 0.7270164787510841, + "grad_norm": 3.6890361309051514, + "learning_rate": 1.5208182213840129e-05, + "loss": 0.0522, + "step": 6706 + }, + { + "epoch": 0.7271248915871639, + "grad_norm": 2.501910448074341, + "learning_rate": 1.5207456840272741e-05, + "loss": 0.0437, + "step": 6707 + }, + { + "epoch": 0.7272333044232437, + "grad_norm": 1.5351797342300415, + "learning_rate": 1.5206731466705355e-05, + "loss": 0.055, + "step": 6708 + }, + { + "epoch": 0.7273417172593235, + "grad_norm": 1.5151523351669312, + "learning_rate": 1.5206006093137969e-05, + "loss": 0.0507, + "step": 6709 + }, + { + "epoch": 0.7274501300954033, + "grad_norm": 1.796477198600769, + "learning_rate": 1.520528071957058e-05, + "loss": 0.0238, + "step": 6710 + }, + { + "epoch": 0.727558542931483, + "grad_norm": 2.233189821243286, + "learning_rate": 1.5204555346003193e-05, + "loss": 0.042, + "step": 6711 + }, + { + "epoch": 0.7276669557675629, + "grad_norm": 2.777512311935425, + "learning_rate": 1.5203829972435807e-05, + "loss": 0.0559, + "step": 6712 + }, + { + "epoch": 0.7277753686036427, + "grad_norm": 2.5562903881073, + "learning_rate": 1.5203104598868419e-05, + "loss": 0.0371, + "step": 6713 + }, + { + "epoch": 0.7278837814397224, + "grad_norm": 2.1675174236297607, + "learning_rate": 1.5202379225301031e-05, + "loss": 0.0231, + "step": 6714 + }, + { + "epoch": 0.7279921942758023, + "grad_norm": 3.1069326400756836, + "learning_rate": 1.5201653851733645e-05, + "loss": 0.039, + "step": 6715 + }, + { + "epoch": 0.7281006071118821, + "grad_norm": 0.8981738090515137, + "learning_rate": 1.5200928478166257e-05, + "loss": 0.0259, + "step": 6716 + }, + { + "epoch": 0.7282090199479618, + "grad_norm": 1.793782114982605, + "learning_rate": 1.520020310459887e-05, + "loss": 0.0412, + "step": 6717 + }, + { + "epoch": 0.7283174327840416, + "grad_norm": 3.2200937271118164, + "learning_rate": 1.5199477731031483e-05, + "loss": 0.0721, + "step": 6718 + }, + { + "epoch": 0.7284258456201215, + "grad_norm": 2.2876012325286865, + "learning_rate": 1.5198752357464095e-05, + "loss": 0.057, + "step": 6719 + }, + { + "epoch": 0.7285342584562012, + "grad_norm": 1.6016408205032349, + "learning_rate": 1.5198026983896707e-05, + "loss": 0.0186, + "step": 6720 + }, + { + "epoch": 0.728642671292281, + "grad_norm": 2.4639525413513184, + "learning_rate": 1.519730161032932e-05, + "loss": 0.0286, + "step": 6721 + }, + { + "epoch": 0.7287510841283608, + "grad_norm": 2.8271231651306152, + "learning_rate": 1.5196576236761933e-05, + "loss": 0.053, + "step": 6722 + }, + { + "epoch": 0.7288594969644406, + "grad_norm": 1.0959464311599731, + "learning_rate": 1.5195850863194545e-05, + "loss": 0.0157, + "step": 6723 + }, + { + "epoch": 0.7289679098005204, + "grad_norm": 1.5732710361480713, + "learning_rate": 1.5195125489627158e-05, + "loss": 0.0356, + "step": 6724 + }, + { + "epoch": 0.7290763226366002, + "grad_norm": 0.723861038684845, + "learning_rate": 1.5194400116059771e-05, + "loss": 0.0082, + "step": 6725 + }, + { + "epoch": 0.7291847354726799, + "grad_norm": 2.3472609519958496, + "learning_rate": 1.5193674742492385e-05, + "loss": 0.0345, + "step": 6726 + }, + { + "epoch": 0.7292931483087598, + "grad_norm": 1.0089025497436523, + "learning_rate": 1.5192949368924999e-05, + "loss": 0.02, + "step": 6727 + }, + { + "epoch": 0.7294015611448396, + "grad_norm": 2.7642688751220703, + "learning_rate": 1.5192223995357611e-05, + "loss": 0.0465, + "step": 6728 + }, + { + "epoch": 0.7295099739809193, + "grad_norm": 6.278065204620361, + "learning_rate": 1.5191498621790223e-05, + "loss": 0.095, + "step": 6729 + }, + { + "epoch": 0.7296183868169991, + "grad_norm": 4.558013439178467, + "learning_rate": 1.5190773248222837e-05, + "loss": 0.0524, + "step": 6730 + }, + { + "epoch": 0.729726799653079, + "grad_norm": 0.8333847522735596, + "learning_rate": 1.519004787465545e-05, + "loss": 0.0164, + "step": 6731 + }, + { + "epoch": 0.7298352124891587, + "grad_norm": 7.0528717041015625, + "learning_rate": 1.5189322501088062e-05, + "loss": 0.1008, + "step": 6732 + }, + { + "epoch": 0.7299436253252385, + "grad_norm": 1.1214733123779297, + "learning_rate": 1.5188597127520674e-05, + "loss": 0.0277, + "step": 6733 + }, + { + "epoch": 0.7300520381613183, + "grad_norm": 2.6479454040527344, + "learning_rate": 1.5187871753953288e-05, + "loss": 0.0228, + "step": 6734 + }, + { + "epoch": 0.7301604509973981, + "grad_norm": 2.205618143081665, + "learning_rate": 1.51871463803859e-05, + "loss": 0.0345, + "step": 6735 + }, + { + "epoch": 0.7302688638334779, + "grad_norm": 3.3147506713867188, + "learning_rate": 1.5186421006818512e-05, + "loss": 0.0616, + "step": 6736 + }, + { + "epoch": 0.7303772766695577, + "grad_norm": 3.9687743186950684, + "learning_rate": 1.5185695633251126e-05, + "loss": 0.0303, + "step": 6737 + }, + { + "epoch": 0.7304856895056374, + "grad_norm": 7.438770771026611, + "learning_rate": 1.5184970259683738e-05, + "loss": 0.031, + "step": 6738 + }, + { + "epoch": 0.7305941023417173, + "grad_norm": 1.0336263179779053, + "learning_rate": 1.518424488611635e-05, + "loss": 0.0135, + "step": 6739 + }, + { + "epoch": 0.730702515177797, + "grad_norm": 6.048058986663818, + "learning_rate": 1.5183519512548964e-05, + "loss": 0.0665, + "step": 6740 + }, + { + "epoch": 0.7308109280138768, + "grad_norm": 1.188639521598816, + "learning_rate": 1.5182794138981576e-05, + "loss": 0.0137, + "step": 6741 + }, + { + "epoch": 0.7309193408499567, + "grad_norm": 1.224031686782837, + "learning_rate": 1.5182068765414188e-05, + "loss": 0.0086, + "step": 6742 + }, + { + "epoch": 0.7310277536860365, + "grad_norm": 1.713287591934204, + "learning_rate": 1.5181343391846804e-05, + "loss": 0.0263, + "step": 6743 + }, + { + "epoch": 0.7311361665221162, + "grad_norm": 1.7000385522842407, + "learning_rate": 1.5180618018279416e-05, + "loss": 0.0317, + "step": 6744 + }, + { + "epoch": 0.731244579358196, + "grad_norm": 4.569529056549072, + "learning_rate": 1.5179892644712028e-05, + "loss": 0.0583, + "step": 6745 + }, + { + "epoch": 0.7313529921942759, + "grad_norm": 5.286359786987305, + "learning_rate": 1.5179167271144642e-05, + "loss": 0.0536, + "step": 6746 + }, + { + "epoch": 0.7314614050303556, + "grad_norm": 1.4433311223983765, + "learning_rate": 1.5178441897577254e-05, + "loss": 0.0208, + "step": 6747 + }, + { + "epoch": 0.7315698178664354, + "grad_norm": 4.632035255432129, + "learning_rate": 1.5177716524009866e-05, + "loss": 0.0244, + "step": 6748 + }, + { + "epoch": 0.7316782307025151, + "grad_norm": 1.2573155164718628, + "learning_rate": 1.517699115044248e-05, + "loss": 0.0189, + "step": 6749 + }, + { + "epoch": 0.731786643538595, + "grad_norm": 1.5941861867904663, + "learning_rate": 1.5176265776875092e-05, + "loss": 0.0388, + "step": 6750 + }, + { + "epoch": 0.7318950563746748, + "grad_norm": 1.6672571897506714, + "learning_rate": 1.5175540403307704e-05, + "loss": 0.0172, + "step": 6751 + }, + { + "epoch": 0.7320034692107545, + "grad_norm": 5.8206682205200195, + "learning_rate": 1.5174815029740318e-05, + "loss": 0.0597, + "step": 6752 + }, + { + "epoch": 0.7321118820468343, + "grad_norm": 2.7459187507629395, + "learning_rate": 1.517408965617293e-05, + "loss": 0.0261, + "step": 6753 + }, + { + "epoch": 0.7322202948829142, + "grad_norm": 3.5945868492126465, + "learning_rate": 1.5173364282605542e-05, + "loss": 0.0748, + "step": 6754 + }, + { + "epoch": 0.7323287077189939, + "grad_norm": 8.458971977233887, + "learning_rate": 1.5172638909038156e-05, + "loss": 0.057, + "step": 6755 + }, + { + "epoch": 0.7324371205550737, + "grad_norm": 6.815972805023193, + "learning_rate": 1.5171913535470768e-05, + "loss": 0.0549, + "step": 6756 + }, + { + "epoch": 0.7325455333911535, + "grad_norm": 2.3126220703125, + "learning_rate": 1.517118816190338e-05, + "loss": 0.0282, + "step": 6757 + }, + { + "epoch": 0.7326539462272333, + "grad_norm": 2.937925338745117, + "learning_rate": 1.5170462788335994e-05, + "loss": 0.0442, + "step": 6758 + }, + { + "epoch": 0.7327623590633131, + "grad_norm": 1.8214120864868164, + "learning_rate": 1.5169737414768606e-05, + "loss": 0.029, + "step": 6759 + }, + { + "epoch": 0.7328707718993929, + "grad_norm": 0.608147382736206, + "learning_rate": 1.516901204120122e-05, + "loss": 0.0092, + "step": 6760 + }, + { + "epoch": 0.7329791847354726, + "grad_norm": 2.187730073928833, + "learning_rate": 1.5168286667633834e-05, + "loss": 0.0358, + "step": 6761 + }, + { + "epoch": 0.7330875975715525, + "grad_norm": 2.2502527236938477, + "learning_rate": 1.5167561294066446e-05, + "loss": 0.0439, + "step": 6762 + }, + { + "epoch": 0.7331960104076323, + "grad_norm": 3.045053005218506, + "learning_rate": 1.5166835920499058e-05, + "loss": 0.029, + "step": 6763 + }, + { + "epoch": 0.733304423243712, + "grad_norm": 2.701554536819458, + "learning_rate": 1.5166110546931672e-05, + "loss": 0.0193, + "step": 6764 + }, + { + "epoch": 0.7334128360797918, + "grad_norm": 1.6401761770248413, + "learning_rate": 1.5165385173364284e-05, + "loss": 0.0285, + "step": 6765 + }, + { + "epoch": 0.7335212489158717, + "grad_norm": 1.929929494857788, + "learning_rate": 1.5164659799796896e-05, + "loss": 0.0278, + "step": 6766 + }, + { + "epoch": 0.7336296617519514, + "grad_norm": 0.971096932888031, + "learning_rate": 1.516393442622951e-05, + "loss": 0.0142, + "step": 6767 + }, + { + "epoch": 0.7337380745880312, + "grad_norm": 5.375793933868408, + "learning_rate": 1.5163209052662122e-05, + "loss": 0.0304, + "step": 6768 + }, + { + "epoch": 0.733846487424111, + "grad_norm": 3.3048131465911865, + "learning_rate": 1.5162483679094735e-05, + "loss": 0.0214, + "step": 6769 + }, + { + "epoch": 0.7339549002601908, + "grad_norm": 1.6129581928253174, + "learning_rate": 1.5161758305527348e-05, + "loss": 0.0185, + "step": 6770 + }, + { + "epoch": 0.7340633130962706, + "grad_norm": 1.319019079208374, + "learning_rate": 1.516103293195996e-05, + "loss": 0.0047, + "step": 6771 + }, + { + "epoch": 0.7341717259323504, + "grad_norm": 1.9892250299453735, + "learning_rate": 1.5160307558392573e-05, + "loss": 0.0456, + "step": 6772 + }, + { + "epoch": 0.7342801387684302, + "grad_norm": 3.088864326477051, + "learning_rate": 1.5159582184825185e-05, + "loss": 0.0321, + "step": 6773 + }, + { + "epoch": 0.73438855160451, + "grad_norm": 4.342159271240234, + "learning_rate": 1.5158856811257799e-05, + "loss": 0.098, + "step": 6774 + }, + { + "epoch": 0.7344969644405898, + "grad_norm": 1.9423121213912964, + "learning_rate": 1.515813143769041e-05, + "loss": 0.0368, + "step": 6775 + }, + { + "epoch": 0.7346053772766695, + "grad_norm": 5.05132532119751, + "learning_rate": 1.5157406064123023e-05, + "loss": 0.0403, + "step": 6776 + }, + { + "epoch": 0.7347137901127494, + "grad_norm": 3.3668339252471924, + "learning_rate": 1.5156680690555637e-05, + "loss": 0.0608, + "step": 6777 + }, + { + "epoch": 0.7348222029488292, + "grad_norm": 8.82043170928955, + "learning_rate": 1.515595531698825e-05, + "loss": 0.0465, + "step": 6778 + }, + { + "epoch": 0.7349306157849089, + "grad_norm": 7.408441066741943, + "learning_rate": 1.5155229943420864e-05, + "loss": 0.054, + "step": 6779 + }, + { + "epoch": 0.7350390286209887, + "grad_norm": 3.5708272457122803, + "learning_rate": 1.5154504569853477e-05, + "loss": 0.0727, + "step": 6780 + }, + { + "epoch": 0.7351474414570686, + "grad_norm": 2.168126344680786, + "learning_rate": 1.5153779196286089e-05, + "loss": 0.0192, + "step": 6781 + }, + { + "epoch": 0.7352558542931483, + "grad_norm": 1.7935357093811035, + "learning_rate": 1.5153053822718703e-05, + "loss": 0.0168, + "step": 6782 + }, + { + "epoch": 0.7353642671292281, + "grad_norm": 1.0476806163787842, + "learning_rate": 1.5152328449151315e-05, + "loss": 0.0142, + "step": 6783 + }, + { + "epoch": 0.7354726799653079, + "grad_norm": 1.7736330032348633, + "learning_rate": 1.5151603075583927e-05, + "loss": 0.0271, + "step": 6784 + }, + { + "epoch": 0.7355810928013877, + "grad_norm": 3.9798028469085693, + "learning_rate": 1.5150877702016539e-05, + "loss": 0.0829, + "step": 6785 + }, + { + "epoch": 0.7356895056374675, + "grad_norm": 3.976670265197754, + "learning_rate": 1.5150152328449153e-05, + "loss": 0.027, + "step": 6786 + }, + { + "epoch": 0.7357979184735473, + "grad_norm": 1.9545964002609253, + "learning_rate": 1.5149426954881765e-05, + "loss": 0.0268, + "step": 6787 + }, + { + "epoch": 0.735906331309627, + "grad_norm": 2.066730499267578, + "learning_rate": 1.5148701581314377e-05, + "loss": 0.0222, + "step": 6788 + }, + { + "epoch": 0.7360147441457069, + "grad_norm": 2.2015397548675537, + "learning_rate": 1.5147976207746991e-05, + "loss": 0.0383, + "step": 6789 + }, + { + "epoch": 0.7361231569817867, + "grad_norm": 4.759432792663574, + "learning_rate": 1.5147250834179603e-05, + "loss": 0.037, + "step": 6790 + }, + { + "epoch": 0.7362315698178664, + "grad_norm": 2.333857297897339, + "learning_rate": 1.5146525460612215e-05, + "loss": 0.021, + "step": 6791 + }, + { + "epoch": 0.7363399826539462, + "grad_norm": 8.203234672546387, + "learning_rate": 1.5145800087044829e-05, + "loss": 0.0738, + "step": 6792 + }, + { + "epoch": 0.736448395490026, + "grad_norm": 2.804830551147461, + "learning_rate": 1.5145074713477441e-05, + "loss": 0.0377, + "step": 6793 + }, + { + "epoch": 0.7365568083261058, + "grad_norm": 4.018276691436768, + "learning_rate": 1.5144349339910053e-05, + "loss": 0.0177, + "step": 6794 + }, + { + "epoch": 0.7366652211621856, + "grad_norm": 1.8953814506530762, + "learning_rate": 1.5143623966342669e-05, + "loss": 0.0251, + "step": 6795 + }, + { + "epoch": 0.7367736339982653, + "grad_norm": 1.4443950653076172, + "learning_rate": 1.5142898592775281e-05, + "loss": 0.0389, + "step": 6796 + }, + { + "epoch": 0.7368820468343452, + "grad_norm": 3.2487549781799316, + "learning_rate": 1.5142173219207895e-05, + "loss": 0.0389, + "step": 6797 + }, + { + "epoch": 0.736990459670425, + "grad_norm": 3.9380104541778564, + "learning_rate": 1.5141447845640507e-05, + "loss": 0.0331, + "step": 6798 + }, + { + "epoch": 0.7370988725065047, + "grad_norm": 3.445918321609497, + "learning_rate": 1.514072247207312e-05, + "loss": 0.0808, + "step": 6799 + }, + { + "epoch": 0.7372072853425846, + "grad_norm": 3.38729190826416, + "learning_rate": 1.5139997098505731e-05, + "loss": 0.0235, + "step": 6800 + }, + { + "epoch": 0.7373156981786644, + "grad_norm": 1.2759398221969604, + "learning_rate": 1.5139271724938345e-05, + "loss": 0.0268, + "step": 6801 + }, + { + "epoch": 0.7374241110147441, + "grad_norm": 1.739288568496704, + "learning_rate": 1.5138546351370957e-05, + "loss": 0.0391, + "step": 6802 + }, + { + "epoch": 0.7375325238508239, + "grad_norm": 2.7893214225769043, + "learning_rate": 1.513782097780357e-05, + "loss": 0.0431, + "step": 6803 + }, + { + "epoch": 0.7376409366869038, + "grad_norm": 1.2981083393096924, + "learning_rate": 1.5137095604236183e-05, + "loss": 0.0187, + "step": 6804 + }, + { + "epoch": 0.7377493495229835, + "grad_norm": 3.1003053188323975, + "learning_rate": 1.5136370230668795e-05, + "loss": 0.0941, + "step": 6805 + }, + { + "epoch": 0.7378577623590633, + "grad_norm": 1.574996829032898, + "learning_rate": 1.5135644857101408e-05, + "loss": 0.0153, + "step": 6806 + }, + { + "epoch": 0.7379661751951431, + "grad_norm": 2.0004160404205322, + "learning_rate": 1.5134919483534021e-05, + "loss": 0.0293, + "step": 6807 + }, + { + "epoch": 0.738074588031223, + "grad_norm": 3.2947194576263428, + "learning_rate": 1.5134194109966634e-05, + "loss": 0.0172, + "step": 6808 + }, + { + "epoch": 0.7381830008673027, + "grad_norm": 2.551668882369995, + "learning_rate": 1.5133468736399246e-05, + "loss": 0.0688, + "step": 6809 + }, + { + "epoch": 0.7382914137033825, + "grad_norm": 4.103377342224121, + "learning_rate": 1.513274336283186e-05, + "loss": 0.0509, + "step": 6810 + }, + { + "epoch": 0.7383998265394622, + "grad_norm": 4.927648067474365, + "learning_rate": 1.5132017989264472e-05, + "loss": 0.0582, + "step": 6811 + }, + { + "epoch": 0.7385082393755421, + "grad_norm": 1.1152660846710205, + "learning_rate": 1.5131292615697086e-05, + "loss": 0.0112, + "step": 6812 + }, + { + "epoch": 0.7386166522116219, + "grad_norm": 3.485210418701172, + "learning_rate": 1.51305672421297e-05, + "loss": 0.0393, + "step": 6813 + }, + { + "epoch": 0.7387250650477016, + "grad_norm": 3.2957379817962646, + "learning_rate": 1.5129841868562312e-05, + "loss": 0.0722, + "step": 6814 + }, + { + "epoch": 0.7388334778837814, + "grad_norm": 4.602722644805908, + "learning_rate": 1.5129116494994924e-05, + "loss": 0.0599, + "step": 6815 + }, + { + "epoch": 0.7389418907198613, + "grad_norm": 0.973288357257843, + "learning_rate": 1.5128391121427537e-05, + "loss": 0.0118, + "step": 6816 + }, + { + "epoch": 0.739050303555941, + "grad_norm": 1.798851490020752, + "learning_rate": 1.512766574786015e-05, + "loss": 0.0408, + "step": 6817 + }, + { + "epoch": 0.7391587163920208, + "grad_norm": 1.3930848836898804, + "learning_rate": 1.5126940374292762e-05, + "loss": 0.014, + "step": 6818 + }, + { + "epoch": 0.7392671292281006, + "grad_norm": 3.1834070682525635, + "learning_rate": 1.5126215000725376e-05, + "loss": 0.0526, + "step": 6819 + }, + { + "epoch": 0.7393755420641804, + "grad_norm": 1.4835104942321777, + "learning_rate": 1.5125489627157988e-05, + "loss": 0.0356, + "step": 6820 + }, + { + "epoch": 0.7394839549002602, + "grad_norm": 3.0060839653015137, + "learning_rate": 1.51247642535906e-05, + "loss": 0.0375, + "step": 6821 + }, + { + "epoch": 0.73959236773634, + "grad_norm": 2.8286452293395996, + "learning_rate": 1.5124038880023214e-05, + "loss": 0.0585, + "step": 6822 + }, + { + "epoch": 0.7397007805724197, + "grad_norm": 2.501243829727173, + "learning_rate": 1.5123313506455826e-05, + "loss": 0.0245, + "step": 6823 + }, + { + "epoch": 0.7398091934084996, + "grad_norm": 4.842895030975342, + "learning_rate": 1.5122588132888438e-05, + "loss": 0.0609, + "step": 6824 + }, + { + "epoch": 0.7399176062445794, + "grad_norm": 2.0219388008117676, + "learning_rate": 1.512186275932105e-05, + "loss": 0.0336, + "step": 6825 + }, + { + "epoch": 0.7400260190806591, + "grad_norm": 1.461854338645935, + "learning_rate": 1.5121137385753664e-05, + "loss": 0.0078, + "step": 6826 + }, + { + "epoch": 0.7401344319167389, + "grad_norm": 1.6615092754364014, + "learning_rate": 1.5120412012186276e-05, + "loss": 0.0362, + "step": 6827 + }, + { + "epoch": 0.7402428447528188, + "grad_norm": 5.191235542297363, + "learning_rate": 1.5119686638618888e-05, + "loss": 0.0748, + "step": 6828 + }, + { + "epoch": 0.7403512575888985, + "grad_norm": 3.5187175273895264, + "learning_rate": 1.5118961265051504e-05, + "loss": 0.0598, + "step": 6829 + }, + { + "epoch": 0.7404596704249783, + "grad_norm": 2.379772424697876, + "learning_rate": 1.5118235891484116e-05, + "loss": 0.0432, + "step": 6830 + }, + { + "epoch": 0.7405680832610582, + "grad_norm": 2.592942476272583, + "learning_rate": 1.511751051791673e-05, + "loss": 0.0805, + "step": 6831 + }, + { + "epoch": 0.7406764960971379, + "grad_norm": 3.4958975315093994, + "learning_rate": 1.5116785144349342e-05, + "loss": 0.0522, + "step": 6832 + }, + { + "epoch": 0.7407849089332177, + "grad_norm": 2.5918333530426025, + "learning_rate": 1.5116059770781954e-05, + "loss": 0.0164, + "step": 6833 + }, + { + "epoch": 0.7408933217692975, + "grad_norm": 0.9230601787567139, + "learning_rate": 1.5115334397214568e-05, + "loss": 0.0076, + "step": 6834 + }, + { + "epoch": 0.7410017346053773, + "grad_norm": 1.1307947635650635, + "learning_rate": 1.511460902364718e-05, + "loss": 0.0071, + "step": 6835 + }, + { + "epoch": 0.7411101474414571, + "grad_norm": 1.4608463048934937, + "learning_rate": 1.5113883650079792e-05, + "loss": 0.0149, + "step": 6836 + }, + { + "epoch": 0.7412185602775369, + "grad_norm": 0.9344396591186523, + "learning_rate": 1.5113158276512404e-05, + "loss": 0.0116, + "step": 6837 + }, + { + "epoch": 0.7413269731136166, + "grad_norm": 1.3922698497772217, + "learning_rate": 1.5112432902945018e-05, + "loss": 0.0191, + "step": 6838 + }, + { + "epoch": 0.7414353859496965, + "grad_norm": 1.4234819412231445, + "learning_rate": 1.511170752937763e-05, + "loss": 0.0275, + "step": 6839 + }, + { + "epoch": 0.7415437987857763, + "grad_norm": 2.654792070388794, + "learning_rate": 1.5110982155810242e-05, + "loss": 0.0136, + "step": 6840 + }, + { + "epoch": 0.741652211621856, + "grad_norm": 1.6284167766571045, + "learning_rate": 1.5110256782242856e-05, + "loss": 0.019, + "step": 6841 + }, + { + "epoch": 0.7417606244579358, + "grad_norm": 2.1164751052856445, + "learning_rate": 1.5109531408675468e-05, + "loss": 0.0375, + "step": 6842 + }, + { + "epoch": 0.7418690372940157, + "grad_norm": 3.342590570449829, + "learning_rate": 1.510880603510808e-05, + "loss": 0.0449, + "step": 6843 + }, + { + "epoch": 0.7419774501300954, + "grad_norm": 2.552125930786133, + "learning_rate": 1.5108080661540694e-05, + "loss": 0.0581, + "step": 6844 + }, + { + "epoch": 0.7420858629661752, + "grad_norm": 1.7454020977020264, + "learning_rate": 1.5107355287973307e-05, + "loss": 0.0234, + "step": 6845 + }, + { + "epoch": 0.7421942758022549, + "grad_norm": 2.7433059215545654, + "learning_rate": 1.5106629914405919e-05, + "loss": 0.0511, + "step": 6846 + }, + { + "epoch": 0.7423026886383348, + "grad_norm": 4.9599080085754395, + "learning_rate": 1.5105904540838534e-05, + "loss": 0.0526, + "step": 6847 + }, + { + "epoch": 0.7424111014744146, + "grad_norm": 1.964926838874817, + "learning_rate": 1.5105179167271146e-05, + "loss": 0.0426, + "step": 6848 + }, + { + "epoch": 0.7425195143104943, + "grad_norm": 1.8257900476455688, + "learning_rate": 1.510445379370376e-05, + "loss": 0.0266, + "step": 6849 + }, + { + "epoch": 0.7426279271465741, + "grad_norm": 2.8428242206573486, + "learning_rate": 1.5103728420136372e-05, + "loss": 0.0596, + "step": 6850 + }, + { + "epoch": 0.742736339982654, + "grad_norm": 2.2884113788604736, + "learning_rate": 1.5103003046568985e-05, + "loss": 0.0265, + "step": 6851 + }, + { + "epoch": 0.7428447528187337, + "grad_norm": 2.691239595413208, + "learning_rate": 1.5102277673001597e-05, + "loss": 0.0219, + "step": 6852 + }, + { + "epoch": 0.7429531656548135, + "grad_norm": 6.10150146484375, + "learning_rate": 1.510155229943421e-05, + "loss": 0.0619, + "step": 6853 + }, + { + "epoch": 0.7430615784908933, + "grad_norm": 5.171630859375, + "learning_rate": 1.5100826925866823e-05, + "loss": 0.0515, + "step": 6854 + }, + { + "epoch": 0.7431699913269731, + "grad_norm": 3.6386091709136963, + "learning_rate": 1.5100101552299435e-05, + "loss": 0.044, + "step": 6855 + }, + { + "epoch": 0.7432784041630529, + "grad_norm": 4.76205587387085, + "learning_rate": 1.5099376178732049e-05, + "loss": 0.047, + "step": 6856 + }, + { + "epoch": 0.7433868169991327, + "grad_norm": 4.027026653289795, + "learning_rate": 1.509865080516466e-05, + "loss": 0.0501, + "step": 6857 + }, + { + "epoch": 0.7434952298352124, + "grad_norm": 5.909848690032959, + "learning_rate": 1.5097925431597273e-05, + "loss": 0.0322, + "step": 6858 + }, + { + "epoch": 0.7436036426712923, + "grad_norm": 2.8652989864349365, + "learning_rate": 1.5097200058029887e-05, + "loss": 0.0495, + "step": 6859 + }, + { + "epoch": 0.7437120555073721, + "grad_norm": 3.17879056930542, + "learning_rate": 1.5096474684462499e-05, + "loss": 0.0205, + "step": 6860 + }, + { + "epoch": 0.7438204683434518, + "grad_norm": 2.2017672061920166, + "learning_rate": 1.5095749310895111e-05, + "loss": 0.044, + "step": 6861 + }, + { + "epoch": 0.7439288811795317, + "grad_norm": 1.5246944427490234, + "learning_rate": 1.5095023937327725e-05, + "loss": 0.0147, + "step": 6862 + }, + { + "epoch": 0.7440372940156115, + "grad_norm": 1.8682267665863037, + "learning_rate": 1.5094298563760337e-05, + "loss": 0.0204, + "step": 6863 + }, + { + "epoch": 0.7441457068516912, + "grad_norm": 1.5657494068145752, + "learning_rate": 1.5093573190192951e-05, + "loss": 0.0182, + "step": 6864 + }, + { + "epoch": 0.744254119687771, + "grad_norm": 1.1791287660598755, + "learning_rate": 1.5092847816625565e-05, + "loss": 0.0142, + "step": 6865 + }, + { + "epoch": 0.7443625325238509, + "grad_norm": 2.823843240737915, + "learning_rate": 1.5092122443058177e-05, + "loss": 0.0683, + "step": 6866 + }, + { + "epoch": 0.7444709453599306, + "grad_norm": 4.43165397644043, + "learning_rate": 1.5091397069490789e-05, + "loss": 0.0318, + "step": 6867 + }, + { + "epoch": 0.7445793581960104, + "grad_norm": 2.9625213146209717, + "learning_rate": 1.5090671695923403e-05, + "loss": 0.0421, + "step": 6868 + }, + { + "epoch": 0.7446877710320902, + "grad_norm": 1.871767282485962, + "learning_rate": 1.5089946322356015e-05, + "loss": 0.017, + "step": 6869 + }, + { + "epoch": 0.74479618386817, + "grad_norm": 2.5552186965942383, + "learning_rate": 1.5089220948788627e-05, + "loss": 0.0483, + "step": 6870 + }, + { + "epoch": 0.7449045967042498, + "grad_norm": 6.702445030212402, + "learning_rate": 1.5088495575221241e-05, + "loss": 0.0624, + "step": 6871 + }, + { + "epoch": 0.7450130095403296, + "grad_norm": 0.9813100695610046, + "learning_rate": 1.5087770201653853e-05, + "loss": 0.0095, + "step": 6872 + }, + { + "epoch": 0.7451214223764093, + "grad_norm": 2.028874635696411, + "learning_rate": 1.5087044828086465e-05, + "loss": 0.0501, + "step": 6873 + }, + { + "epoch": 0.7452298352124892, + "grad_norm": 2.6757476329803467, + "learning_rate": 1.5086319454519079e-05, + "loss": 0.0419, + "step": 6874 + }, + { + "epoch": 0.745338248048569, + "grad_norm": 2.431058406829834, + "learning_rate": 1.5085594080951691e-05, + "loss": 0.0316, + "step": 6875 + }, + { + "epoch": 0.7454466608846487, + "grad_norm": 2.6479485034942627, + "learning_rate": 1.5084868707384303e-05, + "loss": 0.0614, + "step": 6876 + }, + { + "epoch": 0.7455550737207285, + "grad_norm": 1.9924577474594116, + "learning_rate": 1.5084143333816916e-05, + "loss": 0.0283, + "step": 6877 + }, + { + "epoch": 0.7456634865568084, + "grad_norm": 2.33935546875, + "learning_rate": 1.508341796024953e-05, + "loss": 0.0467, + "step": 6878 + }, + { + "epoch": 0.7457718993928881, + "grad_norm": 1.8473480939865112, + "learning_rate": 1.5082692586682141e-05, + "loss": 0.0226, + "step": 6879 + }, + { + "epoch": 0.7458803122289679, + "grad_norm": 1.3642151355743408, + "learning_rate": 1.5081967213114754e-05, + "loss": 0.0126, + "step": 6880 + }, + { + "epoch": 0.7459887250650477, + "grad_norm": 4.725621700286865, + "learning_rate": 1.508124183954737e-05, + "loss": 0.072, + "step": 6881 + }, + { + "epoch": 0.7460971379011275, + "grad_norm": 2.772871971130371, + "learning_rate": 1.5080516465979981e-05, + "loss": 0.034, + "step": 6882 + }, + { + "epoch": 0.7462055507372073, + "grad_norm": 2.6124234199523926, + "learning_rate": 1.5079791092412595e-05, + "loss": 0.0512, + "step": 6883 + }, + { + "epoch": 0.746313963573287, + "grad_norm": 3.068729877471924, + "learning_rate": 1.5079065718845207e-05, + "loss": 0.0355, + "step": 6884 + }, + { + "epoch": 0.7464223764093668, + "grad_norm": 2.9535632133483887, + "learning_rate": 1.507834034527782e-05, + "loss": 0.0262, + "step": 6885 + }, + { + "epoch": 0.7465307892454467, + "grad_norm": 3.331798791885376, + "learning_rate": 1.5077614971710433e-05, + "loss": 0.0075, + "step": 6886 + }, + { + "epoch": 0.7466392020815265, + "grad_norm": 2.119851589202881, + "learning_rate": 1.5076889598143045e-05, + "loss": 0.0219, + "step": 6887 + }, + { + "epoch": 0.7467476149176062, + "grad_norm": 2.480529546737671, + "learning_rate": 1.5076164224575658e-05, + "loss": 0.0333, + "step": 6888 + }, + { + "epoch": 0.7468560277536861, + "grad_norm": 2.5569348335266113, + "learning_rate": 1.507543885100827e-05, + "loss": 0.0182, + "step": 6889 + }, + { + "epoch": 0.7469644405897659, + "grad_norm": 3.113492965698242, + "learning_rate": 1.5074713477440884e-05, + "loss": 0.0373, + "step": 6890 + }, + { + "epoch": 0.7470728534258456, + "grad_norm": 1.728391170501709, + "learning_rate": 1.5073988103873496e-05, + "loss": 0.0349, + "step": 6891 + }, + { + "epoch": 0.7471812662619254, + "grad_norm": 5.859001159667969, + "learning_rate": 1.5073262730306108e-05, + "loss": 0.0796, + "step": 6892 + }, + { + "epoch": 0.7472896790980053, + "grad_norm": 3.0531890392303467, + "learning_rate": 1.5072537356738722e-05, + "loss": 0.0467, + "step": 6893 + }, + { + "epoch": 0.747398091934085, + "grad_norm": 3.312892198562622, + "learning_rate": 1.5071811983171334e-05, + "loss": 0.0508, + "step": 6894 + }, + { + "epoch": 0.7475065047701648, + "grad_norm": 2.1224873065948486, + "learning_rate": 1.5071086609603946e-05, + "loss": 0.0211, + "step": 6895 + }, + { + "epoch": 0.7476149176062445, + "grad_norm": 3.2671754360198975, + "learning_rate": 1.507036123603656e-05, + "loss": 0.047, + "step": 6896 + }, + { + "epoch": 0.7477233304423244, + "grad_norm": 3.5533456802368164, + "learning_rate": 1.5069635862469172e-05, + "loss": 0.0674, + "step": 6897 + }, + { + "epoch": 0.7478317432784042, + "grad_norm": 2.4724373817443848, + "learning_rate": 1.5068910488901787e-05, + "loss": 0.0518, + "step": 6898 + }, + { + "epoch": 0.747940156114484, + "grad_norm": 3.3518292903900146, + "learning_rate": 1.50681851153344e-05, + "loss": 0.0993, + "step": 6899 + }, + { + "epoch": 0.7480485689505637, + "grad_norm": 4.963350296020508, + "learning_rate": 1.5067459741767012e-05, + "loss": 0.0332, + "step": 6900 + }, + { + "epoch": 0.7481569817866436, + "grad_norm": 4.428585529327393, + "learning_rate": 1.5066734368199626e-05, + "loss": 0.0873, + "step": 6901 + }, + { + "epoch": 0.7482653946227233, + "grad_norm": 3.5905539989471436, + "learning_rate": 1.5066008994632238e-05, + "loss": 0.055, + "step": 6902 + }, + { + "epoch": 0.7483738074588031, + "grad_norm": 2.354126453399658, + "learning_rate": 1.506528362106485e-05, + "loss": 0.0314, + "step": 6903 + }, + { + "epoch": 0.7484822202948829, + "grad_norm": 1.4569029808044434, + "learning_rate": 1.5064558247497462e-05, + "loss": 0.0424, + "step": 6904 + }, + { + "epoch": 0.7485906331309627, + "grad_norm": 4.030426979064941, + "learning_rate": 1.5063832873930076e-05, + "loss": 0.0663, + "step": 6905 + }, + { + "epoch": 0.7486990459670425, + "grad_norm": 2.8004775047302246, + "learning_rate": 1.5063107500362688e-05, + "loss": 0.0701, + "step": 6906 + }, + { + "epoch": 0.7488074588031223, + "grad_norm": 0.5033249855041504, + "learning_rate": 1.50623821267953e-05, + "loss": 0.0093, + "step": 6907 + }, + { + "epoch": 0.748915871639202, + "grad_norm": 2.95475435256958, + "learning_rate": 1.5061656753227914e-05, + "loss": 0.0509, + "step": 6908 + }, + { + "epoch": 0.7490242844752819, + "grad_norm": 2.4639852046966553, + "learning_rate": 1.5060931379660526e-05, + "loss": 0.0664, + "step": 6909 + }, + { + "epoch": 0.7491326973113617, + "grad_norm": 1.3079288005828857, + "learning_rate": 1.5060206006093138e-05, + "loss": 0.0291, + "step": 6910 + }, + { + "epoch": 0.7492411101474414, + "grad_norm": 1.6882988214492798, + "learning_rate": 1.5059480632525752e-05, + "loss": 0.0162, + "step": 6911 + }, + { + "epoch": 0.7493495229835212, + "grad_norm": 4.285794734954834, + "learning_rate": 1.5058755258958364e-05, + "loss": 0.0294, + "step": 6912 + }, + { + "epoch": 0.7494579358196011, + "grad_norm": 2.597104549407959, + "learning_rate": 1.5058029885390976e-05, + "loss": 0.0498, + "step": 6913 + }, + { + "epoch": 0.7495663486556808, + "grad_norm": 2.134699821472168, + "learning_rate": 1.505730451182359e-05, + "loss": 0.0308, + "step": 6914 + }, + { + "epoch": 0.7496747614917606, + "grad_norm": 5.598158359527588, + "learning_rate": 1.5056579138256202e-05, + "loss": 0.0412, + "step": 6915 + }, + { + "epoch": 0.7497831743278404, + "grad_norm": 2.207244873046875, + "learning_rate": 1.5055853764688816e-05, + "loss": 0.0719, + "step": 6916 + }, + { + "epoch": 0.7498915871639202, + "grad_norm": 4.071229457855225, + "learning_rate": 1.505512839112143e-05, + "loss": 0.0682, + "step": 6917 + }, + { + "epoch": 0.75, + "grad_norm": 1.6214507818222046, + "learning_rate": 1.5054403017554042e-05, + "loss": 0.0335, + "step": 6918 + }, + { + "epoch": 0.7501084128360798, + "grad_norm": 1.1157615184783936, + "learning_rate": 1.5053677643986654e-05, + "loss": 0.0161, + "step": 6919 + }, + { + "epoch": 0.7502168256721596, + "grad_norm": 1.6097440719604492, + "learning_rate": 1.5052952270419268e-05, + "loss": 0.0323, + "step": 6920 + }, + { + "epoch": 0.7503252385082394, + "grad_norm": 0.8952956199645996, + "learning_rate": 1.505222689685188e-05, + "loss": 0.0065, + "step": 6921 + }, + { + "epoch": 0.7504336513443192, + "grad_norm": 2.4839813709259033, + "learning_rate": 1.5051501523284492e-05, + "loss": 0.0723, + "step": 6922 + }, + { + "epoch": 0.7505420641803989, + "grad_norm": 3.9375884532928467, + "learning_rate": 1.5050776149717106e-05, + "loss": 0.0646, + "step": 6923 + }, + { + "epoch": 0.7506504770164788, + "grad_norm": 2.1522176265716553, + "learning_rate": 1.5050050776149718e-05, + "loss": 0.0492, + "step": 6924 + }, + { + "epoch": 0.7507588898525586, + "grad_norm": 1.250903606414795, + "learning_rate": 1.504932540258233e-05, + "loss": 0.02, + "step": 6925 + }, + { + "epoch": 0.7508673026886383, + "grad_norm": 1.830684781074524, + "learning_rate": 1.5048600029014944e-05, + "loss": 0.0211, + "step": 6926 + }, + { + "epoch": 0.7509757155247181, + "grad_norm": 0.8466300964355469, + "learning_rate": 1.5047874655447557e-05, + "loss": 0.017, + "step": 6927 + }, + { + "epoch": 0.751084128360798, + "grad_norm": 3.7002782821655273, + "learning_rate": 1.5047149281880169e-05, + "loss": 0.0472, + "step": 6928 + }, + { + "epoch": 0.7511925411968777, + "grad_norm": 1.9830766916275024, + "learning_rate": 1.5046423908312781e-05, + "loss": 0.0551, + "step": 6929 + }, + { + "epoch": 0.7513009540329575, + "grad_norm": 0.4268850088119507, + "learning_rate": 1.5045698534745395e-05, + "loss": 0.0039, + "step": 6930 + }, + { + "epoch": 0.7514093668690373, + "grad_norm": 1.68601393699646, + "learning_rate": 1.5044973161178007e-05, + "loss": 0.0327, + "step": 6931 + }, + { + "epoch": 0.7515177797051171, + "grad_norm": 2.090752124786377, + "learning_rate": 1.5044247787610619e-05, + "loss": 0.0315, + "step": 6932 + }, + { + "epoch": 0.7516261925411969, + "grad_norm": 6.636327743530273, + "learning_rate": 1.5043522414043235e-05, + "loss": 0.0903, + "step": 6933 + }, + { + "epoch": 0.7517346053772767, + "grad_norm": 0.801178514957428, + "learning_rate": 1.5042797040475847e-05, + "loss": 0.008, + "step": 6934 + }, + { + "epoch": 0.7518430182133564, + "grad_norm": 0.41737815737724304, + "learning_rate": 1.504207166690846e-05, + "loss": 0.0042, + "step": 6935 + }, + { + "epoch": 0.7519514310494363, + "grad_norm": 2.7248663902282715, + "learning_rate": 1.5041346293341073e-05, + "loss": 0.0445, + "step": 6936 + }, + { + "epoch": 0.752059843885516, + "grad_norm": 0.48417261242866516, + "learning_rate": 1.5040620919773685e-05, + "loss": 0.0173, + "step": 6937 + }, + { + "epoch": 0.7521682567215958, + "grad_norm": 1.714778184890747, + "learning_rate": 1.5039895546206299e-05, + "loss": 0.0409, + "step": 6938 + }, + { + "epoch": 0.7522766695576756, + "grad_norm": 4.180802822113037, + "learning_rate": 1.503917017263891e-05, + "loss": 0.0369, + "step": 6939 + }, + { + "epoch": 0.7523850823937555, + "grad_norm": 3.5136358737945557, + "learning_rate": 1.5038444799071523e-05, + "loss": 0.07, + "step": 6940 + }, + { + "epoch": 0.7524934952298352, + "grad_norm": 2.304281711578369, + "learning_rate": 1.5037719425504135e-05, + "loss": 0.0128, + "step": 6941 + }, + { + "epoch": 0.752601908065915, + "grad_norm": 3.0412344932556152, + "learning_rate": 1.5036994051936749e-05, + "loss": 0.0697, + "step": 6942 + }, + { + "epoch": 0.7527103209019947, + "grad_norm": 1.7026320695877075, + "learning_rate": 1.5036268678369361e-05, + "loss": 0.0258, + "step": 6943 + }, + { + "epoch": 0.7528187337380746, + "grad_norm": 2.5361125469207764, + "learning_rate": 1.5035543304801973e-05, + "loss": 0.0671, + "step": 6944 + }, + { + "epoch": 0.7529271465741544, + "grad_norm": 1.1173344850540161, + "learning_rate": 1.5034817931234587e-05, + "loss": 0.0103, + "step": 6945 + }, + { + "epoch": 0.7530355594102341, + "grad_norm": 4.47458553314209, + "learning_rate": 1.5034092557667199e-05, + "loss": 0.0524, + "step": 6946 + }, + { + "epoch": 0.7531439722463139, + "grad_norm": 3.1938469409942627, + "learning_rate": 1.5033367184099811e-05, + "loss": 0.0584, + "step": 6947 + }, + { + "epoch": 0.7532523850823938, + "grad_norm": 4.302127838134766, + "learning_rate": 1.5032641810532425e-05, + "loss": 0.0395, + "step": 6948 + }, + { + "epoch": 0.7533607979184735, + "grad_norm": 5.547867774963379, + "learning_rate": 1.5031916436965037e-05, + "loss": 0.0423, + "step": 6949 + }, + { + "epoch": 0.7534692107545533, + "grad_norm": 2.640502452850342, + "learning_rate": 1.5031191063397653e-05, + "loss": 0.0769, + "step": 6950 + }, + { + "epoch": 0.7535776235906332, + "grad_norm": 1.7646751403808594, + "learning_rate": 1.5030465689830265e-05, + "loss": 0.0498, + "step": 6951 + }, + { + "epoch": 0.753686036426713, + "grad_norm": 1.8010215759277344, + "learning_rate": 1.5029740316262877e-05, + "loss": 0.0672, + "step": 6952 + }, + { + "epoch": 0.7537944492627927, + "grad_norm": 2.486636161804199, + "learning_rate": 1.5029014942695491e-05, + "loss": 0.0209, + "step": 6953 + }, + { + "epoch": 0.7539028620988725, + "grad_norm": 2.8463242053985596, + "learning_rate": 1.5028289569128103e-05, + "loss": 0.0652, + "step": 6954 + }, + { + "epoch": 0.7540112749349523, + "grad_norm": 1.9014763832092285, + "learning_rate": 1.5027564195560715e-05, + "loss": 0.0357, + "step": 6955 + }, + { + "epoch": 0.7541196877710321, + "grad_norm": 1.3002173900604248, + "learning_rate": 1.5026838821993327e-05, + "loss": 0.0192, + "step": 6956 + }, + { + "epoch": 0.7542281006071119, + "grad_norm": 5.883694171905518, + "learning_rate": 1.5026113448425941e-05, + "loss": 0.0386, + "step": 6957 + }, + { + "epoch": 0.7543365134431916, + "grad_norm": 2.6733522415161133, + "learning_rate": 1.5025388074858553e-05, + "loss": 0.0574, + "step": 6958 + }, + { + "epoch": 0.7544449262792715, + "grad_norm": 2.7445361614227295, + "learning_rate": 1.5024662701291165e-05, + "loss": 0.0335, + "step": 6959 + }, + { + "epoch": 0.7545533391153513, + "grad_norm": 1.3237711191177368, + "learning_rate": 1.502393732772378e-05, + "loss": 0.0139, + "step": 6960 + }, + { + "epoch": 0.754661751951431, + "grad_norm": 5.7629923820495605, + "learning_rate": 1.5023211954156391e-05, + "loss": 0.0735, + "step": 6961 + }, + { + "epoch": 0.7547701647875108, + "grad_norm": 3.2543821334838867, + "learning_rate": 1.5022486580589004e-05, + "loss": 0.0511, + "step": 6962 + }, + { + "epoch": 0.7548785776235907, + "grad_norm": 3.439086437225342, + "learning_rate": 1.5021761207021617e-05, + "loss": 0.0576, + "step": 6963 + }, + { + "epoch": 0.7549869904596704, + "grad_norm": 5.733413219451904, + "learning_rate": 1.502103583345423e-05, + "loss": 0.0715, + "step": 6964 + }, + { + "epoch": 0.7550954032957502, + "grad_norm": 9.509665489196777, + "learning_rate": 1.5020310459886842e-05, + "loss": 0.0374, + "step": 6965 + }, + { + "epoch": 0.75520381613183, + "grad_norm": 2.5418038368225098, + "learning_rate": 1.5019585086319456e-05, + "loss": 0.04, + "step": 6966 + }, + { + "epoch": 0.7553122289679098, + "grad_norm": 2.3677289485931396, + "learning_rate": 1.5018859712752068e-05, + "loss": 0.0223, + "step": 6967 + }, + { + "epoch": 0.7554206418039896, + "grad_norm": 6.645124435424805, + "learning_rate": 1.5018134339184682e-05, + "loss": 0.1031, + "step": 6968 + }, + { + "epoch": 0.7555290546400694, + "grad_norm": 2.5006909370422363, + "learning_rate": 1.5017408965617295e-05, + "loss": 0.0485, + "step": 6969 + }, + { + "epoch": 0.7556374674761491, + "grad_norm": 1.2610487937927246, + "learning_rate": 1.5016683592049908e-05, + "loss": 0.0134, + "step": 6970 + }, + { + "epoch": 0.755745880312229, + "grad_norm": 0.7665954232215881, + "learning_rate": 1.501595821848252e-05, + "loss": 0.0046, + "step": 6971 + }, + { + "epoch": 0.7558542931483088, + "grad_norm": 2.327831983566284, + "learning_rate": 1.5015232844915134e-05, + "loss": 0.0311, + "step": 6972 + }, + { + "epoch": 0.7559627059843885, + "grad_norm": 1.6805483102798462, + "learning_rate": 1.5014507471347746e-05, + "loss": 0.0228, + "step": 6973 + }, + { + "epoch": 0.7560711188204683, + "grad_norm": 3.1990983486175537, + "learning_rate": 1.5013782097780358e-05, + "loss": 0.0235, + "step": 6974 + }, + { + "epoch": 0.7561795316565482, + "grad_norm": 1.1319993734359741, + "learning_rate": 1.5013056724212972e-05, + "loss": 0.0132, + "step": 6975 + }, + { + "epoch": 0.7562879444926279, + "grad_norm": 1.2949992418289185, + "learning_rate": 1.5012331350645584e-05, + "loss": 0.0364, + "step": 6976 + }, + { + "epoch": 0.7563963573287077, + "grad_norm": 3.5639357566833496, + "learning_rate": 1.5011605977078196e-05, + "loss": 0.0371, + "step": 6977 + }, + { + "epoch": 0.7565047701647876, + "grad_norm": 8.15981674194336, + "learning_rate": 1.501088060351081e-05, + "loss": 0.0576, + "step": 6978 + }, + { + "epoch": 0.7566131830008673, + "grad_norm": 2.5989749431610107, + "learning_rate": 1.5010155229943422e-05, + "loss": 0.05, + "step": 6979 + }, + { + "epoch": 0.7567215958369471, + "grad_norm": 2.572223424911499, + "learning_rate": 1.5009429856376034e-05, + "loss": 0.0244, + "step": 6980 + }, + { + "epoch": 0.7568300086730269, + "grad_norm": 4.534915447235107, + "learning_rate": 1.5008704482808646e-05, + "loss": 0.0662, + "step": 6981 + }, + { + "epoch": 0.7569384215091067, + "grad_norm": 2.0559661388397217, + "learning_rate": 1.500797910924126e-05, + "loss": 0.0486, + "step": 6982 + }, + { + "epoch": 0.7570468343451865, + "grad_norm": 5.4362335205078125, + "learning_rate": 1.5007253735673872e-05, + "loss": 0.0215, + "step": 6983 + }, + { + "epoch": 0.7571552471812663, + "grad_norm": 1.6460987329483032, + "learning_rate": 1.5006528362106484e-05, + "loss": 0.0284, + "step": 6984 + }, + { + "epoch": 0.757263660017346, + "grad_norm": 1.6265450716018677, + "learning_rate": 1.50058029885391e-05, + "loss": 0.0117, + "step": 6985 + }, + { + "epoch": 0.7573720728534259, + "grad_norm": 3.7592720985412598, + "learning_rate": 1.5005077614971712e-05, + "loss": 0.0531, + "step": 6986 + }, + { + "epoch": 0.7574804856895057, + "grad_norm": 1.61162269115448, + "learning_rate": 1.5004352241404326e-05, + "loss": 0.0243, + "step": 6987 + }, + { + "epoch": 0.7575888985255854, + "grad_norm": 2.9594104290008545, + "learning_rate": 1.5003626867836938e-05, + "loss": 0.0305, + "step": 6988 + }, + { + "epoch": 0.7576973113616652, + "grad_norm": 3.5979819297790527, + "learning_rate": 1.500290149426955e-05, + "loss": 0.0539, + "step": 6989 + }, + { + "epoch": 0.7578057241977451, + "grad_norm": 1.3644975423812866, + "learning_rate": 1.5002176120702164e-05, + "loss": 0.0232, + "step": 6990 + }, + { + "epoch": 0.7579141370338248, + "grad_norm": 3.91243052482605, + "learning_rate": 1.5001450747134776e-05, + "loss": 0.1131, + "step": 6991 + }, + { + "epoch": 0.7580225498699046, + "grad_norm": 2.6672275066375732, + "learning_rate": 1.5000725373567388e-05, + "loss": 0.0307, + "step": 6992 + }, + { + "epoch": 0.7581309627059843, + "grad_norm": 6.31015157699585, + "learning_rate": 1.5000000000000002e-05, + "loss": 0.0691, + "step": 6993 + }, + { + "epoch": 0.7582393755420642, + "grad_norm": 2.6348164081573486, + "learning_rate": 1.4999274626432614e-05, + "loss": 0.068, + "step": 6994 + }, + { + "epoch": 0.758347788378144, + "grad_norm": 0.8864377737045288, + "learning_rate": 1.4998549252865226e-05, + "loss": 0.0157, + "step": 6995 + }, + { + "epoch": 0.7584562012142237, + "grad_norm": 1.7885667085647583, + "learning_rate": 1.4997823879297839e-05, + "loss": 0.0428, + "step": 6996 + }, + { + "epoch": 0.7585646140503035, + "grad_norm": 2.6904916763305664, + "learning_rate": 1.4997098505730452e-05, + "loss": 0.0492, + "step": 6997 + }, + { + "epoch": 0.7586730268863834, + "grad_norm": 2.7868709564208984, + "learning_rate": 1.4996373132163064e-05, + "loss": 0.0333, + "step": 6998 + }, + { + "epoch": 0.7587814397224631, + "grad_norm": 1.1616328954696655, + "learning_rate": 1.4995647758595677e-05, + "loss": 0.0183, + "step": 6999 + }, + { + "epoch": 0.7588898525585429, + "grad_norm": 2.064457893371582, + "learning_rate": 1.499492238502829e-05, + "loss": 0.0663, + "step": 7000 + }, + { + "epoch": 0.7589982653946227, + "grad_norm": 2.5945911407470703, + "learning_rate": 1.4994197011460903e-05, + "loss": 0.041, + "step": 7001 + }, + { + "epoch": 0.7591066782307025, + "grad_norm": 7.343006134033203, + "learning_rate": 1.4993471637893518e-05, + "loss": 0.0561, + "step": 7002 + }, + { + "epoch": 0.7592150910667823, + "grad_norm": 1.9115451574325562, + "learning_rate": 1.499274626432613e-05, + "loss": 0.0448, + "step": 7003 + }, + { + "epoch": 0.7593235039028621, + "grad_norm": 2.090644359588623, + "learning_rate": 1.4992020890758742e-05, + "loss": 0.0274, + "step": 7004 + }, + { + "epoch": 0.7594319167389418, + "grad_norm": 2.41996169090271, + "learning_rate": 1.4991295517191356e-05, + "loss": 0.0467, + "step": 7005 + }, + { + "epoch": 0.7595403295750217, + "grad_norm": 0.46660932898521423, + "learning_rate": 1.4990570143623968e-05, + "loss": 0.0059, + "step": 7006 + }, + { + "epoch": 0.7596487424111015, + "grad_norm": 1.8066275119781494, + "learning_rate": 1.498984477005658e-05, + "loss": 0.0441, + "step": 7007 + }, + { + "epoch": 0.7597571552471812, + "grad_norm": 2.7991223335266113, + "learning_rate": 1.4989119396489193e-05, + "loss": 0.0674, + "step": 7008 + }, + { + "epoch": 0.7598655680832611, + "grad_norm": 2.3293521404266357, + "learning_rate": 1.4988394022921807e-05, + "loss": 0.0354, + "step": 7009 + }, + { + "epoch": 0.7599739809193409, + "grad_norm": 2.7237234115600586, + "learning_rate": 1.4987668649354419e-05, + "loss": 0.0466, + "step": 7010 + }, + { + "epoch": 0.7600823937554206, + "grad_norm": 3.3369243144989014, + "learning_rate": 1.498694327578703e-05, + "loss": 0.0689, + "step": 7011 + }, + { + "epoch": 0.7601908065915004, + "grad_norm": 1.539096474647522, + "learning_rate": 1.4986217902219645e-05, + "loss": 0.0176, + "step": 7012 + }, + { + "epoch": 0.7602992194275803, + "grad_norm": 3.316786766052246, + "learning_rate": 1.4985492528652257e-05, + "loss": 0.1033, + "step": 7013 + }, + { + "epoch": 0.76040763226366, + "grad_norm": 1.6196253299713135, + "learning_rate": 1.4984767155084869e-05, + "loss": 0.0203, + "step": 7014 + }, + { + "epoch": 0.7605160450997398, + "grad_norm": 1.4182476997375488, + "learning_rate": 1.4984041781517483e-05, + "loss": 0.0358, + "step": 7015 + }, + { + "epoch": 0.7606244579358196, + "grad_norm": 1.135863184928894, + "learning_rate": 1.4983316407950095e-05, + "loss": 0.0107, + "step": 7016 + }, + { + "epoch": 0.7607328707718994, + "grad_norm": 2.0654282569885254, + "learning_rate": 1.4982591034382707e-05, + "loss": 0.0407, + "step": 7017 + }, + { + "epoch": 0.7608412836079792, + "grad_norm": 1.79177987575531, + "learning_rate": 1.4981865660815321e-05, + "loss": 0.031, + "step": 7018 + }, + { + "epoch": 0.760949696444059, + "grad_norm": 2.6302483081817627, + "learning_rate": 1.4981140287247935e-05, + "loss": 0.0116, + "step": 7019 + }, + { + "epoch": 0.7610581092801387, + "grad_norm": 1.7075737714767456, + "learning_rate": 1.4980414913680547e-05, + "loss": 0.0432, + "step": 7020 + }, + { + "epoch": 0.7611665221162186, + "grad_norm": 2.0277798175811768, + "learning_rate": 1.497968954011316e-05, + "loss": 0.0321, + "step": 7021 + }, + { + "epoch": 0.7612749349522984, + "grad_norm": 1.088708519935608, + "learning_rate": 1.4978964166545773e-05, + "loss": 0.0206, + "step": 7022 + }, + { + "epoch": 0.7613833477883781, + "grad_norm": 5.102450847625732, + "learning_rate": 1.4978238792978385e-05, + "loss": 0.0643, + "step": 7023 + }, + { + "epoch": 0.7614917606244579, + "grad_norm": 4.0284013748168945, + "learning_rate": 1.4977513419410999e-05, + "loss": 0.1156, + "step": 7024 + }, + { + "epoch": 0.7616001734605378, + "grad_norm": 3.49843168258667, + "learning_rate": 1.4976788045843611e-05, + "loss": 0.051, + "step": 7025 + }, + { + "epoch": 0.7617085862966175, + "grad_norm": 1.8544772863388062, + "learning_rate": 1.4976062672276223e-05, + "loss": 0.0411, + "step": 7026 + }, + { + "epoch": 0.7618169991326973, + "grad_norm": 1.351499319076538, + "learning_rate": 1.4975337298708837e-05, + "loss": 0.0148, + "step": 7027 + }, + { + "epoch": 0.761925411968777, + "grad_norm": 1.7641880512237549, + "learning_rate": 1.4974611925141449e-05, + "loss": 0.0409, + "step": 7028 + }, + { + "epoch": 0.7620338248048569, + "grad_norm": 2.414318561553955, + "learning_rate": 1.4973886551574061e-05, + "loss": 0.032, + "step": 7029 + }, + { + "epoch": 0.7621422376409367, + "grad_norm": 0.7710043787956238, + "learning_rate": 1.4973161178006675e-05, + "loss": 0.0122, + "step": 7030 + }, + { + "epoch": 0.7622506504770165, + "grad_norm": 2.653524160385132, + "learning_rate": 1.4972435804439287e-05, + "loss": 0.0265, + "step": 7031 + }, + { + "epoch": 0.7623590633130962, + "grad_norm": 0.6909379959106445, + "learning_rate": 1.49717104308719e-05, + "loss": 0.0102, + "step": 7032 + }, + { + "epoch": 0.7624674761491761, + "grad_norm": 3.3013916015625, + "learning_rate": 1.4970985057304512e-05, + "loss": 0.0453, + "step": 7033 + }, + { + "epoch": 0.7625758889852559, + "grad_norm": 2.4934985637664795, + "learning_rate": 1.4970259683737125e-05, + "loss": 0.0702, + "step": 7034 + }, + { + "epoch": 0.7626843018213356, + "grad_norm": 2.6300675868988037, + "learning_rate": 1.4969534310169738e-05, + "loss": 0.0267, + "step": 7035 + }, + { + "epoch": 0.7627927146574154, + "grad_norm": 0.8153594136238098, + "learning_rate": 1.496880893660235e-05, + "loss": 0.0111, + "step": 7036 + }, + { + "epoch": 0.7629011274934953, + "grad_norm": 3.265002965927124, + "learning_rate": 1.4968083563034965e-05, + "loss": 0.038, + "step": 7037 + }, + { + "epoch": 0.763009540329575, + "grad_norm": 6.63699197769165, + "learning_rate": 1.4967358189467577e-05, + "loss": 0.0745, + "step": 7038 + }, + { + "epoch": 0.7631179531656548, + "grad_norm": 2.6655311584472656, + "learning_rate": 1.4966632815900191e-05, + "loss": 0.0404, + "step": 7039 + }, + { + "epoch": 0.7632263660017347, + "grad_norm": 3.3972911834716797, + "learning_rate": 1.4965907442332803e-05, + "loss": 0.0886, + "step": 7040 + }, + { + "epoch": 0.7633347788378144, + "grad_norm": 0.6852679252624512, + "learning_rate": 1.4965182068765415e-05, + "loss": 0.0144, + "step": 7041 + }, + { + "epoch": 0.7634431916738942, + "grad_norm": 4.213459014892578, + "learning_rate": 1.496445669519803e-05, + "loss": 0.0667, + "step": 7042 + }, + { + "epoch": 0.763551604509974, + "grad_norm": 5.175942420959473, + "learning_rate": 1.4963731321630641e-05, + "loss": 0.0469, + "step": 7043 + }, + { + "epoch": 0.7636600173460538, + "grad_norm": 2.912463665008545, + "learning_rate": 1.4963005948063254e-05, + "loss": 0.0391, + "step": 7044 + }, + { + "epoch": 0.7637684301821336, + "grad_norm": 3.800245761871338, + "learning_rate": 1.4962280574495867e-05, + "loss": 0.075, + "step": 7045 + }, + { + "epoch": 0.7638768430182133, + "grad_norm": 2.529370069503784, + "learning_rate": 1.496155520092848e-05, + "loss": 0.0437, + "step": 7046 + }, + { + "epoch": 0.7639852558542931, + "grad_norm": 2.652552366256714, + "learning_rate": 1.4960829827361092e-05, + "loss": 0.0194, + "step": 7047 + }, + { + "epoch": 0.764093668690373, + "grad_norm": 2.298649787902832, + "learning_rate": 1.4960104453793704e-05, + "loss": 0.0137, + "step": 7048 + }, + { + "epoch": 0.7642020815264527, + "grad_norm": 1.5427852869033813, + "learning_rate": 1.4959379080226318e-05, + "loss": 0.0333, + "step": 7049 + }, + { + "epoch": 0.7643104943625325, + "grad_norm": 0.43704214692115784, + "learning_rate": 1.495865370665893e-05, + "loss": 0.0106, + "step": 7050 + }, + { + "epoch": 0.7644189071986123, + "grad_norm": 1.5778722763061523, + "learning_rate": 1.4957928333091542e-05, + "loss": 0.0216, + "step": 7051 + }, + { + "epoch": 0.7645273200346921, + "grad_norm": 2.1977014541625977, + "learning_rate": 1.4957202959524156e-05, + "loss": 0.0284, + "step": 7052 + }, + { + "epoch": 0.7646357328707719, + "grad_norm": 2.1178529262542725, + "learning_rate": 1.4956477585956768e-05, + "loss": 0.0278, + "step": 7053 + }, + { + "epoch": 0.7647441457068517, + "grad_norm": 5.245075225830078, + "learning_rate": 1.4955752212389383e-05, + "loss": 0.0442, + "step": 7054 + }, + { + "epoch": 0.7648525585429314, + "grad_norm": 2.561068534851074, + "learning_rate": 1.4955026838821996e-05, + "loss": 0.0479, + "step": 7055 + }, + { + "epoch": 0.7649609713790113, + "grad_norm": 1.8359326124191284, + "learning_rate": 1.4954301465254608e-05, + "loss": 0.0137, + "step": 7056 + }, + { + "epoch": 0.7650693842150911, + "grad_norm": 1.6988298892974854, + "learning_rate": 1.4953576091687222e-05, + "loss": 0.0139, + "step": 7057 + }, + { + "epoch": 0.7651777970511708, + "grad_norm": 1.550779104232788, + "learning_rate": 1.4952850718119834e-05, + "loss": 0.0105, + "step": 7058 + }, + { + "epoch": 0.7652862098872506, + "grad_norm": 1.844360113143921, + "learning_rate": 1.4952125344552446e-05, + "loss": 0.0181, + "step": 7059 + }, + { + "epoch": 0.7653946227233305, + "grad_norm": 5.014837741851807, + "learning_rate": 1.4951399970985058e-05, + "loss": 0.0383, + "step": 7060 + }, + { + "epoch": 0.7655030355594102, + "grad_norm": 3.2338497638702393, + "learning_rate": 1.4950674597417672e-05, + "loss": 0.0238, + "step": 7061 + }, + { + "epoch": 0.76561144839549, + "grad_norm": 3.995302438735962, + "learning_rate": 1.4949949223850284e-05, + "loss": 0.034, + "step": 7062 + }, + { + "epoch": 0.7657198612315698, + "grad_norm": 0.5685878992080688, + "learning_rate": 1.4949223850282896e-05, + "loss": 0.0064, + "step": 7063 + }, + { + "epoch": 0.7658282740676496, + "grad_norm": 1.621475338935852, + "learning_rate": 1.494849847671551e-05, + "loss": 0.0207, + "step": 7064 + }, + { + "epoch": 0.7659366869037294, + "grad_norm": 4.983532428741455, + "learning_rate": 1.4947773103148122e-05, + "loss": 0.0629, + "step": 7065 + }, + { + "epoch": 0.7660450997398092, + "grad_norm": 1.6365941762924194, + "learning_rate": 1.4947047729580734e-05, + "loss": 0.0152, + "step": 7066 + }, + { + "epoch": 0.766153512575889, + "grad_norm": 1.1338752508163452, + "learning_rate": 1.4946322356013348e-05, + "loss": 0.0181, + "step": 7067 + }, + { + "epoch": 0.7662619254119688, + "grad_norm": 2.4090464115142822, + "learning_rate": 1.494559698244596e-05, + "loss": 0.054, + "step": 7068 + }, + { + "epoch": 0.7663703382480486, + "grad_norm": 2.1589419841766357, + "learning_rate": 1.4944871608878572e-05, + "loss": 0.0392, + "step": 7069 + }, + { + "epoch": 0.7664787510841283, + "grad_norm": 2.0887227058410645, + "learning_rate": 1.4944146235311186e-05, + "loss": 0.037, + "step": 7070 + }, + { + "epoch": 0.7665871639202082, + "grad_norm": 4.222097873687744, + "learning_rate": 1.49434208617438e-05, + "loss": 0.0228, + "step": 7071 + }, + { + "epoch": 0.766695576756288, + "grad_norm": 2.107836961746216, + "learning_rate": 1.4942695488176412e-05, + "loss": 0.0591, + "step": 7072 + }, + { + "epoch": 0.7668039895923677, + "grad_norm": 2.3523101806640625, + "learning_rate": 1.4941970114609026e-05, + "loss": 0.0291, + "step": 7073 + }, + { + "epoch": 0.7669124024284475, + "grad_norm": 2.2774322032928467, + "learning_rate": 1.4941244741041638e-05, + "loss": 0.0494, + "step": 7074 + }, + { + "epoch": 0.7670208152645274, + "grad_norm": 2.1877286434173584, + "learning_rate": 1.494051936747425e-05, + "loss": 0.0342, + "step": 7075 + }, + { + "epoch": 0.7671292281006071, + "grad_norm": 2.026221513748169, + "learning_rate": 1.4939793993906864e-05, + "loss": 0.0206, + "step": 7076 + }, + { + "epoch": 0.7672376409366869, + "grad_norm": 6.385135650634766, + "learning_rate": 1.4939068620339476e-05, + "loss": 0.0514, + "step": 7077 + }, + { + "epoch": 0.7673460537727667, + "grad_norm": 0.8969506621360779, + "learning_rate": 1.4938343246772088e-05, + "loss": 0.0179, + "step": 7078 + }, + { + "epoch": 0.7674544666088465, + "grad_norm": 4.907346725463867, + "learning_rate": 1.4937617873204702e-05, + "loss": 0.021, + "step": 7079 + }, + { + "epoch": 0.7675628794449263, + "grad_norm": 3.046205520629883, + "learning_rate": 1.4936892499637314e-05, + "loss": 0.0572, + "step": 7080 + }, + { + "epoch": 0.7676712922810061, + "grad_norm": 2.5028584003448486, + "learning_rate": 1.4936167126069927e-05, + "loss": 0.0181, + "step": 7081 + }, + { + "epoch": 0.7677797051170858, + "grad_norm": 2.029236078262329, + "learning_rate": 1.493544175250254e-05, + "loss": 0.0394, + "step": 7082 + }, + { + "epoch": 0.7678881179531657, + "grad_norm": 5.648378372192383, + "learning_rate": 1.4934716378935153e-05, + "loss": 0.0412, + "step": 7083 + }, + { + "epoch": 0.7679965307892455, + "grad_norm": 0.9984034895896912, + "learning_rate": 1.4933991005367765e-05, + "loss": 0.0056, + "step": 7084 + }, + { + "epoch": 0.7681049436253252, + "grad_norm": 5.033874034881592, + "learning_rate": 1.4933265631800377e-05, + "loss": 0.0893, + "step": 7085 + }, + { + "epoch": 0.768213356461405, + "grad_norm": 1.5927464962005615, + "learning_rate": 1.493254025823299e-05, + "loss": 0.023, + "step": 7086 + }, + { + "epoch": 0.7683217692974849, + "grad_norm": 3.8449642658233643, + "learning_rate": 1.4931814884665603e-05, + "loss": 0.0428, + "step": 7087 + }, + { + "epoch": 0.7684301821335646, + "grad_norm": 0.7548131942749023, + "learning_rate": 1.4931089511098215e-05, + "loss": 0.0048, + "step": 7088 + }, + { + "epoch": 0.7685385949696444, + "grad_norm": 2.852654457092285, + "learning_rate": 1.493036413753083e-05, + "loss": 0.0452, + "step": 7089 + }, + { + "epoch": 0.7686470078057241, + "grad_norm": 2.1586437225341797, + "learning_rate": 1.4929638763963443e-05, + "loss": 0.0173, + "step": 7090 + }, + { + "epoch": 0.768755420641804, + "grad_norm": 4.028314113616943, + "learning_rate": 1.4928913390396057e-05, + "loss": 0.1056, + "step": 7091 + }, + { + "epoch": 0.7688638334778838, + "grad_norm": 1.4614534378051758, + "learning_rate": 1.4928188016828669e-05, + "loss": 0.0159, + "step": 7092 + }, + { + "epoch": 0.7689722463139635, + "grad_norm": 1.9962416887283325, + "learning_rate": 1.492746264326128e-05, + "loss": 0.0337, + "step": 7093 + }, + { + "epoch": 0.7690806591500433, + "grad_norm": 2.708916187286377, + "learning_rate": 1.4926737269693895e-05, + "loss": 0.021, + "step": 7094 + }, + { + "epoch": 0.7691890719861232, + "grad_norm": 6.075136184692383, + "learning_rate": 1.4926011896126507e-05, + "loss": 0.0941, + "step": 7095 + }, + { + "epoch": 0.769297484822203, + "grad_norm": 0.3736647963523865, + "learning_rate": 1.4925286522559119e-05, + "loss": 0.0051, + "step": 7096 + }, + { + "epoch": 0.7694058976582827, + "grad_norm": 1.6998752355575562, + "learning_rate": 1.4924561148991733e-05, + "loss": 0.016, + "step": 7097 + }, + { + "epoch": 0.7695143104943626, + "grad_norm": 5.889617443084717, + "learning_rate": 1.4923835775424345e-05, + "loss": 0.0532, + "step": 7098 + }, + { + "epoch": 0.7696227233304423, + "grad_norm": 1.5500985383987427, + "learning_rate": 1.4923110401856957e-05, + "loss": 0.0189, + "step": 7099 + }, + { + "epoch": 0.7697311361665221, + "grad_norm": 4.8911919593811035, + "learning_rate": 1.492238502828957e-05, + "loss": 0.1205, + "step": 7100 + }, + { + "epoch": 0.7698395490026019, + "grad_norm": 1.0165715217590332, + "learning_rate": 1.4921659654722183e-05, + "loss": 0.0173, + "step": 7101 + }, + { + "epoch": 0.7699479618386817, + "grad_norm": 2.316009521484375, + "learning_rate": 1.4920934281154795e-05, + "loss": 0.0409, + "step": 7102 + }, + { + "epoch": 0.7700563746747615, + "grad_norm": 1.5225205421447754, + "learning_rate": 1.4920208907587407e-05, + "loss": 0.0098, + "step": 7103 + }, + { + "epoch": 0.7701647875108413, + "grad_norm": 3.3515849113464355, + "learning_rate": 1.4919483534020021e-05, + "loss": 0.0656, + "step": 7104 + }, + { + "epoch": 0.770273200346921, + "grad_norm": 1.908511757850647, + "learning_rate": 1.4918758160452633e-05, + "loss": 0.0528, + "step": 7105 + }, + { + "epoch": 0.7703816131830009, + "grad_norm": 0.40821054577827454, + "learning_rate": 1.4918032786885249e-05, + "loss": 0.0058, + "step": 7106 + }, + { + "epoch": 0.7704900260190807, + "grad_norm": 7.134893894195557, + "learning_rate": 1.4917307413317861e-05, + "loss": 0.049, + "step": 7107 + }, + { + "epoch": 0.7705984388551604, + "grad_norm": 6.294303894042969, + "learning_rate": 1.4916582039750473e-05, + "loss": 0.0969, + "step": 7108 + }, + { + "epoch": 0.7707068516912402, + "grad_norm": 1.170602798461914, + "learning_rate": 1.4915856666183087e-05, + "loss": 0.0076, + "step": 7109 + }, + { + "epoch": 0.7708152645273201, + "grad_norm": 8.539613723754883, + "learning_rate": 1.4915131292615699e-05, + "loss": 0.0749, + "step": 7110 + }, + { + "epoch": 0.7709236773633998, + "grad_norm": 3.997492790222168, + "learning_rate": 1.4914405919048311e-05, + "loss": 0.0565, + "step": 7111 + }, + { + "epoch": 0.7710320901994796, + "grad_norm": 4.011703968048096, + "learning_rate": 1.4913680545480923e-05, + "loss": 0.0728, + "step": 7112 + }, + { + "epoch": 0.7711405030355594, + "grad_norm": 3.3468751907348633, + "learning_rate": 1.4912955171913537e-05, + "loss": 0.0457, + "step": 7113 + }, + { + "epoch": 0.7712489158716392, + "grad_norm": 1.8305943012237549, + "learning_rate": 1.491222979834615e-05, + "loss": 0.0219, + "step": 7114 + }, + { + "epoch": 0.771357328707719, + "grad_norm": 4.629830360412598, + "learning_rate": 1.4911504424778761e-05, + "loss": 0.0519, + "step": 7115 + }, + { + "epoch": 0.7714657415437988, + "grad_norm": 3.2079010009765625, + "learning_rate": 1.4910779051211375e-05, + "loss": 0.0297, + "step": 7116 + }, + { + "epoch": 0.7715741543798785, + "grad_norm": 2.3763670921325684, + "learning_rate": 1.4910053677643987e-05, + "loss": 0.0569, + "step": 7117 + }, + { + "epoch": 0.7716825672159584, + "grad_norm": 0.6462136507034302, + "learning_rate": 1.49093283040766e-05, + "loss": 0.0083, + "step": 7118 + }, + { + "epoch": 0.7717909800520382, + "grad_norm": 0.8507588505744934, + "learning_rate": 1.4908602930509213e-05, + "loss": 0.0085, + "step": 7119 + }, + { + "epoch": 0.7718993928881179, + "grad_norm": 3.314030408859253, + "learning_rate": 1.4907877556941826e-05, + "loss": 0.025, + "step": 7120 + }, + { + "epoch": 0.7720078057241977, + "grad_norm": 6.337638854980469, + "learning_rate": 1.4907152183374438e-05, + "loss": 0.0398, + "step": 7121 + }, + { + "epoch": 0.7721162185602776, + "grad_norm": 3.582717180252075, + "learning_rate": 1.4906426809807052e-05, + "loss": 0.0545, + "step": 7122 + }, + { + "epoch": 0.7722246313963573, + "grad_norm": 3.456608295440674, + "learning_rate": 1.4905701436239665e-05, + "loss": 0.0277, + "step": 7123 + }, + { + "epoch": 0.7723330442324371, + "grad_norm": 2.6784820556640625, + "learning_rate": 1.4904976062672278e-05, + "loss": 0.0251, + "step": 7124 + }, + { + "epoch": 0.772441457068517, + "grad_norm": 1.815152645111084, + "learning_rate": 1.4904250689104891e-05, + "loss": 0.05, + "step": 7125 + }, + { + "epoch": 0.7725498699045967, + "grad_norm": 1.6004600524902344, + "learning_rate": 1.4903525315537504e-05, + "loss": 0.0144, + "step": 7126 + }, + { + "epoch": 0.7726582827406765, + "grad_norm": 2.9487571716308594, + "learning_rate": 1.4902799941970116e-05, + "loss": 0.0484, + "step": 7127 + }, + { + "epoch": 0.7727666955767563, + "grad_norm": 4.101485252380371, + "learning_rate": 1.490207456840273e-05, + "loss": 0.0543, + "step": 7128 + }, + { + "epoch": 0.7728751084128361, + "grad_norm": 0.5329501032829285, + "learning_rate": 1.4901349194835342e-05, + "loss": 0.0077, + "step": 7129 + }, + { + "epoch": 0.7729835212489159, + "grad_norm": 1.8368195295333862, + "learning_rate": 1.4900623821267954e-05, + "loss": 0.0311, + "step": 7130 + }, + { + "epoch": 0.7730919340849957, + "grad_norm": 1.8157910108566284, + "learning_rate": 1.4899898447700568e-05, + "loss": 0.0308, + "step": 7131 + }, + { + "epoch": 0.7732003469210754, + "grad_norm": 4.846690654754639, + "learning_rate": 1.489917307413318e-05, + "loss": 0.0769, + "step": 7132 + }, + { + "epoch": 0.7733087597571553, + "grad_norm": 4.000448226928711, + "learning_rate": 1.4898447700565792e-05, + "loss": 0.0394, + "step": 7133 + }, + { + "epoch": 0.7734171725932351, + "grad_norm": 2.116210699081421, + "learning_rate": 1.4897722326998406e-05, + "loss": 0.0331, + "step": 7134 + }, + { + "epoch": 0.7735255854293148, + "grad_norm": 2.0413079261779785, + "learning_rate": 1.4896996953431018e-05, + "loss": 0.0334, + "step": 7135 + }, + { + "epoch": 0.7736339982653946, + "grad_norm": 2.583834409713745, + "learning_rate": 1.489627157986363e-05, + "loss": 0.026, + "step": 7136 + }, + { + "epoch": 0.7737424111014745, + "grad_norm": 2.4255242347717285, + "learning_rate": 1.4895546206296242e-05, + "loss": 0.0329, + "step": 7137 + }, + { + "epoch": 0.7738508239375542, + "grad_norm": 2.8582024574279785, + "learning_rate": 1.4894820832728856e-05, + "loss": 0.0227, + "step": 7138 + }, + { + "epoch": 0.773959236773634, + "grad_norm": 2.554506301879883, + "learning_rate": 1.4894095459161468e-05, + "loss": 0.0464, + "step": 7139 + }, + { + "epoch": 0.7740676496097137, + "grad_norm": 2.8306667804718018, + "learning_rate": 1.4893370085594084e-05, + "loss": 0.0712, + "step": 7140 + }, + { + "epoch": 0.7741760624457936, + "grad_norm": 3.5061819553375244, + "learning_rate": 1.4892644712026696e-05, + "loss": 0.0322, + "step": 7141 + }, + { + "epoch": 0.7742844752818734, + "grad_norm": 1.5011080503463745, + "learning_rate": 1.4891919338459308e-05, + "loss": 0.0265, + "step": 7142 + }, + { + "epoch": 0.7743928881179531, + "grad_norm": 0.6457360982894897, + "learning_rate": 1.4891193964891922e-05, + "loss": 0.0089, + "step": 7143 + }, + { + "epoch": 0.7745013009540329, + "grad_norm": 1.8163564205169678, + "learning_rate": 1.4890468591324534e-05, + "loss": 0.0402, + "step": 7144 + }, + { + "epoch": 0.7746097137901128, + "grad_norm": 2.9552454948425293, + "learning_rate": 1.4889743217757146e-05, + "loss": 0.0239, + "step": 7145 + }, + { + "epoch": 0.7747181266261925, + "grad_norm": 3.735285758972168, + "learning_rate": 1.488901784418976e-05, + "loss": 0.0714, + "step": 7146 + }, + { + "epoch": 0.7748265394622723, + "grad_norm": 1.7430660724639893, + "learning_rate": 1.4888292470622372e-05, + "loss": 0.0149, + "step": 7147 + }, + { + "epoch": 0.7749349522983521, + "grad_norm": 3.0796959400177, + "learning_rate": 1.4887567097054984e-05, + "loss": 0.0682, + "step": 7148 + }, + { + "epoch": 0.775043365134432, + "grad_norm": 3.9704654216766357, + "learning_rate": 1.4886841723487598e-05, + "loss": 0.0792, + "step": 7149 + }, + { + "epoch": 0.7751517779705117, + "grad_norm": 6.583723068237305, + "learning_rate": 1.488611634992021e-05, + "loss": 0.0412, + "step": 7150 + }, + { + "epoch": 0.7752601908065915, + "grad_norm": 5.07785701751709, + "learning_rate": 1.4885390976352822e-05, + "loss": 0.0518, + "step": 7151 + }, + { + "epoch": 0.7753686036426712, + "grad_norm": 4.364170074462891, + "learning_rate": 1.4884665602785435e-05, + "loss": 0.0438, + "step": 7152 + }, + { + "epoch": 0.7754770164787511, + "grad_norm": 1.8661471605300903, + "learning_rate": 1.4883940229218048e-05, + "loss": 0.0537, + "step": 7153 + }, + { + "epoch": 0.7755854293148309, + "grad_norm": 5.603235244750977, + "learning_rate": 1.488321485565066e-05, + "loss": 0.038, + "step": 7154 + }, + { + "epoch": 0.7756938421509106, + "grad_norm": 1.4277068376541138, + "learning_rate": 1.4882489482083273e-05, + "loss": 0.0401, + "step": 7155 + }, + { + "epoch": 0.7758022549869905, + "grad_norm": 4.368765830993652, + "learning_rate": 1.4881764108515886e-05, + "loss": 0.0304, + "step": 7156 + }, + { + "epoch": 0.7759106678230703, + "grad_norm": 2.9100773334503174, + "learning_rate": 1.4881038734948499e-05, + "loss": 0.0203, + "step": 7157 + }, + { + "epoch": 0.77601908065915, + "grad_norm": 2.5829293727874756, + "learning_rate": 1.4880313361381114e-05, + "loss": 0.0637, + "step": 7158 + }, + { + "epoch": 0.7761274934952298, + "grad_norm": 1.8845280408859253, + "learning_rate": 1.4879587987813726e-05, + "loss": 0.0324, + "step": 7159 + }, + { + "epoch": 0.7762359063313097, + "grad_norm": 1.1194273233413696, + "learning_rate": 1.4878862614246338e-05, + "loss": 0.0147, + "step": 7160 + }, + { + "epoch": 0.7763443191673894, + "grad_norm": 2.3748700618743896, + "learning_rate": 1.4878137240678952e-05, + "loss": 0.0414, + "step": 7161 + }, + { + "epoch": 0.7764527320034692, + "grad_norm": 1.0228155851364136, + "learning_rate": 1.4877411867111564e-05, + "loss": 0.0096, + "step": 7162 + }, + { + "epoch": 0.776561144839549, + "grad_norm": 1.9693766832351685, + "learning_rate": 1.4876686493544177e-05, + "loss": 0.0273, + "step": 7163 + }, + { + "epoch": 0.7766695576756288, + "grad_norm": 1.0996030569076538, + "learning_rate": 1.4875961119976789e-05, + "loss": 0.0379, + "step": 7164 + }, + { + "epoch": 0.7767779705117086, + "grad_norm": 1.81586754322052, + "learning_rate": 1.4875235746409403e-05, + "loss": 0.0339, + "step": 7165 + }, + { + "epoch": 0.7768863833477884, + "grad_norm": 4.100707530975342, + "learning_rate": 1.4874510372842015e-05, + "loss": 0.0337, + "step": 7166 + }, + { + "epoch": 0.7769947961838681, + "grad_norm": 2.4386446475982666, + "learning_rate": 1.4873784999274627e-05, + "loss": 0.024, + "step": 7167 + }, + { + "epoch": 0.777103209019948, + "grad_norm": 1.8376741409301758, + "learning_rate": 1.487305962570724e-05, + "loss": 0.0259, + "step": 7168 + }, + { + "epoch": 0.7772116218560278, + "grad_norm": 1.191521406173706, + "learning_rate": 1.4872334252139853e-05, + "loss": 0.0512, + "step": 7169 + }, + { + "epoch": 0.7773200346921075, + "grad_norm": 1.9961391687393188, + "learning_rate": 1.4871608878572465e-05, + "loss": 0.0208, + "step": 7170 + }, + { + "epoch": 0.7774284475281873, + "grad_norm": 2.0148861408233643, + "learning_rate": 1.4870883505005079e-05, + "loss": 0.0411, + "step": 7171 + }, + { + "epoch": 0.7775368603642672, + "grad_norm": 2.2752981185913086, + "learning_rate": 1.4870158131437691e-05, + "loss": 0.026, + "step": 7172 + }, + { + "epoch": 0.7776452732003469, + "grad_norm": 6.073533535003662, + "learning_rate": 1.4869432757870303e-05, + "loss": 0.0345, + "step": 7173 + }, + { + "epoch": 0.7777536860364267, + "grad_norm": 2.7673168182373047, + "learning_rate": 1.4868707384302917e-05, + "loss": 0.0504, + "step": 7174 + }, + { + "epoch": 0.7778620988725065, + "grad_norm": 2.9976882934570312, + "learning_rate": 1.486798201073553e-05, + "loss": 0.0542, + "step": 7175 + }, + { + "epoch": 0.7779705117085863, + "grad_norm": 1.960453987121582, + "learning_rate": 1.4867256637168143e-05, + "loss": 0.0329, + "step": 7176 + }, + { + "epoch": 0.7780789245446661, + "grad_norm": 3.4617514610290527, + "learning_rate": 1.4866531263600757e-05, + "loss": 0.0951, + "step": 7177 + }, + { + "epoch": 0.7781873373807459, + "grad_norm": 1.6195485591888428, + "learning_rate": 1.4865805890033369e-05, + "loss": 0.0192, + "step": 7178 + }, + { + "epoch": 0.7782957502168256, + "grad_norm": 2.3292362689971924, + "learning_rate": 1.4865080516465981e-05, + "loss": 0.0322, + "step": 7179 + }, + { + "epoch": 0.7784041630529055, + "grad_norm": 2.401826858520508, + "learning_rate": 1.4864355142898595e-05, + "loss": 0.0189, + "step": 7180 + }, + { + "epoch": 0.7785125758889853, + "grad_norm": 3.014495372772217, + "learning_rate": 1.4863629769331207e-05, + "loss": 0.0776, + "step": 7181 + }, + { + "epoch": 0.778620988725065, + "grad_norm": 1.4866117238998413, + "learning_rate": 1.4862904395763819e-05, + "loss": 0.0374, + "step": 7182 + }, + { + "epoch": 0.7787294015611448, + "grad_norm": 2.965590715408325, + "learning_rate": 1.4862179022196433e-05, + "loss": 0.0686, + "step": 7183 + }, + { + "epoch": 0.7788378143972247, + "grad_norm": 2.535609483718872, + "learning_rate": 1.4861453648629045e-05, + "loss": 0.0416, + "step": 7184 + }, + { + "epoch": 0.7789462272333044, + "grad_norm": 2.724886655807495, + "learning_rate": 1.4860728275061657e-05, + "loss": 0.0463, + "step": 7185 + }, + { + "epoch": 0.7790546400693842, + "grad_norm": 1.214533805847168, + "learning_rate": 1.4860002901494271e-05, + "loss": 0.0169, + "step": 7186 + }, + { + "epoch": 0.7791630529054641, + "grad_norm": 1.953575611114502, + "learning_rate": 1.4859277527926883e-05, + "loss": 0.0252, + "step": 7187 + }, + { + "epoch": 0.7792714657415438, + "grad_norm": 1.6454417705535889, + "learning_rate": 1.4858552154359495e-05, + "loss": 0.0191, + "step": 7188 + }, + { + "epoch": 0.7793798785776236, + "grad_norm": 5.456317901611328, + "learning_rate": 1.4857826780792108e-05, + "loss": 0.0289, + "step": 7189 + }, + { + "epoch": 0.7794882914137033, + "grad_norm": 1.404917597770691, + "learning_rate": 1.4857101407224721e-05, + "loss": 0.0269, + "step": 7190 + }, + { + "epoch": 0.7795967042497832, + "grad_norm": 5.257943630218506, + "learning_rate": 1.4856376033657334e-05, + "loss": 0.0664, + "step": 7191 + }, + { + "epoch": 0.779705117085863, + "grad_norm": 2.747854709625244, + "learning_rate": 1.4855650660089949e-05, + "loss": 0.0528, + "step": 7192 + }, + { + "epoch": 0.7798135299219427, + "grad_norm": 1.1113837957382202, + "learning_rate": 1.4854925286522561e-05, + "loss": 0.0112, + "step": 7193 + }, + { + "epoch": 0.7799219427580225, + "grad_norm": 5.4841508865356445, + "learning_rate": 1.4854199912955173e-05, + "loss": 0.0655, + "step": 7194 + }, + { + "epoch": 0.7800303555941024, + "grad_norm": 5.957131862640381, + "learning_rate": 1.4853474539387787e-05, + "loss": 0.0424, + "step": 7195 + }, + { + "epoch": 0.7801387684301822, + "grad_norm": 1.9112788438796997, + "learning_rate": 1.48527491658204e-05, + "loss": 0.0283, + "step": 7196 + }, + { + "epoch": 0.7802471812662619, + "grad_norm": 0.9527618885040283, + "learning_rate": 1.4852023792253011e-05, + "loss": 0.0158, + "step": 7197 + }, + { + "epoch": 0.7803555941023417, + "grad_norm": 8.891447067260742, + "learning_rate": 1.4851298418685625e-05, + "loss": 0.0942, + "step": 7198 + }, + { + "epoch": 0.7804640069384216, + "grad_norm": 2.3627803325653076, + "learning_rate": 1.4850573045118237e-05, + "loss": 0.0529, + "step": 7199 + }, + { + "epoch": 0.7805724197745013, + "grad_norm": 1.5852766036987305, + "learning_rate": 1.484984767155085e-05, + "loss": 0.0196, + "step": 7200 + }, + { + "epoch": 0.7806808326105811, + "grad_norm": 5.051994800567627, + "learning_rate": 1.4849122297983463e-05, + "loss": 0.0439, + "step": 7201 + }, + { + "epoch": 0.7807892454466608, + "grad_norm": 1.9782191514968872, + "learning_rate": 1.4848396924416076e-05, + "loss": 0.0404, + "step": 7202 + }, + { + "epoch": 0.7808976582827407, + "grad_norm": 3.552415370941162, + "learning_rate": 1.4847671550848688e-05, + "loss": 0.0344, + "step": 7203 + }, + { + "epoch": 0.7810060711188205, + "grad_norm": 1.1155728101730347, + "learning_rate": 1.48469461772813e-05, + "loss": 0.0154, + "step": 7204 + }, + { + "epoch": 0.7811144839549002, + "grad_norm": 4.619261264801025, + "learning_rate": 1.4846220803713914e-05, + "loss": 0.0423, + "step": 7205 + }, + { + "epoch": 0.78122289679098, + "grad_norm": 3.6015641689300537, + "learning_rate": 1.4845495430146526e-05, + "loss": 0.0394, + "step": 7206 + }, + { + "epoch": 0.7813313096270599, + "grad_norm": 0.9221594929695129, + "learning_rate": 1.4844770056579138e-05, + "loss": 0.0185, + "step": 7207 + }, + { + "epoch": 0.7814397224631396, + "grad_norm": 2.7153847217559814, + "learning_rate": 1.4844044683011752e-05, + "loss": 0.0285, + "step": 7208 + }, + { + "epoch": 0.7815481352992194, + "grad_norm": 2.0568718910217285, + "learning_rate": 1.4843319309444366e-05, + "loss": 0.024, + "step": 7209 + }, + { + "epoch": 0.7816565481352992, + "grad_norm": 3.5821726322174072, + "learning_rate": 1.484259393587698e-05, + "loss": 0.0459, + "step": 7210 + }, + { + "epoch": 0.781764960971379, + "grad_norm": 3.078188896179199, + "learning_rate": 1.4841868562309592e-05, + "loss": 0.0329, + "step": 7211 + }, + { + "epoch": 0.7818733738074588, + "grad_norm": 1.3874542713165283, + "learning_rate": 1.4841143188742204e-05, + "loss": 0.0277, + "step": 7212 + }, + { + "epoch": 0.7819817866435386, + "grad_norm": 1.6864585876464844, + "learning_rate": 1.4840417815174818e-05, + "loss": 0.0318, + "step": 7213 + }, + { + "epoch": 0.7820901994796184, + "grad_norm": 1.8273521661758423, + "learning_rate": 1.483969244160743e-05, + "loss": 0.0201, + "step": 7214 + }, + { + "epoch": 0.7821986123156982, + "grad_norm": 1.5783238410949707, + "learning_rate": 1.4838967068040042e-05, + "loss": 0.0257, + "step": 7215 + }, + { + "epoch": 0.782307025151778, + "grad_norm": 2.855252981185913, + "learning_rate": 1.4838241694472654e-05, + "loss": 0.0269, + "step": 7216 + }, + { + "epoch": 0.7824154379878577, + "grad_norm": 1.027496576309204, + "learning_rate": 1.4837516320905268e-05, + "loss": 0.0229, + "step": 7217 + }, + { + "epoch": 0.7825238508239376, + "grad_norm": 2.817304849624634, + "learning_rate": 1.483679094733788e-05, + "loss": 0.0415, + "step": 7218 + }, + { + "epoch": 0.7826322636600174, + "grad_norm": 1.7635983228683472, + "learning_rate": 1.4836065573770492e-05, + "loss": 0.0219, + "step": 7219 + }, + { + "epoch": 0.7827406764960971, + "grad_norm": 2.6337478160858154, + "learning_rate": 1.4835340200203106e-05, + "loss": 0.038, + "step": 7220 + }, + { + "epoch": 0.7828490893321769, + "grad_norm": 6.990657329559326, + "learning_rate": 1.4834614826635718e-05, + "loss": 0.0705, + "step": 7221 + }, + { + "epoch": 0.7829575021682568, + "grad_norm": 2.039715051651001, + "learning_rate": 1.483388945306833e-05, + "loss": 0.0079, + "step": 7222 + }, + { + "epoch": 0.7830659150043365, + "grad_norm": 5.440858840942383, + "learning_rate": 1.4833164079500944e-05, + "loss": 0.0668, + "step": 7223 + }, + { + "epoch": 0.7831743278404163, + "grad_norm": 6.466601371765137, + "learning_rate": 1.4832438705933556e-05, + "loss": 0.0326, + "step": 7224 + }, + { + "epoch": 0.7832827406764961, + "grad_norm": 2.601494312286377, + "learning_rate": 1.4831713332366168e-05, + "loss": 0.0282, + "step": 7225 + }, + { + "epoch": 0.7833911535125759, + "grad_norm": 0.5198491811752319, + "learning_rate": 1.4830987958798782e-05, + "loss": 0.0151, + "step": 7226 + }, + { + "epoch": 0.7834995663486557, + "grad_norm": 2.0865979194641113, + "learning_rate": 1.4830262585231396e-05, + "loss": 0.0406, + "step": 7227 + }, + { + "epoch": 0.7836079791847355, + "grad_norm": 1.2536057233810425, + "learning_rate": 1.4829537211664008e-05, + "loss": 0.0332, + "step": 7228 + }, + { + "epoch": 0.7837163920208152, + "grad_norm": 1.3969916105270386, + "learning_rate": 1.4828811838096622e-05, + "loss": 0.0446, + "step": 7229 + }, + { + "epoch": 0.7838248048568951, + "grad_norm": 2.9662647247314453, + "learning_rate": 1.4828086464529234e-05, + "loss": 0.0432, + "step": 7230 + }, + { + "epoch": 0.7839332176929749, + "grad_norm": 3.702552318572998, + "learning_rate": 1.4827361090961846e-05, + "loss": 0.0401, + "step": 7231 + }, + { + "epoch": 0.7840416305290546, + "grad_norm": 3.59220027923584, + "learning_rate": 1.482663571739446e-05, + "loss": 0.037, + "step": 7232 + }, + { + "epoch": 0.7841500433651344, + "grad_norm": 5.902690410614014, + "learning_rate": 1.4825910343827072e-05, + "loss": 0.0331, + "step": 7233 + }, + { + "epoch": 0.7842584562012143, + "grad_norm": 3.738600492477417, + "learning_rate": 1.4825184970259684e-05, + "loss": 0.0553, + "step": 7234 + }, + { + "epoch": 0.784366869037294, + "grad_norm": 0.17621643841266632, + "learning_rate": 1.4824459596692298e-05, + "loss": 0.0021, + "step": 7235 + }, + { + "epoch": 0.7844752818733738, + "grad_norm": 3.6884992122650146, + "learning_rate": 1.482373422312491e-05, + "loss": 0.05, + "step": 7236 + }, + { + "epoch": 0.7845836947094535, + "grad_norm": 0.5442163348197937, + "learning_rate": 1.4823008849557523e-05, + "loss": 0.006, + "step": 7237 + }, + { + "epoch": 0.7846921075455334, + "grad_norm": 1.9410150051116943, + "learning_rate": 1.4822283475990136e-05, + "loss": 0.0206, + "step": 7238 + }, + { + "epoch": 0.7848005203816132, + "grad_norm": 0.8875540494918823, + "learning_rate": 1.4821558102422749e-05, + "loss": 0.0158, + "step": 7239 + }, + { + "epoch": 0.784908933217693, + "grad_norm": 2.742553234100342, + "learning_rate": 1.482083272885536e-05, + "loss": 0.0643, + "step": 7240 + }, + { + "epoch": 0.7850173460537727, + "grad_norm": 2.248521327972412, + "learning_rate": 1.4820107355287975e-05, + "loss": 0.0365, + "step": 7241 + }, + { + "epoch": 0.7851257588898526, + "grad_norm": 1.9694749116897583, + "learning_rate": 1.4819381981720587e-05, + "loss": 0.0688, + "step": 7242 + }, + { + "epoch": 0.7852341717259324, + "grad_norm": 1.4306461811065674, + "learning_rate": 1.4818656608153199e-05, + "loss": 0.0418, + "step": 7243 + }, + { + "epoch": 0.7853425845620121, + "grad_norm": 0.5622298717498779, + "learning_rate": 1.4817931234585814e-05, + "loss": 0.0075, + "step": 7244 + }, + { + "epoch": 0.785450997398092, + "grad_norm": 1.89161217212677, + "learning_rate": 1.4817205861018427e-05, + "loss": 0.0189, + "step": 7245 + }, + { + "epoch": 0.7855594102341718, + "grad_norm": 0.8803873658180237, + "learning_rate": 1.4816480487451039e-05, + "loss": 0.0165, + "step": 7246 + }, + { + "epoch": 0.7856678230702515, + "grad_norm": 3.0707128047943115, + "learning_rate": 1.4815755113883653e-05, + "loss": 0.0257, + "step": 7247 + }, + { + "epoch": 0.7857762359063313, + "grad_norm": 2.6187875270843506, + "learning_rate": 1.4815029740316265e-05, + "loss": 0.0346, + "step": 7248 + }, + { + "epoch": 0.7858846487424112, + "grad_norm": 1.2435921430587769, + "learning_rate": 1.4814304366748877e-05, + "loss": 0.0184, + "step": 7249 + }, + { + "epoch": 0.7859930615784909, + "grad_norm": 8.00239086151123, + "learning_rate": 1.481357899318149e-05, + "loss": 0.0772, + "step": 7250 + }, + { + "epoch": 0.7861014744145707, + "grad_norm": 3.808727502822876, + "learning_rate": 1.4812853619614103e-05, + "loss": 0.0342, + "step": 7251 + }, + { + "epoch": 0.7862098872506504, + "grad_norm": 1.6004741191864014, + "learning_rate": 1.4812128246046715e-05, + "loss": 0.0171, + "step": 7252 + }, + { + "epoch": 0.7863183000867303, + "grad_norm": 6.1819233894348145, + "learning_rate": 1.4811402872479329e-05, + "loss": 0.031, + "step": 7253 + }, + { + "epoch": 0.7864267129228101, + "grad_norm": 7.104381561279297, + "learning_rate": 1.4810677498911941e-05, + "loss": 0.0543, + "step": 7254 + }, + { + "epoch": 0.7865351257588898, + "grad_norm": 3.7121670246124268, + "learning_rate": 1.4809952125344553e-05, + "loss": 0.0568, + "step": 7255 + }, + { + "epoch": 0.7866435385949696, + "grad_norm": 2.5434625148773193, + "learning_rate": 1.4809226751777165e-05, + "loss": 0.031, + "step": 7256 + }, + { + "epoch": 0.7867519514310495, + "grad_norm": 1.4491708278656006, + "learning_rate": 1.4808501378209779e-05, + "loss": 0.0182, + "step": 7257 + }, + { + "epoch": 0.7868603642671292, + "grad_norm": 3.9808735847473145, + "learning_rate": 1.4807776004642391e-05, + "loss": 0.0553, + "step": 7258 + }, + { + "epoch": 0.786968777103209, + "grad_norm": 2.8032922744750977, + "learning_rate": 1.4807050631075003e-05, + "loss": 0.0493, + "step": 7259 + }, + { + "epoch": 0.7870771899392888, + "grad_norm": 2.7927660942077637, + "learning_rate": 1.4806325257507617e-05, + "loss": 0.0415, + "step": 7260 + }, + { + "epoch": 0.7871856027753686, + "grad_norm": 6.855513572692871, + "learning_rate": 1.4805599883940231e-05, + "loss": 0.1005, + "step": 7261 + }, + { + "epoch": 0.7872940156114484, + "grad_norm": 5.300849437713623, + "learning_rate": 1.4804874510372845e-05, + "loss": 0.0735, + "step": 7262 + }, + { + "epoch": 0.7874024284475282, + "grad_norm": 2.779264211654663, + "learning_rate": 1.4804149136805457e-05, + "loss": 0.0144, + "step": 7263 + }, + { + "epoch": 0.7875108412836079, + "grad_norm": 3.3010005950927734, + "learning_rate": 1.4803423763238069e-05, + "loss": 0.0668, + "step": 7264 + }, + { + "epoch": 0.7876192541196878, + "grad_norm": 3.4064619541168213, + "learning_rate": 1.4802698389670683e-05, + "loss": 0.0299, + "step": 7265 + }, + { + "epoch": 0.7877276669557676, + "grad_norm": 1.8243944644927979, + "learning_rate": 1.4801973016103295e-05, + "loss": 0.0219, + "step": 7266 + }, + { + "epoch": 0.7878360797918473, + "grad_norm": 4.498291492462158, + "learning_rate": 1.4801247642535907e-05, + "loss": 0.0514, + "step": 7267 + }, + { + "epoch": 0.7879444926279271, + "grad_norm": 2.339035749435425, + "learning_rate": 1.480052226896852e-05, + "loss": 0.0246, + "step": 7268 + }, + { + "epoch": 0.788052905464007, + "grad_norm": 1.9674608707427979, + "learning_rate": 1.4799796895401133e-05, + "loss": 0.0735, + "step": 7269 + }, + { + "epoch": 0.7881613183000867, + "grad_norm": 2.0698349475860596, + "learning_rate": 1.4799071521833745e-05, + "loss": 0.0571, + "step": 7270 + }, + { + "epoch": 0.7882697311361665, + "grad_norm": 2.5799174308776855, + "learning_rate": 1.4798346148266358e-05, + "loss": 0.0361, + "step": 7271 + }, + { + "epoch": 0.7883781439722463, + "grad_norm": 2.6498422622680664, + "learning_rate": 1.4797620774698971e-05, + "loss": 0.0332, + "step": 7272 + }, + { + "epoch": 0.7884865568083261, + "grad_norm": 1.389792561531067, + "learning_rate": 1.4796895401131583e-05, + "loss": 0.0134, + "step": 7273 + }, + { + "epoch": 0.7885949696444059, + "grad_norm": 2.968893527984619, + "learning_rate": 1.4796170027564196e-05, + "loss": 0.0685, + "step": 7274 + }, + { + "epoch": 0.7887033824804857, + "grad_norm": 1.0499293804168701, + "learning_rate": 1.479544465399681e-05, + "loss": 0.008, + "step": 7275 + }, + { + "epoch": 0.7888117953165655, + "grad_norm": 2.445744037628174, + "learning_rate": 1.4794719280429422e-05, + "loss": 0.0477, + "step": 7276 + }, + { + "epoch": 0.7889202081526453, + "grad_norm": 2.474895477294922, + "learning_rate": 1.4793993906862034e-05, + "loss": 0.0491, + "step": 7277 + }, + { + "epoch": 0.7890286209887251, + "grad_norm": 1.2534019947052002, + "learning_rate": 1.4793268533294648e-05, + "loss": 0.0126, + "step": 7278 + }, + { + "epoch": 0.7891370338248048, + "grad_norm": 0.4601795971393585, + "learning_rate": 1.4792543159727261e-05, + "loss": 0.0048, + "step": 7279 + }, + { + "epoch": 0.7892454466608847, + "grad_norm": 0.9280397295951843, + "learning_rate": 1.4791817786159874e-05, + "loss": 0.02, + "step": 7280 + }, + { + "epoch": 0.7893538594969645, + "grad_norm": 4.077282428741455, + "learning_rate": 1.4791092412592487e-05, + "loss": 0.0272, + "step": 7281 + }, + { + "epoch": 0.7894622723330442, + "grad_norm": 0.5080981254577637, + "learning_rate": 1.47903670390251e-05, + "loss": 0.0093, + "step": 7282 + }, + { + "epoch": 0.789570685169124, + "grad_norm": 1.7629189491271973, + "learning_rate": 1.4789641665457712e-05, + "loss": 0.0249, + "step": 7283 + }, + { + "epoch": 0.7896790980052039, + "grad_norm": 4.184973239898682, + "learning_rate": 1.4788916291890326e-05, + "loss": 0.0839, + "step": 7284 + }, + { + "epoch": 0.7897875108412836, + "grad_norm": 1.717464566230774, + "learning_rate": 1.4788190918322938e-05, + "loss": 0.054, + "step": 7285 + }, + { + "epoch": 0.7898959236773634, + "grad_norm": 3.3832998275756836, + "learning_rate": 1.478746554475555e-05, + "loss": 0.0252, + "step": 7286 + }, + { + "epoch": 0.7900043365134432, + "grad_norm": 3.9383485317230225, + "learning_rate": 1.4786740171188164e-05, + "loss": 0.0306, + "step": 7287 + }, + { + "epoch": 0.790112749349523, + "grad_norm": 3.2722668647766113, + "learning_rate": 1.4786014797620776e-05, + "loss": 0.0484, + "step": 7288 + }, + { + "epoch": 0.7902211621856028, + "grad_norm": 2.984260320663452, + "learning_rate": 1.4785289424053388e-05, + "loss": 0.0724, + "step": 7289 + }, + { + "epoch": 0.7903295750216826, + "grad_norm": 3.0267579555511475, + "learning_rate": 1.4784564050486002e-05, + "loss": 0.0313, + "step": 7290 + }, + { + "epoch": 0.7904379878577623, + "grad_norm": 4.398987770080566, + "learning_rate": 1.4783838676918614e-05, + "loss": 0.0472, + "step": 7291 + }, + { + "epoch": 0.7905464006938422, + "grad_norm": 1.6740131378173828, + "learning_rate": 1.4783113303351226e-05, + "loss": 0.0281, + "step": 7292 + }, + { + "epoch": 0.790654813529922, + "grad_norm": 2.9737465381622314, + "learning_rate": 1.478238792978384e-05, + "loss": 0.0191, + "step": 7293 + }, + { + "epoch": 0.7907632263660017, + "grad_norm": 0.7139529585838318, + "learning_rate": 1.4781662556216452e-05, + "loss": 0.0099, + "step": 7294 + }, + { + "epoch": 0.7908716392020815, + "grad_norm": 2.5596745014190674, + "learning_rate": 1.4780937182649064e-05, + "loss": 0.0288, + "step": 7295 + }, + { + "epoch": 0.7909800520381614, + "grad_norm": 4.689486503601074, + "learning_rate": 1.478021180908168e-05, + "loss": 0.0985, + "step": 7296 + }, + { + "epoch": 0.7910884648742411, + "grad_norm": 3.2683615684509277, + "learning_rate": 1.4779486435514292e-05, + "loss": 0.0214, + "step": 7297 + }, + { + "epoch": 0.7911968777103209, + "grad_norm": 2.1085991859436035, + "learning_rate": 1.4778761061946904e-05, + "loss": 0.0369, + "step": 7298 + }, + { + "epoch": 0.7913052905464006, + "grad_norm": 1.2510831356048584, + "learning_rate": 1.4778035688379518e-05, + "loss": 0.0173, + "step": 7299 + }, + { + "epoch": 0.7914137033824805, + "grad_norm": 1.8584117889404297, + "learning_rate": 1.477731031481213e-05, + "loss": 0.0332, + "step": 7300 + }, + { + "epoch": 0.7915221162185603, + "grad_norm": 5.667847633361816, + "learning_rate": 1.4776584941244742e-05, + "loss": 0.0423, + "step": 7301 + }, + { + "epoch": 0.79163052905464, + "grad_norm": 5.381179332733154, + "learning_rate": 1.4775859567677356e-05, + "loss": 0.0801, + "step": 7302 + }, + { + "epoch": 0.7917389418907199, + "grad_norm": 0.9299309253692627, + "learning_rate": 1.4775134194109968e-05, + "loss": 0.0201, + "step": 7303 + }, + { + "epoch": 0.7918473547267997, + "grad_norm": 1.994738221168518, + "learning_rate": 1.477440882054258e-05, + "loss": 0.0277, + "step": 7304 + }, + { + "epoch": 0.7919557675628794, + "grad_norm": 3.41363525390625, + "learning_rate": 1.4773683446975194e-05, + "loss": 0.0553, + "step": 7305 + }, + { + "epoch": 0.7920641803989592, + "grad_norm": 1.610550045967102, + "learning_rate": 1.4772958073407806e-05, + "loss": 0.0188, + "step": 7306 + }, + { + "epoch": 0.7921725932350391, + "grad_norm": 4.524579048156738, + "learning_rate": 1.4772232699840418e-05, + "loss": 0.0454, + "step": 7307 + }, + { + "epoch": 0.7922810060711188, + "grad_norm": 4.223093509674072, + "learning_rate": 1.477150732627303e-05, + "loss": 0.0953, + "step": 7308 + }, + { + "epoch": 0.7923894189071986, + "grad_norm": 3.243093967437744, + "learning_rate": 1.4770781952705644e-05, + "loss": 0.017, + "step": 7309 + }, + { + "epoch": 0.7924978317432784, + "grad_norm": 2.6651878356933594, + "learning_rate": 1.4770056579138257e-05, + "loss": 0.031, + "step": 7310 + }, + { + "epoch": 0.7926062445793582, + "grad_norm": 3.117210865020752, + "learning_rate": 1.4769331205570869e-05, + "loss": 0.0779, + "step": 7311 + }, + { + "epoch": 0.792714657415438, + "grad_norm": 3.614894151687622, + "learning_rate": 1.4768605832003482e-05, + "loss": 0.0648, + "step": 7312 + }, + { + "epoch": 0.7928230702515178, + "grad_norm": 2.8248953819274902, + "learning_rate": 1.4767880458436096e-05, + "loss": 0.0533, + "step": 7313 + }, + { + "epoch": 0.7929314830875975, + "grad_norm": 4.533317565917969, + "learning_rate": 1.476715508486871e-05, + "loss": 0.058, + "step": 7314 + }, + { + "epoch": 0.7930398959236774, + "grad_norm": 4.4283928871154785, + "learning_rate": 1.4766429711301322e-05, + "loss": 0.0424, + "step": 7315 + }, + { + "epoch": 0.7931483087597572, + "grad_norm": 1.5120707750320435, + "learning_rate": 1.4765704337733934e-05, + "loss": 0.0311, + "step": 7316 + }, + { + "epoch": 0.7932567215958369, + "grad_norm": 0.661862850189209, + "learning_rate": 1.4764978964166548e-05, + "loss": 0.0056, + "step": 7317 + }, + { + "epoch": 0.7933651344319167, + "grad_norm": 1.0245895385742188, + "learning_rate": 1.476425359059916e-05, + "loss": 0.0267, + "step": 7318 + }, + { + "epoch": 0.7934735472679966, + "grad_norm": 0.6727933883666992, + "learning_rate": 1.4763528217031773e-05, + "loss": 0.0133, + "step": 7319 + }, + { + "epoch": 0.7935819601040763, + "grad_norm": 3.187059164047241, + "learning_rate": 1.4762802843464385e-05, + "loss": 0.0163, + "step": 7320 + }, + { + "epoch": 0.7936903729401561, + "grad_norm": 2.7047207355499268, + "learning_rate": 1.4762077469896999e-05, + "loss": 0.034, + "step": 7321 + }, + { + "epoch": 0.7937987857762359, + "grad_norm": 5.477845191955566, + "learning_rate": 1.476135209632961e-05, + "loss": 0.0312, + "step": 7322 + }, + { + "epoch": 0.7939071986123157, + "grad_norm": 1.8906244039535522, + "learning_rate": 1.4760626722762223e-05, + "loss": 0.0104, + "step": 7323 + }, + { + "epoch": 0.7940156114483955, + "grad_norm": 2.6927602291107178, + "learning_rate": 1.4759901349194837e-05, + "loss": 0.0316, + "step": 7324 + }, + { + "epoch": 0.7941240242844753, + "grad_norm": 2.649249792098999, + "learning_rate": 1.4759175975627449e-05, + "loss": 0.0588, + "step": 7325 + }, + { + "epoch": 0.794232437120555, + "grad_norm": 5.326345443725586, + "learning_rate": 1.4758450602060061e-05, + "loss": 0.0319, + "step": 7326 + }, + { + "epoch": 0.7943408499566349, + "grad_norm": 1.3852366209030151, + "learning_rate": 1.4757725228492675e-05, + "loss": 0.0173, + "step": 7327 + }, + { + "epoch": 0.7944492627927147, + "grad_norm": 2.0231282711029053, + "learning_rate": 1.4756999854925287e-05, + "loss": 0.0274, + "step": 7328 + }, + { + "epoch": 0.7945576756287944, + "grad_norm": 7.847790241241455, + "learning_rate": 1.4756274481357899e-05, + "loss": 0.0965, + "step": 7329 + }, + { + "epoch": 0.7946660884648742, + "grad_norm": 3.4567148685455322, + "learning_rate": 1.4755549107790515e-05, + "loss": 0.0374, + "step": 7330 + }, + { + "epoch": 0.7947745013009541, + "grad_norm": 1.8798117637634277, + "learning_rate": 1.4754823734223127e-05, + "loss": 0.0102, + "step": 7331 + }, + { + "epoch": 0.7948829141370338, + "grad_norm": 4.554348945617676, + "learning_rate": 1.4754098360655739e-05, + "loss": 0.0283, + "step": 7332 + }, + { + "epoch": 0.7949913269731136, + "grad_norm": 4.216191291809082, + "learning_rate": 1.4753372987088353e-05, + "loss": 0.0432, + "step": 7333 + }, + { + "epoch": 0.7950997398091935, + "grad_norm": 1.5660018920898438, + "learning_rate": 1.4752647613520965e-05, + "loss": 0.0284, + "step": 7334 + }, + { + "epoch": 0.7952081526452732, + "grad_norm": 3.4279744625091553, + "learning_rate": 1.4751922239953577e-05, + "loss": 0.0881, + "step": 7335 + }, + { + "epoch": 0.795316565481353, + "grad_norm": 2.903308868408203, + "learning_rate": 1.4751196866386191e-05, + "loss": 0.0544, + "step": 7336 + }, + { + "epoch": 0.7954249783174328, + "grad_norm": 4.6922287940979, + "learning_rate": 1.4750471492818803e-05, + "loss": 0.0249, + "step": 7337 + }, + { + "epoch": 0.7955333911535126, + "grad_norm": 2.345313310623169, + "learning_rate": 1.4749746119251415e-05, + "loss": 0.0391, + "step": 7338 + }, + { + "epoch": 0.7956418039895924, + "grad_norm": 2.8396356105804443, + "learning_rate": 1.4749020745684029e-05, + "loss": 0.0544, + "step": 7339 + }, + { + "epoch": 0.7957502168256722, + "grad_norm": 1.7268084287643433, + "learning_rate": 1.4748295372116641e-05, + "loss": 0.0243, + "step": 7340 + }, + { + "epoch": 0.7958586296617519, + "grad_norm": 1.9715940952301025, + "learning_rate": 1.4747569998549253e-05, + "loss": 0.0187, + "step": 7341 + }, + { + "epoch": 0.7959670424978318, + "grad_norm": 2.724940299987793, + "learning_rate": 1.4746844624981867e-05, + "loss": 0.0198, + "step": 7342 + }, + { + "epoch": 0.7960754553339116, + "grad_norm": 3.403642416000366, + "learning_rate": 1.474611925141448e-05, + "loss": 0.0243, + "step": 7343 + }, + { + "epoch": 0.7961838681699913, + "grad_norm": 1.6754637956619263, + "learning_rate": 1.4745393877847091e-05, + "loss": 0.034, + "step": 7344 + }, + { + "epoch": 0.7962922810060711, + "grad_norm": 2.555695056915283, + "learning_rate": 1.4744668504279705e-05, + "loss": 0.0634, + "step": 7345 + }, + { + "epoch": 0.796400693842151, + "grad_norm": 3.9993526935577393, + "learning_rate": 1.4743943130712317e-05, + "loss": 0.0301, + "step": 7346 + }, + { + "epoch": 0.7965091066782307, + "grad_norm": 4.551050662994385, + "learning_rate": 1.474321775714493e-05, + "loss": 0.0513, + "step": 7347 + }, + { + "epoch": 0.7966175195143105, + "grad_norm": 7.551766395568848, + "learning_rate": 1.4742492383577545e-05, + "loss": 0.0544, + "step": 7348 + }, + { + "epoch": 0.7967259323503902, + "grad_norm": 1.1692085266113281, + "learning_rate": 1.4741767010010157e-05, + "loss": 0.0106, + "step": 7349 + }, + { + "epoch": 0.7968343451864701, + "grad_norm": 6.513040542602539, + "learning_rate": 1.474104163644277e-05, + "loss": 0.0807, + "step": 7350 + }, + { + "epoch": 0.7969427580225499, + "grad_norm": 2.732846975326538, + "learning_rate": 1.4740316262875383e-05, + "loss": 0.0399, + "step": 7351 + }, + { + "epoch": 0.7970511708586296, + "grad_norm": 4.7541069984436035, + "learning_rate": 1.4739590889307995e-05, + "loss": 0.045, + "step": 7352 + }, + { + "epoch": 0.7971595836947094, + "grad_norm": 3.839862585067749, + "learning_rate": 1.4738865515740607e-05, + "loss": 0.044, + "step": 7353 + }, + { + "epoch": 0.7972679965307893, + "grad_norm": 0.999291718006134, + "learning_rate": 1.4738140142173221e-05, + "loss": 0.024, + "step": 7354 + }, + { + "epoch": 0.797376409366869, + "grad_norm": 11.970638275146484, + "learning_rate": 1.4737414768605833e-05, + "loss": 0.0454, + "step": 7355 + }, + { + "epoch": 0.7974848222029488, + "grad_norm": 1.17183256149292, + "learning_rate": 1.4736689395038446e-05, + "loss": 0.0316, + "step": 7356 + }, + { + "epoch": 0.7975932350390286, + "grad_norm": 3.006335496902466, + "learning_rate": 1.473596402147106e-05, + "loss": 0.103, + "step": 7357 + }, + { + "epoch": 0.7977016478751084, + "grad_norm": 2.799957275390625, + "learning_rate": 1.4735238647903672e-05, + "loss": 0.0493, + "step": 7358 + }, + { + "epoch": 0.7978100607111882, + "grad_norm": 5.318875789642334, + "learning_rate": 1.4734513274336284e-05, + "loss": 0.0809, + "step": 7359 + }, + { + "epoch": 0.797918473547268, + "grad_norm": 3.8092947006225586, + "learning_rate": 1.4733787900768896e-05, + "loss": 0.0438, + "step": 7360 + }, + { + "epoch": 0.7980268863833477, + "grad_norm": 5.134561538696289, + "learning_rate": 1.473306252720151e-05, + "loss": 0.0879, + "step": 7361 + }, + { + "epoch": 0.7981352992194276, + "grad_norm": 4.112361431121826, + "learning_rate": 1.4732337153634122e-05, + "loss": 0.0383, + "step": 7362 + }, + { + "epoch": 0.7982437120555074, + "grad_norm": 2.1977787017822266, + "learning_rate": 1.4731611780066734e-05, + "loss": 0.0354, + "step": 7363 + }, + { + "epoch": 0.7983521248915871, + "grad_norm": 0.48256731033325195, + "learning_rate": 1.4730886406499348e-05, + "loss": 0.0085, + "step": 7364 + }, + { + "epoch": 0.798460537727667, + "grad_norm": 3.758784532546997, + "learning_rate": 1.4730161032931962e-05, + "loss": 0.0603, + "step": 7365 + }, + { + "epoch": 0.7985689505637468, + "grad_norm": 2.6667978763580322, + "learning_rate": 1.4729435659364576e-05, + "loss": 0.0307, + "step": 7366 + }, + { + "epoch": 0.7986773633998265, + "grad_norm": 1.5987908840179443, + "learning_rate": 1.4728710285797188e-05, + "loss": 0.0214, + "step": 7367 + }, + { + "epoch": 0.7987857762359063, + "grad_norm": 1.7754172086715698, + "learning_rate": 1.47279849122298e-05, + "loss": 0.0162, + "step": 7368 + }, + { + "epoch": 0.7988941890719862, + "grad_norm": 0.8313624858856201, + "learning_rate": 1.4727259538662414e-05, + "loss": 0.0224, + "step": 7369 + }, + { + "epoch": 0.7990026019080659, + "grad_norm": 2.532382011413574, + "learning_rate": 1.4726534165095026e-05, + "loss": 0.0449, + "step": 7370 + }, + { + "epoch": 0.7991110147441457, + "grad_norm": 1.7980880737304688, + "learning_rate": 1.4725808791527638e-05, + "loss": 0.0413, + "step": 7371 + }, + { + "epoch": 0.7992194275802255, + "grad_norm": 3.1873812675476074, + "learning_rate": 1.472508341796025e-05, + "loss": 0.036, + "step": 7372 + }, + { + "epoch": 0.7993278404163053, + "grad_norm": 3.233867883682251, + "learning_rate": 1.4724358044392864e-05, + "loss": 0.0721, + "step": 7373 + }, + { + "epoch": 0.7994362532523851, + "grad_norm": 2.456699848175049, + "learning_rate": 1.4723632670825476e-05, + "loss": 0.0576, + "step": 7374 + }, + { + "epoch": 0.7995446660884649, + "grad_norm": 2.9317808151245117, + "learning_rate": 1.4722907297258088e-05, + "loss": 0.069, + "step": 7375 + }, + { + "epoch": 0.7996530789245446, + "grad_norm": 0.8448598980903625, + "learning_rate": 1.4722181923690702e-05, + "loss": 0.0145, + "step": 7376 + }, + { + "epoch": 0.7997614917606245, + "grad_norm": 1.578770399093628, + "learning_rate": 1.4721456550123314e-05, + "loss": 0.0315, + "step": 7377 + }, + { + "epoch": 0.7998699045967043, + "grad_norm": 4.000262260437012, + "learning_rate": 1.4720731176555926e-05, + "loss": 0.0413, + "step": 7378 + }, + { + "epoch": 0.799978317432784, + "grad_norm": 3.3782074451446533, + "learning_rate": 1.472000580298854e-05, + "loss": 0.0518, + "step": 7379 + }, + { + "epoch": 0.8000867302688638, + "grad_norm": 1.8855093717575073, + "learning_rate": 1.4719280429421152e-05, + "loss": 0.0294, + "step": 7380 + }, + { + "epoch": 0.8001951431049437, + "grad_norm": 0.8238834738731384, + "learning_rate": 1.4718555055853764e-05, + "loss": 0.0131, + "step": 7381 + }, + { + "epoch": 0.8003035559410234, + "grad_norm": 0.6787129044532776, + "learning_rate": 1.471782968228638e-05, + "loss": 0.0056, + "step": 7382 + }, + { + "epoch": 0.8004119687771032, + "grad_norm": 0.9170066118240356, + "learning_rate": 1.4717104308718992e-05, + "loss": 0.0199, + "step": 7383 + }, + { + "epoch": 0.800520381613183, + "grad_norm": 1.0948820114135742, + "learning_rate": 1.4716378935151604e-05, + "loss": 0.0111, + "step": 7384 + }, + { + "epoch": 0.8006287944492628, + "grad_norm": 4.331338405609131, + "learning_rate": 1.4715653561584218e-05, + "loss": 0.0339, + "step": 7385 + }, + { + "epoch": 0.8007372072853426, + "grad_norm": 1.8244320154190063, + "learning_rate": 1.471492818801683e-05, + "loss": 0.0223, + "step": 7386 + }, + { + "epoch": 0.8008456201214224, + "grad_norm": 3.906242847442627, + "learning_rate": 1.4714202814449442e-05, + "loss": 0.0458, + "step": 7387 + }, + { + "epoch": 0.8009540329575021, + "grad_norm": 1.3837815523147583, + "learning_rate": 1.4713477440882056e-05, + "loss": 0.0268, + "step": 7388 + }, + { + "epoch": 0.801062445793582, + "grad_norm": 2.9502809047698975, + "learning_rate": 1.4712752067314668e-05, + "loss": 0.0484, + "step": 7389 + }, + { + "epoch": 0.8011708586296618, + "grad_norm": 3.9696121215820312, + "learning_rate": 1.471202669374728e-05, + "loss": 0.0453, + "step": 7390 + }, + { + "epoch": 0.8012792714657415, + "grad_norm": 2.879714012145996, + "learning_rate": 1.4711301320179894e-05, + "loss": 0.0279, + "step": 7391 + }, + { + "epoch": 0.8013876843018214, + "grad_norm": 1.2540243864059448, + "learning_rate": 1.4710575946612506e-05, + "loss": 0.0146, + "step": 7392 + }, + { + "epoch": 0.8014960971379012, + "grad_norm": 1.6454055309295654, + "learning_rate": 1.4709850573045119e-05, + "loss": 0.0077, + "step": 7393 + }, + { + "epoch": 0.8016045099739809, + "grad_norm": 1.9396872520446777, + "learning_rate": 1.4709125199477732e-05, + "loss": 0.0421, + "step": 7394 + }, + { + "epoch": 0.8017129228100607, + "grad_norm": 7.269184589385986, + "learning_rate": 1.4708399825910345e-05, + "loss": 0.0417, + "step": 7395 + }, + { + "epoch": 0.8018213356461406, + "grad_norm": 0.5714468359947205, + "learning_rate": 1.4707674452342957e-05, + "loss": 0.0094, + "step": 7396 + }, + { + "epoch": 0.8019297484822203, + "grad_norm": 6.068140506744385, + "learning_rate": 1.470694907877557e-05, + "loss": 0.0266, + "step": 7397 + }, + { + "epoch": 0.8020381613183001, + "grad_norm": 1.761820912361145, + "learning_rate": 1.4706223705208183e-05, + "loss": 0.0092, + "step": 7398 + }, + { + "epoch": 0.8021465741543798, + "grad_norm": 4.833492279052734, + "learning_rate": 1.4705498331640797e-05, + "loss": 0.0723, + "step": 7399 + }, + { + "epoch": 0.8022549869904597, + "grad_norm": 2.212031841278076, + "learning_rate": 1.470477295807341e-05, + "loss": 0.025, + "step": 7400 + }, + { + "epoch": 0.8023633998265395, + "grad_norm": 2.7670114040374756, + "learning_rate": 1.4704047584506023e-05, + "loss": 0.0509, + "step": 7401 + }, + { + "epoch": 0.8024718126626192, + "grad_norm": 2.40830659866333, + "learning_rate": 1.4703322210938635e-05, + "loss": 0.0348, + "step": 7402 + }, + { + "epoch": 0.802580225498699, + "grad_norm": 1.3283783197402954, + "learning_rate": 1.4702596837371249e-05, + "loss": 0.0154, + "step": 7403 + }, + { + "epoch": 0.8026886383347789, + "grad_norm": 6.863318920135498, + "learning_rate": 1.470187146380386e-05, + "loss": 0.0188, + "step": 7404 + }, + { + "epoch": 0.8027970511708586, + "grad_norm": 3.152294874191284, + "learning_rate": 1.4701146090236473e-05, + "loss": 0.0222, + "step": 7405 + }, + { + "epoch": 0.8029054640069384, + "grad_norm": 2.051318407058716, + "learning_rate": 1.4700420716669087e-05, + "loss": 0.0215, + "step": 7406 + }, + { + "epoch": 0.8030138768430182, + "grad_norm": 1.4044177532196045, + "learning_rate": 1.4699695343101699e-05, + "loss": 0.0241, + "step": 7407 + }, + { + "epoch": 0.803122289679098, + "grad_norm": 1.9588961601257324, + "learning_rate": 1.4698969969534311e-05, + "loss": 0.0218, + "step": 7408 + }, + { + "epoch": 0.8032307025151778, + "grad_norm": 2.7087814807891846, + "learning_rate": 1.4698244595966925e-05, + "loss": 0.0171, + "step": 7409 + }, + { + "epoch": 0.8033391153512576, + "grad_norm": 3.274188280105591, + "learning_rate": 1.4697519222399537e-05, + "loss": 0.0189, + "step": 7410 + }, + { + "epoch": 0.8034475281873373, + "grad_norm": 3.878110408782959, + "learning_rate": 1.4696793848832149e-05, + "loss": 0.066, + "step": 7411 + }, + { + "epoch": 0.8035559410234172, + "grad_norm": 0.5842037200927734, + "learning_rate": 1.4696068475264761e-05, + "loss": 0.014, + "step": 7412 + }, + { + "epoch": 0.803664353859497, + "grad_norm": 1.8792524337768555, + "learning_rate": 1.4695343101697375e-05, + "loss": 0.0285, + "step": 7413 + }, + { + "epoch": 0.8037727666955767, + "grad_norm": 5.731754302978516, + "learning_rate": 1.4694617728129987e-05, + "loss": 0.0285, + "step": 7414 + }, + { + "epoch": 0.8038811795316565, + "grad_norm": 2.059894561767578, + "learning_rate": 1.46938923545626e-05, + "loss": 0.036, + "step": 7415 + }, + { + "epoch": 0.8039895923677364, + "grad_norm": 3.187483787536621, + "learning_rate": 1.4693166980995213e-05, + "loss": 0.0251, + "step": 7416 + }, + { + "epoch": 0.8040980052038161, + "grad_norm": 0.49868205189704895, + "learning_rate": 1.4692441607427827e-05, + "loss": 0.0076, + "step": 7417 + }, + { + "epoch": 0.8042064180398959, + "grad_norm": 5.260468006134033, + "learning_rate": 1.469171623386044e-05, + "loss": 0.0933, + "step": 7418 + }, + { + "epoch": 0.8043148308759757, + "grad_norm": 4.783598899841309, + "learning_rate": 1.4690990860293053e-05, + "loss": 0.0172, + "step": 7419 + }, + { + "epoch": 0.8044232437120555, + "grad_norm": 7.684121131896973, + "learning_rate": 1.4690265486725665e-05, + "loss": 0.0481, + "step": 7420 + }, + { + "epoch": 0.8045316565481353, + "grad_norm": 3.9170875549316406, + "learning_rate": 1.4689540113158279e-05, + "loss": 0.0635, + "step": 7421 + }, + { + "epoch": 0.8046400693842151, + "grad_norm": 3.40087890625, + "learning_rate": 1.4688814739590891e-05, + "loss": 0.0187, + "step": 7422 + }, + { + "epoch": 0.8047484822202949, + "grad_norm": 3.360121488571167, + "learning_rate": 1.4688089366023503e-05, + "loss": 0.0779, + "step": 7423 + }, + { + "epoch": 0.8048568950563747, + "grad_norm": 3.4409704208374023, + "learning_rate": 1.4687363992456115e-05, + "loss": 0.0558, + "step": 7424 + }, + { + "epoch": 0.8049653078924545, + "grad_norm": 1.2079391479492188, + "learning_rate": 1.468663861888873e-05, + "loss": 0.0262, + "step": 7425 + }, + { + "epoch": 0.8050737207285342, + "grad_norm": 3.1094143390655518, + "learning_rate": 1.4685913245321341e-05, + "loss": 0.0229, + "step": 7426 + }, + { + "epoch": 0.8051821335646141, + "grad_norm": 3.342973232269287, + "learning_rate": 1.4685187871753954e-05, + "loss": 0.0568, + "step": 7427 + }, + { + "epoch": 0.8052905464006939, + "grad_norm": 0.8217092156410217, + "learning_rate": 1.4684462498186567e-05, + "loss": 0.0102, + "step": 7428 + }, + { + "epoch": 0.8053989592367736, + "grad_norm": 1.0624761581420898, + "learning_rate": 1.468373712461918e-05, + "loss": 0.0118, + "step": 7429 + }, + { + "epoch": 0.8055073720728534, + "grad_norm": 3.0242459774017334, + "learning_rate": 1.4683011751051792e-05, + "loss": 0.022, + "step": 7430 + }, + { + "epoch": 0.8056157849089333, + "grad_norm": 3.3426437377929688, + "learning_rate": 1.4682286377484405e-05, + "loss": 0.0417, + "step": 7431 + }, + { + "epoch": 0.805724197745013, + "grad_norm": 8.987101554870605, + "learning_rate": 1.4681561003917018e-05, + "loss": 0.0453, + "step": 7432 + }, + { + "epoch": 0.8058326105810928, + "grad_norm": 4.2114129066467285, + "learning_rate": 1.468083563034963e-05, + "loss": 0.0722, + "step": 7433 + }, + { + "epoch": 0.8059410234171726, + "grad_norm": 6.4394707679748535, + "learning_rate": 1.4680110256782245e-05, + "loss": 0.0268, + "step": 7434 + }, + { + "epoch": 0.8060494362532524, + "grad_norm": 2.215193510055542, + "learning_rate": 1.4679384883214857e-05, + "loss": 0.023, + "step": 7435 + }, + { + "epoch": 0.8061578490893322, + "grad_norm": 6.761296272277832, + "learning_rate": 1.4678659509647471e-05, + "loss": 0.095, + "step": 7436 + }, + { + "epoch": 0.806266261925412, + "grad_norm": 8.563437461853027, + "learning_rate": 1.4677934136080083e-05, + "loss": 0.0567, + "step": 7437 + }, + { + "epoch": 0.8063746747614917, + "grad_norm": 0.755210816860199, + "learning_rate": 1.4677208762512696e-05, + "loss": 0.0138, + "step": 7438 + }, + { + "epoch": 0.8064830875975716, + "grad_norm": 10.294390678405762, + "learning_rate": 1.4676483388945308e-05, + "loss": 0.0676, + "step": 7439 + }, + { + "epoch": 0.8065915004336514, + "grad_norm": 1.6889779567718506, + "learning_rate": 1.4675758015377922e-05, + "loss": 0.044, + "step": 7440 + }, + { + "epoch": 0.8066999132697311, + "grad_norm": 5.335559844970703, + "learning_rate": 1.4675032641810534e-05, + "loss": 0.0682, + "step": 7441 + }, + { + "epoch": 0.8068083261058109, + "grad_norm": 5.853392601013184, + "learning_rate": 1.4674307268243146e-05, + "loss": 0.0557, + "step": 7442 + }, + { + "epoch": 0.8069167389418908, + "grad_norm": 2.9757537841796875, + "learning_rate": 1.467358189467576e-05, + "loss": 0.062, + "step": 7443 + }, + { + "epoch": 0.8070251517779705, + "grad_norm": 1.8573309183120728, + "learning_rate": 1.4672856521108372e-05, + "loss": 0.0551, + "step": 7444 + }, + { + "epoch": 0.8071335646140503, + "grad_norm": 0.6337743401527405, + "learning_rate": 1.4672131147540984e-05, + "loss": 0.0083, + "step": 7445 + }, + { + "epoch": 0.80724197745013, + "grad_norm": 4.107543468475342, + "learning_rate": 1.4671405773973598e-05, + "loss": 0.0582, + "step": 7446 + }, + { + "epoch": 0.8073503902862099, + "grad_norm": 1.1406792402267456, + "learning_rate": 1.467068040040621e-05, + "loss": 0.0053, + "step": 7447 + }, + { + "epoch": 0.8074588031222897, + "grad_norm": 2.1360819339752197, + "learning_rate": 1.4669955026838822e-05, + "loss": 0.0133, + "step": 7448 + }, + { + "epoch": 0.8075672159583694, + "grad_norm": 2.182910680770874, + "learning_rate": 1.4669229653271436e-05, + "loss": 0.0488, + "step": 7449 + }, + { + "epoch": 0.8076756287944492, + "grad_norm": 3.2413344383239746, + "learning_rate": 1.4668504279704048e-05, + "loss": 0.0232, + "step": 7450 + }, + { + "epoch": 0.8077840416305291, + "grad_norm": 1.9500691890716553, + "learning_rate": 1.4667778906136662e-05, + "loss": 0.0389, + "step": 7451 + }, + { + "epoch": 0.8078924544666088, + "grad_norm": 3.14015531539917, + "learning_rate": 1.4667053532569276e-05, + "loss": 0.0424, + "step": 7452 + }, + { + "epoch": 0.8080008673026886, + "grad_norm": 2.645610809326172, + "learning_rate": 1.4666328159001888e-05, + "loss": 0.0617, + "step": 7453 + }, + { + "epoch": 0.8081092801387685, + "grad_norm": 4.518786430358887, + "learning_rate": 1.46656027854345e-05, + "loss": 0.0628, + "step": 7454 + }, + { + "epoch": 0.8082176929748482, + "grad_norm": 0.31590259075164795, + "learning_rate": 1.4664877411867114e-05, + "loss": 0.0053, + "step": 7455 + }, + { + "epoch": 0.808326105810928, + "grad_norm": 0.8084898591041565, + "learning_rate": 1.4664152038299726e-05, + "loss": 0.0127, + "step": 7456 + }, + { + "epoch": 0.8084345186470078, + "grad_norm": 1.0286434888839722, + "learning_rate": 1.4663426664732338e-05, + "loss": 0.0149, + "step": 7457 + }, + { + "epoch": 0.8085429314830876, + "grad_norm": 1.7005419731140137, + "learning_rate": 1.4662701291164952e-05, + "loss": 0.0198, + "step": 7458 + }, + { + "epoch": 0.8086513443191674, + "grad_norm": 3.809764862060547, + "learning_rate": 1.4661975917597564e-05, + "loss": 0.0309, + "step": 7459 + }, + { + "epoch": 0.8087597571552472, + "grad_norm": 2.033198833465576, + "learning_rate": 1.4661250544030176e-05, + "loss": 0.0316, + "step": 7460 + }, + { + "epoch": 0.8088681699913269, + "grad_norm": 0.8397952914237976, + "learning_rate": 1.466052517046279e-05, + "loss": 0.0216, + "step": 7461 + }, + { + "epoch": 0.8089765828274068, + "grad_norm": 1.3727445602416992, + "learning_rate": 1.4659799796895402e-05, + "loss": 0.0114, + "step": 7462 + }, + { + "epoch": 0.8090849956634866, + "grad_norm": 1.7018802165985107, + "learning_rate": 1.4659074423328014e-05, + "loss": 0.014, + "step": 7463 + }, + { + "epoch": 0.8091934084995663, + "grad_norm": 1.2430520057678223, + "learning_rate": 1.4658349049760627e-05, + "loss": 0.0142, + "step": 7464 + }, + { + "epoch": 0.8093018213356461, + "grad_norm": 2.5307140350341797, + "learning_rate": 1.465762367619324e-05, + "loss": 0.0213, + "step": 7465 + }, + { + "epoch": 0.809410234171726, + "grad_norm": 3.7069814205169678, + "learning_rate": 1.4656898302625853e-05, + "loss": 0.037, + "step": 7466 + }, + { + "epoch": 0.8095186470078057, + "grad_norm": 5.160876274108887, + "learning_rate": 1.4656172929058465e-05, + "loss": 0.0762, + "step": 7467 + }, + { + "epoch": 0.8096270598438855, + "grad_norm": 3.779785633087158, + "learning_rate": 1.4655447555491079e-05, + "loss": 0.0357, + "step": 7468 + }, + { + "epoch": 0.8097354726799653, + "grad_norm": 1.7075942754745483, + "learning_rate": 1.4654722181923692e-05, + "loss": 0.0168, + "step": 7469 + }, + { + "epoch": 0.8098438855160451, + "grad_norm": 2.8681702613830566, + "learning_rate": 1.4653996808356306e-05, + "loss": 0.0094, + "step": 7470 + }, + { + "epoch": 0.8099522983521249, + "grad_norm": 4.505555152893066, + "learning_rate": 1.4653271434788918e-05, + "loss": 0.0887, + "step": 7471 + }, + { + "epoch": 0.8100607111882047, + "grad_norm": 6.578101634979248, + "learning_rate": 1.465254606122153e-05, + "loss": 0.0452, + "step": 7472 + }, + { + "epoch": 0.8101691240242844, + "grad_norm": 2.287055253982544, + "learning_rate": 1.4651820687654144e-05, + "loss": 0.0733, + "step": 7473 + }, + { + "epoch": 0.8102775368603643, + "grad_norm": 2.3911337852478027, + "learning_rate": 1.4651095314086756e-05, + "loss": 0.071, + "step": 7474 + }, + { + "epoch": 0.8103859496964441, + "grad_norm": 2.2632596492767334, + "learning_rate": 1.4650369940519369e-05, + "loss": 0.0355, + "step": 7475 + }, + { + "epoch": 0.8104943625325238, + "grad_norm": 6.184582710266113, + "learning_rate": 1.464964456695198e-05, + "loss": 0.0763, + "step": 7476 + }, + { + "epoch": 0.8106027753686036, + "grad_norm": 3.7528350353240967, + "learning_rate": 1.4648919193384595e-05, + "loss": 0.0227, + "step": 7477 + }, + { + "epoch": 0.8107111882046835, + "grad_norm": 2.1816015243530273, + "learning_rate": 1.4648193819817207e-05, + "loss": 0.0525, + "step": 7478 + }, + { + "epoch": 0.8108196010407632, + "grad_norm": 4.2957987785339355, + "learning_rate": 1.4647468446249819e-05, + "loss": 0.1021, + "step": 7479 + }, + { + "epoch": 0.810928013876843, + "grad_norm": 7.076997756958008, + "learning_rate": 1.4646743072682433e-05, + "loss": 0.0746, + "step": 7480 + }, + { + "epoch": 0.8110364267129229, + "grad_norm": 8.571798324584961, + "learning_rate": 1.4646017699115045e-05, + "loss": 0.0505, + "step": 7481 + }, + { + "epoch": 0.8111448395490026, + "grad_norm": 2.5053577423095703, + "learning_rate": 1.4645292325547657e-05, + "loss": 0.047, + "step": 7482 + }, + { + "epoch": 0.8112532523850824, + "grad_norm": 3.112351655960083, + "learning_rate": 1.464456695198027e-05, + "loss": 0.0567, + "step": 7483 + }, + { + "epoch": 0.8113616652211622, + "grad_norm": 2.6368868350982666, + "learning_rate": 1.4643841578412883e-05, + "loss": 0.0533, + "step": 7484 + }, + { + "epoch": 0.811470078057242, + "grad_norm": 2.552844762802124, + "learning_rate": 1.4643116204845495e-05, + "loss": 0.0189, + "step": 7485 + }, + { + "epoch": 0.8115784908933218, + "grad_norm": 2.3163256645202637, + "learning_rate": 1.464239083127811e-05, + "loss": 0.0617, + "step": 7486 + }, + { + "epoch": 0.8116869037294016, + "grad_norm": 1.1630926132202148, + "learning_rate": 1.4641665457710723e-05, + "loss": 0.0214, + "step": 7487 + }, + { + "epoch": 0.8117953165654813, + "grad_norm": 1.664515495300293, + "learning_rate": 1.4640940084143337e-05, + "loss": 0.0335, + "step": 7488 + }, + { + "epoch": 0.8119037294015612, + "grad_norm": 2.1548333168029785, + "learning_rate": 1.4640214710575949e-05, + "loss": 0.024, + "step": 7489 + }, + { + "epoch": 0.812012142237641, + "grad_norm": 0.9812610745429993, + "learning_rate": 1.4639489337008561e-05, + "loss": 0.0141, + "step": 7490 + }, + { + "epoch": 0.8121205550737207, + "grad_norm": 1.3912203311920166, + "learning_rate": 1.4638763963441173e-05, + "loss": 0.0349, + "step": 7491 + }, + { + "epoch": 0.8122289679098005, + "grad_norm": 4.983728885650635, + "learning_rate": 1.4638038589873787e-05, + "loss": 0.0464, + "step": 7492 + }, + { + "epoch": 0.8123373807458804, + "grad_norm": 2.4338619709014893, + "learning_rate": 1.4637313216306399e-05, + "loss": 0.0128, + "step": 7493 + }, + { + "epoch": 0.8124457935819601, + "grad_norm": 0.9159445762634277, + "learning_rate": 1.4636587842739011e-05, + "loss": 0.0077, + "step": 7494 + }, + { + "epoch": 0.8125542064180399, + "grad_norm": 1.3931094408035278, + "learning_rate": 1.4635862469171625e-05, + "loss": 0.0154, + "step": 7495 + }, + { + "epoch": 0.8126626192541196, + "grad_norm": 1.194000482559204, + "learning_rate": 1.4635137095604237e-05, + "loss": 0.0288, + "step": 7496 + }, + { + "epoch": 0.8127710320901995, + "grad_norm": 2.603917121887207, + "learning_rate": 1.463441172203685e-05, + "loss": 0.0374, + "step": 7497 + }, + { + "epoch": 0.8128794449262793, + "grad_norm": 2.1229469776153564, + "learning_rate": 1.4633686348469463e-05, + "loss": 0.0394, + "step": 7498 + }, + { + "epoch": 0.812987857762359, + "grad_norm": 2.453918218612671, + "learning_rate": 1.4632960974902075e-05, + "loss": 0.039, + "step": 7499 + }, + { + "epoch": 0.8130962705984388, + "grad_norm": 5.74997615814209, + "learning_rate": 1.4632235601334687e-05, + "loss": 0.0761, + "step": 7500 + }, + { + "epoch": 0.8132046834345187, + "grad_norm": 3.0498225688934326, + "learning_rate": 1.4631510227767301e-05, + "loss": 0.0215, + "step": 7501 + }, + { + "epoch": 0.8133130962705984, + "grad_norm": 3.27068829536438, + "learning_rate": 1.4630784854199913e-05, + "loss": 0.0297, + "step": 7502 + }, + { + "epoch": 0.8134215091066782, + "grad_norm": 2.6548454761505127, + "learning_rate": 1.4630059480632527e-05, + "loss": 0.0279, + "step": 7503 + }, + { + "epoch": 0.813529921942758, + "grad_norm": 3.5219688415527344, + "learning_rate": 1.4629334107065141e-05, + "loss": 0.0245, + "step": 7504 + }, + { + "epoch": 0.8136383347788378, + "grad_norm": 3.868643045425415, + "learning_rate": 1.4628608733497753e-05, + "loss": 0.0419, + "step": 7505 + }, + { + "epoch": 0.8137467476149176, + "grad_norm": 0.6059176921844482, + "learning_rate": 1.4627883359930365e-05, + "loss": 0.0039, + "step": 7506 + }, + { + "epoch": 0.8138551604509974, + "grad_norm": 3.1934776306152344, + "learning_rate": 1.462715798636298e-05, + "loss": 0.017, + "step": 7507 + }, + { + "epoch": 0.8139635732870771, + "grad_norm": 3.3750290870666504, + "learning_rate": 1.4626432612795591e-05, + "loss": 0.0608, + "step": 7508 + }, + { + "epoch": 0.814071986123157, + "grad_norm": 3.318378210067749, + "learning_rate": 1.4625707239228203e-05, + "loss": 0.0545, + "step": 7509 + }, + { + "epoch": 0.8141803989592368, + "grad_norm": 3.6892945766448975, + "learning_rate": 1.4624981865660817e-05, + "loss": 0.0708, + "step": 7510 + }, + { + "epoch": 0.8142888117953165, + "grad_norm": 4.913559436798096, + "learning_rate": 1.462425649209343e-05, + "loss": 0.0379, + "step": 7511 + }, + { + "epoch": 0.8143972246313964, + "grad_norm": 3.582918167114258, + "learning_rate": 1.4623531118526042e-05, + "loss": 0.0235, + "step": 7512 + }, + { + "epoch": 0.8145056374674762, + "grad_norm": 1.352967381477356, + "learning_rate": 1.4622805744958655e-05, + "loss": 0.0131, + "step": 7513 + }, + { + "epoch": 0.8146140503035559, + "grad_norm": 0.9825380444526672, + "learning_rate": 1.4622080371391268e-05, + "loss": 0.0157, + "step": 7514 + }, + { + "epoch": 0.8147224631396357, + "grad_norm": 2.136942148208618, + "learning_rate": 1.462135499782388e-05, + "loss": 0.0281, + "step": 7515 + }, + { + "epoch": 0.8148308759757156, + "grad_norm": 2.3394205570220947, + "learning_rate": 1.4620629624256492e-05, + "loss": 0.0344, + "step": 7516 + }, + { + "epoch": 0.8149392888117953, + "grad_norm": 3.266730785369873, + "learning_rate": 1.4619904250689106e-05, + "loss": 0.0354, + "step": 7517 + }, + { + "epoch": 0.8150477016478751, + "grad_norm": 6.7171783447265625, + "learning_rate": 1.4619178877121718e-05, + "loss": 0.019, + "step": 7518 + }, + { + "epoch": 0.8151561144839549, + "grad_norm": 0.9255543947219849, + "learning_rate": 1.461845350355433e-05, + "loss": 0.0087, + "step": 7519 + }, + { + "epoch": 0.8152645273200347, + "grad_norm": 3.0809543132781982, + "learning_rate": 1.4617728129986946e-05, + "loss": 0.011, + "step": 7520 + }, + { + "epoch": 0.8153729401561145, + "grad_norm": 3.9142396450042725, + "learning_rate": 1.4617002756419558e-05, + "loss": 0.053, + "step": 7521 + }, + { + "epoch": 0.8154813529921943, + "grad_norm": 1.881406307220459, + "learning_rate": 1.4616277382852172e-05, + "loss": 0.0088, + "step": 7522 + }, + { + "epoch": 0.815589765828274, + "grad_norm": 1.6375586986541748, + "learning_rate": 1.4615552009284784e-05, + "loss": 0.0277, + "step": 7523 + }, + { + "epoch": 0.8156981786643539, + "grad_norm": 2.994612693786621, + "learning_rate": 1.4614826635717396e-05, + "loss": 0.0703, + "step": 7524 + }, + { + "epoch": 0.8158065915004337, + "grad_norm": 5.865778923034668, + "learning_rate": 1.461410126215001e-05, + "loss": 0.027, + "step": 7525 + }, + { + "epoch": 0.8159150043365134, + "grad_norm": 1.1688761711120605, + "learning_rate": 1.4613375888582622e-05, + "loss": 0.01, + "step": 7526 + }, + { + "epoch": 0.8160234171725932, + "grad_norm": 5.630274295806885, + "learning_rate": 1.4612650515015234e-05, + "loss": 0.0349, + "step": 7527 + }, + { + "epoch": 0.8161318300086731, + "grad_norm": 1.4609063863754272, + "learning_rate": 1.4611925141447846e-05, + "loss": 0.019, + "step": 7528 + }, + { + "epoch": 0.8162402428447528, + "grad_norm": 1.9040143489837646, + "learning_rate": 1.461119976788046e-05, + "loss": 0.0418, + "step": 7529 + }, + { + "epoch": 0.8163486556808326, + "grad_norm": 0.599441409111023, + "learning_rate": 1.4610474394313072e-05, + "loss": 0.0089, + "step": 7530 + }, + { + "epoch": 0.8164570685169124, + "grad_norm": 2.4043197631835938, + "learning_rate": 1.4609749020745684e-05, + "loss": 0.0394, + "step": 7531 + }, + { + "epoch": 0.8165654813529922, + "grad_norm": 2.816117525100708, + "learning_rate": 1.4609023647178298e-05, + "loss": 0.0401, + "step": 7532 + }, + { + "epoch": 0.816673894189072, + "grad_norm": 4.440270900726318, + "learning_rate": 1.460829827361091e-05, + "loss": 0.0124, + "step": 7533 + }, + { + "epoch": 0.8167823070251518, + "grad_norm": 10.413166046142578, + "learning_rate": 1.4607572900043522e-05, + "loss": 0.0741, + "step": 7534 + }, + { + "epoch": 0.8168907198612315, + "grad_norm": 8.362428665161133, + "learning_rate": 1.4606847526476136e-05, + "loss": 0.0374, + "step": 7535 + }, + { + "epoch": 0.8169991326973114, + "grad_norm": 5.216920852661133, + "learning_rate": 1.4606122152908748e-05, + "loss": 0.0812, + "step": 7536 + }, + { + "epoch": 0.8171075455333912, + "grad_norm": 5.2106428146362305, + "learning_rate": 1.460539677934136e-05, + "loss": 0.0353, + "step": 7537 + }, + { + "epoch": 0.8172159583694709, + "grad_norm": 3.677908182144165, + "learning_rate": 1.4604671405773976e-05, + "loss": 0.0293, + "step": 7538 + }, + { + "epoch": 0.8173243712055508, + "grad_norm": 3.654573440551758, + "learning_rate": 1.4603946032206588e-05, + "loss": 0.0583, + "step": 7539 + }, + { + "epoch": 0.8174327840416306, + "grad_norm": 4.928186416625977, + "learning_rate": 1.4603220658639202e-05, + "loss": 0.0688, + "step": 7540 + }, + { + "epoch": 0.8175411968777103, + "grad_norm": 1.2752293348312378, + "learning_rate": 1.4602495285071814e-05, + "loss": 0.0162, + "step": 7541 + }, + { + "epoch": 0.8176496097137901, + "grad_norm": 9.350554466247559, + "learning_rate": 1.4601769911504426e-05, + "loss": 0.0481, + "step": 7542 + }, + { + "epoch": 0.81775802254987, + "grad_norm": 1.5565788745880127, + "learning_rate": 1.4601044537937038e-05, + "loss": 0.0257, + "step": 7543 + }, + { + "epoch": 0.8178664353859497, + "grad_norm": 1.8289422988891602, + "learning_rate": 1.4600319164369652e-05, + "loss": 0.0316, + "step": 7544 + }, + { + "epoch": 0.8179748482220295, + "grad_norm": 2.842924118041992, + "learning_rate": 1.4599593790802264e-05, + "loss": 0.064, + "step": 7545 + }, + { + "epoch": 0.8180832610581092, + "grad_norm": 2.5596563816070557, + "learning_rate": 1.4598868417234877e-05, + "loss": 0.0325, + "step": 7546 + }, + { + "epoch": 0.8181916738941891, + "grad_norm": 3.3205623626708984, + "learning_rate": 1.459814304366749e-05, + "loss": 0.035, + "step": 7547 + }, + { + "epoch": 0.8183000867302689, + "grad_norm": 1.5702931880950928, + "learning_rate": 1.4597417670100102e-05, + "loss": 0.03, + "step": 7548 + }, + { + "epoch": 0.8184084995663486, + "grad_norm": 2.9773054122924805, + "learning_rate": 1.4596692296532715e-05, + "loss": 0.035, + "step": 7549 + }, + { + "epoch": 0.8185169124024284, + "grad_norm": 3.2820475101470947, + "learning_rate": 1.4595966922965328e-05, + "loss": 0.0501, + "step": 7550 + }, + { + "epoch": 0.8186253252385083, + "grad_norm": 2.0909626483917236, + "learning_rate": 1.459524154939794e-05, + "loss": 0.0424, + "step": 7551 + }, + { + "epoch": 0.818733738074588, + "grad_norm": 2.5140774250030518, + "learning_rate": 1.4594516175830553e-05, + "loss": 0.0365, + "step": 7552 + }, + { + "epoch": 0.8188421509106678, + "grad_norm": 2.9958853721618652, + "learning_rate": 1.4593790802263167e-05, + "loss": 0.0618, + "step": 7553 + }, + { + "epoch": 0.8189505637467476, + "grad_norm": 3.1913931369781494, + "learning_rate": 1.4593065428695779e-05, + "loss": 0.0555, + "step": 7554 + }, + { + "epoch": 0.8190589765828274, + "grad_norm": 1.3870381116867065, + "learning_rate": 1.4592340055128393e-05, + "loss": 0.0452, + "step": 7555 + }, + { + "epoch": 0.8191673894189072, + "grad_norm": 1.8854713439941406, + "learning_rate": 1.4591614681561006e-05, + "loss": 0.0089, + "step": 7556 + }, + { + "epoch": 0.819275802254987, + "grad_norm": 1.2206634283065796, + "learning_rate": 1.4590889307993619e-05, + "loss": 0.0175, + "step": 7557 + }, + { + "epoch": 0.8193842150910667, + "grad_norm": 4.235910415649414, + "learning_rate": 1.459016393442623e-05, + "loss": 0.0979, + "step": 7558 + }, + { + "epoch": 0.8194926279271466, + "grad_norm": 1.9401626586914062, + "learning_rate": 1.4589438560858845e-05, + "loss": 0.0623, + "step": 7559 + }, + { + "epoch": 0.8196010407632264, + "grad_norm": 1.2863483428955078, + "learning_rate": 1.4588713187291457e-05, + "loss": 0.0307, + "step": 7560 + }, + { + "epoch": 0.8197094535993061, + "grad_norm": 4.6174492835998535, + "learning_rate": 1.4587987813724069e-05, + "loss": 0.0538, + "step": 7561 + }, + { + "epoch": 0.8198178664353859, + "grad_norm": 1.8751379251480103, + "learning_rate": 1.4587262440156683e-05, + "loss": 0.0312, + "step": 7562 + }, + { + "epoch": 0.8199262792714658, + "grad_norm": 1.2698575258255005, + "learning_rate": 1.4586537066589295e-05, + "loss": 0.0281, + "step": 7563 + }, + { + "epoch": 0.8200346921075455, + "grad_norm": 2.79097580909729, + "learning_rate": 1.4585811693021907e-05, + "loss": 0.0432, + "step": 7564 + }, + { + "epoch": 0.8201431049436253, + "grad_norm": 2.544207811355591, + "learning_rate": 1.458508631945452e-05, + "loss": 0.02, + "step": 7565 + }, + { + "epoch": 0.8202515177797051, + "grad_norm": 1.0425843000411987, + "learning_rate": 1.4584360945887133e-05, + "loss": 0.0078, + "step": 7566 + }, + { + "epoch": 0.8203599306157849, + "grad_norm": 1.9594730138778687, + "learning_rate": 1.4583635572319745e-05, + "loss": 0.0396, + "step": 7567 + }, + { + "epoch": 0.8204683434518647, + "grad_norm": 2.6407601833343506, + "learning_rate": 1.4582910198752357e-05, + "loss": 0.064, + "step": 7568 + }, + { + "epoch": 0.8205767562879445, + "grad_norm": 3.9506678581237793, + "learning_rate": 1.4582184825184971e-05, + "loss": 0.0478, + "step": 7569 + }, + { + "epoch": 0.8206851691240243, + "grad_norm": 1.8665238618850708, + "learning_rate": 1.4581459451617583e-05, + "loss": 0.0381, + "step": 7570 + }, + { + "epoch": 0.8207935819601041, + "grad_norm": 2.111774444580078, + "learning_rate": 1.4580734078050195e-05, + "loss": 0.0512, + "step": 7571 + }, + { + "epoch": 0.8209019947961839, + "grad_norm": 3.1290767192840576, + "learning_rate": 1.4580008704482811e-05, + "loss": 0.0617, + "step": 7572 + }, + { + "epoch": 0.8210104076322636, + "grad_norm": 2.091543197631836, + "learning_rate": 1.4579283330915423e-05, + "loss": 0.0232, + "step": 7573 + }, + { + "epoch": 0.8211188204683435, + "grad_norm": 2.432965040206909, + "learning_rate": 1.4578557957348037e-05, + "loss": 0.0408, + "step": 7574 + }, + { + "epoch": 0.8212272333044233, + "grad_norm": 0.8215516209602356, + "learning_rate": 1.4577832583780649e-05, + "loss": 0.0156, + "step": 7575 + }, + { + "epoch": 0.821335646140503, + "grad_norm": 0.5986636281013489, + "learning_rate": 1.4577107210213261e-05, + "loss": 0.0124, + "step": 7576 + }, + { + "epoch": 0.8214440589765828, + "grad_norm": 2.0777666568756104, + "learning_rate": 1.4576381836645875e-05, + "loss": 0.0511, + "step": 7577 + }, + { + "epoch": 0.8215524718126627, + "grad_norm": 1.555815577507019, + "learning_rate": 1.4575656463078487e-05, + "loss": 0.0288, + "step": 7578 + }, + { + "epoch": 0.8216608846487424, + "grad_norm": 3.1729655265808105, + "learning_rate": 1.45749310895111e-05, + "loss": 0.0187, + "step": 7579 + }, + { + "epoch": 0.8217692974848222, + "grad_norm": 4.70890998840332, + "learning_rate": 1.4574205715943711e-05, + "loss": 0.0331, + "step": 7580 + }, + { + "epoch": 0.821877710320902, + "grad_norm": 2.23895001411438, + "learning_rate": 1.4573480342376325e-05, + "loss": 0.037, + "step": 7581 + }, + { + "epoch": 0.8219861231569818, + "grad_norm": 3.7022783756256104, + "learning_rate": 1.4572754968808937e-05, + "loss": 0.0509, + "step": 7582 + }, + { + "epoch": 0.8220945359930616, + "grad_norm": 2.024789333343506, + "learning_rate": 1.457202959524155e-05, + "loss": 0.0307, + "step": 7583 + }, + { + "epoch": 0.8222029488291414, + "grad_norm": 1.5345216989517212, + "learning_rate": 1.4571304221674163e-05, + "loss": 0.0324, + "step": 7584 + }, + { + "epoch": 0.8223113616652211, + "grad_norm": 4.336341857910156, + "learning_rate": 1.4570578848106776e-05, + "loss": 0.061, + "step": 7585 + }, + { + "epoch": 0.822419774501301, + "grad_norm": 1.8076099157333374, + "learning_rate": 1.4569853474539388e-05, + "loss": 0.0158, + "step": 7586 + }, + { + "epoch": 0.8225281873373808, + "grad_norm": 0.8708553314208984, + "learning_rate": 1.4569128100972001e-05, + "loss": 0.0104, + "step": 7587 + }, + { + "epoch": 0.8226366001734605, + "grad_norm": 2.502078056335449, + "learning_rate": 1.4568402727404614e-05, + "loss": 0.0655, + "step": 7588 + }, + { + "epoch": 0.8227450130095403, + "grad_norm": 6.795393943786621, + "learning_rate": 1.4567677353837226e-05, + "loss": 0.0685, + "step": 7589 + }, + { + "epoch": 0.8228534258456202, + "grad_norm": 3.1559672355651855, + "learning_rate": 1.4566951980269841e-05, + "loss": 0.0284, + "step": 7590 + }, + { + "epoch": 0.8229618386816999, + "grad_norm": 0.6633749604225159, + "learning_rate": 1.4566226606702453e-05, + "loss": 0.0083, + "step": 7591 + }, + { + "epoch": 0.8230702515177797, + "grad_norm": 2.8238205909729004, + "learning_rate": 1.4565501233135067e-05, + "loss": 0.0188, + "step": 7592 + }, + { + "epoch": 0.8231786643538594, + "grad_norm": 4.32706356048584, + "learning_rate": 1.456477585956768e-05, + "loss": 0.0703, + "step": 7593 + }, + { + "epoch": 0.8232870771899393, + "grad_norm": 1.0046510696411133, + "learning_rate": 1.4564050486000292e-05, + "loss": 0.007, + "step": 7594 + }, + { + "epoch": 0.8233954900260191, + "grad_norm": 2.6848766803741455, + "learning_rate": 1.4563325112432904e-05, + "loss": 0.0604, + "step": 7595 + }, + { + "epoch": 0.8235039028620988, + "grad_norm": 2.231985330581665, + "learning_rate": 1.4562599738865518e-05, + "loss": 0.0479, + "step": 7596 + }, + { + "epoch": 0.8236123156981786, + "grad_norm": 1.8772141933441162, + "learning_rate": 1.456187436529813e-05, + "loss": 0.0179, + "step": 7597 + }, + { + "epoch": 0.8237207285342585, + "grad_norm": 4.928905487060547, + "learning_rate": 1.4561148991730742e-05, + "loss": 0.0514, + "step": 7598 + }, + { + "epoch": 0.8238291413703382, + "grad_norm": 2.6667091846466064, + "learning_rate": 1.4560423618163356e-05, + "loss": 0.0549, + "step": 7599 + }, + { + "epoch": 0.823937554206418, + "grad_norm": 1.6722919940948486, + "learning_rate": 1.4559698244595968e-05, + "loss": 0.0385, + "step": 7600 + }, + { + "epoch": 0.8240459670424979, + "grad_norm": 1.4931138753890991, + "learning_rate": 1.455897287102858e-05, + "loss": 0.0294, + "step": 7601 + }, + { + "epoch": 0.8241543798785776, + "grad_norm": 6.48368501663208, + "learning_rate": 1.4558247497461194e-05, + "loss": 0.0762, + "step": 7602 + }, + { + "epoch": 0.8242627927146574, + "grad_norm": 3.1588056087493896, + "learning_rate": 1.4557522123893806e-05, + "loss": 0.0587, + "step": 7603 + }, + { + "epoch": 0.8243712055507372, + "grad_norm": 2.75972580909729, + "learning_rate": 1.4556796750326418e-05, + "loss": 0.0492, + "step": 7604 + }, + { + "epoch": 0.824479618386817, + "grad_norm": 2.099327325820923, + "learning_rate": 1.4556071376759032e-05, + "loss": 0.0275, + "step": 7605 + }, + { + "epoch": 0.8245880312228968, + "grad_norm": 2.123668909072876, + "learning_rate": 1.4555346003191644e-05, + "loss": 0.0647, + "step": 7606 + }, + { + "epoch": 0.8246964440589766, + "grad_norm": 2.5944130420684814, + "learning_rate": 1.4554620629624258e-05, + "loss": 0.0592, + "step": 7607 + }, + { + "epoch": 0.8248048568950563, + "grad_norm": 2.108914852142334, + "learning_rate": 1.4553895256056872e-05, + "loss": 0.0642, + "step": 7608 + }, + { + "epoch": 0.8249132697311362, + "grad_norm": 2.010817527770996, + "learning_rate": 1.4553169882489484e-05, + "loss": 0.0305, + "step": 7609 + }, + { + "epoch": 0.825021682567216, + "grad_norm": 2.685663938522339, + "learning_rate": 1.4552444508922096e-05, + "loss": 0.0757, + "step": 7610 + }, + { + "epoch": 0.8251300954032957, + "grad_norm": 3.3161442279815674, + "learning_rate": 1.455171913535471e-05, + "loss": 0.0467, + "step": 7611 + }, + { + "epoch": 0.8252385082393755, + "grad_norm": 0.8738700747489929, + "learning_rate": 1.4550993761787322e-05, + "loss": 0.0354, + "step": 7612 + }, + { + "epoch": 0.8253469210754554, + "grad_norm": 2.6227939128875732, + "learning_rate": 1.4550268388219934e-05, + "loss": 0.0371, + "step": 7613 + }, + { + "epoch": 0.8254553339115351, + "grad_norm": 2.788888454437256, + "learning_rate": 1.4549543014652548e-05, + "loss": 0.0394, + "step": 7614 + }, + { + "epoch": 0.8255637467476149, + "grad_norm": 1.4119529724121094, + "learning_rate": 1.454881764108516e-05, + "loss": 0.0157, + "step": 7615 + }, + { + "epoch": 0.8256721595836947, + "grad_norm": 1.5410587787628174, + "learning_rate": 1.4548092267517772e-05, + "loss": 0.0326, + "step": 7616 + }, + { + "epoch": 0.8257805724197745, + "grad_norm": 1.8513729572296143, + "learning_rate": 1.4547366893950386e-05, + "loss": 0.0316, + "step": 7617 + }, + { + "epoch": 0.8258889852558543, + "grad_norm": 2.5428171157836914, + "learning_rate": 1.4546641520382998e-05, + "loss": 0.0231, + "step": 7618 + }, + { + "epoch": 0.8259973980919341, + "grad_norm": 1.9107699394226074, + "learning_rate": 1.454591614681561e-05, + "loss": 0.034, + "step": 7619 + }, + { + "epoch": 0.8261058109280138, + "grad_norm": 1.6103811264038086, + "learning_rate": 1.4545190773248223e-05, + "loss": 0.0825, + "step": 7620 + }, + { + "epoch": 0.8262142237640937, + "grad_norm": 0.8631187677383423, + "learning_rate": 1.4544465399680836e-05, + "loss": 0.0072, + "step": 7621 + }, + { + "epoch": 0.8263226366001735, + "grad_norm": 1.8391716480255127, + "learning_rate": 1.4543740026113449e-05, + "loss": 0.0165, + "step": 7622 + }, + { + "epoch": 0.8264310494362532, + "grad_norm": 2.102722644805908, + "learning_rate": 1.454301465254606e-05, + "loss": 0.056, + "step": 7623 + }, + { + "epoch": 0.826539462272333, + "grad_norm": 1.7095409631729126, + "learning_rate": 1.4542289278978676e-05, + "loss": 0.0192, + "step": 7624 + }, + { + "epoch": 0.8266478751084129, + "grad_norm": 6.021890640258789, + "learning_rate": 1.4541563905411288e-05, + "loss": 0.0868, + "step": 7625 + }, + { + "epoch": 0.8267562879444926, + "grad_norm": 1.3260810375213623, + "learning_rate": 1.4540838531843902e-05, + "loss": 0.0233, + "step": 7626 + }, + { + "epoch": 0.8268647007805724, + "grad_norm": 2.835767984390259, + "learning_rate": 1.4540113158276514e-05, + "loss": 0.0205, + "step": 7627 + }, + { + "epoch": 0.8269731136166523, + "grad_norm": 2.8763701915740967, + "learning_rate": 1.4539387784709126e-05, + "loss": 0.0397, + "step": 7628 + }, + { + "epoch": 0.827081526452732, + "grad_norm": 2.26567006111145, + "learning_rate": 1.453866241114174e-05, + "loss": 0.0722, + "step": 7629 + }, + { + "epoch": 0.8271899392888118, + "grad_norm": 1.9279849529266357, + "learning_rate": 1.4537937037574352e-05, + "loss": 0.037, + "step": 7630 + }, + { + "epoch": 0.8272983521248916, + "grad_norm": 7.313032150268555, + "learning_rate": 1.4537211664006965e-05, + "loss": 0.0379, + "step": 7631 + }, + { + "epoch": 0.8274067649609714, + "grad_norm": 2.7023143768310547, + "learning_rate": 1.4536486290439578e-05, + "loss": 0.0381, + "step": 7632 + }, + { + "epoch": 0.8275151777970512, + "grad_norm": 4.62299919128418, + "learning_rate": 1.453576091687219e-05, + "loss": 0.05, + "step": 7633 + }, + { + "epoch": 0.827623590633131, + "grad_norm": 1.0039207935333252, + "learning_rate": 1.4535035543304803e-05, + "loss": 0.0083, + "step": 7634 + }, + { + "epoch": 0.8277320034692107, + "grad_norm": 1.517611026763916, + "learning_rate": 1.4534310169737415e-05, + "loss": 0.0281, + "step": 7635 + }, + { + "epoch": 0.8278404163052906, + "grad_norm": 1.4257577657699585, + "learning_rate": 1.4533584796170029e-05, + "loss": 0.0104, + "step": 7636 + }, + { + "epoch": 0.8279488291413704, + "grad_norm": 2.77443790435791, + "learning_rate": 1.4532859422602641e-05, + "loss": 0.0137, + "step": 7637 + }, + { + "epoch": 0.8280572419774501, + "grad_norm": 2.9786338806152344, + "learning_rate": 1.4532134049035253e-05, + "loss": 0.0422, + "step": 7638 + }, + { + "epoch": 0.8281656548135299, + "grad_norm": 1.86813485622406, + "learning_rate": 1.4531408675467867e-05, + "loss": 0.0218, + "step": 7639 + }, + { + "epoch": 0.8282740676496098, + "grad_norm": 2.3342926502227783, + "learning_rate": 1.4530683301900479e-05, + "loss": 0.0428, + "step": 7640 + }, + { + "epoch": 0.8283824804856895, + "grad_norm": 2.3921773433685303, + "learning_rate": 1.4529957928333095e-05, + "loss": 0.0244, + "step": 7641 + }, + { + "epoch": 0.8284908933217693, + "grad_norm": 2.583728313446045, + "learning_rate": 1.4529232554765707e-05, + "loss": 0.0234, + "step": 7642 + }, + { + "epoch": 0.828599306157849, + "grad_norm": 6.578575134277344, + "learning_rate": 1.4528507181198319e-05, + "loss": 0.0484, + "step": 7643 + }, + { + "epoch": 0.8287077189939289, + "grad_norm": 2.647932529449463, + "learning_rate": 1.4527781807630933e-05, + "loss": 0.0363, + "step": 7644 + }, + { + "epoch": 0.8288161318300087, + "grad_norm": 2.621654987335205, + "learning_rate": 1.4527056434063545e-05, + "loss": 0.043, + "step": 7645 + }, + { + "epoch": 0.8289245446660884, + "grad_norm": 5.1866302490234375, + "learning_rate": 1.4526331060496157e-05, + "loss": 0.0445, + "step": 7646 + }, + { + "epoch": 0.8290329575021682, + "grad_norm": 5.70870304107666, + "learning_rate": 1.4525605686928769e-05, + "loss": 0.0265, + "step": 7647 + }, + { + "epoch": 0.8291413703382481, + "grad_norm": 3.697869300842285, + "learning_rate": 1.4524880313361383e-05, + "loss": 0.0274, + "step": 7648 + }, + { + "epoch": 0.8292497831743278, + "grad_norm": 9.059117317199707, + "learning_rate": 1.4524154939793995e-05, + "loss": 0.1016, + "step": 7649 + }, + { + "epoch": 0.8293581960104076, + "grad_norm": 1.1937212944030762, + "learning_rate": 1.4523429566226607e-05, + "loss": 0.0242, + "step": 7650 + }, + { + "epoch": 0.8294666088464874, + "grad_norm": 3.048828601837158, + "learning_rate": 1.4522704192659221e-05, + "loss": 0.0616, + "step": 7651 + }, + { + "epoch": 0.8295750216825672, + "grad_norm": 4.498175144195557, + "learning_rate": 1.4521978819091833e-05, + "loss": 0.0646, + "step": 7652 + }, + { + "epoch": 0.829683434518647, + "grad_norm": 3.037543535232544, + "learning_rate": 1.4521253445524445e-05, + "loss": 0.0614, + "step": 7653 + }, + { + "epoch": 0.8297918473547268, + "grad_norm": 1.251584768295288, + "learning_rate": 1.4520528071957059e-05, + "loss": 0.0308, + "step": 7654 + }, + { + "epoch": 0.8299002601908065, + "grad_norm": 1.5661499500274658, + "learning_rate": 1.4519802698389671e-05, + "loss": 0.0432, + "step": 7655 + }, + { + "epoch": 0.8300086730268864, + "grad_norm": 2.175278663635254, + "learning_rate": 1.4519077324822283e-05, + "loss": 0.0434, + "step": 7656 + }, + { + "epoch": 0.8301170858629662, + "grad_norm": 1.2746717929840088, + "learning_rate": 1.4518351951254897e-05, + "loss": 0.0229, + "step": 7657 + }, + { + "epoch": 0.8302254986990459, + "grad_norm": 1.7791407108306885, + "learning_rate": 1.451762657768751e-05, + "loss": 0.023, + "step": 7658 + }, + { + "epoch": 0.8303339115351258, + "grad_norm": 2.5440220832824707, + "learning_rate": 1.4516901204120123e-05, + "loss": 0.0299, + "step": 7659 + }, + { + "epoch": 0.8304423243712056, + "grad_norm": 1.6549842357635498, + "learning_rate": 1.4516175830552737e-05, + "loss": 0.0308, + "step": 7660 + }, + { + "epoch": 0.8305507372072853, + "grad_norm": 13.417601585388184, + "learning_rate": 1.451545045698535e-05, + "loss": 0.0633, + "step": 7661 + }, + { + "epoch": 0.8306591500433651, + "grad_norm": 2.4436094760894775, + "learning_rate": 1.4514725083417961e-05, + "loss": 0.1105, + "step": 7662 + }, + { + "epoch": 0.830767562879445, + "grad_norm": 4.09722900390625, + "learning_rate": 1.4513999709850575e-05, + "loss": 0.0296, + "step": 7663 + }, + { + "epoch": 0.8308759757155247, + "grad_norm": 4.973878383636475, + "learning_rate": 1.4513274336283187e-05, + "loss": 0.0378, + "step": 7664 + }, + { + "epoch": 0.8309843885516045, + "grad_norm": 0.654225766658783, + "learning_rate": 1.45125489627158e-05, + "loss": 0.0068, + "step": 7665 + }, + { + "epoch": 0.8310928013876843, + "grad_norm": 1.5308928489685059, + "learning_rate": 1.4511823589148413e-05, + "loss": 0.0119, + "step": 7666 + }, + { + "epoch": 0.8312012142237641, + "grad_norm": 1.7247563600540161, + "learning_rate": 1.4511098215581025e-05, + "loss": 0.0226, + "step": 7667 + }, + { + "epoch": 0.8313096270598439, + "grad_norm": 3.5459539890289307, + "learning_rate": 1.4510372842013638e-05, + "loss": 0.0566, + "step": 7668 + }, + { + "epoch": 0.8314180398959237, + "grad_norm": 2.826563835144043, + "learning_rate": 1.4509647468446251e-05, + "loss": 0.0695, + "step": 7669 + }, + { + "epoch": 0.8315264527320034, + "grad_norm": 2.2659053802490234, + "learning_rate": 1.4508922094878864e-05, + "loss": 0.0438, + "step": 7670 + }, + { + "epoch": 0.8316348655680833, + "grad_norm": 1.4873698949813843, + "learning_rate": 1.4508196721311476e-05, + "loss": 0.0264, + "step": 7671 + }, + { + "epoch": 0.8317432784041631, + "grad_norm": 2.7409942150115967, + "learning_rate": 1.4507471347744088e-05, + "loss": 0.0319, + "step": 7672 + }, + { + "epoch": 0.8318516912402428, + "grad_norm": 1.5971659421920776, + "learning_rate": 1.4506745974176702e-05, + "loss": 0.0541, + "step": 7673 + }, + { + "epoch": 0.8319601040763226, + "grad_norm": 3.98539137840271, + "learning_rate": 1.4506020600609314e-05, + "loss": 0.0438, + "step": 7674 + }, + { + "epoch": 0.8320685169124025, + "grad_norm": 1.2681523561477661, + "learning_rate": 1.4505295227041926e-05, + "loss": 0.0266, + "step": 7675 + }, + { + "epoch": 0.8321769297484822, + "grad_norm": 4.004456520080566, + "learning_rate": 1.4504569853474542e-05, + "loss": 0.0439, + "step": 7676 + }, + { + "epoch": 0.832285342584562, + "grad_norm": 2.7462990283966064, + "learning_rate": 1.4503844479907154e-05, + "loss": 0.0698, + "step": 7677 + }, + { + "epoch": 0.8323937554206418, + "grad_norm": 2.375459671020508, + "learning_rate": 1.4503119106339768e-05, + "loss": 0.0195, + "step": 7678 + }, + { + "epoch": 0.8325021682567216, + "grad_norm": 2.249215602874756, + "learning_rate": 1.450239373277238e-05, + "loss": 0.0311, + "step": 7679 + }, + { + "epoch": 0.8326105810928014, + "grad_norm": 3.856605291366577, + "learning_rate": 1.4501668359204992e-05, + "loss": 0.0312, + "step": 7680 + }, + { + "epoch": 0.8327189939288812, + "grad_norm": 0.9704845547676086, + "learning_rate": 1.4500942985637606e-05, + "loss": 0.0197, + "step": 7681 + }, + { + "epoch": 0.8328274067649609, + "grad_norm": 2.164980411529541, + "learning_rate": 1.4500217612070218e-05, + "loss": 0.055, + "step": 7682 + }, + { + "epoch": 0.8329358196010408, + "grad_norm": 2.8754541873931885, + "learning_rate": 1.449949223850283e-05, + "loss": 0.0425, + "step": 7683 + }, + { + "epoch": 0.8330442324371206, + "grad_norm": 0.6995577812194824, + "learning_rate": 1.4498766864935444e-05, + "loss": 0.0101, + "step": 7684 + }, + { + "epoch": 0.8331526452732003, + "grad_norm": 2.5339462757110596, + "learning_rate": 1.4498041491368056e-05, + "loss": 0.0524, + "step": 7685 + }, + { + "epoch": 0.8332610581092801, + "grad_norm": 4.614051818847656, + "learning_rate": 1.4497316117800668e-05, + "loss": 0.026, + "step": 7686 + }, + { + "epoch": 0.83336947094536, + "grad_norm": 3.5328176021575928, + "learning_rate": 1.449659074423328e-05, + "loss": 0.0713, + "step": 7687 + }, + { + "epoch": 0.8334778837814397, + "grad_norm": 1.2250009775161743, + "learning_rate": 1.4495865370665894e-05, + "loss": 0.0296, + "step": 7688 + }, + { + "epoch": 0.8335862966175195, + "grad_norm": 3.4186511039733887, + "learning_rate": 1.4495139997098506e-05, + "loss": 0.0905, + "step": 7689 + }, + { + "epoch": 0.8336947094535994, + "grad_norm": 1.7899284362792969, + "learning_rate": 1.4494414623531118e-05, + "loss": 0.0412, + "step": 7690 + }, + { + "epoch": 0.8338031222896791, + "grad_norm": 1.1291985511779785, + "learning_rate": 1.4493689249963732e-05, + "loss": 0.0135, + "step": 7691 + }, + { + "epoch": 0.8339115351257589, + "grad_norm": 2.825929641723633, + "learning_rate": 1.4492963876396344e-05, + "loss": 0.0296, + "step": 7692 + }, + { + "epoch": 0.8340199479618386, + "grad_norm": 2.1810436248779297, + "learning_rate": 1.449223850282896e-05, + "loss": 0.0384, + "step": 7693 + }, + { + "epoch": 0.8341283607979185, + "grad_norm": 3.1375200748443604, + "learning_rate": 1.4491513129261572e-05, + "loss": 0.0145, + "step": 7694 + }, + { + "epoch": 0.8342367736339983, + "grad_norm": 4.833943843841553, + "learning_rate": 1.4490787755694184e-05, + "loss": 0.0478, + "step": 7695 + }, + { + "epoch": 0.834345186470078, + "grad_norm": 1.0326917171478271, + "learning_rate": 1.4490062382126798e-05, + "loss": 0.0467, + "step": 7696 + }, + { + "epoch": 0.8344535993061578, + "grad_norm": 0.9707934260368347, + "learning_rate": 1.448933700855941e-05, + "loss": 0.023, + "step": 7697 + }, + { + "epoch": 0.8345620121422377, + "grad_norm": 2.296102285385132, + "learning_rate": 1.4488611634992022e-05, + "loss": 0.0429, + "step": 7698 + }, + { + "epoch": 0.8346704249783174, + "grad_norm": 1.6517490148544312, + "learning_rate": 1.4487886261424634e-05, + "loss": 0.0334, + "step": 7699 + }, + { + "epoch": 0.8347788378143972, + "grad_norm": 2.7574362754821777, + "learning_rate": 1.4487160887857248e-05, + "loss": 0.057, + "step": 7700 + }, + { + "epoch": 0.834887250650477, + "grad_norm": 1.8272428512573242, + "learning_rate": 1.448643551428986e-05, + "loss": 0.0559, + "step": 7701 + }, + { + "epoch": 0.8349956634865568, + "grad_norm": 2.7416977882385254, + "learning_rate": 1.4485710140722473e-05, + "loss": 0.019, + "step": 7702 + }, + { + "epoch": 0.8351040763226366, + "grad_norm": 1.130539059638977, + "learning_rate": 1.4484984767155086e-05, + "loss": 0.0167, + "step": 7703 + }, + { + "epoch": 0.8352124891587164, + "grad_norm": 1.3917025327682495, + "learning_rate": 1.4484259393587699e-05, + "loss": 0.0184, + "step": 7704 + }, + { + "epoch": 0.8353209019947961, + "grad_norm": 1.9547743797302246, + "learning_rate": 1.448353402002031e-05, + "loss": 0.0486, + "step": 7705 + }, + { + "epoch": 0.835429314830876, + "grad_norm": 1.2226145267486572, + "learning_rate": 1.4482808646452924e-05, + "loss": 0.0271, + "step": 7706 + }, + { + "epoch": 0.8355377276669558, + "grad_norm": 1.5874848365783691, + "learning_rate": 1.4482083272885537e-05, + "loss": 0.0202, + "step": 7707 + }, + { + "epoch": 0.8356461405030355, + "grad_norm": 2.3124797344207764, + "learning_rate": 1.4481357899318149e-05, + "loss": 0.0275, + "step": 7708 + }, + { + "epoch": 0.8357545533391153, + "grad_norm": 4.686172962188721, + "learning_rate": 1.4480632525750763e-05, + "loss": 0.0352, + "step": 7709 + }, + { + "epoch": 0.8358629661751952, + "grad_norm": 4.442769527435303, + "learning_rate": 1.4479907152183376e-05, + "loss": 0.036, + "step": 7710 + }, + { + "epoch": 0.8359713790112749, + "grad_norm": 3.1839561462402344, + "learning_rate": 1.4479181778615989e-05, + "loss": 0.087, + "step": 7711 + }, + { + "epoch": 0.8360797918473547, + "grad_norm": 2.601142406463623, + "learning_rate": 1.4478456405048602e-05, + "loss": 0.0137, + "step": 7712 + }, + { + "epoch": 0.8361882046834345, + "grad_norm": 3.853170394897461, + "learning_rate": 1.4477731031481215e-05, + "loss": 0.0451, + "step": 7713 + }, + { + "epoch": 0.8362966175195143, + "grad_norm": 1.6385387182235718, + "learning_rate": 1.4477005657913827e-05, + "loss": 0.0222, + "step": 7714 + }, + { + "epoch": 0.8364050303555941, + "grad_norm": 3.5524914264678955, + "learning_rate": 1.447628028434644e-05, + "loss": 0.0397, + "step": 7715 + }, + { + "epoch": 0.8365134431916739, + "grad_norm": 2.475252151489258, + "learning_rate": 1.4475554910779053e-05, + "loss": 0.0443, + "step": 7716 + }, + { + "epoch": 0.8366218560277537, + "grad_norm": 0.979454755783081, + "learning_rate": 1.4474829537211665e-05, + "loss": 0.0078, + "step": 7717 + }, + { + "epoch": 0.8367302688638335, + "grad_norm": 0.948959469795227, + "learning_rate": 1.4474104163644279e-05, + "loss": 0.0254, + "step": 7718 + }, + { + "epoch": 0.8368386816999133, + "grad_norm": 2.9385180473327637, + "learning_rate": 1.447337879007689e-05, + "loss": 0.0205, + "step": 7719 + }, + { + "epoch": 0.836947094535993, + "grad_norm": 8.376883506774902, + "learning_rate": 1.4472653416509503e-05, + "loss": 0.0977, + "step": 7720 + }, + { + "epoch": 0.8370555073720729, + "grad_norm": 3.194550037384033, + "learning_rate": 1.4471928042942117e-05, + "loss": 0.1022, + "step": 7721 + }, + { + "epoch": 0.8371639202081527, + "grad_norm": 2.0021510124206543, + "learning_rate": 1.4471202669374729e-05, + "loss": 0.0445, + "step": 7722 + }, + { + "epoch": 0.8372723330442324, + "grad_norm": 2.1858882904052734, + "learning_rate": 1.4470477295807341e-05, + "loss": 0.0259, + "step": 7723 + }, + { + "epoch": 0.8373807458803122, + "grad_norm": 3.120351791381836, + "learning_rate": 1.4469751922239953e-05, + "loss": 0.0398, + "step": 7724 + }, + { + "epoch": 0.8374891587163921, + "grad_norm": 0.6467716097831726, + "learning_rate": 1.4469026548672567e-05, + "loss": 0.0089, + "step": 7725 + }, + { + "epoch": 0.8375975715524718, + "grad_norm": 2.0375778675079346, + "learning_rate": 1.446830117510518e-05, + "loss": 0.0379, + "step": 7726 + }, + { + "epoch": 0.8377059843885516, + "grad_norm": 2.2931017875671387, + "learning_rate": 1.4467575801537791e-05, + "loss": 0.018, + "step": 7727 + }, + { + "epoch": 0.8378143972246314, + "grad_norm": 3.465770721435547, + "learning_rate": 1.4466850427970407e-05, + "loss": 0.0559, + "step": 7728 + }, + { + "epoch": 0.8379228100607112, + "grad_norm": 1.617780327796936, + "learning_rate": 1.4466125054403019e-05, + "loss": 0.0157, + "step": 7729 + }, + { + "epoch": 0.838031222896791, + "grad_norm": 1.6902738809585571, + "learning_rate": 1.4465399680835633e-05, + "loss": 0.041, + "step": 7730 + }, + { + "epoch": 0.8381396357328708, + "grad_norm": 3.9142093658447266, + "learning_rate": 1.4464674307268245e-05, + "loss": 0.0362, + "step": 7731 + }, + { + "epoch": 0.8382480485689505, + "grad_norm": 6.2219462394714355, + "learning_rate": 1.4463948933700857e-05, + "loss": 0.0292, + "step": 7732 + }, + { + "epoch": 0.8383564614050304, + "grad_norm": 1.9923968315124512, + "learning_rate": 1.4463223560133471e-05, + "loss": 0.0403, + "step": 7733 + }, + { + "epoch": 0.8384648742411102, + "grad_norm": 2.736924648284912, + "learning_rate": 1.4462498186566083e-05, + "loss": 0.0693, + "step": 7734 + }, + { + "epoch": 0.8385732870771899, + "grad_norm": 3.7836859226226807, + "learning_rate": 1.4461772812998695e-05, + "loss": 0.0315, + "step": 7735 + }, + { + "epoch": 0.8386816999132697, + "grad_norm": 7.215294361114502, + "learning_rate": 1.4461047439431309e-05, + "loss": 0.0383, + "step": 7736 + }, + { + "epoch": 0.8387901127493496, + "grad_norm": 1.3358334302902222, + "learning_rate": 1.4460322065863921e-05, + "loss": 0.0594, + "step": 7737 + }, + { + "epoch": 0.8388985255854293, + "grad_norm": 4.1032490730285645, + "learning_rate": 1.4459596692296533e-05, + "loss": 0.0564, + "step": 7738 + }, + { + "epoch": 0.8390069384215091, + "grad_norm": 1.0647828578948975, + "learning_rate": 1.4458871318729146e-05, + "loss": 0.0084, + "step": 7739 + }, + { + "epoch": 0.8391153512575888, + "grad_norm": 3.456160306930542, + "learning_rate": 1.445814594516176e-05, + "loss": 0.0496, + "step": 7740 + }, + { + "epoch": 0.8392237640936687, + "grad_norm": 4.760965824127197, + "learning_rate": 1.4457420571594372e-05, + "loss": 0.0314, + "step": 7741 + }, + { + "epoch": 0.8393321769297485, + "grad_norm": 2.0686287879943848, + "learning_rate": 1.4456695198026984e-05, + "loss": 0.0144, + "step": 7742 + }, + { + "epoch": 0.8394405897658282, + "grad_norm": 2.0892064571380615, + "learning_rate": 1.4455969824459598e-05, + "loss": 0.0255, + "step": 7743 + }, + { + "epoch": 0.839549002601908, + "grad_norm": 2.398925542831421, + "learning_rate": 1.445524445089221e-05, + "loss": 0.0524, + "step": 7744 + }, + { + "epoch": 0.8396574154379879, + "grad_norm": 2.8766415119171143, + "learning_rate": 1.4454519077324825e-05, + "loss": 0.0369, + "step": 7745 + }, + { + "epoch": 0.8397658282740676, + "grad_norm": 2.478670835494995, + "learning_rate": 1.4453793703757437e-05, + "loss": 0.0164, + "step": 7746 + }, + { + "epoch": 0.8398742411101474, + "grad_norm": 3.771129846572876, + "learning_rate": 1.445306833019005e-05, + "loss": 0.09, + "step": 7747 + }, + { + "epoch": 0.8399826539462273, + "grad_norm": 2.1386444568634033, + "learning_rate": 1.4452342956622663e-05, + "loss": 0.0299, + "step": 7748 + }, + { + "epoch": 0.840091066782307, + "grad_norm": 6.908462047576904, + "learning_rate": 1.4451617583055275e-05, + "loss": 0.0534, + "step": 7749 + }, + { + "epoch": 0.8401994796183868, + "grad_norm": 1.6600723266601562, + "learning_rate": 1.4450892209487888e-05, + "loss": 0.0324, + "step": 7750 + }, + { + "epoch": 0.8403078924544666, + "grad_norm": 3.528848886489868, + "learning_rate": 1.44501668359205e-05, + "loss": 0.0559, + "step": 7751 + }, + { + "epoch": 0.8404163052905465, + "grad_norm": 2.8508899211883545, + "learning_rate": 1.4449441462353114e-05, + "loss": 0.0412, + "step": 7752 + }, + { + "epoch": 0.8405247181266262, + "grad_norm": 1.4889405965805054, + "learning_rate": 1.4448716088785726e-05, + "loss": 0.0117, + "step": 7753 + }, + { + "epoch": 0.840633130962706, + "grad_norm": 3.1071054935455322, + "learning_rate": 1.4447990715218338e-05, + "loss": 0.055, + "step": 7754 + }, + { + "epoch": 0.8407415437987857, + "grad_norm": 2.52227783203125, + "learning_rate": 1.4447265341650952e-05, + "loss": 0.0127, + "step": 7755 + }, + { + "epoch": 0.8408499566348656, + "grad_norm": 0.7668010592460632, + "learning_rate": 1.4446539968083564e-05, + "loss": 0.0191, + "step": 7756 + }, + { + "epoch": 0.8409583694709454, + "grad_norm": 1.0669728517532349, + "learning_rate": 1.4445814594516176e-05, + "loss": 0.0148, + "step": 7757 + }, + { + "epoch": 0.8410667823070251, + "grad_norm": 2.6311745643615723, + "learning_rate": 1.444508922094879e-05, + "loss": 0.0582, + "step": 7758 + }, + { + "epoch": 0.8411751951431049, + "grad_norm": 1.0219371318817139, + "learning_rate": 1.4444363847381402e-05, + "loss": 0.0229, + "step": 7759 + }, + { + "epoch": 0.8412836079791848, + "grad_norm": 7.209897518157959, + "learning_rate": 1.4443638473814014e-05, + "loss": 0.0379, + "step": 7760 + }, + { + "epoch": 0.8413920208152645, + "grad_norm": 0.47937747836112976, + "learning_rate": 1.4442913100246628e-05, + "loss": 0.007, + "step": 7761 + }, + { + "epoch": 0.8415004336513443, + "grad_norm": 0.6242983937263489, + "learning_rate": 1.4442187726679242e-05, + "loss": 0.0159, + "step": 7762 + }, + { + "epoch": 0.8416088464874241, + "grad_norm": 6.624815940856934, + "learning_rate": 1.4441462353111854e-05, + "loss": 0.075, + "step": 7763 + }, + { + "epoch": 0.8417172593235039, + "grad_norm": 1.2903333902359009, + "learning_rate": 1.4440736979544468e-05, + "loss": 0.0131, + "step": 7764 + }, + { + "epoch": 0.8418256721595837, + "grad_norm": 1.6097382307052612, + "learning_rate": 1.444001160597708e-05, + "loss": 0.0442, + "step": 7765 + }, + { + "epoch": 0.8419340849956635, + "grad_norm": 4.588756084442139, + "learning_rate": 1.4439286232409692e-05, + "loss": 0.0833, + "step": 7766 + }, + { + "epoch": 0.8420424978317432, + "grad_norm": 2.043297052383423, + "learning_rate": 1.4438560858842306e-05, + "loss": 0.0488, + "step": 7767 + }, + { + "epoch": 0.8421509106678231, + "grad_norm": 0.6301753520965576, + "learning_rate": 1.4437835485274918e-05, + "loss": 0.0071, + "step": 7768 + }, + { + "epoch": 0.8422593235039029, + "grad_norm": 3.1913328170776367, + "learning_rate": 1.443711011170753e-05, + "loss": 0.0216, + "step": 7769 + }, + { + "epoch": 0.8423677363399826, + "grad_norm": 5.898576259613037, + "learning_rate": 1.4436384738140144e-05, + "loss": 0.0472, + "step": 7770 + }, + { + "epoch": 0.8424761491760624, + "grad_norm": 3.3665342330932617, + "learning_rate": 1.4435659364572756e-05, + "loss": 0.034, + "step": 7771 + }, + { + "epoch": 0.8425845620121423, + "grad_norm": 1.4376904964447021, + "learning_rate": 1.4434933991005368e-05, + "loss": 0.0165, + "step": 7772 + }, + { + "epoch": 0.842692974848222, + "grad_norm": 2.7225730419158936, + "learning_rate": 1.4434208617437982e-05, + "loss": 0.0411, + "step": 7773 + }, + { + "epoch": 0.8428013876843018, + "grad_norm": 4.420592784881592, + "learning_rate": 1.4433483243870594e-05, + "loss": 0.0676, + "step": 7774 + }, + { + "epoch": 0.8429098005203816, + "grad_norm": 3.15403151512146, + "learning_rate": 1.4432757870303206e-05, + "loss": 0.0591, + "step": 7775 + }, + { + "epoch": 0.8430182133564614, + "grad_norm": 1.8755394220352173, + "learning_rate": 1.4432032496735819e-05, + "loss": 0.0308, + "step": 7776 + }, + { + "epoch": 0.8431266261925412, + "grad_norm": 4.208742141723633, + "learning_rate": 1.4431307123168432e-05, + "loss": 0.0342, + "step": 7777 + }, + { + "epoch": 0.843235039028621, + "grad_norm": 4.7053704261779785, + "learning_rate": 1.4430581749601045e-05, + "loss": 0.0409, + "step": 7778 + }, + { + "epoch": 0.8433434518647008, + "grad_norm": 1.5048739910125732, + "learning_rate": 1.4429856376033657e-05, + "loss": 0.0265, + "step": 7779 + }, + { + "epoch": 0.8434518647007806, + "grad_norm": 2.092264175415039, + "learning_rate": 1.4429131002466272e-05, + "loss": 0.0435, + "step": 7780 + }, + { + "epoch": 0.8435602775368604, + "grad_norm": 0.9709409475326538, + "learning_rate": 1.4428405628898884e-05, + "loss": 0.0187, + "step": 7781 + }, + { + "epoch": 0.8436686903729401, + "grad_norm": 1.6964929103851318, + "learning_rate": 1.4427680255331498e-05, + "loss": 0.0346, + "step": 7782 + }, + { + "epoch": 0.84377710320902, + "grad_norm": 2.767270088195801, + "learning_rate": 1.442695488176411e-05, + "loss": 0.0253, + "step": 7783 + }, + { + "epoch": 0.8438855160450998, + "grad_norm": 1.1468371152877808, + "learning_rate": 1.4426229508196722e-05, + "loss": 0.0094, + "step": 7784 + }, + { + "epoch": 0.8439939288811795, + "grad_norm": 1.4340543746948242, + "learning_rate": 1.4425504134629336e-05, + "loss": 0.021, + "step": 7785 + }, + { + "epoch": 0.8441023417172593, + "grad_norm": 12.511056900024414, + "learning_rate": 1.4424778761061948e-05, + "loss": 0.1184, + "step": 7786 + }, + { + "epoch": 0.8442107545533392, + "grad_norm": 2.300508499145508, + "learning_rate": 1.442405338749456e-05, + "loss": 0.0302, + "step": 7787 + }, + { + "epoch": 0.8443191673894189, + "grad_norm": 3.97834849357605, + "learning_rate": 1.4423328013927174e-05, + "loss": 0.0768, + "step": 7788 + }, + { + "epoch": 0.8444275802254987, + "grad_norm": 1.475466251373291, + "learning_rate": 1.4422602640359787e-05, + "loss": 0.0286, + "step": 7789 + }, + { + "epoch": 0.8445359930615784, + "grad_norm": 4.236767292022705, + "learning_rate": 1.4421877266792399e-05, + "loss": 0.0312, + "step": 7790 + }, + { + "epoch": 0.8446444058976583, + "grad_norm": 2.7773118019104004, + "learning_rate": 1.4421151893225011e-05, + "loss": 0.0269, + "step": 7791 + }, + { + "epoch": 0.8447528187337381, + "grad_norm": 1.7169722318649292, + "learning_rate": 1.4420426519657625e-05, + "loss": 0.0262, + "step": 7792 + }, + { + "epoch": 0.8448612315698178, + "grad_norm": 6.976860046386719, + "learning_rate": 1.4419701146090237e-05, + "loss": 0.1097, + "step": 7793 + }, + { + "epoch": 0.8449696444058976, + "grad_norm": 1.229834794998169, + "learning_rate": 1.4418975772522849e-05, + "loss": 0.0235, + "step": 7794 + }, + { + "epoch": 0.8450780572419775, + "grad_norm": 3.0984790325164795, + "learning_rate": 1.4418250398955463e-05, + "loss": 0.0327, + "step": 7795 + }, + { + "epoch": 0.8451864700780573, + "grad_norm": 0.6150912046432495, + "learning_rate": 1.4417525025388075e-05, + "loss": 0.0085, + "step": 7796 + }, + { + "epoch": 0.845294882914137, + "grad_norm": 1.4973503351211548, + "learning_rate": 1.441679965182069e-05, + "loss": 0.0233, + "step": 7797 + }, + { + "epoch": 0.8454032957502168, + "grad_norm": 1.1907776594161987, + "learning_rate": 1.4416074278253303e-05, + "loss": 0.0192, + "step": 7798 + }, + { + "epoch": 0.8455117085862967, + "grad_norm": 4.752594470977783, + "learning_rate": 1.4415348904685915e-05, + "loss": 0.0268, + "step": 7799 + }, + { + "epoch": 0.8456201214223764, + "grad_norm": 1.1614174842834473, + "learning_rate": 1.4414623531118529e-05, + "loss": 0.0081, + "step": 7800 + }, + { + "epoch": 0.8457285342584562, + "grad_norm": 18.582843780517578, + "learning_rate": 1.441389815755114e-05, + "loss": 0.0378, + "step": 7801 + }, + { + "epoch": 0.8458369470945359, + "grad_norm": 1.0677576065063477, + "learning_rate": 1.4413172783983753e-05, + "loss": 0.0123, + "step": 7802 + }, + { + "epoch": 0.8459453599306158, + "grad_norm": 4.619366645812988, + "learning_rate": 1.4412447410416365e-05, + "loss": 0.0578, + "step": 7803 + }, + { + "epoch": 0.8460537727666956, + "grad_norm": 3.9320857524871826, + "learning_rate": 1.4411722036848979e-05, + "loss": 0.0481, + "step": 7804 + }, + { + "epoch": 0.8461621856027753, + "grad_norm": 5.854515075683594, + "learning_rate": 1.4410996663281591e-05, + "loss": 0.1046, + "step": 7805 + }, + { + "epoch": 0.8462705984388552, + "grad_norm": 4.201791763305664, + "learning_rate": 1.4410271289714203e-05, + "loss": 0.0135, + "step": 7806 + }, + { + "epoch": 0.846379011274935, + "grad_norm": 2.0636448860168457, + "learning_rate": 1.4409545916146817e-05, + "loss": 0.0246, + "step": 7807 + }, + { + "epoch": 0.8464874241110147, + "grad_norm": 2.1972804069519043, + "learning_rate": 1.440882054257943e-05, + "loss": 0.0603, + "step": 7808 + }, + { + "epoch": 0.8465958369470945, + "grad_norm": 0.13609185814857483, + "learning_rate": 1.4408095169012041e-05, + "loss": 0.0019, + "step": 7809 + }, + { + "epoch": 0.8467042497831744, + "grad_norm": 3.906399726867676, + "learning_rate": 1.4407369795444655e-05, + "loss": 0.0648, + "step": 7810 + }, + { + "epoch": 0.8468126626192541, + "grad_norm": 8.936698913574219, + "learning_rate": 1.4406644421877267e-05, + "loss": 0.0468, + "step": 7811 + }, + { + "epoch": 0.8469210754553339, + "grad_norm": 2.0184688568115234, + "learning_rate": 1.440591904830988e-05, + "loss": 0.0418, + "step": 7812 + }, + { + "epoch": 0.8470294882914137, + "grad_norm": 1.994962453842163, + "learning_rate": 1.4405193674742493e-05, + "loss": 0.0258, + "step": 7813 + }, + { + "epoch": 0.8471379011274935, + "grad_norm": 3.5950775146484375, + "learning_rate": 1.4404468301175107e-05, + "loss": 0.034, + "step": 7814 + }, + { + "epoch": 0.8472463139635733, + "grad_norm": 2.5218446254730225, + "learning_rate": 1.440374292760772e-05, + "loss": 0.0215, + "step": 7815 + }, + { + "epoch": 0.8473547267996531, + "grad_norm": 3.322178602218628, + "learning_rate": 1.4403017554040333e-05, + "loss": 0.0365, + "step": 7816 + }, + { + "epoch": 0.8474631396357328, + "grad_norm": 2.581732749938965, + "learning_rate": 1.4402292180472945e-05, + "loss": 0.0188, + "step": 7817 + }, + { + "epoch": 0.8475715524718127, + "grad_norm": 0.6198875904083252, + "learning_rate": 1.4401566806905557e-05, + "loss": 0.0094, + "step": 7818 + }, + { + "epoch": 0.8476799653078925, + "grad_norm": 1.5628085136413574, + "learning_rate": 1.4400841433338171e-05, + "loss": 0.026, + "step": 7819 + }, + { + "epoch": 0.8477883781439722, + "grad_norm": 3.82014799118042, + "learning_rate": 1.4400116059770783e-05, + "loss": 0.0617, + "step": 7820 + }, + { + "epoch": 0.847896790980052, + "grad_norm": 0.8127856254577637, + "learning_rate": 1.4399390686203396e-05, + "loss": 0.006, + "step": 7821 + }, + { + "epoch": 0.8480052038161319, + "grad_norm": 1.4028599262237549, + "learning_rate": 1.439866531263601e-05, + "loss": 0.0385, + "step": 7822 + }, + { + "epoch": 0.8481136166522116, + "grad_norm": 2.882957935333252, + "learning_rate": 1.4397939939068621e-05, + "loss": 0.0455, + "step": 7823 + }, + { + "epoch": 0.8482220294882914, + "grad_norm": 0.9399594068527222, + "learning_rate": 1.4397214565501234e-05, + "loss": 0.0067, + "step": 7824 + }, + { + "epoch": 0.8483304423243712, + "grad_norm": 5.227859020233154, + "learning_rate": 1.4396489191933847e-05, + "loss": 0.0443, + "step": 7825 + }, + { + "epoch": 0.848438855160451, + "grad_norm": 1.830885887145996, + "learning_rate": 1.439576381836646e-05, + "loss": 0.0452, + "step": 7826 + }, + { + "epoch": 0.8485472679965308, + "grad_norm": 6.246567249298096, + "learning_rate": 1.4395038444799072e-05, + "loss": 0.0614, + "step": 7827 + }, + { + "epoch": 0.8486556808326106, + "grad_norm": 2.0846621990203857, + "learning_rate": 1.4394313071231686e-05, + "loss": 0.0604, + "step": 7828 + }, + { + "epoch": 0.8487640936686903, + "grad_norm": 3.127516746520996, + "learning_rate": 1.4393587697664298e-05, + "loss": 0.0156, + "step": 7829 + }, + { + "epoch": 0.8488725065047702, + "grad_norm": 2.4184441566467285, + "learning_rate": 1.439286232409691e-05, + "loss": 0.0512, + "step": 7830 + }, + { + "epoch": 0.84898091934085, + "grad_norm": 1.6998144388198853, + "learning_rate": 1.4392136950529525e-05, + "loss": 0.0158, + "step": 7831 + }, + { + "epoch": 0.8490893321769297, + "grad_norm": 4.544145584106445, + "learning_rate": 1.4391411576962138e-05, + "loss": 0.0705, + "step": 7832 + }, + { + "epoch": 0.8491977450130095, + "grad_norm": 1.680920958518982, + "learning_rate": 1.439068620339475e-05, + "loss": 0.0299, + "step": 7833 + }, + { + "epoch": 0.8493061578490894, + "grad_norm": 1.011291742324829, + "learning_rate": 1.4389960829827364e-05, + "loss": 0.008, + "step": 7834 + }, + { + "epoch": 0.8494145706851691, + "grad_norm": 3.8860766887664795, + "learning_rate": 1.4389235456259976e-05, + "loss": 0.0773, + "step": 7835 + }, + { + "epoch": 0.8495229835212489, + "grad_norm": 1.3724474906921387, + "learning_rate": 1.4388510082692588e-05, + "loss": 0.0156, + "step": 7836 + }, + { + "epoch": 0.8496313963573288, + "grad_norm": 2.6468231678009033, + "learning_rate": 1.4387784709125202e-05, + "loss": 0.0382, + "step": 7837 + }, + { + "epoch": 0.8497398091934085, + "grad_norm": 0.5746440291404724, + "learning_rate": 1.4387059335557814e-05, + "loss": 0.006, + "step": 7838 + }, + { + "epoch": 0.8498482220294883, + "grad_norm": 0.9809806942939758, + "learning_rate": 1.4386333961990426e-05, + "loss": 0.011, + "step": 7839 + }, + { + "epoch": 0.849956634865568, + "grad_norm": 3.5501089096069336, + "learning_rate": 1.438560858842304e-05, + "loss": 0.0251, + "step": 7840 + }, + { + "epoch": 0.8500650477016479, + "grad_norm": 1.1942278146743774, + "learning_rate": 1.4384883214855652e-05, + "loss": 0.0095, + "step": 7841 + }, + { + "epoch": 0.8501734605377277, + "grad_norm": 4.104964733123779, + "learning_rate": 1.4384157841288264e-05, + "loss": 0.0725, + "step": 7842 + }, + { + "epoch": 0.8502818733738075, + "grad_norm": 6.381147384643555, + "learning_rate": 1.4383432467720876e-05, + "loss": 0.0295, + "step": 7843 + }, + { + "epoch": 0.8503902862098872, + "grad_norm": 2.0958030223846436, + "learning_rate": 1.438270709415349e-05, + "loss": 0.0559, + "step": 7844 + }, + { + "epoch": 0.8504986990459671, + "grad_norm": 1.3061503171920776, + "learning_rate": 1.4381981720586102e-05, + "loss": 0.0167, + "step": 7845 + }, + { + "epoch": 0.8506071118820469, + "grad_norm": 4.696508884429932, + "learning_rate": 1.4381256347018714e-05, + "loss": 0.0205, + "step": 7846 + }, + { + "epoch": 0.8507155247181266, + "grad_norm": 3.5079169273376465, + "learning_rate": 1.4380530973451328e-05, + "loss": 0.0475, + "step": 7847 + }, + { + "epoch": 0.8508239375542064, + "grad_norm": 2.6185851097106934, + "learning_rate": 1.437980559988394e-05, + "loss": 0.0247, + "step": 7848 + }, + { + "epoch": 0.8509323503902863, + "grad_norm": 3.8638126850128174, + "learning_rate": 1.4379080226316556e-05, + "loss": 0.0668, + "step": 7849 + }, + { + "epoch": 0.851040763226366, + "grad_norm": 2.571821451187134, + "learning_rate": 1.4378354852749168e-05, + "loss": 0.0468, + "step": 7850 + }, + { + "epoch": 0.8511491760624458, + "grad_norm": 4.505301475524902, + "learning_rate": 1.437762947918178e-05, + "loss": 0.0906, + "step": 7851 + }, + { + "epoch": 0.8512575888985255, + "grad_norm": 1.7465318441390991, + "learning_rate": 1.4376904105614394e-05, + "loss": 0.0176, + "step": 7852 + }, + { + "epoch": 0.8513660017346054, + "grad_norm": 1.304469347000122, + "learning_rate": 1.4376178732047006e-05, + "loss": 0.0198, + "step": 7853 + }, + { + "epoch": 0.8514744145706852, + "grad_norm": 2.416959285736084, + "learning_rate": 1.4375453358479618e-05, + "loss": 0.0327, + "step": 7854 + }, + { + "epoch": 0.8515828274067649, + "grad_norm": 2.333500623703003, + "learning_rate": 1.437472798491223e-05, + "loss": 0.0371, + "step": 7855 + }, + { + "epoch": 0.8516912402428447, + "grad_norm": 5.655284404754639, + "learning_rate": 1.4374002611344844e-05, + "loss": 0.0829, + "step": 7856 + }, + { + "epoch": 0.8517996530789246, + "grad_norm": 8.614811897277832, + "learning_rate": 1.4373277237777456e-05, + "loss": 0.0275, + "step": 7857 + }, + { + "epoch": 0.8519080659150043, + "grad_norm": 3.2939112186431885, + "learning_rate": 1.4372551864210069e-05, + "loss": 0.0196, + "step": 7858 + }, + { + "epoch": 0.8520164787510841, + "grad_norm": 3.5342228412628174, + "learning_rate": 1.4371826490642682e-05, + "loss": 0.0405, + "step": 7859 + }, + { + "epoch": 0.8521248915871639, + "grad_norm": 2.8568379878997803, + "learning_rate": 1.4371101117075295e-05, + "loss": 0.0496, + "step": 7860 + }, + { + "epoch": 0.8522333044232437, + "grad_norm": 6.459378719329834, + "learning_rate": 1.4370375743507907e-05, + "loss": 0.0191, + "step": 7861 + }, + { + "epoch": 0.8523417172593235, + "grad_norm": 2.274736166000366, + "learning_rate": 1.436965036994052e-05, + "loss": 0.0528, + "step": 7862 + }, + { + "epoch": 0.8524501300954033, + "grad_norm": 0.760412335395813, + "learning_rate": 1.4368924996373133e-05, + "loss": 0.0165, + "step": 7863 + }, + { + "epoch": 0.852558542931483, + "grad_norm": 2.398339033126831, + "learning_rate": 1.4368199622805745e-05, + "loss": 0.0364, + "step": 7864 + }, + { + "epoch": 0.8526669557675629, + "grad_norm": 1.2623645067214966, + "learning_rate": 1.4367474249238359e-05, + "loss": 0.0256, + "step": 7865 + }, + { + "epoch": 0.8527753686036427, + "grad_norm": 1.066795825958252, + "learning_rate": 1.4366748875670972e-05, + "loss": 0.0205, + "step": 7866 + }, + { + "epoch": 0.8528837814397224, + "grad_norm": 1.4200782775878906, + "learning_rate": 1.4366023502103585e-05, + "loss": 0.0155, + "step": 7867 + }, + { + "epoch": 0.8529921942758023, + "grad_norm": 0.5442965030670166, + "learning_rate": 1.4365298128536198e-05, + "loss": 0.0044, + "step": 7868 + }, + { + "epoch": 0.8531006071118821, + "grad_norm": 3.343388795852661, + "learning_rate": 1.436457275496881e-05, + "loss": 0.0284, + "step": 7869 + }, + { + "epoch": 0.8532090199479618, + "grad_norm": 5.12826681137085, + "learning_rate": 1.4363847381401423e-05, + "loss": 0.03, + "step": 7870 + }, + { + "epoch": 0.8533174327840416, + "grad_norm": 1.8511780500411987, + "learning_rate": 1.4363122007834037e-05, + "loss": 0.0201, + "step": 7871 + }, + { + "epoch": 0.8534258456201215, + "grad_norm": 2.226949453353882, + "learning_rate": 1.4362396634266649e-05, + "loss": 0.0312, + "step": 7872 + }, + { + "epoch": 0.8535342584562012, + "grad_norm": 1.8751212358474731, + "learning_rate": 1.4361671260699261e-05, + "loss": 0.0114, + "step": 7873 + }, + { + "epoch": 0.853642671292281, + "grad_norm": 1.191179871559143, + "learning_rate": 1.4360945887131875e-05, + "loss": 0.0185, + "step": 7874 + }, + { + "epoch": 0.8537510841283608, + "grad_norm": 2.3877806663513184, + "learning_rate": 1.4360220513564487e-05, + "loss": 0.0585, + "step": 7875 + }, + { + "epoch": 0.8538594969644406, + "grad_norm": 2.0774645805358887, + "learning_rate": 1.4359495139997099e-05, + "loss": 0.0309, + "step": 7876 + }, + { + "epoch": 0.8539679098005204, + "grad_norm": 3.0537006855010986, + "learning_rate": 1.4358769766429713e-05, + "loss": 0.0359, + "step": 7877 + }, + { + "epoch": 0.8540763226366002, + "grad_norm": 7.482642650604248, + "learning_rate": 1.4358044392862325e-05, + "loss": 0.0305, + "step": 7878 + }, + { + "epoch": 0.8541847354726799, + "grad_norm": 3.3957812786102295, + "learning_rate": 1.4357319019294937e-05, + "loss": 0.0808, + "step": 7879 + }, + { + "epoch": 0.8542931483087598, + "grad_norm": 2.543931722640991, + "learning_rate": 1.4356593645727551e-05, + "loss": 0.0085, + "step": 7880 + }, + { + "epoch": 0.8544015611448396, + "grad_norm": 5.875619888305664, + "learning_rate": 1.4355868272160163e-05, + "loss": 0.0822, + "step": 7881 + }, + { + "epoch": 0.8545099739809193, + "grad_norm": 3.75769305229187, + "learning_rate": 1.4355142898592775e-05, + "loss": 0.0617, + "step": 7882 + }, + { + "epoch": 0.8546183868169991, + "grad_norm": 3.5077409744262695, + "learning_rate": 1.435441752502539e-05, + "loss": 0.0175, + "step": 7883 + }, + { + "epoch": 0.854726799653079, + "grad_norm": 2.227557897567749, + "learning_rate": 1.4353692151458003e-05, + "loss": 0.0552, + "step": 7884 + }, + { + "epoch": 0.8548352124891587, + "grad_norm": 4.8203349113464355, + "learning_rate": 1.4352966777890615e-05, + "loss": 0.0651, + "step": 7885 + }, + { + "epoch": 0.8549436253252385, + "grad_norm": 4.041932582855225, + "learning_rate": 1.4352241404323229e-05, + "loss": 0.0269, + "step": 7886 + }, + { + "epoch": 0.8550520381613183, + "grad_norm": 4.194319248199463, + "learning_rate": 1.4351516030755841e-05, + "loss": 0.0289, + "step": 7887 + }, + { + "epoch": 0.8551604509973981, + "grad_norm": 8.310619354248047, + "learning_rate": 1.4350790657188453e-05, + "loss": 0.0941, + "step": 7888 + }, + { + "epoch": 0.8552688638334779, + "grad_norm": 3.6021065711975098, + "learning_rate": 1.4350065283621067e-05, + "loss": 0.0167, + "step": 7889 + }, + { + "epoch": 0.8553772766695577, + "grad_norm": 4.401678085327148, + "learning_rate": 1.4349339910053679e-05, + "loss": 0.0388, + "step": 7890 + }, + { + "epoch": 0.8554856895056374, + "grad_norm": 8.598514556884766, + "learning_rate": 1.4348614536486291e-05, + "loss": 0.0531, + "step": 7891 + }, + { + "epoch": 0.8555941023417173, + "grad_norm": 4.2111358642578125, + "learning_rate": 1.4347889162918905e-05, + "loss": 0.0641, + "step": 7892 + }, + { + "epoch": 0.855702515177797, + "grad_norm": 3.388364315032959, + "learning_rate": 1.4347163789351517e-05, + "loss": 0.0245, + "step": 7893 + }, + { + "epoch": 0.8558109280138768, + "grad_norm": 1.8728500604629517, + "learning_rate": 1.434643841578413e-05, + "loss": 0.0476, + "step": 7894 + }, + { + "epoch": 0.8559193408499567, + "grad_norm": 1.5026918649673462, + "learning_rate": 1.4345713042216742e-05, + "loss": 0.0338, + "step": 7895 + }, + { + "epoch": 0.8560277536860365, + "grad_norm": 1.3171316385269165, + "learning_rate": 1.4344987668649355e-05, + "loss": 0.0135, + "step": 7896 + }, + { + "epoch": 0.8561361665221162, + "grad_norm": 2.72672176361084, + "learning_rate": 1.4344262295081968e-05, + "loss": 0.0192, + "step": 7897 + }, + { + "epoch": 0.856244579358196, + "grad_norm": 3.287897825241089, + "learning_rate": 1.434353692151458e-05, + "loss": 0.0418, + "step": 7898 + }, + { + "epoch": 0.8563529921942759, + "grad_norm": 2.912839412689209, + "learning_rate": 1.4342811547947194e-05, + "loss": 0.0393, + "step": 7899 + }, + { + "epoch": 0.8564614050303556, + "grad_norm": 0.8123520016670227, + "learning_rate": 1.4342086174379807e-05, + "loss": 0.0108, + "step": 7900 + }, + { + "epoch": 0.8565698178664354, + "grad_norm": 1.5728330612182617, + "learning_rate": 1.4341360800812421e-05, + "loss": 0.024, + "step": 7901 + }, + { + "epoch": 0.8566782307025151, + "grad_norm": 2.0778541564941406, + "learning_rate": 1.4340635427245033e-05, + "loss": 0.0144, + "step": 7902 + }, + { + "epoch": 0.856786643538595, + "grad_norm": 2.402223825454712, + "learning_rate": 1.4339910053677645e-05, + "loss": 0.0357, + "step": 7903 + }, + { + "epoch": 0.8568950563746748, + "grad_norm": 1.5333987474441528, + "learning_rate": 1.433918468011026e-05, + "loss": 0.037, + "step": 7904 + }, + { + "epoch": 0.8570034692107545, + "grad_norm": 2.9616901874542236, + "learning_rate": 1.4338459306542871e-05, + "loss": 0.0512, + "step": 7905 + }, + { + "epoch": 0.8571118820468343, + "grad_norm": 3.4067704677581787, + "learning_rate": 1.4337733932975484e-05, + "loss": 0.0542, + "step": 7906 + }, + { + "epoch": 0.8572202948829142, + "grad_norm": 0.7370985150337219, + "learning_rate": 1.4337008559408096e-05, + "loss": 0.0087, + "step": 7907 + }, + { + "epoch": 0.8573287077189939, + "grad_norm": 1.3576719760894775, + "learning_rate": 1.433628318584071e-05, + "loss": 0.0095, + "step": 7908 + }, + { + "epoch": 0.8574371205550737, + "grad_norm": 5.346457004547119, + "learning_rate": 1.4335557812273322e-05, + "loss": 0.0655, + "step": 7909 + }, + { + "epoch": 0.8575455333911535, + "grad_norm": 1.857419729232788, + "learning_rate": 1.4334832438705934e-05, + "loss": 0.0628, + "step": 7910 + }, + { + "epoch": 0.8576539462272333, + "grad_norm": 7.7005815505981445, + "learning_rate": 1.4334107065138548e-05, + "loss": 0.0572, + "step": 7911 + }, + { + "epoch": 0.8577623590633131, + "grad_norm": 2.252260684967041, + "learning_rate": 1.433338169157116e-05, + "loss": 0.0542, + "step": 7912 + }, + { + "epoch": 0.8578707718993929, + "grad_norm": 1.4418315887451172, + "learning_rate": 1.4332656318003772e-05, + "loss": 0.0322, + "step": 7913 + }, + { + "epoch": 0.8579791847354726, + "grad_norm": 2.509261131286621, + "learning_rate": 1.4331930944436386e-05, + "loss": 0.0482, + "step": 7914 + }, + { + "epoch": 0.8580875975715525, + "grad_norm": 2.421544075012207, + "learning_rate": 1.4331205570868998e-05, + "loss": 0.0198, + "step": 7915 + }, + { + "epoch": 0.8581960104076323, + "grad_norm": 8.245369911193848, + "learning_rate": 1.433048019730161e-05, + "loss": 0.0868, + "step": 7916 + }, + { + "epoch": 0.858304423243712, + "grad_norm": 2.0613250732421875, + "learning_rate": 1.4329754823734224e-05, + "loss": 0.0141, + "step": 7917 + }, + { + "epoch": 0.8584128360797918, + "grad_norm": 0.7271268367767334, + "learning_rate": 1.4329029450166838e-05, + "loss": 0.0176, + "step": 7918 + }, + { + "epoch": 0.8585212489158717, + "grad_norm": 4.258963584899902, + "learning_rate": 1.432830407659945e-05, + "loss": 0.0599, + "step": 7919 + }, + { + "epoch": 0.8586296617519514, + "grad_norm": 3.650184392929077, + "learning_rate": 1.4327578703032064e-05, + "loss": 0.0431, + "step": 7920 + }, + { + "epoch": 0.8587380745880312, + "grad_norm": 4.528315544128418, + "learning_rate": 1.4326853329464676e-05, + "loss": 0.048, + "step": 7921 + }, + { + "epoch": 0.858846487424111, + "grad_norm": 3.0805792808532715, + "learning_rate": 1.4326127955897288e-05, + "loss": 0.0184, + "step": 7922 + }, + { + "epoch": 0.8589549002601908, + "grad_norm": 2.9207918643951416, + "learning_rate": 1.4325402582329902e-05, + "loss": 0.0221, + "step": 7923 + }, + { + "epoch": 0.8590633130962706, + "grad_norm": 4.687262058258057, + "learning_rate": 1.4324677208762514e-05, + "loss": 0.0254, + "step": 7924 + }, + { + "epoch": 0.8591717259323504, + "grad_norm": 0.5306021571159363, + "learning_rate": 1.4323951835195126e-05, + "loss": 0.0036, + "step": 7925 + }, + { + "epoch": 0.8592801387684302, + "grad_norm": 1.5192605257034302, + "learning_rate": 1.432322646162774e-05, + "loss": 0.0452, + "step": 7926 + }, + { + "epoch": 0.85938855160451, + "grad_norm": 2.149482488632202, + "learning_rate": 1.4322501088060352e-05, + "loss": 0.0539, + "step": 7927 + }, + { + "epoch": 0.8594969644405898, + "grad_norm": 2.514404535293579, + "learning_rate": 1.4321775714492964e-05, + "loss": 0.047, + "step": 7928 + }, + { + "epoch": 0.8596053772766695, + "grad_norm": 3.1017355918884277, + "learning_rate": 1.4321050340925578e-05, + "loss": 0.0457, + "step": 7929 + }, + { + "epoch": 0.8597137901127494, + "grad_norm": 3.960538148880005, + "learning_rate": 1.432032496735819e-05, + "loss": 0.0329, + "step": 7930 + }, + { + "epoch": 0.8598222029488292, + "grad_norm": 3.1817591190338135, + "learning_rate": 1.4319599593790802e-05, + "loss": 0.0446, + "step": 7931 + }, + { + "epoch": 0.8599306157849089, + "grad_norm": 0.884964644908905, + "learning_rate": 1.4318874220223416e-05, + "loss": 0.0116, + "step": 7932 + }, + { + "epoch": 0.8600390286209887, + "grad_norm": 1.7113118171691895, + "learning_rate": 1.4318148846656028e-05, + "loss": 0.0188, + "step": 7933 + }, + { + "epoch": 0.8601474414570686, + "grad_norm": 2.585540533065796, + "learning_rate": 1.431742347308864e-05, + "loss": 0.0092, + "step": 7934 + }, + { + "epoch": 0.8602558542931483, + "grad_norm": 2.165289878845215, + "learning_rate": 1.4316698099521256e-05, + "loss": 0.0388, + "step": 7935 + }, + { + "epoch": 0.8603642671292281, + "grad_norm": 3.3495943546295166, + "learning_rate": 1.4315972725953868e-05, + "loss": 0.0286, + "step": 7936 + }, + { + "epoch": 0.8604726799653079, + "grad_norm": 3.0867693424224854, + "learning_rate": 1.431524735238648e-05, + "loss": 0.0506, + "step": 7937 + }, + { + "epoch": 0.8605810928013877, + "grad_norm": 1.505769968032837, + "learning_rate": 1.4314521978819094e-05, + "loss": 0.026, + "step": 7938 + }, + { + "epoch": 0.8606895056374675, + "grad_norm": 4.461441993713379, + "learning_rate": 1.4313796605251706e-05, + "loss": 0.047, + "step": 7939 + }, + { + "epoch": 0.8607979184735473, + "grad_norm": 2.9051320552825928, + "learning_rate": 1.4313071231684319e-05, + "loss": 0.043, + "step": 7940 + }, + { + "epoch": 0.860906331309627, + "grad_norm": 1.8841675519943237, + "learning_rate": 1.4312345858116932e-05, + "loss": 0.033, + "step": 7941 + }, + { + "epoch": 0.8610147441457069, + "grad_norm": 2.567692518234253, + "learning_rate": 1.4311620484549544e-05, + "loss": 0.0132, + "step": 7942 + }, + { + "epoch": 0.8611231569817867, + "grad_norm": 2.047226905822754, + "learning_rate": 1.4310895110982157e-05, + "loss": 0.0326, + "step": 7943 + }, + { + "epoch": 0.8612315698178664, + "grad_norm": 1.3074296712875366, + "learning_rate": 1.431016973741477e-05, + "loss": 0.0509, + "step": 7944 + }, + { + "epoch": 0.8613399826539462, + "grad_norm": 2.216212511062622, + "learning_rate": 1.4309444363847383e-05, + "loss": 0.0416, + "step": 7945 + }, + { + "epoch": 0.861448395490026, + "grad_norm": 0.2364555448293686, + "learning_rate": 1.4308718990279995e-05, + "loss": 0.0032, + "step": 7946 + }, + { + "epoch": 0.8615568083261058, + "grad_norm": 6.4078755378723145, + "learning_rate": 1.4307993616712607e-05, + "loss": 0.0414, + "step": 7947 + }, + { + "epoch": 0.8616652211621856, + "grad_norm": 2.1328303813934326, + "learning_rate": 1.430726824314522e-05, + "loss": 0.0211, + "step": 7948 + }, + { + "epoch": 0.8617736339982653, + "grad_norm": 6.467708110809326, + "learning_rate": 1.4306542869577833e-05, + "loss": 0.069, + "step": 7949 + }, + { + "epoch": 0.8618820468343452, + "grad_norm": 1.336592435836792, + "learning_rate": 1.4305817496010445e-05, + "loss": 0.0219, + "step": 7950 + }, + { + "epoch": 0.861990459670425, + "grad_norm": 4.886775970458984, + "learning_rate": 1.4305092122443059e-05, + "loss": 0.0593, + "step": 7951 + }, + { + "epoch": 0.8620988725065047, + "grad_norm": 4.479347229003906, + "learning_rate": 1.4304366748875673e-05, + "loss": 0.022, + "step": 7952 + }, + { + "epoch": 0.8622072853425846, + "grad_norm": 1.3871519565582275, + "learning_rate": 1.4303641375308287e-05, + "loss": 0.0132, + "step": 7953 + }, + { + "epoch": 0.8623156981786644, + "grad_norm": 3.7969322204589844, + "learning_rate": 1.4302916001740899e-05, + "loss": 0.0373, + "step": 7954 + }, + { + "epoch": 0.8624241110147441, + "grad_norm": 2.7746410369873047, + "learning_rate": 1.430219062817351e-05, + "loss": 0.0305, + "step": 7955 + }, + { + "epoch": 0.8625325238508239, + "grad_norm": 1.81478750705719, + "learning_rate": 1.4301465254606125e-05, + "loss": 0.0503, + "step": 7956 + }, + { + "epoch": 0.8626409366869038, + "grad_norm": 0.9114680886268616, + "learning_rate": 1.4300739881038737e-05, + "loss": 0.0073, + "step": 7957 + }, + { + "epoch": 0.8627493495229835, + "grad_norm": 4.426782608032227, + "learning_rate": 1.4300014507471349e-05, + "loss": 0.0484, + "step": 7958 + }, + { + "epoch": 0.8628577623590633, + "grad_norm": 1.475464940071106, + "learning_rate": 1.4299289133903961e-05, + "loss": 0.0383, + "step": 7959 + }, + { + "epoch": 0.8629661751951431, + "grad_norm": 6.71256160736084, + "learning_rate": 1.4298563760336575e-05, + "loss": 0.0546, + "step": 7960 + }, + { + "epoch": 0.863074588031223, + "grad_norm": 2.6416752338409424, + "learning_rate": 1.4297838386769187e-05, + "loss": 0.112, + "step": 7961 + }, + { + "epoch": 0.8631830008673027, + "grad_norm": 4.468672275543213, + "learning_rate": 1.42971130132018e-05, + "loss": 0.0546, + "step": 7962 + }, + { + "epoch": 0.8632914137033825, + "grad_norm": 1.649457335472107, + "learning_rate": 1.4296387639634413e-05, + "loss": 0.0317, + "step": 7963 + }, + { + "epoch": 0.8633998265394622, + "grad_norm": 5.474328517913818, + "learning_rate": 1.4295662266067025e-05, + "loss": 0.0703, + "step": 7964 + }, + { + "epoch": 0.8635082393755421, + "grad_norm": 2.8965513706207275, + "learning_rate": 1.4294936892499637e-05, + "loss": 0.0453, + "step": 7965 + }, + { + "epoch": 0.8636166522116219, + "grad_norm": 1.8866665363311768, + "learning_rate": 1.4294211518932251e-05, + "loss": 0.0218, + "step": 7966 + }, + { + "epoch": 0.8637250650477016, + "grad_norm": 2.7889275550842285, + "learning_rate": 1.4293486145364863e-05, + "loss": 0.035, + "step": 7967 + }, + { + "epoch": 0.8638334778837814, + "grad_norm": 4.094773769378662, + "learning_rate": 1.4292760771797475e-05, + "loss": 0.0461, + "step": 7968 + }, + { + "epoch": 0.8639418907198613, + "grad_norm": 1.426754117012024, + "learning_rate": 1.429203539823009e-05, + "loss": 0.011, + "step": 7969 + }, + { + "epoch": 0.864050303555941, + "grad_norm": 2.0648205280303955, + "learning_rate": 1.4291310024662703e-05, + "loss": 0.0537, + "step": 7970 + }, + { + "epoch": 0.8641587163920208, + "grad_norm": 4.195618629455566, + "learning_rate": 1.4290584651095315e-05, + "loss": 0.0547, + "step": 7971 + }, + { + "epoch": 0.8642671292281006, + "grad_norm": 1.8323825597763062, + "learning_rate": 1.4289859277527929e-05, + "loss": 0.0112, + "step": 7972 + }, + { + "epoch": 0.8643755420641804, + "grad_norm": 1.1312495470046997, + "learning_rate": 1.4289133903960541e-05, + "loss": 0.0108, + "step": 7973 + }, + { + "epoch": 0.8644839549002602, + "grad_norm": 4.610406398773193, + "learning_rate": 1.4288408530393153e-05, + "loss": 0.0481, + "step": 7974 + }, + { + "epoch": 0.86459236773634, + "grad_norm": 1.8559489250183105, + "learning_rate": 1.4287683156825767e-05, + "loss": 0.0378, + "step": 7975 + }, + { + "epoch": 0.8647007805724197, + "grad_norm": 5.073514938354492, + "learning_rate": 1.428695778325838e-05, + "loss": 0.0445, + "step": 7976 + }, + { + "epoch": 0.8648091934084996, + "grad_norm": 6.153916358947754, + "learning_rate": 1.4286232409690992e-05, + "loss": 0.0446, + "step": 7977 + }, + { + "epoch": 0.8649176062445794, + "grad_norm": 0.7037168145179749, + "learning_rate": 1.4285507036123605e-05, + "loss": 0.0093, + "step": 7978 + }, + { + "epoch": 0.8650260190806591, + "grad_norm": 0.7595009803771973, + "learning_rate": 1.4284781662556218e-05, + "loss": 0.0126, + "step": 7979 + }, + { + "epoch": 0.8651344319167389, + "grad_norm": 0.8300931453704834, + "learning_rate": 1.428405628898883e-05, + "loss": 0.0045, + "step": 7980 + }, + { + "epoch": 0.8652428447528188, + "grad_norm": 1.3292790651321411, + "learning_rate": 1.4283330915421443e-05, + "loss": 0.0099, + "step": 7981 + }, + { + "epoch": 0.8653512575888985, + "grad_norm": 7.49363899230957, + "learning_rate": 1.4282605541854056e-05, + "loss": 0.0337, + "step": 7982 + }, + { + "epoch": 0.8654596704249783, + "grad_norm": 4.0521345138549805, + "learning_rate": 1.4281880168286668e-05, + "loss": 0.0544, + "step": 7983 + }, + { + "epoch": 0.8655680832610582, + "grad_norm": 2.8904573917388916, + "learning_rate": 1.4281154794719282e-05, + "loss": 0.0242, + "step": 7984 + }, + { + "epoch": 0.8656764960971379, + "grad_norm": 2.3111212253570557, + "learning_rate": 1.4280429421151894e-05, + "loss": 0.0332, + "step": 7985 + }, + { + "epoch": 0.8657849089332177, + "grad_norm": 6.046841144561768, + "learning_rate": 1.4279704047584506e-05, + "loss": 0.0453, + "step": 7986 + }, + { + "epoch": 0.8658933217692975, + "grad_norm": 4.434520721435547, + "learning_rate": 1.4278978674017121e-05, + "loss": 0.0398, + "step": 7987 + }, + { + "epoch": 0.8660017346053773, + "grad_norm": 5.272397994995117, + "learning_rate": 1.4278253300449734e-05, + "loss": 0.0375, + "step": 7988 + }, + { + "epoch": 0.8661101474414571, + "grad_norm": 3.17649507522583, + "learning_rate": 1.4277527926882346e-05, + "loss": 0.0181, + "step": 7989 + }, + { + "epoch": 0.8662185602775369, + "grad_norm": 1.9770466089248657, + "learning_rate": 1.427680255331496e-05, + "loss": 0.0298, + "step": 7990 + }, + { + "epoch": 0.8663269731136166, + "grad_norm": 4.37705659866333, + "learning_rate": 1.4276077179747572e-05, + "loss": 0.0865, + "step": 7991 + }, + { + "epoch": 0.8664353859496965, + "grad_norm": 2.7172749042510986, + "learning_rate": 1.4275351806180184e-05, + "loss": 0.0428, + "step": 7992 + }, + { + "epoch": 0.8665437987857763, + "grad_norm": 2.8756325244903564, + "learning_rate": 1.4274626432612798e-05, + "loss": 0.0407, + "step": 7993 + }, + { + "epoch": 0.866652211621856, + "grad_norm": 1.973703145980835, + "learning_rate": 1.427390105904541e-05, + "loss": 0.0268, + "step": 7994 + }, + { + "epoch": 0.8667606244579358, + "grad_norm": 7.730795383453369, + "learning_rate": 1.4273175685478022e-05, + "loss": 0.0402, + "step": 7995 + }, + { + "epoch": 0.8668690372940157, + "grad_norm": 3.4862921237945557, + "learning_rate": 1.4272450311910636e-05, + "loss": 0.0339, + "step": 7996 + }, + { + "epoch": 0.8669774501300954, + "grad_norm": 1.2074073553085327, + "learning_rate": 1.4271724938343248e-05, + "loss": 0.0235, + "step": 7997 + }, + { + "epoch": 0.8670858629661752, + "grad_norm": 3.9618494510650635, + "learning_rate": 1.427099956477586e-05, + "loss": 0.044, + "step": 7998 + }, + { + "epoch": 0.8671942758022549, + "grad_norm": 1.8907667398452759, + "learning_rate": 1.4270274191208472e-05, + "loss": 0.0182, + "step": 7999 + }, + { + "epoch": 0.8673026886383348, + "grad_norm": 6.01931619644165, + "learning_rate": 1.4269548817641086e-05, + "loss": 0.0561, + "step": 8000 + }, + { + "epoch": 0.8674111014744146, + "grad_norm": 1.4781064987182617, + "learning_rate": 1.4268823444073698e-05, + "loss": 0.0126, + "step": 8001 + }, + { + "epoch": 0.8675195143104943, + "grad_norm": 4.126604080200195, + "learning_rate": 1.426809807050631e-05, + "loss": 0.0476, + "step": 8002 + }, + { + "epoch": 0.8676279271465741, + "grad_norm": 1.4719938039779663, + "learning_rate": 1.4267372696938924e-05, + "loss": 0.0285, + "step": 8003 + }, + { + "epoch": 0.867736339982654, + "grad_norm": 4.364186763763428, + "learning_rate": 1.4266647323371538e-05, + "loss": 0.0861, + "step": 8004 + }, + { + "epoch": 0.8678447528187337, + "grad_norm": 6.41081428527832, + "learning_rate": 1.4265921949804152e-05, + "loss": 0.036, + "step": 8005 + }, + { + "epoch": 0.8679531656548135, + "grad_norm": 7.8169264793396, + "learning_rate": 1.4265196576236764e-05, + "loss": 0.0413, + "step": 8006 + }, + { + "epoch": 0.8680615784908933, + "grad_norm": 3.0241174697875977, + "learning_rate": 1.4264471202669376e-05, + "loss": 0.0347, + "step": 8007 + }, + { + "epoch": 0.8681699913269731, + "grad_norm": 1.7187522649765015, + "learning_rate": 1.426374582910199e-05, + "loss": 0.0579, + "step": 8008 + }, + { + "epoch": 0.8682784041630529, + "grad_norm": 0.5590873956680298, + "learning_rate": 1.4263020455534602e-05, + "loss": 0.0057, + "step": 8009 + }, + { + "epoch": 0.8683868169991327, + "grad_norm": 0.5295031070709229, + "learning_rate": 1.4262295081967214e-05, + "loss": 0.0089, + "step": 8010 + }, + { + "epoch": 0.8684952298352124, + "grad_norm": 2.9369022846221924, + "learning_rate": 1.4261569708399826e-05, + "loss": 0.069, + "step": 8011 + }, + { + "epoch": 0.8686036426712923, + "grad_norm": 1.9200690984725952, + "learning_rate": 1.426084433483244e-05, + "loss": 0.053, + "step": 8012 + }, + { + "epoch": 0.8687120555073721, + "grad_norm": 3.173733711242676, + "learning_rate": 1.4260118961265052e-05, + "loss": 0.039, + "step": 8013 + }, + { + "epoch": 0.8688204683434518, + "grad_norm": 1.3937958478927612, + "learning_rate": 1.4259393587697665e-05, + "loss": 0.0267, + "step": 8014 + }, + { + "epoch": 0.8689288811795317, + "grad_norm": 0.8004685640335083, + "learning_rate": 1.4258668214130278e-05, + "loss": 0.0095, + "step": 8015 + }, + { + "epoch": 0.8690372940156115, + "grad_norm": 1.726497769355774, + "learning_rate": 1.425794284056289e-05, + "loss": 0.0104, + "step": 8016 + }, + { + "epoch": 0.8691457068516912, + "grad_norm": 1.454379916191101, + "learning_rate": 1.4257217466995503e-05, + "loss": 0.0352, + "step": 8017 + }, + { + "epoch": 0.869254119687771, + "grad_norm": 2.4096763134002686, + "learning_rate": 1.4256492093428117e-05, + "loss": 0.0328, + "step": 8018 + }, + { + "epoch": 0.8693625325238509, + "grad_norm": 4.362844467163086, + "learning_rate": 1.4255766719860729e-05, + "loss": 0.0289, + "step": 8019 + }, + { + "epoch": 0.8694709453599306, + "grad_norm": 4.002511024475098, + "learning_rate": 1.425504134629334e-05, + "loss": 0.0295, + "step": 8020 + }, + { + "epoch": 0.8695793581960104, + "grad_norm": 3.798621892929077, + "learning_rate": 1.4254315972725956e-05, + "loss": 0.0286, + "step": 8021 + }, + { + "epoch": 0.8696877710320902, + "grad_norm": 1.831995964050293, + "learning_rate": 1.4253590599158568e-05, + "loss": 0.0485, + "step": 8022 + }, + { + "epoch": 0.86979618386817, + "grad_norm": 2.0084691047668457, + "learning_rate": 1.4252865225591182e-05, + "loss": 0.0207, + "step": 8023 + }, + { + "epoch": 0.8699045967042498, + "grad_norm": 2.2252769470214844, + "learning_rate": 1.4252139852023794e-05, + "loss": 0.029, + "step": 8024 + }, + { + "epoch": 0.8700130095403296, + "grad_norm": 3.4625980854034424, + "learning_rate": 1.4251414478456407e-05, + "loss": 0.0263, + "step": 8025 + }, + { + "epoch": 0.8701214223764093, + "grad_norm": 5.8266425132751465, + "learning_rate": 1.4250689104889019e-05, + "loss": 0.0691, + "step": 8026 + }, + { + "epoch": 0.8702298352124892, + "grad_norm": 2.0979743003845215, + "learning_rate": 1.4249963731321633e-05, + "loss": 0.0519, + "step": 8027 + }, + { + "epoch": 0.870338248048569, + "grad_norm": 1.3338253498077393, + "learning_rate": 1.4249238357754245e-05, + "loss": 0.0119, + "step": 8028 + }, + { + "epoch": 0.8704466608846487, + "grad_norm": 2.486506223678589, + "learning_rate": 1.4248512984186857e-05, + "loss": 0.0397, + "step": 8029 + }, + { + "epoch": 0.8705550737207285, + "grad_norm": 5.641447067260742, + "learning_rate": 1.424778761061947e-05, + "loss": 0.0444, + "step": 8030 + }, + { + "epoch": 0.8706634865568084, + "grad_norm": 5.563135623931885, + "learning_rate": 1.4247062237052083e-05, + "loss": 0.0326, + "step": 8031 + }, + { + "epoch": 0.8707718993928881, + "grad_norm": 0.914780855178833, + "learning_rate": 1.4246336863484695e-05, + "loss": 0.0083, + "step": 8032 + }, + { + "epoch": 0.8708803122289679, + "grad_norm": 5.336762428283691, + "learning_rate": 1.4245611489917309e-05, + "loss": 0.0192, + "step": 8033 + }, + { + "epoch": 0.8709887250650477, + "grad_norm": 0.9229397773742676, + "learning_rate": 1.4244886116349921e-05, + "loss": 0.0107, + "step": 8034 + }, + { + "epoch": 0.8710971379011275, + "grad_norm": 4.800426959991455, + "learning_rate": 1.4244160742782533e-05, + "loss": 0.0209, + "step": 8035 + }, + { + "epoch": 0.8712055507372073, + "grad_norm": 11.083303451538086, + "learning_rate": 1.4243435369215147e-05, + "loss": 0.0965, + "step": 8036 + }, + { + "epoch": 0.871313963573287, + "grad_norm": 5.42661714553833, + "learning_rate": 1.4242709995647759e-05, + "loss": 0.0281, + "step": 8037 + }, + { + "epoch": 0.8714223764093668, + "grad_norm": 4.321871280670166, + "learning_rate": 1.4241984622080371e-05, + "loss": 0.069, + "step": 8038 + }, + { + "epoch": 0.8715307892454467, + "grad_norm": 1.9252147674560547, + "learning_rate": 1.4241259248512987e-05, + "loss": 0.0277, + "step": 8039 + }, + { + "epoch": 0.8716392020815265, + "grad_norm": 2.663621187210083, + "learning_rate": 1.4240533874945599e-05, + "loss": 0.0087, + "step": 8040 + }, + { + "epoch": 0.8717476149176062, + "grad_norm": 1.3736023902893066, + "learning_rate": 1.4239808501378211e-05, + "loss": 0.0109, + "step": 8041 + }, + { + "epoch": 0.8718560277536861, + "grad_norm": 3.192681312561035, + "learning_rate": 1.4239083127810825e-05, + "loss": 0.0273, + "step": 8042 + }, + { + "epoch": 0.8719644405897659, + "grad_norm": 4.5956711769104, + "learning_rate": 1.4238357754243437e-05, + "loss": 0.0354, + "step": 8043 + }, + { + "epoch": 0.8720728534258456, + "grad_norm": 0.19816190004348755, + "learning_rate": 1.423763238067605e-05, + "loss": 0.0039, + "step": 8044 + }, + { + "epoch": 0.8721812662619254, + "grad_norm": 3.715514659881592, + "learning_rate": 1.4236907007108663e-05, + "loss": 0.0583, + "step": 8045 + }, + { + "epoch": 0.8722896790980053, + "grad_norm": 5.1728034019470215, + "learning_rate": 1.4236181633541275e-05, + "loss": 0.0493, + "step": 8046 + }, + { + "epoch": 0.872398091934085, + "grad_norm": 3.220126152038574, + "learning_rate": 1.4235456259973887e-05, + "loss": 0.0344, + "step": 8047 + }, + { + "epoch": 0.8725065047701648, + "grad_norm": 2.192103385925293, + "learning_rate": 1.4234730886406501e-05, + "loss": 0.0261, + "step": 8048 + }, + { + "epoch": 0.8726149176062445, + "grad_norm": 5.530479907989502, + "learning_rate": 1.4234005512839113e-05, + "loss": 0.0956, + "step": 8049 + }, + { + "epoch": 0.8727233304423244, + "grad_norm": 4.064207077026367, + "learning_rate": 1.4233280139271725e-05, + "loss": 0.0398, + "step": 8050 + }, + { + "epoch": 0.8728317432784042, + "grad_norm": 4.3106889724731445, + "learning_rate": 1.4232554765704338e-05, + "loss": 0.0206, + "step": 8051 + }, + { + "epoch": 0.872940156114484, + "grad_norm": 5.409969329833984, + "learning_rate": 1.4231829392136951e-05, + "loss": 0.048, + "step": 8052 + }, + { + "epoch": 0.8730485689505637, + "grad_norm": 0.729382336139679, + "learning_rate": 1.4231104018569564e-05, + "loss": 0.0148, + "step": 8053 + }, + { + "epoch": 0.8731569817866436, + "grad_norm": 3.5952699184417725, + "learning_rate": 1.4230378645002176e-05, + "loss": 0.07, + "step": 8054 + }, + { + "epoch": 0.8732653946227233, + "grad_norm": 6.0624895095825195, + "learning_rate": 1.422965327143479e-05, + "loss": 0.0312, + "step": 8055 + }, + { + "epoch": 0.8733738074588031, + "grad_norm": 2.4765589237213135, + "learning_rate": 1.4228927897867403e-05, + "loss": 0.0476, + "step": 8056 + }, + { + "epoch": 0.8734822202948829, + "grad_norm": 1.8202593326568604, + "learning_rate": 1.4228202524300017e-05, + "loss": 0.0145, + "step": 8057 + }, + { + "epoch": 0.8735906331309627, + "grad_norm": 8.579437255859375, + "learning_rate": 1.422747715073263e-05, + "loss": 0.0266, + "step": 8058 + }, + { + "epoch": 0.8736990459670425, + "grad_norm": 3.6471474170684814, + "learning_rate": 1.4226751777165241e-05, + "loss": 0.0338, + "step": 8059 + }, + { + "epoch": 0.8738074588031223, + "grad_norm": 5.379640102386475, + "learning_rate": 1.4226026403597855e-05, + "loss": 0.0481, + "step": 8060 + }, + { + "epoch": 0.873915871639202, + "grad_norm": 7.652167320251465, + "learning_rate": 1.4225301030030467e-05, + "loss": 0.0492, + "step": 8061 + }, + { + "epoch": 0.8740242844752819, + "grad_norm": 0.7928423285484314, + "learning_rate": 1.422457565646308e-05, + "loss": 0.0133, + "step": 8062 + }, + { + "epoch": 0.8741326973113617, + "grad_norm": 3.019301176071167, + "learning_rate": 1.4223850282895692e-05, + "loss": 0.0392, + "step": 8063 + }, + { + "epoch": 0.8742411101474414, + "grad_norm": 2.615589141845703, + "learning_rate": 1.4223124909328306e-05, + "loss": 0.0273, + "step": 8064 + }, + { + "epoch": 0.8743495229835212, + "grad_norm": 1.0863593816757202, + "learning_rate": 1.4222399535760918e-05, + "loss": 0.0136, + "step": 8065 + }, + { + "epoch": 0.8744579358196011, + "grad_norm": 1.9341484308242798, + "learning_rate": 1.422167416219353e-05, + "loss": 0.0206, + "step": 8066 + }, + { + "epoch": 0.8745663486556808, + "grad_norm": 5.123457431793213, + "learning_rate": 1.4220948788626144e-05, + "loss": 0.0528, + "step": 8067 + }, + { + "epoch": 0.8746747614917606, + "grad_norm": 1.7236984968185425, + "learning_rate": 1.4220223415058756e-05, + "loss": 0.0122, + "step": 8068 + }, + { + "epoch": 0.8747831743278404, + "grad_norm": 4.51174259185791, + "learning_rate": 1.4219498041491368e-05, + "loss": 0.0456, + "step": 8069 + }, + { + "epoch": 0.8748915871639202, + "grad_norm": 1.562206745147705, + "learning_rate": 1.4218772667923982e-05, + "loss": 0.0156, + "step": 8070 + }, + { + "epoch": 0.875, + "grad_norm": 2.314427614212036, + "learning_rate": 1.4218047294356594e-05, + "loss": 0.0313, + "step": 8071 + }, + { + "epoch": 0.8751084128360798, + "grad_norm": 3.936433792114258, + "learning_rate": 1.4217321920789206e-05, + "loss": 0.0328, + "step": 8072 + }, + { + "epoch": 0.8752168256721596, + "grad_norm": 2.0195021629333496, + "learning_rate": 1.4216596547221822e-05, + "loss": 0.0398, + "step": 8073 + }, + { + "epoch": 0.8753252385082394, + "grad_norm": 4.12320613861084, + "learning_rate": 1.4215871173654434e-05, + "loss": 0.0188, + "step": 8074 + }, + { + "epoch": 0.8754336513443192, + "grad_norm": 3.0054001808166504, + "learning_rate": 1.4215145800087048e-05, + "loss": 0.0443, + "step": 8075 + }, + { + "epoch": 0.8755420641803989, + "grad_norm": 6.599697589874268, + "learning_rate": 1.421442042651966e-05, + "loss": 0.0962, + "step": 8076 + }, + { + "epoch": 0.8756504770164788, + "grad_norm": 2.3718421459198, + "learning_rate": 1.4213695052952272e-05, + "loss": 0.0328, + "step": 8077 + }, + { + "epoch": 0.8757588898525586, + "grad_norm": 1.9595767259597778, + "learning_rate": 1.4212969679384884e-05, + "loss": 0.0187, + "step": 8078 + }, + { + "epoch": 0.8758673026886383, + "grad_norm": 1.090933918952942, + "learning_rate": 1.4212244305817498e-05, + "loss": 0.0412, + "step": 8079 + }, + { + "epoch": 0.8759757155247181, + "grad_norm": 1.7367082834243774, + "learning_rate": 1.421151893225011e-05, + "loss": 0.0359, + "step": 8080 + }, + { + "epoch": 0.876084128360798, + "grad_norm": 0.815008819103241, + "learning_rate": 1.4210793558682722e-05, + "loss": 0.0177, + "step": 8081 + }, + { + "epoch": 0.8761925411968777, + "grad_norm": 3.6490068435668945, + "learning_rate": 1.4210068185115336e-05, + "loss": 0.048, + "step": 8082 + }, + { + "epoch": 0.8763009540329575, + "grad_norm": 3.1223134994506836, + "learning_rate": 1.4209342811547948e-05, + "loss": 0.0723, + "step": 8083 + }, + { + "epoch": 0.8764093668690373, + "grad_norm": 3.181519031524658, + "learning_rate": 1.420861743798056e-05, + "loss": 0.0727, + "step": 8084 + }, + { + "epoch": 0.8765177797051171, + "grad_norm": 2.6436495780944824, + "learning_rate": 1.4207892064413174e-05, + "loss": 0.0415, + "step": 8085 + }, + { + "epoch": 0.8766261925411969, + "grad_norm": 1.5125778913497925, + "learning_rate": 1.4207166690845786e-05, + "loss": 0.0133, + "step": 8086 + }, + { + "epoch": 0.8767346053772767, + "grad_norm": 2.2957100868225098, + "learning_rate": 1.4206441317278398e-05, + "loss": 0.0329, + "step": 8087 + }, + { + "epoch": 0.8768430182133564, + "grad_norm": 2.7202341556549072, + "learning_rate": 1.4205715943711012e-05, + "loss": 0.0212, + "step": 8088 + }, + { + "epoch": 0.8769514310494363, + "grad_norm": 6.10230016708374, + "learning_rate": 1.4204990570143624e-05, + "loss": 0.0237, + "step": 8089 + }, + { + "epoch": 0.877059843885516, + "grad_norm": 0.8007824420928955, + "learning_rate": 1.4204265196576237e-05, + "loss": 0.0061, + "step": 8090 + }, + { + "epoch": 0.8771682567215958, + "grad_norm": 0.9383057355880737, + "learning_rate": 1.4203539823008852e-05, + "loss": 0.0099, + "step": 8091 + }, + { + "epoch": 0.8772766695576756, + "grad_norm": 2.4465301036834717, + "learning_rate": 1.4202814449441464e-05, + "loss": 0.0418, + "step": 8092 + }, + { + "epoch": 0.8773850823937555, + "grad_norm": 9.287232398986816, + "learning_rate": 1.4202089075874076e-05, + "loss": 0.0643, + "step": 8093 + }, + { + "epoch": 0.8774934952298352, + "grad_norm": 5.525873184204102, + "learning_rate": 1.420136370230669e-05, + "loss": 0.0159, + "step": 8094 + }, + { + "epoch": 0.877601908065915, + "grad_norm": 2.5937018394470215, + "learning_rate": 1.4200638328739302e-05, + "loss": 0.0333, + "step": 8095 + }, + { + "epoch": 0.8777103209019947, + "grad_norm": 2.1100547313690186, + "learning_rate": 1.4199912955171915e-05, + "loss": 0.0375, + "step": 8096 + }, + { + "epoch": 0.8778187337380746, + "grad_norm": 5.172200679779053, + "learning_rate": 1.4199187581604528e-05, + "loss": 0.05, + "step": 8097 + }, + { + "epoch": 0.8779271465741544, + "grad_norm": 2.057159423828125, + "learning_rate": 1.419846220803714e-05, + "loss": 0.0381, + "step": 8098 + }, + { + "epoch": 0.8780355594102341, + "grad_norm": 3.6056032180786133, + "learning_rate": 1.4197736834469753e-05, + "loss": 0.0216, + "step": 8099 + }, + { + "epoch": 0.8781439722463139, + "grad_norm": 4.345184326171875, + "learning_rate": 1.4197011460902366e-05, + "loss": 0.0499, + "step": 8100 + }, + { + "epoch": 0.8782523850823938, + "grad_norm": 3.0804624557495117, + "learning_rate": 1.4196286087334979e-05, + "loss": 0.0308, + "step": 8101 + }, + { + "epoch": 0.8783607979184735, + "grad_norm": 0.3881496489048004, + "learning_rate": 1.419556071376759e-05, + "loss": 0.0043, + "step": 8102 + }, + { + "epoch": 0.8784692107545533, + "grad_norm": 0.7614729404449463, + "learning_rate": 1.4194835340200203e-05, + "loss": 0.0353, + "step": 8103 + }, + { + "epoch": 0.8785776235906332, + "grad_norm": 1.5191359519958496, + "learning_rate": 1.4194109966632817e-05, + "loss": 0.0175, + "step": 8104 + }, + { + "epoch": 0.878686036426713, + "grad_norm": 3.7749412059783936, + "learning_rate": 1.4193384593065429e-05, + "loss": 0.0536, + "step": 8105 + }, + { + "epoch": 0.8787944492627927, + "grad_norm": 2.7214560508728027, + "learning_rate": 1.4192659219498041e-05, + "loss": 0.0303, + "step": 8106 + }, + { + "epoch": 0.8789028620988725, + "grad_norm": 1.6478720903396606, + "learning_rate": 1.4191933845930655e-05, + "loss": 0.0063, + "step": 8107 + }, + { + "epoch": 0.8790112749349523, + "grad_norm": 1.9963879585266113, + "learning_rate": 1.4191208472363269e-05, + "loss": 0.0122, + "step": 8108 + }, + { + "epoch": 0.8791196877710321, + "grad_norm": 2.8423500061035156, + "learning_rate": 1.4190483098795883e-05, + "loss": 0.0459, + "step": 8109 + }, + { + "epoch": 0.8792281006071119, + "grad_norm": 2.4887118339538574, + "learning_rate": 1.4189757725228495e-05, + "loss": 0.0397, + "step": 8110 + }, + { + "epoch": 0.8793365134431916, + "grad_norm": 1.7338216304779053, + "learning_rate": 1.4189032351661107e-05, + "loss": 0.0253, + "step": 8111 + }, + { + "epoch": 0.8794449262792715, + "grad_norm": 2.0312564373016357, + "learning_rate": 1.418830697809372e-05, + "loss": 0.0167, + "step": 8112 + }, + { + "epoch": 0.8795533391153513, + "grad_norm": 2.309907913208008, + "learning_rate": 1.4187581604526333e-05, + "loss": 0.0163, + "step": 8113 + }, + { + "epoch": 0.879661751951431, + "grad_norm": 2.6956863403320312, + "learning_rate": 1.4186856230958945e-05, + "loss": 0.0392, + "step": 8114 + }, + { + "epoch": 0.8797701647875108, + "grad_norm": 1.0241786241531372, + "learning_rate": 1.4186130857391557e-05, + "loss": 0.0316, + "step": 8115 + }, + { + "epoch": 0.8798785776235907, + "grad_norm": 1.7261027097702026, + "learning_rate": 1.4185405483824171e-05, + "loss": 0.0151, + "step": 8116 + }, + { + "epoch": 0.8799869904596704, + "grad_norm": 4.07289457321167, + "learning_rate": 1.4184680110256783e-05, + "loss": 0.0488, + "step": 8117 + }, + { + "epoch": 0.8800954032957502, + "grad_norm": 2.8484842777252197, + "learning_rate": 1.4183954736689395e-05, + "loss": 0.0605, + "step": 8118 + }, + { + "epoch": 0.88020381613183, + "grad_norm": 2.36439847946167, + "learning_rate": 1.4183229363122009e-05, + "loss": 0.0624, + "step": 8119 + }, + { + "epoch": 0.8803122289679098, + "grad_norm": 3.9905757904052734, + "learning_rate": 1.4182503989554621e-05, + "loss": 0.0816, + "step": 8120 + }, + { + "epoch": 0.8804206418039896, + "grad_norm": 0.7506633996963501, + "learning_rate": 1.4181778615987233e-05, + "loss": 0.0084, + "step": 8121 + }, + { + "epoch": 0.8805290546400694, + "grad_norm": 4.849565029144287, + "learning_rate": 1.4181053242419847e-05, + "loss": 0.0526, + "step": 8122 + }, + { + "epoch": 0.8806374674761491, + "grad_norm": 4.705129146575928, + "learning_rate": 1.418032786885246e-05, + "loss": 0.0946, + "step": 8123 + }, + { + "epoch": 0.880745880312229, + "grad_norm": 3.7280917167663574, + "learning_rate": 1.4179602495285071e-05, + "loss": 0.0408, + "step": 8124 + }, + { + "epoch": 0.8808542931483088, + "grad_norm": 2.348576545715332, + "learning_rate": 1.4178877121717687e-05, + "loss": 0.0294, + "step": 8125 + }, + { + "epoch": 0.8809627059843885, + "grad_norm": 2.6470015048980713, + "learning_rate": 1.4178151748150299e-05, + "loss": 0.0451, + "step": 8126 + }, + { + "epoch": 0.8810711188204683, + "grad_norm": 2.187103033065796, + "learning_rate": 1.4177426374582913e-05, + "loss": 0.0252, + "step": 8127 + }, + { + "epoch": 0.8811795316565482, + "grad_norm": 0.9906470775604248, + "learning_rate": 1.4176701001015525e-05, + "loss": 0.0112, + "step": 8128 + }, + { + "epoch": 0.8812879444926279, + "grad_norm": 9.511249542236328, + "learning_rate": 1.4175975627448137e-05, + "loss": 0.0601, + "step": 8129 + }, + { + "epoch": 0.8813963573287077, + "grad_norm": 1.7483736276626587, + "learning_rate": 1.417525025388075e-05, + "loss": 0.0202, + "step": 8130 + }, + { + "epoch": 0.8815047701647876, + "grad_norm": 3.9009652137756348, + "learning_rate": 1.4174524880313363e-05, + "loss": 0.0435, + "step": 8131 + }, + { + "epoch": 0.8816131830008673, + "grad_norm": 4.477999210357666, + "learning_rate": 1.4173799506745975e-05, + "loss": 0.0557, + "step": 8132 + }, + { + "epoch": 0.8817215958369471, + "grad_norm": 2.8291385173797607, + "learning_rate": 1.4173074133178588e-05, + "loss": 0.0454, + "step": 8133 + }, + { + "epoch": 0.8818300086730269, + "grad_norm": 3.304182529449463, + "learning_rate": 1.4172348759611201e-05, + "loss": 0.0704, + "step": 8134 + }, + { + "epoch": 0.8819384215091067, + "grad_norm": 2.9368510246276855, + "learning_rate": 1.4171623386043814e-05, + "loss": 0.0163, + "step": 8135 + }, + { + "epoch": 0.8820468343451865, + "grad_norm": 2.714245080947876, + "learning_rate": 1.4170898012476426e-05, + "loss": 0.0909, + "step": 8136 + }, + { + "epoch": 0.8821552471812663, + "grad_norm": 3.389958381652832, + "learning_rate": 1.417017263890904e-05, + "loss": 0.0268, + "step": 8137 + }, + { + "epoch": 0.882263660017346, + "grad_norm": 4.7069830894470215, + "learning_rate": 1.4169447265341652e-05, + "loss": 0.0444, + "step": 8138 + }, + { + "epoch": 0.8823720728534259, + "grad_norm": 2.516436815261841, + "learning_rate": 1.4168721891774264e-05, + "loss": 0.0311, + "step": 8139 + }, + { + "epoch": 0.8824804856895057, + "grad_norm": 1.0590834617614746, + "learning_rate": 1.4167996518206878e-05, + "loss": 0.0301, + "step": 8140 + }, + { + "epoch": 0.8825888985255854, + "grad_norm": 0.6395785212516785, + "learning_rate": 1.416727114463949e-05, + "loss": 0.008, + "step": 8141 + }, + { + "epoch": 0.8826973113616652, + "grad_norm": 3.4620423316955566, + "learning_rate": 1.4166545771072104e-05, + "loss": 0.0281, + "step": 8142 + }, + { + "epoch": 0.8828057241977451, + "grad_norm": 2.935607433319092, + "learning_rate": 1.4165820397504717e-05, + "loss": 0.0406, + "step": 8143 + }, + { + "epoch": 0.8829141370338248, + "grad_norm": 4.401345252990723, + "learning_rate": 1.416509502393733e-05, + "loss": 0.0212, + "step": 8144 + }, + { + "epoch": 0.8830225498699046, + "grad_norm": 0.5647295117378235, + "learning_rate": 1.4164369650369942e-05, + "loss": 0.0079, + "step": 8145 + }, + { + "epoch": 0.8831309627059843, + "grad_norm": 2.473067045211792, + "learning_rate": 1.4163644276802556e-05, + "loss": 0.0243, + "step": 8146 + }, + { + "epoch": 0.8832393755420642, + "grad_norm": 8.424570083618164, + "learning_rate": 1.4162918903235168e-05, + "loss": 0.0564, + "step": 8147 + }, + { + "epoch": 0.883347788378144, + "grad_norm": 2.0886871814727783, + "learning_rate": 1.416219352966778e-05, + "loss": 0.0405, + "step": 8148 + }, + { + "epoch": 0.8834562012142237, + "grad_norm": 2.5879416465759277, + "learning_rate": 1.4161468156100394e-05, + "loss": 0.0304, + "step": 8149 + }, + { + "epoch": 0.8835646140503035, + "grad_norm": 4.727646350860596, + "learning_rate": 1.4160742782533006e-05, + "loss": 0.1042, + "step": 8150 + }, + { + "epoch": 0.8836730268863834, + "grad_norm": 1.9059433937072754, + "learning_rate": 1.4160017408965618e-05, + "loss": 0.037, + "step": 8151 + }, + { + "epoch": 0.8837814397224631, + "grad_norm": 1.6435415744781494, + "learning_rate": 1.4159292035398232e-05, + "loss": 0.0369, + "step": 8152 + }, + { + "epoch": 0.8838898525585429, + "grad_norm": 1.580568790435791, + "learning_rate": 1.4158566661830844e-05, + "loss": 0.016, + "step": 8153 + }, + { + "epoch": 0.8839982653946227, + "grad_norm": 2.0795249938964844, + "learning_rate": 1.4157841288263456e-05, + "loss": 0.023, + "step": 8154 + }, + { + "epoch": 0.8841066782307025, + "grad_norm": 3.0455148220062256, + "learning_rate": 1.4157115914696068e-05, + "loss": 0.0447, + "step": 8155 + }, + { + "epoch": 0.8842150910667823, + "grad_norm": 0.7289374470710754, + "learning_rate": 1.4156390541128682e-05, + "loss": 0.0197, + "step": 8156 + }, + { + "epoch": 0.8843235039028621, + "grad_norm": 1.312934398651123, + "learning_rate": 1.4155665167561294e-05, + "loss": 0.0191, + "step": 8157 + }, + { + "epoch": 0.8844319167389418, + "grad_norm": 1.5560188293457031, + "learning_rate": 1.4154939793993906e-05, + "loss": 0.0174, + "step": 8158 + }, + { + "epoch": 0.8845403295750217, + "grad_norm": 7.621614456176758, + "learning_rate": 1.415421442042652e-05, + "loss": 0.0827, + "step": 8159 + }, + { + "epoch": 0.8846487424111015, + "grad_norm": 5.801401138305664, + "learning_rate": 1.4153489046859134e-05, + "loss": 0.0451, + "step": 8160 + }, + { + "epoch": 0.8847571552471812, + "grad_norm": 0.821002185344696, + "learning_rate": 1.4152763673291748e-05, + "loss": 0.0053, + "step": 8161 + }, + { + "epoch": 0.8848655680832611, + "grad_norm": 3.2111148834228516, + "learning_rate": 1.415203829972436e-05, + "loss": 0.0535, + "step": 8162 + }, + { + "epoch": 0.8849739809193409, + "grad_norm": 0.9641271233558655, + "learning_rate": 1.4151312926156972e-05, + "loss": 0.0139, + "step": 8163 + }, + { + "epoch": 0.8850823937554206, + "grad_norm": 0.7371121048927307, + "learning_rate": 1.4150587552589586e-05, + "loss": 0.0097, + "step": 8164 + }, + { + "epoch": 0.8851908065915004, + "grad_norm": 5.145103931427002, + "learning_rate": 1.4149862179022198e-05, + "loss": 0.0442, + "step": 8165 + }, + { + "epoch": 0.8852992194275803, + "grad_norm": 3.4555656909942627, + "learning_rate": 1.414913680545481e-05, + "loss": 0.0426, + "step": 8166 + }, + { + "epoch": 0.88540763226366, + "grad_norm": 2.2625460624694824, + "learning_rate": 1.4148411431887422e-05, + "loss": 0.0286, + "step": 8167 + }, + { + "epoch": 0.8855160450997398, + "grad_norm": 3.107111692428589, + "learning_rate": 1.4147686058320036e-05, + "loss": 0.0308, + "step": 8168 + }, + { + "epoch": 0.8856244579358196, + "grad_norm": 4.1133928298950195, + "learning_rate": 1.4146960684752648e-05, + "loss": 0.0153, + "step": 8169 + }, + { + "epoch": 0.8857328707718994, + "grad_norm": 3.633727788925171, + "learning_rate": 1.414623531118526e-05, + "loss": 0.0525, + "step": 8170 + }, + { + "epoch": 0.8858412836079792, + "grad_norm": 3.58784556388855, + "learning_rate": 1.4145509937617874e-05, + "loss": 0.0205, + "step": 8171 + }, + { + "epoch": 0.885949696444059, + "grad_norm": 0.861797571182251, + "learning_rate": 1.4144784564050487e-05, + "loss": 0.0346, + "step": 8172 + }, + { + "epoch": 0.8860581092801387, + "grad_norm": 7.225543022155762, + "learning_rate": 1.4144059190483099e-05, + "loss": 0.0514, + "step": 8173 + }, + { + "epoch": 0.8861665221162186, + "grad_norm": 5.029970645904541, + "learning_rate": 1.4143333816915713e-05, + "loss": 0.0474, + "step": 8174 + }, + { + "epoch": 0.8862749349522984, + "grad_norm": 4.14299201965332, + "learning_rate": 1.4142608443348325e-05, + "loss": 0.0875, + "step": 8175 + }, + { + "epoch": 0.8863833477883781, + "grad_norm": 1.314704418182373, + "learning_rate": 1.4141883069780937e-05, + "loss": 0.028, + "step": 8176 + }, + { + "epoch": 0.8864917606244579, + "grad_norm": 0.9844270944595337, + "learning_rate": 1.4141157696213552e-05, + "loss": 0.0237, + "step": 8177 + }, + { + "epoch": 0.8866001734605378, + "grad_norm": 3.147526264190674, + "learning_rate": 1.4140432322646164e-05, + "loss": 0.0362, + "step": 8178 + }, + { + "epoch": 0.8867085862966175, + "grad_norm": 3.85404109954834, + "learning_rate": 1.4139706949078778e-05, + "loss": 0.0282, + "step": 8179 + }, + { + "epoch": 0.8868169991326973, + "grad_norm": 3.5942940711975098, + "learning_rate": 1.413898157551139e-05, + "loss": 0.0364, + "step": 8180 + }, + { + "epoch": 0.886925411968777, + "grad_norm": 2.244324207305908, + "learning_rate": 1.4138256201944003e-05, + "loss": 0.0241, + "step": 8181 + }, + { + "epoch": 0.8870338248048569, + "grad_norm": 3.5700981616973877, + "learning_rate": 1.4137530828376615e-05, + "loss": 0.0614, + "step": 8182 + }, + { + "epoch": 0.8871422376409367, + "grad_norm": 8.395803451538086, + "learning_rate": 1.4136805454809229e-05, + "loss": 0.0689, + "step": 8183 + }, + { + "epoch": 0.8872506504770165, + "grad_norm": 3.419193744659424, + "learning_rate": 1.413608008124184e-05, + "loss": 0.0308, + "step": 8184 + }, + { + "epoch": 0.8873590633130962, + "grad_norm": 1.9990322589874268, + "learning_rate": 1.4135354707674453e-05, + "loss": 0.0238, + "step": 8185 + }, + { + "epoch": 0.8874674761491761, + "grad_norm": 5.15859317779541, + "learning_rate": 1.4134629334107067e-05, + "loss": 0.0228, + "step": 8186 + }, + { + "epoch": 0.8875758889852559, + "grad_norm": 4.388890743255615, + "learning_rate": 1.4133903960539679e-05, + "loss": 0.0417, + "step": 8187 + }, + { + "epoch": 0.8876843018213356, + "grad_norm": 2.184662342071533, + "learning_rate": 1.4133178586972291e-05, + "loss": 0.0226, + "step": 8188 + }, + { + "epoch": 0.8877927146574154, + "grad_norm": 7.697894096374512, + "learning_rate": 1.4132453213404905e-05, + "loss": 0.0347, + "step": 8189 + }, + { + "epoch": 0.8879011274934953, + "grad_norm": 1.4115116596221924, + "learning_rate": 1.4131727839837517e-05, + "loss": 0.0392, + "step": 8190 + }, + { + "epoch": 0.888009540329575, + "grad_norm": 4.8858795166015625, + "learning_rate": 1.4131002466270129e-05, + "loss": 0.0397, + "step": 8191 + }, + { + "epoch": 0.8881179531656548, + "grad_norm": 2.812225580215454, + "learning_rate": 1.4130277092702743e-05, + "loss": 0.0491, + "step": 8192 + }, + { + "epoch": 0.8882263660017347, + "grad_norm": 2.62455153465271, + "learning_rate": 1.4129551719135355e-05, + "loss": 0.0496, + "step": 8193 + }, + { + "epoch": 0.8883347788378144, + "grad_norm": 2.732896566390991, + "learning_rate": 1.4128826345567969e-05, + "loss": 0.0175, + "step": 8194 + }, + { + "epoch": 0.8884431916738942, + "grad_norm": 2.7297933101654053, + "learning_rate": 1.4128100972000583e-05, + "loss": 0.0256, + "step": 8195 + }, + { + "epoch": 0.888551604509974, + "grad_norm": 3.67106294631958, + "learning_rate": 1.4127375598433195e-05, + "loss": 0.0472, + "step": 8196 + }, + { + "epoch": 0.8886600173460538, + "grad_norm": 1.4374045133590698, + "learning_rate": 1.4126650224865807e-05, + "loss": 0.0227, + "step": 8197 + }, + { + "epoch": 0.8887684301821336, + "grad_norm": 4.918849945068359, + "learning_rate": 1.4125924851298421e-05, + "loss": 0.0184, + "step": 8198 + }, + { + "epoch": 0.8888768430182133, + "grad_norm": 1.114795207977295, + "learning_rate": 1.4125199477731033e-05, + "loss": 0.0165, + "step": 8199 + }, + { + "epoch": 0.8889852558542931, + "grad_norm": 2.612481117248535, + "learning_rate": 1.4124474104163645e-05, + "loss": 0.0401, + "step": 8200 + }, + { + "epoch": 0.889093668690373, + "grad_norm": 2.352902412414551, + "learning_rate": 1.4123748730596259e-05, + "loss": 0.0087, + "step": 8201 + }, + { + "epoch": 0.8892020815264527, + "grad_norm": 5.588131427764893, + "learning_rate": 1.4123023357028871e-05, + "loss": 0.0986, + "step": 8202 + }, + { + "epoch": 0.8893104943625325, + "grad_norm": 2.0635342597961426, + "learning_rate": 1.4122297983461483e-05, + "loss": 0.0257, + "step": 8203 + }, + { + "epoch": 0.8894189071986123, + "grad_norm": 2.4894328117370605, + "learning_rate": 1.4121572609894097e-05, + "loss": 0.0133, + "step": 8204 + }, + { + "epoch": 0.8895273200346921, + "grad_norm": 4.181028366088867, + "learning_rate": 1.412084723632671e-05, + "loss": 0.0819, + "step": 8205 + }, + { + "epoch": 0.8896357328707719, + "grad_norm": 6.74847936630249, + "learning_rate": 1.4120121862759321e-05, + "loss": 0.061, + "step": 8206 + }, + { + "epoch": 0.8897441457068517, + "grad_norm": 3.871793270111084, + "learning_rate": 1.4119396489191934e-05, + "loss": 0.0262, + "step": 8207 + }, + { + "epoch": 0.8898525585429314, + "grad_norm": 3.4960849285125732, + "learning_rate": 1.4118671115624547e-05, + "loss": 0.05, + "step": 8208 + }, + { + "epoch": 0.8899609713790113, + "grad_norm": 4.3872599601745605, + "learning_rate": 1.411794574205716e-05, + "loss": 0.1068, + "step": 8209 + }, + { + "epoch": 0.8900693842150911, + "grad_norm": 2.012056589126587, + "learning_rate": 1.4117220368489772e-05, + "loss": 0.0238, + "step": 8210 + }, + { + "epoch": 0.8901777970511708, + "grad_norm": 4.1055402755737305, + "learning_rate": 1.4116494994922387e-05, + "loss": 0.0369, + "step": 8211 + }, + { + "epoch": 0.8902862098872506, + "grad_norm": 2.899153709411621, + "learning_rate": 1.4115769621355e-05, + "loss": 0.0227, + "step": 8212 + }, + { + "epoch": 0.8903946227233305, + "grad_norm": 1.3254992961883545, + "learning_rate": 1.4115044247787613e-05, + "loss": 0.0154, + "step": 8213 + }, + { + "epoch": 0.8905030355594102, + "grad_norm": 7.4821648597717285, + "learning_rate": 1.4114318874220225e-05, + "loss": 0.0739, + "step": 8214 + }, + { + "epoch": 0.89061144839549, + "grad_norm": 1.5745426416397095, + "learning_rate": 1.4113593500652838e-05, + "loss": 0.0289, + "step": 8215 + }, + { + "epoch": 0.8907198612315698, + "grad_norm": 4.187367916107178, + "learning_rate": 1.4112868127085451e-05, + "loss": 0.1123, + "step": 8216 + }, + { + "epoch": 0.8908282740676496, + "grad_norm": 3.607309103012085, + "learning_rate": 1.4112142753518063e-05, + "loss": 0.0364, + "step": 8217 + }, + { + "epoch": 0.8909366869037294, + "grad_norm": 1.662791132926941, + "learning_rate": 1.4111417379950676e-05, + "loss": 0.0335, + "step": 8218 + }, + { + "epoch": 0.8910450997398092, + "grad_norm": 3.5446763038635254, + "learning_rate": 1.4110692006383288e-05, + "loss": 0.038, + "step": 8219 + }, + { + "epoch": 0.891153512575889, + "grad_norm": 2.875978469848633, + "learning_rate": 1.4109966632815902e-05, + "loss": 0.0504, + "step": 8220 + }, + { + "epoch": 0.8912619254119688, + "grad_norm": 6.809833526611328, + "learning_rate": 1.4109241259248514e-05, + "loss": 0.0471, + "step": 8221 + }, + { + "epoch": 0.8913703382480486, + "grad_norm": 2.5542514324188232, + "learning_rate": 1.4108515885681126e-05, + "loss": 0.0384, + "step": 8222 + }, + { + "epoch": 0.8914787510841283, + "grad_norm": 0.8577737212181091, + "learning_rate": 1.410779051211374e-05, + "loss": 0.0093, + "step": 8223 + }, + { + "epoch": 0.8915871639202082, + "grad_norm": 6.514819145202637, + "learning_rate": 1.4107065138546352e-05, + "loss": 0.0668, + "step": 8224 + }, + { + "epoch": 0.891695576756288, + "grad_norm": 1.5531800985336304, + "learning_rate": 1.4106339764978964e-05, + "loss": 0.0237, + "step": 8225 + }, + { + "epoch": 0.8918039895923677, + "grad_norm": 2.1796927452087402, + "learning_rate": 1.4105614391411578e-05, + "loss": 0.0198, + "step": 8226 + }, + { + "epoch": 0.8919124024284475, + "grad_norm": 4.340536117553711, + "learning_rate": 1.410488901784419e-05, + "loss": 0.0265, + "step": 8227 + }, + { + "epoch": 0.8920208152645274, + "grad_norm": 3.7541697025299072, + "learning_rate": 1.4104163644276802e-05, + "loss": 0.0096, + "step": 8228 + }, + { + "epoch": 0.8921292281006071, + "grad_norm": 3.217391014099121, + "learning_rate": 1.4103438270709418e-05, + "loss": 0.0591, + "step": 8229 + }, + { + "epoch": 0.8922376409366869, + "grad_norm": 3.6494638919830322, + "learning_rate": 1.410271289714203e-05, + "loss": 0.0541, + "step": 8230 + }, + { + "epoch": 0.8923460537727667, + "grad_norm": 5.286374092102051, + "learning_rate": 1.4101987523574644e-05, + "loss": 0.071, + "step": 8231 + }, + { + "epoch": 0.8924544666088465, + "grad_norm": 1.6361194849014282, + "learning_rate": 1.4101262150007256e-05, + "loss": 0.0384, + "step": 8232 + }, + { + "epoch": 0.8925628794449263, + "grad_norm": 3.9301917552948, + "learning_rate": 1.4100536776439868e-05, + "loss": 0.0851, + "step": 8233 + }, + { + "epoch": 0.8926712922810061, + "grad_norm": 2.2370147705078125, + "learning_rate": 1.409981140287248e-05, + "loss": 0.0468, + "step": 8234 + }, + { + "epoch": 0.8927797051170858, + "grad_norm": 3.241680145263672, + "learning_rate": 1.4099086029305094e-05, + "loss": 0.0536, + "step": 8235 + }, + { + "epoch": 0.8928881179531657, + "grad_norm": 2.064021348953247, + "learning_rate": 1.4098360655737706e-05, + "loss": 0.0316, + "step": 8236 + }, + { + "epoch": 0.8929965307892455, + "grad_norm": 4.439774036407471, + "learning_rate": 1.4097635282170318e-05, + "loss": 0.087, + "step": 8237 + }, + { + "epoch": 0.8931049436253252, + "grad_norm": 2.818328380584717, + "learning_rate": 1.4096909908602932e-05, + "loss": 0.0374, + "step": 8238 + }, + { + "epoch": 0.893213356461405, + "grad_norm": 3.725044012069702, + "learning_rate": 1.4096184535035544e-05, + "loss": 0.0588, + "step": 8239 + }, + { + "epoch": 0.8933217692974849, + "grad_norm": 0.6285514831542969, + "learning_rate": 1.4095459161468156e-05, + "loss": 0.0076, + "step": 8240 + }, + { + "epoch": 0.8934301821335646, + "grad_norm": 0.9239868521690369, + "learning_rate": 1.409473378790077e-05, + "loss": 0.0178, + "step": 8241 + }, + { + "epoch": 0.8935385949696444, + "grad_norm": 3.2385401725769043, + "learning_rate": 1.4094008414333382e-05, + "loss": 0.0306, + "step": 8242 + }, + { + "epoch": 0.8936470078057241, + "grad_norm": 4.167422294616699, + "learning_rate": 1.4093283040765994e-05, + "loss": 0.0432, + "step": 8243 + }, + { + "epoch": 0.893755420641804, + "grad_norm": 2.517362356185913, + "learning_rate": 1.4092557667198608e-05, + "loss": 0.0407, + "step": 8244 + }, + { + "epoch": 0.8938638334778838, + "grad_norm": 3.045186758041382, + "learning_rate": 1.409183229363122e-05, + "loss": 0.0282, + "step": 8245 + }, + { + "epoch": 0.8939722463139635, + "grad_norm": 4.564958095550537, + "learning_rate": 1.4091106920063834e-05, + "loss": 0.0542, + "step": 8246 + }, + { + "epoch": 0.8940806591500433, + "grad_norm": 6.25771951675415, + "learning_rate": 1.4090381546496448e-05, + "loss": 0.0163, + "step": 8247 + }, + { + "epoch": 0.8941890719861232, + "grad_norm": 6.117558002471924, + "learning_rate": 1.408965617292906e-05, + "loss": 0.0493, + "step": 8248 + }, + { + "epoch": 0.894297484822203, + "grad_norm": 2.7841148376464844, + "learning_rate": 1.4088930799361672e-05, + "loss": 0.0509, + "step": 8249 + }, + { + "epoch": 0.8944058976582827, + "grad_norm": 3.190276861190796, + "learning_rate": 1.4088205425794286e-05, + "loss": 0.0375, + "step": 8250 + }, + { + "epoch": 0.8945143104943626, + "grad_norm": 6.831657886505127, + "learning_rate": 1.4087480052226898e-05, + "loss": 0.0821, + "step": 8251 + }, + { + "epoch": 0.8946227233304423, + "grad_norm": 2.058582067489624, + "learning_rate": 1.408675467865951e-05, + "loss": 0.0211, + "step": 8252 + }, + { + "epoch": 0.8947311361665221, + "grad_norm": 1.2170650959014893, + "learning_rate": 1.4086029305092124e-05, + "loss": 0.0337, + "step": 8253 + }, + { + "epoch": 0.8948395490026019, + "grad_norm": 0.6889882683753967, + "learning_rate": 1.4085303931524737e-05, + "loss": 0.0127, + "step": 8254 + }, + { + "epoch": 0.8949479618386817, + "grad_norm": 1.995898962020874, + "learning_rate": 1.4084578557957349e-05, + "loss": 0.0353, + "step": 8255 + }, + { + "epoch": 0.8950563746747615, + "grad_norm": 3.116739511489868, + "learning_rate": 1.4083853184389962e-05, + "loss": 0.0407, + "step": 8256 + }, + { + "epoch": 0.8951647875108413, + "grad_norm": 2.7445900440216064, + "learning_rate": 1.4083127810822575e-05, + "loss": 0.0465, + "step": 8257 + }, + { + "epoch": 0.895273200346921, + "grad_norm": 3.6000866889953613, + "learning_rate": 1.4082402437255187e-05, + "loss": 0.041, + "step": 8258 + }, + { + "epoch": 0.8953816131830009, + "grad_norm": 0.9442522525787354, + "learning_rate": 1.4081677063687799e-05, + "loss": 0.01, + "step": 8259 + }, + { + "epoch": 0.8954900260190807, + "grad_norm": 1.4445059299468994, + "learning_rate": 1.4080951690120413e-05, + "loss": 0.0173, + "step": 8260 + }, + { + "epoch": 0.8955984388551604, + "grad_norm": 4.083212375640869, + "learning_rate": 1.4080226316553025e-05, + "loss": 0.0235, + "step": 8261 + }, + { + "epoch": 0.8957068516912402, + "grad_norm": 1.5218348503112793, + "learning_rate": 1.4079500942985637e-05, + "loss": 0.026, + "step": 8262 + }, + { + "epoch": 0.8958152645273201, + "grad_norm": 1.9311972856521606, + "learning_rate": 1.4078775569418253e-05, + "loss": 0.0188, + "step": 8263 + }, + { + "epoch": 0.8959236773633998, + "grad_norm": 3.4977304935455322, + "learning_rate": 1.4078050195850865e-05, + "loss": 0.0698, + "step": 8264 + }, + { + "epoch": 0.8960320901994796, + "grad_norm": 2.792728900909424, + "learning_rate": 1.4077324822283479e-05, + "loss": 0.0617, + "step": 8265 + }, + { + "epoch": 0.8961405030355594, + "grad_norm": 2.2409777641296387, + "learning_rate": 1.407659944871609e-05, + "loss": 0.0204, + "step": 8266 + }, + { + "epoch": 0.8962489158716392, + "grad_norm": 2.026326894760132, + "learning_rate": 1.4075874075148703e-05, + "loss": 0.0406, + "step": 8267 + }, + { + "epoch": 0.896357328707719, + "grad_norm": 6.794897556304932, + "learning_rate": 1.4075148701581317e-05, + "loss": 0.0608, + "step": 8268 + }, + { + "epoch": 0.8964657415437988, + "grad_norm": 3.8285958766937256, + "learning_rate": 1.4074423328013929e-05, + "loss": 0.0343, + "step": 8269 + }, + { + "epoch": 0.8965741543798785, + "grad_norm": 5.262848377227783, + "learning_rate": 1.4073697954446541e-05, + "loss": 0.0557, + "step": 8270 + }, + { + "epoch": 0.8966825672159584, + "grad_norm": 0.9171281456947327, + "learning_rate": 1.4072972580879155e-05, + "loss": 0.0184, + "step": 8271 + }, + { + "epoch": 0.8967909800520382, + "grad_norm": 3.4985833168029785, + "learning_rate": 1.4072247207311767e-05, + "loss": 0.0449, + "step": 8272 + }, + { + "epoch": 0.8968993928881179, + "grad_norm": 2.280780076980591, + "learning_rate": 1.4071521833744379e-05, + "loss": 0.0511, + "step": 8273 + }, + { + "epoch": 0.8970078057241977, + "grad_norm": 1.5873397588729858, + "learning_rate": 1.4070796460176991e-05, + "loss": 0.0174, + "step": 8274 + }, + { + "epoch": 0.8971162185602776, + "grad_norm": 1.622533917427063, + "learning_rate": 1.4070071086609605e-05, + "loss": 0.0305, + "step": 8275 + }, + { + "epoch": 0.8972246313963573, + "grad_norm": 1.979678750038147, + "learning_rate": 1.4069345713042217e-05, + "loss": 0.0386, + "step": 8276 + }, + { + "epoch": 0.8973330442324371, + "grad_norm": 3.31545090675354, + "learning_rate": 1.406862033947483e-05, + "loss": 0.0218, + "step": 8277 + }, + { + "epoch": 0.897441457068517, + "grad_norm": 1.2455799579620361, + "learning_rate": 1.4067894965907443e-05, + "loss": 0.0106, + "step": 8278 + }, + { + "epoch": 0.8975498699045967, + "grad_norm": 2.9444215297698975, + "learning_rate": 1.4067169592340055e-05, + "loss": 0.0428, + "step": 8279 + }, + { + "epoch": 0.8976582827406765, + "grad_norm": 5.366804122924805, + "learning_rate": 1.4066444218772667e-05, + "loss": 0.0499, + "step": 8280 + }, + { + "epoch": 0.8977666955767563, + "grad_norm": 7.160105228424072, + "learning_rate": 1.4065718845205283e-05, + "loss": 0.0401, + "step": 8281 + }, + { + "epoch": 0.8978751084128361, + "grad_norm": 1.9013944864273071, + "learning_rate": 1.4064993471637895e-05, + "loss": 0.0413, + "step": 8282 + }, + { + "epoch": 0.8979835212489159, + "grad_norm": 3.923009157180786, + "learning_rate": 1.4064268098070509e-05, + "loss": 0.0453, + "step": 8283 + }, + { + "epoch": 0.8980919340849957, + "grad_norm": 1.5775058269500732, + "learning_rate": 1.4063542724503121e-05, + "loss": 0.0236, + "step": 8284 + }, + { + "epoch": 0.8982003469210754, + "grad_norm": 5.134922027587891, + "learning_rate": 1.4062817350935733e-05, + "loss": 0.0415, + "step": 8285 + }, + { + "epoch": 0.8983087597571553, + "grad_norm": 1.1480906009674072, + "learning_rate": 1.4062091977368345e-05, + "loss": 0.0186, + "step": 8286 + }, + { + "epoch": 0.8984171725932351, + "grad_norm": 1.5007569789886475, + "learning_rate": 1.406136660380096e-05, + "loss": 0.0223, + "step": 8287 + }, + { + "epoch": 0.8985255854293148, + "grad_norm": 2.6970925331115723, + "learning_rate": 1.4060641230233571e-05, + "loss": 0.032, + "step": 8288 + }, + { + "epoch": 0.8986339982653946, + "grad_norm": 5.773308277130127, + "learning_rate": 1.4059915856666184e-05, + "loss": 0.0433, + "step": 8289 + }, + { + "epoch": 0.8987424111014745, + "grad_norm": 3.1914479732513428, + "learning_rate": 1.4059190483098797e-05, + "loss": 0.0537, + "step": 8290 + }, + { + "epoch": 0.8988508239375542, + "grad_norm": 3.7280309200286865, + "learning_rate": 1.405846510953141e-05, + "loss": 0.0356, + "step": 8291 + }, + { + "epoch": 0.898959236773634, + "grad_norm": 1.7022732496261597, + "learning_rate": 1.4057739735964022e-05, + "loss": 0.01, + "step": 8292 + }, + { + "epoch": 0.8990676496097137, + "grad_norm": 2.4121527671813965, + "learning_rate": 1.4057014362396636e-05, + "loss": 0.018, + "step": 8293 + }, + { + "epoch": 0.8991760624457936, + "grad_norm": 1.3924740552902222, + "learning_rate": 1.4056288988829248e-05, + "loss": 0.0205, + "step": 8294 + }, + { + "epoch": 0.8992844752818734, + "grad_norm": 0.3389495015144348, + "learning_rate": 1.405556361526186e-05, + "loss": 0.0052, + "step": 8295 + }, + { + "epoch": 0.8993928881179531, + "grad_norm": 6.110579013824463, + "learning_rate": 1.4054838241694474e-05, + "loss": 0.0882, + "step": 8296 + }, + { + "epoch": 0.8995013009540329, + "grad_norm": 1.9771289825439453, + "learning_rate": 1.4054112868127086e-05, + "loss": 0.0185, + "step": 8297 + }, + { + "epoch": 0.8996097137901128, + "grad_norm": 0.8169770240783691, + "learning_rate": 1.40533874945597e-05, + "loss": 0.011, + "step": 8298 + }, + { + "epoch": 0.8997181266261925, + "grad_norm": 2.734151601791382, + "learning_rate": 1.4052662120992313e-05, + "loss": 0.0319, + "step": 8299 + }, + { + "epoch": 0.8998265394622723, + "grad_norm": 5.609870433807373, + "learning_rate": 1.4051936747424926e-05, + "loss": 0.0591, + "step": 8300 + }, + { + "epoch": 0.8999349522983521, + "grad_norm": 2.507356882095337, + "learning_rate": 1.4051211373857538e-05, + "loss": 0.0279, + "step": 8301 + }, + { + "epoch": 0.900043365134432, + "grad_norm": 1.9649852514266968, + "learning_rate": 1.4050486000290152e-05, + "loss": 0.0217, + "step": 8302 + }, + { + "epoch": 0.9001517779705117, + "grad_norm": 1.7771289348602295, + "learning_rate": 1.4049760626722764e-05, + "loss": 0.0129, + "step": 8303 + }, + { + "epoch": 0.9002601908065915, + "grad_norm": 3.0334715843200684, + "learning_rate": 1.4049035253155376e-05, + "loss": 0.061, + "step": 8304 + }, + { + "epoch": 0.9003686036426712, + "grad_norm": 2.7761647701263428, + "learning_rate": 1.404830987958799e-05, + "loss": 0.018, + "step": 8305 + }, + { + "epoch": 0.9004770164787511, + "grad_norm": 4.558225631713867, + "learning_rate": 1.4047584506020602e-05, + "loss": 0.0198, + "step": 8306 + }, + { + "epoch": 0.9005854293148309, + "grad_norm": 3.168891191482544, + "learning_rate": 1.4046859132453214e-05, + "loss": 0.0411, + "step": 8307 + }, + { + "epoch": 0.9006938421509106, + "grad_norm": 2.302689790725708, + "learning_rate": 1.4046133758885828e-05, + "loss": 0.0568, + "step": 8308 + }, + { + "epoch": 0.9008022549869905, + "grad_norm": 2.972642660140991, + "learning_rate": 1.404540838531844e-05, + "loss": 0.0277, + "step": 8309 + }, + { + "epoch": 0.9009106678230703, + "grad_norm": 1.4862194061279297, + "learning_rate": 1.4044683011751052e-05, + "loss": 0.0318, + "step": 8310 + }, + { + "epoch": 0.90101908065915, + "grad_norm": 1.5363081693649292, + "learning_rate": 1.4043957638183664e-05, + "loss": 0.0125, + "step": 8311 + }, + { + "epoch": 0.9011274934952298, + "grad_norm": 2.4841387271881104, + "learning_rate": 1.4043232264616278e-05, + "loss": 0.0317, + "step": 8312 + }, + { + "epoch": 0.9012359063313097, + "grad_norm": 0.8528346419334412, + "learning_rate": 1.404250689104889e-05, + "loss": 0.0095, + "step": 8313 + }, + { + "epoch": 0.9013443191673894, + "grad_norm": 2.8826956748962402, + "learning_rate": 1.4041781517481502e-05, + "loss": 0.015, + "step": 8314 + }, + { + "epoch": 0.9014527320034692, + "grad_norm": 0.5402315855026245, + "learning_rate": 1.4041056143914118e-05, + "loss": 0.0062, + "step": 8315 + }, + { + "epoch": 0.901561144839549, + "grad_norm": 0.39570245146751404, + "learning_rate": 1.404033077034673e-05, + "loss": 0.0046, + "step": 8316 + }, + { + "epoch": 0.9016695576756288, + "grad_norm": 0.48465242981910706, + "learning_rate": 1.4039605396779344e-05, + "loss": 0.0055, + "step": 8317 + }, + { + "epoch": 0.9017779705117086, + "grad_norm": 1.5928453207015991, + "learning_rate": 1.4038880023211956e-05, + "loss": 0.0422, + "step": 8318 + }, + { + "epoch": 0.9018863833477884, + "grad_norm": 1.8426361083984375, + "learning_rate": 1.4038154649644568e-05, + "loss": 0.0154, + "step": 8319 + }, + { + "epoch": 0.9019947961838681, + "grad_norm": 5.027517795562744, + "learning_rate": 1.4037429276077182e-05, + "loss": 0.0483, + "step": 8320 + }, + { + "epoch": 0.902103209019948, + "grad_norm": 1.7771984338760376, + "learning_rate": 1.4036703902509794e-05, + "loss": 0.0156, + "step": 8321 + }, + { + "epoch": 0.9022116218560278, + "grad_norm": 6.023350238800049, + "learning_rate": 1.4035978528942406e-05, + "loss": 0.0763, + "step": 8322 + }, + { + "epoch": 0.9023200346921075, + "grad_norm": 5.037346839904785, + "learning_rate": 1.403525315537502e-05, + "loss": 0.0719, + "step": 8323 + }, + { + "epoch": 0.9024284475281873, + "grad_norm": 1.3765473365783691, + "learning_rate": 1.4034527781807632e-05, + "loss": 0.0171, + "step": 8324 + }, + { + "epoch": 0.9025368603642672, + "grad_norm": 5.254991054534912, + "learning_rate": 1.4033802408240244e-05, + "loss": 0.0329, + "step": 8325 + }, + { + "epoch": 0.9026452732003469, + "grad_norm": 3.196200370788574, + "learning_rate": 1.4033077034672857e-05, + "loss": 0.0233, + "step": 8326 + }, + { + "epoch": 0.9027536860364267, + "grad_norm": 4.589845180511475, + "learning_rate": 1.403235166110547e-05, + "loss": 0.0303, + "step": 8327 + }, + { + "epoch": 0.9028620988725065, + "grad_norm": 2.7324113845825195, + "learning_rate": 1.4031626287538083e-05, + "loss": 0.0344, + "step": 8328 + }, + { + "epoch": 0.9029705117085863, + "grad_norm": 2.049126386642456, + "learning_rate": 1.4030900913970695e-05, + "loss": 0.0099, + "step": 8329 + }, + { + "epoch": 0.9030789245446661, + "grad_norm": 6.71043062210083, + "learning_rate": 1.4030175540403309e-05, + "loss": 0.0784, + "step": 8330 + }, + { + "epoch": 0.9031873373807459, + "grad_norm": 6.584545135498047, + "learning_rate": 1.402945016683592e-05, + "loss": 0.0905, + "step": 8331 + }, + { + "epoch": 0.9032957502168256, + "grad_norm": 1.7851992845535278, + "learning_rate": 1.4028724793268536e-05, + "loss": 0.0186, + "step": 8332 + }, + { + "epoch": 0.9034041630529055, + "grad_norm": 4.256053924560547, + "learning_rate": 1.4027999419701148e-05, + "loss": 0.0507, + "step": 8333 + }, + { + "epoch": 0.9035125758889853, + "grad_norm": 4.421395301818848, + "learning_rate": 1.402727404613376e-05, + "loss": 0.0453, + "step": 8334 + }, + { + "epoch": 0.903620988725065, + "grad_norm": 3.1021037101745605, + "learning_rate": 1.4026548672566374e-05, + "loss": 0.0321, + "step": 8335 + }, + { + "epoch": 0.9037294015611448, + "grad_norm": 4.069814205169678, + "learning_rate": 1.4025823298998986e-05, + "loss": 0.1117, + "step": 8336 + }, + { + "epoch": 0.9038378143972247, + "grad_norm": 6.978459358215332, + "learning_rate": 1.4025097925431599e-05, + "loss": 0.0446, + "step": 8337 + }, + { + "epoch": 0.9039462272333044, + "grad_norm": 3.0451009273529053, + "learning_rate": 1.402437255186421e-05, + "loss": 0.0348, + "step": 8338 + }, + { + "epoch": 0.9040546400693842, + "grad_norm": 2.8020546436309814, + "learning_rate": 1.4023647178296825e-05, + "loss": 0.044, + "step": 8339 + }, + { + "epoch": 0.9041630529054641, + "grad_norm": 4.14419412612915, + "learning_rate": 1.4022921804729437e-05, + "loss": 0.0365, + "step": 8340 + }, + { + "epoch": 0.9042714657415438, + "grad_norm": 1.4713126420974731, + "learning_rate": 1.4022196431162049e-05, + "loss": 0.0232, + "step": 8341 + }, + { + "epoch": 0.9043798785776236, + "grad_norm": 2.476170778274536, + "learning_rate": 1.4021471057594663e-05, + "loss": 0.0345, + "step": 8342 + }, + { + "epoch": 0.9044882914137033, + "grad_norm": 4.523934841156006, + "learning_rate": 1.4020745684027275e-05, + "loss": 0.0176, + "step": 8343 + }, + { + "epoch": 0.9045967042497832, + "grad_norm": 1.3135902881622314, + "learning_rate": 1.4020020310459887e-05, + "loss": 0.0201, + "step": 8344 + }, + { + "epoch": 0.904705117085863, + "grad_norm": 3.4832935333251953, + "learning_rate": 1.4019294936892501e-05, + "loss": 0.0182, + "step": 8345 + }, + { + "epoch": 0.9048135299219427, + "grad_norm": 0.8566215634346008, + "learning_rate": 1.4018569563325113e-05, + "loss": 0.0166, + "step": 8346 + }, + { + "epoch": 0.9049219427580225, + "grad_norm": 2.0454509258270264, + "learning_rate": 1.4017844189757725e-05, + "loss": 0.0736, + "step": 8347 + }, + { + "epoch": 0.9050303555941024, + "grad_norm": 3.369948387145996, + "learning_rate": 1.4017118816190339e-05, + "loss": 0.0651, + "step": 8348 + }, + { + "epoch": 0.9051387684301822, + "grad_norm": 1.4488424062728882, + "learning_rate": 1.4016393442622951e-05, + "loss": 0.0105, + "step": 8349 + }, + { + "epoch": 0.9052471812662619, + "grad_norm": 3.3316550254821777, + "learning_rate": 1.4015668069055565e-05, + "loss": 0.0466, + "step": 8350 + }, + { + "epoch": 0.9053555941023417, + "grad_norm": 1.5843784809112549, + "learning_rate": 1.4014942695488179e-05, + "loss": 0.0292, + "step": 8351 + }, + { + "epoch": 0.9054640069384216, + "grad_norm": 2.54097056388855, + "learning_rate": 1.4014217321920791e-05, + "loss": 0.0375, + "step": 8352 + }, + { + "epoch": 0.9055724197745013, + "grad_norm": 2.9824271202087402, + "learning_rate": 1.4013491948353403e-05, + "loss": 0.072, + "step": 8353 + }, + { + "epoch": 0.9056808326105811, + "grad_norm": 3.255727767944336, + "learning_rate": 1.4012766574786017e-05, + "loss": 0.0429, + "step": 8354 + }, + { + "epoch": 0.9057892454466608, + "grad_norm": 3.223811388015747, + "learning_rate": 1.4012041201218629e-05, + "loss": 0.0493, + "step": 8355 + }, + { + "epoch": 0.9058976582827407, + "grad_norm": 2.4308948516845703, + "learning_rate": 1.4011315827651241e-05, + "loss": 0.042, + "step": 8356 + }, + { + "epoch": 0.9060060711188205, + "grad_norm": 2.2864267826080322, + "learning_rate": 1.4010590454083855e-05, + "loss": 0.0519, + "step": 8357 + }, + { + "epoch": 0.9061144839549002, + "grad_norm": 2.421469211578369, + "learning_rate": 1.4009865080516467e-05, + "loss": 0.0398, + "step": 8358 + }, + { + "epoch": 0.90622289679098, + "grad_norm": 2.23480486869812, + "learning_rate": 1.400913970694908e-05, + "loss": 0.0178, + "step": 8359 + }, + { + "epoch": 0.9063313096270599, + "grad_norm": 2.4703941345214844, + "learning_rate": 1.4008414333381693e-05, + "loss": 0.0537, + "step": 8360 + }, + { + "epoch": 0.9064397224631396, + "grad_norm": 1.3697996139526367, + "learning_rate": 1.4007688959814305e-05, + "loss": 0.017, + "step": 8361 + }, + { + "epoch": 0.9065481352992194, + "grad_norm": 1.203723430633545, + "learning_rate": 1.4006963586246917e-05, + "loss": 0.0123, + "step": 8362 + }, + { + "epoch": 0.9066565481352992, + "grad_norm": 1.3816462755203247, + "learning_rate": 1.400623821267953e-05, + "loss": 0.0178, + "step": 8363 + }, + { + "epoch": 0.906764960971379, + "grad_norm": 1.996351957321167, + "learning_rate": 1.4005512839112143e-05, + "loss": 0.0362, + "step": 8364 + }, + { + "epoch": 0.9068733738074588, + "grad_norm": 7.953176975250244, + "learning_rate": 1.4004787465544756e-05, + "loss": 0.0989, + "step": 8365 + }, + { + "epoch": 0.9069817866435386, + "grad_norm": 1.5689573287963867, + "learning_rate": 1.4004062091977368e-05, + "loss": 0.0441, + "step": 8366 + }, + { + "epoch": 0.9070901994796184, + "grad_norm": 2.8643364906311035, + "learning_rate": 1.4003336718409983e-05, + "loss": 0.0315, + "step": 8367 + }, + { + "epoch": 0.9071986123156982, + "grad_norm": 4.359369277954102, + "learning_rate": 1.4002611344842595e-05, + "loss": 0.0184, + "step": 8368 + }, + { + "epoch": 0.907307025151778, + "grad_norm": 5.657686233520508, + "learning_rate": 1.400188597127521e-05, + "loss": 0.0253, + "step": 8369 + }, + { + "epoch": 0.9074154379878577, + "grad_norm": 5.587958812713623, + "learning_rate": 1.4001160597707821e-05, + "loss": 0.0549, + "step": 8370 + }, + { + "epoch": 0.9075238508239376, + "grad_norm": 1.413216471672058, + "learning_rate": 1.4000435224140434e-05, + "loss": 0.02, + "step": 8371 + }, + { + "epoch": 0.9076322636600174, + "grad_norm": 1.5797715187072754, + "learning_rate": 1.3999709850573047e-05, + "loss": 0.0403, + "step": 8372 + }, + { + "epoch": 0.9077406764960971, + "grad_norm": 2.6975436210632324, + "learning_rate": 1.399898447700566e-05, + "loss": 0.0482, + "step": 8373 + }, + { + "epoch": 0.9078490893321769, + "grad_norm": 1.1023377180099487, + "learning_rate": 1.3998259103438272e-05, + "loss": 0.0127, + "step": 8374 + }, + { + "epoch": 0.9079575021682568, + "grad_norm": 9.26219654083252, + "learning_rate": 1.3997533729870885e-05, + "loss": 0.0715, + "step": 8375 + }, + { + "epoch": 0.9080659150043365, + "grad_norm": 2.410871744155884, + "learning_rate": 1.3996808356303498e-05, + "loss": 0.0504, + "step": 8376 + }, + { + "epoch": 0.9081743278404163, + "grad_norm": 3.686931848526001, + "learning_rate": 1.399608298273611e-05, + "loss": 0.0155, + "step": 8377 + }, + { + "epoch": 0.9082827406764961, + "grad_norm": 5.3265180587768555, + "learning_rate": 1.3995357609168722e-05, + "loss": 0.1037, + "step": 8378 + }, + { + "epoch": 0.9083911535125759, + "grad_norm": 2.6756138801574707, + "learning_rate": 1.3994632235601336e-05, + "loss": 0.0396, + "step": 8379 + }, + { + "epoch": 0.9084995663486557, + "grad_norm": 2.0785434246063232, + "learning_rate": 1.3993906862033948e-05, + "loss": 0.0302, + "step": 8380 + }, + { + "epoch": 0.9086079791847355, + "grad_norm": 0.8396346569061279, + "learning_rate": 1.399318148846656e-05, + "loss": 0.025, + "step": 8381 + }, + { + "epoch": 0.9087163920208152, + "grad_norm": 2.0035929679870605, + "learning_rate": 1.3992456114899174e-05, + "loss": 0.0204, + "step": 8382 + }, + { + "epoch": 0.9088248048568951, + "grad_norm": 2.963052988052368, + "learning_rate": 1.3991730741331786e-05, + "loss": 0.057, + "step": 8383 + }, + { + "epoch": 0.9089332176929749, + "grad_norm": 1.8634177446365356, + "learning_rate": 1.3991005367764402e-05, + "loss": 0.0383, + "step": 8384 + }, + { + "epoch": 0.9090416305290546, + "grad_norm": 3.2357611656188965, + "learning_rate": 1.3990279994197014e-05, + "loss": 0.0393, + "step": 8385 + }, + { + "epoch": 0.9091500433651344, + "grad_norm": 3.6230156421661377, + "learning_rate": 1.3989554620629626e-05, + "loss": 0.0331, + "step": 8386 + }, + { + "epoch": 0.9092584562012143, + "grad_norm": 0.9975314140319824, + "learning_rate": 1.398882924706224e-05, + "loss": 0.0248, + "step": 8387 + }, + { + "epoch": 0.909366869037294, + "grad_norm": 2.2959156036376953, + "learning_rate": 1.3988103873494852e-05, + "loss": 0.0488, + "step": 8388 + }, + { + "epoch": 0.9094752818733738, + "grad_norm": 0.7901853919029236, + "learning_rate": 1.3987378499927464e-05, + "loss": 0.0099, + "step": 8389 + }, + { + "epoch": 0.9095836947094535, + "grad_norm": 1.569171667098999, + "learning_rate": 1.3986653126360076e-05, + "loss": 0.0212, + "step": 8390 + }, + { + "epoch": 0.9096921075455334, + "grad_norm": 1.3371672630310059, + "learning_rate": 1.398592775279269e-05, + "loss": 0.0262, + "step": 8391 + }, + { + "epoch": 0.9098005203816132, + "grad_norm": 0.8651427626609802, + "learning_rate": 1.3985202379225302e-05, + "loss": 0.0154, + "step": 8392 + }, + { + "epoch": 0.909908933217693, + "grad_norm": 3.1030380725860596, + "learning_rate": 1.3984477005657914e-05, + "loss": 0.0727, + "step": 8393 + }, + { + "epoch": 0.9100173460537727, + "grad_norm": 3.5123207569122314, + "learning_rate": 1.3983751632090528e-05, + "loss": 0.047, + "step": 8394 + }, + { + "epoch": 0.9101257588898526, + "grad_norm": 2.099821090698242, + "learning_rate": 1.398302625852314e-05, + "loss": 0.048, + "step": 8395 + }, + { + "epoch": 0.9102341717259324, + "grad_norm": 1.6457762718200684, + "learning_rate": 1.3982300884955752e-05, + "loss": 0.0192, + "step": 8396 + }, + { + "epoch": 0.9103425845620121, + "grad_norm": 1.2651816606521606, + "learning_rate": 1.3981575511388366e-05, + "loss": 0.0236, + "step": 8397 + }, + { + "epoch": 0.910450997398092, + "grad_norm": 1.2562614679336548, + "learning_rate": 1.3980850137820978e-05, + "loss": 0.0196, + "step": 8398 + }, + { + "epoch": 0.9105594102341718, + "grad_norm": 1.1374123096466064, + "learning_rate": 1.398012476425359e-05, + "loss": 0.0266, + "step": 8399 + }, + { + "epoch": 0.9106678230702515, + "grad_norm": 1.2911728620529175, + "learning_rate": 1.3979399390686204e-05, + "loss": 0.0345, + "step": 8400 + }, + { + "epoch": 0.9107762359063313, + "grad_norm": 1.5075355768203735, + "learning_rate": 1.3978674017118818e-05, + "loss": 0.0082, + "step": 8401 + }, + { + "epoch": 0.9108846487424112, + "grad_norm": 0.23309792578220367, + "learning_rate": 1.397794864355143e-05, + "loss": 0.0028, + "step": 8402 + }, + { + "epoch": 0.9109930615784909, + "grad_norm": 1.610909342765808, + "learning_rate": 1.3977223269984044e-05, + "loss": 0.0427, + "step": 8403 + }, + { + "epoch": 0.9111014744145707, + "grad_norm": 0.7765282392501831, + "learning_rate": 1.3976497896416656e-05, + "loss": 0.013, + "step": 8404 + }, + { + "epoch": 0.9112098872506504, + "grad_norm": 2.8082146644592285, + "learning_rate": 1.3975772522849268e-05, + "loss": 0.0454, + "step": 8405 + }, + { + "epoch": 0.9113183000867303, + "grad_norm": 6.097225666046143, + "learning_rate": 1.3975047149281882e-05, + "loss": 0.043, + "step": 8406 + }, + { + "epoch": 0.9114267129228101, + "grad_norm": 6.285368919372559, + "learning_rate": 1.3974321775714494e-05, + "loss": 0.0604, + "step": 8407 + }, + { + "epoch": 0.9115351257588898, + "grad_norm": 3.8050084114074707, + "learning_rate": 1.3973596402147107e-05, + "loss": 0.0233, + "step": 8408 + }, + { + "epoch": 0.9116435385949696, + "grad_norm": 11.87686538696289, + "learning_rate": 1.397287102857972e-05, + "loss": 0.02, + "step": 8409 + }, + { + "epoch": 0.9117519514310495, + "grad_norm": 4.664642333984375, + "learning_rate": 1.3972145655012333e-05, + "loss": 0.0874, + "step": 8410 + }, + { + "epoch": 0.9118603642671292, + "grad_norm": 4.319272041320801, + "learning_rate": 1.3971420281444945e-05, + "loss": 0.0174, + "step": 8411 + }, + { + "epoch": 0.911968777103209, + "grad_norm": 1.2757232189178467, + "learning_rate": 1.3970694907877559e-05, + "loss": 0.0331, + "step": 8412 + }, + { + "epoch": 0.9120771899392888, + "grad_norm": 5.735389232635498, + "learning_rate": 1.396996953431017e-05, + "loss": 0.1202, + "step": 8413 + }, + { + "epoch": 0.9121856027753686, + "grad_norm": 0.7851817011833191, + "learning_rate": 1.3969244160742783e-05, + "loss": 0.0108, + "step": 8414 + }, + { + "epoch": 0.9122940156114484, + "grad_norm": 0.9095591306686401, + "learning_rate": 1.3968518787175395e-05, + "loss": 0.0056, + "step": 8415 + }, + { + "epoch": 0.9124024284475282, + "grad_norm": 1.2571004629135132, + "learning_rate": 1.3967793413608009e-05, + "loss": 0.0175, + "step": 8416 + }, + { + "epoch": 0.9125108412836079, + "grad_norm": 2.9266505241394043, + "learning_rate": 1.3967068040040621e-05, + "loss": 0.0413, + "step": 8417 + }, + { + "epoch": 0.9126192541196878, + "grad_norm": 5.112222671508789, + "learning_rate": 1.3966342666473233e-05, + "loss": 0.0448, + "step": 8418 + }, + { + "epoch": 0.9127276669557676, + "grad_norm": 2.77168607711792, + "learning_rate": 1.3965617292905849e-05, + "loss": 0.0468, + "step": 8419 + }, + { + "epoch": 0.9128360797918473, + "grad_norm": 6.708448886871338, + "learning_rate": 1.396489191933846e-05, + "loss": 0.0495, + "step": 8420 + }, + { + "epoch": 0.9129444926279271, + "grad_norm": 1.460741639137268, + "learning_rate": 1.3964166545771075e-05, + "loss": 0.0163, + "step": 8421 + }, + { + "epoch": 0.913052905464007, + "grad_norm": 1.4521501064300537, + "learning_rate": 1.3963441172203687e-05, + "loss": 0.0265, + "step": 8422 + }, + { + "epoch": 0.9131613183000867, + "grad_norm": 3.800154209136963, + "learning_rate": 1.3962715798636299e-05, + "loss": 0.0236, + "step": 8423 + }, + { + "epoch": 0.9132697311361665, + "grad_norm": 5.9727582931518555, + "learning_rate": 1.3961990425068913e-05, + "loss": 0.0352, + "step": 8424 + }, + { + "epoch": 0.9133781439722463, + "grad_norm": 6.312608242034912, + "learning_rate": 1.3961265051501525e-05, + "loss": 0.0817, + "step": 8425 + }, + { + "epoch": 0.9134865568083261, + "grad_norm": 2.2784066200256348, + "learning_rate": 1.3960539677934137e-05, + "loss": 0.0446, + "step": 8426 + }, + { + "epoch": 0.9135949696444059, + "grad_norm": 2.5310511589050293, + "learning_rate": 1.395981430436675e-05, + "loss": 0.0286, + "step": 8427 + }, + { + "epoch": 0.9137033824804857, + "grad_norm": 4.879404067993164, + "learning_rate": 1.3959088930799363e-05, + "loss": 0.0341, + "step": 8428 + }, + { + "epoch": 0.9138117953165655, + "grad_norm": 1.6890206336975098, + "learning_rate": 1.3958363557231975e-05, + "loss": 0.0163, + "step": 8429 + }, + { + "epoch": 0.9139202081526453, + "grad_norm": 1.898097038269043, + "learning_rate": 1.3957638183664587e-05, + "loss": 0.0546, + "step": 8430 + }, + { + "epoch": 0.9140286209887251, + "grad_norm": 2.987457036972046, + "learning_rate": 1.3956912810097201e-05, + "loss": 0.0235, + "step": 8431 + }, + { + "epoch": 0.9141370338248048, + "grad_norm": 0.9947543740272522, + "learning_rate": 1.3956187436529813e-05, + "loss": 0.0162, + "step": 8432 + }, + { + "epoch": 0.9142454466608847, + "grad_norm": 0.5241755843162537, + "learning_rate": 1.3955462062962425e-05, + "loss": 0.003, + "step": 8433 + }, + { + "epoch": 0.9143538594969645, + "grad_norm": 1.0250399112701416, + "learning_rate": 1.395473668939504e-05, + "loss": 0.022, + "step": 8434 + }, + { + "epoch": 0.9144622723330442, + "grad_norm": 0.46598124504089355, + "learning_rate": 1.3954011315827651e-05, + "loss": 0.0073, + "step": 8435 + }, + { + "epoch": 0.914570685169124, + "grad_norm": 3.141772747039795, + "learning_rate": 1.3953285942260267e-05, + "loss": 0.0558, + "step": 8436 + }, + { + "epoch": 0.9146790980052039, + "grad_norm": 1.6632139682769775, + "learning_rate": 1.3952560568692879e-05, + "loss": 0.03, + "step": 8437 + }, + { + "epoch": 0.9147875108412836, + "grad_norm": 0.7600255012512207, + "learning_rate": 1.3951835195125491e-05, + "loss": 0.0133, + "step": 8438 + }, + { + "epoch": 0.9148959236773634, + "grad_norm": 2.888866424560547, + "learning_rate": 1.3951109821558105e-05, + "loss": 0.0139, + "step": 8439 + }, + { + "epoch": 0.9150043365134432, + "grad_norm": 2.6745340824127197, + "learning_rate": 1.3950384447990717e-05, + "loss": 0.0223, + "step": 8440 + }, + { + "epoch": 0.915112749349523, + "grad_norm": 3.6253085136413574, + "learning_rate": 1.394965907442333e-05, + "loss": 0.0762, + "step": 8441 + }, + { + "epoch": 0.9152211621856028, + "grad_norm": 2.7503013610839844, + "learning_rate": 1.3948933700855941e-05, + "loss": 0.061, + "step": 8442 + }, + { + "epoch": 0.9153295750216826, + "grad_norm": 3.7586421966552734, + "learning_rate": 1.3948208327288555e-05, + "loss": 0.0309, + "step": 8443 + }, + { + "epoch": 0.9154379878577623, + "grad_norm": 4.503740310668945, + "learning_rate": 1.3947482953721167e-05, + "loss": 0.0424, + "step": 8444 + }, + { + "epoch": 0.9155464006938422, + "grad_norm": 1.6820027828216553, + "learning_rate": 1.394675758015378e-05, + "loss": 0.028, + "step": 8445 + }, + { + "epoch": 0.915654813529922, + "grad_norm": 0.7517967224121094, + "learning_rate": 1.3946032206586393e-05, + "loss": 0.0077, + "step": 8446 + }, + { + "epoch": 0.9157632263660017, + "grad_norm": 3.2900264263153076, + "learning_rate": 1.3945306833019006e-05, + "loss": 0.0462, + "step": 8447 + }, + { + "epoch": 0.9158716392020815, + "grad_norm": 0.400535523891449, + "learning_rate": 1.3944581459451618e-05, + "loss": 0.0052, + "step": 8448 + }, + { + "epoch": 0.9159800520381614, + "grad_norm": 0.8359771370887756, + "learning_rate": 1.3943856085884232e-05, + "loss": 0.0112, + "step": 8449 + }, + { + "epoch": 0.9160884648742411, + "grad_norm": 5.346131801605225, + "learning_rate": 1.3943130712316844e-05, + "loss": 0.0376, + "step": 8450 + }, + { + "epoch": 0.9161968777103209, + "grad_norm": 2.9303290843963623, + "learning_rate": 1.3942405338749456e-05, + "loss": 0.034, + "step": 8451 + }, + { + "epoch": 0.9163052905464006, + "grad_norm": 1.0719205141067505, + "learning_rate": 1.394167996518207e-05, + "loss": 0.0086, + "step": 8452 + }, + { + "epoch": 0.9164137033824805, + "grad_norm": 3.5398037433624268, + "learning_rate": 1.3940954591614683e-05, + "loss": 0.056, + "step": 8453 + }, + { + "epoch": 0.9165221162185603, + "grad_norm": 6.3816070556640625, + "learning_rate": 1.3940229218047296e-05, + "loss": 0.0402, + "step": 8454 + }, + { + "epoch": 0.91663052905464, + "grad_norm": 4.347390174865723, + "learning_rate": 1.393950384447991e-05, + "loss": 0.0529, + "step": 8455 + }, + { + "epoch": 0.9167389418907199, + "grad_norm": 4.652406692504883, + "learning_rate": 1.3938778470912522e-05, + "loss": 0.0486, + "step": 8456 + }, + { + "epoch": 0.9168473547267997, + "grad_norm": 3.1865625381469727, + "learning_rate": 1.3938053097345134e-05, + "loss": 0.0446, + "step": 8457 + }, + { + "epoch": 0.9169557675628794, + "grad_norm": 3.7809767723083496, + "learning_rate": 1.3937327723777748e-05, + "loss": 0.016, + "step": 8458 + }, + { + "epoch": 0.9170641803989592, + "grad_norm": 2.0268964767456055, + "learning_rate": 1.393660235021036e-05, + "loss": 0.0278, + "step": 8459 + }, + { + "epoch": 0.9171725932350391, + "grad_norm": 1.2483668327331543, + "learning_rate": 1.3935876976642972e-05, + "loss": 0.0237, + "step": 8460 + }, + { + "epoch": 0.9172810060711188, + "grad_norm": 5.363368511199951, + "learning_rate": 1.3935151603075586e-05, + "loss": 0.0494, + "step": 8461 + }, + { + "epoch": 0.9173894189071986, + "grad_norm": 1.5439043045043945, + "learning_rate": 1.3934426229508198e-05, + "loss": 0.0316, + "step": 8462 + }, + { + "epoch": 0.9174978317432784, + "grad_norm": 2.0525894165039062, + "learning_rate": 1.393370085594081e-05, + "loss": 0.04, + "step": 8463 + }, + { + "epoch": 0.9176062445793582, + "grad_norm": 0.40995433926582336, + "learning_rate": 1.3932975482373424e-05, + "loss": 0.0047, + "step": 8464 + }, + { + "epoch": 0.917714657415438, + "grad_norm": 3.813476085662842, + "learning_rate": 1.3932250108806036e-05, + "loss": 0.0346, + "step": 8465 + }, + { + "epoch": 0.9178230702515178, + "grad_norm": 4.442056179046631, + "learning_rate": 1.3931524735238648e-05, + "loss": 0.0304, + "step": 8466 + }, + { + "epoch": 0.9179314830875975, + "grad_norm": 2.6280055046081543, + "learning_rate": 1.3930799361671262e-05, + "loss": 0.0374, + "step": 8467 + }, + { + "epoch": 0.9180398959236774, + "grad_norm": 5.99978494644165, + "learning_rate": 1.3930073988103874e-05, + "loss": 0.0663, + "step": 8468 + }, + { + "epoch": 0.9181483087597572, + "grad_norm": 4.510543346405029, + "learning_rate": 1.3929348614536486e-05, + "loss": 0.0962, + "step": 8469 + }, + { + "epoch": 0.9182567215958369, + "grad_norm": 2.7554080486297607, + "learning_rate": 1.3928623240969098e-05, + "loss": 0.08, + "step": 8470 + }, + { + "epoch": 0.9183651344319167, + "grad_norm": 2.514972686767578, + "learning_rate": 1.3927897867401714e-05, + "loss": 0.0422, + "step": 8471 + }, + { + "epoch": 0.9184735472679966, + "grad_norm": 2.3576180934906006, + "learning_rate": 1.3927172493834326e-05, + "loss": 0.0508, + "step": 8472 + }, + { + "epoch": 0.9185819601040763, + "grad_norm": 1.902227759361267, + "learning_rate": 1.392644712026694e-05, + "loss": 0.0339, + "step": 8473 + }, + { + "epoch": 0.9186903729401561, + "grad_norm": 4.76362943649292, + "learning_rate": 1.3925721746699552e-05, + "loss": 0.0546, + "step": 8474 + }, + { + "epoch": 0.9187987857762359, + "grad_norm": 1.631369709968567, + "learning_rate": 1.3924996373132164e-05, + "loss": 0.0609, + "step": 8475 + }, + { + "epoch": 0.9189071986123157, + "grad_norm": 3.276959180831909, + "learning_rate": 1.3924270999564778e-05, + "loss": 0.0566, + "step": 8476 + }, + { + "epoch": 0.9190156114483955, + "grad_norm": 3.060988426208496, + "learning_rate": 1.392354562599739e-05, + "loss": 0.0464, + "step": 8477 + }, + { + "epoch": 0.9191240242844753, + "grad_norm": 3.9001383781433105, + "learning_rate": 1.3922820252430002e-05, + "loss": 0.0722, + "step": 8478 + }, + { + "epoch": 0.919232437120555, + "grad_norm": 2.7481842041015625, + "learning_rate": 1.3922094878862616e-05, + "loss": 0.0254, + "step": 8479 + }, + { + "epoch": 0.9193408499566349, + "grad_norm": 3.2228198051452637, + "learning_rate": 1.3921369505295228e-05, + "loss": 0.0183, + "step": 8480 + }, + { + "epoch": 0.9194492627927147, + "grad_norm": 5.009462356567383, + "learning_rate": 1.392064413172784e-05, + "loss": 0.0447, + "step": 8481 + }, + { + "epoch": 0.9195576756287944, + "grad_norm": 2.4413154125213623, + "learning_rate": 1.3919918758160453e-05, + "loss": 0.0303, + "step": 8482 + }, + { + "epoch": 0.9196660884648742, + "grad_norm": 1.668933629989624, + "learning_rate": 1.3919193384593066e-05, + "loss": 0.0331, + "step": 8483 + }, + { + "epoch": 0.9197745013009541, + "grad_norm": 5.039419174194336, + "learning_rate": 1.3918468011025679e-05, + "loss": 0.0625, + "step": 8484 + }, + { + "epoch": 0.9198829141370338, + "grad_norm": 1.1684705018997192, + "learning_rate": 1.391774263745829e-05, + "loss": 0.0241, + "step": 8485 + }, + { + "epoch": 0.9199913269731136, + "grad_norm": 1.1543831825256348, + "learning_rate": 1.3917017263890905e-05, + "loss": 0.0274, + "step": 8486 + }, + { + "epoch": 0.9200997398091935, + "grad_norm": 3.4961392879486084, + "learning_rate": 1.3916291890323517e-05, + "loss": 0.0615, + "step": 8487 + }, + { + "epoch": 0.9202081526452732, + "grad_norm": 1.2600023746490479, + "learning_rate": 1.3915566516756132e-05, + "loss": 0.0386, + "step": 8488 + }, + { + "epoch": 0.920316565481353, + "grad_norm": 3.4596173763275146, + "learning_rate": 1.3914841143188744e-05, + "loss": 0.0314, + "step": 8489 + }, + { + "epoch": 0.9204249783174328, + "grad_norm": 2.304016351699829, + "learning_rate": 1.3914115769621357e-05, + "loss": 0.0292, + "step": 8490 + }, + { + "epoch": 0.9205333911535126, + "grad_norm": 1.0195846557617188, + "learning_rate": 1.391339039605397e-05, + "loss": 0.0217, + "step": 8491 + }, + { + "epoch": 0.9206418039895924, + "grad_norm": 1.9847793579101562, + "learning_rate": 1.3912665022486582e-05, + "loss": 0.0503, + "step": 8492 + }, + { + "epoch": 0.9207502168256722, + "grad_norm": 2.3105673789978027, + "learning_rate": 1.3911939648919195e-05, + "loss": 0.0226, + "step": 8493 + }, + { + "epoch": 0.9208586296617519, + "grad_norm": 0.7644813060760498, + "learning_rate": 1.3911214275351807e-05, + "loss": 0.0208, + "step": 8494 + }, + { + "epoch": 0.9209670424978318, + "grad_norm": 1.5255510807037354, + "learning_rate": 1.391048890178442e-05, + "loss": 0.0349, + "step": 8495 + }, + { + "epoch": 0.9210754553339116, + "grad_norm": 1.3901594877243042, + "learning_rate": 1.3909763528217033e-05, + "loss": 0.0132, + "step": 8496 + }, + { + "epoch": 0.9211838681699913, + "grad_norm": 2.250382900238037, + "learning_rate": 1.3909038154649645e-05, + "loss": 0.0282, + "step": 8497 + }, + { + "epoch": 0.9212922810060711, + "grad_norm": 1.39760422706604, + "learning_rate": 1.3908312781082259e-05, + "loss": 0.0224, + "step": 8498 + }, + { + "epoch": 0.921400693842151, + "grad_norm": 2.0609405040740967, + "learning_rate": 1.3907587407514871e-05, + "loss": 0.0759, + "step": 8499 + }, + { + "epoch": 0.9215091066782307, + "grad_norm": 2.3160881996154785, + "learning_rate": 1.3906862033947483e-05, + "loss": 0.0375, + "step": 8500 + }, + { + "epoch": 0.9216175195143105, + "grad_norm": 1.9063149690628052, + "learning_rate": 1.3906136660380097e-05, + "loss": 0.0336, + "step": 8501 + }, + { + "epoch": 0.9217259323503902, + "grad_norm": 2.8560941219329834, + "learning_rate": 1.3905411286812709e-05, + "loss": 0.0313, + "step": 8502 + }, + { + "epoch": 0.9218343451864701, + "grad_norm": 0.6314084529876709, + "learning_rate": 1.3904685913245321e-05, + "loss": 0.007, + "step": 8503 + }, + { + "epoch": 0.9219427580225499, + "grad_norm": 5.412745952606201, + "learning_rate": 1.3903960539677935e-05, + "loss": 0.0488, + "step": 8504 + }, + { + "epoch": 0.9220511708586296, + "grad_norm": 4.3545403480529785, + "learning_rate": 1.3903235166110549e-05, + "loss": 0.0274, + "step": 8505 + }, + { + "epoch": 0.9221595836947094, + "grad_norm": 8.283913612365723, + "learning_rate": 1.3902509792543161e-05, + "loss": 0.0416, + "step": 8506 + }, + { + "epoch": 0.9222679965307893, + "grad_norm": 1.7499359846115112, + "learning_rate": 1.3901784418975775e-05, + "loss": 0.0393, + "step": 8507 + }, + { + "epoch": 0.922376409366869, + "grad_norm": 1.4723316431045532, + "learning_rate": 1.3901059045408387e-05, + "loss": 0.0345, + "step": 8508 + }, + { + "epoch": 0.9224848222029488, + "grad_norm": 6.843486785888672, + "learning_rate": 1.3900333671840999e-05, + "loss": 0.0502, + "step": 8509 + }, + { + "epoch": 0.9225932350390286, + "grad_norm": 4.728048801422119, + "learning_rate": 1.3899608298273613e-05, + "loss": 0.0504, + "step": 8510 + }, + { + "epoch": 0.9227016478751084, + "grad_norm": 3.001657724380493, + "learning_rate": 1.3898882924706225e-05, + "loss": 0.0343, + "step": 8511 + }, + { + "epoch": 0.9228100607111882, + "grad_norm": 5.466350555419922, + "learning_rate": 1.3898157551138837e-05, + "loss": 0.06, + "step": 8512 + }, + { + "epoch": 0.922918473547268, + "grad_norm": 2.936077356338501, + "learning_rate": 1.3897432177571451e-05, + "loss": 0.0521, + "step": 8513 + }, + { + "epoch": 0.9230268863833477, + "grad_norm": 0.7408561706542969, + "learning_rate": 1.3896706804004063e-05, + "loss": 0.0098, + "step": 8514 + }, + { + "epoch": 0.9231352992194276, + "grad_norm": 2.38419508934021, + "learning_rate": 1.3895981430436675e-05, + "loss": 0.0279, + "step": 8515 + }, + { + "epoch": 0.9232437120555074, + "grad_norm": 4.58666467666626, + "learning_rate": 1.389525605686929e-05, + "loss": 0.0846, + "step": 8516 + }, + { + "epoch": 0.9233521248915871, + "grad_norm": 10.348797798156738, + "learning_rate": 1.3894530683301901e-05, + "loss": 0.0422, + "step": 8517 + }, + { + "epoch": 0.923460537727667, + "grad_norm": 3.2230947017669678, + "learning_rate": 1.3893805309734513e-05, + "loss": 0.0645, + "step": 8518 + }, + { + "epoch": 0.9235689505637468, + "grad_norm": 3.204894781112671, + "learning_rate": 1.3893079936167127e-05, + "loss": 0.0265, + "step": 8519 + }, + { + "epoch": 0.9236773633998265, + "grad_norm": 3.2018353939056396, + "learning_rate": 1.389235456259974e-05, + "loss": 0.0605, + "step": 8520 + }, + { + "epoch": 0.9237857762359063, + "grad_norm": 2.640467405319214, + "learning_rate": 1.3891629189032352e-05, + "loss": 0.0444, + "step": 8521 + }, + { + "epoch": 0.9238941890719862, + "grad_norm": 2.935930013656616, + "learning_rate": 1.3890903815464967e-05, + "loss": 0.0613, + "step": 8522 + }, + { + "epoch": 0.9240026019080659, + "grad_norm": 1.105961561203003, + "learning_rate": 1.389017844189758e-05, + "loss": 0.0174, + "step": 8523 + }, + { + "epoch": 0.9241110147441457, + "grad_norm": 4.632648468017578, + "learning_rate": 1.3889453068330191e-05, + "loss": 0.0382, + "step": 8524 + }, + { + "epoch": 0.9242194275802255, + "grad_norm": 1.5653190612792969, + "learning_rate": 1.3888727694762805e-05, + "loss": 0.0347, + "step": 8525 + }, + { + "epoch": 0.9243278404163053, + "grad_norm": 4.926639080047607, + "learning_rate": 1.3888002321195417e-05, + "loss": 0.0346, + "step": 8526 + }, + { + "epoch": 0.9244362532523851, + "grad_norm": 1.3088148832321167, + "learning_rate": 1.388727694762803e-05, + "loss": 0.008, + "step": 8527 + }, + { + "epoch": 0.9245446660884649, + "grad_norm": 2.7849578857421875, + "learning_rate": 1.3886551574060643e-05, + "loss": 0.0463, + "step": 8528 + }, + { + "epoch": 0.9246530789245446, + "grad_norm": 3.1930429935455322, + "learning_rate": 1.3885826200493256e-05, + "loss": 0.0457, + "step": 8529 + }, + { + "epoch": 0.9247614917606245, + "grad_norm": 1.6945810317993164, + "learning_rate": 1.3885100826925868e-05, + "loss": 0.0215, + "step": 8530 + }, + { + "epoch": 0.9248699045967043, + "grad_norm": 3.72428035736084, + "learning_rate": 1.3884375453358481e-05, + "loss": 0.051, + "step": 8531 + }, + { + "epoch": 0.924978317432784, + "grad_norm": 2.824763298034668, + "learning_rate": 1.3883650079791094e-05, + "loss": 0.0238, + "step": 8532 + }, + { + "epoch": 0.9250867302688638, + "grad_norm": 4.700382232666016, + "learning_rate": 1.3882924706223706e-05, + "loss": 0.0721, + "step": 8533 + }, + { + "epoch": 0.9251951431049437, + "grad_norm": 1.130699634552002, + "learning_rate": 1.3882199332656318e-05, + "loss": 0.0168, + "step": 8534 + }, + { + "epoch": 0.9253035559410234, + "grad_norm": 3.0887129306793213, + "learning_rate": 1.3881473959088932e-05, + "loss": 0.0848, + "step": 8535 + }, + { + "epoch": 0.9254119687771032, + "grad_norm": 3.58156418800354, + "learning_rate": 1.3880748585521544e-05, + "loss": 0.0201, + "step": 8536 + }, + { + "epoch": 0.925520381613183, + "grad_norm": 2.234769105911255, + "learning_rate": 1.3880023211954156e-05, + "loss": 0.0326, + "step": 8537 + }, + { + "epoch": 0.9256287944492628, + "grad_norm": 5.124780654907227, + "learning_rate": 1.387929783838677e-05, + "loss": 0.0523, + "step": 8538 + }, + { + "epoch": 0.9257372072853426, + "grad_norm": 3.2038705348968506, + "learning_rate": 1.3878572464819382e-05, + "loss": 0.0199, + "step": 8539 + }, + { + "epoch": 0.9258456201214224, + "grad_norm": 1.347934603691101, + "learning_rate": 1.3877847091251998e-05, + "loss": 0.0144, + "step": 8540 + }, + { + "epoch": 0.9259540329575021, + "grad_norm": 1.059982180595398, + "learning_rate": 1.387712171768461e-05, + "loss": 0.0164, + "step": 8541 + }, + { + "epoch": 0.926062445793582, + "grad_norm": 3.016829252243042, + "learning_rate": 1.3876396344117222e-05, + "loss": 0.0171, + "step": 8542 + }, + { + "epoch": 0.9261708586296618, + "grad_norm": 2.5528311729431152, + "learning_rate": 1.3875670970549836e-05, + "loss": 0.0273, + "step": 8543 + }, + { + "epoch": 0.9262792714657415, + "grad_norm": 2.2986464500427246, + "learning_rate": 1.3874945596982448e-05, + "loss": 0.0413, + "step": 8544 + }, + { + "epoch": 0.9263876843018214, + "grad_norm": 2.297376871109009, + "learning_rate": 1.387422022341506e-05, + "loss": 0.0444, + "step": 8545 + }, + { + "epoch": 0.9264960971379012, + "grad_norm": 0.7410364747047424, + "learning_rate": 1.3873494849847672e-05, + "loss": 0.0081, + "step": 8546 + }, + { + "epoch": 0.9266045099739809, + "grad_norm": 4.93111515045166, + "learning_rate": 1.3872769476280286e-05, + "loss": 0.0212, + "step": 8547 + }, + { + "epoch": 0.9267129228100607, + "grad_norm": 4.0666680335998535, + "learning_rate": 1.3872044102712898e-05, + "loss": 0.0689, + "step": 8548 + }, + { + "epoch": 0.9268213356461406, + "grad_norm": 2.2384722232818604, + "learning_rate": 1.387131872914551e-05, + "loss": 0.0161, + "step": 8549 + }, + { + "epoch": 0.9269297484822203, + "grad_norm": 2.2763254642486572, + "learning_rate": 1.3870593355578124e-05, + "loss": 0.0364, + "step": 8550 + }, + { + "epoch": 0.9270381613183001, + "grad_norm": 3.067729949951172, + "learning_rate": 1.3869867982010736e-05, + "loss": 0.0346, + "step": 8551 + }, + { + "epoch": 0.9271465741543798, + "grad_norm": 3.2523603439331055, + "learning_rate": 1.3869142608443348e-05, + "loss": 0.0484, + "step": 8552 + }, + { + "epoch": 0.9272549869904597, + "grad_norm": 2.4922521114349365, + "learning_rate": 1.3868417234875962e-05, + "loss": 0.0251, + "step": 8553 + }, + { + "epoch": 0.9273633998265395, + "grad_norm": 3.0084455013275146, + "learning_rate": 1.3867691861308574e-05, + "loss": 0.0274, + "step": 8554 + }, + { + "epoch": 0.9274718126626192, + "grad_norm": 1.4451777935028076, + "learning_rate": 1.3866966487741186e-05, + "loss": 0.0148, + "step": 8555 + }, + { + "epoch": 0.927580225498699, + "grad_norm": 2.8306992053985596, + "learning_rate": 1.38662411141738e-05, + "loss": 0.0297, + "step": 8556 + }, + { + "epoch": 0.9276886383347789, + "grad_norm": 3.86063551902771, + "learning_rate": 1.3865515740606414e-05, + "loss": 0.0683, + "step": 8557 + }, + { + "epoch": 0.9277970511708586, + "grad_norm": 0.4612574577331543, + "learning_rate": 1.3864790367039026e-05, + "loss": 0.0037, + "step": 8558 + }, + { + "epoch": 0.9279054640069384, + "grad_norm": 2.3339521884918213, + "learning_rate": 1.386406499347164e-05, + "loss": 0.032, + "step": 8559 + }, + { + "epoch": 0.9280138768430182, + "grad_norm": 4.128164768218994, + "learning_rate": 1.3863339619904252e-05, + "loss": 0.0587, + "step": 8560 + }, + { + "epoch": 0.928122289679098, + "grad_norm": 5.883138179779053, + "learning_rate": 1.3862614246336864e-05, + "loss": 0.0725, + "step": 8561 + }, + { + "epoch": 0.9282307025151778, + "grad_norm": 1.1442203521728516, + "learning_rate": 1.3861888872769478e-05, + "loss": 0.0135, + "step": 8562 + }, + { + "epoch": 0.9283391153512576, + "grad_norm": 1.1425457000732422, + "learning_rate": 1.386116349920209e-05, + "loss": 0.0087, + "step": 8563 + }, + { + "epoch": 0.9284475281873373, + "grad_norm": 1.6979331970214844, + "learning_rate": 1.3860438125634703e-05, + "loss": 0.0172, + "step": 8564 + }, + { + "epoch": 0.9285559410234172, + "grad_norm": 3.6668009757995605, + "learning_rate": 1.3859712752067316e-05, + "loss": 0.0295, + "step": 8565 + }, + { + "epoch": 0.928664353859497, + "grad_norm": 4.244620323181152, + "learning_rate": 1.3858987378499929e-05, + "loss": 0.078, + "step": 8566 + }, + { + "epoch": 0.9287727666955767, + "grad_norm": 5.346982955932617, + "learning_rate": 1.385826200493254e-05, + "loss": 0.051, + "step": 8567 + }, + { + "epoch": 0.9288811795316565, + "grad_norm": 1.332585096359253, + "learning_rate": 1.3857536631365155e-05, + "loss": 0.0377, + "step": 8568 + }, + { + "epoch": 0.9289895923677364, + "grad_norm": 2.6991660594940186, + "learning_rate": 1.3856811257797767e-05, + "loss": 0.063, + "step": 8569 + }, + { + "epoch": 0.9290980052038161, + "grad_norm": 4.764258861541748, + "learning_rate": 1.3856085884230379e-05, + "loss": 0.0601, + "step": 8570 + }, + { + "epoch": 0.9292064180398959, + "grad_norm": 2.0964066982269287, + "learning_rate": 1.3855360510662993e-05, + "loss": 0.0211, + "step": 8571 + }, + { + "epoch": 0.9293148308759757, + "grad_norm": 11.2772855758667, + "learning_rate": 1.3854635137095605e-05, + "loss": 0.0427, + "step": 8572 + }, + { + "epoch": 0.9294232437120555, + "grad_norm": 11.518863677978516, + "learning_rate": 1.3853909763528217e-05, + "loss": 0.0617, + "step": 8573 + }, + { + "epoch": 0.9295316565481353, + "grad_norm": 4.495183944702148, + "learning_rate": 1.3853184389960832e-05, + "loss": 0.0908, + "step": 8574 + }, + { + "epoch": 0.9296400693842151, + "grad_norm": 9.407407760620117, + "learning_rate": 1.3852459016393445e-05, + "loss": 0.0842, + "step": 8575 + }, + { + "epoch": 0.9297484822202949, + "grad_norm": 4.680434703826904, + "learning_rate": 1.3851733642826057e-05, + "loss": 0.0547, + "step": 8576 + }, + { + "epoch": 0.9298568950563747, + "grad_norm": 1.0200347900390625, + "learning_rate": 1.385100826925867e-05, + "loss": 0.0201, + "step": 8577 + }, + { + "epoch": 0.9299653078924545, + "grad_norm": 1.207375168800354, + "learning_rate": 1.3850282895691283e-05, + "loss": 0.0148, + "step": 8578 + }, + { + "epoch": 0.9300737207285342, + "grad_norm": 1.2470118999481201, + "learning_rate": 1.3849557522123895e-05, + "loss": 0.008, + "step": 8579 + }, + { + "epoch": 0.9301821335646141, + "grad_norm": 4.145768165588379, + "learning_rate": 1.3848832148556509e-05, + "loss": 0.0395, + "step": 8580 + }, + { + "epoch": 0.9302905464006939, + "grad_norm": 3.8287723064422607, + "learning_rate": 1.3848106774989121e-05, + "loss": 0.0438, + "step": 8581 + }, + { + "epoch": 0.9303989592367736, + "grad_norm": 3.783295154571533, + "learning_rate": 1.3847381401421733e-05, + "loss": 0.0552, + "step": 8582 + }, + { + "epoch": 0.9305073720728534, + "grad_norm": 0.9919529557228088, + "learning_rate": 1.3846656027854347e-05, + "loss": 0.0096, + "step": 8583 + }, + { + "epoch": 0.9306157849089333, + "grad_norm": 2.9193124771118164, + "learning_rate": 1.3845930654286959e-05, + "loss": 0.0823, + "step": 8584 + }, + { + "epoch": 0.930724197745013, + "grad_norm": 1.846256971359253, + "learning_rate": 1.3845205280719571e-05, + "loss": 0.0407, + "step": 8585 + }, + { + "epoch": 0.9308326105810928, + "grad_norm": 3.9566192626953125, + "learning_rate": 1.3844479907152183e-05, + "loss": 0.0141, + "step": 8586 + }, + { + "epoch": 0.9309410234171726, + "grad_norm": 2.215907335281372, + "learning_rate": 1.3843754533584797e-05, + "loss": 0.0214, + "step": 8587 + }, + { + "epoch": 0.9310494362532524, + "grad_norm": 2.568105459213257, + "learning_rate": 1.384302916001741e-05, + "loss": 0.0373, + "step": 8588 + }, + { + "epoch": 0.9311578490893322, + "grad_norm": 2.329993724822998, + "learning_rate": 1.3842303786450021e-05, + "loss": 0.0433, + "step": 8589 + }, + { + "epoch": 0.931266261925412, + "grad_norm": 1.5816622972488403, + "learning_rate": 1.3841578412882635e-05, + "loss": 0.0387, + "step": 8590 + }, + { + "epoch": 0.9313746747614917, + "grad_norm": 4.305905342102051, + "learning_rate": 1.3840853039315247e-05, + "loss": 0.0185, + "step": 8591 + }, + { + "epoch": 0.9314830875975716, + "grad_norm": 2.424520254135132, + "learning_rate": 1.3840127665747863e-05, + "loss": 0.0273, + "step": 8592 + }, + { + "epoch": 0.9315915004336514, + "grad_norm": 1.8384439945220947, + "learning_rate": 1.3839402292180475e-05, + "loss": 0.041, + "step": 8593 + }, + { + "epoch": 0.9316999132697311, + "grad_norm": 2.9513230323791504, + "learning_rate": 1.3838676918613087e-05, + "loss": 0.0269, + "step": 8594 + }, + { + "epoch": 0.9318083261058109, + "grad_norm": 1.925074815750122, + "learning_rate": 1.3837951545045701e-05, + "loss": 0.0233, + "step": 8595 + }, + { + "epoch": 0.9319167389418908, + "grad_norm": 1.4447447061538696, + "learning_rate": 1.3837226171478313e-05, + "loss": 0.0141, + "step": 8596 + }, + { + "epoch": 0.9320251517779705, + "grad_norm": 4.433757781982422, + "learning_rate": 1.3836500797910925e-05, + "loss": 0.0315, + "step": 8597 + }, + { + "epoch": 0.9321335646140503, + "grad_norm": 4.324197769165039, + "learning_rate": 1.3835775424343537e-05, + "loss": 0.0535, + "step": 8598 + }, + { + "epoch": 0.93224197745013, + "grad_norm": 1.1538876295089722, + "learning_rate": 1.3835050050776151e-05, + "loss": 0.0227, + "step": 8599 + }, + { + "epoch": 0.9323503902862099, + "grad_norm": 0.7205594182014465, + "learning_rate": 1.3834324677208763e-05, + "loss": 0.0154, + "step": 8600 + }, + { + "epoch": 0.9324588031222897, + "grad_norm": 3.731473207473755, + "learning_rate": 1.3833599303641376e-05, + "loss": 0.0549, + "step": 8601 + }, + { + "epoch": 0.9325672159583694, + "grad_norm": 1.8112514019012451, + "learning_rate": 1.383287393007399e-05, + "loss": 0.0323, + "step": 8602 + }, + { + "epoch": 0.9326756287944492, + "grad_norm": 1.5164012908935547, + "learning_rate": 1.3832148556506602e-05, + "loss": 0.0145, + "step": 8603 + }, + { + "epoch": 0.9327840416305291, + "grad_norm": 3.9600749015808105, + "learning_rate": 1.3831423182939214e-05, + "loss": 0.0363, + "step": 8604 + }, + { + "epoch": 0.9328924544666088, + "grad_norm": 0.8458036184310913, + "learning_rate": 1.3830697809371828e-05, + "loss": 0.0133, + "step": 8605 + }, + { + "epoch": 0.9330008673026886, + "grad_norm": 3.7169573307037354, + "learning_rate": 1.382997243580444e-05, + "loss": 0.0511, + "step": 8606 + }, + { + "epoch": 0.9331092801387685, + "grad_norm": 7.400829792022705, + "learning_rate": 1.3829247062237052e-05, + "loss": 0.0396, + "step": 8607 + }, + { + "epoch": 0.9332176929748482, + "grad_norm": 2.697540283203125, + "learning_rate": 1.3828521688669666e-05, + "loss": 0.0595, + "step": 8608 + }, + { + "epoch": 0.933326105810928, + "grad_norm": 2.767688035964966, + "learning_rate": 1.382779631510228e-05, + "loss": 0.0301, + "step": 8609 + }, + { + "epoch": 0.9334345186470078, + "grad_norm": 2.469000816345215, + "learning_rate": 1.3827070941534892e-05, + "loss": 0.0153, + "step": 8610 + }, + { + "epoch": 0.9335429314830876, + "grad_norm": 6.34702205657959, + "learning_rate": 1.3826345567967505e-05, + "loss": 0.0421, + "step": 8611 + }, + { + "epoch": 0.9336513443191674, + "grad_norm": 3.588881015777588, + "learning_rate": 1.3825620194400118e-05, + "loss": 0.0617, + "step": 8612 + }, + { + "epoch": 0.9337597571552472, + "grad_norm": 2.5135488510131836, + "learning_rate": 1.382489482083273e-05, + "loss": 0.0211, + "step": 8613 + }, + { + "epoch": 0.9338681699913269, + "grad_norm": 3.2300374507904053, + "learning_rate": 1.3824169447265344e-05, + "loss": 0.0563, + "step": 8614 + }, + { + "epoch": 0.9339765828274068, + "grad_norm": 2.1990575790405273, + "learning_rate": 1.3823444073697956e-05, + "loss": 0.0214, + "step": 8615 + }, + { + "epoch": 0.9340849956634866, + "grad_norm": 2.505262613296509, + "learning_rate": 1.3822718700130568e-05, + "loss": 0.0377, + "step": 8616 + }, + { + "epoch": 0.9341934084995663, + "grad_norm": 2.197030544281006, + "learning_rate": 1.3821993326563182e-05, + "loss": 0.0139, + "step": 8617 + }, + { + "epoch": 0.9343018213356461, + "grad_norm": 1.6092253923416138, + "learning_rate": 1.3821267952995794e-05, + "loss": 0.0252, + "step": 8618 + }, + { + "epoch": 0.934410234171726, + "grad_norm": 2.5790467262268066, + "learning_rate": 1.3820542579428406e-05, + "loss": 0.0337, + "step": 8619 + }, + { + "epoch": 0.9345186470078057, + "grad_norm": 2.1696014404296875, + "learning_rate": 1.381981720586102e-05, + "loss": 0.0315, + "step": 8620 + }, + { + "epoch": 0.9346270598438855, + "grad_norm": 4.443599700927734, + "learning_rate": 1.3819091832293632e-05, + "loss": 0.0299, + "step": 8621 + }, + { + "epoch": 0.9347354726799653, + "grad_norm": 2.3260912895202637, + "learning_rate": 1.3818366458726244e-05, + "loss": 0.0337, + "step": 8622 + }, + { + "epoch": 0.9348438855160451, + "grad_norm": 4.664198875427246, + "learning_rate": 1.3817641085158858e-05, + "loss": 0.0657, + "step": 8623 + }, + { + "epoch": 0.9349522983521249, + "grad_norm": 2.3706552982330322, + "learning_rate": 1.381691571159147e-05, + "loss": 0.0136, + "step": 8624 + }, + { + "epoch": 0.9350607111882047, + "grad_norm": 1.7036980390548706, + "learning_rate": 1.3816190338024082e-05, + "loss": 0.0138, + "step": 8625 + }, + { + "epoch": 0.9351691240242844, + "grad_norm": 2.872187852859497, + "learning_rate": 1.3815464964456698e-05, + "loss": 0.0237, + "step": 8626 + }, + { + "epoch": 0.9352775368603643, + "grad_norm": 2.239619016647339, + "learning_rate": 1.381473959088931e-05, + "loss": 0.0396, + "step": 8627 + }, + { + "epoch": 0.9353859496964441, + "grad_norm": 1.3765431642532349, + "learning_rate": 1.3814014217321922e-05, + "loss": 0.0145, + "step": 8628 + }, + { + "epoch": 0.9354943625325238, + "grad_norm": 2.4817025661468506, + "learning_rate": 1.3813288843754536e-05, + "loss": 0.029, + "step": 8629 + }, + { + "epoch": 0.9356027753686036, + "grad_norm": 1.0484141111373901, + "learning_rate": 1.3812563470187148e-05, + "loss": 0.0146, + "step": 8630 + }, + { + "epoch": 0.9357111882046835, + "grad_norm": 2.3375303745269775, + "learning_rate": 1.381183809661976e-05, + "loss": 0.0185, + "step": 8631 + }, + { + "epoch": 0.9358196010407632, + "grad_norm": 7.1971940994262695, + "learning_rate": 1.3811112723052374e-05, + "loss": 0.072, + "step": 8632 + }, + { + "epoch": 0.935928013876843, + "grad_norm": 6.046617031097412, + "learning_rate": 1.3810387349484986e-05, + "loss": 0.0588, + "step": 8633 + }, + { + "epoch": 0.9360364267129229, + "grad_norm": 3.065606117248535, + "learning_rate": 1.3809661975917598e-05, + "loss": 0.0345, + "step": 8634 + }, + { + "epoch": 0.9361448395490026, + "grad_norm": 2.8507816791534424, + "learning_rate": 1.3808936602350212e-05, + "loss": 0.0334, + "step": 8635 + }, + { + "epoch": 0.9362532523850824, + "grad_norm": 3.9796464443206787, + "learning_rate": 1.3808211228782824e-05, + "loss": 0.0394, + "step": 8636 + }, + { + "epoch": 0.9363616652211622, + "grad_norm": 1.2220994234085083, + "learning_rate": 1.3807485855215436e-05, + "loss": 0.0061, + "step": 8637 + }, + { + "epoch": 0.936470078057242, + "grad_norm": 6.0984578132629395, + "learning_rate": 1.3806760481648049e-05, + "loss": 0.0407, + "step": 8638 + }, + { + "epoch": 0.9365784908933218, + "grad_norm": 3.448197841644287, + "learning_rate": 1.3806035108080662e-05, + "loss": 0.0728, + "step": 8639 + }, + { + "epoch": 0.9366869037294016, + "grad_norm": 5.182494640350342, + "learning_rate": 1.3805309734513275e-05, + "loss": 0.0743, + "step": 8640 + }, + { + "epoch": 0.9367953165654813, + "grad_norm": 1.9862041473388672, + "learning_rate": 1.3804584360945887e-05, + "loss": 0.0194, + "step": 8641 + }, + { + "epoch": 0.9369037294015612, + "grad_norm": 1.548352837562561, + "learning_rate": 1.38038589873785e-05, + "loss": 0.0288, + "step": 8642 + }, + { + "epoch": 0.937012142237641, + "grad_norm": 3.5409648418426514, + "learning_rate": 1.3803133613811114e-05, + "loss": 0.0434, + "step": 8643 + }, + { + "epoch": 0.9371205550737207, + "grad_norm": 4.408899307250977, + "learning_rate": 1.3802408240243728e-05, + "loss": 0.0185, + "step": 8644 + }, + { + "epoch": 0.9372289679098005, + "grad_norm": 3.690617322921753, + "learning_rate": 1.380168286667634e-05, + "loss": 0.0773, + "step": 8645 + }, + { + "epoch": 0.9373373807458804, + "grad_norm": 1.742215633392334, + "learning_rate": 1.3800957493108953e-05, + "loss": 0.0499, + "step": 8646 + }, + { + "epoch": 0.9374457935819601, + "grad_norm": 1.589892864227295, + "learning_rate": 1.3800232119541566e-05, + "loss": 0.0217, + "step": 8647 + }, + { + "epoch": 0.9375542064180399, + "grad_norm": 3.1525015830993652, + "learning_rate": 1.3799506745974179e-05, + "loss": 0.056, + "step": 8648 + }, + { + "epoch": 0.9376626192541196, + "grad_norm": 2.6854093074798584, + "learning_rate": 1.379878137240679e-05, + "loss": 0.0239, + "step": 8649 + }, + { + "epoch": 0.9377710320901995, + "grad_norm": 3.96545147895813, + "learning_rate": 1.3798055998839403e-05, + "loss": 0.0412, + "step": 8650 + }, + { + "epoch": 0.9378794449262793, + "grad_norm": 3.3627517223358154, + "learning_rate": 1.3797330625272017e-05, + "loss": 0.0624, + "step": 8651 + }, + { + "epoch": 0.937987857762359, + "grad_norm": 2.4829659461975098, + "learning_rate": 1.3796605251704629e-05, + "loss": 0.045, + "step": 8652 + }, + { + "epoch": 0.9380962705984388, + "grad_norm": 2.9079713821411133, + "learning_rate": 1.3795879878137241e-05, + "loss": 0.0776, + "step": 8653 + }, + { + "epoch": 0.9382046834345187, + "grad_norm": 4.2344255447387695, + "learning_rate": 1.3795154504569855e-05, + "loss": 0.0449, + "step": 8654 + }, + { + "epoch": 0.9383130962705984, + "grad_norm": 2.953770637512207, + "learning_rate": 1.3794429131002467e-05, + "loss": 0.0631, + "step": 8655 + }, + { + "epoch": 0.9384215091066782, + "grad_norm": 0.49131178855895996, + "learning_rate": 1.3793703757435079e-05, + "loss": 0.0067, + "step": 8656 + }, + { + "epoch": 0.938529921942758, + "grad_norm": 1.208130121231079, + "learning_rate": 1.3792978383867693e-05, + "loss": 0.0292, + "step": 8657 + }, + { + "epoch": 0.9386383347788378, + "grad_norm": 4.879807472229004, + "learning_rate": 1.3792253010300305e-05, + "loss": 0.0465, + "step": 8658 + }, + { + "epoch": 0.9387467476149176, + "grad_norm": 3.2110745906829834, + "learning_rate": 1.3791527636732917e-05, + "loss": 0.012, + "step": 8659 + }, + { + "epoch": 0.9388551604509974, + "grad_norm": 2.7101590633392334, + "learning_rate": 1.3790802263165531e-05, + "loss": 0.0346, + "step": 8660 + }, + { + "epoch": 0.9389635732870771, + "grad_norm": 1.8438022136688232, + "learning_rate": 1.3790076889598145e-05, + "loss": 0.0181, + "step": 8661 + }, + { + "epoch": 0.939071986123157, + "grad_norm": 1.1457862854003906, + "learning_rate": 1.3789351516030759e-05, + "loss": 0.0131, + "step": 8662 + }, + { + "epoch": 0.9391803989592368, + "grad_norm": 0.8550861477851868, + "learning_rate": 1.378862614246337e-05, + "loss": 0.0127, + "step": 8663 + }, + { + "epoch": 0.9392888117953165, + "grad_norm": 3.5854814052581787, + "learning_rate": 1.3787900768895983e-05, + "loss": 0.0803, + "step": 8664 + }, + { + "epoch": 0.9393972246313964, + "grad_norm": 5.500837326049805, + "learning_rate": 1.3787175395328595e-05, + "loss": 0.0409, + "step": 8665 + }, + { + "epoch": 0.9395056374674762, + "grad_norm": 0.5856916308403015, + "learning_rate": 1.3786450021761209e-05, + "loss": 0.0065, + "step": 8666 + }, + { + "epoch": 0.9396140503035559, + "grad_norm": 2.1580100059509277, + "learning_rate": 1.3785724648193821e-05, + "loss": 0.036, + "step": 8667 + }, + { + "epoch": 0.9397224631396357, + "grad_norm": 1.4886258840560913, + "learning_rate": 1.3784999274626433e-05, + "loss": 0.0179, + "step": 8668 + }, + { + "epoch": 0.9398308759757156, + "grad_norm": 0.6828023791313171, + "learning_rate": 1.3784273901059047e-05, + "loss": 0.007, + "step": 8669 + }, + { + "epoch": 0.9399392888117953, + "grad_norm": 0.9334819912910461, + "learning_rate": 1.378354852749166e-05, + "loss": 0.014, + "step": 8670 + }, + { + "epoch": 0.9400477016478751, + "grad_norm": 2.125150203704834, + "learning_rate": 1.3782823153924271e-05, + "loss": 0.0265, + "step": 8671 + }, + { + "epoch": 0.9401561144839549, + "grad_norm": 2.996809959411621, + "learning_rate": 1.3782097780356885e-05, + "loss": 0.0321, + "step": 8672 + }, + { + "epoch": 0.9402645273200347, + "grad_norm": 2.4085233211517334, + "learning_rate": 1.3781372406789497e-05, + "loss": 0.0144, + "step": 8673 + }, + { + "epoch": 0.9403729401561145, + "grad_norm": 5.369088172912598, + "learning_rate": 1.378064703322211e-05, + "loss": 0.0779, + "step": 8674 + }, + { + "epoch": 0.9404813529921943, + "grad_norm": 11.783510208129883, + "learning_rate": 1.3779921659654723e-05, + "loss": 0.0339, + "step": 8675 + }, + { + "epoch": 0.940589765828274, + "grad_norm": 4.220808029174805, + "learning_rate": 1.3779196286087335e-05, + "loss": 0.0483, + "step": 8676 + }, + { + "epoch": 0.9406981786643539, + "grad_norm": 6.404341220855713, + "learning_rate": 1.3778470912519948e-05, + "loss": 0.0536, + "step": 8677 + }, + { + "epoch": 0.9408065915004337, + "grad_norm": 4.394832611083984, + "learning_rate": 1.3777745538952563e-05, + "loss": 0.0477, + "step": 8678 + }, + { + "epoch": 0.9409150043365134, + "grad_norm": 4.348500728607178, + "learning_rate": 1.3777020165385175e-05, + "loss": 0.0537, + "step": 8679 + }, + { + "epoch": 0.9410234171725932, + "grad_norm": 7.643219470977783, + "learning_rate": 1.3776294791817787e-05, + "loss": 0.0684, + "step": 8680 + }, + { + "epoch": 0.9411318300086731, + "grad_norm": 3.3667397499084473, + "learning_rate": 1.3775569418250401e-05, + "loss": 0.0512, + "step": 8681 + }, + { + "epoch": 0.9412402428447528, + "grad_norm": 4.264876842498779, + "learning_rate": 1.3774844044683013e-05, + "loss": 0.0552, + "step": 8682 + }, + { + "epoch": 0.9413486556808326, + "grad_norm": 2.4247872829437256, + "learning_rate": 1.3774118671115626e-05, + "loss": 0.0409, + "step": 8683 + }, + { + "epoch": 0.9414570685169124, + "grad_norm": 3.756927490234375, + "learning_rate": 1.377339329754824e-05, + "loss": 0.0758, + "step": 8684 + }, + { + "epoch": 0.9415654813529922, + "grad_norm": 1.4405584335327148, + "learning_rate": 1.3772667923980852e-05, + "loss": 0.0142, + "step": 8685 + }, + { + "epoch": 0.941673894189072, + "grad_norm": 4.616700649261475, + "learning_rate": 1.3771942550413464e-05, + "loss": 0.0743, + "step": 8686 + }, + { + "epoch": 0.9417823070251518, + "grad_norm": 3.1854331493377686, + "learning_rate": 1.3771217176846078e-05, + "loss": 0.0303, + "step": 8687 + }, + { + "epoch": 0.9418907198612315, + "grad_norm": 1.6011382341384888, + "learning_rate": 1.377049180327869e-05, + "loss": 0.0223, + "step": 8688 + }, + { + "epoch": 0.9419991326973114, + "grad_norm": 1.1783686876296997, + "learning_rate": 1.3769766429711302e-05, + "loss": 0.0295, + "step": 8689 + }, + { + "epoch": 0.9421075455333912, + "grad_norm": 13.060320854187012, + "learning_rate": 1.3769041056143914e-05, + "loss": 0.0712, + "step": 8690 + }, + { + "epoch": 0.9422159583694709, + "grad_norm": 3.3373165130615234, + "learning_rate": 1.3768315682576528e-05, + "loss": 0.0195, + "step": 8691 + }, + { + "epoch": 0.9423243712055508, + "grad_norm": 1.9401984214782715, + "learning_rate": 1.376759030900914e-05, + "loss": 0.0449, + "step": 8692 + }, + { + "epoch": 0.9424327840416306, + "grad_norm": 2.333695888519287, + "learning_rate": 1.3766864935441752e-05, + "loss": 0.0315, + "step": 8693 + }, + { + "epoch": 0.9425411968777103, + "grad_norm": 2.7442421913146973, + "learning_rate": 1.3766139561874366e-05, + "loss": 0.039, + "step": 8694 + }, + { + "epoch": 0.9426496097137901, + "grad_norm": 1.2892529964447021, + "learning_rate": 1.376541418830698e-05, + "loss": 0.0265, + "step": 8695 + }, + { + "epoch": 0.94275802254987, + "grad_norm": 2.467742919921875, + "learning_rate": 1.3764688814739594e-05, + "loss": 0.0379, + "step": 8696 + }, + { + "epoch": 0.9428664353859497, + "grad_norm": 2.222163438796997, + "learning_rate": 1.3763963441172206e-05, + "loss": 0.029, + "step": 8697 + }, + { + "epoch": 0.9429748482220295, + "grad_norm": 5.572494029998779, + "learning_rate": 1.3763238067604818e-05, + "loss": 0.02, + "step": 8698 + }, + { + "epoch": 0.9430832610581092, + "grad_norm": 6.926926136016846, + "learning_rate": 1.3762512694037432e-05, + "loss": 0.0836, + "step": 8699 + }, + { + "epoch": 0.9431916738941891, + "grad_norm": 1.2604461908340454, + "learning_rate": 1.3761787320470044e-05, + "loss": 0.0243, + "step": 8700 + }, + { + "epoch": 0.9433000867302689, + "grad_norm": 3.0656216144561768, + "learning_rate": 1.3761061946902656e-05, + "loss": 0.0513, + "step": 8701 + }, + { + "epoch": 0.9434084995663486, + "grad_norm": 3.557474136352539, + "learning_rate": 1.3760336573335268e-05, + "loss": 0.0227, + "step": 8702 + }, + { + "epoch": 0.9435169124024284, + "grad_norm": 2.7854995727539062, + "learning_rate": 1.3759611199767882e-05, + "loss": 0.0261, + "step": 8703 + }, + { + "epoch": 0.9436253252385083, + "grad_norm": 3.5218122005462646, + "learning_rate": 1.3758885826200494e-05, + "loss": 0.0363, + "step": 8704 + }, + { + "epoch": 0.943733738074588, + "grad_norm": 0.8250637054443359, + "learning_rate": 1.3758160452633106e-05, + "loss": 0.0167, + "step": 8705 + }, + { + "epoch": 0.9438421509106678, + "grad_norm": 0.6618899703025818, + "learning_rate": 1.375743507906572e-05, + "loss": 0.0061, + "step": 8706 + }, + { + "epoch": 0.9439505637467476, + "grad_norm": 4.289603233337402, + "learning_rate": 1.3756709705498332e-05, + "loss": 0.0685, + "step": 8707 + }, + { + "epoch": 0.9440589765828274, + "grad_norm": 2.8098015785217285, + "learning_rate": 1.3755984331930944e-05, + "loss": 0.0558, + "step": 8708 + }, + { + "epoch": 0.9441673894189072, + "grad_norm": 1.5383641719818115, + "learning_rate": 1.3755258958363558e-05, + "loss": 0.0058, + "step": 8709 + }, + { + "epoch": 0.944275802254987, + "grad_norm": 2.87109375, + "learning_rate": 1.375453358479617e-05, + "loss": 0.0409, + "step": 8710 + }, + { + "epoch": 0.9443842150910667, + "grad_norm": 6.598599910736084, + "learning_rate": 1.3753808211228783e-05, + "loss": 0.0786, + "step": 8711 + }, + { + "epoch": 0.9444926279271466, + "grad_norm": 2.021245241165161, + "learning_rate": 1.3753082837661398e-05, + "loss": 0.0194, + "step": 8712 + }, + { + "epoch": 0.9446010407632264, + "grad_norm": 2.11641001701355, + "learning_rate": 1.375235746409401e-05, + "loss": 0.0147, + "step": 8713 + }, + { + "epoch": 0.9447094535993061, + "grad_norm": 4.208502769470215, + "learning_rate": 1.3751632090526624e-05, + "loss": 0.068, + "step": 8714 + }, + { + "epoch": 0.9448178664353859, + "grad_norm": 1.1199215650558472, + "learning_rate": 1.3750906716959236e-05, + "loss": 0.0183, + "step": 8715 + }, + { + "epoch": 0.9449262792714658, + "grad_norm": 2.0982348918914795, + "learning_rate": 1.3750181343391848e-05, + "loss": 0.0224, + "step": 8716 + }, + { + "epoch": 0.9450346921075455, + "grad_norm": 2.5238730907440186, + "learning_rate": 1.374945596982446e-05, + "loss": 0.0391, + "step": 8717 + }, + { + "epoch": 0.9451431049436253, + "grad_norm": 5.100780487060547, + "learning_rate": 1.3748730596257074e-05, + "loss": 0.0357, + "step": 8718 + }, + { + "epoch": 0.9452515177797051, + "grad_norm": 2.679482936859131, + "learning_rate": 1.3748005222689686e-05, + "loss": 0.026, + "step": 8719 + }, + { + "epoch": 0.9453599306157849, + "grad_norm": 3.6109044551849365, + "learning_rate": 1.3747279849122299e-05, + "loss": 0.0214, + "step": 8720 + }, + { + "epoch": 0.9454683434518647, + "grad_norm": 1.455071210861206, + "learning_rate": 1.3746554475554912e-05, + "loss": 0.0154, + "step": 8721 + }, + { + "epoch": 0.9455767562879445, + "grad_norm": 0.7192318439483643, + "learning_rate": 1.3745829101987525e-05, + "loss": 0.0081, + "step": 8722 + }, + { + "epoch": 0.9456851691240243, + "grad_norm": 2.516632556915283, + "learning_rate": 1.3745103728420137e-05, + "loss": 0.0512, + "step": 8723 + }, + { + "epoch": 0.9457935819601041, + "grad_norm": 1.597502589225769, + "learning_rate": 1.374437835485275e-05, + "loss": 0.0228, + "step": 8724 + }, + { + "epoch": 0.9459019947961839, + "grad_norm": 3.038886308670044, + "learning_rate": 1.3743652981285363e-05, + "loss": 0.027, + "step": 8725 + }, + { + "epoch": 0.9460104076322636, + "grad_norm": 1.1820393800735474, + "learning_rate": 1.3742927607717975e-05, + "loss": 0.02, + "step": 8726 + }, + { + "epoch": 0.9461188204683435, + "grad_norm": 0.5173026323318481, + "learning_rate": 1.3742202234150589e-05, + "loss": 0.0115, + "step": 8727 + }, + { + "epoch": 0.9462272333044233, + "grad_norm": 4.084953308105469, + "learning_rate": 1.37414768605832e-05, + "loss": 0.0134, + "step": 8728 + }, + { + "epoch": 0.946335646140503, + "grad_norm": 4.3158721923828125, + "learning_rate": 1.3740751487015813e-05, + "loss": 0.0735, + "step": 8729 + }, + { + "epoch": 0.9464440589765828, + "grad_norm": 4.143754482269287, + "learning_rate": 1.3740026113448428e-05, + "loss": 0.0545, + "step": 8730 + }, + { + "epoch": 0.9465524718126627, + "grad_norm": 0.37706372141838074, + "learning_rate": 1.373930073988104e-05, + "loss": 0.0031, + "step": 8731 + }, + { + "epoch": 0.9466608846487424, + "grad_norm": 2.160548210144043, + "learning_rate": 1.3738575366313653e-05, + "loss": 0.0359, + "step": 8732 + }, + { + "epoch": 0.9467692974848222, + "grad_norm": 2.371265172958374, + "learning_rate": 1.3737849992746267e-05, + "loss": 0.0423, + "step": 8733 + }, + { + "epoch": 0.946877710320902, + "grad_norm": 0.7921820282936096, + "learning_rate": 1.3737124619178879e-05, + "loss": 0.0249, + "step": 8734 + }, + { + "epoch": 0.9469861231569818, + "grad_norm": 4.374550819396973, + "learning_rate": 1.3736399245611491e-05, + "loss": 0.0641, + "step": 8735 + }, + { + "epoch": 0.9470945359930616, + "grad_norm": 2.374562978744507, + "learning_rate": 1.3735673872044105e-05, + "loss": 0.0253, + "step": 8736 + }, + { + "epoch": 0.9472029488291414, + "grad_norm": 2.854729652404785, + "learning_rate": 1.3734948498476717e-05, + "loss": 0.0326, + "step": 8737 + }, + { + "epoch": 0.9473113616652211, + "grad_norm": 1.8157366514205933, + "learning_rate": 1.3734223124909329e-05, + "loss": 0.0346, + "step": 8738 + }, + { + "epoch": 0.947419774501301, + "grad_norm": 5.442349910736084, + "learning_rate": 1.3733497751341943e-05, + "loss": 0.0915, + "step": 8739 + }, + { + "epoch": 0.9475281873373808, + "grad_norm": 4.239284515380859, + "learning_rate": 1.3732772377774555e-05, + "loss": 0.0807, + "step": 8740 + }, + { + "epoch": 0.9476366001734605, + "grad_norm": 5.029367446899414, + "learning_rate": 1.3732047004207167e-05, + "loss": 0.0482, + "step": 8741 + }, + { + "epoch": 0.9477450130095403, + "grad_norm": 4.373657703399658, + "learning_rate": 1.373132163063978e-05, + "loss": 0.0472, + "step": 8742 + }, + { + "epoch": 0.9478534258456202, + "grad_norm": 5.742046356201172, + "learning_rate": 1.3730596257072393e-05, + "loss": 0.0824, + "step": 8743 + }, + { + "epoch": 0.9479618386816999, + "grad_norm": 6.073207378387451, + "learning_rate": 1.3729870883505005e-05, + "loss": 0.071, + "step": 8744 + }, + { + "epoch": 0.9480702515177797, + "grad_norm": 3.3062021732330322, + "learning_rate": 1.3729145509937617e-05, + "loss": 0.0583, + "step": 8745 + }, + { + "epoch": 0.9481786643538594, + "grad_norm": 2.439680576324463, + "learning_rate": 1.3728420136370231e-05, + "loss": 0.0528, + "step": 8746 + }, + { + "epoch": 0.9482870771899393, + "grad_norm": 7.750209331512451, + "learning_rate": 1.3727694762802845e-05, + "loss": 0.066, + "step": 8747 + }, + { + "epoch": 0.9483954900260191, + "grad_norm": 4.429505348205566, + "learning_rate": 1.3726969389235459e-05, + "loss": 0.0729, + "step": 8748 + }, + { + "epoch": 0.9485039028620988, + "grad_norm": 2.326303720474243, + "learning_rate": 1.3726244015668071e-05, + "loss": 0.0474, + "step": 8749 + }, + { + "epoch": 0.9486123156981786, + "grad_norm": 2.1385884284973145, + "learning_rate": 1.3725518642100683e-05, + "loss": 0.0161, + "step": 8750 + }, + { + "epoch": 0.9487207285342585, + "grad_norm": 2.5577731132507324, + "learning_rate": 1.3724793268533297e-05, + "loss": 0.052, + "step": 8751 + }, + { + "epoch": 0.9488291413703382, + "grad_norm": 1.418915867805481, + "learning_rate": 1.372406789496591e-05, + "loss": 0.0246, + "step": 8752 + }, + { + "epoch": 0.948937554206418, + "grad_norm": 1.8836743831634521, + "learning_rate": 1.3723342521398521e-05, + "loss": 0.0324, + "step": 8753 + }, + { + "epoch": 0.9490459670424979, + "grad_norm": 1.2911263704299927, + "learning_rate": 1.3722617147831133e-05, + "loss": 0.0194, + "step": 8754 + }, + { + "epoch": 0.9491543798785776, + "grad_norm": 3.3495841026306152, + "learning_rate": 1.3721891774263747e-05, + "loss": 0.0336, + "step": 8755 + }, + { + "epoch": 0.9492627927146574, + "grad_norm": 2.604976177215576, + "learning_rate": 1.372116640069636e-05, + "loss": 0.0457, + "step": 8756 + }, + { + "epoch": 0.9493712055507372, + "grad_norm": 1.3462547063827515, + "learning_rate": 1.3720441027128972e-05, + "loss": 0.0153, + "step": 8757 + }, + { + "epoch": 0.949479618386817, + "grad_norm": 2.391394853591919, + "learning_rate": 1.3719715653561585e-05, + "loss": 0.0462, + "step": 8758 + }, + { + "epoch": 0.9495880312228968, + "grad_norm": 1.3941394090652466, + "learning_rate": 1.3718990279994198e-05, + "loss": 0.0173, + "step": 8759 + }, + { + "epoch": 0.9496964440589766, + "grad_norm": 2.8001315593719482, + "learning_rate": 1.371826490642681e-05, + "loss": 0.0288, + "step": 8760 + }, + { + "epoch": 0.9498048568950563, + "grad_norm": 4.491488456726074, + "learning_rate": 1.3717539532859424e-05, + "loss": 0.0306, + "step": 8761 + }, + { + "epoch": 0.9499132697311362, + "grad_norm": 1.1155809164047241, + "learning_rate": 1.3716814159292036e-05, + "loss": 0.0121, + "step": 8762 + }, + { + "epoch": 0.950021682567216, + "grad_norm": 2.173912763595581, + "learning_rate": 1.3716088785724648e-05, + "loss": 0.0335, + "step": 8763 + }, + { + "epoch": 0.9501300954032957, + "grad_norm": 2.084991216659546, + "learning_rate": 1.3715363412157263e-05, + "loss": 0.0246, + "step": 8764 + }, + { + "epoch": 0.9502385082393755, + "grad_norm": 2.625124216079712, + "learning_rate": 1.3714638038589876e-05, + "loss": 0.0597, + "step": 8765 + }, + { + "epoch": 0.9503469210754554, + "grad_norm": 1.4882903099060059, + "learning_rate": 1.371391266502249e-05, + "loss": 0.0151, + "step": 8766 + }, + { + "epoch": 0.9504553339115351, + "grad_norm": 1.2398340702056885, + "learning_rate": 1.3713187291455101e-05, + "loss": 0.0196, + "step": 8767 + }, + { + "epoch": 0.9505637467476149, + "grad_norm": 1.8697466850280762, + "learning_rate": 1.3712461917887714e-05, + "loss": 0.0416, + "step": 8768 + }, + { + "epoch": 0.9506721595836947, + "grad_norm": 2.4413671493530273, + "learning_rate": 1.3711736544320326e-05, + "loss": 0.018, + "step": 8769 + }, + { + "epoch": 0.9507805724197745, + "grad_norm": 2.9759392738342285, + "learning_rate": 1.371101117075294e-05, + "loss": 0.0418, + "step": 8770 + }, + { + "epoch": 0.9508889852558543, + "grad_norm": 0.5930295586585999, + "learning_rate": 1.3710285797185552e-05, + "loss": 0.0083, + "step": 8771 + }, + { + "epoch": 0.9509973980919341, + "grad_norm": 2.8033618927001953, + "learning_rate": 1.3709560423618164e-05, + "loss": 0.0222, + "step": 8772 + }, + { + "epoch": 0.9511058109280138, + "grad_norm": 4.86038064956665, + "learning_rate": 1.3708835050050778e-05, + "loss": 0.086, + "step": 8773 + }, + { + "epoch": 0.9512142237640937, + "grad_norm": 3.514348268508911, + "learning_rate": 1.370810967648339e-05, + "loss": 0.0513, + "step": 8774 + }, + { + "epoch": 0.9513226366001735, + "grad_norm": 2.6551918983459473, + "learning_rate": 1.3707384302916002e-05, + "loss": 0.0308, + "step": 8775 + }, + { + "epoch": 0.9514310494362532, + "grad_norm": 3.8023412227630615, + "learning_rate": 1.3706658929348616e-05, + "loss": 0.0165, + "step": 8776 + }, + { + "epoch": 0.951539462272333, + "grad_norm": 0.6298784613609314, + "learning_rate": 1.3705933555781228e-05, + "loss": 0.0074, + "step": 8777 + }, + { + "epoch": 0.9516478751084129, + "grad_norm": 3.857577323913574, + "learning_rate": 1.370520818221384e-05, + "loss": 0.0588, + "step": 8778 + }, + { + "epoch": 0.9517562879444926, + "grad_norm": 2.1985583305358887, + "learning_rate": 1.3704482808646454e-05, + "loss": 0.0558, + "step": 8779 + }, + { + "epoch": 0.9518647007805724, + "grad_norm": 0.8554001450538635, + "learning_rate": 1.3703757435079066e-05, + "loss": 0.0046, + "step": 8780 + }, + { + "epoch": 0.9519731136166523, + "grad_norm": 2.799971580505371, + "learning_rate": 1.3703032061511678e-05, + "loss": 0.044, + "step": 8781 + }, + { + "epoch": 0.952081526452732, + "grad_norm": 2.129002332687378, + "learning_rate": 1.3702306687944294e-05, + "loss": 0.0221, + "step": 8782 + }, + { + "epoch": 0.9521899392888118, + "grad_norm": 1.8250370025634766, + "learning_rate": 1.3701581314376906e-05, + "loss": 0.0463, + "step": 8783 + }, + { + "epoch": 0.9522983521248916, + "grad_norm": 1.162560224533081, + "learning_rate": 1.3700855940809518e-05, + "loss": 0.0106, + "step": 8784 + }, + { + "epoch": 0.9524067649609714, + "grad_norm": 5.445535659790039, + "learning_rate": 1.3700130567242132e-05, + "loss": 0.0707, + "step": 8785 + }, + { + "epoch": 0.9525151777970512, + "grad_norm": 2.325725555419922, + "learning_rate": 1.3699405193674744e-05, + "loss": 0.0354, + "step": 8786 + }, + { + "epoch": 0.952623590633131, + "grad_norm": 2.706845998764038, + "learning_rate": 1.3698679820107356e-05, + "loss": 0.0272, + "step": 8787 + }, + { + "epoch": 0.9527320034692107, + "grad_norm": 0.6528693437576294, + "learning_rate": 1.369795444653997e-05, + "loss": 0.0047, + "step": 8788 + }, + { + "epoch": 0.9528404163052906, + "grad_norm": 4.0993547439575195, + "learning_rate": 1.3697229072972582e-05, + "loss": 0.0564, + "step": 8789 + }, + { + "epoch": 0.9529488291413704, + "grad_norm": 1.095982313156128, + "learning_rate": 1.3696503699405194e-05, + "loss": 0.019, + "step": 8790 + }, + { + "epoch": 0.9530572419774501, + "grad_norm": 1.8680896759033203, + "learning_rate": 1.3695778325837808e-05, + "loss": 0.0425, + "step": 8791 + }, + { + "epoch": 0.9531656548135299, + "grad_norm": 5.358338356018066, + "learning_rate": 1.369505295227042e-05, + "loss": 0.0486, + "step": 8792 + }, + { + "epoch": 0.9532740676496098, + "grad_norm": 1.5953505039215088, + "learning_rate": 1.3694327578703032e-05, + "loss": 0.0116, + "step": 8793 + }, + { + "epoch": 0.9533824804856895, + "grad_norm": 1.2300220727920532, + "learning_rate": 1.3693602205135645e-05, + "loss": 0.0526, + "step": 8794 + }, + { + "epoch": 0.9534908933217693, + "grad_norm": 1.8456320762634277, + "learning_rate": 1.3692876831568258e-05, + "loss": 0.0274, + "step": 8795 + }, + { + "epoch": 0.953599306157849, + "grad_norm": 1.1919381618499756, + "learning_rate": 1.369215145800087e-05, + "loss": 0.0136, + "step": 8796 + }, + { + "epoch": 0.9537077189939289, + "grad_norm": 3.2162349224090576, + "learning_rate": 1.3691426084433483e-05, + "loss": 0.0613, + "step": 8797 + }, + { + "epoch": 0.9538161318300087, + "grad_norm": 3.5163681507110596, + "learning_rate": 1.3690700710866097e-05, + "loss": 0.0551, + "step": 8798 + }, + { + "epoch": 0.9539245446660884, + "grad_norm": 4.91391658782959, + "learning_rate": 1.368997533729871e-05, + "loss": 0.0268, + "step": 8799 + }, + { + "epoch": 0.9540329575021682, + "grad_norm": 1.4208767414093018, + "learning_rate": 1.3689249963731324e-05, + "loss": 0.0299, + "step": 8800 + }, + { + "epoch": 0.9541413703382481, + "grad_norm": 2.8005850315093994, + "learning_rate": 1.3688524590163936e-05, + "loss": 0.0583, + "step": 8801 + }, + { + "epoch": 0.9542497831743278, + "grad_norm": 2.6698966026306152, + "learning_rate": 1.3687799216596549e-05, + "loss": 0.0294, + "step": 8802 + }, + { + "epoch": 0.9543581960104076, + "grad_norm": 2.2200076580047607, + "learning_rate": 1.3687073843029162e-05, + "loss": 0.0256, + "step": 8803 + }, + { + "epoch": 0.9544666088464874, + "grad_norm": 0.6191549301147461, + "learning_rate": 1.3686348469461775e-05, + "loss": 0.0146, + "step": 8804 + }, + { + "epoch": 0.9545750216825672, + "grad_norm": 1.9056721925735474, + "learning_rate": 1.3685623095894387e-05, + "loss": 0.0365, + "step": 8805 + }, + { + "epoch": 0.954683434518647, + "grad_norm": 4.03994083404541, + "learning_rate": 1.3684897722326999e-05, + "loss": 0.0459, + "step": 8806 + }, + { + "epoch": 0.9547918473547268, + "grad_norm": 1.5168955326080322, + "learning_rate": 1.3684172348759613e-05, + "loss": 0.018, + "step": 8807 + }, + { + "epoch": 0.9549002601908065, + "grad_norm": 2.3176684379577637, + "learning_rate": 1.3683446975192225e-05, + "loss": 0.0312, + "step": 8808 + }, + { + "epoch": 0.9550086730268864, + "grad_norm": 6.924596309661865, + "learning_rate": 1.3682721601624837e-05, + "loss": 0.0224, + "step": 8809 + }, + { + "epoch": 0.9551170858629662, + "grad_norm": 0.482411652803421, + "learning_rate": 1.368199622805745e-05, + "loss": 0.0043, + "step": 8810 + }, + { + "epoch": 0.9552254986990459, + "grad_norm": 2.677380323410034, + "learning_rate": 1.3681270854490063e-05, + "loss": 0.0284, + "step": 8811 + }, + { + "epoch": 0.9553339115351258, + "grad_norm": 4.570540428161621, + "learning_rate": 1.3680545480922675e-05, + "loss": 0.0226, + "step": 8812 + }, + { + "epoch": 0.9554423243712056, + "grad_norm": 0.37221166491508484, + "learning_rate": 1.3679820107355289e-05, + "loss": 0.0075, + "step": 8813 + }, + { + "epoch": 0.9555507372072853, + "grad_norm": 1.0962142944335938, + "learning_rate": 1.3679094733787901e-05, + "loss": 0.0095, + "step": 8814 + }, + { + "epoch": 0.9556591500433651, + "grad_norm": 2.9128527641296387, + "learning_rate": 1.3678369360220513e-05, + "loss": 0.0388, + "step": 8815 + }, + { + "epoch": 0.955767562879445, + "grad_norm": 3.8865175247192383, + "learning_rate": 1.3677643986653129e-05, + "loss": 0.0758, + "step": 8816 + }, + { + "epoch": 0.9558759757155247, + "grad_norm": 0.9535856246948242, + "learning_rate": 1.3676918613085741e-05, + "loss": 0.0123, + "step": 8817 + }, + { + "epoch": 0.9559843885516045, + "grad_norm": 2.3970870971679688, + "learning_rate": 1.3676193239518355e-05, + "loss": 0.0206, + "step": 8818 + }, + { + "epoch": 0.9560928013876843, + "grad_norm": 5.793380260467529, + "learning_rate": 1.3675467865950967e-05, + "loss": 0.0585, + "step": 8819 + }, + { + "epoch": 0.9562012142237641, + "grad_norm": 1.4903054237365723, + "learning_rate": 1.3674742492383579e-05, + "loss": 0.0329, + "step": 8820 + }, + { + "epoch": 0.9563096270598439, + "grad_norm": 4.531562805175781, + "learning_rate": 1.3674017118816191e-05, + "loss": 0.0304, + "step": 8821 + }, + { + "epoch": 0.9564180398959237, + "grad_norm": 2.2041704654693604, + "learning_rate": 1.3673291745248805e-05, + "loss": 0.0542, + "step": 8822 + }, + { + "epoch": 0.9565264527320034, + "grad_norm": 4.222274303436279, + "learning_rate": 1.3672566371681417e-05, + "loss": 0.0897, + "step": 8823 + }, + { + "epoch": 0.9566348655680833, + "grad_norm": 4.611809253692627, + "learning_rate": 1.367184099811403e-05, + "loss": 0.0454, + "step": 8824 + }, + { + "epoch": 0.9567432784041631, + "grad_norm": 2.9809207916259766, + "learning_rate": 1.3671115624546643e-05, + "loss": 0.0375, + "step": 8825 + }, + { + "epoch": 0.9568516912402428, + "grad_norm": 0.5710883140563965, + "learning_rate": 1.3670390250979255e-05, + "loss": 0.0059, + "step": 8826 + }, + { + "epoch": 0.9569601040763226, + "grad_norm": 3.875929355621338, + "learning_rate": 1.3669664877411867e-05, + "loss": 0.0725, + "step": 8827 + }, + { + "epoch": 0.9570685169124025, + "grad_norm": 7.84041166305542, + "learning_rate": 1.3668939503844481e-05, + "loss": 0.0766, + "step": 8828 + }, + { + "epoch": 0.9571769297484822, + "grad_norm": 0.9123333096504211, + "learning_rate": 1.3668214130277093e-05, + "loss": 0.0067, + "step": 8829 + }, + { + "epoch": 0.957285342584562, + "grad_norm": 1.4951438903808594, + "learning_rate": 1.3667488756709705e-05, + "loss": 0.0159, + "step": 8830 + }, + { + "epoch": 0.9573937554206418, + "grad_norm": 3.923917293548584, + "learning_rate": 1.366676338314232e-05, + "loss": 0.0735, + "step": 8831 + }, + { + "epoch": 0.9575021682567216, + "grad_norm": 1.5346674919128418, + "learning_rate": 1.3666038009574931e-05, + "loss": 0.0276, + "step": 8832 + }, + { + "epoch": 0.9576105810928014, + "grad_norm": 1.9821244478225708, + "learning_rate": 1.3665312636007545e-05, + "loss": 0.0428, + "step": 8833 + }, + { + "epoch": 0.9577189939288812, + "grad_norm": 2.42559814453125, + "learning_rate": 1.3664587262440159e-05, + "loss": 0.025, + "step": 8834 + }, + { + "epoch": 0.9578274067649609, + "grad_norm": 3.063298463821411, + "learning_rate": 1.3663861888872771e-05, + "loss": 0.052, + "step": 8835 + }, + { + "epoch": 0.9579358196010408, + "grad_norm": 4.7157487869262695, + "learning_rate": 1.3663136515305383e-05, + "loss": 0.0904, + "step": 8836 + }, + { + "epoch": 0.9580442324371206, + "grad_norm": 2.2673897743225098, + "learning_rate": 1.3662411141737997e-05, + "loss": 0.0448, + "step": 8837 + }, + { + "epoch": 0.9581526452732003, + "grad_norm": 1.2233039140701294, + "learning_rate": 1.366168576817061e-05, + "loss": 0.0069, + "step": 8838 + }, + { + "epoch": 0.9582610581092801, + "grad_norm": 3.213486909866333, + "learning_rate": 1.3660960394603222e-05, + "loss": 0.0529, + "step": 8839 + }, + { + "epoch": 0.95836947094536, + "grad_norm": 2.9236369132995605, + "learning_rate": 1.3660235021035835e-05, + "loss": 0.0311, + "step": 8840 + }, + { + "epoch": 0.9584778837814397, + "grad_norm": 4.237668991088867, + "learning_rate": 1.3659509647468448e-05, + "loss": 0.0207, + "step": 8841 + }, + { + "epoch": 0.9585862966175195, + "grad_norm": 3.8697094917297363, + "learning_rate": 1.365878427390106e-05, + "loss": 0.0302, + "step": 8842 + }, + { + "epoch": 0.9586947094535994, + "grad_norm": 4.74583101272583, + "learning_rate": 1.3658058900333674e-05, + "loss": 0.0222, + "step": 8843 + }, + { + "epoch": 0.9588031222896791, + "grad_norm": 1.7947701215744019, + "learning_rate": 1.3657333526766286e-05, + "loss": 0.0404, + "step": 8844 + }, + { + "epoch": 0.9589115351257589, + "grad_norm": 1.4655593633651733, + "learning_rate": 1.3656608153198898e-05, + "loss": 0.0222, + "step": 8845 + }, + { + "epoch": 0.9590199479618386, + "grad_norm": 5.576328277587891, + "learning_rate": 1.365588277963151e-05, + "loss": 0.0739, + "step": 8846 + }, + { + "epoch": 0.9591283607979185, + "grad_norm": 4.066137790679932, + "learning_rate": 1.3655157406064124e-05, + "loss": 0.0628, + "step": 8847 + }, + { + "epoch": 0.9592367736339983, + "grad_norm": 1.4564383029937744, + "learning_rate": 1.3654432032496736e-05, + "loss": 0.0326, + "step": 8848 + }, + { + "epoch": 0.959345186470078, + "grad_norm": 2.212949514389038, + "learning_rate": 1.3653706658929348e-05, + "loss": 0.0408, + "step": 8849 + }, + { + "epoch": 0.9594535993061578, + "grad_norm": 3.8090875148773193, + "learning_rate": 1.3652981285361962e-05, + "loss": 0.0518, + "step": 8850 + }, + { + "epoch": 0.9595620121422377, + "grad_norm": 2.250668525695801, + "learning_rate": 1.3652255911794576e-05, + "loss": 0.0437, + "step": 8851 + }, + { + "epoch": 0.9596704249783174, + "grad_norm": 1.6673139333724976, + "learning_rate": 1.365153053822719e-05, + "loss": 0.0296, + "step": 8852 + }, + { + "epoch": 0.9597788378143972, + "grad_norm": 2.1054539680480957, + "learning_rate": 1.3650805164659802e-05, + "loss": 0.0426, + "step": 8853 + }, + { + "epoch": 0.959887250650477, + "grad_norm": 1.643102765083313, + "learning_rate": 1.3650079791092414e-05, + "loss": 0.0372, + "step": 8854 + }, + { + "epoch": 0.9599956634865568, + "grad_norm": 2.5123331546783447, + "learning_rate": 1.3649354417525028e-05, + "loss": 0.0172, + "step": 8855 + }, + { + "epoch": 0.9601040763226366, + "grad_norm": 0.9863014221191406, + "learning_rate": 1.364862904395764e-05, + "loss": 0.0157, + "step": 8856 + }, + { + "epoch": 0.9602124891587164, + "grad_norm": 1.3259793519973755, + "learning_rate": 1.3647903670390252e-05, + "loss": 0.0415, + "step": 8857 + }, + { + "epoch": 0.9603209019947961, + "grad_norm": 3.6058833599090576, + "learning_rate": 1.3647178296822866e-05, + "loss": 0.0655, + "step": 8858 + }, + { + "epoch": 0.960429314830876, + "grad_norm": 2.719345808029175, + "learning_rate": 1.3646452923255478e-05, + "loss": 0.0368, + "step": 8859 + }, + { + "epoch": 0.9605377276669558, + "grad_norm": 4.492651462554932, + "learning_rate": 1.364572754968809e-05, + "loss": 0.0709, + "step": 8860 + }, + { + "epoch": 0.9606461405030355, + "grad_norm": 2.198688507080078, + "learning_rate": 1.3645002176120702e-05, + "loss": 0.044, + "step": 8861 + }, + { + "epoch": 0.9607545533391153, + "grad_norm": 4.0990891456604, + "learning_rate": 1.3644276802553316e-05, + "loss": 0.0354, + "step": 8862 + }, + { + "epoch": 0.9608629661751952, + "grad_norm": 1.6997272968292236, + "learning_rate": 1.3643551428985928e-05, + "loss": 0.0223, + "step": 8863 + }, + { + "epoch": 0.9609713790112749, + "grad_norm": 1.6940566301345825, + "learning_rate": 1.364282605541854e-05, + "loss": 0.0276, + "step": 8864 + }, + { + "epoch": 0.9610797918473547, + "grad_norm": 2.438579797744751, + "learning_rate": 1.3642100681851154e-05, + "loss": 0.0507, + "step": 8865 + }, + { + "epoch": 0.9611882046834345, + "grad_norm": 1.8320825099945068, + "learning_rate": 1.3641375308283766e-05, + "loss": 0.0157, + "step": 8866 + }, + { + "epoch": 0.9612966175195143, + "grad_norm": 2.1757516860961914, + "learning_rate": 1.3640649934716379e-05, + "loss": 0.0184, + "step": 8867 + }, + { + "epoch": 0.9614050303555941, + "grad_norm": 2.206751585006714, + "learning_rate": 1.3639924561148994e-05, + "loss": 0.0401, + "step": 8868 + }, + { + "epoch": 0.9615134431916739, + "grad_norm": 1.9773002862930298, + "learning_rate": 1.3639199187581606e-05, + "loss": 0.0253, + "step": 8869 + }, + { + "epoch": 0.9616218560277537, + "grad_norm": 1.5730582475662231, + "learning_rate": 1.363847381401422e-05, + "loss": 0.0121, + "step": 8870 + }, + { + "epoch": 0.9617302688638335, + "grad_norm": 3.1541314125061035, + "learning_rate": 1.3637748440446832e-05, + "loss": 0.0768, + "step": 8871 + }, + { + "epoch": 0.9618386816999133, + "grad_norm": 1.064860463142395, + "learning_rate": 1.3637023066879444e-05, + "loss": 0.0177, + "step": 8872 + }, + { + "epoch": 0.961947094535993, + "grad_norm": 1.6613603830337524, + "learning_rate": 1.3636297693312056e-05, + "loss": 0.0338, + "step": 8873 + }, + { + "epoch": 0.9620555073720729, + "grad_norm": 6.6101531982421875, + "learning_rate": 1.363557231974467e-05, + "loss": 0.044, + "step": 8874 + }, + { + "epoch": 0.9621639202081527, + "grad_norm": 2.3142812252044678, + "learning_rate": 1.3634846946177282e-05, + "loss": 0.0236, + "step": 8875 + }, + { + "epoch": 0.9622723330442324, + "grad_norm": 1.7492775917053223, + "learning_rate": 1.3634121572609895e-05, + "loss": 0.0229, + "step": 8876 + }, + { + "epoch": 0.9623807458803122, + "grad_norm": 0.6889079809188843, + "learning_rate": 1.3633396199042508e-05, + "loss": 0.0106, + "step": 8877 + }, + { + "epoch": 0.9624891587163921, + "grad_norm": 2.8186421394348145, + "learning_rate": 1.363267082547512e-05, + "loss": 0.0458, + "step": 8878 + }, + { + "epoch": 0.9625975715524718, + "grad_norm": 2.173252582550049, + "learning_rate": 1.3631945451907733e-05, + "loss": 0.0361, + "step": 8879 + }, + { + "epoch": 0.9627059843885516, + "grad_norm": 2.9429705142974854, + "learning_rate": 1.3631220078340347e-05, + "loss": 0.0549, + "step": 8880 + }, + { + "epoch": 0.9628143972246314, + "grad_norm": 3.648930549621582, + "learning_rate": 1.3630494704772959e-05, + "loss": 0.0468, + "step": 8881 + }, + { + "epoch": 0.9629228100607112, + "grad_norm": 2.67004132270813, + "learning_rate": 1.362976933120557e-05, + "loss": 0.057, + "step": 8882 + }, + { + "epoch": 0.963031222896791, + "grad_norm": 1.0572335720062256, + "learning_rate": 1.3629043957638185e-05, + "loss": 0.0176, + "step": 8883 + }, + { + "epoch": 0.9631396357328708, + "grad_norm": 0.8595595359802246, + "learning_rate": 1.3628318584070797e-05, + "loss": 0.0202, + "step": 8884 + }, + { + "epoch": 0.9632480485689505, + "grad_norm": 1.9233484268188477, + "learning_rate": 1.362759321050341e-05, + "loss": 0.046, + "step": 8885 + }, + { + "epoch": 0.9633564614050304, + "grad_norm": 2.8714962005615234, + "learning_rate": 1.3626867836936024e-05, + "loss": 0.0391, + "step": 8886 + }, + { + "epoch": 0.9634648742411102, + "grad_norm": 4.081872463226318, + "learning_rate": 1.3626142463368637e-05, + "loss": 0.0591, + "step": 8887 + }, + { + "epoch": 0.9635732870771899, + "grad_norm": 1.528430700302124, + "learning_rate": 1.3625417089801249e-05, + "loss": 0.0169, + "step": 8888 + }, + { + "epoch": 0.9636816999132697, + "grad_norm": 1.7850044965744019, + "learning_rate": 1.3624691716233863e-05, + "loss": 0.0287, + "step": 8889 + }, + { + "epoch": 0.9637901127493496, + "grad_norm": 0.7606122493743896, + "learning_rate": 1.3623966342666475e-05, + "loss": 0.0178, + "step": 8890 + }, + { + "epoch": 0.9638985255854293, + "grad_norm": 0.8682780265808105, + "learning_rate": 1.3623240969099087e-05, + "loss": 0.0164, + "step": 8891 + }, + { + "epoch": 0.9640069384215091, + "grad_norm": 2.291215181350708, + "learning_rate": 1.36225155955317e-05, + "loss": 0.0331, + "step": 8892 + }, + { + "epoch": 0.9641153512575888, + "grad_norm": 2.6328189373016357, + "learning_rate": 1.3621790221964313e-05, + "loss": 0.0409, + "step": 8893 + }, + { + "epoch": 0.9642237640936687, + "grad_norm": 2.1734368801116943, + "learning_rate": 1.3621064848396925e-05, + "loss": 0.0349, + "step": 8894 + }, + { + "epoch": 0.9643321769297485, + "grad_norm": 2.5064146518707275, + "learning_rate": 1.3620339474829539e-05, + "loss": 0.0425, + "step": 8895 + }, + { + "epoch": 0.9644405897658282, + "grad_norm": 6.038628101348877, + "learning_rate": 1.3619614101262151e-05, + "loss": 0.0453, + "step": 8896 + }, + { + "epoch": 0.964549002601908, + "grad_norm": 2.667942762374878, + "learning_rate": 1.3618888727694763e-05, + "loss": 0.0441, + "step": 8897 + }, + { + "epoch": 0.9646574154379879, + "grad_norm": 2.3937199115753174, + "learning_rate": 1.3618163354127375e-05, + "loss": 0.0248, + "step": 8898 + }, + { + "epoch": 0.9647658282740676, + "grad_norm": 2.090954542160034, + "learning_rate": 1.3617437980559989e-05, + "loss": 0.0299, + "step": 8899 + }, + { + "epoch": 0.9648742411101474, + "grad_norm": 1.1070536375045776, + "learning_rate": 1.3616712606992601e-05, + "loss": 0.015, + "step": 8900 + }, + { + "epoch": 0.9649826539462273, + "grad_norm": 0.42120540142059326, + "learning_rate": 1.3615987233425213e-05, + "loss": 0.0031, + "step": 8901 + }, + { + "epoch": 0.965091066782307, + "grad_norm": 1.4701509475708008, + "learning_rate": 1.3615261859857829e-05, + "loss": 0.0206, + "step": 8902 + }, + { + "epoch": 0.9651994796183868, + "grad_norm": 1.5268417596817017, + "learning_rate": 1.3614536486290441e-05, + "loss": 0.0176, + "step": 8903 + }, + { + "epoch": 0.9653078924544666, + "grad_norm": 2.5863821506500244, + "learning_rate": 1.3613811112723055e-05, + "loss": 0.0449, + "step": 8904 + }, + { + "epoch": 0.9654163052905465, + "grad_norm": 7.094207763671875, + "learning_rate": 1.3613085739155667e-05, + "loss": 0.0675, + "step": 8905 + }, + { + "epoch": 0.9655247181266262, + "grad_norm": 3.7243568897247314, + "learning_rate": 1.361236036558828e-05, + "loss": 0.0208, + "step": 8906 + }, + { + "epoch": 0.965633130962706, + "grad_norm": 4.805793285369873, + "learning_rate": 1.3611634992020893e-05, + "loss": 0.0589, + "step": 8907 + }, + { + "epoch": 0.9657415437987857, + "grad_norm": 1.5477594137191772, + "learning_rate": 1.3610909618453505e-05, + "loss": 0.0398, + "step": 8908 + }, + { + "epoch": 0.9658499566348656, + "grad_norm": 3.609875440597534, + "learning_rate": 1.3610184244886117e-05, + "loss": 0.1013, + "step": 8909 + }, + { + "epoch": 0.9659583694709454, + "grad_norm": 0.6954930424690247, + "learning_rate": 1.3609458871318731e-05, + "loss": 0.0058, + "step": 8910 + }, + { + "epoch": 0.9660667823070251, + "grad_norm": 1.0432699918746948, + "learning_rate": 1.3608733497751343e-05, + "loss": 0.0171, + "step": 8911 + }, + { + "epoch": 0.9661751951431049, + "grad_norm": 6.518730640411377, + "learning_rate": 1.3608008124183955e-05, + "loss": 0.0297, + "step": 8912 + }, + { + "epoch": 0.9662836079791848, + "grad_norm": 3.834095001220703, + "learning_rate": 1.3607282750616568e-05, + "loss": 0.0488, + "step": 8913 + }, + { + "epoch": 0.9663920208152645, + "grad_norm": 2.9924826622009277, + "learning_rate": 1.3606557377049181e-05, + "loss": 0.0269, + "step": 8914 + }, + { + "epoch": 0.9665004336513443, + "grad_norm": 2.3355703353881836, + "learning_rate": 1.3605832003481794e-05, + "loss": 0.0206, + "step": 8915 + }, + { + "epoch": 0.9666088464874241, + "grad_norm": 3.309283971786499, + "learning_rate": 1.3605106629914406e-05, + "loss": 0.0272, + "step": 8916 + }, + { + "epoch": 0.9667172593235039, + "grad_norm": 2.345611333847046, + "learning_rate": 1.360438125634702e-05, + "loss": 0.0293, + "step": 8917 + }, + { + "epoch": 0.9668256721595837, + "grad_norm": 2.6100614070892334, + "learning_rate": 1.3603655882779632e-05, + "loss": 0.0392, + "step": 8918 + }, + { + "epoch": 0.9669340849956635, + "grad_norm": 2.2686452865600586, + "learning_rate": 1.3602930509212244e-05, + "loss": 0.0874, + "step": 8919 + }, + { + "epoch": 0.9670424978317432, + "grad_norm": 1.727089524269104, + "learning_rate": 1.360220513564486e-05, + "loss": 0.0289, + "step": 8920 + }, + { + "epoch": 0.9671509106678231, + "grad_norm": 1.2771246433258057, + "learning_rate": 1.3601479762077472e-05, + "loss": 0.041, + "step": 8921 + }, + { + "epoch": 0.9672593235039029, + "grad_norm": 1.239546537399292, + "learning_rate": 1.3600754388510085e-05, + "loss": 0.0214, + "step": 8922 + }, + { + "epoch": 0.9673677363399826, + "grad_norm": 2.2060697078704834, + "learning_rate": 1.3600029014942698e-05, + "loss": 0.0287, + "step": 8923 + }, + { + "epoch": 0.9674761491760624, + "grad_norm": 2.1377651691436768, + "learning_rate": 1.359930364137531e-05, + "loss": 0.0261, + "step": 8924 + }, + { + "epoch": 0.9675845620121423, + "grad_norm": 3.9529035091400146, + "learning_rate": 1.3598578267807922e-05, + "loss": 0.0236, + "step": 8925 + }, + { + "epoch": 0.967692974848222, + "grad_norm": 2.8849363327026367, + "learning_rate": 1.3597852894240536e-05, + "loss": 0.0077, + "step": 8926 + }, + { + "epoch": 0.9678013876843018, + "grad_norm": 3.200531482696533, + "learning_rate": 1.3597127520673148e-05, + "loss": 0.0369, + "step": 8927 + }, + { + "epoch": 0.9679098005203816, + "grad_norm": 5.337920665740967, + "learning_rate": 1.359640214710576e-05, + "loss": 0.0804, + "step": 8928 + }, + { + "epoch": 0.9680182133564614, + "grad_norm": 1.4607571363449097, + "learning_rate": 1.3595676773538374e-05, + "loss": 0.0273, + "step": 8929 + }, + { + "epoch": 0.9681266261925412, + "grad_norm": 1.4298828840255737, + "learning_rate": 1.3594951399970986e-05, + "loss": 0.0117, + "step": 8930 + }, + { + "epoch": 0.968235039028621, + "grad_norm": 2.1254591941833496, + "learning_rate": 1.3594226026403598e-05, + "loss": 0.0458, + "step": 8931 + }, + { + "epoch": 0.9683434518647008, + "grad_norm": 1.1853749752044678, + "learning_rate": 1.3593500652836212e-05, + "loss": 0.0115, + "step": 8932 + }, + { + "epoch": 0.9684518647007806, + "grad_norm": 2.946035861968994, + "learning_rate": 1.3592775279268824e-05, + "loss": 0.0719, + "step": 8933 + }, + { + "epoch": 0.9685602775368604, + "grad_norm": 2.0292203426361084, + "learning_rate": 1.3592049905701436e-05, + "loss": 0.0242, + "step": 8934 + }, + { + "epoch": 0.9686686903729401, + "grad_norm": 3.630452871322632, + "learning_rate": 1.359132453213405e-05, + "loss": 0.0441, + "step": 8935 + }, + { + "epoch": 0.96877710320902, + "grad_norm": 1.9541417360305786, + "learning_rate": 1.3590599158566662e-05, + "loss": 0.015, + "step": 8936 + }, + { + "epoch": 0.9688855160450998, + "grad_norm": 1.410569190979004, + "learning_rate": 1.3589873784999276e-05, + "loss": 0.0286, + "step": 8937 + }, + { + "epoch": 0.9689939288811795, + "grad_norm": 3.1401877403259277, + "learning_rate": 1.358914841143189e-05, + "loss": 0.0482, + "step": 8938 + }, + { + "epoch": 0.9691023417172593, + "grad_norm": 7.860505104064941, + "learning_rate": 1.3588423037864502e-05, + "loss": 0.0654, + "step": 8939 + }, + { + "epoch": 0.9692107545533392, + "grad_norm": 1.0066132545471191, + "learning_rate": 1.3587697664297114e-05, + "loss": 0.0131, + "step": 8940 + }, + { + "epoch": 0.9693191673894189, + "grad_norm": 4.096842288970947, + "learning_rate": 1.3586972290729728e-05, + "loss": 0.0716, + "step": 8941 + }, + { + "epoch": 0.9694275802254987, + "grad_norm": 3.3850080966949463, + "learning_rate": 1.358624691716234e-05, + "loss": 0.0454, + "step": 8942 + }, + { + "epoch": 0.9695359930615784, + "grad_norm": 2.100177526473999, + "learning_rate": 1.3585521543594952e-05, + "loss": 0.0223, + "step": 8943 + }, + { + "epoch": 0.9696444058976583, + "grad_norm": 2.9457333087921143, + "learning_rate": 1.3584796170027566e-05, + "loss": 0.024, + "step": 8944 + }, + { + "epoch": 0.9697528187337381, + "grad_norm": 1.2029186487197876, + "learning_rate": 1.3584070796460178e-05, + "loss": 0.0288, + "step": 8945 + }, + { + "epoch": 0.9698612315698178, + "grad_norm": 1.9859061241149902, + "learning_rate": 1.358334542289279e-05, + "loss": 0.0118, + "step": 8946 + }, + { + "epoch": 0.9699696444058976, + "grad_norm": 2.634540557861328, + "learning_rate": 1.3582620049325404e-05, + "loss": 0.0607, + "step": 8947 + }, + { + "epoch": 0.9700780572419775, + "grad_norm": 1.6147620677947998, + "learning_rate": 1.3581894675758016e-05, + "loss": 0.022, + "step": 8948 + }, + { + "epoch": 0.9701864700780573, + "grad_norm": 3.3354411125183105, + "learning_rate": 1.3581169302190628e-05, + "loss": 0.0394, + "step": 8949 + }, + { + "epoch": 0.970294882914137, + "grad_norm": 4.655428409576416, + "learning_rate": 1.358044392862324e-05, + "loss": 0.0641, + "step": 8950 + }, + { + "epoch": 0.9704032957502168, + "grad_norm": 4.724836349487305, + "learning_rate": 1.3579718555055854e-05, + "loss": 0.031, + "step": 8951 + }, + { + "epoch": 0.9705117085862967, + "grad_norm": 1.1729846000671387, + "learning_rate": 1.3578993181488467e-05, + "loss": 0.013, + "step": 8952 + }, + { + "epoch": 0.9706201214223764, + "grad_norm": 2.0005557537078857, + "learning_rate": 1.3578267807921079e-05, + "loss": 0.0326, + "step": 8953 + }, + { + "epoch": 0.9707285342584562, + "grad_norm": 4.587646484375, + "learning_rate": 1.3577542434353694e-05, + "loss": 0.0434, + "step": 8954 + }, + { + "epoch": 0.9708369470945359, + "grad_norm": 10.241863250732422, + "learning_rate": 1.3576817060786306e-05, + "loss": 0.0463, + "step": 8955 + }, + { + "epoch": 0.9709453599306158, + "grad_norm": 5.111763954162598, + "learning_rate": 1.357609168721892e-05, + "loss": 0.0435, + "step": 8956 + }, + { + "epoch": 0.9710537727666956, + "grad_norm": 2.807638168334961, + "learning_rate": 1.3575366313651532e-05, + "loss": 0.0423, + "step": 8957 + }, + { + "epoch": 0.9711621856027753, + "grad_norm": 1.2474229335784912, + "learning_rate": 1.3574640940084145e-05, + "loss": 0.0143, + "step": 8958 + }, + { + "epoch": 0.9712705984388552, + "grad_norm": 4.8621344566345215, + "learning_rate": 1.3573915566516758e-05, + "loss": 0.0472, + "step": 8959 + }, + { + "epoch": 0.971379011274935, + "grad_norm": 1.383542776107788, + "learning_rate": 1.357319019294937e-05, + "loss": 0.0178, + "step": 8960 + }, + { + "epoch": 0.9714874241110147, + "grad_norm": 2.8337955474853516, + "learning_rate": 1.3572464819381983e-05, + "loss": 0.0371, + "step": 8961 + }, + { + "epoch": 0.9715958369470945, + "grad_norm": 2.253937244415283, + "learning_rate": 1.3571739445814597e-05, + "loss": 0.0117, + "step": 8962 + }, + { + "epoch": 0.9717042497831744, + "grad_norm": 2.8338167667388916, + "learning_rate": 1.3571014072247209e-05, + "loss": 0.0491, + "step": 8963 + }, + { + "epoch": 0.9718126626192541, + "grad_norm": 1.5414025783538818, + "learning_rate": 1.357028869867982e-05, + "loss": 0.0283, + "step": 8964 + }, + { + "epoch": 0.9719210754553339, + "grad_norm": 2.3996801376342773, + "learning_rate": 1.3569563325112433e-05, + "loss": 0.045, + "step": 8965 + }, + { + "epoch": 0.9720294882914137, + "grad_norm": 2.2299883365631104, + "learning_rate": 1.3568837951545047e-05, + "loss": 0.0427, + "step": 8966 + }, + { + "epoch": 0.9721379011274935, + "grad_norm": 2.9810373783111572, + "learning_rate": 1.3568112577977659e-05, + "loss": 0.0163, + "step": 8967 + }, + { + "epoch": 0.9722463139635733, + "grad_norm": 1.2880902290344238, + "learning_rate": 1.3567387204410271e-05, + "loss": 0.0302, + "step": 8968 + }, + { + "epoch": 0.9723547267996531, + "grad_norm": 4.929921627044678, + "learning_rate": 1.3566661830842885e-05, + "loss": 0.0846, + "step": 8969 + }, + { + "epoch": 0.9724631396357328, + "grad_norm": 4.019143581390381, + "learning_rate": 1.3565936457275497e-05, + "loss": 0.0994, + "step": 8970 + }, + { + "epoch": 0.9725715524718127, + "grad_norm": 2.06046199798584, + "learning_rate": 1.356521108370811e-05, + "loss": 0.0275, + "step": 8971 + }, + { + "epoch": 0.9726799653078925, + "grad_norm": 3.893878936767578, + "learning_rate": 1.3564485710140725e-05, + "loss": 0.062, + "step": 8972 + }, + { + "epoch": 0.9727883781439722, + "grad_norm": 1.1575039625167847, + "learning_rate": 1.3563760336573337e-05, + "loss": 0.0189, + "step": 8973 + }, + { + "epoch": 0.972896790980052, + "grad_norm": 1.0886038541793823, + "learning_rate": 1.356303496300595e-05, + "loss": 0.0081, + "step": 8974 + }, + { + "epoch": 0.9730052038161319, + "grad_norm": 1.9321688413619995, + "learning_rate": 1.3562309589438563e-05, + "loss": 0.0194, + "step": 8975 + }, + { + "epoch": 0.9731136166522116, + "grad_norm": 1.191667914390564, + "learning_rate": 1.3561584215871175e-05, + "loss": 0.0238, + "step": 8976 + }, + { + "epoch": 0.9732220294882914, + "grad_norm": 2.6108784675598145, + "learning_rate": 1.3560858842303787e-05, + "loss": 0.0411, + "step": 8977 + }, + { + "epoch": 0.9733304423243712, + "grad_norm": 2.139984607696533, + "learning_rate": 1.3560133468736401e-05, + "loss": 0.0596, + "step": 8978 + }, + { + "epoch": 0.973438855160451, + "grad_norm": 5.238001823425293, + "learning_rate": 1.3559408095169013e-05, + "loss": 0.0648, + "step": 8979 + }, + { + "epoch": 0.9735472679965308, + "grad_norm": 5.995332717895508, + "learning_rate": 1.3558682721601625e-05, + "loss": 0.0377, + "step": 8980 + }, + { + "epoch": 0.9736556808326106, + "grad_norm": 4.831324577331543, + "learning_rate": 1.3557957348034239e-05, + "loss": 0.0444, + "step": 8981 + }, + { + "epoch": 0.9737640936686903, + "grad_norm": 3.800861358642578, + "learning_rate": 1.3557231974466851e-05, + "loss": 0.0559, + "step": 8982 + }, + { + "epoch": 0.9738725065047702, + "grad_norm": 1.14506196975708, + "learning_rate": 1.3556506600899463e-05, + "loss": 0.0132, + "step": 8983 + }, + { + "epoch": 0.97398091934085, + "grad_norm": 1.9734073877334595, + "learning_rate": 1.3555781227332077e-05, + "loss": 0.0275, + "step": 8984 + }, + { + "epoch": 0.9740893321769297, + "grad_norm": 3.9755659103393555, + "learning_rate": 1.355505585376469e-05, + "loss": 0.023, + "step": 8985 + }, + { + "epoch": 0.9741977450130095, + "grad_norm": 2.703421115875244, + "learning_rate": 1.3554330480197302e-05, + "loss": 0.081, + "step": 8986 + }, + { + "epoch": 0.9743061578490894, + "grad_norm": 2.131044626235962, + "learning_rate": 1.3553605106629915e-05, + "loss": 0.026, + "step": 8987 + }, + { + "epoch": 0.9744145706851691, + "grad_norm": 2.355328321456909, + "learning_rate": 1.3552879733062527e-05, + "loss": 0.0427, + "step": 8988 + }, + { + "epoch": 0.9745229835212489, + "grad_norm": 3.670884370803833, + "learning_rate": 1.3552154359495141e-05, + "loss": 0.0336, + "step": 8989 + }, + { + "epoch": 0.9746313963573288, + "grad_norm": 1.2157554626464844, + "learning_rate": 1.3551428985927755e-05, + "loss": 0.024, + "step": 8990 + }, + { + "epoch": 0.9747398091934085, + "grad_norm": 5.075969219207764, + "learning_rate": 1.3550703612360367e-05, + "loss": 0.061, + "step": 8991 + }, + { + "epoch": 0.9748482220294883, + "grad_norm": 1.871977686882019, + "learning_rate": 1.354997823879298e-05, + "loss": 0.0334, + "step": 8992 + }, + { + "epoch": 0.974956634865568, + "grad_norm": 1.936237096786499, + "learning_rate": 1.3549252865225593e-05, + "loss": 0.012, + "step": 8993 + }, + { + "epoch": 0.9750650477016479, + "grad_norm": 4.025847434997559, + "learning_rate": 1.3548527491658205e-05, + "loss": 0.0727, + "step": 8994 + }, + { + "epoch": 0.9751734605377277, + "grad_norm": 1.833248496055603, + "learning_rate": 1.3547802118090818e-05, + "loss": 0.065, + "step": 8995 + }, + { + "epoch": 0.9752818733738075, + "grad_norm": 8.040950775146484, + "learning_rate": 1.3547076744523431e-05, + "loss": 0.0685, + "step": 8996 + }, + { + "epoch": 0.9753902862098872, + "grad_norm": 2.6865761280059814, + "learning_rate": 1.3546351370956044e-05, + "loss": 0.0143, + "step": 8997 + }, + { + "epoch": 0.9754986990459671, + "grad_norm": 2.173337936401367, + "learning_rate": 1.3545625997388656e-05, + "loss": 0.0453, + "step": 8998 + }, + { + "epoch": 0.9756071118820469, + "grad_norm": 1.3969767093658447, + "learning_rate": 1.354490062382127e-05, + "loss": 0.0158, + "step": 8999 + }, + { + "epoch": 0.9757155247181266, + "grad_norm": 2.112328052520752, + "learning_rate": 1.3544175250253882e-05, + "loss": 0.0662, + "step": 9000 + }, + { + "epoch": 0.9758239375542064, + "grad_norm": 3.878114938735962, + "learning_rate": 1.3543449876686494e-05, + "loss": 0.0255, + "step": 9001 + }, + { + "epoch": 0.9759323503902863, + "grad_norm": 2.5912857055664062, + "learning_rate": 1.3542724503119106e-05, + "loss": 0.0234, + "step": 9002 + }, + { + "epoch": 0.976040763226366, + "grad_norm": 2.36749267578125, + "learning_rate": 1.354199912955172e-05, + "loss": 0.0184, + "step": 9003 + }, + { + "epoch": 0.9761491760624458, + "grad_norm": 1.154564380645752, + "learning_rate": 1.3541273755984332e-05, + "loss": 0.0233, + "step": 9004 + }, + { + "epoch": 0.9762575888985255, + "grad_norm": 1.93625807762146, + "learning_rate": 1.3540548382416944e-05, + "loss": 0.0234, + "step": 9005 + }, + { + "epoch": 0.9763660017346054, + "grad_norm": 1.4719481468200684, + "learning_rate": 1.353982300884956e-05, + "loss": 0.017, + "step": 9006 + }, + { + "epoch": 0.9764744145706852, + "grad_norm": 9.991081237792969, + "learning_rate": 1.3539097635282172e-05, + "loss": 0.0697, + "step": 9007 + }, + { + "epoch": 0.9765828274067649, + "grad_norm": 2.7270147800445557, + "learning_rate": 1.3538372261714786e-05, + "loss": 0.046, + "step": 9008 + }, + { + "epoch": 0.9766912402428447, + "grad_norm": 0.8256828784942627, + "learning_rate": 1.3537646888147398e-05, + "loss": 0.0108, + "step": 9009 + }, + { + "epoch": 0.9767996530789246, + "grad_norm": 1.9632959365844727, + "learning_rate": 1.353692151458001e-05, + "loss": 0.0472, + "step": 9010 + }, + { + "epoch": 0.9769080659150043, + "grad_norm": 4.897665977478027, + "learning_rate": 1.3536196141012624e-05, + "loss": 0.0532, + "step": 9011 + }, + { + "epoch": 0.9770164787510841, + "grad_norm": 2.8334383964538574, + "learning_rate": 1.3535470767445236e-05, + "loss": 0.0285, + "step": 9012 + }, + { + "epoch": 0.9771248915871639, + "grad_norm": 1.4012959003448486, + "learning_rate": 1.3534745393877848e-05, + "loss": 0.0323, + "step": 9013 + }, + { + "epoch": 0.9772333044232437, + "grad_norm": 2.1127030849456787, + "learning_rate": 1.3534020020310462e-05, + "loss": 0.0433, + "step": 9014 + }, + { + "epoch": 0.9773417172593235, + "grad_norm": 1.9473252296447754, + "learning_rate": 1.3533294646743074e-05, + "loss": 0.0412, + "step": 9015 + }, + { + "epoch": 0.9774501300954033, + "grad_norm": 2.8276429176330566, + "learning_rate": 1.3532569273175686e-05, + "loss": 0.0252, + "step": 9016 + }, + { + "epoch": 0.977558542931483, + "grad_norm": 1.5552780628204346, + "learning_rate": 1.3531843899608298e-05, + "loss": 0.0375, + "step": 9017 + }, + { + "epoch": 0.9776669557675629, + "grad_norm": 3.7950873374938965, + "learning_rate": 1.3531118526040912e-05, + "loss": 0.0541, + "step": 9018 + }, + { + "epoch": 0.9777753686036427, + "grad_norm": 2.1337976455688477, + "learning_rate": 1.3530393152473524e-05, + "loss": 0.0142, + "step": 9019 + }, + { + "epoch": 0.9778837814397224, + "grad_norm": 4.193657398223877, + "learning_rate": 1.3529667778906136e-05, + "loss": 0.078, + "step": 9020 + }, + { + "epoch": 0.9779921942758023, + "grad_norm": 3.911177396774292, + "learning_rate": 1.352894240533875e-05, + "loss": 0.0253, + "step": 9021 + }, + { + "epoch": 0.9781006071118821, + "grad_norm": 0.34766536951065063, + "learning_rate": 1.3528217031771362e-05, + "loss": 0.0051, + "step": 9022 + }, + { + "epoch": 0.9782090199479618, + "grad_norm": 0.8849215507507324, + "learning_rate": 1.3527491658203978e-05, + "loss": 0.0151, + "step": 9023 + }, + { + "epoch": 0.9783174327840416, + "grad_norm": 1.512416958808899, + "learning_rate": 1.352676628463659e-05, + "loss": 0.0314, + "step": 9024 + }, + { + "epoch": 0.9784258456201215, + "grad_norm": 2.1263647079467773, + "learning_rate": 1.3526040911069202e-05, + "loss": 0.0613, + "step": 9025 + }, + { + "epoch": 0.9785342584562012, + "grad_norm": 2.4127907752990723, + "learning_rate": 1.3525315537501816e-05, + "loss": 0.0356, + "step": 9026 + }, + { + "epoch": 0.978642671292281, + "grad_norm": 0.49108654260635376, + "learning_rate": 1.3524590163934428e-05, + "loss": 0.0073, + "step": 9027 + }, + { + "epoch": 0.9787510841283608, + "grad_norm": 3.562370538711548, + "learning_rate": 1.352386479036704e-05, + "loss": 0.0179, + "step": 9028 + }, + { + "epoch": 0.9788594969644406, + "grad_norm": 3.0526487827301025, + "learning_rate": 1.3523139416799652e-05, + "loss": 0.0532, + "step": 9029 + }, + { + "epoch": 0.9789679098005204, + "grad_norm": 3.4491918087005615, + "learning_rate": 1.3522414043232266e-05, + "loss": 0.0212, + "step": 9030 + }, + { + "epoch": 0.9790763226366002, + "grad_norm": 1.8999606370925903, + "learning_rate": 1.3521688669664878e-05, + "loss": 0.0572, + "step": 9031 + }, + { + "epoch": 0.9791847354726799, + "grad_norm": 1.5383471250534058, + "learning_rate": 1.352096329609749e-05, + "loss": 0.0273, + "step": 9032 + }, + { + "epoch": 0.9792931483087598, + "grad_norm": 2.438178300857544, + "learning_rate": 1.3520237922530104e-05, + "loss": 0.0287, + "step": 9033 + }, + { + "epoch": 0.9794015611448396, + "grad_norm": 2.7922210693359375, + "learning_rate": 1.3519512548962717e-05, + "loss": 0.0594, + "step": 9034 + }, + { + "epoch": 0.9795099739809193, + "grad_norm": 2.8669002056121826, + "learning_rate": 1.3518787175395329e-05, + "loss": 0.0281, + "step": 9035 + }, + { + "epoch": 0.9796183868169991, + "grad_norm": 3.2271080017089844, + "learning_rate": 1.3518061801827943e-05, + "loss": 0.0463, + "step": 9036 + }, + { + "epoch": 0.979726799653079, + "grad_norm": 1.7671337127685547, + "learning_rate": 1.3517336428260555e-05, + "loss": 0.0376, + "step": 9037 + }, + { + "epoch": 0.9798352124891587, + "grad_norm": 15.489291191101074, + "learning_rate": 1.3516611054693167e-05, + "loss": 0.025, + "step": 9038 + }, + { + "epoch": 0.9799436253252385, + "grad_norm": 1.3113291263580322, + "learning_rate": 1.351588568112578e-05, + "loss": 0.022, + "step": 9039 + }, + { + "epoch": 0.9800520381613183, + "grad_norm": 1.731619954109192, + "learning_rate": 1.3515160307558393e-05, + "loss": 0.0282, + "step": 9040 + }, + { + "epoch": 0.9801604509973981, + "grad_norm": 0.6047652363777161, + "learning_rate": 1.3514434933991007e-05, + "loss": 0.0032, + "step": 9041 + }, + { + "epoch": 0.9802688638334779, + "grad_norm": 2.3079705238342285, + "learning_rate": 1.351370956042362e-05, + "loss": 0.0282, + "step": 9042 + }, + { + "epoch": 0.9803772766695577, + "grad_norm": 2.313474655151367, + "learning_rate": 1.3512984186856233e-05, + "loss": 0.0207, + "step": 9043 + }, + { + "epoch": 0.9804856895056374, + "grad_norm": 2.91409969329834, + "learning_rate": 1.3512258813288845e-05, + "loss": 0.0453, + "step": 9044 + }, + { + "epoch": 0.9805941023417173, + "grad_norm": 2.901327610015869, + "learning_rate": 1.3511533439721459e-05, + "loss": 0.0198, + "step": 9045 + }, + { + "epoch": 0.980702515177797, + "grad_norm": 4.7180962562561035, + "learning_rate": 1.351080806615407e-05, + "loss": 0.0525, + "step": 9046 + }, + { + "epoch": 0.9808109280138768, + "grad_norm": 1.0519962310791016, + "learning_rate": 1.3510082692586683e-05, + "loss": 0.024, + "step": 9047 + }, + { + "epoch": 0.9809193408499567, + "grad_norm": 1.998060703277588, + "learning_rate": 1.3509357319019297e-05, + "loss": 0.0228, + "step": 9048 + }, + { + "epoch": 0.9810277536860365, + "grad_norm": 2.0501527786254883, + "learning_rate": 1.3508631945451909e-05, + "loss": 0.0396, + "step": 9049 + }, + { + "epoch": 0.9811361665221162, + "grad_norm": 1.8871486186981201, + "learning_rate": 1.3507906571884521e-05, + "loss": 0.0295, + "step": 9050 + }, + { + "epoch": 0.981244579358196, + "grad_norm": 4.924339771270752, + "learning_rate": 1.3507181198317135e-05, + "loss": 0.0696, + "step": 9051 + }, + { + "epoch": 0.9813529921942759, + "grad_norm": 2.0039713382720947, + "learning_rate": 1.3506455824749747e-05, + "loss": 0.0395, + "step": 9052 + }, + { + "epoch": 0.9814614050303556, + "grad_norm": 2.317927122116089, + "learning_rate": 1.350573045118236e-05, + "loss": 0.0355, + "step": 9053 + }, + { + "epoch": 0.9815698178664354, + "grad_norm": 8.343865394592285, + "learning_rate": 1.3505005077614971e-05, + "loss": 0.0498, + "step": 9054 + }, + { + "epoch": 0.9816782307025151, + "grad_norm": 1.73792564868927, + "learning_rate": 1.3504279704047585e-05, + "loss": 0.021, + "step": 9055 + }, + { + "epoch": 0.981786643538595, + "grad_norm": 2.512516736984253, + "learning_rate": 1.3503554330480197e-05, + "loss": 0.0464, + "step": 9056 + }, + { + "epoch": 0.9818950563746748, + "grad_norm": 5.76796817779541, + "learning_rate": 1.350282895691281e-05, + "loss": 0.0238, + "step": 9057 + }, + { + "epoch": 0.9820034692107545, + "grad_norm": 1.9001004695892334, + "learning_rate": 1.3502103583345425e-05, + "loss": 0.0088, + "step": 9058 + }, + { + "epoch": 0.9821118820468343, + "grad_norm": 0.8163529634475708, + "learning_rate": 1.3501378209778037e-05, + "loss": 0.0062, + "step": 9059 + }, + { + "epoch": 0.9822202948829142, + "grad_norm": 5.349799633026123, + "learning_rate": 1.3500652836210651e-05, + "loss": 0.0586, + "step": 9060 + }, + { + "epoch": 0.9823287077189939, + "grad_norm": 4.21513032913208, + "learning_rate": 1.3499927462643263e-05, + "loss": 0.053, + "step": 9061 + }, + { + "epoch": 0.9824371205550737, + "grad_norm": 2.8520073890686035, + "learning_rate": 1.3499202089075875e-05, + "loss": 0.0611, + "step": 9062 + }, + { + "epoch": 0.9825455333911535, + "grad_norm": 2.4588563442230225, + "learning_rate": 1.3498476715508489e-05, + "loss": 0.0351, + "step": 9063 + }, + { + "epoch": 0.9826539462272333, + "grad_norm": 1.1445426940917969, + "learning_rate": 1.3497751341941101e-05, + "loss": 0.0155, + "step": 9064 + }, + { + "epoch": 0.9827623590633131, + "grad_norm": 3.306607961654663, + "learning_rate": 1.3497025968373713e-05, + "loss": 0.0467, + "step": 9065 + }, + { + "epoch": 0.9828707718993929, + "grad_norm": 1.687203288078308, + "learning_rate": 1.3496300594806327e-05, + "loss": 0.034, + "step": 9066 + }, + { + "epoch": 0.9829791847354726, + "grad_norm": 1.1286364793777466, + "learning_rate": 1.349557522123894e-05, + "loss": 0.0092, + "step": 9067 + }, + { + "epoch": 0.9830875975715525, + "grad_norm": 1.300818681716919, + "learning_rate": 1.3494849847671551e-05, + "loss": 0.0359, + "step": 9068 + }, + { + "epoch": 0.9831960104076323, + "grad_norm": 5.891154766082764, + "learning_rate": 1.3494124474104164e-05, + "loss": 0.1048, + "step": 9069 + }, + { + "epoch": 0.983304423243712, + "grad_norm": 3.3072032928466797, + "learning_rate": 1.3493399100536777e-05, + "loss": 0.037, + "step": 9070 + }, + { + "epoch": 0.9834128360797918, + "grad_norm": 3.1701085567474365, + "learning_rate": 1.349267372696939e-05, + "loss": 0.0282, + "step": 9071 + }, + { + "epoch": 0.9835212489158717, + "grad_norm": 1.1444268226623535, + "learning_rate": 1.3491948353402002e-05, + "loss": 0.0122, + "step": 9072 + }, + { + "epoch": 0.9836296617519514, + "grad_norm": 6.246299743652344, + "learning_rate": 1.3491222979834616e-05, + "loss": 0.0336, + "step": 9073 + }, + { + "epoch": 0.9837380745880312, + "grad_norm": 5.5343017578125, + "learning_rate": 1.3490497606267228e-05, + "loss": 0.012, + "step": 9074 + }, + { + "epoch": 0.983846487424111, + "grad_norm": 3.808913469314575, + "learning_rate": 1.3489772232699843e-05, + "loss": 0.0536, + "step": 9075 + }, + { + "epoch": 0.9839549002601908, + "grad_norm": 3.2995800971984863, + "learning_rate": 1.3489046859132455e-05, + "loss": 0.0401, + "step": 9076 + }, + { + "epoch": 0.9840633130962706, + "grad_norm": 4.440751075744629, + "learning_rate": 1.3488321485565068e-05, + "loss": 0.078, + "step": 9077 + }, + { + "epoch": 0.9841717259323504, + "grad_norm": 0.8556493520736694, + "learning_rate": 1.3487596111997681e-05, + "loss": 0.0059, + "step": 9078 + }, + { + "epoch": 0.9842801387684302, + "grad_norm": 1.2906279563903809, + "learning_rate": 1.3486870738430294e-05, + "loss": 0.0175, + "step": 9079 + }, + { + "epoch": 0.98438855160451, + "grad_norm": 2.0366690158843994, + "learning_rate": 1.3486145364862906e-05, + "loss": 0.0225, + "step": 9080 + }, + { + "epoch": 0.9844969644405898, + "grad_norm": 2.537835121154785, + "learning_rate": 1.3485419991295518e-05, + "loss": 0.0738, + "step": 9081 + }, + { + "epoch": 0.9846053772766695, + "grad_norm": 3.0483052730560303, + "learning_rate": 1.3484694617728132e-05, + "loss": 0.037, + "step": 9082 + }, + { + "epoch": 0.9847137901127494, + "grad_norm": 1.3144201040267944, + "learning_rate": 1.3483969244160744e-05, + "loss": 0.0167, + "step": 9083 + }, + { + "epoch": 0.9848222029488292, + "grad_norm": 4.859297752380371, + "learning_rate": 1.3483243870593356e-05, + "loss": 0.0871, + "step": 9084 + }, + { + "epoch": 0.9849306157849089, + "grad_norm": 2.649836540222168, + "learning_rate": 1.348251849702597e-05, + "loss": 0.0262, + "step": 9085 + }, + { + "epoch": 0.9850390286209887, + "grad_norm": 1.2053687572479248, + "learning_rate": 1.3481793123458582e-05, + "loss": 0.0117, + "step": 9086 + }, + { + "epoch": 0.9851474414570686, + "grad_norm": 1.4305537939071655, + "learning_rate": 1.3481067749891194e-05, + "loss": 0.0158, + "step": 9087 + }, + { + "epoch": 0.9852558542931483, + "grad_norm": 4.179466724395752, + "learning_rate": 1.3480342376323808e-05, + "loss": 0.026, + "step": 9088 + }, + { + "epoch": 0.9853642671292281, + "grad_norm": 7.35892915725708, + "learning_rate": 1.347961700275642e-05, + "loss": 0.0478, + "step": 9089 + }, + { + "epoch": 0.9854726799653079, + "grad_norm": 3.1673946380615234, + "learning_rate": 1.3478891629189032e-05, + "loss": 0.0209, + "step": 9090 + }, + { + "epoch": 0.9855810928013877, + "grad_norm": 3.8302831649780273, + "learning_rate": 1.3478166255621646e-05, + "loss": 0.0617, + "step": 9091 + }, + { + "epoch": 0.9856895056374675, + "grad_norm": 1.001213788986206, + "learning_rate": 1.347744088205426e-05, + "loss": 0.0107, + "step": 9092 + }, + { + "epoch": 0.9857979184735473, + "grad_norm": 1.3959245681762695, + "learning_rate": 1.3476715508486872e-05, + "loss": 0.0213, + "step": 9093 + }, + { + "epoch": 0.985906331309627, + "grad_norm": 1.8189024925231934, + "learning_rate": 1.3475990134919486e-05, + "loss": 0.0381, + "step": 9094 + }, + { + "epoch": 0.9860147441457069, + "grad_norm": 1.3861052989959717, + "learning_rate": 1.3475264761352098e-05, + "loss": 0.0207, + "step": 9095 + }, + { + "epoch": 0.9861231569817867, + "grad_norm": 5.702711582183838, + "learning_rate": 1.347453938778471e-05, + "loss": 0.0753, + "step": 9096 + }, + { + "epoch": 0.9862315698178664, + "grad_norm": 1.5634264945983887, + "learning_rate": 1.3473814014217324e-05, + "loss": 0.0288, + "step": 9097 + }, + { + "epoch": 0.9863399826539462, + "grad_norm": 6.61164665222168, + "learning_rate": 1.3473088640649936e-05, + "loss": 0.0189, + "step": 9098 + }, + { + "epoch": 0.986448395490026, + "grad_norm": 2.368988513946533, + "learning_rate": 1.3472363267082548e-05, + "loss": 0.0623, + "step": 9099 + }, + { + "epoch": 0.9865568083261058, + "grad_norm": 0.3237082064151764, + "learning_rate": 1.3471637893515162e-05, + "loss": 0.0059, + "step": 9100 + }, + { + "epoch": 0.9866652211621856, + "grad_norm": 4.390634536743164, + "learning_rate": 1.3470912519947774e-05, + "loss": 0.0529, + "step": 9101 + }, + { + "epoch": 0.9867736339982653, + "grad_norm": 1.4066991806030273, + "learning_rate": 1.3470187146380386e-05, + "loss": 0.0176, + "step": 9102 + }, + { + "epoch": 0.9868820468343452, + "grad_norm": 2.2742910385131836, + "learning_rate": 1.3469461772813e-05, + "loss": 0.0347, + "step": 9103 + }, + { + "epoch": 0.986990459670425, + "grad_norm": 2.6876087188720703, + "learning_rate": 1.3468736399245612e-05, + "loss": 0.0342, + "step": 9104 + }, + { + "epoch": 0.9870988725065047, + "grad_norm": 1.855434775352478, + "learning_rate": 1.3468011025678224e-05, + "loss": 0.0123, + "step": 9105 + }, + { + "epoch": 0.9872072853425846, + "grad_norm": 8.665816307067871, + "learning_rate": 1.3467285652110838e-05, + "loss": 0.0207, + "step": 9106 + }, + { + "epoch": 0.9873156981786644, + "grad_norm": 1.539131760597229, + "learning_rate": 1.346656027854345e-05, + "loss": 0.0171, + "step": 9107 + }, + { + "epoch": 0.9874241110147441, + "grad_norm": 0.2535606622695923, + "learning_rate": 1.3465834904976063e-05, + "loss": 0.0028, + "step": 9108 + }, + { + "epoch": 0.9875325238508239, + "grad_norm": 3.1815569400787354, + "learning_rate": 1.3465109531408675e-05, + "loss": 0.0454, + "step": 9109 + }, + { + "epoch": 0.9876409366869038, + "grad_norm": 2.3288400173187256, + "learning_rate": 1.346438415784129e-05, + "loss": 0.0372, + "step": 9110 + }, + { + "epoch": 0.9877493495229835, + "grad_norm": 2.2758116722106934, + "learning_rate": 1.3463658784273902e-05, + "loss": 0.0343, + "step": 9111 + }, + { + "epoch": 0.9878577623590633, + "grad_norm": 6.095326900482178, + "learning_rate": 1.3462933410706516e-05, + "loss": 0.0343, + "step": 9112 + }, + { + "epoch": 0.9879661751951431, + "grad_norm": 2.2783310413360596, + "learning_rate": 1.3462208037139128e-05, + "loss": 0.065, + "step": 9113 + }, + { + "epoch": 0.988074588031223, + "grad_norm": 1.9057471752166748, + "learning_rate": 1.346148266357174e-05, + "loss": 0.0387, + "step": 9114 + }, + { + "epoch": 0.9881830008673027, + "grad_norm": 1.641620397567749, + "learning_rate": 1.3460757290004354e-05, + "loss": 0.0101, + "step": 9115 + }, + { + "epoch": 0.9882914137033825, + "grad_norm": 6.1596879959106445, + "learning_rate": 1.3460031916436967e-05, + "loss": 0.0173, + "step": 9116 + }, + { + "epoch": 0.9883998265394622, + "grad_norm": 3.486483335494995, + "learning_rate": 1.3459306542869579e-05, + "loss": 0.0556, + "step": 9117 + }, + { + "epoch": 0.9885082393755421, + "grad_norm": 3.3892481327056885, + "learning_rate": 1.3458581169302193e-05, + "loss": 0.0397, + "step": 9118 + }, + { + "epoch": 0.9886166522116219, + "grad_norm": 3.2889578342437744, + "learning_rate": 1.3457855795734805e-05, + "loss": 0.0447, + "step": 9119 + }, + { + "epoch": 0.9887250650477016, + "grad_norm": 4.3360114097595215, + "learning_rate": 1.3457130422167417e-05, + "loss": 0.0393, + "step": 9120 + }, + { + "epoch": 0.9888334778837814, + "grad_norm": 1.052935004234314, + "learning_rate": 1.3456405048600029e-05, + "loss": 0.013, + "step": 9121 + }, + { + "epoch": 0.9889418907198613, + "grad_norm": 3.293581008911133, + "learning_rate": 1.3455679675032643e-05, + "loss": 0.0124, + "step": 9122 + }, + { + "epoch": 0.989050303555941, + "grad_norm": 6.915004253387451, + "learning_rate": 1.3454954301465255e-05, + "loss": 0.043, + "step": 9123 + }, + { + "epoch": 0.9891587163920208, + "grad_norm": 2.3093314170837402, + "learning_rate": 1.3454228927897867e-05, + "loss": 0.0533, + "step": 9124 + }, + { + "epoch": 0.9892671292281006, + "grad_norm": 6.460171222686768, + "learning_rate": 1.3453503554330481e-05, + "loss": 0.0591, + "step": 9125 + }, + { + "epoch": 0.9893755420641804, + "grad_norm": 1.673353910446167, + "learning_rate": 1.3452778180763093e-05, + "loss": 0.0333, + "step": 9126 + }, + { + "epoch": 0.9894839549002602, + "grad_norm": 0.4139632284641266, + "learning_rate": 1.3452052807195709e-05, + "loss": 0.0044, + "step": 9127 + }, + { + "epoch": 0.98959236773634, + "grad_norm": 1.0554137229919434, + "learning_rate": 1.345132743362832e-05, + "loss": 0.0389, + "step": 9128 + }, + { + "epoch": 0.9897007805724197, + "grad_norm": 1.724918007850647, + "learning_rate": 1.3450602060060933e-05, + "loss": 0.0239, + "step": 9129 + }, + { + "epoch": 0.9898091934084996, + "grad_norm": 3.197364091873169, + "learning_rate": 1.3449876686493547e-05, + "loss": 0.0268, + "step": 9130 + }, + { + "epoch": 0.9899176062445794, + "grad_norm": 4.402266025543213, + "learning_rate": 1.3449151312926159e-05, + "loss": 0.0583, + "step": 9131 + }, + { + "epoch": 0.9900260190806591, + "grad_norm": 2.107455253601074, + "learning_rate": 1.3448425939358771e-05, + "loss": 0.0167, + "step": 9132 + }, + { + "epoch": 0.9901344319167389, + "grad_norm": 1.0654751062393188, + "learning_rate": 1.3447700565791383e-05, + "loss": 0.015, + "step": 9133 + }, + { + "epoch": 0.9902428447528188, + "grad_norm": 1.0876644849777222, + "learning_rate": 1.3446975192223997e-05, + "loss": 0.01, + "step": 9134 + }, + { + "epoch": 0.9903512575888985, + "grad_norm": 3.341845989227295, + "learning_rate": 1.3446249818656609e-05, + "loss": 0.0414, + "step": 9135 + }, + { + "epoch": 0.9904596704249783, + "grad_norm": 3.4141247272491455, + "learning_rate": 1.3445524445089221e-05, + "loss": 0.0492, + "step": 9136 + }, + { + "epoch": 0.9905680832610582, + "grad_norm": 3.1243896484375, + "learning_rate": 1.3444799071521835e-05, + "loss": 0.0527, + "step": 9137 + }, + { + "epoch": 0.9906764960971379, + "grad_norm": 0.732563316822052, + "learning_rate": 1.3444073697954447e-05, + "loss": 0.0223, + "step": 9138 + }, + { + "epoch": 0.9907849089332177, + "grad_norm": 4.1925482749938965, + "learning_rate": 1.344334832438706e-05, + "loss": 0.0403, + "step": 9139 + }, + { + "epoch": 0.9908933217692975, + "grad_norm": 0.5052802562713623, + "learning_rate": 1.3442622950819673e-05, + "loss": 0.0072, + "step": 9140 + }, + { + "epoch": 0.9910017346053773, + "grad_norm": 0.7651629447937012, + "learning_rate": 1.3441897577252285e-05, + "loss": 0.0058, + "step": 9141 + }, + { + "epoch": 0.9911101474414571, + "grad_norm": 1.9590442180633545, + "learning_rate": 1.3441172203684898e-05, + "loss": 0.0165, + "step": 9142 + }, + { + "epoch": 0.9912185602775369, + "grad_norm": 2.029169797897339, + "learning_rate": 1.3440446830117511e-05, + "loss": 0.0253, + "step": 9143 + }, + { + "epoch": 0.9913269731136166, + "grad_norm": 2.1641366481781006, + "learning_rate": 1.3439721456550125e-05, + "loss": 0.0208, + "step": 9144 + }, + { + "epoch": 0.9914353859496965, + "grad_norm": 2.7373147010803223, + "learning_rate": 1.3438996082982737e-05, + "loss": 0.0245, + "step": 9145 + }, + { + "epoch": 0.9915437987857763, + "grad_norm": 4.6439361572265625, + "learning_rate": 1.3438270709415351e-05, + "loss": 0.0428, + "step": 9146 + }, + { + "epoch": 0.991652211621856, + "grad_norm": 1.6654999256134033, + "learning_rate": 1.3437545335847963e-05, + "loss": 0.0206, + "step": 9147 + }, + { + "epoch": 0.9917606244579358, + "grad_norm": 3.7115981578826904, + "learning_rate": 1.3436819962280575e-05, + "loss": 0.0472, + "step": 9148 + }, + { + "epoch": 0.9918690372940157, + "grad_norm": 1.9577929973602295, + "learning_rate": 1.343609458871319e-05, + "loss": 0.052, + "step": 9149 + }, + { + "epoch": 0.9919774501300954, + "grad_norm": 3.6225574016571045, + "learning_rate": 1.3435369215145801e-05, + "loss": 0.0377, + "step": 9150 + }, + { + "epoch": 0.9920858629661752, + "grad_norm": 11.036279678344727, + "learning_rate": 1.3434643841578414e-05, + "loss": 0.0407, + "step": 9151 + }, + { + "epoch": 0.9921942758022549, + "grad_norm": 9.569491386413574, + "learning_rate": 1.3433918468011027e-05, + "loss": 0.0629, + "step": 9152 + }, + { + "epoch": 0.9923026886383348, + "grad_norm": 0.993134081363678, + "learning_rate": 1.343319309444364e-05, + "loss": 0.0063, + "step": 9153 + }, + { + "epoch": 0.9924111014744146, + "grad_norm": 3.3953940868377686, + "learning_rate": 1.3432467720876252e-05, + "loss": 0.0433, + "step": 9154 + }, + { + "epoch": 0.9925195143104943, + "grad_norm": 3.835908889770508, + "learning_rate": 1.3431742347308866e-05, + "loss": 0.0659, + "step": 9155 + }, + { + "epoch": 0.9926279271465741, + "grad_norm": 13.490124702453613, + "learning_rate": 1.3431016973741478e-05, + "loss": 0.044, + "step": 9156 + }, + { + "epoch": 0.992736339982654, + "grad_norm": 2.668637990951538, + "learning_rate": 1.343029160017409e-05, + "loss": 0.0389, + "step": 9157 + }, + { + "epoch": 0.9928447528187337, + "grad_norm": 1.6357649564743042, + "learning_rate": 1.3429566226606704e-05, + "loss": 0.0258, + "step": 9158 + }, + { + "epoch": 0.9929531656548135, + "grad_norm": 0.2717982828617096, + "learning_rate": 1.3428840853039316e-05, + "loss": 0.0034, + "step": 9159 + }, + { + "epoch": 0.9930615784908933, + "grad_norm": 3.476224899291992, + "learning_rate": 1.3428115479471928e-05, + "loss": 0.0136, + "step": 9160 + }, + { + "epoch": 0.9931699913269731, + "grad_norm": 2.3203699588775635, + "learning_rate": 1.342739010590454e-05, + "loss": 0.0185, + "step": 9161 + }, + { + "epoch": 0.9932784041630529, + "grad_norm": 1.8624248504638672, + "learning_rate": 1.3426664732337156e-05, + "loss": 0.0289, + "step": 9162 + }, + { + "epoch": 0.9933868169991327, + "grad_norm": 8.826269149780273, + "learning_rate": 1.3425939358769768e-05, + "loss": 0.0579, + "step": 9163 + }, + { + "epoch": 0.9934952298352124, + "grad_norm": 3.1335878372192383, + "learning_rate": 1.3425213985202382e-05, + "loss": 0.0432, + "step": 9164 + }, + { + "epoch": 0.9936036426712923, + "grad_norm": 3.6050238609313965, + "learning_rate": 1.3424488611634994e-05, + "loss": 0.0343, + "step": 9165 + }, + { + "epoch": 0.9937120555073721, + "grad_norm": 2.908069610595703, + "learning_rate": 1.3423763238067606e-05, + "loss": 0.0524, + "step": 9166 + }, + { + "epoch": 0.9938204683434518, + "grad_norm": 0.40847307443618774, + "learning_rate": 1.342303786450022e-05, + "loss": 0.0057, + "step": 9167 + }, + { + "epoch": 0.9939288811795317, + "grad_norm": 1.138314127922058, + "learning_rate": 1.3422312490932832e-05, + "loss": 0.0146, + "step": 9168 + }, + { + "epoch": 0.9940372940156115, + "grad_norm": 3.374288320541382, + "learning_rate": 1.3421587117365444e-05, + "loss": 0.0135, + "step": 9169 + }, + { + "epoch": 0.9941457068516912, + "grad_norm": 5.15782356262207, + "learning_rate": 1.3420861743798058e-05, + "loss": 0.0706, + "step": 9170 + }, + { + "epoch": 0.994254119687771, + "grad_norm": 2.7204458713531494, + "learning_rate": 1.342013637023067e-05, + "loss": 0.0539, + "step": 9171 + }, + { + "epoch": 0.9943625325238509, + "grad_norm": 3.4442903995513916, + "learning_rate": 1.3419410996663282e-05, + "loss": 0.0285, + "step": 9172 + }, + { + "epoch": 0.9944709453599306, + "grad_norm": 6.425852298736572, + "learning_rate": 1.3418685623095894e-05, + "loss": 0.0559, + "step": 9173 + }, + { + "epoch": 0.9945793581960104, + "grad_norm": 1.9708062410354614, + "learning_rate": 1.3417960249528508e-05, + "loss": 0.0262, + "step": 9174 + }, + { + "epoch": 0.9946877710320902, + "grad_norm": 3.817274808883667, + "learning_rate": 1.341723487596112e-05, + "loss": 0.0214, + "step": 9175 + }, + { + "epoch": 0.99479618386817, + "grad_norm": 2.2434146404266357, + "learning_rate": 1.3416509502393732e-05, + "loss": 0.0819, + "step": 9176 + }, + { + "epoch": 0.9949045967042498, + "grad_norm": 2.1082992553710938, + "learning_rate": 1.3415784128826346e-05, + "loss": 0.0087, + "step": 9177 + }, + { + "epoch": 0.9950130095403296, + "grad_norm": 2.578388214111328, + "learning_rate": 1.3415058755258958e-05, + "loss": 0.0382, + "step": 9178 + }, + { + "epoch": 0.9951214223764093, + "grad_norm": 1.272395372390747, + "learning_rate": 1.3414333381691574e-05, + "loss": 0.0252, + "step": 9179 + }, + { + "epoch": 0.9952298352124892, + "grad_norm": 0.3004473149776459, + "learning_rate": 1.3413608008124186e-05, + "loss": 0.0048, + "step": 9180 + }, + { + "epoch": 0.995338248048569, + "grad_norm": 3.923668622970581, + "learning_rate": 1.3412882634556798e-05, + "loss": 0.0624, + "step": 9181 + }, + { + "epoch": 0.9954466608846487, + "grad_norm": 5.748037815093994, + "learning_rate": 1.3412157260989412e-05, + "loss": 0.0657, + "step": 9182 + }, + { + "epoch": 0.9955550737207285, + "grad_norm": 1.627102017402649, + "learning_rate": 1.3411431887422024e-05, + "loss": 0.0331, + "step": 9183 + }, + { + "epoch": 0.9956634865568084, + "grad_norm": 0.7357792258262634, + "learning_rate": 1.3410706513854636e-05, + "loss": 0.0111, + "step": 9184 + }, + { + "epoch": 0.9957718993928881, + "grad_norm": 5.3691840171813965, + "learning_rate": 1.3409981140287248e-05, + "loss": 0.0476, + "step": 9185 + }, + { + "epoch": 0.9958803122289679, + "grad_norm": 3.214963674545288, + "learning_rate": 1.3409255766719862e-05, + "loss": 0.0321, + "step": 9186 + }, + { + "epoch": 0.9959887250650477, + "grad_norm": 0.7253509163856506, + "learning_rate": 1.3408530393152474e-05, + "loss": 0.0178, + "step": 9187 + }, + { + "epoch": 0.9960971379011275, + "grad_norm": 1.5984035730361938, + "learning_rate": 1.3407805019585087e-05, + "loss": 0.024, + "step": 9188 + }, + { + "epoch": 0.9962055507372073, + "grad_norm": 4.211014747619629, + "learning_rate": 1.34070796460177e-05, + "loss": 0.0532, + "step": 9189 + }, + { + "epoch": 0.996313963573287, + "grad_norm": 3.602658271789551, + "learning_rate": 1.3406354272450313e-05, + "loss": 0.0448, + "step": 9190 + }, + { + "epoch": 0.9964223764093668, + "grad_norm": 2.150390386581421, + "learning_rate": 1.3405628898882925e-05, + "loss": 0.028, + "step": 9191 + }, + { + "epoch": 0.9965307892454467, + "grad_norm": 1.0745795965194702, + "learning_rate": 1.3404903525315539e-05, + "loss": 0.0147, + "step": 9192 + }, + { + "epoch": 0.9966392020815265, + "grad_norm": 1.4816557168960571, + "learning_rate": 1.340417815174815e-05, + "loss": 0.0433, + "step": 9193 + }, + { + "epoch": 0.9967476149176062, + "grad_norm": 2.670391082763672, + "learning_rate": 1.3403452778180763e-05, + "loss": 0.0263, + "step": 9194 + }, + { + "epoch": 0.9968560277536861, + "grad_norm": 1.8945658206939697, + "learning_rate": 1.3402727404613377e-05, + "loss": 0.0204, + "step": 9195 + }, + { + "epoch": 0.9969644405897659, + "grad_norm": 2.731471538543701, + "learning_rate": 1.340200203104599e-05, + "loss": 0.0122, + "step": 9196 + }, + { + "epoch": 0.9970728534258456, + "grad_norm": 6.528893947601318, + "learning_rate": 1.3401276657478603e-05, + "loss": 0.0483, + "step": 9197 + }, + { + "epoch": 0.9971812662619254, + "grad_norm": 5.012412071228027, + "learning_rate": 1.3400551283911217e-05, + "loss": 0.0337, + "step": 9198 + }, + { + "epoch": 0.9972896790980053, + "grad_norm": 3.0206644535064697, + "learning_rate": 1.3399825910343829e-05, + "loss": 0.0171, + "step": 9199 + }, + { + "epoch": 0.997398091934085, + "grad_norm": 2.1890530586242676, + "learning_rate": 1.339910053677644e-05, + "loss": 0.0591, + "step": 9200 + }, + { + "epoch": 0.9975065047701648, + "grad_norm": 4.991154193878174, + "learning_rate": 1.3398375163209055e-05, + "loss": 0.032, + "step": 9201 + }, + { + "epoch": 0.9976149176062445, + "grad_norm": 1.3406016826629639, + "learning_rate": 1.3397649789641667e-05, + "loss": 0.0083, + "step": 9202 + }, + { + "epoch": 0.9977233304423244, + "grad_norm": 3.8390374183654785, + "learning_rate": 1.3396924416074279e-05, + "loss": 0.0557, + "step": 9203 + }, + { + "epoch": 0.9978317432784042, + "grad_norm": 5.421045780181885, + "learning_rate": 1.3396199042506893e-05, + "loss": 0.0293, + "step": 9204 + }, + { + "epoch": 0.997940156114484, + "grad_norm": 2.3463854789733887, + "learning_rate": 1.3395473668939505e-05, + "loss": 0.0218, + "step": 9205 + }, + { + "epoch": 0.9980485689505637, + "grad_norm": 1.5904332399368286, + "learning_rate": 1.3394748295372117e-05, + "loss": 0.0393, + "step": 9206 + }, + { + "epoch": 0.9981569817866436, + "grad_norm": 2.5820558071136475, + "learning_rate": 1.3394022921804731e-05, + "loss": 0.0369, + "step": 9207 + }, + { + "epoch": 0.9982653946227233, + "grad_norm": 3.2563912868499756, + "learning_rate": 1.3393297548237343e-05, + "loss": 0.044, + "step": 9208 + }, + { + "epoch": 0.9983738074588031, + "grad_norm": 6.894834041595459, + "learning_rate": 1.3392572174669955e-05, + "loss": 0.0225, + "step": 9209 + }, + { + "epoch": 0.9984822202948829, + "grad_norm": 0.6876869797706604, + "learning_rate": 1.3391846801102569e-05, + "loss": 0.0102, + "step": 9210 + }, + { + "epoch": 0.9985906331309627, + "grad_norm": 2.142308473587036, + "learning_rate": 1.3391121427535181e-05, + "loss": 0.0455, + "step": 9211 + }, + { + "epoch": 0.9986990459670425, + "grad_norm": 2.0169217586517334, + "learning_rate": 1.3390396053967793e-05, + "loss": 0.0337, + "step": 9212 + }, + { + "epoch": 0.9988074588031223, + "grad_norm": 1.7866837978363037, + "learning_rate": 1.3389670680400409e-05, + "loss": 0.0154, + "step": 9213 + }, + { + "epoch": 0.998915871639202, + "grad_norm": 2.7004573345184326, + "learning_rate": 1.3388945306833021e-05, + "loss": 0.0459, + "step": 9214 + }, + { + "epoch": 0.9990242844752819, + "grad_norm": 4.470727443695068, + "learning_rate": 1.3388219933265633e-05, + "loss": 0.0251, + "step": 9215 + }, + { + "epoch": 0.9991326973113617, + "grad_norm": 3.3358447551727295, + "learning_rate": 1.3387494559698247e-05, + "loss": 0.0446, + "step": 9216 + }, + { + "epoch": 0.9992411101474414, + "grad_norm": 4.154749870300293, + "learning_rate": 1.3386769186130859e-05, + "loss": 0.0676, + "step": 9217 + }, + { + "epoch": 0.9993495229835212, + "grad_norm": 2.06245493888855, + "learning_rate": 1.3386043812563471e-05, + "loss": 0.0093, + "step": 9218 + }, + { + "epoch": 0.9994579358196011, + "grad_norm": 1.3406084775924683, + "learning_rate": 1.3385318438996085e-05, + "loss": 0.0181, + "step": 9219 + }, + { + "epoch": 0.9995663486556808, + "grad_norm": 0.8761276602745056, + "learning_rate": 1.3384593065428697e-05, + "loss": 0.0174, + "step": 9220 + }, + { + "epoch": 0.9996747614917606, + "grad_norm": 1.6961619853973389, + "learning_rate": 1.338386769186131e-05, + "loss": 0.04, + "step": 9221 + }, + { + "epoch": 0.9997831743278404, + "grad_norm": 1.3865456581115723, + "learning_rate": 1.3383142318293923e-05, + "loss": 0.0229, + "step": 9222 + }, + { + "epoch": 0.9998915871639202, + "grad_norm": 5.342336654663086, + "learning_rate": 1.3382416944726535e-05, + "loss": 0.0473, + "step": 9223 + }, + { + "epoch": 1.0, + "grad_norm": 3.263672351837158, + "learning_rate": 1.3381691571159147e-05, + "loss": 0.0298, + "step": 9224 + }, + { + "epoch": 1.0001084128360798, + "grad_norm": 2.090317487716675, + "learning_rate": 1.338096619759176e-05, + "loss": 0.0216, + "step": 9225 + }, + { + "epoch": 1.0002168256721595, + "grad_norm": 0.43028223514556885, + "learning_rate": 1.3380240824024373e-05, + "loss": 0.0052, + "step": 9226 + }, + { + "epoch": 1.0003252385082393, + "grad_norm": 0.7791002988815308, + "learning_rate": 1.3379515450456986e-05, + "loss": 0.0078, + "step": 9227 + }, + { + "epoch": 1.0004336513443193, + "grad_norm": 1.959311604499817, + "learning_rate": 1.3378790076889598e-05, + "loss": 0.0399, + "step": 9228 + }, + { + "epoch": 1.000542064180399, + "grad_norm": 0.8919258117675781, + "learning_rate": 1.3378064703322212e-05, + "loss": 0.0114, + "step": 9229 + }, + { + "epoch": 1.0006504770164788, + "grad_norm": 1.9426615238189697, + "learning_rate": 1.3377339329754824e-05, + "loss": 0.0208, + "step": 9230 + }, + { + "epoch": 1.0007588898525586, + "grad_norm": 0.4217337667942047, + "learning_rate": 1.337661395618744e-05, + "loss": 0.0051, + "step": 9231 + }, + { + "epoch": 1.0008673026886383, + "grad_norm": 1.8073465824127197, + "learning_rate": 1.3375888582620051e-05, + "loss": 0.0246, + "step": 9232 + }, + { + "epoch": 1.000975715524718, + "grad_norm": 1.54167640209198, + "learning_rate": 1.3375163209052664e-05, + "loss": 0.0255, + "step": 9233 + }, + { + "epoch": 1.0010841283607979, + "grad_norm": 2.613482713699341, + "learning_rate": 1.3374437835485277e-05, + "loss": 0.0245, + "step": 9234 + }, + { + "epoch": 1.0011925411968776, + "grad_norm": 1.201854944229126, + "learning_rate": 1.337371246191789e-05, + "loss": 0.0141, + "step": 9235 + }, + { + "epoch": 1.0013009540329576, + "grad_norm": 2.6634747982025146, + "learning_rate": 1.3372987088350502e-05, + "loss": 0.0394, + "step": 9236 + }, + { + "epoch": 1.0014093668690374, + "grad_norm": 1.2729456424713135, + "learning_rate": 1.3372261714783114e-05, + "loss": 0.0146, + "step": 9237 + }, + { + "epoch": 1.0015177797051171, + "grad_norm": 11.94725513458252, + "learning_rate": 1.3371536341215728e-05, + "loss": 0.0331, + "step": 9238 + }, + { + "epoch": 1.001626192541197, + "grad_norm": 1.2996151447296143, + "learning_rate": 1.337081096764834e-05, + "loss": 0.0135, + "step": 9239 + }, + { + "epoch": 1.0017346053772767, + "grad_norm": 4.38715934753418, + "learning_rate": 1.3370085594080952e-05, + "loss": 0.0922, + "step": 9240 + }, + { + "epoch": 1.0018430182133564, + "grad_norm": 1.532747745513916, + "learning_rate": 1.3369360220513566e-05, + "loss": 0.0169, + "step": 9241 + }, + { + "epoch": 1.0019514310494362, + "grad_norm": 5.177864074707031, + "learning_rate": 1.3368634846946178e-05, + "loss": 0.0229, + "step": 9242 + }, + { + "epoch": 1.002059843885516, + "grad_norm": 1.5488766431808472, + "learning_rate": 1.336790947337879e-05, + "loss": 0.0199, + "step": 9243 + }, + { + "epoch": 1.002168256721596, + "grad_norm": 1.940401554107666, + "learning_rate": 1.3367184099811404e-05, + "loss": 0.0421, + "step": 9244 + }, + { + "epoch": 1.0022766695576757, + "grad_norm": 2.1750566959381104, + "learning_rate": 1.3366458726244016e-05, + "loss": 0.0572, + "step": 9245 + }, + { + "epoch": 1.0023850823937555, + "grad_norm": 2.2152082920074463, + "learning_rate": 1.3365733352676628e-05, + "loss": 0.0466, + "step": 9246 + }, + { + "epoch": 1.0024934952298352, + "grad_norm": 0.46551159024238586, + "learning_rate": 1.3365007979109242e-05, + "loss": 0.0064, + "step": 9247 + }, + { + "epoch": 1.002601908065915, + "grad_norm": 1.347625494003296, + "learning_rate": 1.3364282605541856e-05, + "loss": 0.0217, + "step": 9248 + }, + { + "epoch": 1.0027103209019947, + "grad_norm": 6.930816173553467, + "learning_rate": 1.336355723197447e-05, + "loss": 0.0145, + "step": 9249 + }, + { + "epoch": 1.0028187337380745, + "grad_norm": 1.5547314882278442, + "learning_rate": 1.3362831858407082e-05, + "loss": 0.0369, + "step": 9250 + }, + { + "epoch": 1.0029271465741543, + "grad_norm": 0.9879695773124695, + "learning_rate": 1.3362106484839694e-05, + "loss": 0.011, + "step": 9251 + }, + { + "epoch": 1.0030355594102343, + "grad_norm": 2.972745656967163, + "learning_rate": 1.3361381111272306e-05, + "loss": 0.0291, + "step": 9252 + }, + { + "epoch": 1.003143972246314, + "grad_norm": 2.657994031906128, + "learning_rate": 1.336065573770492e-05, + "loss": 0.0435, + "step": 9253 + }, + { + "epoch": 1.0032523850823938, + "grad_norm": 0.25849926471710205, + "learning_rate": 1.3359930364137532e-05, + "loss": 0.0045, + "step": 9254 + }, + { + "epoch": 1.0033607979184735, + "grad_norm": 1.921229362487793, + "learning_rate": 1.3359204990570144e-05, + "loss": 0.0112, + "step": 9255 + }, + { + "epoch": 1.0034692107545533, + "grad_norm": 7.290126800537109, + "learning_rate": 1.3358479617002758e-05, + "loss": 0.0227, + "step": 9256 + }, + { + "epoch": 1.003577623590633, + "grad_norm": 2.901625633239746, + "learning_rate": 1.335775424343537e-05, + "loss": 0.0331, + "step": 9257 + }, + { + "epoch": 1.0036860364267128, + "grad_norm": 1.3846831321716309, + "learning_rate": 1.3357028869867982e-05, + "loss": 0.0318, + "step": 9258 + }, + { + "epoch": 1.0037944492627928, + "grad_norm": 3.9267876148223877, + "learning_rate": 1.3356303496300596e-05, + "loss": 0.0453, + "step": 9259 + }, + { + "epoch": 1.0039028620988726, + "grad_norm": 2.269362211227417, + "learning_rate": 1.3355578122733208e-05, + "loss": 0.0334, + "step": 9260 + }, + { + "epoch": 1.0040112749349523, + "grad_norm": 1.5713984966278076, + "learning_rate": 1.335485274916582e-05, + "loss": 0.0274, + "step": 9261 + }, + { + "epoch": 1.004119687771032, + "grad_norm": 1.1960965394973755, + "learning_rate": 1.3354127375598434e-05, + "loss": 0.0392, + "step": 9262 + }, + { + "epoch": 1.0042281006071119, + "grad_norm": 0.42447414994239807, + "learning_rate": 1.3353402002031046e-05, + "loss": 0.0056, + "step": 9263 + }, + { + "epoch": 1.0043365134431916, + "grad_norm": 0.2571236491203308, + "learning_rate": 1.3352676628463659e-05, + "loss": 0.0033, + "step": 9264 + }, + { + "epoch": 1.0044449262792714, + "grad_norm": 0.44083210825920105, + "learning_rate": 1.3351951254896274e-05, + "loss": 0.0027, + "step": 9265 + }, + { + "epoch": 1.0045533391153512, + "grad_norm": 1.9022506475448608, + "learning_rate": 1.3351225881328886e-05, + "loss": 0.0188, + "step": 9266 + }, + { + "epoch": 1.0046617519514311, + "grad_norm": 2.8218994140625, + "learning_rate": 1.3350500507761498e-05, + "loss": 0.0304, + "step": 9267 + }, + { + "epoch": 1.004770164787511, + "grad_norm": 3.183429002761841, + "learning_rate": 1.3349775134194112e-05, + "loss": 0.0503, + "step": 9268 + }, + { + "epoch": 1.0048785776235907, + "grad_norm": 1.9211933612823486, + "learning_rate": 1.3349049760626724e-05, + "loss": 0.044, + "step": 9269 + }, + { + "epoch": 1.0049869904596704, + "grad_norm": 3.6063175201416016, + "learning_rate": 1.3348324387059337e-05, + "loss": 0.0248, + "step": 9270 + }, + { + "epoch": 1.0050954032957502, + "grad_norm": 0.6688576340675354, + "learning_rate": 1.334759901349195e-05, + "loss": 0.0039, + "step": 9271 + }, + { + "epoch": 1.00520381613183, + "grad_norm": 0.9018802642822266, + "learning_rate": 1.3346873639924563e-05, + "loss": 0.0059, + "step": 9272 + }, + { + "epoch": 1.0053122289679097, + "grad_norm": 1.6995795965194702, + "learning_rate": 1.3346148266357175e-05, + "loss": 0.0334, + "step": 9273 + }, + { + "epoch": 1.0054206418039895, + "grad_norm": 7.564227104187012, + "learning_rate": 1.3345422892789789e-05, + "loss": 0.0134, + "step": 9274 + }, + { + "epoch": 1.0055290546400695, + "grad_norm": 1.0658785104751587, + "learning_rate": 1.33446975192224e-05, + "loss": 0.012, + "step": 9275 + }, + { + "epoch": 1.0056374674761492, + "grad_norm": 0.9185298085212708, + "learning_rate": 1.3343972145655013e-05, + "loss": 0.0133, + "step": 9276 + }, + { + "epoch": 1.005745880312229, + "grad_norm": 2.6648733615875244, + "learning_rate": 1.3343246772087625e-05, + "loss": 0.0124, + "step": 9277 + }, + { + "epoch": 1.0058542931483088, + "grad_norm": 1.2241474390029907, + "learning_rate": 1.3342521398520239e-05, + "loss": 0.0206, + "step": 9278 + }, + { + "epoch": 1.0059627059843885, + "grad_norm": 2.9929075241088867, + "learning_rate": 1.3341796024952851e-05, + "loss": 0.015, + "step": 9279 + }, + { + "epoch": 1.0060711188204683, + "grad_norm": 0.23785929381847382, + "learning_rate": 1.3341070651385463e-05, + "loss": 0.0038, + "step": 9280 + }, + { + "epoch": 1.006179531656548, + "grad_norm": 4.799199104309082, + "learning_rate": 1.3340345277818077e-05, + "loss": 0.0964, + "step": 9281 + }, + { + "epoch": 1.0062879444926278, + "grad_norm": 0.2377571016550064, + "learning_rate": 1.3339619904250689e-05, + "loss": 0.0036, + "step": 9282 + }, + { + "epoch": 1.0063963573287078, + "grad_norm": 2.28068470954895, + "learning_rate": 1.3338894530683305e-05, + "loss": 0.0525, + "step": 9283 + }, + { + "epoch": 1.0065047701647876, + "grad_norm": 3.9727320671081543, + "learning_rate": 1.3338169157115917e-05, + "loss": 0.0717, + "step": 9284 + }, + { + "epoch": 1.0066131830008673, + "grad_norm": 0.14648646116256714, + "learning_rate": 1.3337443783548529e-05, + "loss": 0.0016, + "step": 9285 + }, + { + "epoch": 1.006721595836947, + "grad_norm": 0.41646096110343933, + "learning_rate": 1.3336718409981143e-05, + "loss": 0.0056, + "step": 9286 + }, + { + "epoch": 1.0068300086730269, + "grad_norm": 0.8322235345840454, + "learning_rate": 1.3335993036413755e-05, + "loss": 0.0136, + "step": 9287 + }, + { + "epoch": 1.0069384215091066, + "grad_norm": 1.5413799285888672, + "learning_rate": 1.3335267662846367e-05, + "loss": 0.0111, + "step": 9288 + }, + { + "epoch": 1.0070468343451864, + "grad_norm": 1.8780972957611084, + "learning_rate": 1.333454228927898e-05, + "loss": 0.0267, + "step": 9289 + }, + { + "epoch": 1.0071552471812664, + "grad_norm": 2.910874366760254, + "learning_rate": 1.3333816915711593e-05, + "loss": 0.0331, + "step": 9290 + }, + { + "epoch": 1.0072636600173461, + "grad_norm": 3.0940213203430176, + "learning_rate": 1.3333091542144205e-05, + "loss": 0.0383, + "step": 9291 + }, + { + "epoch": 1.007372072853426, + "grad_norm": 3.5723330974578857, + "learning_rate": 1.3332366168576817e-05, + "loss": 0.0215, + "step": 9292 + }, + { + "epoch": 1.0074804856895057, + "grad_norm": 3.595047950744629, + "learning_rate": 1.3331640795009431e-05, + "loss": 0.051, + "step": 9293 + }, + { + "epoch": 1.0075888985255854, + "grad_norm": 4.597019195556641, + "learning_rate": 1.3330915421442043e-05, + "loss": 0.0681, + "step": 9294 + }, + { + "epoch": 1.0076973113616652, + "grad_norm": 3.2451255321502686, + "learning_rate": 1.3330190047874655e-05, + "loss": 0.0729, + "step": 9295 + }, + { + "epoch": 1.007805724197745, + "grad_norm": 2.9106082916259766, + "learning_rate": 1.332946467430727e-05, + "loss": 0.0198, + "step": 9296 + }, + { + "epoch": 1.0079141370338247, + "grad_norm": 0.22642049193382263, + "learning_rate": 1.3328739300739881e-05, + "loss": 0.0024, + "step": 9297 + }, + { + "epoch": 1.0080225498699047, + "grad_norm": 1.4692795276641846, + "learning_rate": 1.3328013927172494e-05, + "loss": 0.0182, + "step": 9298 + }, + { + "epoch": 1.0081309627059845, + "grad_norm": 4.1376447677612305, + "learning_rate": 1.3327288553605107e-05, + "loss": 0.0399, + "step": 9299 + }, + { + "epoch": 1.0082393755420642, + "grad_norm": 0.21464093029499054, + "learning_rate": 1.3326563180037721e-05, + "loss": 0.0021, + "step": 9300 + }, + { + "epoch": 1.008347788378144, + "grad_norm": 1.187662959098816, + "learning_rate": 1.3325837806470335e-05, + "loss": 0.0151, + "step": 9301 + }, + { + "epoch": 1.0084562012142237, + "grad_norm": 1.0252171754837036, + "learning_rate": 1.3325112432902947e-05, + "loss": 0.0162, + "step": 9302 + }, + { + "epoch": 1.0085646140503035, + "grad_norm": 1.1953128576278687, + "learning_rate": 1.332438705933556e-05, + "loss": 0.0069, + "step": 9303 + }, + { + "epoch": 1.0086730268863833, + "grad_norm": 3.010662317276001, + "learning_rate": 1.3323661685768171e-05, + "loss": 0.0147, + "step": 9304 + }, + { + "epoch": 1.008781439722463, + "grad_norm": 0.8814054727554321, + "learning_rate": 1.3322936312200785e-05, + "loss": 0.0068, + "step": 9305 + }, + { + "epoch": 1.008889852558543, + "grad_norm": 1.7111115455627441, + "learning_rate": 1.3322210938633397e-05, + "loss": 0.0257, + "step": 9306 + }, + { + "epoch": 1.0089982653946228, + "grad_norm": 2.41446590423584, + "learning_rate": 1.332148556506601e-05, + "loss": 0.0201, + "step": 9307 + }, + { + "epoch": 1.0091066782307025, + "grad_norm": 1.7674806118011475, + "learning_rate": 1.3320760191498623e-05, + "loss": 0.0069, + "step": 9308 + }, + { + "epoch": 1.0092150910667823, + "grad_norm": 2.2057340145111084, + "learning_rate": 1.3320034817931236e-05, + "loss": 0.0259, + "step": 9309 + }, + { + "epoch": 1.009323503902862, + "grad_norm": 4.41972017288208, + "learning_rate": 1.3319309444363848e-05, + "loss": 0.0278, + "step": 9310 + }, + { + "epoch": 1.0094319167389418, + "grad_norm": 6.951976776123047, + "learning_rate": 1.3318584070796462e-05, + "loss": 0.0134, + "step": 9311 + }, + { + "epoch": 1.0095403295750216, + "grad_norm": 6.6200079917907715, + "learning_rate": 1.3317858697229074e-05, + "loss": 0.0339, + "step": 9312 + }, + { + "epoch": 1.0096487424111014, + "grad_norm": 1.1984825134277344, + "learning_rate": 1.3317133323661686e-05, + "loss": 0.0123, + "step": 9313 + }, + { + "epoch": 1.0097571552471813, + "grad_norm": 0.6427276134490967, + "learning_rate": 1.33164079500943e-05, + "loss": 0.0077, + "step": 9314 + }, + { + "epoch": 1.0098655680832611, + "grad_norm": 1.6934797763824463, + "learning_rate": 1.3315682576526912e-05, + "loss": 0.0519, + "step": 9315 + }, + { + "epoch": 1.0099739809193409, + "grad_norm": 4.131855010986328, + "learning_rate": 1.3314957202959524e-05, + "loss": 0.0348, + "step": 9316 + }, + { + "epoch": 1.0100823937554206, + "grad_norm": 2.315430164337158, + "learning_rate": 1.331423182939214e-05, + "loss": 0.0242, + "step": 9317 + }, + { + "epoch": 1.0101908065915004, + "grad_norm": 0.9566731452941895, + "learning_rate": 1.3313506455824752e-05, + "loss": 0.0255, + "step": 9318 + }, + { + "epoch": 1.0102992194275802, + "grad_norm": 2.480764865875244, + "learning_rate": 1.3312781082257364e-05, + "loss": 0.0247, + "step": 9319 + }, + { + "epoch": 1.01040763226366, + "grad_norm": 1.554291844367981, + "learning_rate": 1.3312055708689978e-05, + "loss": 0.0213, + "step": 9320 + }, + { + "epoch": 1.01051604509974, + "grad_norm": 3.870537519454956, + "learning_rate": 1.331133033512259e-05, + "loss": 0.0712, + "step": 9321 + }, + { + "epoch": 1.0106244579358197, + "grad_norm": 0.9166957139968872, + "learning_rate": 1.3310604961555202e-05, + "loss": 0.0087, + "step": 9322 + }, + { + "epoch": 1.0107328707718994, + "grad_norm": 2.2351293563842773, + "learning_rate": 1.3309879587987816e-05, + "loss": 0.0133, + "step": 9323 + }, + { + "epoch": 1.0108412836079792, + "grad_norm": 2.537343740463257, + "learning_rate": 1.3309154214420428e-05, + "loss": 0.0345, + "step": 9324 + }, + { + "epoch": 1.010949696444059, + "grad_norm": 3.626647472381592, + "learning_rate": 1.330842884085304e-05, + "loss": 0.0253, + "step": 9325 + }, + { + "epoch": 1.0110581092801387, + "grad_norm": 1.104966640472412, + "learning_rate": 1.3307703467285654e-05, + "loss": 0.0222, + "step": 9326 + }, + { + "epoch": 1.0111665221162185, + "grad_norm": 1.6430498361587524, + "learning_rate": 1.3306978093718266e-05, + "loss": 0.0265, + "step": 9327 + }, + { + "epoch": 1.0112749349522983, + "grad_norm": 6.614035606384277, + "learning_rate": 1.3306252720150878e-05, + "loss": 0.0581, + "step": 9328 + }, + { + "epoch": 1.0113833477883782, + "grad_norm": 2.0080623626708984, + "learning_rate": 1.330552734658349e-05, + "loss": 0.0242, + "step": 9329 + }, + { + "epoch": 1.011491760624458, + "grad_norm": 1.0575305223464966, + "learning_rate": 1.3304801973016104e-05, + "loss": 0.0144, + "step": 9330 + }, + { + "epoch": 1.0116001734605378, + "grad_norm": 1.8413658142089844, + "learning_rate": 1.3304076599448716e-05, + "loss": 0.0139, + "step": 9331 + }, + { + "epoch": 1.0117085862966175, + "grad_norm": 2.9649577140808105, + "learning_rate": 1.3303351225881328e-05, + "loss": 0.0216, + "step": 9332 + }, + { + "epoch": 1.0118169991326973, + "grad_norm": 7.4007248878479, + "learning_rate": 1.3302625852313942e-05, + "loss": 0.0227, + "step": 9333 + }, + { + "epoch": 1.011925411968777, + "grad_norm": 1.690480351448059, + "learning_rate": 1.3301900478746556e-05, + "loss": 0.015, + "step": 9334 + }, + { + "epoch": 1.0120338248048568, + "grad_norm": 4.039403438568115, + "learning_rate": 1.330117510517917e-05, + "loss": 0.0219, + "step": 9335 + }, + { + "epoch": 1.0121422376409366, + "grad_norm": 0.4893222451210022, + "learning_rate": 1.3300449731611782e-05, + "loss": 0.008, + "step": 9336 + }, + { + "epoch": 1.0122506504770166, + "grad_norm": 0.35159382224082947, + "learning_rate": 1.3299724358044394e-05, + "loss": 0.0041, + "step": 9337 + }, + { + "epoch": 1.0123590633130963, + "grad_norm": 1.569234013557434, + "learning_rate": 1.3298998984477008e-05, + "loss": 0.0259, + "step": 9338 + }, + { + "epoch": 1.012467476149176, + "grad_norm": 1.7654422521591187, + "learning_rate": 1.329827361090962e-05, + "loss": 0.02, + "step": 9339 + }, + { + "epoch": 1.0125758889852559, + "grad_norm": 1.7162877321243286, + "learning_rate": 1.3297548237342232e-05, + "loss": 0.0091, + "step": 9340 + }, + { + "epoch": 1.0126843018213356, + "grad_norm": 0.8302964568138123, + "learning_rate": 1.3296822863774844e-05, + "loss": 0.0103, + "step": 9341 + }, + { + "epoch": 1.0127927146574154, + "grad_norm": 0.5845112204551697, + "learning_rate": 1.3296097490207458e-05, + "loss": 0.0025, + "step": 9342 + }, + { + "epoch": 1.0129011274934951, + "grad_norm": 1.3628607988357544, + "learning_rate": 1.329537211664007e-05, + "loss": 0.0217, + "step": 9343 + }, + { + "epoch": 1.0130095403295751, + "grad_norm": 2.8496253490448, + "learning_rate": 1.3294646743072683e-05, + "loss": 0.0162, + "step": 9344 + }, + { + "epoch": 1.013117953165655, + "grad_norm": 0.289695143699646, + "learning_rate": 1.3293921369505296e-05, + "loss": 0.0024, + "step": 9345 + }, + { + "epoch": 1.0132263660017347, + "grad_norm": 3.9596996307373047, + "learning_rate": 1.3293195995937909e-05, + "loss": 0.0227, + "step": 9346 + }, + { + "epoch": 1.0133347788378144, + "grad_norm": 1.6275138854980469, + "learning_rate": 1.329247062237052e-05, + "loss": 0.022, + "step": 9347 + }, + { + "epoch": 1.0134431916738942, + "grad_norm": 2.7896740436553955, + "learning_rate": 1.3291745248803135e-05, + "loss": 0.0285, + "step": 9348 + }, + { + "epoch": 1.013551604509974, + "grad_norm": 3.313952922821045, + "learning_rate": 1.3291019875235747e-05, + "loss": 0.0422, + "step": 9349 + }, + { + "epoch": 1.0136600173460537, + "grad_norm": 2.886186122894287, + "learning_rate": 1.3290294501668359e-05, + "loss": 0.0203, + "step": 9350 + }, + { + "epoch": 1.0137684301821335, + "grad_norm": 0.8949986696243286, + "learning_rate": 1.3289569128100973e-05, + "loss": 0.0113, + "step": 9351 + }, + { + "epoch": 1.0138768430182135, + "grad_norm": 4.485427379608154, + "learning_rate": 1.3288843754533587e-05, + "loss": 0.0275, + "step": 9352 + }, + { + "epoch": 1.0139852558542932, + "grad_norm": 5.84232759475708, + "learning_rate": 1.32881183809662e-05, + "loss": 0.0792, + "step": 9353 + }, + { + "epoch": 1.014093668690373, + "grad_norm": 1.873491883277893, + "learning_rate": 1.3287393007398813e-05, + "loss": 0.0046, + "step": 9354 + }, + { + "epoch": 1.0142020815264527, + "grad_norm": 4.983179092407227, + "learning_rate": 1.3286667633831425e-05, + "loss": 0.0439, + "step": 9355 + }, + { + "epoch": 1.0143104943625325, + "grad_norm": 4.300745010375977, + "learning_rate": 1.3285942260264037e-05, + "loss": 0.0226, + "step": 9356 + }, + { + "epoch": 1.0144189071986123, + "grad_norm": 2.8750970363616943, + "learning_rate": 1.328521688669665e-05, + "loss": 0.0054, + "step": 9357 + }, + { + "epoch": 1.014527320034692, + "grad_norm": 1.0465267896652222, + "learning_rate": 1.3284491513129263e-05, + "loss": 0.007, + "step": 9358 + }, + { + "epoch": 1.0146357328707718, + "grad_norm": 4.427806854248047, + "learning_rate": 1.3283766139561875e-05, + "loss": 0.0747, + "step": 9359 + }, + { + "epoch": 1.0147441457068518, + "grad_norm": 2.565159320831299, + "learning_rate": 1.3283040765994489e-05, + "loss": 0.0143, + "step": 9360 + }, + { + "epoch": 1.0148525585429315, + "grad_norm": 1.0939221382141113, + "learning_rate": 1.3282315392427101e-05, + "loss": 0.004, + "step": 9361 + }, + { + "epoch": 1.0149609713790113, + "grad_norm": 2.4235289096832275, + "learning_rate": 1.3281590018859713e-05, + "loss": 0.0131, + "step": 9362 + }, + { + "epoch": 1.015069384215091, + "grad_norm": 0.34638288617134094, + "learning_rate": 1.3280864645292327e-05, + "loss": 0.002, + "step": 9363 + }, + { + "epoch": 1.0151777970511708, + "grad_norm": 3.490722417831421, + "learning_rate": 1.3280139271724939e-05, + "loss": 0.0551, + "step": 9364 + }, + { + "epoch": 1.0152862098872506, + "grad_norm": 1.14731764793396, + "learning_rate": 1.3279413898157551e-05, + "loss": 0.0065, + "step": 9365 + }, + { + "epoch": 1.0153946227233304, + "grad_norm": 3.597241163253784, + "learning_rate": 1.3278688524590165e-05, + "loss": 0.0164, + "step": 9366 + }, + { + "epoch": 1.0155030355594101, + "grad_norm": 0.3451208770275116, + "learning_rate": 1.3277963151022777e-05, + "loss": 0.0015, + "step": 9367 + }, + { + "epoch": 1.0156114483954901, + "grad_norm": 2.8023877143859863, + "learning_rate": 1.327723777745539e-05, + "loss": 0.0214, + "step": 9368 + }, + { + "epoch": 1.0157198612315699, + "grad_norm": 3.0335073471069336, + "learning_rate": 1.3276512403888005e-05, + "loss": 0.0081, + "step": 9369 + }, + { + "epoch": 1.0158282740676496, + "grad_norm": 1.7878156900405884, + "learning_rate": 1.3275787030320617e-05, + "loss": 0.0226, + "step": 9370 + }, + { + "epoch": 1.0159366869037294, + "grad_norm": 0.7011999487876892, + "learning_rate": 1.3275061656753229e-05, + "loss": 0.0049, + "step": 9371 + }, + { + "epoch": 1.0160450997398092, + "grad_norm": 2.36696457862854, + "learning_rate": 1.3274336283185843e-05, + "loss": 0.0329, + "step": 9372 + }, + { + "epoch": 1.016153512575889, + "grad_norm": 1.077376365661621, + "learning_rate": 1.3273610909618455e-05, + "loss": 0.0039, + "step": 9373 + }, + { + "epoch": 1.0162619254119687, + "grad_norm": 2.5195486545562744, + "learning_rate": 1.3272885536051067e-05, + "loss": 0.0091, + "step": 9374 + }, + { + "epoch": 1.0163703382480487, + "grad_norm": 1.4250826835632324, + "learning_rate": 1.3272160162483681e-05, + "loss": 0.0611, + "step": 9375 + }, + { + "epoch": 1.0164787510841284, + "grad_norm": 5.3216423988342285, + "learning_rate": 1.3271434788916293e-05, + "loss": 0.0683, + "step": 9376 + }, + { + "epoch": 1.0165871639202082, + "grad_norm": 3.759678602218628, + "learning_rate": 1.3270709415348905e-05, + "loss": 0.0135, + "step": 9377 + }, + { + "epoch": 1.016695576756288, + "grad_norm": 2.2229959964752197, + "learning_rate": 1.326998404178152e-05, + "loss": 0.0365, + "step": 9378 + }, + { + "epoch": 1.0168039895923677, + "grad_norm": 12.993019104003906, + "learning_rate": 1.3269258668214131e-05, + "loss": 0.0472, + "step": 9379 + }, + { + "epoch": 1.0169124024284475, + "grad_norm": 2.4876046180725098, + "learning_rate": 1.3268533294646743e-05, + "loss": 0.0155, + "step": 9380 + }, + { + "epoch": 1.0170208152645273, + "grad_norm": 0.18532946705818176, + "learning_rate": 1.3267807921079356e-05, + "loss": 0.0015, + "step": 9381 + }, + { + "epoch": 1.017129228100607, + "grad_norm": 3.8319754600524902, + "learning_rate": 1.326708254751197e-05, + "loss": 0.044, + "step": 9382 + }, + { + "epoch": 1.017237640936687, + "grad_norm": 3.2897109985351562, + "learning_rate": 1.3266357173944582e-05, + "loss": 0.0192, + "step": 9383 + }, + { + "epoch": 1.0173460537727668, + "grad_norm": 4.382720947265625, + "learning_rate": 1.3265631800377194e-05, + "loss": 0.024, + "step": 9384 + }, + { + "epoch": 1.0174544666088465, + "grad_norm": 1.460736632347107, + "learning_rate": 1.3264906426809808e-05, + "loss": 0.0156, + "step": 9385 + }, + { + "epoch": 1.0175628794449263, + "grad_norm": 2.151554822921753, + "learning_rate": 1.3264181053242421e-05, + "loss": 0.0039, + "step": 9386 + }, + { + "epoch": 1.017671292281006, + "grad_norm": 5.289944648742676, + "learning_rate": 1.3263455679675035e-05, + "loss": 0.0235, + "step": 9387 + }, + { + "epoch": 1.0177797051170858, + "grad_norm": 2.8999743461608887, + "learning_rate": 1.3262730306107647e-05, + "loss": 0.0252, + "step": 9388 + }, + { + "epoch": 1.0178881179531656, + "grad_norm": 1.2558412551879883, + "learning_rate": 1.326200493254026e-05, + "loss": 0.0074, + "step": 9389 + }, + { + "epoch": 1.0179965307892453, + "grad_norm": 3.742034435272217, + "learning_rate": 1.3261279558972873e-05, + "loss": 0.0283, + "step": 9390 + }, + { + "epoch": 1.0181049436253253, + "grad_norm": 4.722580432891846, + "learning_rate": 1.3260554185405486e-05, + "loss": 0.03, + "step": 9391 + }, + { + "epoch": 1.018213356461405, + "grad_norm": 4.176451206207275, + "learning_rate": 1.3259828811838098e-05, + "loss": 0.026, + "step": 9392 + }, + { + "epoch": 1.0183217692974849, + "grad_norm": 3.580923080444336, + "learning_rate": 1.325910343827071e-05, + "loss": 0.0206, + "step": 9393 + }, + { + "epoch": 1.0184301821335646, + "grad_norm": 0.1216217428445816, + "learning_rate": 1.3258378064703324e-05, + "loss": 0.0006, + "step": 9394 + }, + { + "epoch": 1.0185385949696444, + "grad_norm": 1.7245556116104126, + "learning_rate": 1.3257652691135936e-05, + "loss": 0.0101, + "step": 9395 + }, + { + "epoch": 1.0186470078057241, + "grad_norm": 1.956231951713562, + "learning_rate": 1.3256927317568548e-05, + "loss": 0.0209, + "step": 9396 + }, + { + "epoch": 1.018755420641804, + "grad_norm": 1.1300545930862427, + "learning_rate": 1.3256201944001162e-05, + "loss": 0.0057, + "step": 9397 + }, + { + "epoch": 1.0188638334778837, + "grad_norm": 0.7246788740158081, + "learning_rate": 1.3255476570433774e-05, + "loss": 0.0027, + "step": 9398 + }, + { + "epoch": 1.0189722463139637, + "grad_norm": 1.352586269378662, + "learning_rate": 1.3254751196866386e-05, + "loss": 0.0165, + "step": 9399 + }, + { + "epoch": 1.0190806591500434, + "grad_norm": 2.016444444656372, + "learning_rate": 1.3254025823299e-05, + "loss": 0.0162, + "step": 9400 + }, + { + "epoch": 1.0191890719861232, + "grad_norm": 1.500136375427246, + "learning_rate": 1.3253300449731612e-05, + "loss": 0.0077, + "step": 9401 + }, + { + "epoch": 1.019297484822203, + "grad_norm": 1.2532891035079956, + "learning_rate": 1.3252575076164224e-05, + "loss": 0.0319, + "step": 9402 + }, + { + "epoch": 1.0194058976582827, + "grad_norm": 2.6953577995300293, + "learning_rate": 1.325184970259684e-05, + "loss": 0.1094, + "step": 9403 + }, + { + "epoch": 1.0195143104943625, + "grad_norm": 0.7145906686782837, + "learning_rate": 1.3251124329029452e-05, + "loss": 0.0037, + "step": 9404 + }, + { + "epoch": 1.0196227233304422, + "grad_norm": 3.817883253097534, + "learning_rate": 1.3250398955462066e-05, + "loss": 0.1089, + "step": 9405 + }, + { + "epoch": 1.0197311361665222, + "grad_norm": 2.4096033573150635, + "learning_rate": 1.3249673581894678e-05, + "loss": 0.0232, + "step": 9406 + }, + { + "epoch": 1.019839549002602, + "grad_norm": 1.5752055644989014, + "learning_rate": 1.324894820832729e-05, + "loss": 0.0304, + "step": 9407 + }, + { + "epoch": 1.0199479618386817, + "grad_norm": 2.753203868865967, + "learning_rate": 1.3248222834759902e-05, + "loss": 0.0409, + "step": 9408 + }, + { + "epoch": 1.0200563746747615, + "grad_norm": 4.69536018371582, + "learning_rate": 1.3247497461192516e-05, + "loss": 0.0899, + "step": 9409 + }, + { + "epoch": 1.0201647875108413, + "grad_norm": 1.5082591772079468, + "learning_rate": 1.3246772087625128e-05, + "loss": 0.0149, + "step": 9410 + }, + { + "epoch": 1.020273200346921, + "grad_norm": 5.379029273986816, + "learning_rate": 1.324604671405774e-05, + "loss": 0.0356, + "step": 9411 + }, + { + "epoch": 1.0203816131830008, + "grad_norm": 3.3036420345306396, + "learning_rate": 1.3245321340490354e-05, + "loss": 0.0273, + "step": 9412 + }, + { + "epoch": 1.0204900260190806, + "grad_norm": 3.132098436355591, + "learning_rate": 1.3244595966922966e-05, + "loss": 0.0155, + "step": 9413 + }, + { + "epoch": 1.0205984388551606, + "grad_norm": 0.46937817335128784, + "learning_rate": 1.3243870593355578e-05, + "loss": 0.0047, + "step": 9414 + }, + { + "epoch": 1.0207068516912403, + "grad_norm": 0.30098336935043335, + "learning_rate": 1.3243145219788192e-05, + "loss": 0.003, + "step": 9415 + }, + { + "epoch": 1.02081526452732, + "grad_norm": 3.613107442855835, + "learning_rate": 1.3242419846220804e-05, + "loss": 0.0196, + "step": 9416 + }, + { + "epoch": 1.0209236773633998, + "grad_norm": 1.3251768350601196, + "learning_rate": 1.3241694472653417e-05, + "loss": 0.0141, + "step": 9417 + }, + { + "epoch": 1.0210320901994796, + "grad_norm": 0.4024938941001892, + "learning_rate": 1.324096909908603e-05, + "loss": 0.0026, + "step": 9418 + }, + { + "epoch": 1.0211405030355594, + "grad_norm": 1.4952269792556763, + "learning_rate": 1.3240243725518643e-05, + "loss": 0.0124, + "step": 9419 + }, + { + "epoch": 1.0212489158716391, + "grad_norm": 1.0167229175567627, + "learning_rate": 1.3239518351951255e-05, + "loss": 0.0106, + "step": 9420 + }, + { + "epoch": 1.021357328707719, + "grad_norm": 1.8880043029785156, + "learning_rate": 1.323879297838387e-05, + "loss": 0.0348, + "step": 9421 + }, + { + "epoch": 1.0214657415437989, + "grad_norm": 4.953146934509277, + "learning_rate": 1.3238067604816482e-05, + "loss": 0.0512, + "step": 9422 + }, + { + "epoch": 1.0215741543798786, + "grad_norm": 6.720941066741943, + "learning_rate": 1.3237342231249094e-05, + "loss": 0.0177, + "step": 9423 + }, + { + "epoch": 1.0216825672159584, + "grad_norm": 1.6376980543136597, + "learning_rate": 1.3236616857681708e-05, + "loss": 0.0112, + "step": 9424 + }, + { + "epoch": 1.0217909800520382, + "grad_norm": 0.7140413522720337, + "learning_rate": 1.323589148411432e-05, + "loss": 0.0056, + "step": 9425 + }, + { + "epoch": 1.021899392888118, + "grad_norm": 1.313881278038025, + "learning_rate": 1.3235166110546933e-05, + "loss": 0.006, + "step": 9426 + }, + { + "epoch": 1.0220078057241977, + "grad_norm": 1.5670381784439087, + "learning_rate": 1.3234440736979546e-05, + "loss": 0.0135, + "step": 9427 + }, + { + "epoch": 1.0221162185602775, + "grad_norm": 0.9809780120849609, + "learning_rate": 1.3233715363412159e-05, + "loss": 0.0106, + "step": 9428 + }, + { + "epoch": 1.0222246313963572, + "grad_norm": 1.739552617073059, + "learning_rate": 1.323298998984477e-05, + "loss": 0.0167, + "step": 9429 + }, + { + "epoch": 1.0223330442324372, + "grad_norm": 2.251674175262451, + "learning_rate": 1.3232264616277385e-05, + "loss": 0.0318, + "step": 9430 + }, + { + "epoch": 1.022441457068517, + "grad_norm": 2.896998405456543, + "learning_rate": 1.3231539242709997e-05, + "loss": 0.0162, + "step": 9431 + }, + { + "epoch": 1.0225498699045967, + "grad_norm": 1.9568415880203247, + "learning_rate": 1.3230813869142609e-05, + "loss": 0.0271, + "step": 9432 + }, + { + "epoch": 1.0226582827406765, + "grad_norm": 2.2305166721343994, + "learning_rate": 1.3230088495575221e-05, + "loss": 0.0173, + "step": 9433 + }, + { + "epoch": 1.0227666955767563, + "grad_norm": 0.907880961894989, + "learning_rate": 1.3229363122007835e-05, + "loss": 0.0111, + "step": 9434 + }, + { + "epoch": 1.022875108412836, + "grad_norm": 1.0778015851974487, + "learning_rate": 1.3228637748440447e-05, + "loss": 0.0048, + "step": 9435 + }, + { + "epoch": 1.0229835212489158, + "grad_norm": 2.8678247928619385, + "learning_rate": 1.3227912374873059e-05, + "loss": 0.0189, + "step": 9436 + }, + { + "epoch": 1.0230919340849958, + "grad_norm": 2.0692331790924072, + "learning_rate": 1.3227187001305673e-05, + "loss": 0.0227, + "step": 9437 + }, + { + "epoch": 1.0232003469210755, + "grad_norm": 0.27789658308029175, + "learning_rate": 1.3226461627738287e-05, + "loss": 0.0032, + "step": 9438 + }, + { + "epoch": 1.0233087597571553, + "grad_norm": 3.6080729961395264, + "learning_rate": 1.32257362541709e-05, + "loss": 0.0349, + "step": 9439 + }, + { + "epoch": 1.023417172593235, + "grad_norm": 2.5668258666992188, + "learning_rate": 1.3225010880603513e-05, + "loss": 0.0472, + "step": 9440 + }, + { + "epoch": 1.0235255854293148, + "grad_norm": 1.2367454767227173, + "learning_rate": 1.3224285507036125e-05, + "loss": 0.0095, + "step": 9441 + }, + { + "epoch": 1.0236339982653946, + "grad_norm": 2.498795986175537, + "learning_rate": 1.3223560133468739e-05, + "loss": 0.0117, + "step": 9442 + }, + { + "epoch": 1.0237424111014743, + "grad_norm": 2.0935728549957275, + "learning_rate": 1.3222834759901351e-05, + "loss": 0.0534, + "step": 9443 + }, + { + "epoch": 1.023850823937554, + "grad_norm": 1.8412178754806519, + "learning_rate": 1.3222109386333963e-05, + "loss": 0.0087, + "step": 9444 + }, + { + "epoch": 1.023959236773634, + "grad_norm": 2.1618504524230957, + "learning_rate": 1.3221384012766575e-05, + "loss": 0.0456, + "step": 9445 + }, + { + "epoch": 1.0240676496097139, + "grad_norm": 6.870763301849365, + "learning_rate": 1.3220658639199189e-05, + "loss": 0.0621, + "step": 9446 + }, + { + "epoch": 1.0241760624457936, + "grad_norm": 4.141903877258301, + "learning_rate": 1.3219933265631801e-05, + "loss": 0.0117, + "step": 9447 + }, + { + "epoch": 1.0242844752818734, + "grad_norm": 1.285343050956726, + "learning_rate": 1.3219207892064413e-05, + "loss": 0.0062, + "step": 9448 + }, + { + "epoch": 1.0243928881179531, + "grad_norm": 3.880573034286499, + "learning_rate": 1.3218482518497027e-05, + "loss": 0.0179, + "step": 9449 + }, + { + "epoch": 1.024501300954033, + "grad_norm": 2.291929006576538, + "learning_rate": 1.321775714492964e-05, + "loss": 0.0036, + "step": 9450 + }, + { + "epoch": 1.0246097137901127, + "grad_norm": 3.0323684215545654, + "learning_rate": 1.3217031771362251e-05, + "loss": 0.0391, + "step": 9451 + }, + { + "epoch": 1.0247181266261924, + "grad_norm": 3.673429489135742, + "learning_rate": 1.3216306397794865e-05, + "loss": 0.0189, + "step": 9452 + }, + { + "epoch": 1.0248265394622724, + "grad_norm": 1.819890022277832, + "learning_rate": 1.3215581024227477e-05, + "loss": 0.0085, + "step": 9453 + }, + { + "epoch": 1.0249349522983522, + "grad_norm": 0.8544552326202393, + "learning_rate": 1.321485565066009e-05, + "loss": 0.0125, + "step": 9454 + }, + { + "epoch": 1.025043365134432, + "grad_norm": 2.084407329559326, + "learning_rate": 1.3214130277092705e-05, + "loss": 0.0152, + "step": 9455 + }, + { + "epoch": 1.0251517779705117, + "grad_norm": 2.4418866634368896, + "learning_rate": 1.3213404903525317e-05, + "loss": 0.0111, + "step": 9456 + }, + { + "epoch": 1.0252601908065915, + "grad_norm": 4.749178886413574, + "learning_rate": 1.3212679529957931e-05, + "loss": 0.0593, + "step": 9457 + }, + { + "epoch": 1.0253686036426712, + "grad_norm": 2.8102493286132812, + "learning_rate": 1.3211954156390543e-05, + "loss": 0.0322, + "step": 9458 + }, + { + "epoch": 1.025477016478751, + "grad_norm": 2.780932664871216, + "learning_rate": 1.3211228782823155e-05, + "loss": 0.0132, + "step": 9459 + }, + { + "epoch": 1.0255854293148308, + "grad_norm": 0.4300728738307953, + "learning_rate": 1.3210503409255767e-05, + "loss": 0.0017, + "step": 9460 + }, + { + "epoch": 1.0256938421509108, + "grad_norm": 0.37488192319869995, + "learning_rate": 1.3209778035688381e-05, + "loss": 0.0036, + "step": 9461 + }, + { + "epoch": 1.0258022549869905, + "grad_norm": 3.7877020835876465, + "learning_rate": 1.3209052662120993e-05, + "loss": 0.0264, + "step": 9462 + }, + { + "epoch": 1.0259106678230703, + "grad_norm": 2.933492422103882, + "learning_rate": 1.3208327288553606e-05, + "loss": 0.0058, + "step": 9463 + }, + { + "epoch": 1.02601908065915, + "grad_norm": 1.309349536895752, + "learning_rate": 1.320760191498622e-05, + "loss": 0.0036, + "step": 9464 + }, + { + "epoch": 1.0261274934952298, + "grad_norm": 4.990745544433594, + "learning_rate": 1.3206876541418832e-05, + "loss": 0.0717, + "step": 9465 + }, + { + "epoch": 1.0262359063313096, + "grad_norm": 6.775963306427002, + "learning_rate": 1.3206151167851444e-05, + "loss": 0.0353, + "step": 9466 + }, + { + "epoch": 1.0263443191673893, + "grad_norm": 0.25914129614830017, + "learning_rate": 1.3205425794284058e-05, + "loss": 0.0024, + "step": 9467 + }, + { + "epoch": 1.0264527320034693, + "grad_norm": 2.2397799491882324, + "learning_rate": 1.320470042071667e-05, + "loss": 0.0303, + "step": 9468 + }, + { + "epoch": 1.026561144839549, + "grad_norm": 7.468449592590332, + "learning_rate": 1.3203975047149282e-05, + "loss": 0.0222, + "step": 9469 + }, + { + "epoch": 1.0266695576756288, + "grad_norm": 6.039738178253174, + "learning_rate": 1.3203249673581896e-05, + "loss": 0.0304, + "step": 9470 + }, + { + "epoch": 1.0267779705117086, + "grad_norm": 0.5818811655044556, + "learning_rate": 1.3202524300014508e-05, + "loss": 0.0036, + "step": 9471 + }, + { + "epoch": 1.0268863833477884, + "grad_norm": 0.7991334795951843, + "learning_rate": 1.320179892644712e-05, + "loss": 0.0086, + "step": 9472 + }, + { + "epoch": 1.0269947961838681, + "grad_norm": 1.3546605110168457, + "learning_rate": 1.3201073552879736e-05, + "loss": 0.0239, + "step": 9473 + }, + { + "epoch": 1.027103209019948, + "grad_norm": 3.6435184478759766, + "learning_rate": 1.3200348179312348e-05, + "loss": 0.0323, + "step": 9474 + }, + { + "epoch": 1.0272116218560277, + "grad_norm": 0.43417519330978394, + "learning_rate": 1.319962280574496e-05, + "loss": 0.0037, + "step": 9475 + }, + { + "epoch": 1.0273200346921076, + "grad_norm": 3.33212947845459, + "learning_rate": 1.3198897432177574e-05, + "loss": 0.0811, + "step": 9476 + }, + { + "epoch": 1.0274284475281874, + "grad_norm": 2.7151248455047607, + "learning_rate": 1.3198172058610186e-05, + "loss": 0.0113, + "step": 9477 + }, + { + "epoch": 1.0275368603642672, + "grad_norm": 2.750929355621338, + "learning_rate": 1.3197446685042798e-05, + "loss": 0.0243, + "step": 9478 + }, + { + "epoch": 1.027645273200347, + "grad_norm": 3.6332216262817383, + "learning_rate": 1.3196721311475412e-05, + "loss": 0.0674, + "step": 9479 + }, + { + "epoch": 1.0277536860364267, + "grad_norm": 3.6895081996917725, + "learning_rate": 1.3195995937908024e-05, + "loss": 0.0317, + "step": 9480 + }, + { + "epoch": 1.0278620988725065, + "grad_norm": 1.747079610824585, + "learning_rate": 1.3195270564340636e-05, + "loss": 0.0058, + "step": 9481 + }, + { + "epoch": 1.0279705117085862, + "grad_norm": 5.163487434387207, + "learning_rate": 1.319454519077325e-05, + "loss": 0.0504, + "step": 9482 + }, + { + "epoch": 1.028078924544666, + "grad_norm": 0.1716814786195755, + "learning_rate": 1.3193819817205862e-05, + "loss": 0.0013, + "step": 9483 + }, + { + "epoch": 1.028187337380746, + "grad_norm": 23.876110076904297, + "learning_rate": 1.3193094443638474e-05, + "loss": 0.0683, + "step": 9484 + }, + { + "epoch": 1.0282957502168257, + "grad_norm": 2.8362877368927, + "learning_rate": 1.3192369070071086e-05, + "loss": 0.0093, + "step": 9485 + }, + { + "epoch": 1.0284041630529055, + "grad_norm": 1.9595043659210205, + "learning_rate": 1.31916436965037e-05, + "loss": 0.0116, + "step": 9486 + }, + { + "epoch": 1.0285125758889853, + "grad_norm": 1.9747129678726196, + "learning_rate": 1.3190918322936312e-05, + "loss": 0.0121, + "step": 9487 + }, + { + "epoch": 1.028620988725065, + "grad_norm": 0.3265377879142761, + "learning_rate": 1.3190192949368924e-05, + "loss": 0.0018, + "step": 9488 + }, + { + "epoch": 1.0287294015611448, + "grad_norm": 0.5983511209487915, + "learning_rate": 1.3189467575801538e-05, + "loss": 0.0031, + "step": 9489 + }, + { + "epoch": 1.0288378143972245, + "grad_norm": 4.19911003112793, + "learning_rate": 1.3188742202234152e-05, + "loss": 0.0432, + "step": 9490 + }, + { + "epoch": 1.0289462272333045, + "grad_norm": 3.018266439437866, + "learning_rate": 1.3188016828666766e-05, + "loss": 0.0136, + "step": 9491 + }, + { + "epoch": 1.0290546400693843, + "grad_norm": 0.5147871971130371, + "learning_rate": 1.3187291455099378e-05, + "loss": 0.0067, + "step": 9492 + }, + { + "epoch": 1.029163052905464, + "grad_norm": 4.236519813537598, + "learning_rate": 1.318656608153199e-05, + "loss": 0.0273, + "step": 9493 + }, + { + "epoch": 1.0292714657415438, + "grad_norm": 1.7871907949447632, + "learning_rate": 1.3185840707964604e-05, + "loss": 0.0197, + "step": 9494 + }, + { + "epoch": 1.0293798785776236, + "grad_norm": 1.2151414155960083, + "learning_rate": 1.3185115334397216e-05, + "loss": 0.0071, + "step": 9495 + }, + { + "epoch": 1.0294882914137033, + "grad_norm": 1.4609218835830688, + "learning_rate": 1.3184389960829828e-05, + "loss": 0.0238, + "step": 9496 + }, + { + "epoch": 1.0295967042497831, + "grad_norm": 3.2545530796051025, + "learning_rate": 1.3183664587262442e-05, + "loss": 0.0252, + "step": 9497 + }, + { + "epoch": 1.0297051170858629, + "grad_norm": 1.0144455432891846, + "learning_rate": 1.3182939213695054e-05, + "loss": 0.0121, + "step": 9498 + }, + { + "epoch": 1.0298135299219429, + "grad_norm": 4.3016533851623535, + "learning_rate": 1.3182213840127666e-05, + "loss": 0.0141, + "step": 9499 + }, + { + "epoch": 1.0299219427580226, + "grad_norm": 1.9744436740875244, + "learning_rate": 1.3181488466560279e-05, + "loss": 0.0227, + "step": 9500 + }, + { + "epoch": 1.0300303555941024, + "grad_norm": 1.3174666166305542, + "learning_rate": 1.3180763092992892e-05, + "loss": 0.0084, + "step": 9501 + }, + { + "epoch": 1.0301387684301822, + "grad_norm": 0.5996721386909485, + "learning_rate": 1.3180037719425505e-05, + "loss": 0.0043, + "step": 9502 + }, + { + "epoch": 1.030247181266262, + "grad_norm": 5.895175457000732, + "learning_rate": 1.3179312345858117e-05, + "loss": 0.0471, + "step": 9503 + }, + { + "epoch": 1.0303555941023417, + "grad_norm": 1.4839346408843994, + "learning_rate": 1.317858697229073e-05, + "loss": 0.0145, + "step": 9504 + }, + { + "epoch": 1.0304640069384214, + "grad_norm": 0.14136378467082977, + "learning_rate": 1.3177861598723343e-05, + "loss": 0.001, + "step": 9505 + }, + { + "epoch": 1.0305724197745012, + "grad_norm": 2.836693286895752, + "learning_rate": 1.3177136225155955e-05, + "loss": 0.0076, + "step": 9506 + }, + { + "epoch": 1.0306808326105812, + "grad_norm": 2.6257004737854004, + "learning_rate": 1.317641085158857e-05, + "loss": 0.0432, + "step": 9507 + }, + { + "epoch": 1.030789245446661, + "grad_norm": 3.104152202606201, + "learning_rate": 1.3175685478021183e-05, + "loss": 0.0442, + "step": 9508 + }, + { + "epoch": 1.0308976582827407, + "grad_norm": 6.386959552764893, + "learning_rate": 1.3174960104453796e-05, + "loss": 0.0709, + "step": 9509 + }, + { + "epoch": 1.0310060711188205, + "grad_norm": 3.5833473205566406, + "learning_rate": 1.3174234730886409e-05, + "loss": 0.0323, + "step": 9510 + }, + { + "epoch": 1.0311144839549002, + "grad_norm": 1.1136432886123657, + "learning_rate": 1.317350935731902e-05, + "loss": 0.0038, + "step": 9511 + }, + { + "epoch": 1.03122289679098, + "grad_norm": 2.6158533096313477, + "learning_rate": 1.3172783983751633e-05, + "loss": 0.0667, + "step": 9512 + }, + { + "epoch": 1.0313313096270598, + "grad_norm": 1.8294525146484375, + "learning_rate": 1.3172058610184247e-05, + "loss": 0.019, + "step": 9513 + }, + { + "epoch": 1.0314397224631395, + "grad_norm": 1.487972378730774, + "learning_rate": 1.3171333236616859e-05, + "loss": 0.048, + "step": 9514 + }, + { + "epoch": 1.0315481352992195, + "grad_norm": 0.8816925883293152, + "learning_rate": 1.3170607863049471e-05, + "loss": 0.0105, + "step": 9515 + }, + { + "epoch": 1.0316565481352993, + "grad_norm": 0.20235145092010498, + "learning_rate": 1.3169882489482085e-05, + "loss": 0.0017, + "step": 9516 + }, + { + "epoch": 1.031764960971379, + "grad_norm": 2.0457122325897217, + "learning_rate": 1.3169157115914697e-05, + "loss": 0.0163, + "step": 9517 + }, + { + "epoch": 1.0318733738074588, + "grad_norm": 4.737603664398193, + "learning_rate": 1.3168431742347309e-05, + "loss": 0.0346, + "step": 9518 + }, + { + "epoch": 1.0319817866435386, + "grad_norm": 0.15958982706069946, + "learning_rate": 1.3167706368779923e-05, + "loss": 0.0016, + "step": 9519 + }, + { + "epoch": 1.0320901994796183, + "grad_norm": 4.799680709838867, + "learning_rate": 1.3166980995212535e-05, + "loss": 0.0174, + "step": 9520 + }, + { + "epoch": 1.032198612315698, + "grad_norm": 1.022735834121704, + "learning_rate": 1.3166255621645147e-05, + "loss": 0.008, + "step": 9521 + }, + { + "epoch": 1.0323070251517779, + "grad_norm": 0.7288884520530701, + "learning_rate": 1.3165530248077761e-05, + "loss": 0.0133, + "step": 9522 + }, + { + "epoch": 1.0324154379878578, + "grad_norm": 5.010593891143799, + "learning_rate": 1.3164804874510373e-05, + "loss": 0.0287, + "step": 9523 + }, + { + "epoch": 1.0325238508239376, + "grad_norm": 4.475753307342529, + "learning_rate": 1.3164079500942987e-05, + "loss": 0.041, + "step": 9524 + }, + { + "epoch": 1.0326322636600174, + "grad_norm": 1.0464750528335571, + "learning_rate": 1.3163354127375601e-05, + "loss": 0.0029, + "step": 9525 + }, + { + "epoch": 1.0327406764960971, + "grad_norm": 3.530505895614624, + "learning_rate": 1.3162628753808213e-05, + "loss": 0.0206, + "step": 9526 + }, + { + "epoch": 1.032849089332177, + "grad_norm": 2.827594041824341, + "learning_rate": 1.3161903380240825e-05, + "loss": 0.0077, + "step": 9527 + }, + { + "epoch": 1.0329575021682567, + "grad_norm": 0.31615927815437317, + "learning_rate": 1.3161178006673439e-05, + "loss": 0.002, + "step": 9528 + }, + { + "epoch": 1.0330659150043364, + "grad_norm": 0.6828850507736206, + "learning_rate": 1.3160452633106051e-05, + "loss": 0.0073, + "step": 9529 + }, + { + "epoch": 1.0331743278404164, + "grad_norm": 1.8957700729370117, + "learning_rate": 1.3159727259538663e-05, + "loss": 0.0191, + "step": 9530 + }, + { + "epoch": 1.0332827406764962, + "grad_norm": 2.399704694747925, + "learning_rate": 1.3159001885971277e-05, + "loss": 0.0286, + "step": 9531 + }, + { + "epoch": 1.033391153512576, + "grad_norm": 2.3036088943481445, + "learning_rate": 1.315827651240389e-05, + "loss": 0.0089, + "step": 9532 + }, + { + "epoch": 1.0334995663486557, + "grad_norm": 2.388286828994751, + "learning_rate": 1.3157551138836501e-05, + "loss": 0.004, + "step": 9533 + }, + { + "epoch": 1.0336079791847355, + "grad_norm": 2.6129841804504395, + "learning_rate": 1.3156825765269115e-05, + "loss": 0.0653, + "step": 9534 + }, + { + "epoch": 1.0337163920208152, + "grad_norm": 4.378616809844971, + "learning_rate": 1.3156100391701727e-05, + "loss": 0.0104, + "step": 9535 + }, + { + "epoch": 1.033824804856895, + "grad_norm": 1.2949475049972534, + "learning_rate": 1.315537501813434e-05, + "loss": 0.0032, + "step": 9536 + }, + { + "epoch": 1.0339332176929747, + "grad_norm": 9.627583503723145, + "learning_rate": 1.3154649644566952e-05, + "loss": 0.0213, + "step": 9537 + }, + { + "epoch": 1.0340416305290547, + "grad_norm": 2.926309585571289, + "learning_rate": 1.3153924270999565e-05, + "loss": 0.0138, + "step": 9538 + }, + { + "epoch": 1.0341500433651345, + "grad_norm": 2.823669910430908, + "learning_rate": 1.3153198897432178e-05, + "loss": 0.0434, + "step": 9539 + }, + { + "epoch": 1.0342584562012143, + "grad_norm": 5.235680103302002, + "learning_rate": 1.315247352386479e-05, + "loss": 0.0248, + "step": 9540 + }, + { + "epoch": 1.034366869037294, + "grad_norm": 2.9549975395202637, + "learning_rate": 1.3151748150297404e-05, + "loss": 0.0181, + "step": 9541 + }, + { + "epoch": 1.0344752818733738, + "grad_norm": 1.8625260591506958, + "learning_rate": 1.3151022776730017e-05, + "loss": 0.0189, + "step": 9542 + }, + { + "epoch": 1.0345836947094535, + "grad_norm": 3.766920804977417, + "learning_rate": 1.3150297403162631e-05, + "loss": 0.0199, + "step": 9543 + }, + { + "epoch": 1.0346921075455333, + "grad_norm": 3.999290943145752, + "learning_rate": 1.3149572029595243e-05, + "loss": 0.0311, + "step": 9544 + }, + { + "epoch": 1.034800520381613, + "grad_norm": 2.334874391555786, + "learning_rate": 1.3148846656027856e-05, + "loss": 0.0157, + "step": 9545 + }, + { + "epoch": 1.034908933217693, + "grad_norm": 3.0194907188415527, + "learning_rate": 1.314812128246047e-05, + "loss": 0.047, + "step": 9546 + }, + { + "epoch": 1.0350173460537728, + "grad_norm": 4.0991997718811035, + "learning_rate": 1.3147395908893082e-05, + "loss": 0.0168, + "step": 9547 + }, + { + "epoch": 1.0351257588898526, + "grad_norm": 1.2058863639831543, + "learning_rate": 1.3146670535325694e-05, + "loss": 0.0068, + "step": 9548 + }, + { + "epoch": 1.0352341717259324, + "grad_norm": 2.1653146743774414, + "learning_rate": 1.3145945161758308e-05, + "loss": 0.0317, + "step": 9549 + }, + { + "epoch": 1.0353425845620121, + "grad_norm": 2.9162938594818115, + "learning_rate": 1.314521978819092e-05, + "loss": 0.0257, + "step": 9550 + }, + { + "epoch": 1.0354509973980919, + "grad_norm": 1.7722725868225098, + "learning_rate": 1.3144494414623532e-05, + "loss": 0.0236, + "step": 9551 + }, + { + "epoch": 1.0355594102341716, + "grad_norm": 0.2590125799179077, + "learning_rate": 1.3143769041056144e-05, + "loss": 0.0028, + "step": 9552 + }, + { + "epoch": 1.0356678230702516, + "grad_norm": 5.760369777679443, + "learning_rate": 1.3143043667488758e-05, + "loss": 0.0572, + "step": 9553 + }, + { + "epoch": 1.0357762359063314, + "grad_norm": 2.312770366668701, + "learning_rate": 1.314231829392137e-05, + "loss": 0.0049, + "step": 9554 + }, + { + "epoch": 1.0358846487424112, + "grad_norm": 5.434932231903076, + "learning_rate": 1.3141592920353982e-05, + "loss": 0.0769, + "step": 9555 + }, + { + "epoch": 1.035993061578491, + "grad_norm": 7.208318710327148, + "learning_rate": 1.3140867546786596e-05, + "loss": 0.0929, + "step": 9556 + }, + { + "epoch": 1.0361014744145707, + "grad_norm": 2.9517014026641846, + "learning_rate": 1.3140142173219208e-05, + "loss": 0.0114, + "step": 9557 + }, + { + "epoch": 1.0362098872506504, + "grad_norm": 5.508167266845703, + "learning_rate": 1.313941679965182e-05, + "loss": 0.0538, + "step": 9558 + }, + { + "epoch": 1.0363183000867302, + "grad_norm": 4.463249206542969, + "learning_rate": 1.3138691426084436e-05, + "loss": 0.0032, + "step": 9559 + }, + { + "epoch": 1.03642671292281, + "grad_norm": 0.5609736442565918, + "learning_rate": 1.3137966052517048e-05, + "loss": 0.0071, + "step": 9560 + }, + { + "epoch": 1.03653512575889, + "grad_norm": 2.7134783267974854, + "learning_rate": 1.3137240678949662e-05, + "loss": 0.0289, + "step": 9561 + }, + { + "epoch": 1.0366435385949697, + "grad_norm": 4.619690895080566, + "learning_rate": 1.3136515305382274e-05, + "loss": 0.0596, + "step": 9562 + }, + { + "epoch": 1.0367519514310495, + "grad_norm": 4.621611595153809, + "learning_rate": 1.3135789931814886e-05, + "loss": 0.0149, + "step": 9563 + }, + { + "epoch": 1.0368603642671292, + "grad_norm": 7.453864097595215, + "learning_rate": 1.3135064558247498e-05, + "loss": 0.0415, + "step": 9564 + }, + { + "epoch": 1.036968777103209, + "grad_norm": 3.311875104904175, + "learning_rate": 1.3134339184680112e-05, + "loss": 0.0253, + "step": 9565 + }, + { + "epoch": 1.0370771899392888, + "grad_norm": 6.752494812011719, + "learning_rate": 1.3133613811112724e-05, + "loss": 0.0488, + "step": 9566 + }, + { + "epoch": 1.0371856027753685, + "grad_norm": 2.2036123275756836, + "learning_rate": 1.3132888437545336e-05, + "loss": 0.0106, + "step": 9567 + }, + { + "epoch": 1.0372940156114483, + "grad_norm": 0.07384271174669266, + "learning_rate": 1.313216306397795e-05, + "loss": 0.0009, + "step": 9568 + }, + { + "epoch": 1.0374024284475283, + "grad_norm": 0.44767245650291443, + "learning_rate": 1.3131437690410562e-05, + "loss": 0.0042, + "step": 9569 + }, + { + "epoch": 1.037510841283608, + "grad_norm": 3.720950126647949, + "learning_rate": 1.3130712316843174e-05, + "loss": 0.0069, + "step": 9570 + }, + { + "epoch": 1.0376192541196878, + "grad_norm": 0.43038830161094666, + "learning_rate": 1.3129986943275788e-05, + "loss": 0.0044, + "step": 9571 + }, + { + "epoch": 1.0377276669557676, + "grad_norm": 1.6563421487808228, + "learning_rate": 1.31292615697084e-05, + "loss": 0.0423, + "step": 9572 + }, + { + "epoch": 1.0378360797918473, + "grad_norm": 3.229098320007324, + "learning_rate": 1.3128536196141013e-05, + "loss": 0.0377, + "step": 9573 + }, + { + "epoch": 1.037944492627927, + "grad_norm": 2.3815903663635254, + "learning_rate": 1.3127810822573626e-05, + "loss": 0.0121, + "step": 9574 + }, + { + "epoch": 1.0380529054640069, + "grad_norm": 1.103410005569458, + "learning_rate": 1.3127085449006239e-05, + "loss": 0.0111, + "step": 9575 + }, + { + "epoch": 1.0381613183000868, + "grad_norm": 6.90800142288208, + "learning_rate": 1.3126360075438852e-05, + "loss": 0.0918, + "step": 9576 + }, + { + "epoch": 1.0382697311361666, + "grad_norm": 2.5451412200927734, + "learning_rate": 1.3125634701871466e-05, + "loss": 0.0252, + "step": 9577 + }, + { + "epoch": 1.0383781439722464, + "grad_norm": 1.39750337600708, + "learning_rate": 1.3124909328304078e-05, + "loss": 0.0045, + "step": 9578 + }, + { + "epoch": 1.0384865568083261, + "grad_norm": 2.6598939895629883, + "learning_rate": 1.312418395473669e-05, + "loss": 0.0152, + "step": 9579 + }, + { + "epoch": 1.038594969644406, + "grad_norm": 0.9184389114379883, + "learning_rate": 1.3123458581169304e-05, + "loss": 0.0076, + "step": 9580 + }, + { + "epoch": 1.0387033824804857, + "grad_norm": 0.9636974334716797, + "learning_rate": 1.3122733207601916e-05, + "loss": 0.0032, + "step": 9581 + }, + { + "epoch": 1.0388117953165654, + "grad_norm": 1.9106290340423584, + "learning_rate": 1.3122007834034529e-05, + "loss": 0.0267, + "step": 9582 + }, + { + "epoch": 1.0389202081526452, + "grad_norm": 3.7689716815948486, + "learning_rate": 1.3121282460467142e-05, + "loss": 0.0396, + "step": 9583 + }, + { + "epoch": 1.0390286209887252, + "grad_norm": 0.4455714523792267, + "learning_rate": 1.3120557086899755e-05, + "loss": 0.0029, + "step": 9584 + }, + { + "epoch": 1.039137033824805, + "grad_norm": 1.436852216720581, + "learning_rate": 1.3119831713332367e-05, + "loss": 0.0124, + "step": 9585 + }, + { + "epoch": 1.0392454466608847, + "grad_norm": 2.504395008087158, + "learning_rate": 1.311910633976498e-05, + "loss": 0.0153, + "step": 9586 + }, + { + "epoch": 1.0393538594969645, + "grad_norm": 2.634329319000244, + "learning_rate": 1.3118380966197593e-05, + "loss": 0.024, + "step": 9587 + }, + { + "epoch": 1.0394622723330442, + "grad_norm": 0.04729940742254257, + "learning_rate": 1.3117655592630205e-05, + "loss": 0.0006, + "step": 9588 + }, + { + "epoch": 1.039570685169124, + "grad_norm": 1.5515424013137817, + "learning_rate": 1.3116930219062817e-05, + "loss": 0.0129, + "step": 9589 + }, + { + "epoch": 1.0396790980052038, + "grad_norm": 3.642504930496216, + "learning_rate": 1.311620484549543e-05, + "loss": 0.0218, + "step": 9590 + }, + { + "epoch": 1.0397875108412835, + "grad_norm": 2.586061477661133, + "learning_rate": 1.3115479471928043e-05, + "loss": 0.0137, + "step": 9591 + }, + { + "epoch": 1.0398959236773635, + "grad_norm": 4.531463146209717, + "learning_rate": 1.3114754098360655e-05, + "loss": 0.0089, + "step": 9592 + }, + { + "epoch": 1.0400043365134433, + "grad_norm": 4.939754486083984, + "learning_rate": 1.311402872479327e-05, + "loss": 0.0249, + "step": 9593 + }, + { + "epoch": 1.040112749349523, + "grad_norm": 2.2224764823913574, + "learning_rate": 1.3113303351225883e-05, + "loss": 0.0073, + "step": 9594 + }, + { + "epoch": 1.0402211621856028, + "grad_norm": 4.176815509796143, + "learning_rate": 1.3112577977658497e-05, + "loss": 0.0378, + "step": 9595 + }, + { + "epoch": 1.0403295750216826, + "grad_norm": 3.4347825050354004, + "learning_rate": 1.3111852604091109e-05, + "loss": 0.0388, + "step": 9596 + }, + { + "epoch": 1.0404379878577623, + "grad_norm": 5.67831563949585, + "learning_rate": 1.3111127230523721e-05, + "loss": 0.0712, + "step": 9597 + }, + { + "epoch": 1.040546400693842, + "grad_norm": 2.95428729057312, + "learning_rate": 1.3110401856956335e-05, + "loss": 0.0225, + "step": 9598 + }, + { + "epoch": 1.0406548135299218, + "grad_norm": 1.446640968322754, + "learning_rate": 1.3109676483388947e-05, + "loss": 0.0097, + "step": 9599 + }, + { + "epoch": 1.0407632263660018, + "grad_norm": 8.04585075378418, + "learning_rate": 1.3108951109821559e-05, + "loss": 0.0135, + "step": 9600 + }, + { + "epoch": 1.0408716392020816, + "grad_norm": 1.732962727546692, + "learning_rate": 1.3108225736254173e-05, + "loss": 0.0162, + "step": 9601 + }, + { + "epoch": 1.0409800520381614, + "grad_norm": 2.654958724975586, + "learning_rate": 1.3107500362686785e-05, + "loss": 0.0475, + "step": 9602 + }, + { + "epoch": 1.0410884648742411, + "grad_norm": 0.3966701328754425, + "learning_rate": 1.3106774989119397e-05, + "loss": 0.0046, + "step": 9603 + }, + { + "epoch": 1.0411968777103209, + "grad_norm": 5.016984462738037, + "learning_rate": 1.310604961555201e-05, + "loss": 0.0324, + "step": 9604 + }, + { + "epoch": 1.0413052905464006, + "grad_norm": 0.459143728017807, + "learning_rate": 1.3105324241984623e-05, + "loss": 0.0022, + "step": 9605 + }, + { + "epoch": 1.0414137033824804, + "grad_norm": 4.999817848205566, + "learning_rate": 1.3104598868417235e-05, + "loss": 0.0536, + "step": 9606 + }, + { + "epoch": 1.0415221162185602, + "grad_norm": 2.205812692642212, + "learning_rate": 1.3103873494849847e-05, + "loss": 0.0242, + "step": 9607 + }, + { + "epoch": 1.0416305290546402, + "grad_norm": 2.6767289638519287, + "learning_rate": 1.3103148121282461e-05, + "loss": 0.0515, + "step": 9608 + }, + { + "epoch": 1.04173894189072, + "grad_norm": 10.136765480041504, + "learning_rate": 1.3102422747715073e-05, + "loss": 0.0639, + "step": 9609 + }, + { + "epoch": 1.0418473547267997, + "grad_norm": 0.5638269186019897, + "learning_rate": 1.3101697374147686e-05, + "loss": 0.0024, + "step": 9610 + }, + { + "epoch": 1.0419557675628794, + "grad_norm": 4.526321887969971, + "learning_rate": 1.3100972000580301e-05, + "loss": 0.0424, + "step": 9611 + }, + { + "epoch": 1.0420641803989592, + "grad_norm": 3.784024238586426, + "learning_rate": 1.3100246627012913e-05, + "loss": 0.0197, + "step": 9612 + }, + { + "epoch": 1.042172593235039, + "grad_norm": 3.5705840587615967, + "learning_rate": 1.3099521253445527e-05, + "loss": 0.0488, + "step": 9613 + }, + { + "epoch": 1.0422810060711187, + "grad_norm": 6.392776966094971, + "learning_rate": 1.309879587987814e-05, + "loss": 0.0473, + "step": 9614 + }, + { + "epoch": 1.0423894189071987, + "grad_norm": 0.47550079226493835, + "learning_rate": 1.3098070506310751e-05, + "loss": 0.0049, + "step": 9615 + }, + { + "epoch": 1.0424978317432785, + "grad_norm": 0.7022923827171326, + "learning_rate": 1.3097345132743363e-05, + "loss": 0.0067, + "step": 9616 + }, + { + "epoch": 1.0426062445793582, + "grad_norm": 0.46771296858787537, + "learning_rate": 1.3096619759175977e-05, + "loss": 0.0027, + "step": 9617 + }, + { + "epoch": 1.042714657415438, + "grad_norm": 3.700401782989502, + "learning_rate": 1.309589438560859e-05, + "loss": 0.0116, + "step": 9618 + }, + { + "epoch": 1.0428230702515178, + "grad_norm": 0.8040894865989685, + "learning_rate": 1.3095169012041202e-05, + "loss": 0.0088, + "step": 9619 + }, + { + "epoch": 1.0429314830875975, + "grad_norm": 1.3244301080703735, + "learning_rate": 1.3094443638473815e-05, + "loss": 0.038, + "step": 9620 + }, + { + "epoch": 1.0430398959236773, + "grad_norm": 0.806926965713501, + "learning_rate": 1.3093718264906428e-05, + "loss": 0.0075, + "step": 9621 + }, + { + "epoch": 1.043148308759757, + "grad_norm": 5.730484962463379, + "learning_rate": 1.309299289133904e-05, + "loss": 0.0711, + "step": 9622 + }, + { + "epoch": 1.043256721595837, + "grad_norm": 1.1369876861572266, + "learning_rate": 1.3092267517771654e-05, + "loss": 0.0148, + "step": 9623 + }, + { + "epoch": 1.0433651344319168, + "grad_norm": 2.2863690853118896, + "learning_rate": 1.3091542144204266e-05, + "loss": 0.0112, + "step": 9624 + }, + { + "epoch": 1.0434735472679966, + "grad_norm": 2.2446506023406982, + "learning_rate": 1.3090816770636878e-05, + "loss": 0.0215, + "step": 9625 + }, + { + "epoch": 1.0435819601040763, + "grad_norm": 13.966079711914062, + "learning_rate": 1.3090091397069492e-05, + "loss": 0.0246, + "step": 9626 + }, + { + "epoch": 1.043690372940156, + "grad_norm": 8.931628227233887, + "learning_rate": 1.3089366023502104e-05, + "loss": 0.0222, + "step": 9627 + }, + { + "epoch": 1.0437987857762359, + "grad_norm": 2.491696357727051, + "learning_rate": 1.3088640649934718e-05, + "loss": 0.0346, + "step": 9628 + }, + { + "epoch": 1.0439071986123156, + "grad_norm": 0.488510400056839, + "learning_rate": 1.3087915276367332e-05, + "loss": 0.0025, + "step": 9629 + }, + { + "epoch": 1.0440156114483954, + "grad_norm": 4.185714244842529, + "learning_rate": 1.3087189902799944e-05, + "loss": 0.012, + "step": 9630 + }, + { + "epoch": 1.0441240242844754, + "grad_norm": 0.5878823399543762, + "learning_rate": 1.3086464529232556e-05, + "loss": 0.0057, + "step": 9631 + }, + { + "epoch": 1.0442324371205551, + "grad_norm": 3.957395315170288, + "learning_rate": 1.308573915566517e-05, + "loss": 0.0243, + "step": 9632 + }, + { + "epoch": 1.044340849956635, + "grad_norm": 0.49074557423591614, + "learning_rate": 1.3085013782097782e-05, + "loss": 0.0031, + "step": 9633 + }, + { + "epoch": 1.0444492627927147, + "grad_norm": 1.3745478391647339, + "learning_rate": 1.3084288408530394e-05, + "loss": 0.0103, + "step": 9634 + }, + { + "epoch": 1.0445576756287944, + "grad_norm": 7.010217666625977, + "learning_rate": 1.3083563034963008e-05, + "loss": 0.0502, + "step": 9635 + }, + { + "epoch": 1.0446660884648742, + "grad_norm": 1.9822090864181519, + "learning_rate": 1.308283766139562e-05, + "loss": 0.0111, + "step": 9636 + }, + { + "epoch": 1.044774501300954, + "grad_norm": 3.015834331512451, + "learning_rate": 1.3082112287828232e-05, + "loss": 0.0384, + "step": 9637 + }, + { + "epoch": 1.044882914137034, + "grad_norm": 3.8797152042388916, + "learning_rate": 1.3081386914260846e-05, + "loss": 0.023, + "step": 9638 + }, + { + "epoch": 1.0449913269731137, + "grad_norm": 1.5294798612594604, + "learning_rate": 1.3080661540693458e-05, + "loss": 0.0118, + "step": 9639 + }, + { + "epoch": 1.0450997398091935, + "grad_norm": 1.0301700830459595, + "learning_rate": 1.307993616712607e-05, + "loss": 0.0081, + "step": 9640 + }, + { + "epoch": 1.0452081526452732, + "grad_norm": 3.0316686630249023, + "learning_rate": 1.3079210793558682e-05, + "loss": 0.0096, + "step": 9641 + }, + { + "epoch": 1.045316565481353, + "grad_norm": 4.794443130493164, + "learning_rate": 1.3078485419991296e-05, + "loss": 0.041, + "step": 9642 + }, + { + "epoch": 1.0454249783174328, + "grad_norm": 3.44374942779541, + "learning_rate": 1.3077760046423908e-05, + "loss": 0.0314, + "step": 9643 + }, + { + "epoch": 1.0455333911535125, + "grad_norm": 8.800759315490723, + "learning_rate": 1.307703467285652e-05, + "loss": 0.0533, + "step": 9644 + }, + { + "epoch": 1.0456418039895923, + "grad_norm": 0.851231038570404, + "learning_rate": 1.3076309299289136e-05, + "loss": 0.0061, + "step": 9645 + }, + { + "epoch": 1.0457502168256723, + "grad_norm": 1.3145511150360107, + "learning_rate": 1.3075583925721748e-05, + "loss": 0.0058, + "step": 9646 + }, + { + "epoch": 1.045858629661752, + "grad_norm": 1.5607351064682007, + "learning_rate": 1.3074858552154362e-05, + "loss": 0.0123, + "step": 9647 + }, + { + "epoch": 1.0459670424978318, + "grad_norm": 2.653270959854126, + "learning_rate": 1.3074133178586974e-05, + "loss": 0.0152, + "step": 9648 + }, + { + "epoch": 1.0460754553339116, + "grad_norm": 4.52673864364624, + "learning_rate": 1.3073407805019586e-05, + "loss": 0.0681, + "step": 9649 + }, + { + "epoch": 1.0461838681699913, + "grad_norm": 4.024427890777588, + "learning_rate": 1.30726824314522e-05, + "loss": 0.0109, + "step": 9650 + }, + { + "epoch": 1.046292281006071, + "grad_norm": 5.561961650848389, + "learning_rate": 1.3071957057884812e-05, + "loss": 0.0961, + "step": 9651 + }, + { + "epoch": 1.0464006938421508, + "grad_norm": 2.4048709869384766, + "learning_rate": 1.3071231684317424e-05, + "loss": 0.0074, + "step": 9652 + }, + { + "epoch": 1.0465091066782306, + "grad_norm": 4.881707668304443, + "learning_rate": 1.3070506310750038e-05, + "loss": 0.0393, + "step": 9653 + }, + { + "epoch": 1.0466175195143106, + "grad_norm": 2.9111502170562744, + "learning_rate": 1.306978093718265e-05, + "loss": 0.021, + "step": 9654 + }, + { + "epoch": 1.0467259323503904, + "grad_norm": 2.9614551067352295, + "learning_rate": 1.3069055563615263e-05, + "loss": 0.0264, + "step": 9655 + }, + { + "epoch": 1.0468343451864701, + "grad_norm": 1.1010103225708008, + "learning_rate": 1.3068330190047875e-05, + "loss": 0.0034, + "step": 9656 + }, + { + "epoch": 1.0469427580225499, + "grad_norm": 0.28847265243530273, + "learning_rate": 1.3067604816480488e-05, + "loss": 0.0022, + "step": 9657 + }, + { + "epoch": 1.0470511708586296, + "grad_norm": 2.1636457443237305, + "learning_rate": 1.30668794429131e-05, + "loss": 0.0164, + "step": 9658 + }, + { + "epoch": 1.0471595836947094, + "grad_norm": 2.1508376598358154, + "learning_rate": 1.3066154069345713e-05, + "loss": 0.0117, + "step": 9659 + }, + { + "epoch": 1.0472679965307892, + "grad_norm": 3.1717493534088135, + "learning_rate": 1.3065428695778327e-05, + "loss": 0.0101, + "step": 9660 + }, + { + "epoch": 1.047376409366869, + "grad_norm": 0.5581274032592773, + "learning_rate": 1.3064703322210939e-05, + "loss": 0.0058, + "step": 9661 + }, + { + "epoch": 1.047484822202949, + "grad_norm": 3.932213544845581, + "learning_rate": 1.3063977948643551e-05, + "loss": 0.031, + "step": 9662 + }, + { + "epoch": 1.0475932350390287, + "grad_norm": 0.029185833409428596, + "learning_rate": 1.3063252575076166e-05, + "loss": 0.0004, + "step": 9663 + }, + { + "epoch": 1.0477016478751084, + "grad_norm": 0.2708066403865814, + "learning_rate": 1.3062527201508779e-05, + "loss": 0.0022, + "step": 9664 + }, + { + "epoch": 1.0478100607111882, + "grad_norm": 0.8112379312515259, + "learning_rate": 1.3061801827941392e-05, + "loss": 0.0062, + "step": 9665 + }, + { + "epoch": 1.047918473547268, + "grad_norm": 2.429957866668701, + "learning_rate": 1.3061076454374005e-05, + "loss": 0.0111, + "step": 9666 + }, + { + "epoch": 1.0480268863833477, + "grad_norm": 2.0663280487060547, + "learning_rate": 1.3060351080806617e-05, + "loss": 0.0071, + "step": 9667 + }, + { + "epoch": 1.0481352992194275, + "grad_norm": 0.8058027029037476, + "learning_rate": 1.3059625707239229e-05, + "loss": 0.0054, + "step": 9668 + }, + { + "epoch": 1.0482437120555073, + "grad_norm": 3.678502321243286, + "learning_rate": 1.3058900333671843e-05, + "loss": 0.0461, + "step": 9669 + }, + { + "epoch": 1.0483521248915872, + "grad_norm": 1.2106748819351196, + "learning_rate": 1.3058174960104455e-05, + "loss": 0.0238, + "step": 9670 + }, + { + "epoch": 1.048460537727667, + "grad_norm": 6.096736431121826, + "learning_rate": 1.3057449586537067e-05, + "loss": 0.0281, + "step": 9671 + }, + { + "epoch": 1.0485689505637468, + "grad_norm": 2.9138965606689453, + "learning_rate": 1.305672421296968e-05, + "loss": 0.0462, + "step": 9672 + }, + { + "epoch": 1.0486773633998265, + "grad_norm": 5.463394641876221, + "learning_rate": 1.3055998839402293e-05, + "loss": 0.0077, + "step": 9673 + }, + { + "epoch": 1.0487857762359063, + "grad_norm": 1.9843932390213013, + "learning_rate": 1.3055273465834905e-05, + "loss": 0.0346, + "step": 9674 + }, + { + "epoch": 1.048894189071986, + "grad_norm": 4.090080738067627, + "learning_rate": 1.3054548092267519e-05, + "loss": 0.0271, + "step": 9675 + }, + { + "epoch": 1.0490026019080658, + "grad_norm": 3.5714545249938965, + "learning_rate": 1.3053822718700131e-05, + "loss": 0.0441, + "step": 9676 + }, + { + "epoch": 1.0491110147441458, + "grad_norm": 5.521245956420898, + "learning_rate": 1.3053097345132743e-05, + "loss": 0.0182, + "step": 9677 + }, + { + "epoch": 1.0492194275802256, + "grad_norm": 5.197751045227051, + "learning_rate": 1.3052371971565357e-05, + "loss": 0.0136, + "step": 9678 + }, + { + "epoch": 1.0493278404163053, + "grad_norm": 1.597524642944336, + "learning_rate": 1.305164659799797e-05, + "loss": 0.0196, + "step": 9679 + }, + { + "epoch": 1.049436253252385, + "grad_norm": 1.45485258102417, + "learning_rate": 1.3050921224430583e-05, + "loss": 0.0077, + "step": 9680 + }, + { + "epoch": 1.0495446660884649, + "grad_norm": 7.784924507141113, + "learning_rate": 1.3050195850863197e-05, + "loss": 0.0603, + "step": 9681 + }, + { + "epoch": 1.0496530789245446, + "grad_norm": 0.9974430203437805, + "learning_rate": 1.3049470477295809e-05, + "loss": 0.0081, + "step": 9682 + }, + { + "epoch": 1.0497614917606244, + "grad_norm": 3.4537124633789062, + "learning_rate": 1.3048745103728421e-05, + "loss": 0.0085, + "step": 9683 + }, + { + "epoch": 1.0498699045967042, + "grad_norm": 2.9831552505493164, + "learning_rate": 1.3048019730161035e-05, + "loss": 0.0471, + "step": 9684 + }, + { + "epoch": 1.0499783174327841, + "grad_norm": 2.547170877456665, + "learning_rate": 1.3047294356593647e-05, + "loss": 0.0505, + "step": 9685 + }, + { + "epoch": 1.050086730268864, + "grad_norm": 4.572047710418701, + "learning_rate": 1.304656898302626e-05, + "loss": 0.0186, + "step": 9686 + }, + { + "epoch": 1.0501951431049437, + "grad_norm": 6.744280815124512, + "learning_rate": 1.3045843609458873e-05, + "loss": 0.037, + "step": 9687 + }, + { + "epoch": 1.0503035559410234, + "grad_norm": 2.1851871013641357, + "learning_rate": 1.3045118235891485e-05, + "loss": 0.0172, + "step": 9688 + }, + { + "epoch": 1.0504119687771032, + "grad_norm": 3.8777575492858887, + "learning_rate": 1.3044392862324097e-05, + "loss": 0.0261, + "step": 9689 + }, + { + "epoch": 1.050520381613183, + "grad_norm": 3.660984992980957, + "learning_rate": 1.3043667488756711e-05, + "loss": 0.0315, + "step": 9690 + }, + { + "epoch": 1.0506287944492627, + "grad_norm": 0.5201823115348816, + "learning_rate": 1.3042942115189323e-05, + "loss": 0.0016, + "step": 9691 + }, + { + "epoch": 1.0507372072853425, + "grad_norm": 1.9124720096588135, + "learning_rate": 1.3042216741621936e-05, + "loss": 0.0387, + "step": 9692 + }, + { + "epoch": 1.0508456201214225, + "grad_norm": 2.939878463745117, + "learning_rate": 1.304149136805455e-05, + "loss": 0.0128, + "step": 9693 + }, + { + "epoch": 1.0509540329575022, + "grad_norm": 0.9632169008255005, + "learning_rate": 1.3040765994487162e-05, + "loss": 0.0055, + "step": 9694 + }, + { + "epoch": 1.051062445793582, + "grad_norm": 4.557284355163574, + "learning_rate": 1.3040040620919774e-05, + "loss": 0.0168, + "step": 9695 + }, + { + "epoch": 1.0511708586296618, + "grad_norm": 3.699101448059082, + "learning_rate": 1.3039315247352386e-05, + "loss": 0.0127, + "step": 9696 + }, + { + "epoch": 1.0512792714657415, + "grad_norm": 4.081125259399414, + "learning_rate": 1.3038589873785001e-05, + "loss": 0.0097, + "step": 9697 + }, + { + "epoch": 1.0513876843018213, + "grad_norm": 2.1245222091674805, + "learning_rate": 1.3037864500217613e-05, + "loss": 0.0414, + "step": 9698 + }, + { + "epoch": 1.051496097137901, + "grad_norm": 0.518394947052002, + "learning_rate": 1.3037139126650227e-05, + "loss": 0.0026, + "step": 9699 + }, + { + "epoch": 1.051604509973981, + "grad_norm": 4.870002269744873, + "learning_rate": 1.303641375308284e-05, + "loss": 0.0292, + "step": 9700 + }, + { + "epoch": 1.0517129228100608, + "grad_norm": 6.6674113273620605, + "learning_rate": 1.3035688379515452e-05, + "loss": 0.0244, + "step": 9701 + }, + { + "epoch": 1.0518213356461406, + "grad_norm": 1.8903782367706299, + "learning_rate": 1.3034963005948065e-05, + "loss": 0.0156, + "step": 9702 + }, + { + "epoch": 1.0519297484822203, + "grad_norm": 1.1027103662490845, + "learning_rate": 1.3034237632380678e-05, + "loss": 0.0087, + "step": 9703 + }, + { + "epoch": 1.0520381613183, + "grad_norm": 1.1548618078231812, + "learning_rate": 1.303351225881329e-05, + "loss": 0.0109, + "step": 9704 + }, + { + "epoch": 1.0521465741543798, + "grad_norm": 0.13332870602607727, + "learning_rate": 1.3032786885245904e-05, + "loss": 0.0012, + "step": 9705 + }, + { + "epoch": 1.0522549869904596, + "grad_norm": 4.2781267166137695, + "learning_rate": 1.3032061511678516e-05, + "loss": 0.0548, + "step": 9706 + }, + { + "epoch": 1.0523633998265394, + "grad_norm": 1.3620363473892212, + "learning_rate": 1.3031336138111128e-05, + "loss": 0.0122, + "step": 9707 + }, + { + "epoch": 1.0524718126626194, + "grad_norm": 3.0491318702697754, + "learning_rate": 1.303061076454374e-05, + "loss": 0.0036, + "step": 9708 + }, + { + "epoch": 1.0525802254986991, + "grad_norm": 0.6609632968902588, + "learning_rate": 1.3029885390976354e-05, + "loss": 0.0031, + "step": 9709 + }, + { + "epoch": 1.0526886383347789, + "grad_norm": 4.357736587524414, + "learning_rate": 1.3029160017408966e-05, + "loss": 0.0461, + "step": 9710 + }, + { + "epoch": 1.0527970511708586, + "grad_norm": 3.892932176589966, + "learning_rate": 1.3028434643841578e-05, + "loss": 0.0317, + "step": 9711 + }, + { + "epoch": 1.0529054640069384, + "grad_norm": 0.3604851961135864, + "learning_rate": 1.3027709270274192e-05, + "loss": 0.0028, + "step": 9712 + }, + { + "epoch": 1.0530138768430182, + "grad_norm": 2.976585626602173, + "learning_rate": 1.3026983896706804e-05, + "loss": 0.0301, + "step": 9713 + }, + { + "epoch": 1.053122289679098, + "grad_norm": 4.173663139343262, + "learning_rate": 1.302625852313942e-05, + "loss": 0.0139, + "step": 9714 + }, + { + "epoch": 1.0532307025151777, + "grad_norm": 3.338691234588623, + "learning_rate": 1.3025533149572032e-05, + "loss": 0.0244, + "step": 9715 + }, + { + "epoch": 1.0533391153512577, + "grad_norm": 4.078210353851318, + "learning_rate": 1.3024807776004644e-05, + "loss": 0.0188, + "step": 9716 + }, + { + "epoch": 1.0534475281873374, + "grad_norm": 2.9195244312286377, + "learning_rate": 1.3024082402437258e-05, + "loss": 0.0133, + "step": 9717 + }, + { + "epoch": 1.0535559410234172, + "grad_norm": 7.209257125854492, + "learning_rate": 1.302335702886987e-05, + "loss": 0.0323, + "step": 9718 + }, + { + "epoch": 1.053664353859497, + "grad_norm": 6.207608699798584, + "learning_rate": 1.3022631655302482e-05, + "loss": 0.0347, + "step": 9719 + }, + { + "epoch": 1.0537727666955767, + "grad_norm": 2.896144151687622, + "learning_rate": 1.3021906281735094e-05, + "loss": 0.0112, + "step": 9720 + }, + { + "epoch": 1.0538811795316565, + "grad_norm": 0.39565756916999817, + "learning_rate": 1.3021180908167708e-05, + "loss": 0.0039, + "step": 9721 + }, + { + "epoch": 1.0539895923677363, + "grad_norm": 3.256970167160034, + "learning_rate": 1.302045553460032e-05, + "loss": 0.034, + "step": 9722 + }, + { + "epoch": 1.0540980052038162, + "grad_norm": 5.305150032043457, + "learning_rate": 1.3019730161032932e-05, + "loss": 0.0455, + "step": 9723 + }, + { + "epoch": 1.054206418039896, + "grad_norm": 0.7377846240997314, + "learning_rate": 1.3019004787465546e-05, + "loss": 0.0104, + "step": 9724 + }, + { + "epoch": 1.0543148308759758, + "grad_norm": 5.6527419090271, + "learning_rate": 1.3018279413898158e-05, + "loss": 0.0418, + "step": 9725 + }, + { + "epoch": 1.0544232437120555, + "grad_norm": 2.0357120037078857, + "learning_rate": 1.301755404033077e-05, + "loss": 0.0389, + "step": 9726 + }, + { + "epoch": 1.0545316565481353, + "grad_norm": 2.9760806560516357, + "learning_rate": 1.3016828666763384e-05, + "loss": 0.0478, + "step": 9727 + }, + { + "epoch": 1.054640069384215, + "grad_norm": 0.23494091629981995, + "learning_rate": 1.3016103293195996e-05, + "loss": 0.0027, + "step": 9728 + }, + { + "epoch": 1.0547484822202948, + "grad_norm": 2.147491931915283, + "learning_rate": 1.3015377919628609e-05, + "loss": 0.0242, + "step": 9729 + }, + { + "epoch": 1.0548568950563746, + "grad_norm": 5.840658664703369, + "learning_rate": 1.3014652546061222e-05, + "loss": 0.038, + "step": 9730 + }, + { + "epoch": 1.0549653078924546, + "grad_norm": 3.18574857711792, + "learning_rate": 1.3013927172493835e-05, + "loss": 0.0164, + "step": 9731 + }, + { + "epoch": 1.0550737207285343, + "grad_norm": 4.426056861877441, + "learning_rate": 1.3013201798926448e-05, + "loss": 0.0277, + "step": 9732 + }, + { + "epoch": 1.055182133564614, + "grad_norm": 1.7245723009109497, + "learning_rate": 1.3012476425359062e-05, + "loss": 0.0455, + "step": 9733 + }, + { + "epoch": 1.0552905464006939, + "grad_norm": 0.9395841360092163, + "learning_rate": 1.3011751051791674e-05, + "loss": 0.0079, + "step": 9734 + }, + { + "epoch": 1.0553989592367736, + "grad_norm": 1.186076283454895, + "learning_rate": 1.3011025678224286e-05, + "loss": 0.01, + "step": 9735 + }, + { + "epoch": 1.0555073720728534, + "grad_norm": 0.7256186604499817, + "learning_rate": 1.30103003046569e-05, + "loss": 0.0081, + "step": 9736 + }, + { + "epoch": 1.0556157849089332, + "grad_norm": 0.24933260679244995, + "learning_rate": 1.3009574931089512e-05, + "loss": 0.0015, + "step": 9737 + }, + { + "epoch": 1.055724197745013, + "grad_norm": 2.7544825077056885, + "learning_rate": 1.3008849557522125e-05, + "loss": 0.0108, + "step": 9738 + }, + { + "epoch": 1.055832610581093, + "grad_norm": 0.6612237691879272, + "learning_rate": 1.3008124183954738e-05, + "loss": 0.0052, + "step": 9739 + }, + { + "epoch": 1.0559410234171727, + "grad_norm": 0.16658346354961395, + "learning_rate": 1.300739881038735e-05, + "loss": 0.0013, + "step": 9740 + }, + { + "epoch": 1.0560494362532524, + "grad_norm": 1.7885116338729858, + "learning_rate": 1.3006673436819963e-05, + "loss": 0.0367, + "step": 9741 + }, + { + "epoch": 1.0561578490893322, + "grad_norm": 2.1524882316589355, + "learning_rate": 1.3005948063252577e-05, + "loss": 0.0245, + "step": 9742 + }, + { + "epoch": 1.056266261925412, + "grad_norm": 4.509793758392334, + "learning_rate": 1.3005222689685189e-05, + "loss": 0.0656, + "step": 9743 + }, + { + "epoch": 1.0563746747614917, + "grad_norm": 2.778512477874756, + "learning_rate": 1.3004497316117801e-05, + "loss": 0.01, + "step": 9744 + }, + { + "epoch": 1.0564830875975715, + "grad_norm": 2.6931819915771484, + "learning_rate": 1.3003771942550415e-05, + "loss": 0.0286, + "step": 9745 + }, + { + "epoch": 1.0565915004336512, + "grad_norm": 2.9986562728881836, + "learning_rate": 1.3003046568983027e-05, + "loss": 0.0212, + "step": 9746 + }, + { + "epoch": 1.0566999132697312, + "grad_norm": 2.945216417312622, + "learning_rate": 1.3002321195415639e-05, + "loss": 0.0166, + "step": 9747 + }, + { + "epoch": 1.056808326105811, + "grad_norm": 4.987485885620117, + "learning_rate": 1.3001595821848251e-05, + "loss": 0.03, + "step": 9748 + }, + { + "epoch": 1.0569167389418908, + "grad_norm": 0.22160720825195312, + "learning_rate": 1.3000870448280867e-05, + "loss": 0.0017, + "step": 9749 + }, + { + "epoch": 1.0570251517779705, + "grad_norm": 3.657862424850464, + "learning_rate": 1.3000145074713479e-05, + "loss": 0.0128, + "step": 9750 + }, + { + "epoch": 1.0571335646140503, + "grad_norm": 0.03447316959500313, + "learning_rate": 1.2999419701146093e-05, + "loss": 0.0002, + "step": 9751 + }, + { + "epoch": 1.05724197745013, + "grad_norm": 2.833768606185913, + "learning_rate": 1.2998694327578705e-05, + "loss": 0.028, + "step": 9752 + }, + { + "epoch": 1.0573503902862098, + "grad_norm": 2.264411449432373, + "learning_rate": 1.2997968954011317e-05, + "loss": 0.0171, + "step": 9753 + }, + { + "epoch": 1.0574588031222896, + "grad_norm": 2.0730202198028564, + "learning_rate": 1.299724358044393e-05, + "loss": 0.0084, + "step": 9754 + }, + { + "epoch": 1.0575672159583696, + "grad_norm": 1.3524513244628906, + "learning_rate": 1.2996518206876543e-05, + "loss": 0.0067, + "step": 9755 + }, + { + "epoch": 1.0576756287944493, + "grad_norm": 0.5102637410163879, + "learning_rate": 1.2995792833309155e-05, + "loss": 0.0018, + "step": 9756 + }, + { + "epoch": 1.057784041630529, + "grad_norm": 2.2553441524505615, + "learning_rate": 1.2995067459741769e-05, + "loss": 0.0293, + "step": 9757 + }, + { + "epoch": 1.0578924544666088, + "grad_norm": 0.14218880236148834, + "learning_rate": 1.2994342086174381e-05, + "loss": 0.0011, + "step": 9758 + }, + { + "epoch": 1.0580008673026886, + "grad_norm": 3.2857165336608887, + "learning_rate": 1.2993616712606993e-05, + "loss": 0.0189, + "step": 9759 + }, + { + "epoch": 1.0581092801387684, + "grad_norm": 3.655541181564331, + "learning_rate": 1.2992891339039605e-05, + "loss": 0.017, + "step": 9760 + }, + { + "epoch": 1.0582176929748481, + "grad_norm": 5.21370267868042, + "learning_rate": 1.299216596547222e-05, + "loss": 0.0409, + "step": 9761 + }, + { + "epoch": 1.0583261058109281, + "grad_norm": 3.50141978263855, + "learning_rate": 1.2991440591904831e-05, + "loss": 0.029, + "step": 9762 + }, + { + "epoch": 1.0584345186470079, + "grad_norm": 1.0344538688659668, + "learning_rate": 1.2990715218337443e-05, + "loss": 0.011, + "step": 9763 + }, + { + "epoch": 1.0585429314830876, + "grad_norm": 0.2497597336769104, + "learning_rate": 1.2989989844770057e-05, + "loss": 0.0022, + "step": 9764 + }, + { + "epoch": 1.0586513443191674, + "grad_norm": 1.813018798828125, + "learning_rate": 1.298926447120267e-05, + "loss": 0.0182, + "step": 9765 + }, + { + "epoch": 1.0587597571552472, + "grad_norm": 1.9354408979415894, + "learning_rate": 1.2988539097635285e-05, + "loss": 0.0189, + "step": 9766 + }, + { + "epoch": 1.058868169991327, + "grad_norm": 0.6743407249450684, + "learning_rate": 1.2987813724067897e-05, + "loss": 0.005, + "step": 9767 + }, + { + "epoch": 1.0589765828274067, + "grad_norm": 0.9707982540130615, + "learning_rate": 1.298708835050051e-05, + "loss": 0.004, + "step": 9768 + }, + { + "epoch": 1.0590849956634865, + "grad_norm": 4.938325881958008, + "learning_rate": 1.2986362976933123e-05, + "loss": 0.0102, + "step": 9769 + }, + { + "epoch": 1.0591934084995664, + "grad_norm": 3.9242660999298096, + "learning_rate": 1.2985637603365735e-05, + "loss": 0.0264, + "step": 9770 + }, + { + "epoch": 1.0593018213356462, + "grad_norm": 3.292689800262451, + "learning_rate": 1.2984912229798347e-05, + "loss": 0.0282, + "step": 9771 + }, + { + "epoch": 1.059410234171726, + "grad_norm": 5.263095855712891, + "learning_rate": 1.298418685623096e-05, + "loss": 0.0379, + "step": 9772 + }, + { + "epoch": 1.0595186470078057, + "grad_norm": 1.7570066452026367, + "learning_rate": 1.2983461482663573e-05, + "loss": 0.0282, + "step": 9773 + }, + { + "epoch": 1.0596270598438855, + "grad_norm": 5.307700157165527, + "learning_rate": 1.2982736109096185e-05, + "loss": 0.0482, + "step": 9774 + }, + { + "epoch": 1.0597354726799653, + "grad_norm": 5.467892169952393, + "learning_rate": 1.2982010735528798e-05, + "loss": 0.0554, + "step": 9775 + }, + { + "epoch": 1.059843885516045, + "grad_norm": 3.4813241958618164, + "learning_rate": 1.2981285361961411e-05, + "loss": 0.0383, + "step": 9776 + }, + { + "epoch": 1.0599522983521248, + "grad_norm": 3.5555083751678467, + "learning_rate": 1.2980559988394024e-05, + "loss": 0.0467, + "step": 9777 + }, + { + "epoch": 1.0600607111882048, + "grad_norm": 1.2783443927764893, + "learning_rate": 1.2979834614826636e-05, + "loss": 0.0071, + "step": 9778 + }, + { + "epoch": 1.0601691240242845, + "grad_norm": 0.03446350619196892, + "learning_rate": 1.297910924125925e-05, + "loss": 0.0002, + "step": 9779 + }, + { + "epoch": 1.0602775368603643, + "grad_norm": 2.813283920288086, + "learning_rate": 1.2978383867691862e-05, + "loss": 0.0212, + "step": 9780 + }, + { + "epoch": 1.060385949696444, + "grad_norm": 0.329120934009552, + "learning_rate": 1.2977658494124474e-05, + "loss": 0.0025, + "step": 9781 + }, + { + "epoch": 1.0604943625325238, + "grad_norm": 5.000430583953857, + "learning_rate": 1.2976933120557088e-05, + "loss": 0.0279, + "step": 9782 + }, + { + "epoch": 1.0606027753686036, + "grad_norm": 5.057465553283691, + "learning_rate": 1.29762077469897e-05, + "loss": 0.0497, + "step": 9783 + }, + { + "epoch": 1.0607111882046834, + "grad_norm": 0.5195196866989136, + "learning_rate": 1.2975482373422314e-05, + "loss": 0.0038, + "step": 9784 + }, + { + "epoch": 1.0608196010407633, + "grad_norm": 0.9592354893684387, + "learning_rate": 1.2974756999854928e-05, + "loss": 0.0021, + "step": 9785 + }, + { + "epoch": 1.060928013876843, + "grad_norm": 2.4405016899108887, + "learning_rate": 1.297403162628754e-05, + "loss": 0.0387, + "step": 9786 + }, + { + "epoch": 1.0610364267129229, + "grad_norm": 4.107253551483154, + "learning_rate": 1.2973306252720152e-05, + "loss": 0.0295, + "step": 9787 + }, + { + "epoch": 1.0611448395490026, + "grad_norm": 0.7941290140151978, + "learning_rate": 1.2972580879152766e-05, + "loss": 0.0081, + "step": 9788 + }, + { + "epoch": 1.0612532523850824, + "grad_norm": 0.662915050983429, + "learning_rate": 1.2971855505585378e-05, + "loss": 0.0036, + "step": 9789 + }, + { + "epoch": 1.0613616652211622, + "grad_norm": 1.2539347410202026, + "learning_rate": 1.297113013201799e-05, + "loss": 0.0121, + "step": 9790 + }, + { + "epoch": 1.061470078057242, + "grad_norm": 4.862504482269287, + "learning_rate": 1.2970404758450604e-05, + "loss": 0.0174, + "step": 9791 + }, + { + "epoch": 1.0615784908933217, + "grad_norm": 4.05817174911499, + "learning_rate": 1.2969679384883216e-05, + "loss": 0.0512, + "step": 9792 + }, + { + "epoch": 1.0616869037294017, + "grad_norm": 1.11787748336792, + "learning_rate": 1.2968954011315828e-05, + "loss": 0.0093, + "step": 9793 + }, + { + "epoch": 1.0617953165654814, + "grad_norm": 5.443536281585693, + "learning_rate": 1.2968228637748442e-05, + "loss": 0.1058, + "step": 9794 + }, + { + "epoch": 1.0619037294015612, + "grad_norm": 4.616586208343506, + "learning_rate": 1.2967503264181054e-05, + "loss": 0.0238, + "step": 9795 + }, + { + "epoch": 1.062012142237641, + "grad_norm": 0.9318910241127014, + "learning_rate": 1.2966777890613666e-05, + "loss": 0.03, + "step": 9796 + }, + { + "epoch": 1.0621205550737207, + "grad_norm": 0.8645606637001038, + "learning_rate": 1.296605251704628e-05, + "loss": 0.0052, + "step": 9797 + }, + { + "epoch": 1.0622289679098005, + "grad_norm": 3.53139066696167, + "learning_rate": 1.2965327143478892e-05, + "loss": 0.0699, + "step": 9798 + }, + { + "epoch": 1.0623373807458802, + "grad_norm": 3.3563225269317627, + "learning_rate": 1.2964601769911504e-05, + "loss": 0.0099, + "step": 9799 + }, + { + "epoch": 1.06244579358196, + "grad_norm": 3.9967079162597656, + "learning_rate": 1.2963876396344116e-05, + "loss": 0.0141, + "step": 9800 + }, + { + "epoch": 1.06255420641804, + "grad_norm": 2.9865338802337646, + "learning_rate": 1.2963151022776732e-05, + "loss": 0.0236, + "step": 9801 + }, + { + "epoch": 1.0626626192541198, + "grad_norm": 4.608270645141602, + "learning_rate": 1.2962425649209344e-05, + "loss": 0.0376, + "step": 9802 + }, + { + "epoch": 1.0627710320901995, + "grad_norm": 2.2648379802703857, + "learning_rate": 1.2961700275641958e-05, + "loss": 0.0133, + "step": 9803 + }, + { + "epoch": 1.0628794449262793, + "grad_norm": 0.38861411809921265, + "learning_rate": 1.296097490207457e-05, + "loss": 0.0019, + "step": 9804 + }, + { + "epoch": 1.062987857762359, + "grad_norm": 4.1139445304870605, + "learning_rate": 1.2960249528507182e-05, + "loss": 0.0251, + "step": 9805 + }, + { + "epoch": 1.0630962705984388, + "grad_norm": 1.8840434551239014, + "learning_rate": 1.2959524154939796e-05, + "loss": 0.0117, + "step": 9806 + }, + { + "epoch": 1.0632046834345186, + "grad_norm": 2.534461259841919, + "learning_rate": 1.2958798781372408e-05, + "loss": 0.0283, + "step": 9807 + }, + { + "epoch": 1.0633130962705986, + "grad_norm": 2.999166965484619, + "learning_rate": 1.295807340780502e-05, + "loss": 0.019, + "step": 9808 + }, + { + "epoch": 1.0634215091066783, + "grad_norm": 3.0616228580474854, + "learning_rate": 1.2957348034237634e-05, + "loss": 0.0153, + "step": 9809 + }, + { + "epoch": 1.063529921942758, + "grad_norm": 1.2597479820251465, + "learning_rate": 1.2956622660670246e-05, + "loss": 0.0035, + "step": 9810 + }, + { + "epoch": 1.0636383347788378, + "grad_norm": 2.884744882583618, + "learning_rate": 1.2955897287102859e-05, + "loss": 0.0325, + "step": 9811 + }, + { + "epoch": 1.0637467476149176, + "grad_norm": 0.41919514536857605, + "learning_rate": 1.295517191353547e-05, + "loss": 0.0028, + "step": 9812 + }, + { + "epoch": 1.0638551604509974, + "grad_norm": 1.651247262954712, + "learning_rate": 1.2954446539968084e-05, + "loss": 0.0593, + "step": 9813 + }, + { + "epoch": 1.0639635732870771, + "grad_norm": 0.17371895909309387, + "learning_rate": 1.2953721166400697e-05, + "loss": 0.0012, + "step": 9814 + }, + { + "epoch": 1.064071986123157, + "grad_norm": 4.3390092849731445, + "learning_rate": 1.2952995792833309e-05, + "loss": 0.015, + "step": 9815 + }, + { + "epoch": 1.0641803989592367, + "grad_norm": 1.5647300481796265, + "learning_rate": 1.2952270419265923e-05, + "loss": 0.011, + "step": 9816 + }, + { + "epoch": 1.0642888117953166, + "grad_norm": 0.8562095165252686, + "learning_rate": 1.2951545045698535e-05, + "loss": 0.0039, + "step": 9817 + }, + { + "epoch": 1.0643972246313964, + "grad_norm": 1.477183222770691, + "learning_rate": 1.295081967213115e-05, + "loss": 0.0078, + "step": 9818 + }, + { + "epoch": 1.0645056374674762, + "grad_norm": 1.8591934442520142, + "learning_rate": 1.2950094298563762e-05, + "loss": 0.0278, + "step": 9819 + }, + { + "epoch": 1.064614050303556, + "grad_norm": 2.0862107276916504, + "learning_rate": 1.2949368924996375e-05, + "loss": 0.0225, + "step": 9820 + }, + { + "epoch": 1.0647224631396357, + "grad_norm": 4.201657772064209, + "learning_rate": 1.2948643551428988e-05, + "loss": 0.0311, + "step": 9821 + }, + { + "epoch": 1.0648308759757155, + "grad_norm": 3.388580083847046, + "learning_rate": 1.29479181778616e-05, + "loss": 0.0448, + "step": 9822 + }, + { + "epoch": 1.0649392888117952, + "grad_norm": 2.0859785079956055, + "learning_rate": 1.2947192804294213e-05, + "loss": 0.0102, + "step": 9823 + }, + { + "epoch": 1.0650477016478752, + "grad_norm": 3.0322957038879395, + "learning_rate": 1.2946467430726825e-05, + "loss": 0.0223, + "step": 9824 + }, + { + "epoch": 1.065156114483955, + "grad_norm": 1.5103093385696411, + "learning_rate": 1.2945742057159439e-05, + "loss": 0.0105, + "step": 9825 + }, + { + "epoch": 1.0652645273200347, + "grad_norm": 1.1632415056228638, + "learning_rate": 1.294501668359205e-05, + "loss": 0.0082, + "step": 9826 + }, + { + "epoch": 1.0653729401561145, + "grad_norm": 0.7650364637374878, + "learning_rate": 1.2944291310024663e-05, + "loss": 0.0027, + "step": 9827 + }, + { + "epoch": 1.0654813529921943, + "grad_norm": 1.3435990810394287, + "learning_rate": 1.2943565936457277e-05, + "loss": 0.0333, + "step": 9828 + }, + { + "epoch": 1.065589765828274, + "grad_norm": 1.9156732559204102, + "learning_rate": 1.2942840562889889e-05, + "loss": 0.0193, + "step": 9829 + }, + { + "epoch": 1.0656981786643538, + "grad_norm": 7.762242794036865, + "learning_rate": 1.2942115189322501e-05, + "loss": 0.027, + "step": 9830 + }, + { + "epoch": 1.0658065915004336, + "grad_norm": 7.425129413604736, + "learning_rate": 1.2941389815755115e-05, + "loss": 0.0231, + "step": 9831 + }, + { + "epoch": 1.0659150043365135, + "grad_norm": 1.0044101476669312, + "learning_rate": 1.2940664442187727e-05, + "loss": 0.0081, + "step": 9832 + }, + { + "epoch": 1.0660234171725933, + "grad_norm": 1.5091191530227661, + "learning_rate": 1.293993906862034e-05, + "loss": 0.0039, + "step": 9833 + }, + { + "epoch": 1.066131830008673, + "grad_norm": 1.621502161026001, + "learning_rate": 1.2939213695052953e-05, + "loss": 0.0066, + "step": 9834 + }, + { + "epoch": 1.0662402428447528, + "grad_norm": 11.543394088745117, + "learning_rate": 1.2938488321485567e-05, + "loss": 0.0393, + "step": 9835 + }, + { + "epoch": 1.0663486556808326, + "grad_norm": 3.3331689834594727, + "learning_rate": 1.2937762947918179e-05, + "loss": 0.0353, + "step": 9836 + }, + { + "epoch": 1.0664570685169124, + "grad_norm": 3.0752687454223633, + "learning_rate": 1.2937037574350793e-05, + "loss": 0.0129, + "step": 9837 + }, + { + "epoch": 1.0665654813529921, + "grad_norm": 0.06221609190106392, + "learning_rate": 1.2936312200783405e-05, + "loss": 0.0004, + "step": 9838 + }, + { + "epoch": 1.0666738941890719, + "grad_norm": 3.4207377433776855, + "learning_rate": 1.2935586827216017e-05, + "loss": 0.0274, + "step": 9839 + }, + { + "epoch": 1.0667823070251519, + "grad_norm": 11.988481521606445, + "learning_rate": 1.2934861453648631e-05, + "loss": 0.0203, + "step": 9840 + }, + { + "epoch": 1.0668907198612316, + "grad_norm": 1.0517531633377075, + "learning_rate": 1.2934136080081243e-05, + "loss": 0.0113, + "step": 9841 + }, + { + "epoch": 1.0669991326973114, + "grad_norm": 1.2865164279937744, + "learning_rate": 1.2933410706513855e-05, + "loss": 0.008, + "step": 9842 + }, + { + "epoch": 1.0671075455333912, + "grad_norm": 0.11984287202358246, + "learning_rate": 1.2932685332946469e-05, + "loss": 0.0014, + "step": 9843 + }, + { + "epoch": 1.067215958369471, + "grad_norm": 1.8081369400024414, + "learning_rate": 1.2931959959379081e-05, + "loss": 0.0164, + "step": 9844 + }, + { + "epoch": 1.0673243712055507, + "grad_norm": 1.485372543334961, + "learning_rate": 1.2931234585811693e-05, + "loss": 0.0295, + "step": 9845 + }, + { + "epoch": 1.0674327840416304, + "grad_norm": 0.4234057068824768, + "learning_rate": 1.2930509212244307e-05, + "loss": 0.0012, + "step": 9846 + }, + { + "epoch": 1.0675411968777104, + "grad_norm": 0.37517431378364563, + "learning_rate": 1.292978383867692e-05, + "loss": 0.001, + "step": 9847 + }, + { + "epoch": 1.0676496097137902, + "grad_norm": 2.3773210048675537, + "learning_rate": 1.2929058465109532e-05, + "loss": 0.0146, + "step": 9848 + }, + { + "epoch": 1.06775802254987, + "grad_norm": 2.056351661682129, + "learning_rate": 1.2928333091542145e-05, + "loss": 0.0268, + "step": 9849 + }, + { + "epoch": 1.0678664353859497, + "grad_norm": 1.1039689779281616, + "learning_rate": 1.2927607717974758e-05, + "loss": 0.0081, + "step": 9850 + }, + { + "epoch": 1.0679748482220295, + "grad_norm": 5.370610237121582, + "learning_rate": 1.292688234440737e-05, + "loss": 0.0456, + "step": 9851 + }, + { + "epoch": 1.0680832610581092, + "grad_norm": 1.0348899364471436, + "learning_rate": 1.2926156970839982e-05, + "loss": 0.0071, + "step": 9852 + }, + { + "epoch": 1.068191673894189, + "grad_norm": 0.2143176943063736, + "learning_rate": 1.2925431597272597e-05, + "loss": 0.0017, + "step": 9853 + }, + { + "epoch": 1.0683000867302688, + "grad_norm": 8.30761432647705, + "learning_rate": 1.292470622370521e-05, + "loss": 0.0364, + "step": 9854 + }, + { + "epoch": 1.0684084995663488, + "grad_norm": 1.1529144048690796, + "learning_rate": 1.2923980850137823e-05, + "loss": 0.0074, + "step": 9855 + }, + { + "epoch": 1.0685169124024285, + "grad_norm": 1.9924708604812622, + "learning_rate": 1.2923255476570435e-05, + "loss": 0.0069, + "step": 9856 + }, + { + "epoch": 1.0686253252385083, + "grad_norm": 2.588531494140625, + "learning_rate": 1.2922530103003048e-05, + "loss": 0.028, + "step": 9857 + }, + { + "epoch": 1.068733738074588, + "grad_norm": 3.892995834350586, + "learning_rate": 1.2921804729435661e-05, + "loss": 0.0272, + "step": 9858 + }, + { + "epoch": 1.0688421509106678, + "grad_norm": 3.949800968170166, + "learning_rate": 1.2921079355868274e-05, + "loss": 0.0053, + "step": 9859 + }, + { + "epoch": 1.0689505637467476, + "grad_norm": 5.186405181884766, + "learning_rate": 1.2920353982300886e-05, + "loss": 0.0063, + "step": 9860 + }, + { + "epoch": 1.0690589765828273, + "grad_norm": 4.293986797332764, + "learning_rate": 1.29196286087335e-05, + "loss": 0.0104, + "step": 9861 + }, + { + "epoch": 1.069167389418907, + "grad_norm": 2.630086898803711, + "learning_rate": 1.2918903235166112e-05, + "loss": 0.0173, + "step": 9862 + }, + { + "epoch": 1.069275802254987, + "grad_norm": 5.767346382141113, + "learning_rate": 1.2918177861598724e-05, + "loss": 0.0481, + "step": 9863 + }, + { + "epoch": 1.0693842150910668, + "grad_norm": 3.194920539855957, + "learning_rate": 1.2917452488031336e-05, + "loss": 0.0397, + "step": 9864 + }, + { + "epoch": 1.0694926279271466, + "grad_norm": 0.2946666479110718, + "learning_rate": 1.291672711446395e-05, + "loss": 0.0015, + "step": 9865 + }, + { + "epoch": 1.0696010407632264, + "grad_norm": 2.440474033355713, + "learning_rate": 1.2916001740896562e-05, + "loss": 0.0028, + "step": 9866 + }, + { + "epoch": 1.0697094535993061, + "grad_norm": 1.1309338808059692, + "learning_rate": 1.2915276367329174e-05, + "loss": 0.0068, + "step": 9867 + }, + { + "epoch": 1.069817866435386, + "grad_norm": 3.2739882469177246, + "learning_rate": 1.2914550993761788e-05, + "loss": 0.0546, + "step": 9868 + }, + { + "epoch": 1.0699262792714657, + "grad_norm": 3.8982791900634766, + "learning_rate": 1.29138256201944e-05, + "loss": 0.0186, + "step": 9869 + }, + { + "epoch": 1.0700346921075456, + "grad_norm": 7.765040874481201, + "learning_rate": 1.2913100246627016e-05, + "loss": 0.0091, + "step": 9870 + }, + { + "epoch": 1.0701431049436254, + "grad_norm": 3.7249581813812256, + "learning_rate": 1.2912374873059628e-05, + "loss": 0.012, + "step": 9871 + }, + { + "epoch": 1.0702515177797052, + "grad_norm": 1.3497551679611206, + "learning_rate": 1.291164949949224e-05, + "loss": 0.0178, + "step": 9872 + }, + { + "epoch": 1.070359930615785, + "grad_norm": 1.7104679346084595, + "learning_rate": 1.2910924125924854e-05, + "loss": 0.0199, + "step": 9873 + }, + { + "epoch": 1.0704683434518647, + "grad_norm": 3.172987937927246, + "learning_rate": 1.2910198752357466e-05, + "loss": 0.0398, + "step": 9874 + }, + { + "epoch": 1.0705767562879445, + "grad_norm": 2.5528876781463623, + "learning_rate": 1.2909473378790078e-05, + "loss": 0.038, + "step": 9875 + }, + { + "epoch": 1.0706851691240242, + "grad_norm": 0.5729954242706299, + "learning_rate": 1.290874800522269e-05, + "loss": 0.004, + "step": 9876 + }, + { + "epoch": 1.070793581960104, + "grad_norm": 1.5889581441879272, + "learning_rate": 1.2908022631655304e-05, + "loss": 0.0214, + "step": 9877 + }, + { + "epoch": 1.0709019947961838, + "grad_norm": 1.8542770147323608, + "learning_rate": 1.2907297258087916e-05, + "loss": 0.0214, + "step": 9878 + }, + { + "epoch": 1.0710104076322637, + "grad_norm": 1.133204460144043, + "learning_rate": 1.2906571884520528e-05, + "loss": 0.0044, + "step": 9879 + }, + { + "epoch": 1.0711188204683435, + "grad_norm": 3.445004940032959, + "learning_rate": 1.2905846510953142e-05, + "loss": 0.0718, + "step": 9880 + }, + { + "epoch": 1.0712272333044233, + "grad_norm": 5.825837135314941, + "learning_rate": 1.2905121137385754e-05, + "loss": 0.0827, + "step": 9881 + }, + { + "epoch": 1.071335646140503, + "grad_norm": 1.7614126205444336, + "learning_rate": 1.2904395763818366e-05, + "loss": 0.0031, + "step": 9882 + }, + { + "epoch": 1.0714440589765828, + "grad_norm": 0.45025309920310974, + "learning_rate": 1.290367039025098e-05, + "loss": 0.0025, + "step": 9883 + }, + { + "epoch": 1.0715524718126626, + "grad_norm": 11.172690391540527, + "learning_rate": 1.2902945016683592e-05, + "loss": 0.0343, + "step": 9884 + }, + { + "epoch": 1.0716608846487423, + "grad_norm": 4.321569442749023, + "learning_rate": 1.2902219643116205e-05, + "loss": 0.0751, + "step": 9885 + }, + { + "epoch": 1.0717692974848223, + "grad_norm": 1.3681563138961792, + "learning_rate": 1.2901494269548818e-05, + "loss": 0.0202, + "step": 9886 + }, + { + "epoch": 1.071877710320902, + "grad_norm": 0.12191576510667801, + "learning_rate": 1.2900768895981432e-05, + "loss": 0.001, + "step": 9887 + }, + { + "epoch": 1.0719861231569818, + "grad_norm": 1.8789304494857788, + "learning_rate": 1.2900043522414046e-05, + "loss": 0.0194, + "step": 9888 + }, + { + "epoch": 1.0720945359930616, + "grad_norm": 2.742020845413208, + "learning_rate": 1.2899318148846658e-05, + "loss": 0.0609, + "step": 9889 + }, + { + "epoch": 1.0722029488291414, + "grad_norm": 0.9493685364723206, + "learning_rate": 1.289859277527927e-05, + "loss": 0.0047, + "step": 9890 + }, + { + "epoch": 1.0723113616652211, + "grad_norm": 1.9433414936065674, + "learning_rate": 1.2897867401711883e-05, + "loss": 0.0464, + "step": 9891 + }, + { + "epoch": 1.0724197745013009, + "grad_norm": 0.681700587272644, + "learning_rate": 1.2897142028144496e-05, + "loss": 0.0053, + "step": 9892 + }, + { + "epoch": 1.0725281873373806, + "grad_norm": 1.8113795518875122, + "learning_rate": 1.2896416654577108e-05, + "loss": 0.0088, + "step": 9893 + }, + { + "epoch": 1.0726366001734606, + "grad_norm": 8.115792274475098, + "learning_rate": 1.289569128100972e-05, + "loss": 0.0213, + "step": 9894 + }, + { + "epoch": 1.0727450130095404, + "grad_norm": 7.063379287719727, + "learning_rate": 1.2894965907442334e-05, + "loss": 0.0166, + "step": 9895 + }, + { + "epoch": 1.0728534258456202, + "grad_norm": 0.2523278594017029, + "learning_rate": 1.2894240533874947e-05, + "loss": 0.0026, + "step": 9896 + }, + { + "epoch": 1.0729618386817, + "grad_norm": 1.6092413663864136, + "learning_rate": 1.2893515160307559e-05, + "loss": 0.0167, + "step": 9897 + }, + { + "epoch": 1.0730702515177797, + "grad_norm": 0.6538949608802795, + "learning_rate": 1.2892789786740173e-05, + "loss": 0.0034, + "step": 9898 + }, + { + "epoch": 1.0731786643538594, + "grad_norm": 5.39265775680542, + "learning_rate": 1.2892064413172785e-05, + "loss": 0.0228, + "step": 9899 + }, + { + "epoch": 1.0732870771899392, + "grad_norm": 3.9766604900360107, + "learning_rate": 1.2891339039605397e-05, + "loss": 0.0264, + "step": 9900 + }, + { + "epoch": 1.073395490026019, + "grad_norm": 2.0788984298706055, + "learning_rate": 1.289061366603801e-05, + "loss": 0.0104, + "step": 9901 + }, + { + "epoch": 1.073503902862099, + "grad_norm": 3.1487538814544678, + "learning_rate": 1.2889888292470623e-05, + "loss": 0.0353, + "step": 9902 + }, + { + "epoch": 1.0736123156981787, + "grad_norm": 0.26118701696395874, + "learning_rate": 1.2889162918903235e-05, + "loss": 0.0018, + "step": 9903 + }, + { + "epoch": 1.0737207285342585, + "grad_norm": 0.8615208268165588, + "learning_rate": 1.288843754533585e-05, + "loss": 0.0035, + "step": 9904 + }, + { + "epoch": 1.0738291413703382, + "grad_norm": 3.0149283409118652, + "learning_rate": 1.2887712171768463e-05, + "loss": 0.025, + "step": 9905 + }, + { + "epoch": 1.073937554206418, + "grad_norm": 1.4355676174163818, + "learning_rate": 1.2886986798201075e-05, + "loss": 0.0313, + "step": 9906 + }, + { + "epoch": 1.0740459670424978, + "grad_norm": 14.608979225158691, + "learning_rate": 1.2886261424633689e-05, + "loss": 0.0119, + "step": 9907 + }, + { + "epoch": 1.0741543798785775, + "grad_norm": 0.6189460754394531, + "learning_rate": 1.28855360510663e-05, + "loss": 0.0034, + "step": 9908 + }, + { + "epoch": 1.0742627927146575, + "grad_norm": 1.45028555393219, + "learning_rate": 1.2884810677498913e-05, + "loss": 0.0073, + "step": 9909 + }, + { + "epoch": 1.0743712055507373, + "grad_norm": 2.9419877529144287, + "learning_rate": 1.2884085303931527e-05, + "loss": 0.017, + "step": 9910 + }, + { + "epoch": 1.074479618386817, + "grad_norm": 2.021641969680786, + "learning_rate": 1.2883359930364139e-05, + "loss": 0.0083, + "step": 9911 + }, + { + "epoch": 1.0745880312228968, + "grad_norm": 10.559528350830078, + "learning_rate": 1.2882634556796751e-05, + "loss": 0.0345, + "step": 9912 + }, + { + "epoch": 1.0746964440589766, + "grad_norm": 6.386816024780273, + "learning_rate": 1.2881909183229365e-05, + "loss": 0.0387, + "step": 9913 + }, + { + "epoch": 1.0748048568950563, + "grad_norm": 13.043672561645508, + "learning_rate": 1.2881183809661977e-05, + "loss": 0.0401, + "step": 9914 + }, + { + "epoch": 1.074913269731136, + "grad_norm": 2.2442948818206787, + "learning_rate": 1.288045843609459e-05, + "loss": 0.0034, + "step": 9915 + }, + { + "epoch": 1.0750216825672159, + "grad_norm": 5.027309894561768, + "learning_rate": 1.2879733062527201e-05, + "loss": 0.0135, + "step": 9916 + }, + { + "epoch": 1.0751300954032958, + "grad_norm": 4.1211256980896, + "learning_rate": 1.2879007688959815e-05, + "loss": 0.0521, + "step": 9917 + }, + { + "epoch": 1.0752385082393756, + "grad_norm": 2.153651714324951, + "learning_rate": 1.2878282315392427e-05, + "loss": 0.0377, + "step": 9918 + }, + { + "epoch": 1.0753469210754554, + "grad_norm": 0.9386270642280579, + "learning_rate": 1.287755694182504e-05, + "loss": 0.0035, + "step": 9919 + }, + { + "epoch": 1.0754553339115351, + "grad_norm": 0.8857551217079163, + "learning_rate": 1.2876831568257653e-05, + "loss": 0.0034, + "step": 9920 + }, + { + "epoch": 1.075563746747615, + "grad_norm": 7.861253261566162, + "learning_rate": 1.2876106194690265e-05, + "loss": 0.0353, + "step": 9921 + }, + { + "epoch": 1.0756721595836947, + "grad_norm": 0.404358834028244, + "learning_rate": 1.2875380821122881e-05, + "loss": 0.0025, + "step": 9922 + }, + { + "epoch": 1.0757805724197744, + "grad_norm": 2.1751458644866943, + "learning_rate": 1.2874655447555493e-05, + "loss": 0.022, + "step": 9923 + }, + { + "epoch": 1.0758889852558542, + "grad_norm": 2.7880070209503174, + "learning_rate": 1.2873930073988105e-05, + "loss": 0.0205, + "step": 9924 + }, + { + "epoch": 1.0759973980919342, + "grad_norm": 1.7018450498580933, + "learning_rate": 1.2873204700420719e-05, + "loss": 0.0132, + "step": 9925 + }, + { + "epoch": 1.076105810928014, + "grad_norm": 2.6970255374908447, + "learning_rate": 1.2872479326853331e-05, + "loss": 0.0147, + "step": 9926 + }, + { + "epoch": 1.0762142237640937, + "grad_norm": 3.71162748336792, + "learning_rate": 1.2871753953285943e-05, + "loss": 0.0543, + "step": 9927 + }, + { + "epoch": 1.0763226366001735, + "grad_norm": 3.009406328201294, + "learning_rate": 1.2871028579718556e-05, + "loss": 0.0385, + "step": 9928 + }, + { + "epoch": 1.0764310494362532, + "grad_norm": 2.600724697113037, + "learning_rate": 1.287030320615117e-05, + "loss": 0.0085, + "step": 9929 + }, + { + "epoch": 1.076539462272333, + "grad_norm": 3.364687919616699, + "learning_rate": 1.2869577832583782e-05, + "loss": 0.0104, + "step": 9930 + }, + { + "epoch": 1.0766478751084128, + "grad_norm": 0.8728182315826416, + "learning_rate": 1.2868852459016394e-05, + "loss": 0.0028, + "step": 9931 + }, + { + "epoch": 1.0767562879444927, + "grad_norm": 3.9293508529663086, + "learning_rate": 1.2868127085449007e-05, + "loss": 0.0477, + "step": 9932 + }, + { + "epoch": 1.0768647007805725, + "grad_norm": 3.185703992843628, + "learning_rate": 1.286740171188162e-05, + "loss": 0.0675, + "step": 9933 + }, + { + "epoch": 1.0769731136166523, + "grad_norm": 0.45358511805534363, + "learning_rate": 1.2866676338314232e-05, + "loss": 0.0022, + "step": 9934 + }, + { + "epoch": 1.077081526452732, + "grad_norm": 4.657661437988281, + "learning_rate": 1.2865950964746846e-05, + "loss": 0.046, + "step": 9935 + }, + { + "epoch": 1.0771899392888118, + "grad_norm": 3.891643524169922, + "learning_rate": 1.2865225591179458e-05, + "loss": 0.0087, + "step": 9936 + }, + { + "epoch": 1.0772983521248916, + "grad_norm": 4.245467662811279, + "learning_rate": 1.286450021761207e-05, + "loss": 0.0166, + "step": 9937 + }, + { + "epoch": 1.0774067649609713, + "grad_norm": 0.14598627388477325, + "learning_rate": 1.2863774844044684e-05, + "loss": 0.0015, + "step": 9938 + }, + { + "epoch": 1.077515177797051, + "grad_norm": 2.3526766300201416, + "learning_rate": 1.2863049470477298e-05, + "loss": 0.0452, + "step": 9939 + }, + { + "epoch": 1.0776235906331308, + "grad_norm": 1.3630528450012207, + "learning_rate": 1.2862324096909911e-05, + "loss": 0.0147, + "step": 9940 + }, + { + "epoch": 1.0777320034692108, + "grad_norm": 1.243538737297058, + "learning_rate": 1.2861598723342524e-05, + "loss": 0.0058, + "step": 9941 + }, + { + "epoch": 1.0778404163052906, + "grad_norm": 0.9909841418266296, + "learning_rate": 1.2860873349775136e-05, + "loss": 0.0053, + "step": 9942 + }, + { + "epoch": 1.0779488291413704, + "grad_norm": 3.6473679542541504, + "learning_rate": 1.2860147976207748e-05, + "loss": 0.0212, + "step": 9943 + }, + { + "epoch": 1.0780572419774501, + "grad_norm": 6.737545490264893, + "learning_rate": 1.2859422602640362e-05, + "loss": 0.0823, + "step": 9944 + }, + { + "epoch": 1.0781656548135299, + "grad_norm": 3.8939640522003174, + "learning_rate": 1.2858697229072974e-05, + "loss": 0.0415, + "step": 9945 + }, + { + "epoch": 1.0782740676496096, + "grad_norm": 2.2160065174102783, + "learning_rate": 1.2857971855505586e-05, + "loss": 0.0101, + "step": 9946 + }, + { + "epoch": 1.0783824804856894, + "grad_norm": 0.4603744149208069, + "learning_rate": 1.28572464819382e-05, + "loss": 0.0017, + "step": 9947 + }, + { + "epoch": 1.0784908933217694, + "grad_norm": 2.393056869506836, + "learning_rate": 1.2856521108370812e-05, + "loss": 0.0372, + "step": 9948 + }, + { + "epoch": 1.0785993061578492, + "grad_norm": 0.6679983735084534, + "learning_rate": 1.2855795734803424e-05, + "loss": 0.0077, + "step": 9949 + }, + { + "epoch": 1.078707718993929, + "grad_norm": 0.7264618873596191, + "learning_rate": 1.2855070361236038e-05, + "loss": 0.0031, + "step": 9950 + }, + { + "epoch": 1.0788161318300087, + "grad_norm": 1.274093508720398, + "learning_rate": 1.285434498766865e-05, + "loss": 0.0355, + "step": 9951 + }, + { + "epoch": 1.0789245446660884, + "grad_norm": 4.880984783172607, + "learning_rate": 1.2853619614101262e-05, + "loss": 0.0726, + "step": 9952 + }, + { + "epoch": 1.0790329575021682, + "grad_norm": 3.309927225112915, + "learning_rate": 1.2852894240533876e-05, + "loss": 0.0112, + "step": 9953 + }, + { + "epoch": 1.079141370338248, + "grad_norm": 0.19643695652484894, + "learning_rate": 1.2852168866966488e-05, + "loss": 0.0009, + "step": 9954 + }, + { + "epoch": 1.0792497831743277, + "grad_norm": 3.4190773963928223, + "learning_rate": 1.28514434933991e-05, + "loss": 0.0066, + "step": 9955 + }, + { + "epoch": 1.0793581960104077, + "grad_norm": 3.834362268447876, + "learning_rate": 1.2850718119831716e-05, + "loss": 0.0243, + "step": 9956 + }, + { + "epoch": 1.0794666088464875, + "grad_norm": 3.73468279838562, + "learning_rate": 1.2849992746264328e-05, + "loss": 0.0737, + "step": 9957 + }, + { + "epoch": 1.0795750216825672, + "grad_norm": 0.8603652119636536, + "learning_rate": 1.284926737269694e-05, + "loss": 0.0066, + "step": 9958 + }, + { + "epoch": 1.079683434518647, + "grad_norm": 3.342792510986328, + "learning_rate": 1.2848541999129554e-05, + "loss": 0.0389, + "step": 9959 + }, + { + "epoch": 1.0797918473547268, + "grad_norm": 0.5023723244667053, + "learning_rate": 1.2847816625562166e-05, + "loss": 0.0028, + "step": 9960 + }, + { + "epoch": 1.0799002601908065, + "grad_norm": 0.31955185532569885, + "learning_rate": 1.2847091251994778e-05, + "loss": 0.0018, + "step": 9961 + }, + { + "epoch": 1.0800086730268863, + "grad_norm": 5.123091220855713, + "learning_rate": 1.2846365878427392e-05, + "loss": 0.0308, + "step": 9962 + }, + { + "epoch": 1.080117085862966, + "grad_norm": 0.4991840124130249, + "learning_rate": 1.2845640504860004e-05, + "loss": 0.0064, + "step": 9963 + }, + { + "epoch": 1.080225498699046, + "grad_norm": 2.182586431503296, + "learning_rate": 1.2844915131292616e-05, + "loss": 0.0133, + "step": 9964 + }, + { + "epoch": 1.0803339115351258, + "grad_norm": 2.5860605239868164, + "learning_rate": 1.284418975772523e-05, + "loss": 0.032, + "step": 9965 + }, + { + "epoch": 1.0804423243712056, + "grad_norm": 0.3267560601234436, + "learning_rate": 1.2843464384157842e-05, + "loss": 0.0044, + "step": 9966 + }, + { + "epoch": 1.0805507372072853, + "grad_norm": 1.6903090476989746, + "learning_rate": 1.2842739010590455e-05, + "loss": 0.0098, + "step": 9967 + }, + { + "epoch": 1.080659150043365, + "grad_norm": 2.4064152240753174, + "learning_rate": 1.2842013637023067e-05, + "loss": 0.0373, + "step": 9968 + }, + { + "epoch": 1.0807675628794449, + "grad_norm": 6.24027156829834, + "learning_rate": 1.284128826345568e-05, + "loss": 0.0405, + "step": 9969 + }, + { + "epoch": 1.0808759757155246, + "grad_norm": 4.410862445831299, + "learning_rate": 1.2840562889888293e-05, + "loss": 0.0446, + "step": 9970 + }, + { + "epoch": 1.0809843885516046, + "grad_norm": 2.168830633163452, + "learning_rate": 1.2839837516320905e-05, + "loss": 0.0259, + "step": 9971 + }, + { + "epoch": 1.0810928013876844, + "grad_norm": 4.746983051300049, + "learning_rate": 1.2839112142753519e-05, + "loss": 0.0229, + "step": 9972 + }, + { + "epoch": 1.0812012142237641, + "grad_norm": 3.9711172580718994, + "learning_rate": 1.283838676918613e-05, + "loss": 0.0262, + "step": 9973 + }, + { + "epoch": 1.081309627059844, + "grad_norm": 1.1091651916503906, + "learning_rate": 1.2837661395618746e-05, + "loss": 0.0128, + "step": 9974 + }, + { + "epoch": 1.0814180398959237, + "grad_norm": 1.937779188156128, + "learning_rate": 1.2836936022051358e-05, + "loss": 0.0561, + "step": 9975 + }, + { + "epoch": 1.0815264527320034, + "grad_norm": 0.7768162488937378, + "learning_rate": 1.283621064848397e-05, + "loss": 0.0119, + "step": 9976 + }, + { + "epoch": 1.0816348655680832, + "grad_norm": 1.3132396936416626, + "learning_rate": 1.2835485274916584e-05, + "loss": 0.0286, + "step": 9977 + }, + { + "epoch": 1.081743278404163, + "grad_norm": 1.788223147392273, + "learning_rate": 1.2834759901349197e-05, + "loss": 0.0183, + "step": 9978 + }, + { + "epoch": 1.081851691240243, + "grad_norm": 0.39442309737205505, + "learning_rate": 1.2834034527781809e-05, + "loss": 0.0044, + "step": 9979 + }, + { + "epoch": 1.0819601040763227, + "grad_norm": 2.3557844161987305, + "learning_rate": 1.2833309154214421e-05, + "loss": 0.0081, + "step": 9980 + }, + { + "epoch": 1.0820685169124025, + "grad_norm": 2.471142053604126, + "learning_rate": 1.2832583780647035e-05, + "loss": 0.0322, + "step": 9981 + }, + { + "epoch": 1.0821769297484822, + "grad_norm": 2.451589584350586, + "learning_rate": 1.2831858407079647e-05, + "loss": 0.0212, + "step": 9982 + }, + { + "epoch": 1.082285342584562, + "grad_norm": 0.4645666778087616, + "learning_rate": 1.2831133033512259e-05, + "loss": 0.0034, + "step": 9983 + }, + { + "epoch": 1.0823937554206418, + "grad_norm": 2.1195383071899414, + "learning_rate": 1.2830407659944873e-05, + "loss": 0.0556, + "step": 9984 + }, + { + "epoch": 1.0825021682567215, + "grad_norm": 2.463881015777588, + "learning_rate": 1.2829682286377485e-05, + "loss": 0.0225, + "step": 9985 + }, + { + "epoch": 1.0826105810928013, + "grad_norm": 4.0231733322143555, + "learning_rate": 1.2828956912810097e-05, + "loss": 0.0231, + "step": 9986 + }, + { + "epoch": 1.0827189939288813, + "grad_norm": 0.36293143033981323, + "learning_rate": 1.2828231539242711e-05, + "loss": 0.0046, + "step": 9987 + }, + { + "epoch": 1.082827406764961, + "grad_norm": 1.5441174507141113, + "learning_rate": 1.2827506165675323e-05, + "loss": 0.0064, + "step": 9988 + }, + { + "epoch": 1.0829358196010408, + "grad_norm": 3.4634995460510254, + "learning_rate": 1.2826780792107935e-05, + "loss": 0.0419, + "step": 9989 + }, + { + "epoch": 1.0830442324371206, + "grad_norm": 0.9702847599983215, + "learning_rate": 1.2826055418540549e-05, + "loss": 0.003, + "step": 9990 + }, + { + "epoch": 1.0831526452732003, + "grad_norm": 4.063960075378418, + "learning_rate": 1.2825330044973163e-05, + "loss": 0.0075, + "step": 9991 + }, + { + "epoch": 1.08326105810928, + "grad_norm": 3.582078218460083, + "learning_rate": 1.2824604671405777e-05, + "loss": 0.0144, + "step": 9992 + }, + { + "epoch": 1.0833694709453598, + "grad_norm": 2.1913204193115234, + "learning_rate": 1.2823879297838389e-05, + "loss": 0.0188, + "step": 9993 + }, + { + "epoch": 1.0834778837814398, + "grad_norm": 1.7198697328567505, + "learning_rate": 1.2823153924271001e-05, + "loss": 0.0248, + "step": 9994 + }, + { + "epoch": 1.0835862966175196, + "grad_norm": 5.310633659362793, + "learning_rate": 1.2822428550703613e-05, + "loss": 0.0414, + "step": 9995 + }, + { + "epoch": 1.0836947094535994, + "grad_norm": 4.603381156921387, + "learning_rate": 1.2821703177136227e-05, + "loss": 0.0226, + "step": 9996 + }, + { + "epoch": 1.0838031222896791, + "grad_norm": 5.023976802825928, + "learning_rate": 1.282097780356884e-05, + "loss": 0.0637, + "step": 9997 + }, + { + "epoch": 1.0839115351257589, + "grad_norm": 2.2861337661743164, + "learning_rate": 1.2820252430001451e-05, + "loss": 0.0712, + "step": 9998 + }, + { + "epoch": 1.0840199479618386, + "grad_norm": 0.06889315694570541, + "learning_rate": 1.2819527056434065e-05, + "loss": 0.0006, + "step": 9999 + }, + { + "epoch": 1.0841283607979184, + "grad_norm": 6.562941551208496, + "learning_rate": 1.2818801682866677e-05, + "loss": 0.0211, + "step": 10000 + }, + { + "epoch": 1.0842367736339982, + "grad_norm": 2.8691482543945312, + "learning_rate": 1.281807630929929e-05, + "loss": 0.0248, + "step": 10001 + }, + { + "epoch": 1.0843451864700782, + "grad_norm": 0.8408215641975403, + "learning_rate": 1.2817350935731903e-05, + "loss": 0.0053, + "step": 10002 + }, + { + "epoch": 1.084453599306158, + "grad_norm": 0.6695507764816284, + "learning_rate": 1.2816625562164515e-05, + "loss": 0.0066, + "step": 10003 + }, + { + "epoch": 1.0845620121422377, + "grad_norm": 1.591231346130371, + "learning_rate": 1.2815900188597128e-05, + "loss": 0.0157, + "step": 10004 + }, + { + "epoch": 1.0846704249783174, + "grad_norm": 0.6067041158676147, + "learning_rate": 1.2815174815029741e-05, + "loss": 0.0064, + "step": 10005 + }, + { + "epoch": 1.0847788378143972, + "grad_norm": 2.0425431728363037, + "learning_rate": 1.2814449441462354e-05, + "loss": 0.014, + "step": 10006 + }, + { + "epoch": 1.084887250650477, + "grad_norm": 0.2978549301624298, + "learning_rate": 1.2813724067894966e-05, + "loss": 0.0022, + "step": 10007 + }, + { + "epoch": 1.0849956634865567, + "grad_norm": 13.835149765014648, + "learning_rate": 1.2812998694327581e-05, + "loss": 0.0287, + "step": 10008 + }, + { + "epoch": 1.0851040763226365, + "grad_norm": 2.0309805870056152, + "learning_rate": 1.2812273320760193e-05, + "loss": 0.0125, + "step": 10009 + }, + { + "epoch": 1.0852124891587165, + "grad_norm": 2.789281129837036, + "learning_rate": 1.2811547947192805e-05, + "loss": 0.0135, + "step": 10010 + }, + { + "epoch": 1.0853209019947962, + "grad_norm": 2.6090965270996094, + "learning_rate": 1.281082257362542e-05, + "loss": 0.0455, + "step": 10011 + }, + { + "epoch": 1.085429314830876, + "grad_norm": 1.6601552963256836, + "learning_rate": 1.2810097200058031e-05, + "loss": 0.0188, + "step": 10012 + }, + { + "epoch": 1.0855377276669558, + "grad_norm": 4.3228912353515625, + "learning_rate": 1.2809371826490644e-05, + "loss": 0.071, + "step": 10013 + }, + { + "epoch": 1.0856461405030355, + "grad_norm": 1.0037578344345093, + "learning_rate": 1.2808646452923257e-05, + "loss": 0.0046, + "step": 10014 + }, + { + "epoch": 1.0857545533391153, + "grad_norm": 4.528359413146973, + "learning_rate": 1.280792107935587e-05, + "loss": 0.0335, + "step": 10015 + }, + { + "epoch": 1.085862966175195, + "grad_norm": 0.23017823696136475, + "learning_rate": 1.2807195705788482e-05, + "loss": 0.0014, + "step": 10016 + }, + { + "epoch": 1.085971379011275, + "grad_norm": 3.2162952423095703, + "learning_rate": 1.2806470332221096e-05, + "loss": 0.0255, + "step": 10017 + }, + { + "epoch": 1.0860797918473548, + "grad_norm": 0.6214327812194824, + "learning_rate": 1.2805744958653708e-05, + "loss": 0.0012, + "step": 10018 + }, + { + "epoch": 1.0861882046834346, + "grad_norm": 0.1524006426334381, + "learning_rate": 1.280501958508632e-05, + "loss": 0.0019, + "step": 10019 + }, + { + "epoch": 1.0862966175195143, + "grad_norm": 0.5612609386444092, + "learning_rate": 1.2804294211518932e-05, + "loss": 0.0022, + "step": 10020 + }, + { + "epoch": 1.086405030355594, + "grad_norm": 1.0037575960159302, + "learning_rate": 1.2803568837951546e-05, + "loss": 0.0176, + "step": 10021 + }, + { + "epoch": 1.0865134431916739, + "grad_norm": 2.4832441806793213, + "learning_rate": 1.2802843464384158e-05, + "loss": 0.0328, + "step": 10022 + }, + { + "epoch": 1.0866218560277536, + "grad_norm": 0.06443511694669724, + "learning_rate": 1.280211809081677e-05, + "loss": 0.0007, + "step": 10023 + }, + { + "epoch": 1.0867302688638334, + "grad_norm": 1.5524108409881592, + "learning_rate": 1.2801392717249384e-05, + "loss": 0.0099, + "step": 10024 + }, + { + "epoch": 1.0868386816999132, + "grad_norm": 2.63883376121521, + "learning_rate": 1.2800667343681998e-05, + "loss": 0.0441, + "step": 10025 + }, + { + "epoch": 1.0869470945359931, + "grad_norm": 0.15037205815315247, + "learning_rate": 1.2799941970114612e-05, + "loss": 0.0019, + "step": 10026 + }, + { + "epoch": 1.087055507372073, + "grad_norm": 3.539262056350708, + "learning_rate": 1.2799216596547224e-05, + "loss": 0.0142, + "step": 10027 + }, + { + "epoch": 1.0871639202081527, + "grad_norm": 3.7019033432006836, + "learning_rate": 1.2798491222979836e-05, + "loss": 0.0343, + "step": 10028 + }, + { + "epoch": 1.0872723330442324, + "grad_norm": 1.2455297708511353, + "learning_rate": 1.279776584941245e-05, + "loss": 0.0297, + "step": 10029 + }, + { + "epoch": 1.0873807458803122, + "grad_norm": 0.9419097900390625, + "learning_rate": 1.2797040475845062e-05, + "loss": 0.0039, + "step": 10030 + }, + { + "epoch": 1.087489158716392, + "grad_norm": 2.1243364810943604, + "learning_rate": 1.2796315102277674e-05, + "loss": 0.0127, + "step": 10031 + }, + { + "epoch": 1.0875975715524717, + "grad_norm": 0.977845311164856, + "learning_rate": 1.2795589728710286e-05, + "loss": 0.0224, + "step": 10032 + }, + { + "epoch": 1.0877059843885517, + "grad_norm": 1.4567339420318604, + "learning_rate": 1.27948643551429e-05, + "loss": 0.0292, + "step": 10033 + }, + { + "epoch": 1.0878143972246315, + "grad_norm": 0.34003886580467224, + "learning_rate": 1.2794138981575512e-05, + "loss": 0.003, + "step": 10034 + }, + { + "epoch": 1.0879228100607112, + "grad_norm": 2.769033432006836, + "learning_rate": 1.2793413608008124e-05, + "loss": 0.0197, + "step": 10035 + }, + { + "epoch": 1.088031222896791, + "grad_norm": 0.16021665930747986, + "learning_rate": 1.2792688234440738e-05, + "loss": 0.0013, + "step": 10036 + }, + { + "epoch": 1.0881396357328708, + "grad_norm": 1.773360013961792, + "learning_rate": 1.279196286087335e-05, + "loss": 0.0391, + "step": 10037 + }, + { + "epoch": 1.0882480485689505, + "grad_norm": 3.7176268100738525, + "learning_rate": 1.2791237487305962e-05, + "loss": 0.045, + "step": 10038 + }, + { + "epoch": 1.0883564614050303, + "grad_norm": 3.549219846725464, + "learning_rate": 1.2790512113738576e-05, + "loss": 0.0467, + "step": 10039 + }, + { + "epoch": 1.08846487424111, + "grad_norm": 2.7902276515960693, + "learning_rate": 1.2789786740171188e-05, + "loss": 0.0123, + "step": 10040 + }, + { + "epoch": 1.08857328707719, + "grad_norm": 1.6138418912887573, + "learning_rate": 1.27890613666038e-05, + "loss": 0.019, + "step": 10041 + }, + { + "epoch": 1.0886816999132698, + "grad_norm": 1.812343716621399, + "learning_rate": 1.2788335993036414e-05, + "loss": 0.0091, + "step": 10042 + }, + { + "epoch": 1.0887901127493496, + "grad_norm": 4.244384765625, + "learning_rate": 1.2787610619469028e-05, + "loss": 0.0157, + "step": 10043 + }, + { + "epoch": 1.0888985255854293, + "grad_norm": 1.7866770029067993, + "learning_rate": 1.2786885245901642e-05, + "loss": 0.0325, + "step": 10044 + }, + { + "epoch": 1.089006938421509, + "grad_norm": 0.592184841632843, + "learning_rate": 1.2786159872334254e-05, + "loss": 0.0047, + "step": 10045 + }, + { + "epoch": 1.0891153512575888, + "grad_norm": 3.4189605712890625, + "learning_rate": 1.2785434498766866e-05, + "loss": 0.0191, + "step": 10046 + }, + { + "epoch": 1.0892237640936686, + "grad_norm": 6.920853614807129, + "learning_rate": 1.2784709125199479e-05, + "loss": 0.0401, + "step": 10047 + }, + { + "epoch": 1.0893321769297484, + "grad_norm": 5.314445495605469, + "learning_rate": 1.2783983751632092e-05, + "loss": 0.0098, + "step": 10048 + }, + { + "epoch": 1.0894405897658284, + "grad_norm": 1.9401413202285767, + "learning_rate": 1.2783258378064704e-05, + "loss": 0.0105, + "step": 10049 + }, + { + "epoch": 1.0895490026019081, + "grad_norm": 2.247743844985962, + "learning_rate": 1.2782533004497317e-05, + "loss": 0.0514, + "step": 10050 + }, + { + "epoch": 1.0896574154379879, + "grad_norm": 4.021987438201904, + "learning_rate": 1.278180763092993e-05, + "loss": 0.0317, + "step": 10051 + }, + { + "epoch": 1.0897658282740676, + "grad_norm": 2.0238542556762695, + "learning_rate": 1.2781082257362543e-05, + "loss": 0.0095, + "step": 10052 + }, + { + "epoch": 1.0898742411101474, + "grad_norm": 2.805187702178955, + "learning_rate": 1.2780356883795155e-05, + "loss": 0.0471, + "step": 10053 + }, + { + "epoch": 1.0899826539462272, + "grad_norm": 1.3336372375488281, + "learning_rate": 1.2779631510227769e-05, + "loss": 0.009, + "step": 10054 + }, + { + "epoch": 1.090091066782307, + "grad_norm": 1.1732560396194458, + "learning_rate": 1.277890613666038e-05, + "loss": 0.0095, + "step": 10055 + }, + { + "epoch": 1.090199479618387, + "grad_norm": 3.974730968475342, + "learning_rate": 1.2778180763092993e-05, + "loss": 0.0247, + "step": 10056 + }, + { + "epoch": 1.0903078924544667, + "grad_norm": 3.5329337120056152, + "learning_rate": 1.2777455389525607e-05, + "loss": 0.03, + "step": 10057 + }, + { + "epoch": 1.0904163052905465, + "grad_norm": 3.0943503379821777, + "learning_rate": 1.2776730015958219e-05, + "loss": 0.028, + "step": 10058 + }, + { + "epoch": 1.0905247181266262, + "grad_norm": 0.48203688859939575, + "learning_rate": 1.2776004642390831e-05, + "loss": 0.0019, + "step": 10059 + }, + { + "epoch": 1.090633130962706, + "grad_norm": 4.702168941497803, + "learning_rate": 1.2775279268823447e-05, + "loss": 0.0469, + "step": 10060 + }, + { + "epoch": 1.0907415437987857, + "grad_norm": 4.015799522399902, + "learning_rate": 1.2774553895256059e-05, + "loss": 0.01, + "step": 10061 + }, + { + "epoch": 1.0908499566348655, + "grad_norm": 3.456878662109375, + "learning_rate": 1.277382852168867e-05, + "loss": 0.0281, + "step": 10062 + }, + { + "epoch": 1.0909583694709453, + "grad_norm": 0.9316082000732422, + "learning_rate": 1.2773103148121285e-05, + "loss": 0.0048, + "step": 10063 + }, + { + "epoch": 1.0910667823070253, + "grad_norm": 0.9011953473091125, + "learning_rate": 1.2772377774553897e-05, + "loss": 0.0047, + "step": 10064 + }, + { + "epoch": 1.091175195143105, + "grad_norm": 1.81675124168396, + "learning_rate": 1.2771652400986509e-05, + "loss": 0.0115, + "step": 10065 + }, + { + "epoch": 1.0912836079791848, + "grad_norm": 0.05056634917855263, + "learning_rate": 1.2770927027419123e-05, + "loss": 0.0004, + "step": 10066 + }, + { + "epoch": 1.0913920208152645, + "grad_norm": 4.403650760650635, + "learning_rate": 1.2770201653851735e-05, + "loss": 0.0627, + "step": 10067 + }, + { + "epoch": 1.0915004336513443, + "grad_norm": 0.7925182580947876, + "learning_rate": 1.2769476280284347e-05, + "loss": 0.007, + "step": 10068 + }, + { + "epoch": 1.091608846487424, + "grad_norm": 4.221546649932861, + "learning_rate": 1.2768750906716961e-05, + "loss": 0.0466, + "step": 10069 + }, + { + "epoch": 1.0917172593235038, + "grad_norm": 1.4511384963989258, + "learning_rate": 1.2768025533149573e-05, + "loss": 0.0143, + "step": 10070 + }, + { + "epoch": 1.0918256721595836, + "grad_norm": 3.0821800231933594, + "learning_rate": 1.2767300159582185e-05, + "loss": 0.0444, + "step": 10071 + }, + { + "epoch": 1.0919340849956636, + "grad_norm": 0.34466612339019775, + "learning_rate": 1.2766574786014797e-05, + "loss": 0.0035, + "step": 10072 + }, + { + "epoch": 1.0920424978317433, + "grad_norm": 4.54889440536499, + "learning_rate": 1.2765849412447411e-05, + "loss": 0.0363, + "step": 10073 + }, + { + "epoch": 1.092150910667823, + "grad_norm": 0.4006764590740204, + "learning_rate": 1.2765124038880023e-05, + "loss": 0.0021, + "step": 10074 + }, + { + "epoch": 1.0922593235039029, + "grad_norm": 4.101487636566162, + "learning_rate": 1.2764398665312635e-05, + "loss": 0.032, + "step": 10075 + }, + { + "epoch": 1.0923677363399826, + "grad_norm": 4.493408203125, + "learning_rate": 1.276367329174525e-05, + "loss": 0.0654, + "step": 10076 + }, + { + "epoch": 1.0924761491760624, + "grad_norm": 2.176694869995117, + "learning_rate": 1.2762947918177863e-05, + "loss": 0.0157, + "step": 10077 + }, + { + "epoch": 1.0925845620121422, + "grad_norm": 0.3701590895652771, + "learning_rate": 1.2762222544610477e-05, + "loss": 0.0028, + "step": 10078 + }, + { + "epoch": 1.0926929748482221, + "grad_norm": 3.1437463760375977, + "learning_rate": 1.2761497171043089e-05, + "loss": 0.0126, + "step": 10079 + }, + { + "epoch": 1.092801387684302, + "grad_norm": 4.643587589263916, + "learning_rate": 1.2760771797475701e-05, + "loss": 0.0118, + "step": 10080 + }, + { + "epoch": 1.0929098005203817, + "grad_norm": 4.278164386749268, + "learning_rate": 1.2760046423908315e-05, + "loss": 0.0449, + "step": 10081 + }, + { + "epoch": 1.0930182133564614, + "grad_norm": 0.24420060217380524, + "learning_rate": 1.2759321050340927e-05, + "loss": 0.0016, + "step": 10082 + }, + { + "epoch": 1.0931266261925412, + "grad_norm": 0.19508413970470428, + "learning_rate": 1.275859567677354e-05, + "loss": 0.0017, + "step": 10083 + }, + { + "epoch": 1.093235039028621, + "grad_norm": 2.9350788593292236, + "learning_rate": 1.2757870303206153e-05, + "loss": 0.0219, + "step": 10084 + }, + { + "epoch": 1.0933434518647007, + "grad_norm": 2.4302258491516113, + "learning_rate": 1.2757144929638765e-05, + "loss": 0.0206, + "step": 10085 + }, + { + "epoch": 1.0934518647007805, + "grad_norm": 1.8715157508850098, + "learning_rate": 1.2756419556071378e-05, + "loss": 0.0083, + "step": 10086 + }, + { + "epoch": 1.0935602775368602, + "grad_norm": 5.9356794357299805, + "learning_rate": 1.275569418250399e-05, + "loss": 0.08, + "step": 10087 + }, + { + "epoch": 1.0936686903729402, + "grad_norm": 3.6474967002868652, + "learning_rate": 1.2754968808936603e-05, + "loss": 0.0171, + "step": 10088 + }, + { + "epoch": 1.09377710320902, + "grad_norm": 3.422241687774658, + "learning_rate": 1.2754243435369216e-05, + "loss": 0.0123, + "step": 10089 + }, + { + "epoch": 1.0938855160450998, + "grad_norm": 0.7618327736854553, + "learning_rate": 1.2753518061801828e-05, + "loss": 0.0054, + "step": 10090 + }, + { + "epoch": 1.0939939288811795, + "grad_norm": 1.0851718187332153, + "learning_rate": 1.2752792688234442e-05, + "loss": 0.0032, + "step": 10091 + }, + { + "epoch": 1.0941023417172593, + "grad_norm": 1.6523823738098145, + "learning_rate": 1.2752067314667054e-05, + "loss": 0.0115, + "step": 10092 + }, + { + "epoch": 1.094210754553339, + "grad_norm": 2.640871047973633, + "learning_rate": 1.2751341941099666e-05, + "loss": 0.0175, + "step": 10093 + }, + { + "epoch": 1.0943191673894188, + "grad_norm": 3.9240143299102783, + "learning_rate": 1.2750616567532281e-05, + "loss": 0.0506, + "step": 10094 + }, + { + "epoch": 1.0944275802254988, + "grad_norm": 5.433993339538574, + "learning_rate": 1.2749891193964894e-05, + "loss": 0.0266, + "step": 10095 + }, + { + "epoch": 1.0945359930615786, + "grad_norm": 1.0542099475860596, + "learning_rate": 1.2749165820397507e-05, + "loss": 0.0115, + "step": 10096 + }, + { + "epoch": 1.0946444058976583, + "grad_norm": 0.9387018084526062, + "learning_rate": 1.274844044683012e-05, + "loss": 0.0156, + "step": 10097 + }, + { + "epoch": 1.094752818733738, + "grad_norm": 2.957519292831421, + "learning_rate": 1.2747715073262732e-05, + "loss": 0.0224, + "step": 10098 + }, + { + "epoch": 1.0948612315698178, + "grad_norm": 0.4051058292388916, + "learning_rate": 1.2746989699695344e-05, + "loss": 0.0029, + "step": 10099 + }, + { + "epoch": 1.0949696444058976, + "grad_norm": 2.68487548828125, + "learning_rate": 1.2746264326127958e-05, + "loss": 0.0482, + "step": 10100 + }, + { + "epoch": 1.0950780572419774, + "grad_norm": 1.7548797130584717, + "learning_rate": 1.274553895256057e-05, + "loss": 0.0096, + "step": 10101 + }, + { + "epoch": 1.0951864700780571, + "grad_norm": 6.856265068054199, + "learning_rate": 1.2744813578993182e-05, + "loss": 0.0539, + "step": 10102 + }, + { + "epoch": 1.0952948829141371, + "grad_norm": 2.0372352600097656, + "learning_rate": 1.2744088205425796e-05, + "loss": 0.0037, + "step": 10103 + }, + { + "epoch": 1.0954032957502169, + "grad_norm": 1.7782000303268433, + "learning_rate": 1.2743362831858408e-05, + "loss": 0.0022, + "step": 10104 + }, + { + "epoch": 1.0955117085862967, + "grad_norm": 3.6572659015655518, + "learning_rate": 1.274263745829102e-05, + "loss": 0.0312, + "step": 10105 + }, + { + "epoch": 1.0956201214223764, + "grad_norm": 2.3388750553131104, + "learning_rate": 1.2741912084723634e-05, + "loss": 0.0302, + "step": 10106 + }, + { + "epoch": 1.0957285342584562, + "grad_norm": 0.29054558277130127, + "learning_rate": 1.2741186711156246e-05, + "loss": 0.0018, + "step": 10107 + }, + { + "epoch": 1.095836947094536, + "grad_norm": 1.1603727340698242, + "learning_rate": 1.2740461337588858e-05, + "loss": 0.0193, + "step": 10108 + }, + { + "epoch": 1.0959453599306157, + "grad_norm": 1.2900668382644653, + "learning_rate": 1.2739735964021472e-05, + "loss": 0.0103, + "step": 10109 + }, + { + "epoch": 1.0960537727666955, + "grad_norm": 3.548489570617676, + "learning_rate": 1.2739010590454084e-05, + "loss": 0.0581, + "step": 10110 + }, + { + "epoch": 1.0961621856027755, + "grad_norm": 10.002979278564453, + "learning_rate": 1.2738285216886696e-05, + "loss": 0.0688, + "step": 10111 + }, + { + "epoch": 1.0962705984388552, + "grad_norm": 3.708585023880005, + "learning_rate": 1.2737559843319312e-05, + "loss": 0.0533, + "step": 10112 + }, + { + "epoch": 1.096379011274935, + "grad_norm": 8.24851131439209, + "learning_rate": 1.2736834469751924e-05, + "loss": 0.0442, + "step": 10113 + }, + { + "epoch": 1.0964874241110147, + "grad_norm": 4.03413200378418, + "learning_rate": 1.2736109096184536e-05, + "loss": 0.0242, + "step": 10114 + }, + { + "epoch": 1.0965958369470945, + "grad_norm": 2.077897548675537, + "learning_rate": 1.273538372261715e-05, + "loss": 0.0091, + "step": 10115 + }, + { + "epoch": 1.0967042497831743, + "grad_norm": 1.625030755996704, + "learning_rate": 1.2734658349049762e-05, + "loss": 0.006, + "step": 10116 + }, + { + "epoch": 1.096812662619254, + "grad_norm": 3.3242509365081787, + "learning_rate": 1.2733932975482374e-05, + "loss": 0.0768, + "step": 10117 + }, + { + "epoch": 1.096921075455334, + "grad_norm": 1.6318027973175049, + "learning_rate": 1.2733207601914988e-05, + "loss": 0.0217, + "step": 10118 + }, + { + "epoch": 1.0970294882914138, + "grad_norm": 3.42779803276062, + "learning_rate": 1.27324822283476e-05, + "loss": 0.0324, + "step": 10119 + }, + { + "epoch": 1.0971379011274935, + "grad_norm": 3.127987861633301, + "learning_rate": 1.2731756854780212e-05, + "loss": 0.008, + "step": 10120 + }, + { + "epoch": 1.0972463139635733, + "grad_norm": 0.913298487663269, + "learning_rate": 1.2731031481212826e-05, + "loss": 0.0052, + "step": 10121 + }, + { + "epoch": 1.097354726799653, + "grad_norm": 0.33424559235572815, + "learning_rate": 1.2730306107645438e-05, + "loss": 0.0017, + "step": 10122 + }, + { + "epoch": 1.0974631396357328, + "grad_norm": 6.3980937004089355, + "learning_rate": 1.272958073407805e-05, + "loss": 0.0358, + "step": 10123 + }, + { + "epoch": 1.0975715524718126, + "grad_norm": 2.2223331928253174, + "learning_rate": 1.2728855360510663e-05, + "loss": 0.0375, + "step": 10124 + }, + { + "epoch": 1.0976799653078924, + "grad_norm": 0.12028948962688446, + "learning_rate": 1.2728129986943277e-05, + "loss": 0.0012, + "step": 10125 + }, + { + "epoch": 1.0977883781439723, + "grad_norm": 3.323575258255005, + "learning_rate": 1.2727404613375889e-05, + "loss": 0.0532, + "step": 10126 + }, + { + "epoch": 1.097896790980052, + "grad_norm": 0.11264148354530334, + "learning_rate": 1.27266792398085e-05, + "loss": 0.0009, + "step": 10127 + }, + { + "epoch": 1.0980052038161319, + "grad_norm": 1.1254230737686157, + "learning_rate": 1.2725953866241115e-05, + "loss": 0.0075, + "step": 10128 + }, + { + "epoch": 1.0981136166522116, + "grad_norm": 3.9813594818115234, + "learning_rate": 1.2725228492673728e-05, + "loss": 0.028, + "step": 10129 + }, + { + "epoch": 1.0982220294882914, + "grad_norm": 1.5732179880142212, + "learning_rate": 1.2724503119106342e-05, + "loss": 0.0548, + "step": 10130 + }, + { + "epoch": 1.0983304423243712, + "grad_norm": 3.3860116004943848, + "learning_rate": 1.2723777745538954e-05, + "loss": 0.0597, + "step": 10131 + }, + { + "epoch": 1.098438855160451, + "grad_norm": 1.775453805923462, + "learning_rate": 1.2723052371971567e-05, + "loss": 0.0115, + "step": 10132 + }, + { + "epoch": 1.0985472679965307, + "grad_norm": 4.194432258605957, + "learning_rate": 1.272232699840418e-05, + "loss": 0.0604, + "step": 10133 + }, + { + "epoch": 1.0986556808326107, + "grad_norm": 13.551997184753418, + "learning_rate": 1.2721601624836793e-05, + "loss": 0.0145, + "step": 10134 + }, + { + "epoch": 1.0987640936686904, + "grad_norm": 10.609891891479492, + "learning_rate": 1.2720876251269405e-05, + "loss": 0.0561, + "step": 10135 + }, + { + "epoch": 1.0988725065047702, + "grad_norm": 0.04591526463627815, + "learning_rate": 1.2720150877702019e-05, + "loss": 0.0005, + "step": 10136 + }, + { + "epoch": 1.09898091934085, + "grad_norm": 2.1228716373443604, + "learning_rate": 1.271942550413463e-05, + "loss": 0.0249, + "step": 10137 + }, + { + "epoch": 1.0990893321769297, + "grad_norm": 4.254547595977783, + "learning_rate": 1.2718700130567243e-05, + "loss": 0.0396, + "step": 10138 + }, + { + "epoch": 1.0991977450130095, + "grad_norm": 1.9328454732894897, + "learning_rate": 1.2717974756999855e-05, + "loss": 0.0338, + "step": 10139 + }, + { + "epoch": 1.0993061578490892, + "grad_norm": 1.987633228302002, + "learning_rate": 1.2717249383432469e-05, + "loss": 0.0123, + "step": 10140 + }, + { + "epoch": 1.0994145706851692, + "grad_norm": 0.33529314398765564, + "learning_rate": 1.2716524009865081e-05, + "loss": 0.004, + "step": 10141 + }, + { + "epoch": 1.099522983521249, + "grad_norm": 1.283994436264038, + "learning_rate": 1.2715798636297693e-05, + "loss": 0.0183, + "step": 10142 + }, + { + "epoch": 1.0996313963573288, + "grad_norm": 5.0968194007873535, + "learning_rate": 1.2715073262730307e-05, + "loss": 0.0246, + "step": 10143 + }, + { + "epoch": 1.0997398091934085, + "grad_norm": 2.2846460342407227, + "learning_rate": 1.2714347889162919e-05, + "loss": 0.0178, + "step": 10144 + }, + { + "epoch": 1.0998482220294883, + "grad_norm": 4.236953258514404, + "learning_rate": 1.2713622515595531e-05, + "loss": 0.0311, + "step": 10145 + }, + { + "epoch": 1.099956634865568, + "grad_norm": 5.081821918487549, + "learning_rate": 1.2712897142028147e-05, + "loss": 0.0083, + "step": 10146 + }, + { + "epoch": 1.1000650477016478, + "grad_norm": 1.0751383304595947, + "learning_rate": 1.2712171768460759e-05, + "loss": 0.0087, + "step": 10147 + }, + { + "epoch": 1.1001734605377276, + "grad_norm": 0.7005839943885803, + "learning_rate": 1.2711446394893373e-05, + "loss": 0.0078, + "step": 10148 + }, + { + "epoch": 1.1002818733738076, + "grad_norm": 3.209693193435669, + "learning_rate": 1.2710721021325985e-05, + "loss": 0.0296, + "step": 10149 + }, + { + "epoch": 1.1003902862098873, + "grad_norm": 1.8439486026763916, + "learning_rate": 1.2709995647758597e-05, + "loss": 0.0257, + "step": 10150 + }, + { + "epoch": 1.100498699045967, + "grad_norm": 4.317352771759033, + "learning_rate": 1.270927027419121e-05, + "loss": 0.0367, + "step": 10151 + }, + { + "epoch": 1.1006071118820469, + "grad_norm": 3.228891134262085, + "learning_rate": 1.2708544900623823e-05, + "loss": 0.0193, + "step": 10152 + }, + { + "epoch": 1.1007155247181266, + "grad_norm": 4.914162635803223, + "learning_rate": 1.2707819527056435e-05, + "loss": 0.0327, + "step": 10153 + }, + { + "epoch": 1.1008239375542064, + "grad_norm": 0.16366201639175415, + "learning_rate": 1.2707094153489047e-05, + "loss": 0.0021, + "step": 10154 + }, + { + "epoch": 1.1009323503902861, + "grad_norm": 1.5768712759017944, + "learning_rate": 1.2706368779921661e-05, + "loss": 0.0223, + "step": 10155 + }, + { + "epoch": 1.101040763226366, + "grad_norm": 0.3966556489467621, + "learning_rate": 1.2705643406354273e-05, + "loss": 0.006, + "step": 10156 + }, + { + "epoch": 1.101149176062446, + "grad_norm": 1.2316192388534546, + "learning_rate": 1.2704918032786885e-05, + "loss": 0.0088, + "step": 10157 + }, + { + "epoch": 1.1012575888985257, + "grad_norm": 4.447566032409668, + "learning_rate": 1.27041926592195e-05, + "loss": 0.0217, + "step": 10158 + }, + { + "epoch": 1.1013660017346054, + "grad_norm": 2.408970594406128, + "learning_rate": 1.2703467285652111e-05, + "loss": 0.0067, + "step": 10159 + }, + { + "epoch": 1.1014744145706852, + "grad_norm": 0.20786887407302856, + "learning_rate": 1.2702741912084724e-05, + "loss": 0.0023, + "step": 10160 + }, + { + "epoch": 1.101582827406765, + "grad_norm": 6.046696186065674, + "learning_rate": 1.2702016538517337e-05, + "loss": 0.0368, + "step": 10161 + }, + { + "epoch": 1.1016912402428447, + "grad_norm": 1.3861736059188843, + "learning_rate": 1.270129116494995e-05, + "loss": 0.0058, + "step": 10162 + }, + { + "epoch": 1.1017996530789245, + "grad_norm": 5.024414539337158, + "learning_rate": 1.2700565791382562e-05, + "loss": 0.027, + "step": 10163 + }, + { + "epoch": 1.1019080659150045, + "grad_norm": 4.173727989196777, + "learning_rate": 1.2699840417815177e-05, + "loss": 0.1063, + "step": 10164 + }, + { + "epoch": 1.1020164787510842, + "grad_norm": 0.27667802572250366, + "learning_rate": 1.269911504424779e-05, + "loss": 0.0018, + "step": 10165 + }, + { + "epoch": 1.102124891587164, + "grad_norm": 2.5336287021636963, + "learning_rate": 1.2698389670680402e-05, + "loss": 0.012, + "step": 10166 + }, + { + "epoch": 1.1022333044232437, + "grad_norm": 2.7631187438964844, + "learning_rate": 1.2697664297113015e-05, + "loss": 0.023, + "step": 10167 + }, + { + "epoch": 1.1023417172593235, + "grad_norm": 4.1801371574401855, + "learning_rate": 1.2696938923545627e-05, + "loss": 0.0302, + "step": 10168 + }, + { + "epoch": 1.1024501300954033, + "grad_norm": 0.09599796682596207, + "learning_rate": 1.269621354997824e-05, + "loss": 0.001, + "step": 10169 + }, + { + "epoch": 1.102558542931483, + "grad_norm": 0.10820082575082779, + "learning_rate": 1.2695488176410853e-05, + "loss": 0.001, + "step": 10170 + }, + { + "epoch": 1.1026669557675628, + "grad_norm": 2.6173903942108154, + "learning_rate": 1.2694762802843466e-05, + "loss": 0.018, + "step": 10171 + }, + { + "epoch": 1.1027753686036426, + "grad_norm": 1.1547459363937378, + "learning_rate": 1.2694037429276078e-05, + "loss": 0.0069, + "step": 10172 + }, + { + "epoch": 1.1028837814397225, + "grad_norm": 0.9795272350311279, + "learning_rate": 1.2693312055708692e-05, + "loss": 0.0176, + "step": 10173 + }, + { + "epoch": 1.1029921942758023, + "grad_norm": 0.1447223722934723, + "learning_rate": 1.2692586682141304e-05, + "loss": 0.0013, + "step": 10174 + }, + { + "epoch": 1.103100607111882, + "grad_norm": 3.2369086742401123, + "learning_rate": 1.2691861308573916e-05, + "loss": 0.0324, + "step": 10175 + }, + { + "epoch": 1.1032090199479618, + "grad_norm": 9.55897331237793, + "learning_rate": 1.2691135935006528e-05, + "loss": 0.0799, + "step": 10176 + }, + { + "epoch": 1.1033174327840416, + "grad_norm": 7.35979700088501, + "learning_rate": 1.2690410561439142e-05, + "loss": 0.0506, + "step": 10177 + }, + { + "epoch": 1.1034258456201214, + "grad_norm": 3.581557273864746, + "learning_rate": 1.2689685187871754e-05, + "loss": 0.0242, + "step": 10178 + }, + { + "epoch": 1.1035342584562011, + "grad_norm": 0.36655130982398987, + "learning_rate": 1.2688959814304366e-05, + "loss": 0.0046, + "step": 10179 + }, + { + "epoch": 1.103642671292281, + "grad_norm": 10.789104461669922, + "learning_rate": 1.268823444073698e-05, + "loss": 0.0685, + "step": 10180 + }, + { + "epoch": 1.1037510841283609, + "grad_norm": 4.334464073181152, + "learning_rate": 1.2687509067169594e-05, + "loss": 0.0264, + "step": 10181 + }, + { + "epoch": 1.1038594969644406, + "grad_norm": 2.2468183040618896, + "learning_rate": 1.2686783693602208e-05, + "loss": 0.0103, + "step": 10182 + }, + { + "epoch": 1.1039679098005204, + "grad_norm": 0.5220861434936523, + "learning_rate": 1.268605832003482e-05, + "loss": 0.0028, + "step": 10183 + }, + { + "epoch": 1.1040763226366002, + "grad_norm": 7.6618170738220215, + "learning_rate": 1.2685332946467432e-05, + "loss": 0.0554, + "step": 10184 + }, + { + "epoch": 1.10418473547268, + "grad_norm": 8.34553337097168, + "learning_rate": 1.2684607572900046e-05, + "loss": 0.0094, + "step": 10185 + }, + { + "epoch": 1.1042931483087597, + "grad_norm": 0.7339574694633484, + "learning_rate": 1.2683882199332658e-05, + "loss": 0.0042, + "step": 10186 + }, + { + "epoch": 1.1044015611448394, + "grad_norm": 0.3307735025882721, + "learning_rate": 1.268315682576527e-05, + "loss": 0.0015, + "step": 10187 + }, + { + "epoch": 1.1045099739809194, + "grad_norm": 2.233412265777588, + "learning_rate": 1.2682431452197884e-05, + "loss": 0.0104, + "step": 10188 + }, + { + "epoch": 1.1046183868169992, + "grad_norm": 8.529111862182617, + "learning_rate": 1.2681706078630496e-05, + "loss": 0.0146, + "step": 10189 + }, + { + "epoch": 1.104726799653079, + "grad_norm": 3.958578109741211, + "learning_rate": 1.2680980705063108e-05, + "loss": 0.0858, + "step": 10190 + }, + { + "epoch": 1.1048352124891587, + "grad_norm": 5.268475532531738, + "learning_rate": 1.268025533149572e-05, + "loss": 0.0093, + "step": 10191 + }, + { + "epoch": 1.1049436253252385, + "grad_norm": 1.7461951971054077, + "learning_rate": 1.2679529957928334e-05, + "loss": 0.027, + "step": 10192 + }, + { + "epoch": 1.1050520381613183, + "grad_norm": 2.1919503211975098, + "learning_rate": 1.2678804584360946e-05, + "loss": 0.0209, + "step": 10193 + }, + { + "epoch": 1.105160450997398, + "grad_norm": 8.951277732849121, + "learning_rate": 1.2678079210793558e-05, + "loss": 0.0636, + "step": 10194 + }, + { + "epoch": 1.1052688638334778, + "grad_norm": 3.79563570022583, + "learning_rate": 1.2677353837226172e-05, + "loss": 0.0747, + "step": 10195 + }, + { + "epoch": 1.1053772766695578, + "grad_norm": 1.427872657775879, + "learning_rate": 1.2676628463658784e-05, + "loss": 0.0112, + "step": 10196 + }, + { + "epoch": 1.1054856895056375, + "grad_norm": 2.035123348236084, + "learning_rate": 1.2675903090091397e-05, + "loss": 0.024, + "step": 10197 + }, + { + "epoch": 1.1055941023417173, + "grad_norm": 9.845048904418945, + "learning_rate": 1.2675177716524012e-05, + "loss": 0.1372, + "step": 10198 + }, + { + "epoch": 1.105702515177797, + "grad_norm": 1.5244044065475464, + "learning_rate": 1.2674452342956624e-05, + "loss": 0.031, + "step": 10199 + }, + { + "epoch": 1.1058109280138768, + "grad_norm": 1.9349756240844727, + "learning_rate": 1.2673726969389238e-05, + "loss": 0.0212, + "step": 10200 + }, + { + "epoch": 1.1059193408499566, + "grad_norm": 1.1660690307617188, + "learning_rate": 1.267300159582185e-05, + "loss": 0.0064, + "step": 10201 + }, + { + "epoch": 1.1060277536860363, + "grad_norm": 1.573014497756958, + "learning_rate": 1.2672276222254462e-05, + "loss": 0.0071, + "step": 10202 + }, + { + "epoch": 1.1061361665221163, + "grad_norm": 1.699020504951477, + "learning_rate": 1.2671550848687075e-05, + "loss": 0.0268, + "step": 10203 + }, + { + "epoch": 1.106244579358196, + "grad_norm": 5.115821361541748, + "learning_rate": 1.2670825475119688e-05, + "loss": 0.0205, + "step": 10204 + }, + { + "epoch": 1.1063529921942759, + "grad_norm": 1.2937870025634766, + "learning_rate": 1.26701001015523e-05, + "loss": 0.0146, + "step": 10205 + }, + { + "epoch": 1.1064614050303556, + "grad_norm": 2.6294469833374023, + "learning_rate": 1.2669374727984913e-05, + "loss": 0.0205, + "step": 10206 + }, + { + "epoch": 1.1065698178664354, + "grad_norm": 1.6337225437164307, + "learning_rate": 1.2668649354417526e-05, + "loss": 0.0065, + "step": 10207 + }, + { + "epoch": 1.1066782307025151, + "grad_norm": 0.20608209073543549, + "learning_rate": 1.2667923980850139e-05, + "loss": 0.0013, + "step": 10208 + }, + { + "epoch": 1.106786643538595, + "grad_norm": 1.3708641529083252, + "learning_rate": 1.266719860728275e-05, + "loss": 0.0166, + "step": 10209 + }, + { + "epoch": 1.1068950563746747, + "grad_norm": 1.0780367851257324, + "learning_rate": 1.2666473233715365e-05, + "loss": 0.0196, + "step": 10210 + }, + { + "epoch": 1.1070034692107547, + "grad_norm": 4.208450794219971, + "learning_rate": 1.2665747860147977e-05, + "loss": 0.0787, + "step": 10211 + }, + { + "epoch": 1.1071118820468344, + "grad_norm": 1.9008315801620483, + "learning_rate": 1.2665022486580589e-05, + "loss": 0.0443, + "step": 10212 + }, + { + "epoch": 1.1072202948829142, + "grad_norm": 0.5593231320381165, + "learning_rate": 1.2664297113013203e-05, + "loss": 0.0043, + "step": 10213 + }, + { + "epoch": 1.107328707718994, + "grad_norm": 0.9259340167045593, + "learning_rate": 1.2663571739445815e-05, + "loss": 0.0159, + "step": 10214 + }, + { + "epoch": 1.1074371205550737, + "grad_norm": 0.5494740605354309, + "learning_rate": 1.2662846365878429e-05, + "loss": 0.0035, + "step": 10215 + }, + { + "epoch": 1.1075455333911535, + "grad_norm": 0.7374987006187439, + "learning_rate": 1.2662120992311043e-05, + "loss": 0.0036, + "step": 10216 + }, + { + "epoch": 1.1076539462272332, + "grad_norm": 5.274246692657471, + "learning_rate": 1.2661395618743655e-05, + "loss": 0.0505, + "step": 10217 + }, + { + "epoch": 1.107762359063313, + "grad_norm": 5.255888938903809, + "learning_rate": 1.2660670245176267e-05, + "loss": 0.0356, + "step": 10218 + }, + { + "epoch": 1.107870771899393, + "grad_norm": 4.813749313354492, + "learning_rate": 1.265994487160888e-05, + "loss": 0.034, + "step": 10219 + }, + { + "epoch": 1.1079791847354727, + "grad_norm": 0.9652076959609985, + "learning_rate": 1.2659219498041493e-05, + "loss": 0.0174, + "step": 10220 + }, + { + "epoch": 1.1080875975715525, + "grad_norm": 0.6862908601760864, + "learning_rate": 1.2658494124474105e-05, + "loss": 0.0077, + "step": 10221 + }, + { + "epoch": 1.1081960104076323, + "grad_norm": 2.168825626373291, + "learning_rate": 1.2657768750906719e-05, + "loss": 0.0243, + "step": 10222 + }, + { + "epoch": 1.108304423243712, + "grad_norm": 4.463935375213623, + "learning_rate": 1.2657043377339331e-05, + "loss": 0.0373, + "step": 10223 + }, + { + "epoch": 1.1084128360797918, + "grad_norm": 4.3121771812438965, + "learning_rate": 1.2656318003771943e-05, + "loss": 0.0496, + "step": 10224 + }, + { + "epoch": 1.1085212489158716, + "grad_norm": 2.2172884941101074, + "learning_rate": 1.2655592630204557e-05, + "loss": 0.0427, + "step": 10225 + }, + { + "epoch": 1.1086296617519515, + "grad_norm": 0.6527466773986816, + "learning_rate": 1.2654867256637169e-05, + "loss": 0.0087, + "step": 10226 + }, + { + "epoch": 1.1087380745880313, + "grad_norm": 2.2828164100646973, + "learning_rate": 1.2654141883069781e-05, + "loss": 0.0083, + "step": 10227 + }, + { + "epoch": 1.108846487424111, + "grad_norm": 2.411937713623047, + "learning_rate": 1.2653416509502393e-05, + "loss": 0.0291, + "step": 10228 + }, + { + "epoch": 1.1089549002601908, + "grad_norm": 0.6941301822662354, + "learning_rate": 1.2652691135935007e-05, + "loss": 0.0099, + "step": 10229 + }, + { + "epoch": 1.1090633130962706, + "grad_norm": 0.7322220802307129, + "learning_rate": 1.265196576236762e-05, + "loss": 0.0088, + "step": 10230 + }, + { + "epoch": 1.1091717259323504, + "grad_norm": 3.429954767227173, + "learning_rate": 1.2651240388800231e-05, + "loss": 0.0337, + "step": 10231 + }, + { + "epoch": 1.1092801387684301, + "grad_norm": 1.4456894397735596, + "learning_rate": 1.2650515015232845e-05, + "loss": 0.0085, + "step": 10232 + }, + { + "epoch": 1.1093885516045099, + "grad_norm": 1.7358111143112183, + "learning_rate": 1.264978964166546e-05, + "loss": 0.0116, + "step": 10233 + }, + { + "epoch": 1.1094969644405897, + "grad_norm": 1.185176134109497, + "learning_rate": 1.2649064268098073e-05, + "loss": 0.0093, + "step": 10234 + }, + { + "epoch": 1.1096053772766696, + "grad_norm": 9.221698760986328, + "learning_rate": 1.2648338894530685e-05, + "loss": 0.0403, + "step": 10235 + }, + { + "epoch": 1.1097137901127494, + "grad_norm": 1.580173373222351, + "learning_rate": 1.2647613520963297e-05, + "loss": 0.0135, + "step": 10236 + }, + { + "epoch": 1.1098222029488292, + "grad_norm": 4.289511203765869, + "learning_rate": 1.2646888147395911e-05, + "loss": 0.045, + "step": 10237 + }, + { + "epoch": 1.109930615784909, + "grad_norm": 1.3069252967834473, + "learning_rate": 1.2646162773828523e-05, + "loss": 0.0041, + "step": 10238 + }, + { + "epoch": 1.1100390286209887, + "grad_norm": 5.915219783782959, + "learning_rate": 1.2645437400261135e-05, + "loss": 0.0242, + "step": 10239 + }, + { + "epoch": 1.1101474414570685, + "grad_norm": 4.060632228851318, + "learning_rate": 1.264471202669375e-05, + "loss": 0.0533, + "step": 10240 + }, + { + "epoch": 1.1102558542931482, + "grad_norm": 2.6692233085632324, + "learning_rate": 1.2643986653126361e-05, + "loss": 0.0049, + "step": 10241 + }, + { + "epoch": 1.1103642671292282, + "grad_norm": 1.676923155784607, + "learning_rate": 1.2643261279558974e-05, + "loss": 0.0111, + "step": 10242 + }, + { + "epoch": 1.110472679965308, + "grad_norm": 2.1495778560638428, + "learning_rate": 1.2642535905991586e-05, + "loss": 0.0081, + "step": 10243 + }, + { + "epoch": 1.1105810928013877, + "grad_norm": 1.9622013568878174, + "learning_rate": 1.26418105324242e-05, + "loss": 0.0119, + "step": 10244 + }, + { + "epoch": 1.1106895056374675, + "grad_norm": 2.6239984035491943, + "learning_rate": 1.2641085158856812e-05, + "loss": 0.0444, + "step": 10245 + }, + { + "epoch": 1.1107979184735473, + "grad_norm": 1.413711667060852, + "learning_rate": 1.2640359785289424e-05, + "loss": 0.0045, + "step": 10246 + }, + { + "epoch": 1.110906331309627, + "grad_norm": 5.77454137802124, + "learning_rate": 1.2639634411722038e-05, + "loss": 0.0348, + "step": 10247 + }, + { + "epoch": 1.1110147441457068, + "grad_norm": 2.268737554550171, + "learning_rate": 1.263890903815465e-05, + "loss": 0.0744, + "step": 10248 + }, + { + "epoch": 1.1111231569817865, + "grad_norm": 0.2916097044944763, + "learning_rate": 1.2638183664587262e-05, + "loss": 0.0028, + "step": 10249 + }, + { + "epoch": 1.1112315698178665, + "grad_norm": 0.22631794214248657, + "learning_rate": 1.2637458291019877e-05, + "loss": 0.0021, + "step": 10250 + }, + { + "epoch": 1.1113399826539463, + "grad_norm": 2.099576711654663, + "learning_rate": 1.263673291745249e-05, + "loss": 0.012, + "step": 10251 + }, + { + "epoch": 1.111448395490026, + "grad_norm": 0.5919299721717834, + "learning_rate": 1.2636007543885103e-05, + "loss": 0.0057, + "step": 10252 + }, + { + "epoch": 1.1115568083261058, + "grad_norm": 2.7535510063171387, + "learning_rate": 1.2635282170317716e-05, + "loss": 0.0617, + "step": 10253 + }, + { + "epoch": 1.1116652211621856, + "grad_norm": 4.5971856117248535, + "learning_rate": 1.2634556796750328e-05, + "loss": 0.0326, + "step": 10254 + }, + { + "epoch": 1.1117736339982653, + "grad_norm": 0.53492271900177, + "learning_rate": 1.263383142318294e-05, + "loss": 0.0036, + "step": 10255 + }, + { + "epoch": 1.111882046834345, + "grad_norm": 5.85207462310791, + "learning_rate": 1.2633106049615554e-05, + "loss": 0.0556, + "step": 10256 + }, + { + "epoch": 1.1119904596704249, + "grad_norm": 0.29306361079216003, + "learning_rate": 1.2632380676048166e-05, + "loss": 0.0019, + "step": 10257 + }, + { + "epoch": 1.1120988725065049, + "grad_norm": 2.3949551582336426, + "learning_rate": 1.2631655302480778e-05, + "loss": 0.0296, + "step": 10258 + }, + { + "epoch": 1.1122072853425846, + "grad_norm": 2.7598612308502197, + "learning_rate": 1.2630929928913392e-05, + "loss": 0.0632, + "step": 10259 + }, + { + "epoch": 1.1123156981786644, + "grad_norm": 14.332667350769043, + "learning_rate": 1.2630204555346004e-05, + "loss": 0.0435, + "step": 10260 + }, + { + "epoch": 1.1124241110147441, + "grad_norm": 1.7271099090576172, + "learning_rate": 1.2629479181778616e-05, + "loss": 0.0081, + "step": 10261 + }, + { + "epoch": 1.112532523850824, + "grad_norm": 3.300029754638672, + "learning_rate": 1.262875380821123e-05, + "loss": 0.0438, + "step": 10262 + }, + { + "epoch": 1.1126409366869037, + "grad_norm": 3.6534297466278076, + "learning_rate": 1.2628028434643842e-05, + "loss": 0.0087, + "step": 10263 + }, + { + "epoch": 1.1127493495229834, + "grad_norm": 0.09416160732507706, + "learning_rate": 1.2627303061076454e-05, + "loss": 0.0009, + "step": 10264 + }, + { + "epoch": 1.1128577623590634, + "grad_norm": 0.17677170038223267, + "learning_rate": 1.2626577687509068e-05, + "loss": 0.0007, + "step": 10265 + }, + { + "epoch": 1.1129661751951432, + "grad_norm": 0.3093985617160797, + "learning_rate": 1.262585231394168e-05, + "loss": 0.0009, + "step": 10266 + }, + { + "epoch": 1.113074588031223, + "grad_norm": 5.624025821685791, + "learning_rate": 1.2625126940374294e-05, + "loss": 0.0298, + "step": 10267 + }, + { + "epoch": 1.1131830008673027, + "grad_norm": 4.449479579925537, + "learning_rate": 1.2624401566806908e-05, + "loss": 0.0348, + "step": 10268 + }, + { + "epoch": 1.1132914137033825, + "grad_norm": 1.0431010723114014, + "learning_rate": 1.262367619323952e-05, + "loss": 0.0141, + "step": 10269 + }, + { + "epoch": 1.1133998265394622, + "grad_norm": 3.583038091659546, + "learning_rate": 1.2622950819672132e-05, + "loss": 0.0375, + "step": 10270 + }, + { + "epoch": 1.113508239375542, + "grad_norm": 3.918065309524536, + "learning_rate": 1.2622225446104746e-05, + "loss": 0.0226, + "step": 10271 + }, + { + "epoch": 1.1136166522116218, + "grad_norm": 1.3822004795074463, + "learning_rate": 1.2621500072537358e-05, + "loss": 0.047, + "step": 10272 + }, + { + "epoch": 1.1137250650477017, + "grad_norm": 1.3532004356384277, + "learning_rate": 1.262077469896997e-05, + "loss": 0.0098, + "step": 10273 + }, + { + "epoch": 1.1138334778837815, + "grad_norm": 2.644845962524414, + "learning_rate": 1.2620049325402584e-05, + "loss": 0.0226, + "step": 10274 + }, + { + "epoch": 1.1139418907198613, + "grad_norm": 2.0988423824310303, + "learning_rate": 1.2619323951835196e-05, + "loss": 0.0442, + "step": 10275 + }, + { + "epoch": 1.114050303555941, + "grad_norm": 1.586164116859436, + "learning_rate": 1.2618598578267808e-05, + "loss": 0.0098, + "step": 10276 + }, + { + "epoch": 1.1141587163920208, + "grad_norm": 2.868159294128418, + "learning_rate": 1.2617873204700422e-05, + "loss": 0.0269, + "step": 10277 + }, + { + "epoch": 1.1142671292281006, + "grad_norm": 2.118804454803467, + "learning_rate": 1.2617147831133034e-05, + "loss": 0.0106, + "step": 10278 + }, + { + "epoch": 1.1143755420641803, + "grad_norm": 2.4666202068328857, + "learning_rate": 1.2616422457565647e-05, + "loss": 0.029, + "step": 10279 + }, + { + "epoch": 1.11448395490026, + "grad_norm": 5.2343292236328125, + "learning_rate": 1.2615697083998259e-05, + "loss": 0.0127, + "step": 10280 + }, + { + "epoch": 1.11459236773634, + "grad_norm": 2.348787784576416, + "learning_rate": 1.2614971710430873e-05, + "loss": 0.0095, + "step": 10281 + }, + { + "epoch": 1.1147007805724198, + "grad_norm": 2.3581161499023438, + "learning_rate": 1.2614246336863485e-05, + "loss": 0.0345, + "step": 10282 + }, + { + "epoch": 1.1148091934084996, + "grad_norm": 0.4299543499946594, + "learning_rate": 1.2613520963296097e-05, + "loss": 0.0048, + "step": 10283 + }, + { + "epoch": 1.1149176062445794, + "grad_norm": 0.6346377730369568, + "learning_rate": 1.261279558972871e-05, + "loss": 0.0025, + "step": 10284 + }, + { + "epoch": 1.1150260190806591, + "grad_norm": 2.658402681350708, + "learning_rate": 1.2612070216161324e-05, + "loss": 0.0157, + "step": 10285 + }, + { + "epoch": 1.1151344319167389, + "grad_norm": 1.9532601833343506, + "learning_rate": 1.2611344842593938e-05, + "loss": 0.0149, + "step": 10286 + }, + { + "epoch": 1.1152428447528187, + "grad_norm": 1.157078504562378, + "learning_rate": 1.261061946902655e-05, + "loss": 0.0052, + "step": 10287 + }, + { + "epoch": 1.1153512575888986, + "grad_norm": 3.3451454639434814, + "learning_rate": 1.2609894095459163e-05, + "loss": 0.0516, + "step": 10288 + }, + { + "epoch": 1.1154596704249784, + "grad_norm": 0.8753933310508728, + "learning_rate": 1.2609168721891776e-05, + "loss": 0.0062, + "step": 10289 + }, + { + "epoch": 1.1155680832610582, + "grad_norm": 0.41297638416290283, + "learning_rate": 1.2608443348324389e-05, + "loss": 0.0018, + "step": 10290 + }, + { + "epoch": 1.115676496097138, + "grad_norm": 3.45290470123291, + "learning_rate": 1.2607717974757e-05, + "loss": 0.0302, + "step": 10291 + }, + { + "epoch": 1.1157849089332177, + "grad_norm": 1.8983768224716187, + "learning_rate": 1.2606992601189615e-05, + "loss": 0.0112, + "step": 10292 + }, + { + "epoch": 1.1158933217692975, + "grad_norm": 3.1393868923187256, + "learning_rate": 1.2606267227622227e-05, + "loss": 0.029, + "step": 10293 + }, + { + "epoch": 1.1160017346053772, + "grad_norm": 3.000072956085205, + "learning_rate": 1.2605541854054839e-05, + "loss": 0.0129, + "step": 10294 + }, + { + "epoch": 1.116110147441457, + "grad_norm": 0.296774685382843, + "learning_rate": 1.2604816480487451e-05, + "loss": 0.0044, + "step": 10295 + }, + { + "epoch": 1.116218560277537, + "grad_norm": 3.9000771045684814, + "learning_rate": 1.2604091106920065e-05, + "loss": 0.0436, + "step": 10296 + }, + { + "epoch": 1.1163269731136167, + "grad_norm": 3.1116154193878174, + "learning_rate": 1.2603365733352677e-05, + "loss": 0.0319, + "step": 10297 + }, + { + "epoch": 1.1164353859496965, + "grad_norm": 1.049696683883667, + "learning_rate": 1.2602640359785289e-05, + "loss": 0.0037, + "step": 10298 + }, + { + "epoch": 1.1165437987857763, + "grad_norm": 1.0804600715637207, + "learning_rate": 1.2601914986217903e-05, + "loss": 0.017, + "step": 10299 + }, + { + "epoch": 1.116652211621856, + "grad_norm": 1.7980408668518066, + "learning_rate": 1.2601189612650515e-05, + "loss": 0.0114, + "step": 10300 + }, + { + "epoch": 1.1167606244579358, + "grad_norm": 5.013002872467041, + "learning_rate": 1.2600464239083127e-05, + "loss": 0.0085, + "step": 10301 + }, + { + "epoch": 1.1168690372940155, + "grad_norm": 6.826665878295898, + "learning_rate": 1.2599738865515743e-05, + "loss": 0.0282, + "step": 10302 + }, + { + "epoch": 1.1169774501300953, + "grad_norm": 7.385800361633301, + "learning_rate": 1.2599013491948355e-05, + "loss": 0.0408, + "step": 10303 + }, + { + "epoch": 1.1170858629661753, + "grad_norm": 2.5563924312591553, + "learning_rate": 1.2598288118380969e-05, + "loss": 0.0903, + "step": 10304 + }, + { + "epoch": 1.117194275802255, + "grad_norm": 1.2798287868499756, + "learning_rate": 1.2597562744813581e-05, + "loss": 0.013, + "step": 10305 + }, + { + "epoch": 1.1173026886383348, + "grad_norm": 4.012268543243408, + "learning_rate": 1.2596837371246193e-05, + "loss": 0.0473, + "step": 10306 + }, + { + "epoch": 1.1174111014744146, + "grad_norm": 1.191202998161316, + "learning_rate": 1.2596111997678805e-05, + "loss": 0.0049, + "step": 10307 + }, + { + "epoch": 1.1175195143104943, + "grad_norm": 0.2391948252916336, + "learning_rate": 1.2595386624111419e-05, + "loss": 0.002, + "step": 10308 + }, + { + "epoch": 1.117627927146574, + "grad_norm": 2.232689619064331, + "learning_rate": 1.2594661250544031e-05, + "loss": 0.0061, + "step": 10309 + }, + { + "epoch": 1.1177363399826539, + "grad_norm": 2.207911968231201, + "learning_rate": 1.2593935876976643e-05, + "loss": 0.0131, + "step": 10310 + }, + { + "epoch": 1.1178447528187339, + "grad_norm": 1.8440946340560913, + "learning_rate": 1.2593210503409257e-05, + "loss": 0.0162, + "step": 10311 + }, + { + "epoch": 1.1179531656548136, + "grad_norm": 3.613603353500366, + "learning_rate": 1.259248512984187e-05, + "loss": 0.0288, + "step": 10312 + }, + { + "epoch": 1.1180615784908934, + "grad_norm": 1.7133461236953735, + "learning_rate": 1.2591759756274481e-05, + "loss": 0.0121, + "step": 10313 + }, + { + "epoch": 1.1181699913269731, + "grad_norm": 3.44966459274292, + "learning_rate": 1.2591034382707095e-05, + "loss": 0.0247, + "step": 10314 + }, + { + "epoch": 1.118278404163053, + "grad_norm": 9.411883354187012, + "learning_rate": 1.2590309009139707e-05, + "loss": 0.0558, + "step": 10315 + }, + { + "epoch": 1.1183868169991327, + "grad_norm": 0.6365850567817688, + "learning_rate": 1.258958363557232e-05, + "loss": 0.0036, + "step": 10316 + }, + { + "epoch": 1.1184952298352124, + "grad_norm": 1.8463106155395508, + "learning_rate": 1.2588858262004933e-05, + "loss": 0.0322, + "step": 10317 + }, + { + "epoch": 1.1186036426712922, + "grad_norm": 0.9373073577880859, + "learning_rate": 1.2588132888437546e-05, + "loss": 0.0164, + "step": 10318 + }, + { + "epoch": 1.118712055507372, + "grad_norm": 2.6147451400756836, + "learning_rate": 1.258740751487016e-05, + "loss": 0.0189, + "step": 10319 + }, + { + "epoch": 1.118820468343452, + "grad_norm": 2.266547679901123, + "learning_rate": 1.2586682141302773e-05, + "loss": 0.0443, + "step": 10320 + }, + { + "epoch": 1.1189288811795317, + "grad_norm": 4.22499418258667, + "learning_rate": 1.2585956767735385e-05, + "loss": 0.0654, + "step": 10321 + }, + { + "epoch": 1.1190372940156115, + "grad_norm": 0.12885139882564545, + "learning_rate": 1.2585231394167998e-05, + "loss": 0.0013, + "step": 10322 + }, + { + "epoch": 1.1191457068516912, + "grad_norm": 0.7459959387779236, + "learning_rate": 1.2584506020600611e-05, + "loss": 0.0042, + "step": 10323 + }, + { + "epoch": 1.119254119687771, + "grad_norm": 3.6303563117980957, + "learning_rate": 1.2583780647033223e-05, + "loss": 0.0334, + "step": 10324 + }, + { + "epoch": 1.1193625325238508, + "grad_norm": 2.6749465465545654, + "learning_rate": 1.2583055273465836e-05, + "loss": 0.0256, + "step": 10325 + }, + { + "epoch": 1.1194709453599305, + "grad_norm": 2.5122272968292236, + "learning_rate": 1.258232989989845e-05, + "loss": 0.0054, + "step": 10326 + }, + { + "epoch": 1.1195793581960105, + "grad_norm": 2.851895332336426, + "learning_rate": 1.2581604526331062e-05, + "loss": 0.0433, + "step": 10327 + }, + { + "epoch": 1.1196877710320903, + "grad_norm": 4.535192966461182, + "learning_rate": 1.2580879152763674e-05, + "loss": 0.0281, + "step": 10328 + }, + { + "epoch": 1.11979618386817, + "grad_norm": 0.7245651483535767, + "learning_rate": 1.2580153779196288e-05, + "loss": 0.0048, + "step": 10329 + }, + { + "epoch": 1.1199045967042498, + "grad_norm": 2.0218894481658936, + "learning_rate": 1.25794284056289e-05, + "loss": 0.0071, + "step": 10330 + }, + { + "epoch": 1.1200130095403296, + "grad_norm": 0.5973336100578308, + "learning_rate": 1.2578703032061512e-05, + "loss": 0.01, + "step": 10331 + }, + { + "epoch": 1.1201214223764093, + "grad_norm": 18.280298233032227, + "learning_rate": 1.2577977658494126e-05, + "loss": 0.0479, + "step": 10332 + }, + { + "epoch": 1.120229835212489, + "grad_norm": 0.8211955428123474, + "learning_rate": 1.2577252284926738e-05, + "loss": 0.0049, + "step": 10333 + }, + { + "epoch": 1.1203382480485689, + "grad_norm": 3.6163320541381836, + "learning_rate": 1.257652691135935e-05, + "loss": 0.0133, + "step": 10334 + }, + { + "epoch": 1.1204466608846488, + "grad_norm": 3.1438729763031006, + "learning_rate": 1.2575801537791962e-05, + "loss": 0.0366, + "step": 10335 + }, + { + "epoch": 1.1205550737207286, + "grad_norm": 0.3941115438938141, + "learning_rate": 1.2575076164224578e-05, + "loss": 0.004, + "step": 10336 + }, + { + "epoch": 1.1206634865568084, + "grad_norm": 0.5712740421295166, + "learning_rate": 1.257435079065719e-05, + "loss": 0.0053, + "step": 10337 + }, + { + "epoch": 1.1207718993928881, + "grad_norm": 0.7742627859115601, + "learning_rate": 1.2573625417089804e-05, + "loss": 0.0072, + "step": 10338 + }, + { + "epoch": 1.120880312228968, + "grad_norm": 1.3531818389892578, + "learning_rate": 1.2572900043522416e-05, + "loss": 0.0107, + "step": 10339 + }, + { + "epoch": 1.1209887250650477, + "grad_norm": 0.4982919692993164, + "learning_rate": 1.2572174669955028e-05, + "loss": 0.0033, + "step": 10340 + }, + { + "epoch": 1.1210971379011274, + "grad_norm": 0.11382073909044266, + "learning_rate": 1.2571449296387642e-05, + "loss": 0.0008, + "step": 10341 + }, + { + "epoch": 1.1212055507372072, + "grad_norm": 2.938082218170166, + "learning_rate": 1.2570723922820254e-05, + "loss": 0.018, + "step": 10342 + }, + { + "epoch": 1.1213139635732872, + "grad_norm": 4.39736270904541, + "learning_rate": 1.2569998549252866e-05, + "loss": 0.0583, + "step": 10343 + }, + { + "epoch": 1.121422376409367, + "grad_norm": 2.050408363342285, + "learning_rate": 1.256927317568548e-05, + "loss": 0.0413, + "step": 10344 + }, + { + "epoch": 1.1215307892454467, + "grad_norm": 0.5961617231369019, + "learning_rate": 1.2568547802118092e-05, + "loss": 0.0041, + "step": 10345 + }, + { + "epoch": 1.1216392020815265, + "grad_norm": 0.2098778784275055, + "learning_rate": 1.2567822428550704e-05, + "loss": 0.0015, + "step": 10346 + }, + { + "epoch": 1.1217476149176062, + "grad_norm": 3.2519497871398926, + "learning_rate": 1.2567097054983316e-05, + "loss": 0.018, + "step": 10347 + }, + { + "epoch": 1.121856027753686, + "grad_norm": 1.1259660720825195, + "learning_rate": 1.256637168141593e-05, + "loss": 0.0103, + "step": 10348 + }, + { + "epoch": 1.1219644405897657, + "grad_norm": 2.115053415298462, + "learning_rate": 1.2565646307848542e-05, + "loss": 0.0175, + "step": 10349 + }, + { + "epoch": 1.1220728534258457, + "grad_norm": 0.26164713501930237, + "learning_rate": 1.2564920934281154e-05, + "loss": 0.0011, + "step": 10350 + }, + { + "epoch": 1.1221812662619255, + "grad_norm": 1.1800506114959717, + "learning_rate": 1.2564195560713768e-05, + "loss": 0.006, + "step": 10351 + }, + { + "epoch": 1.1222896790980053, + "grad_norm": 2.8302385807037354, + "learning_rate": 1.256347018714638e-05, + "loss": 0.0496, + "step": 10352 + }, + { + "epoch": 1.122398091934085, + "grad_norm": 6.705056667327881, + "learning_rate": 1.2562744813578993e-05, + "loss": 0.042, + "step": 10353 + }, + { + "epoch": 1.1225065047701648, + "grad_norm": 0.36867254972457886, + "learning_rate": 1.2562019440011608e-05, + "loss": 0.0012, + "step": 10354 + }, + { + "epoch": 1.1226149176062445, + "grad_norm": 1.1409404277801514, + "learning_rate": 1.256129406644422e-05, + "loss": 0.0305, + "step": 10355 + }, + { + "epoch": 1.1227233304423243, + "grad_norm": 0.9734519720077515, + "learning_rate": 1.2560568692876834e-05, + "loss": 0.0028, + "step": 10356 + }, + { + "epoch": 1.122831743278404, + "grad_norm": 0.5512242913246155, + "learning_rate": 1.2559843319309446e-05, + "loss": 0.005, + "step": 10357 + }, + { + "epoch": 1.122940156114484, + "grad_norm": 4.3284783363342285, + "learning_rate": 1.2559117945742058e-05, + "loss": 0.0984, + "step": 10358 + }, + { + "epoch": 1.1230485689505638, + "grad_norm": 0.12063323706388474, + "learning_rate": 1.255839257217467e-05, + "loss": 0.0007, + "step": 10359 + }, + { + "epoch": 1.1231569817866436, + "grad_norm": 5.6766533851623535, + "learning_rate": 1.2557667198607284e-05, + "loss": 0.038, + "step": 10360 + }, + { + "epoch": 1.1232653946227233, + "grad_norm": 1.3108208179473877, + "learning_rate": 1.2556941825039897e-05, + "loss": 0.0044, + "step": 10361 + }, + { + "epoch": 1.123373807458803, + "grad_norm": 1.248775839805603, + "learning_rate": 1.2556216451472509e-05, + "loss": 0.0313, + "step": 10362 + }, + { + "epoch": 1.1234822202948829, + "grad_norm": 0.4003758132457733, + "learning_rate": 1.2555491077905123e-05, + "loss": 0.0016, + "step": 10363 + }, + { + "epoch": 1.1235906331309626, + "grad_norm": 3.1425576210021973, + "learning_rate": 1.2554765704337735e-05, + "loss": 0.0229, + "step": 10364 + }, + { + "epoch": 1.1236990459670424, + "grad_norm": 0.8150647282600403, + "learning_rate": 1.2554040330770347e-05, + "loss": 0.0123, + "step": 10365 + }, + { + "epoch": 1.1238074588031224, + "grad_norm": 0.2704700529575348, + "learning_rate": 1.255331495720296e-05, + "loss": 0.0016, + "step": 10366 + }, + { + "epoch": 1.1239158716392021, + "grad_norm": 0.9388219118118286, + "learning_rate": 1.2552589583635573e-05, + "loss": 0.0017, + "step": 10367 + }, + { + "epoch": 1.124024284475282, + "grad_norm": 1.3233481645584106, + "learning_rate": 1.2551864210068185e-05, + "loss": 0.0097, + "step": 10368 + }, + { + "epoch": 1.1241326973113617, + "grad_norm": 4.981418132781982, + "learning_rate": 1.2551138836500799e-05, + "loss": 0.011, + "step": 10369 + }, + { + "epoch": 1.1242411101474414, + "grad_norm": 5.024017333984375, + "learning_rate": 1.2550413462933411e-05, + "loss": 0.0511, + "step": 10370 + }, + { + "epoch": 1.1243495229835212, + "grad_norm": 0.13312716782093048, + "learning_rate": 1.2549688089366025e-05, + "loss": 0.0005, + "step": 10371 + }, + { + "epoch": 1.124457935819601, + "grad_norm": 8.085797309875488, + "learning_rate": 1.2548962715798639e-05, + "loss": 0.0492, + "step": 10372 + }, + { + "epoch": 1.124566348655681, + "grad_norm": 2.290024757385254, + "learning_rate": 1.254823734223125e-05, + "loss": 0.016, + "step": 10373 + }, + { + "epoch": 1.1246747614917607, + "grad_norm": 2.0613698959350586, + "learning_rate": 1.2547511968663863e-05, + "loss": 0.0064, + "step": 10374 + }, + { + "epoch": 1.1247831743278405, + "grad_norm": 0.02706063911318779, + "learning_rate": 1.2546786595096477e-05, + "loss": 0.0003, + "step": 10375 + }, + { + "epoch": 1.1248915871639202, + "grad_norm": 2.8505570888519287, + "learning_rate": 1.2546061221529089e-05, + "loss": 0.0265, + "step": 10376 + }, + { + "epoch": 1.125, + "grad_norm": 0.17708931863307953, + "learning_rate": 1.2545335847961701e-05, + "loss": 0.0018, + "step": 10377 + }, + { + "epoch": 1.1251084128360798, + "grad_norm": 0.485818475484848, + "learning_rate": 1.2544610474394315e-05, + "loss": 0.0059, + "step": 10378 + }, + { + "epoch": 1.1252168256721595, + "grad_norm": 1.8922743797302246, + "learning_rate": 1.2543885100826927e-05, + "loss": 0.0255, + "step": 10379 + }, + { + "epoch": 1.1253252385082393, + "grad_norm": 9.750527381896973, + "learning_rate": 1.2543159727259539e-05, + "loss": 0.0258, + "step": 10380 + }, + { + "epoch": 1.125433651344319, + "grad_norm": 2.3685495853424072, + "learning_rate": 1.2542434353692153e-05, + "loss": 0.0219, + "step": 10381 + }, + { + "epoch": 1.125542064180399, + "grad_norm": 3.988415479660034, + "learning_rate": 1.2541708980124765e-05, + "loss": 0.0116, + "step": 10382 + }, + { + "epoch": 1.1256504770164788, + "grad_norm": 1.5664050579071045, + "learning_rate": 1.2540983606557377e-05, + "loss": 0.0295, + "step": 10383 + }, + { + "epoch": 1.1257588898525586, + "grad_norm": 0.10507801175117493, + "learning_rate": 1.2540258232989991e-05, + "loss": 0.001, + "step": 10384 + }, + { + "epoch": 1.1258673026886383, + "grad_norm": 11.081219673156738, + "learning_rate": 1.2539532859422603e-05, + "loss": 0.0107, + "step": 10385 + }, + { + "epoch": 1.125975715524718, + "grad_norm": 5.3087968826293945, + "learning_rate": 1.2538807485855215e-05, + "loss": 0.0107, + "step": 10386 + }, + { + "epoch": 1.1260841283607979, + "grad_norm": 1.5419609546661377, + "learning_rate": 1.2538082112287827e-05, + "loss": 0.0195, + "step": 10387 + }, + { + "epoch": 1.1261925411968776, + "grad_norm": 0.9951841831207275, + "learning_rate": 1.2537356738720443e-05, + "loss": 0.0176, + "step": 10388 + }, + { + "epoch": 1.1263009540329576, + "grad_norm": 3.938720226287842, + "learning_rate": 1.2536631365153055e-05, + "loss": 0.0853, + "step": 10389 + }, + { + "epoch": 1.1264093668690374, + "grad_norm": 3.7845020294189453, + "learning_rate": 1.2535905991585669e-05, + "loss": 0.0357, + "step": 10390 + }, + { + "epoch": 1.1265177797051171, + "grad_norm": 0.536945104598999, + "learning_rate": 1.2535180618018281e-05, + "loss": 0.0049, + "step": 10391 + }, + { + "epoch": 1.126626192541197, + "grad_norm": 2.7023260593414307, + "learning_rate": 1.2534455244450893e-05, + "loss": 0.0135, + "step": 10392 + }, + { + "epoch": 1.1267346053772767, + "grad_norm": 2.011730909347534, + "learning_rate": 1.2533729870883507e-05, + "loss": 0.0243, + "step": 10393 + }, + { + "epoch": 1.1268430182133564, + "grad_norm": 0.2870978116989136, + "learning_rate": 1.253300449731612e-05, + "loss": 0.0025, + "step": 10394 + }, + { + "epoch": 1.1269514310494362, + "grad_norm": 0.5083059072494507, + "learning_rate": 1.2532279123748731e-05, + "loss": 0.0084, + "step": 10395 + }, + { + "epoch": 1.1270598438855162, + "grad_norm": 0.08183609694242477, + "learning_rate": 1.2531553750181345e-05, + "loss": 0.0007, + "step": 10396 + }, + { + "epoch": 1.127168256721596, + "grad_norm": 3.890216827392578, + "learning_rate": 1.2530828376613957e-05, + "loss": 0.0591, + "step": 10397 + }, + { + "epoch": 1.1272766695576757, + "grad_norm": 1.7075811624526978, + "learning_rate": 1.253010300304657e-05, + "loss": 0.0337, + "step": 10398 + }, + { + "epoch": 1.1273850823937555, + "grad_norm": 12.453351020812988, + "learning_rate": 1.2529377629479182e-05, + "loss": 0.0115, + "step": 10399 + }, + { + "epoch": 1.1274934952298352, + "grad_norm": 8.713900566101074, + "learning_rate": 1.2528652255911796e-05, + "loss": 0.0197, + "step": 10400 + }, + { + "epoch": 1.127601908065915, + "grad_norm": 3.8616559505462646, + "learning_rate": 1.2527926882344408e-05, + "loss": 0.0167, + "step": 10401 + }, + { + "epoch": 1.1277103209019947, + "grad_norm": 1.0953795909881592, + "learning_rate": 1.252720150877702e-05, + "loss": 0.006, + "step": 10402 + }, + { + "epoch": 1.1278187337380745, + "grad_norm": 3.4939661026000977, + "learning_rate": 1.2526476135209634e-05, + "loss": 0.0127, + "step": 10403 + }, + { + "epoch": 1.1279271465741543, + "grad_norm": 3.842339277267456, + "learning_rate": 1.2525750761642246e-05, + "loss": 0.0244, + "step": 10404 + }, + { + "epoch": 1.1280355594102343, + "grad_norm": 3.2255666255950928, + "learning_rate": 1.2525025388074861e-05, + "loss": 0.0279, + "step": 10405 + }, + { + "epoch": 1.128143972246314, + "grad_norm": 4.538171291351318, + "learning_rate": 1.2524300014507473e-05, + "loss": 0.0436, + "step": 10406 + }, + { + "epoch": 1.1282523850823938, + "grad_norm": 1.5988668203353882, + "learning_rate": 1.2523574640940086e-05, + "loss": 0.0048, + "step": 10407 + }, + { + "epoch": 1.1283607979184735, + "grad_norm": 2.4138283729553223, + "learning_rate": 1.25228492673727e-05, + "loss": 0.0163, + "step": 10408 + }, + { + "epoch": 1.1284692107545533, + "grad_norm": 5.256359577178955, + "learning_rate": 1.2522123893805312e-05, + "loss": 0.026, + "step": 10409 + }, + { + "epoch": 1.128577623590633, + "grad_norm": 1.9938725233078003, + "learning_rate": 1.2521398520237924e-05, + "loss": 0.0036, + "step": 10410 + }, + { + "epoch": 1.1286860364267128, + "grad_norm": 0.8947387337684631, + "learning_rate": 1.2520673146670536e-05, + "loss": 0.0042, + "step": 10411 + }, + { + "epoch": 1.1287944492627928, + "grad_norm": 1.0231592655181885, + "learning_rate": 1.251994777310315e-05, + "loss": 0.0072, + "step": 10412 + }, + { + "epoch": 1.1289028620988726, + "grad_norm": 1.0518407821655273, + "learning_rate": 1.2519222399535762e-05, + "loss": 0.0126, + "step": 10413 + }, + { + "epoch": 1.1290112749349523, + "grad_norm": 2.306643009185791, + "learning_rate": 1.2518497025968374e-05, + "loss": 0.0184, + "step": 10414 + }, + { + "epoch": 1.129119687771032, + "grad_norm": 1.496917486190796, + "learning_rate": 1.2517771652400988e-05, + "loss": 0.0138, + "step": 10415 + }, + { + "epoch": 1.1292281006071119, + "grad_norm": 2.778191328048706, + "learning_rate": 1.25170462788336e-05, + "loss": 0.0153, + "step": 10416 + }, + { + "epoch": 1.1293365134431916, + "grad_norm": 4.388197898864746, + "learning_rate": 1.2516320905266212e-05, + "loss": 0.0232, + "step": 10417 + }, + { + "epoch": 1.1294449262792714, + "grad_norm": 4.163319110870361, + "learning_rate": 1.2515595531698826e-05, + "loss": 0.0102, + "step": 10418 + }, + { + "epoch": 1.1295533391153512, + "grad_norm": 0.4878508746623993, + "learning_rate": 1.2514870158131438e-05, + "loss": 0.0023, + "step": 10419 + }, + { + "epoch": 1.129661751951431, + "grad_norm": 4.926235198974609, + "learning_rate": 1.251414478456405e-05, + "loss": 0.0265, + "step": 10420 + }, + { + "epoch": 1.129770164787511, + "grad_norm": 3.1332736015319824, + "learning_rate": 1.2513419410996664e-05, + "loss": 0.0067, + "step": 10421 + }, + { + "epoch": 1.1298785776235907, + "grad_norm": 1.3838874101638794, + "learning_rate": 1.2512694037429276e-05, + "loss": 0.0151, + "step": 10422 + }, + { + "epoch": 1.1299869904596704, + "grad_norm": 1.86652672290802, + "learning_rate": 1.251196866386189e-05, + "loss": 0.006, + "step": 10423 + }, + { + "epoch": 1.1300954032957502, + "grad_norm": 1.8534901142120361, + "learning_rate": 1.2511243290294504e-05, + "loss": 0.0101, + "step": 10424 + }, + { + "epoch": 1.13020381613183, + "grad_norm": 2.1340126991271973, + "learning_rate": 1.2510517916727116e-05, + "loss": 0.0263, + "step": 10425 + }, + { + "epoch": 1.1303122289679097, + "grad_norm": 7.822855472564697, + "learning_rate": 1.2509792543159728e-05, + "loss": 0.0518, + "step": 10426 + }, + { + "epoch": 1.1304206418039895, + "grad_norm": 1.4275591373443604, + "learning_rate": 1.2509067169592342e-05, + "loss": 0.0168, + "step": 10427 + }, + { + "epoch": 1.1305290546400695, + "grad_norm": 2.6632046699523926, + "learning_rate": 1.2508341796024954e-05, + "loss": 0.0493, + "step": 10428 + }, + { + "epoch": 1.1306374674761492, + "grad_norm": 2.2804672718048096, + "learning_rate": 1.2507616422457566e-05, + "loss": 0.0104, + "step": 10429 + }, + { + "epoch": 1.130745880312229, + "grad_norm": 3.410876512527466, + "learning_rate": 1.250689104889018e-05, + "loss": 0.0219, + "step": 10430 + }, + { + "epoch": 1.1308542931483088, + "grad_norm": 3.793426752090454, + "learning_rate": 1.2506165675322792e-05, + "loss": 0.0135, + "step": 10431 + }, + { + "epoch": 1.1309627059843885, + "grad_norm": 5.293542385101318, + "learning_rate": 1.2505440301755404e-05, + "loss": 0.0351, + "step": 10432 + }, + { + "epoch": 1.1310711188204683, + "grad_norm": 3.027252435684204, + "learning_rate": 1.2504714928188018e-05, + "loss": 0.0235, + "step": 10433 + }, + { + "epoch": 1.131179531656548, + "grad_norm": 0.6606712341308594, + "learning_rate": 1.250398955462063e-05, + "loss": 0.0034, + "step": 10434 + }, + { + "epoch": 1.131287944492628, + "grad_norm": 2.3067104816436768, + "learning_rate": 1.2503264181053243e-05, + "loss": 0.0137, + "step": 10435 + }, + { + "epoch": 1.1313963573287078, + "grad_norm": 8.800302505493164, + "learning_rate": 1.2502538807485856e-05, + "loss": 0.0286, + "step": 10436 + }, + { + "epoch": 1.1315047701647876, + "grad_norm": 5.642056465148926, + "learning_rate": 1.2501813433918469e-05, + "loss": 0.0686, + "step": 10437 + }, + { + "epoch": 1.1316131830008673, + "grad_norm": 3.8944060802459717, + "learning_rate": 1.250108806035108e-05, + "loss": 0.0238, + "step": 10438 + }, + { + "epoch": 1.131721595836947, + "grad_norm": 3.9151201248168945, + "learning_rate": 1.2500362686783693e-05, + "loss": 0.0102, + "step": 10439 + }, + { + "epoch": 1.1318300086730269, + "grad_norm": 5.802786827087402, + "learning_rate": 1.2499637313216308e-05, + "loss": 0.0233, + "step": 10440 + }, + { + "epoch": 1.1319384215091066, + "grad_norm": 1.1443102359771729, + "learning_rate": 1.249891193964892e-05, + "loss": 0.0029, + "step": 10441 + }, + { + "epoch": 1.1320468343451864, + "grad_norm": 2.0255002975463867, + "learning_rate": 1.2498186566081534e-05, + "loss": 0.0164, + "step": 10442 + }, + { + "epoch": 1.1321552471812661, + "grad_norm": 1.1342777013778687, + "learning_rate": 1.2497461192514146e-05, + "loss": 0.0202, + "step": 10443 + }, + { + "epoch": 1.1322636600173461, + "grad_norm": 3.6223020553588867, + "learning_rate": 1.2496735818946759e-05, + "loss": 0.0418, + "step": 10444 + }, + { + "epoch": 1.132372072853426, + "grad_norm": 4.388412952423096, + "learning_rate": 1.2496010445379372e-05, + "loss": 0.0147, + "step": 10445 + }, + { + "epoch": 1.1324804856895057, + "grad_norm": 1.1229276657104492, + "learning_rate": 1.2495285071811985e-05, + "loss": 0.0113, + "step": 10446 + }, + { + "epoch": 1.1325888985255854, + "grad_norm": 1.7231072187423706, + "learning_rate": 1.2494559698244597e-05, + "loss": 0.0121, + "step": 10447 + }, + { + "epoch": 1.1326973113616652, + "grad_norm": 1.7711806297302246, + "learning_rate": 1.249383432467721e-05, + "loss": 0.0309, + "step": 10448 + }, + { + "epoch": 1.132805724197745, + "grad_norm": 0.15331439673900604, + "learning_rate": 1.2493108951109823e-05, + "loss": 0.0014, + "step": 10449 + }, + { + "epoch": 1.1329141370338247, + "grad_norm": 1.4910777807235718, + "learning_rate": 1.2492383577542435e-05, + "loss": 0.0199, + "step": 10450 + }, + { + "epoch": 1.1330225498699047, + "grad_norm": 0.1508302539587021, + "learning_rate": 1.2491658203975047e-05, + "loss": 0.001, + "step": 10451 + }, + { + "epoch": 1.1331309627059845, + "grad_norm": 1.5211964845657349, + "learning_rate": 1.2490932830407661e-05, + "loss": 0.0058, + "step": 10452 + }, + { + "epoch": 1.1332393755420642, + "grad_norm": 3.233747959136963, + "learning_rate": 1.2490207456840273e-05, + "loss": 0.0385, + "step": 10453 + }, + { + "epoch": 1.133347788378144, + "grad_norm": 0.13653752207756042, + "learning_rate": 1.2489482083272885e-05, + "loss": 0.0011, + "step": 10454 + }, + { + "epoch": 1.1334562012142237, + "grad_norm": 1.310117483139038, + "learning_rate": 1.2488756709705499e-05, + "loss": 0.0088, + "step": 10455 + }, + { + "epoch": 1.1335646140503035, + "grad_norm": 2.8639988899230957, + "learning_rate": 1.2488031336138111e-05, + "loss": 0.0143, + "step": 10456 + }, + { + "epoch": 1.1336730268863833, + "grad_norm": 1.7932069301605225, + "learning_rate": 1.2487305962570727e-05, + "loss": 0.0122, + "step": 10457 + }, + { + "epoch": 1.1337814397224633, + "grad_norm": 1.651576280593872, + "learning_rate": 1.2486580589003339e-05, + "loss": 0.0098, + "step": 10458 + }, + { + "epoch": 1.133889852558543, + "grad_norm": 3.288120746612549, + "learning_rate": 1.2485855215435951e-05, + "loss": 0.0174, + "step": 10459 + }, + { + "epoch": 1.1339982653946228, + "grad_norm": 2.061448574066162, + "learning_rate": 1.2485129841868565e-05, + "loss": 0.0332, + "step": 10460 + }, + { + "epoch": 1.1341066782307025, + "grad_norm": 5.49696683883667, + "learning_rate": 1.2484404468301177e-05, + "loss": 0.0351, + "step": 10461 + }, + { + "epoch": 1.1342150910667823, + "grad_norm": 2.022017478942871, + "learning_rate": 1.2483679094733789e-05, + "loss": 0.0201, + "step": 10462 + }, + { + "epoch": 1.134323503902862, + "grad_norm": 1.352828025817871, + "learning_rate": 1.2482953721166401e-05, + "loss": 0.0338, + "step": 10463 + }, + { + "epoch": 1.1344319167389418, + "grad_norm": 1.115476131439209, + "learning_rate": 1.2482228347599015e-05, + "loss": 0.0035, + "step": 10464 + }, + { + "epoch": 1.1345403295750216, + "grad_norm": 2.198533535003662, + "learning_rate": 1.2481502974031627e-05, + "loss": 0.0055, + "step": 10465 + }, + { + "epoch": 1.1346487424111014, + "grad_norm": 0.9715263843536377, + "learning_rate": 1.248077760046424e-05, + "loss": 0.0031, + "step": 10466 + }, + { + "epoch": 1.1347571552471813, + "grad_norm": 2.789468288421631, + "learning_rate": 1.2480052226896853e-05, + "loss": 0.0511, + "step": 10467 + }, + { + "epoch": 1.1348655680832611, + "grad_norm": 1.1758778095245361, + "learning_rate": 1.2479326853329465e-05, + "loss": 0.0055, + "step": 10468 + }, + { + "epoch": 1.1349739809193409, + "grad_norm": 0.06486776471138, + "learning_rate": 1.2478601479762077e-05, + "loss": 0.0008, + "step": 10469 + }, + { + "epoch": 1.1350823937554206, + "grad_norm": 3.061410903930664, + "learning_rate": 1.2477876106194691e-05, + "loss": 0.0173, + "step": 10470 + }, + { + "epoch": 1.1351908065915004, + "grad_norm": 0.12137952446937561, + "learning_rate": 1.2477150732627303e-05, + "loss": 0.0012, + "step": 10471 + }, + { + "epoch": 1.1352992194275802, + "grad_norm": 0.2553182542324066, + "learning_rate": 1.2476425359059916e-05, + "loss": 0.0023, + "step": 10472 + }, + { + "epoch": 1.13540763226366, + "grad_norm": 1.4653456211090088, + "learning_rate": 1.247569998549253e-05, + "loss": 0.0279, + "step": 10473 + }, + { + "epoch": 1.13551604509974, + "grad_norm": 3.0087966918945312, + "learning_rate": 1.2474974611925142e-05, + "loss": 0.0488, + "step": 10474 + }, + { + "epoch": 1.1356244579358197, + "grad_norm": 5.502133369445801, + "learning_rate": 1.2474249238357755e-05, + "loss": 0.021, + "step": 10475 + }, + { + "epoch": 1.1357328707718994, + "grad_norm": 12.825187683105469, + "learning_rate": 1.247352386479037e-05, + "loss": 0.0231, + "step": 10476 + }, + { + "epoch": 1.1358412836079792, + "grad_norm": 0.056013185530900955, + "learning_rate": 1.2472798491222981e-05, + "loss": 0.0005, + "step": 10477 + }, + { + "epoch": 1.135949696444059, + "grad_norm": 0.7900618314743042, + "learning_rate": 1.2472073117655594e-05, + "loss": 0.0091, + "step": 10478 + }, + { + "epoch": 1.1360581092801387, + "grad_norm": 1.3729685544967651, + "learning_rate": 1.2471347744088207e-05, + "loss": 0.0052, + "step": 10479 + }, + { + "epoch": 1.1361665221162185, + "grad_norm": 2.6317248344421387, + "learning_rate": 1.247062237052082e-05, + "loss": 0.0248, + "step": 10480 + }, + { + "epoch": 1.1362749349522985, + "grad_norm": 0.6340425610542297, + "learning_rate": 1.2469896996953432e-05, + "loss": 0.0046, + "step": 10481 + }, + { + "epoch": 1.1363833477883782, + "grad_norm": 2.460214614868164, + "learning_rate": 1.2469171623386045e-05, + "loss": 0.0313, + "step": 10482 + }, + { + "epoch": 1.136491760624458, + "grad_norm": 0.3389621078968048, + "learning_rate": 1.2468446249818658e-05, + "loss": 0.0034, + "step": 10483 + }, + { + "epoch": 1.1366001734605378, + "grad_norm": 0.12373770028352737, + "learning_rate": 1.246772087625127e-05, + "loss": 0.001, + "step": 10484 + }, + { + "epoch": 1.1367085862966175, + "grad_norm": 2.3107330799102783, + "learning_rate": 1.2466995502683884e-05, + "loss": 0.0109, + "step": 10485 + }, + { + "epoch": 1.1368169991326973, + "grad_norm": 4.688377857208252, + "learning_rate": 1.2466270129116496e-05, + "loss": 0.0654, + "step": 10486 + }, + { + "epoch": 1.136925411968777, + "grad_norm": 1.3869283199310303, + "learning_rate": 1.2465544755549108e-05, + "loss": 0.0237, + "step": 10487 + }, + { + "epoch": 1.1370338248048568, + "grad_norm": 0.3116285502910614, + "learning_rate": 1.2464819381981722e-05, + "loss": 0.0024, + "step": 10488 + }, + { + "epoch": 1.1371422376409366, + "grad_norm": 9.641010284423828, + "learning_rate": 1.2464094008414334e-05, + "loss": 0.0271, + "step": 10489 + }, + { + "epoch": 1.1372506504770166, + "grad_norm": 2.4640023708343506, + "learning_rate": 1.2463368634846946e-05, + "loss": 0.0114, + "step": 10490 + }, + { + "epoch": 1.1373590633130963, + "grad_norm": 0.2548443377017975, + "learning_rate": 1.2462643261279558e-05, + "loss": 0.0024, + "step": 10491 + }, + { + "epoch": 1.137467476149176, + "grad_norm": 0.08828099071979523, + "learning_rate": 1.2461917887712174e-05, + "loss": 0.0007, + "step": 10492 + }, + { + "epoch": 1.1375758889852559, + "grad_norm": 5.940776824951172, + "learning_rate": 1.2461192514144786e-05, + "loss": 0.052, + "step": 10493 + }, + { + "epoch": 1.1376843018213356, + "grad_norm": 1.4086817502975464, + "learning_rate": 1.24604671405774e-05, + "loss": 0.0164, + "step": 10494 + }, + { + "epoch": 1.1377927146574154, + "grad_norm": 0.9456499814987183, + "learning_rate": 1.2459741767010012e-05, + "loss": 0.0063, + "step": 10495 + }, + { + "epoch": 1.1379011274934951, + "grad_norm": 2.873215436935425, + "learning_rate": 1.2459016393442624e-05, + "loss": 0.0134, + "step": 10496 + }, + { + "epoch": 1.1380095403295751, + "grad_norm": 3.5465950965881348, + "learning_rate": 1.2458291019875238e-05, + "loss": 0.0106, + "step": 10497 + }, + { + "epoch": 1.138117953165655, + "grad_norm": 5.6985626220703125, + "learning_rate": 1.245756564630785e-05, + "loss": 0.0373, + "step": 10498 + }, + { + "epoch": 1.1382263660017347, + "grad_norm": 2.148146867752075, + "learning_rate": 1.2456840272740462e-05, + "loss": 0.0379, + "step": 10499 + }, + { + "epoch": 1.1383347788378144, + "grad_norm": 3.130943775177002, + "learning_rate": 1.2456114899173076e-05, + "loss": 0.0231, + "step": 10500 + }, + { + "epoch": 1.1384431916738942, + "grad_norm": 0.6957602500915527, + "learning_rate": 1.2455389525605688e-05, + "loss": 0.003, + "step": 10501 + }, + { + "epoch": 1.138551604509974, + "grad_norm": 3.75000262260437, + "learning_rate": 1.24546641520383e-05, + "loss": 0.0269, + "step": 10502 + }, + { + "epoch": 1.1386600173460537, + "grad_norm": 2.882845640182495, + "learning_rate": 1.2453938778470912e-05, + "loss": 0.0176, + "step": 10503 + }, + { + "epoch": 1.1387684301821335, + "grad_norm": 3.09200119972229, + "learning_rate": 1.2453213404903526e-05, + "loss": 0.0526, + "step": 10504 + }, + { + "epoch": 1.1388768430182132, + "grad_norm": 0.43793201446533203, + "learning_rate": 1.2452488031336138e-05, + "loss": 0.0019, + "step": 10505 + }, + { + "epoch": 1.1389852558542932, + "grad_norm": 0.12523511052131653, + "learning_rate": 1.245176265776875e-05, + "loss": 0.001, + "step": 10506 + }, + { + "epoch": 1.139093668690373, + "grad_norm": 2.7970805168151855, + "learning_rate": 1.2451037284201364e-05, + "loss": 0.0444, + "step": 10507 + }, + { + "epoch": 1.1392020815264527, + "grad_norm": 4.8501129150390625, + "learning_rate": 1.2450311910633976e-05, + "loss": 0.0235, + "step": 10508 + }, + { + "epoch": 1.1393104943625325, + "grad_norm": 2.493410110473633, + "learning_rate": 1.2449586537066592e-05, + "loss": 0.0158, + "step": 10509 + }, + { + "epoch": 1.1394189071986123, + "grad_norm": 3.022639751434326, + "learning_rate": 1.2448861163499204e-05, + "loss": 0.0259, + "step": 10510 + }, + { + "epoch": 1.139527320034692, + "grad_norm": 1.5737484693527222, + "learning_rate": 1.2448135789931816e-05, + "loss": 0.0105, + "step": 10511 + }, + { + "epoch": 1.1396357328707718, + "grad_norm": 0.07132431864738464, + "learning_rate": 1.244741041636443e-05, + "loss": 0.0007, + "step": 10512 + }, + { + "epoch": 1.1397441457068518, + "grad_norm": 5.764310359954834, + "learning_rate": 1.2446685042797042e-05, + "loss": 0.0238, + "step": 10513 + }, + { + "epoch": 1.1398525585429315, + "grad_norm": 1.3518571853637695, + "learning_rate": 1.2445959669229654e-05, + "loss": 0.0113, + "step": 10514 + }, + { + "epoch": 1.1399609713790113, + "grad_norm": 0.044961679726839066, + "learning_rate": 1.2445234295662267e-05, + "loss": 0.0006, + "step": 10515 + }, + { + "epoch": 1.140069384215091, + "grad_norm": 3.3136913776397705, + "learning_rate": 1.244450892209488e-05, + "loss": 0.0379, + "step": 10516 + }, + { + "epoch": 1.1401777970511708, + "grad_norm": 6.0540361404418945, + "learning_rate": 1.2443783548527493e-05, + "loss": 0.0731, + "step": 10517 + }, + { + "epoch": 1.1402862098872506, + "grad_norm": 0.32061606645584106, + "learning_rate": 1.2443058174960105e-05, + "loss": 0.0022, + "step": 10518 + }, + { + "epoch": 1.1403946227233304, + "grad_norm": 2.930694818496704, + "learning_rate": 1.2442332801392719e-05, + "loss": 0.0286, + "step": 10519 + }, + { + "epoch": 1.1405030355594103, + "grad_norm": 0.41300439834594727, + "learning_rate": 1.244160742782533e-05, + "loss": 0.0038, + "step": 10520 + }, + { + "epoch": 1.1406114483954901, + "grad_norm": 4.783082008361816, + "learning_rate": 1.2440882054257943e-05, + "loss": 0.0318, + "step": 10521 + }, + { + "epoch": 1.1407198612315699, + "grad_norm": 4.417159080505371, + "learning_rate": 1.2440156680690557e-05, + "loss": 0.096, + "step": 10522 + }, + { + "epoch": 1.1408282740676496, + "grad_norm": 1.3172262907028198, + "learning_rate": 1.2439431307123169e-05, + "loss": 0.0069, + "step": 10523 + }, + { + "epoch": 1.1409366869037294, + "grad_norm": 1.8009284734725952, + "learning_rate": 1.2438705933555781e-05, + "loss": 0.0465, + "step": 10524 + }, + { + "epoch": 1.1410450997398092, + "grad_norm": 5.508368015289307, + "learning_rate": 1.2437980559988395e-05, + "loss": 0.0782, + "step": 10525 + }, + { + "epoch": 1.141153512575889, + "grad_norm": 2.0798630714416504, + "learning_rate": 1.2437255186421009e-05, + "loss": 0.0262, + "step": 10526 + }, + { + "epoch": 1.1412619254119687, + "grad_norm": 2.9865715503692627, + "learning_rate": 1.2436529812853622e-05, + "loss": 0.0073, + "step": 10527 + }, + { + "epoch": 1.1413703382480485, + "grad_norm": 7.658822059631348, + "learning_rate": 1.2435804439286235e-05, + "loss": 0.0288, + "step": 10528 + }, + { + "epoch": 1.1414787510841284, + "grad_norm": 0.7282237410545349, + "learning_rate": 1.2435079065718847e-05, + "loss": 0.0051, + "step": 10529 + }, + { + "epoch": 1.1415871639202082, + "grad_norm": 3.419572353363037, + "learning_rate": 1.2434353692151459e-05, + "loss": 0.0372, + "step": 10530 + }, + { + "epoch": 1.141695576756288, + "grad_norm": 0.6220067143440247, + "learning_rate": 1.2433628318584073e-05, + "loss": 0.0045, + "step": 10531 + }, + { + "epoch": 1.1418039895923677, + "grad_norm": 2.369121551513672, + "learning_rate": 1.2432902945016685e-05, + "loss": 0.0078, + "step": 10532 + }, + { + "epoch": 1.1419124024284475, + "grad_norm": 6.324095726013184, + "learning_rate": 1.2432177571449297e-05, + "loss": 0.0162, + "step": 10533 + }, + { + "epoch": 1.1420208152645273, + "grad_norm": 8.172477722167969, + "learning_rate": 1.243145219788191e-05, + "loss": 0.0351, + "step": 10534 + }, + { + "epoch": 1.142129228100607, + "grad_norm": 0.5131189227104187, + "learning_rate": 1.2430726824314523e-05, + "loss": 0.0037, + "step": 10535 + }, + { + "epoch": 1.142237640936687, + "grad_norm": 7.541801452636719, + "learning_rate": 1.2430001450747135e-05, + "loss": 0.038, + "step": 10536 + }, + { + "epoch": 1.1423460537727668, + "grad_norm": 7.022790908813477, + "learning_rate": 1.2429276077179749e-05, + "loss": 0.0273, + "step": 10537 + }, + { + "epoch": 1.1424544666088465, + "grad_norm": 2.2276816368103027, + "learning_rate": 1.2428550703612361e-05, + "loss": 0.0439, + "step": 10538 + }, + { + "epoch": 1.1425628794449263, + "grad_norm": 0.8088498115539551, + "learning_rate": 1.2427825330044973e-05, + "loss": 0.0049, + "step": 10539 + }, + { + "epoch": 1.142671292281006, + "grad_norm": 2.7064645290374756, + "learning_rate": 1.2427099956477587e-05, + "loss": 0.0554, + "step": 10540 + }, + { + "epoch": 1.1427797051170858, + "grad_norm": 8.87732219696045, + "learning_rate": 1.24263745829102e-05, + "loss": 0.0468, + "step": 10541 + }, + { + "epoch": 1.1428881179531656, + "grad_norm": 6.618129730224609, + "learning_rate": 1.2425649209342811e-05, + "loss": 0.0227, + "step": 10542 + }, + { + "epoch": 1.1429965307892456, + "grad_norm": 12.779276847839355, + "learning_rate": 1.2424923835775424e-05, + "loss": 0.0424, + "step": 10543 + }, + { + "epoch": 1.1431049436253253, + "grad_norm": 1.0685380697250366, + "learning_rate": 1.2424198462208039e-05, + "loss": 0.0065, + "step": 10544 + }, + { + "epoch": 1.143213356461405, + "grad_norm": 2.8401083946228027, + "learning_rate": 1.2423473088640651e-05, + "loss": 0.0398, + "step": 10545 + }, + { + "epoch": 1.1433217692974849, + "grad_norm": 3.8100156784057617, + "learning_rate": 1.2422747715073265e-05, + "loss": 0.0113, + "step": 10546 + }, + { + "epoch": 1.1434301821335646, + "grad_norm": 2.63846755027771, + "learning_rate": 1.2422022341505877e-05, + "loss": 0.0214, + "step": 10547 + }, + { + "epoch": 1.1435385949696444, + "grad_norm": 3.378232002258301, + "learning_rate": 1.242129696793849e-05, + "loss": 0.0368, + "step": 10548 + }, + { + "epoch": 1.1436470078057241, + "grad_norm": 0.8743346333503723, + "learning_rate": 1.2420571594371103e-05, + "loss": 0.0033, + "step": 10549 + }, + { + "epoch": 1.143755420641804, + "grad_norm": 5.144850254058838, + "learning_rate": 1.2419846220803715e-05, + "loss": 0.0402, + "step": 10550 + }, + { + "epoch": 1.1438638334778837, + "grad_norm": 0.5597625374794006, + "learning_rate": 1.2419120847236327e-05, + "loss": 0.0058, + "step": 10551 + }, + { + "epoch": 1.1439722463139637, + "grad_norm": 5.679676532745361, + "learning_rate": 1.2418395473668941e-05, + "loss": 0.0428, + "step": 10552 + }, + { + "epoch": 1.1440806591500434, + "grad_norm": 4.176504135131836, + "learning_rate": 1.2417670100101553e-05, + "loss": 0.044, + "step": 10553 + }, + { + "epoch": 1.1441890719861232, + "grad_norm": 2.3692634105682373, + "learning_rate": 1.2416944726534166e-05, + "loss": 0.038, + "step": 10554 + }, + { + "epoch": 1.144297484822203, + "grad_norm": 0.13885198533535004, + "learning_rate": 1.2416219352966778e-05, + "loss": 0.001, + "step": 10555 + }, + { + "epoch": 1.1444058976582827, + "grad_norm": 1.7397737503051758, + "learning_rate": 1.2415493979399392e-05, + "loss": 0.0237, + "step": 10556 + }, + { + "epoch": 1.1445143104943625, + "grad_norm": 0.4437062442302704, + "learning_rate": 1.2414768605832004e-05, + "loss": 0.0047, + "step": 10557 + }, + { + "epoch": 1.1446227233304422, + "grad_norm": 1.8645254373550415, + "learning_rate": 1.2414043232264616e-05, + "loss": 0.016, + "step": 10558 + }, + { + "epoch": 1.1447311361665222, + "grad_norm": 0.3177691400051117, + "learning_rate": 1.241331785869723e-05, + "loss": 0.0032, + "step": 10559 + }, + { + "epoch": 1.144839549002602, + "grad_norm": 5.898500919342041, + "learning_rate": 1.2412592485129842e-05, + "loss": 0.0394, + "step": 10560 + }, + { + "epoch": 1.1449479618386817, + "grad_norm": 1.1564921140670776, + "learning_rate": 1.2411867111562457e-05, + "loss": 0.0103, + "step": 10561 + }, + { + "epoch": 1.1450563746747615, + "grad_norm": 3.1622819900512695, + "learning_rate": 1.241114173799507e-05, + "loss": 0.0441, + "step": 10562 + }, + { + "epoch": 1.1451647875108413, + "grad_norm": 4.671721458435059, + "learning_rate": 1.2410416364427682e-05, + "loss": 0.0423, + "step": 10563 + }, + { + "epoch": 1.145273200346921, + "grad_norm": 2.8934895992279053, + "learning_rate": 1.2409690990860295e-05, + "loss": 0.0297, + "step": 10564 + }, + { + "epoch": 1.1453816131830008, + "grad_norm": 2.919051170349121, + "learning_rate": 1.2408965617292908e-05, + "loss": 0.0246, + "step": 10565 + }, + { + "epoch": 1.1454900260190806, + "grad_norm": 2.2072227001190186, + "learning_rate": 1.240824024372552e-05, + "loss": 0.0364, + "step": 10566 + }, + { + "epoch": 1.1455984388551603, + "grad_norm": 0.3474150002002716, + "learning_rate": 1.2407514870158132e-05, + "loss": 0.0036, + "step": 10567 + }, + { + "epoch": 1.1457068516912403, + "grad_norm": 6.581076622009277, + "learning_rate": 1.2406789496590746e-05, + "loss": 0.0614, + "step": 10568 + }, + { + "epoch": 1.14581526452732, + "grad_norm": 5.527686595916748, + "learning_rate": 1.2406064123023358e-05, + "loss": 0.0405, + "step": 10569 + }, + { + "epoch": 1.1459236773633998, + "grad_norm": 4.439646244049072, + "learning_rate": 1.240533874945597e-05, + "loss": 0.0155, + "step": 10570 + }, + { + "epoch": 1.1460320901994796, + "grad_norm": 2.813629388809204, + "learning_rate": 1.2404613375888584e-05, + "loss": 0.0333, + "step": 10571 + }, + { + "epoch": 1.1461405030355594, + "grad_norm": 6.696311950683594, + "learning_rate": 1.2403888002321196e-05, + "loss": 0.0496, + "step": 10572 + }, + { + "epoch": 1.1462489158716391, + "grad_norm": 1.4801316261291504, + "learning_rate": 1.2403162628753808e-05, + "loss": 0.0058, + "step": 10573 + }, + { + "epoch": 1.146357328707719, + "grad_norm": 0.19377732276916504, + "learning_rate": 1.2402437255186422e-05, + "loss": 0.0021, + "step": 10574 + }, + { + "epoch": 1.1464657415437989, + "grad_norm": 0.3294000029563904, + "learning_rate": 1.2401711881619034e-05, + "loss": 0.0019, + "step": 10575 + }, + { + "epoch": 1.1465741543798786, + "grad_norm": 3.4490702152252197, + "learning_rate": 1.2400986508051646e-05, + "loss": 0.0212, + "step": 10576 + }, + { + "epoch": 1.1466825672159584, + "grad_norm": 1.2702746391296387, + "learning_rate": 1.240026113448426e-05, + "loss": 0.0196, + "step": 10577 + }, + { + "epoch": 1.1467909800520382, + "grad_norm": 1.8990757465362549, + "learning_rate": 1.2399535760916874e-05, + "loss": 0.0491, + "step": 10578 + }, + { + "epoch": 1.146899392888118, + "grad_norm": 3.781750202178955, + "learning_rate": 1.2398810387349488e-05, + "loss": 0.022, + "step": 10579 + }, + { + "epoch": 1.1470078057241977, + "grad_norm": 0.09059783071279526, + "learning_rate": 1.23980850137821e-05, + "loss": 0.0013, + "step": 10580 + }, + { + "epoch": 1.1471162185602775, + "grad_norm": 12.430246353149414, + "learning_rate": 1.2397359640214712e-05, + "loss": 0.0332, + "step": 10581 + }, + { + "epoch": 1.1472246313963574, + "grad_norm": 0.950738251209259, + "learning_rate": 1.2396634266647324e-05, + "loss": 0.0064, + "step": 10582 + }, + { + "epoch": 1.1473330442324372, + "grad_norm": 1.6782125234603882, + "learning_rate": 1.2395908893079938e-05, + "loss": 0.0336, + "step": 10583 + }, + { + "epoch": 1.147441457068517, + "grad_norm": 2.9093050956726074, + "learning_rate": 1.239518351951255e-05, + "loss": 0.0206, + "step": 10584 + }, + { + "epoch": 1.1475498699045967, + "grad_norm": 2.481363296508789, + "learning_rate": 1.2394458145945162e-05, + "loss": 0.0246, + "step": 10585 + }, + { + "epoch": 1.1476582827406765, + "grad_norm": 1.6322767734527588, + "learning_rate": 1.2393732772377776e-05, + "loss": 0.012, + "step": 10586 + }, + { + "epoch": 1.1477666955767563, + "grad_norm": 2.0267772674560547, + "learning_rate": 1.2393007398810388e-05, + "loss": 0.0176, + "step": 10587 + }, + { + "epoch": 1.147875108412836, + "grad_norm": 0.16055409610271454, + "learning_rate": 1.2392282025243e-05, + "loss": 0.0015, + "step": 10588 + }, + { + "epoch": 1.1479835212489158, + "grad_norm": 3.705143451690674, + "learning_rate": 1.2391556651675614e-05, + "loss": 0.0178, + "step": 10589 + }, + { + "epoch": 1.1480919340849955, + "grad_norm": 2.6657943725585938, + "learning_rate": 1.2390831278108226e-05, + "loss": 0.0293, + "step": 10590 + }, + { + "epoch": 1.1482003469210755, + "grad_norm": 3.7416012287139893, + "learning_rate": 1.2390105904540839e-05, + "loss": 0.0161, + "step": 10591 + }, + { + "epoch": 1.1483087597571553, + "grad_norm": 3.979271173477173, + "learning_rate": 1.2389380530973452e-05, + "loss": 0.0264, + "step": 10592 + }, + { + "epoch": 1.148417172593235, + "grad_norm": 5.258120059967041, + "learning_rate": 1.2388655157406065e-05, + "loss": 0.0244, + "step": 10593 + }, + { + "epoch": 1.1485255854293148, + "grad_norm": 2.722315549850464, + "learning_rate": 1.2387929783838677e-05, + "loss": 0.0408, + "step": 10594 + }, + { + "epoch": 1.1486339982653946, + "grad_norm": 4.331416130065918, + "learning_rate": 1.2387204410271292e-05, + "loss": 0.0491, + "step": 10595 + }, + { + "epoch": 1.1487424111014743, + "grad_norm": 2.0005874633789062, + "learning_rate": 1.2386479036703904e-05, + "loss": 0.0226, + "step": 10596 + }, + { + "epoch": 1.148850823937554, + "grad_norm": 0.660638153553009, + "learning_rate": 1.2385753663136517e-05, + "loss": 0.0048, + "step": 10597 + }, + { + "epoch": 1.148959236773634, + "grad_norm": 1.3886324167251587, + "learning_rate": 1.238502828956913e-05, + "loss": 0.0037, + "step": 10598 + }, + { + "epoch": 1.1490676496097139, + "grad_norm": 3.8301169872283936, + "learning_rate": 1.2384302916001743e-05, + "loss": 0.0089, + "step": 10599 + }, + { + "epoch": 1.1491760624457936, + "grad_norm": 4.694914817810059, + "learning_rate": 1.2383577542434355e-05, + "loss": 0.0375, + "step": 10600 + }, + { + "epoch": 1.1492844752818734, + "grad_norm": 2.306316375732422, + "learning_rate": 1.2382852168866968e-05, + "loss": 0.0094, + "step": 10601 + }, + { + "epoch": 1.1493928881179531, + "grad_norm": 0.23366442322731018, + "learning_rate": 1.238212679529958e-05, + "loss": 0.0016, + "step": 10602 + }, + { + "epoch": 1.149501300954033, + "grad_norm": 1.1701046228408813, + "learning_rate": 1.2381401421732193e-05, + "loss": 0.0233, + "step": 10603 + }, + { + "epoch": 1.1496097137901127, + "grad_norm": 2.2087783813476562, + "learning_rate": 1.2380676048164807e-05, + "loss": 0.0392, + "step": 10604 + }, + { + "epoch": 1.1497181266261927, + "grad_norm": 4.669911861419678, + "learning_rate": 1.2379950674597419e-05, + "loss": 0.016, + "step": 10605 + }, + { + "epoch": 1.1498265394622724, + "grad_norm": 1.4044429063796997, + "learning_rate": 1.2379225301030031e-05, + "loss": 0.0156, + "step": 10606 + }, + { + "epoch": 1.1499349522983522, + "grad_norm": 1.757011890411377, + "learning_rate": 1.2378499927462643e-05, + "loss": 0.034, + "step": 10607 + }, + { + "epoch": 1.150043365134432, + "grad_norm": 3.17077374458313, + "learning_rate": 1.2377774553895257e-05, + "loss": 0.0353, + "step": 10608 + }, + { + "epoch": 1.1501517779705117, + "grad_norm": 0.9660161137580872, + "learning_rate": 1.2377049180327869e-05, + "loss": 0.0075, + "step": 10609 + }, + { + "epoch": 1.1502601908065915, + "grad_norm": 1.2415030002593994, + "learning_rate": 1.2376323806760481e-05, + "loss": 0.0167, + "step": 10610 + }, + { + "epoch": 1.1503686036426712, + "grad_norm": 1.357675552368164, + "learning_rate": 1.2375598433193095e-05, + "loss": 0.0144, + "step": 10611 + }, + { + "epoch": 1.150477016478751, + "grad_norm": 1.6037304401397705, + "learning_rate": 1.2374873059625707e-05, + "loss": 0.011, + "step": 10612 + }, + { + "epoch": 1.1505854293148308, + "grad_norm": 4.964379787445068, + "learning_rate": 1.2374147686058323e-05, + "loss": 0.0222, + "step": 10613 + }, + { + "epoch": 1.1506938421509108, + "grad_norm": 2.9494144916534424, + "learning_rate": 1.2373422312490935e-05, + "loss": 0.0083, + "step": 10614 + }, + { + "epoch": 1.1508022549869905, + "grad_norm": 0.22076618671417236, + "learning_rate": 1.2372696938923547e-05, + "loss": 0.0015, + "step": 10615 + }, + { + "epoch": 1.1509106678230703, + "grad_norm": 3.0717146396636963, + "learning_rate": 1.237197156535616e-05, + "loss": 0.017, + "step": 10616 + }, + { + "epoch": 1.15101908065915, + "grad_norm": 1.622320294380188, + "learning_rate": 1.2371246191788773e-05, + "loss": 0.0066, + "step": 10617 + }, + { + "epoch": 1.1511274934952298, + "grad_norm": 0.7066470980644226, + "learning_rate": 1.2370520818221385e-05, + "loss": 0.0069, + "step": 10618 + }, + { + "epoch": 1.1512359063313096, + "grad_norm": 2.539029121398926, + "learning_rate": 1.2369795444653997e-05, + "loss": 0.0093, + "step": 10619 + }, + { + "epoch": 1.1513443191673893, + "grad_norm": 2.3437578678131104, + "learning_rate": 1.2369070071086611e-05, + "loss": 0.0199, + "step": 10620 + }, + { + "epoch": 1.1514527320034693, + "grad_norm": 1.2366231679916382, + "learning_rate": 1.2368344697519223e-05, + "loss": 0.0229, + "step": 10621 + }, + { + "epoch": 1.151561144839549, + "grad_norm": 1.482199788093567, + "learning_rate": 1.2367619323951835e-05, + "loss": 0.0127, + "step": 10622 + }, + { + "epoch": 1.1516695576756288, + "grad_norm": 4.686888217926025, + "learning_rate": 1.236689395038445e-05, + "loss": 0.0419, + "step": 10623 + }, + { + "epoch": 1.1517779705117086, + "grad_norm": 2.157151222229004, + "learning_rate": 1.2366168576817061e-05, + "loss": 0.0445, + "step": 10624 + }, + { + "epoch": 1.1518863833477884, + "grad_norm": 1.5888174772262573, + "learning_rate": 1.2365443203249673e-05, + "loss": 0.0275, + "step": 10625 + }, + { + "epoch": 1.1519947961838681, + "grad_norm": 10.978565216064453, + "learning_rate": 1.2364717829682287e-05, + "loss": 0.0527, + "step": 10626 + }, + { + "epoch": 1.152103209019948, + "grad_norm": 1.0315221548080444, + "learning_rate": 1.23639924561149e-05, + "loss": 0.01, + "step": 10627 + }, + { + "epoch": 1.1522116218560277, + "grad_norm": 5.939178466796875, + "learning_rate": 1.2363267082547512e-05, + "loss": 0.0359, + "step": 10628 + }, + { + "epoch": 1.1523200346921076, + "grad_norm": 3.675218105316162, + "learning_rate": 1.2362541708980125e-05, + "loss": 0.0573, + "step": 10629 + }, + { + "epoch": 1.1524284475281874, + "grad_norm": 1.3978179693222046, + "learning_rate": 1.236181633541274e-05, + "loss": 0.0135, + "step": 10630 + }, + { + "epoch": 1.1525368603642672, + "grad_norm": 4.6784281730651855, + "learning_rate": 1.2361090961845353e-05, + "loss": 0.0158, + "step": 10631 + }, + { + "epoch": 1.152645273200347, + "grad_norm": 0.1494358777999878, + "learning_rate": 1.2360365588277965e-05, + "loss": 0.0013, + "step": 10632 + }, + { + "epoch": 1.1527536860364267, + "grad_norm": 1.8012932538986206, + "learning_rate": 1.2359640214710577e-05, + "loss": 0.0278, + "step": 10633 + }, + { + "epoch": 1.1528620988725065, + "grad_norm": 4.155951023101807, + "learning_rate": 1.235891484114319e-05, + "loss": 0.0296, + "step": 10634 + }, + { + "epoch": 1.1529705117085862, + "grad_norm": 2.976898670196533, + "learning_rate": 1.2358189467575803e-05, + "loss": 0.0173, + "step": 10635 + }, + { + "epoch": 1.153078924544666, + "grad_norm": 2.188105583190918, + "learning_rate": 1.2357464094008416e-05, + "loss": 0.0141, + "step": 10636 + }, + { + "epoch": 1.153187337380746, + "grad_norm": 0.9634979963302612, + "learning_rate": 1.2356738720441028e-05, + "loss": 0.0026, + "step": 10637 + }, + { + "epoch": 1.1532957502168257, + "grad_norm": 3.0373289585113525, + "learning_rate": 1.2356013346873642e-05, + "loss": 0.0278, + "step": 10638 + }, + { + "epoch": 1.1534041630529055, + "grad_norm": 2.091172933578491, + "learning_rate": 1.2355287973306254e-05, + "loss": 0.0305, + "step": 10639 + }, + { + "epoch": 1.1535125758889853, + "grad_norm": 0.7407464981079102, + "learning_rate": 1.2354562599738866e-05, + "loss": 0.0131, + "step": 10640 + }, + { + "epoch": 1.153620988725065, + "grad_norm": 0.26332664489746094, + "learning_rate": 1.235383722617148e-05, + "loss": 0.0019, + "step": 10641 + }, + { + "epoch": 1.1537294015611448, + "grad_norm": 1.3483699560165405, + "learning_rate": 1.2353111852604092e-05, + "loss": 0.0234, + "step": 10642 + }, + { + "epoch": 1.1538378143972245, + "grad_norm": 8.522867202758789, + "learning_rate": 1.2352386479036704e-05, + "loss": 0.0773, + "step": 10643 + }, + { + "epoch": 1.1539462272333045, + "grad_norm": 5.749290943145752, + "learning_rate": 1.2351661105469318e-05, + "loss": 0.0237, + "step": 10644 + }, + { + "epoch": 1.1540546400693843, + "grad_norm": 0.11390066146850586, + "learning_rate": 1.235093573190193e-05, + "loss": 0.001, + "step": 10645 + }, + { + "epoch": 1.154163052905464, + "grad_norm": 2.0122838020324707, + "learning_rate": 1.2350210358334542e-05, + "loss": 0.0174, + "step": 10646 + }, + { + "epoch": 1.1542714657415438, + "grad_norm": 1.114821195602417, + "learning_rate": 1.2349484984767158e-05, + "loss": 0.0139, + "step": 10647 + }, + { + "epoch": 1.1543798785776236, + "grad_norm": 1.4995425939559937, + "learning_rate": 1.234875961119977e-05, + "loss": 0.0119, + "step": 10648 + }, + { + "epoch": 1.1544882914137033, + "grad_norm": 2.271369695663452, + "learning_rate": 1.2348034237632382e-05, + "loss": 0.0448, + "step": 10649 + }, + { + "epoch": 1.1545967042497831, + "grad_norm": 4.873913764953613, + "learning_rate": 1.2347308864064996e-05, + "loss": 0.0453, + "step": 10650 + }, + { + "epoch": 1.1547051170858629, + "grad_norm": 1.0955698490142822, + "learning_rate": 1.2346583490497608e-05, + "loss": 0.0131, + "step": 10651 + }, + { + "epoch": 1.1548135299219426, + "grad_norm": 1.262264370918274, + "learning_rate": 1.234585811693022e-05, + "loss": 0.0103, + "step": 10652 + }, + { + "epoch": 1.1549219427580226, + "grad_norm": 1.2340686321258545, + "learning_rate": 1.2345132743362834e-05, + "loss": 0.0099, + "step": 10653 + }, + { + "epoch": 1.1550303555941024, + "grad_norm": 2.378216505050659, + "learning_rate": 1.2344407369795446e-05, + "loss": 0.0482, + "step": 10654 + }, + { + "epoch": 1.1551387684301822, + "grad_norm": 0.9072058200836182, + "learning_rate": 1.2343681996228058e-05, + "loss": 0.0126, + "step": 10655 + }, + { + "epoch": 1.155247181266262, + "grad_norm": 1.4891494512557983, + "learning_rate": 1.2342956622660672e-05, + "loss": 0.0104, + "step": 10656 + }, + { + "epoch": 1.1553555941023417, + "grad_norm": 4.212276458740234, + "learning_rate": 1.2342231249093284e-05, + "loss": 0.0145, + "step": 10657 + }, + { + "epoch": 1.1554640069384214, + "grad_norm": 4.962377548217773, + "learning_rate": 1.2341505875525896e-05, + "loss": 0.0633, + "step": 10658 + }, + { + "epoch": 1.1555724197745012, + "grad_norm": 0.2760832607746124, + "learning_rate": 1.2340780501958508e-05, + "loss": 0.0026, + "step": 10659 + }, + { + "epoch": 1.1556808326105812, + "grad_norm": 0.41949889063835144, + "learning_rate": 1.2340055128391122e-05, + "loss": 0.0055, + "step": 10660 + }, + { + "epoch": 1.155789245446661, + "grad_norm": 0.17083154618740082, + "learning_rate": 1.2339329754823734e-05, + "loss": 0.0018, + "step": 10661 + }, + { + "epoch": 1.1558976582827407, + "grad_norm": 3.2282750606536865, + "learning_rate": 1.2338604381256346e-05, + "loss": 0.0415, + "step": 10662 + }, + { + "epoch": 1.1560060711188205, + "grad_norm": 4.070411205291748, + "learning_rate": 1.233787900768896e-05, + "loss": 0.0473, + "step": 10663 + }, + { + "epoch": 1.1561144839549002, + "grad_norm": 3.1696321964263916, + "learning_rate": 1.2337153634121572e-05, + "loss": 0.0408, + "step": 10664 + }, + { + "epoch": 1.15622289679098, + "grad_norm": 0.3756750822067261, + "learning_rate": 1.2336428260554188e-05, + "loss": 0.0026, + "step": 10665 + }, + { + "epoch": 1.1563313096270598, + "grad_norm": 1.3778599500656128, + "learning_rate": 1.23357028869868e-05, + "loss": 0.0196, + "step": 10666 + }, + { + "epoch": 1.1564397224631398, + "grad_norm": 0.904708981513977, + "learning_rate": 1.2334977513419412e-05, + "loss": 0.0132, + "step": 10667 + }, + { + "epoch": 1.1565481352992195, + "grad_norm": 1.6472934484481812, + "learning_rate": 1.2334252139852026e-05, + "loss": 0.0086, + "step": 10668 + }, + { + "epoch": 1.1566565481352993, + "grad_norm": 2.897707939147949, + "learning_rate": 1.2333526766284638e-05, + "loss": 0.0256, + "step": 10669 + }, + { + "epoch": 1.156764960971379, + "grad_norm": 8.847631454467773, + "learning_rate": 1.233280139271725e-05, + "loss": 0.0448, + "step": 10670 + }, + { + "epoch": 1.1568733738074588, + "grad_norm": 0.8423620462417603, + "learning_rate": 1.2332076019149863e-05, + "loss": 0.0024, + "step": 10671 + }, + { + "epoch": 1.1569817866435386, + "grad_norm": 4.805181503295898, + "learning_rate": 1.2331350645582476e-05, + "loss": 0.0232, + "step": 10672 + }, + { + "epoch": 1.1570901994796183, + "grad_norm": 7.322365760803223, + "learning_rate": 1.2330625272015089e-05, + "loss": 0.0663, + "step": 10673 + }, + { + "epoch": 1.157198612315698, + "grad_norm": 1.2432138919830322, + "learning_rate": 1.23298998984477e-05, + "loss": 0.0061, + "step": 10674 + }, + { + "epoch": 1.1573070251517779, + "grad_norm": 1.2346810102462769, + "learning_rate": 1.2329174524880315e-05, + "loss": 0.0081, + "step": 10675 + }, + { + "epoch": 1.1574154379878578, + "grad_norm": 1.0252699851989746, + "learning_rate": 1.2328449151312927e-05, + "loss": 0.0096, + "step": 10676 + }, + { + "epoch": 1.1575238508239376, + "grad_norm": 0.22069844603538513, + "learning_rate": 1.2327723777745539e-05, + "loss": 0.0016, + "step": 10677 + }, + { + "epoch": 1.1576322636600174, + "grad_norm": 1.1366244554519653, + "learning_rate": 1.2326998404178153e-05, + "loss": 0.0079, + "step": 10678 + }, + { + "epoch": 1.1577406764960971, + "grad_norm": 0.1904645562171936, + "learning_rate": 1.2326273030610765e-05, + "loss": 0.0012, + "step": 10679 + }, + { + "epoch": 1.157849089332177, + "grad_norm": 1.9315080642700195, + "learning_rate": 1.2325547657043377e-05, + "loss": 0.0182, + "step": 10680 + }, + { + "epoch": 1.1579575021682567, + "grad_norm": 2.4485702514648438, + "learning_rate": 1.232482228347599e-05, + "loss": 0.0207, + "step": 10681 + }, + { + "epoch": 1.1580659150043364, + "grad_norm": 0.11476253718137741, + "learning_rate": 1.2324096909908605e-05, + "loss": 0.0012, + "step": 10682 + }, + { + "epoch": 1.1581743278404164, + "grad_norm": 0.501018226146698, + "learning_rate": 1.2323371536341218e-05, + "loss": 0.0013, + "step": 10683 + }, + { + "epoch": 1.1582827406764962, + "grad_norm": 1.4625177383422852, + "learning_rate": 1.232264616277383e-05, + "loss": 0.0111, + "step": 10684 + }, + { + "epoch": 1.158391153512576, + "grad_norm": 1.6412147283554077, + "learning_rate": 1.2321920789206443e-05, + "loss": 0.0133, + "step": 10685 + }, + { + "epoch": 1.1584995663486557, + "grad_norm": 2.5339317321777344, + "learning_rate": 1.2321195415639055e-05, + "loss": 0.0472, + "step": 10686 + }, + { + "epoch": 1.1586079791847355, + "grad_norm": 4.414684295654297, + "learning_rate": 1.2320470042071669e-05, + "loss": 0.0247, + "step": 10687 + }, + { + "epoch": 1.1587163920208152, + "grad_norm": 2.7836453914642334, + "learning_rate": 1.2319744668504281e-05, + "loss": 0.0416, + "step": 10688 + }, + { + "epoch": 1.158824804856895, + "grad_norm": 5.303212642669678, + "learning_rate": 1.2319019294936893e-05, + "loss": 0.0317, + "step": 10689 + }, + { + "epoch": 1.158933217692975, + "grad_norm": 2.9861717224121094, + "learning_rate": 1.2318293921369507e-05, + "loss": 0.0257, + "step": 10690 + }, + { + "epoch": 1.1590416305290547, + "grad_norm": 9.227189064025879, + "learning_rate": 1.2317568547802119e-05, + "loss": 0.0356, + "step": 10691 + }, + { + "epoch": 1.1591500433651345, + "grad_norm": 1.4068794250488281, + "learning_rate": 1.2316843174234731e-05, + "loss": 0.0115, + "step": 10692 + }, + { + "epoch": 1.1592584562012143, + "grad_norm": 0.14560459554195404, + "learning_rate": 1.2316117800667345e-05, + "loss": 0.0007, + "step": 10693 + }, + { + "epoch": 1.159366869037294, + "grad_norm": 9.871265411376953, + "learning_rate": 1.2315392427099957e-05, + "loss": 0.0354, + "step": 10694 + }, + { + "epoch": 1.1594752818733738, + "grad_norm": 3.0117344856262207, + "learning_rate": 1.231466705353257e-05, + "loss": 0.0141, + "step": 10695 + }, + { + "epoch": 1.1595836947094535, + "grad_norm": 2.7265734672546387, + "learning_rate": 1.2313941679965183e-05, + "loss": 0.0269, + "step": 10696 + }, + { + "epoch": 1.1596921075455333, + "grad_norm": 5.679281711578369, + "learning_rate": 1.2313216306397795e-05, + "loss": 0.0138, + "step": 10697 + }, + { + "epoch": 1.159800520381613, + "grad_norm": 0.7453668713569641, + "learning_rate": 1.2312490932830407e-05, + "loss": 0.0074, + "step": 10698 + }, + { + "epoch": 1.159908933217693, + "grad_norm": 3.0735747814178467, + "learning_rate": 1.2311765559263023e-05, + "loss": 0.0081, + "step": 10699 + }, + { + "epoch": 1.1600173460537728, + "grad_norm": 0.9462940096855164, + "learning_rate": 1.2311040185695635e-05, + "loss": 0.0242, + "step": 10700 + }, + { + "epoch": 1.1601257588898526, + "grad_norm": 0.18517550826072693, + "learning_rate": 1.2310314812128247e-05, + "loss": 0.0008, + "step": 10701 + }, + { + "epoch": 1.1602341717259324, + "grad_norm": 1.0490038394927979, + "learning_rate": 1.2309589438560861e-05, + "loss": 0.0078, + "step": 10702 + }, + { + "epoch": 1.1603425845620121, + "grad_norm": 3.2226805686950684, + "learning_rate": 1.2308864064993473e-05, + "loss": 0.0901, + "step": 10703 + }, + { + "epoch": 1.1604509973980919, + "grad_norm": 7.668471813201904, + "learning_rate": 1.2308138691426085e-05, + "loss": 0.0717, + "step": 10704 + }, + { + "epoch": 1.1605594102341716, + "grad_norm": 3.7984511852264404, + "learning_rate": 1.23074133178587e-05, + "loss": 0.0203, + "step": 10705 + }, + { + "epoch": 1.1606678230702516, + "grad_norm": 4.2666730880737305, + "learning_rate": 1.2306687944291311e-05, + "loss": 0.0363, + "step": 10706 + }, + { + "epoch": 1.1607762359063314, + "grad_norm": 3.537769317626953, + "learning_rate": 1.2305962570723923e-05, + "loss": 0.05, + "step": 10707 + }, + { + "epoch": 1.1608846487424112, + "grad_norm": 4.922683238983154, + "learning_rate": 1.2305237197156537e-05, + "loss": 0.097, + "step": 10708 + }, + { + "epoch": 1.160993061578491, + "grad_norm": 1.3655755519866943, + "learning_rate": 1.230451182358915e-05, + "loss": 0.0031, + "step": 10709 + }, + { + "epoch": 1.1611014744145707, + "grad_norm": 0.2931571900844574, + "learning_rate": 1.2303786450021762e-05, + "loss": 0.0021, + "step": 10710 + }, + { + "epoch": 1.1612098872506504, + "grad_norm": 1.1709083318710327, + "learning_rate": 1.2303061076454374e-05, + "loss": 0.0035, + "step": 10711 + }, + { + "epoch": 1.1613183000867302, + "grad_norm": 0.1960158795118332, + "learning_rate": 1.2302335702886988e-05, + "loss": 0.001, + "step": 10712 + }, + { + "epoch": 1.16142671292281, + "grad_norm": 3.948054313659668, + "learning_rate": 1.23016103293196e-05, + "loss": 0.032, + "step": 10713 + }, + { + "epoch": 1.1615351257588897, + "grad_norm": 2.2712557315826416, + "learning_rate": 1.2300884955752212e-05, + "loss": 0.0215, + "step": 10714 + }, + { + "epoch": 1.1616435385949697, + "grad_norm": 1.4103881120681763, + "learning_rate": 1.2300159582184826e-05, + "loss": 0.0329, + "step": 10715 + }, + { + "epoch": 1.1617519514310495, + "grad_norm": 3.514883041381836, + "learning_rate": 1.229943420861744e-05, + "loss": 0.0188, + "step": 10716 + }, + { + "epoch": 1.1618603642671292, + "grad_norm": 4.668564796447754, + "learning_rate": 1.2298708835050053e-05, + "loss": 0.0324, + "step": 10717 + }, + { + "epoch": 1.161968777103209, + "grad_norm": 2.5757222175598145, + "learning_rate": 1.2297983461482665e-05, + "loss": 0.0085, + "step": 10718 + }, + { + "epoch": 1.1620771899392888, + "grad_norm": 0.17504912614822388, + "learning_rate": 1.2297258087915278e-05, + "loss": 0.0014, + "step": 10719 + }, + { + "epoch": 1.1621856027753685, + "grad_norm": 1.7273012399673462, + "learning_rate": 1.2296532714347891e-05, + "loss": 0.0377, + "step": 10720 + }, + { + "epoch": 1.1622940156114483, + "grad_norm": 3.8379251956939697, + "learning_rate": 1.2295807340780504e-05, + "loss": 0.0176, + "step": 10721 + }, + { + "epoch": 1.1624024284475283, + "grad_norm": 2.395920753479004, + "learning_rate": 1.2295081967213116e-05, + "loss": 0.0224, + "step": 10722 + }, + { + "epoch": 1.162510841283608, + "grad_norm": 1.5120869874954224, + "learning_rate": 1.229435659364573e-05, + "loss": 0.0086, + "step": 10723 + }, + { + "epoch": 1.1626192541196878, + "grad_norm": 5.162423610687256, + "learning_rate": 1.2293631220078342e-05, + "loss": 0.0185, + "step": 10724 + }, + { + "epoch": 1.1627276669557676, + "grad_norm": 1.3883498907089233, + "learning_rate": 1.2292905846510954e-05, + "loss": 0.0066, + "step": 10725 + }, + { + "epoch": 1.1628360797918473, + "grad_norm": 2.2028133869171143, + "learning_rate": 1.2292180472943566e-05, + "loss": 0.0125, + "step": 10726 + }, + { + "epoch": 1.162944492627927, + "grad_norm": 2.6081011295318604, + "learning_rate": 1.229145509937618e-05, + "loss": 0.0618, + "step": 10727 + }, + { + "epoch": 1.1630529054640069, + "grad_norm": 3.0108630657196045, + "learning_rate": 1.2290729725808792e-05, + "loss": 0.0249, + "step": 10728 + }, + { + "epoch": 1.1631613183000868, + "grad_norm": 0.24123787879943848, + "learning_rate": 1.2290004352241404e-05, + "loss": 0.0029, + "step": 10729 + }, + { + "epoch": 1.1632697311361666, + "grad_norm": 4.900505065917969, + "learning_rate": 1.2289278978674018e-05, + "loss": 0.047, + "step": 10730 + }, + { + "epoch": 1.1633781439722464, + "grad_norm": 1.5363402366638184, + "learning_rate": 1.228855360510663e-05, + "loss": 0.0248, + "step": 10731 + }, + { + "epoch": 1.1634865568083261, + "grad_norm": 3.093320369720459, + "learning_rate": 1.2287828231539242e-05, + "loss": 0.0718, + "step": 10732 + }, + { + "epoch": 1.163594969644406, + "grad_norm": 4.049489498138428, + "learning_rate": 1.2287102857971856e-05, + "loss": 0.0481, + "step": 10733 + }, + { + "epoch": 1.1637033824804857, + "grad_norm": 0.37174880504608154, + "learning_rate": 1.228637748440447e-05, + "loss": 0.0039, + "step": 10734 + }, + { + "epoch": 1.1638117953165654, + "grad_norm": 3.9074883460998535, + "learning_rate": 1.2285652110837084e-05, + "loss": 0.018, + "step": 10735 + }, + { + "epoch": 1.1639202081526452, + "grad_norm": 7.4912109375, + "learning_rate": 1.2284926737269696e-05, + "loss": 0.0744, + "step": 10736 + }, + { + "epoch": 1.164028620988725, + "grad_norm": 2.0629870891571045, + "learning_rate": 1.2284201363702308e-05, + "loss": 0.0388, + "step": 10737 + }, + { + "epoch": 1.164137033824805, + "grad_norm": 4.269862651824951, + "learning_rate": 1.228347599013492e-05, + "loss": 0.0754, + "step": 10738 + }, + { + "epoch": 1.1642454466608847, + "grad_norm": 1.3108526468276978, + "learning_rate": 1.2282750616567534e-05, + "loss": 0.0067, + "step": 10739 + }, + { + "epoch": 1.1643538594969645, + "grad_norm": 0.9598301649093628, + "learning_rate": 1.2282025243000146e-05, + "loss": 0.005, + "step": 10740 + }, + { + "epoch": 1.1644622723330442, + "grad_norm": 2.3309266567230225, + "learning_rate": 1.2281299869432758e-05, + "loss": 0.0388, + "step": 10741 + }, + { + "epoch": 1.164570685169124, + "grad_norm": 2.5600500106811523, + "learning_rate": 1.2280574495865372e-05, + "loss": 0.0846, + "step": 10742 + }, + { + "epoch": 1.1646790980052038, + "grad_norm": 0.810566782951355, + "learning_rate": 1.2279849122297984e-05, + "loss": 0.0108, + "step": 10743 + }, + { + "epoch": 1.1647875108412835, + "grad_norm": 2.580857753753662, + "learning_rate": 1.2279123748730596e-05, + "loss": 0.0116, + "step": 10744 + }, + { + "epoch": 1.1648959236773635, + "grad_norm": 1.1908650398254395, + "learning_rate": 1.227839837516321e-05, + "loss": 0.012, + "step": 10745 + }, + { + "epoch": 1.1650043365134433, + "grad_norm": 3.395934581756592, + "learning_rate": 1.2277673001595822e-05, + "loss": 0.0559, + "step": 10746 + }, + { + "epoch": 1.165112749349523, + "grad_norm": 0.9373562335968018, + "learning_rate": 1.2276947628028435e-05, + "loss": 0.0092, + "step": 10747 + }, + { + "epoch": 1.1652211621856028, + "grad_norm": 1.548380732536316, + "learning_rate": 1.2276222254461048e-05, + "loss": 0.0232, + "step": 10748 + }, + { + "epoch": 1.1653295750216826, + "grad_norm": 2.268845319747925, + "learning_rate": 1.227549688089366e-05, + "loss": 0.0232, + "step": 10749 + }, + { + "epoch": 1.1654379878577623, + "grad_norm": 0.713431715965271, + "learning_rate": 1.2274771507326273e-05, + "loss": 0.0033, + "step": 10750 + }, + { + "epoch": 1.165546400693842, + "grad_norm": 1.0240733623504639, + "learning_rate": 1.2274046133758888e-05, + "loss": 0.0114, + "step": 10751 + }, + { + "epoch": 1.165654813529922, + "grad_norm": 0.2666237950325012, + "learning_rate": 1.22733207601915e-05, + "loss": 0.0032, + "step": 10752 + }, + { + "epoch": 1.1657632263660018, + "grad_norm": 2.307831048965454, + "learning_rate": 1.2272595386624113e-05, + "loss": 0.0157, + "step": 10753 + }, + { + "epoch": 1.1658716392020816, + "grad_norm": 2.6865859031677246, + "learning_rate": 1.2271870013056726e-05, + "loss": 0.014, + "step": 10754 + }, + { + "epoch": 1.1659800520381614, + "grad_norm": 0.7404294013977051, + "learning_rate": 1.2271144639489339e-05, + "loss": 0.0183, + "step": 10755 + }, + { + "epoch": 1.1660884648742411, + "grad_norm": 0.475744366645813, + "learning_rate": 1.227041926592195e-05, + "loss": 0.0032, + "step": 10756 + }, + { + "epoch": 1.1661968777103209, + "grad_norm": 1.5064566135406494, + "learning_rate": 1.2269693892354564e-05, + "loss": 0.0049, + "step": 10757 + }, + { + "epoch": 1.1663052905464006, + "grad_norm": 7.682575225830078, + "learning_rate": 1.2268968518787177e-05, + "loss": 0.0388, + "step": 10758 + }, + { + "epoch": 1.1664137033824804, + "grad_norm": 0.9802988767623901, + "learning_rate": 1.2268243145219789e-05, + "loss": 0.0081, + "step": 10759 + }, + { + "epoch": 1.1665221162185602, + "grad_norm": 3.7472732067108154, + "learning_rate": 1.2267517771652403e-05, + "loss": 0.0192, + "step": 10760 + }, + { + "epoch": 1.1666305290546402, + "grad_norm": 1.2500760555267334, + "learning_rate": 1.2266792398085015e-05, + "loss": 0.0143, + "step": 10761 + }, + { + "epoch": 1.16673894189072, + "grad_norm": 0.6737989783287048, + "learning_rate": 1.2266067024517627e-05, + "loss": 0.0078, + "step": 10762 + }, + { + "epoch": 1.1668473547267997, + "grad_norm": 1.9330629110336304, + "learning_rate": 1.2265341650950239e-05, + "loss": 0.0172, + "step": 10763 + }, + { + "epoch": 1.1669557675628794, + "grad_norm": 0.9924888014793396, + "learning_rate": 1.2264616277382853e-05, + "loss": 0.0038, + "step": 10764 + }, + { + "epoch": 1.1670641803989592, + "grad_norm": 4.1156721115112305, + "learning_rate": 1.2263890903815465e-05, + "loss": 0.0121, + "step": 10765 + }, + { + "epoch": 1.167172593235039, + "grad_norm": 2.089001178741455, + "learning_rate": 1.2263165530248077e-05, + "loss": 0.0181, + "step": 10766 + }, + { + "epoch": 1.1672810060711187, + "grad_norm": 0.6226394176483154, + "learning_rate": 1.2262440156680691e-05, + "loss": 0.0017, + "step": 10767 + }, + { + "epoch": 1.1673894189071987, + "grad_norm": 0.6915224194526672, + "learning_rate": 1.2261714783113305e-05, + "loss": 0.0042, + "step": 10768 + }, + { + "epoch": 1.1674978317432785, + "grad_norm": 3.703533172607422, + "learning_rate": 1.2260989409545919e-05, + "loss": 0.0416, + "step": 10769 + }, + { + "epoch": 1.1676062445793582, + "grad_norm": 0.4358938932418823, + "learning_rate": 1.226026403597853e-05, + "loss": 0.004, + "step": 10770 + }, + { + "epoch": 1.167714657415438, + "grad_norm": 3.056002616882324, + "learning_rate": 1.2259538662411143e-05, + "loss": 0.0504, + "step": 10771 + }, + { + "epoch": 1.1678230702515178, + "grad_norm": 0.10252916067838669, + "learning_rate": 1.2258813288843757e-05, + "loss": 0.0007, + "step": 10772 + }, + { + "epoch": 1.1679314830875975, + "grad_norm": 4.384588718414307, + "learning_rate": 1.2258087915276369e-05, + "loss": 0.0152, + "step": 10773 + }, + { + "epoch": 1.1680398959236773, + "grad_norm": 0.15525999665260315, + "learning_rate": 1.2257362541708981e-05, + "loss": 0.0009, + "step": 10774 + }, + { + "epoch": 1.168148308759757, + "grad_norm": 0.555203914642334, + "learning_rate": 1.2256637168141595e-05, + "loss": 0.003, + "step": 10775 + }, + { + "epoch": 1.1682567215958368, + "grad_norm": 4.453968048095703, + "learning_rate": 1.2255911794574207e-05, + "loss": 0.0076, + "step": 10776 + }, + { + "epoch": 1.1683651344319168, + "grad_norm": 0.5761930346488953, + "learning_rate": 1.225518642100682e-05, + "loss": 0.0007, + "step": 10777 + }, + { + "epoch": 1.1684735472679966, + "grad_norm": 5.195566177368164, + "learning_rate": 1.2254461047439431e-05, + "loss": 0.0308, + "step": 10778 + }, + { + "epoch": 1.1685819601040763, + "grad_norm": 2.7743327617645264, + "learning_rate": 1.2253735673872045e-05, + "loss": 0.0126, + "step": 10779 + }, + { + "epoch": 1.168690372940156, + "grad_norm": 0.9464302659034729, + "learning_rate": 1.2253010300304657e-05, + "loss": 0.0032, + "step": 10780 + }, + { + "epoch": 1.1687987857762359, + "grad_norm": 0.88682621717453, + "learning_rate": 1.225228492673727e-05, + "loss": 0.0025, + "step": 10781 + }, + { + "epoch": 1.1689071986123156, + "grad_norm": 6.281367778778076, + "learning_rate": 1.2251559553169883e-05, + "loss": 0.013, + "step": 10782 + }, + { + "epoch": 1.1690156114483954, + "grad_norm": 0.8770917057991028, + "learning_rate": 1.2250834179602495e-05, + "loss": 0.0045, + "step": 10783 + }, + { + "epoch": 1.1691240242844754, + "grad_norm": 3.0756354331970215, + "learning_rate": 1.2250108806035108e-05, + "loss": 0.0118, + "step": 10784 + }, + { + "epoch": 1.1692324371205551, + "grad_norm": 2.763733148574829, + "learning_rate": 1.2249383432467721e-05, + "loss": 0.0221, + "step": 10785 + }, + { + "epoch": 1.169340849956635, + "grad_norm": 0.34327179193496704, + "learning_rate": 1.2248658058900335e-05, + "loss": 0.0015, + "step": 10786 + }, + { + "epoch": 1.1694492627927147, + "grad_norm": 4.554755210876465, + "learning_rate": 1.2247932685332949e-05, + "loss": 0.0152, + "step": 10787 + }, + { + "epoch": 1.1695576756287944, + "grad_norm": 1.8248990774154663, + "learning_rate": 1.2247207311765561e-05, + "loss": 0.046, + "step": 10788 + }, + { + "epoch": 1.1696660884648742, + "grad_norm": 0.10901188105344772, + "learning_rate": 1.2246481938198173e-05, + "loss": 0.0011, + "step": 10789 + }, + { + "epoch": 1.169774501300954, + "grad_norm": 2.5980989933013916, + "learning_rate": 1.2245756564630786e-05, + "loss": 0.0427, + "step": 10790 + }, + { + "epoch": 1.169882914137034, + "grad_norm": 1.4779480695724487, + "learning_rate": 1.22450311910634e-05, + "loss": 0.0194, + "step": 10791 + }, + { + "epoch": 1.1699913269731137, + "grad_norm": 1.635900855064392, + "learning_rate": 1.2244305817496012e-05, + "loss": 0.0117, + "step": 10792 + }, + { + "epoch": 1.1700997398091935, + "grad_norm": 5.991448879241943, + "learning_rate": 1.2243580443928624e-05, + "loss": 0.0556, + "step": 10793 + }, + { + "epoch": 1.1702081526452732, + "grad_norm": 4.229211807250977, + "learning_rate": 1.2242855070361238e-05, + "loss": 0.0369, + "step": 10794 + }, + { + "epoch": 1.170316565481353, + "grad_norm": 0.5257624983787537, + "learning_rate": 1.224212969679385e-05, + "loss": 0.0084, + "step": 10795 + }, + { + "epoch": 1.1704249783174328, + "grad_norm": 3.4594640731811523, + "learning_rate": 1.2241404323226462e-05, + "loss": 0.0413, + "step": 10796 + }, + { + "epoch": 1.1705333911535125, + "grad_norm": 3.358440399169922, + "learning_rate": 1.2240678949659076e-05, + "loss": 0.0276, + "step": 10797 + }, + { + "epoch": 1.1706418039895923, + "grad_norm": 6.882850170135498, + "learning_rate": 1.2239953576091688e-05, + "loss": 0.0246, + "step": 10798 + }, + { + "epoch": 1.170750216825672, + "grad_norm": 3.7097830772399902, + "learning_rate": 1.22392282025243e-05, + "loss": 0.0473, + "step": 10799 + }, + { + "epoch": 1.170858629661752, + "grad_norm": 0.18751226365566254, + "learning_rate": 1.2238502828956914e-05, + "loss": 0.0006, + "step": 10800 + }, + { + "epoch": 1.1709670424978318, + "grad_norm": 1.191802978515625, + "learning_rate": 1.2237777455389526e-05, + "loss": 0.0063, + "step": 10801 + }, + { + "epoch": 1.1710754553339116, + "grad_norm": 7.3010358810424805, + "learning_rate": 1.2237052081822138e-05, + "loss": 0.0367, + "step": 10802 + }, + { + "epoch": 1.1711838681699913, + "grad_norm": 2.4413344860076904, + "learning_rate": 1.2236326708254754e-05, + "loss": 0.0093, + "step": 10803 + }, + { + "epoch": 1.171292281006071, + "grad_norm": 6.005721569061279, + "learning_rate": 1.2235601334687366e-05, + "loss": 0.0276, + "step": 10804 + }, + { + "epoch": 1.1714006938421508, + "grad_norm": 1.110290288925171, + "learning_rate": 1.2234875961119978e-05, + "loss": 0.0204, + "step": 10805 + }, + { + "epoch": 1.1715091066782306, + "grad_norm": 7.738246440887451, + "learning_rate": 1.2234150587552592e-05, + "loss": 0.0508, + "step": 10806 + }, + { + "epoch": 1.1716175195143106, + "grad_norm": 0.381146103143692, + "learning_rate": 1.2233425213985204e-05, + "loss": 0.0026, + "step": 10807 + }, + { + "epoch": 1.1717259323503904, + "grad_norm": 7.251605987548828, + "learning_rate": 1.2232699840417816e-05, + "loss": 0.0186, + "step": 10808 + }, + { + "epoch": 1.1718343451864701, + "grad_norm": 0.5635120868682861, + "learning_rate": 1.223197446685043e-05, + "loss": 0.004, + "step": 10809 + }, + { + "epoch": 1.1719427580225499, + "grad_norm": 1.8142399787902832, + "learning_rate": 1.2231249093283042e-05, + "loss": 0.0255, + "step": 10810 + }, + { + "epoch": 1.1720511708586296, + "grad_norm": 1.870002269744873, + "learning_rate": 1.2230523719715654e-05, + "loss": 0.005, + "step": 10811 + }, + { + "epoch": 1.1721595836947094, + "grad_norm": 0.9798648953437805, + "learning_rate": 1.2229798346148268e-05, + "loss": 0.0068, + "step": 10812 + }, + { + "epoch": 1.1722679965307892, + "grad_norm": 0.9636462926864624, + "learning_rate": 1.222907297258088e-05, + "loss": 0.011, + "step": 10813 + }, + { + "epoch": 1.1723764093668692, + "grad_norm": 1.815579891204834, + "learning_rate": 1.2228347599013492e-05, + "loss": 0.0082, + "step": 10814 + }, + { + "epoch": 1.172484822202949, + "grad_norm": 0.28261858224868774, + "learning_rate": 1.2227622225446104e-05, + "loss": 0.0023, + "step": 10815 + }, + { + "epoch": 1.1725932350390287, + "grad_norm": 6.491742134094238, + "learning_rate": 1.2226896851878718e-05, + "loss": 0.0269, + "step": 10816 + }, + { + "epoch": 1.1727016478751084, + "grad_norm": 1.8914709091186523, + "learning_rate": 1.222617147831133e-05, + "loss": 0.0324, + "step": 10817 + }, + { + "epoch": 1.1728100607111882, + "grad_norm": 0.28135132789611816, + "learning_rate": 1.2225446104743943e-05, + "loss": 0.0028, + "step": 10818 + }, + { + "epoch": 1.172918473547268, + "grad_norm": 4.037976264953613, + "learning_rate": 1.2224720731176556e-05, + "loss": 0.0304, + "step": 10819 + }, + { + "epoch": 1.1730268863833477, + "grad_norm": 1.322960376739502, + "learning_rate": 1.222399535760917e-05, + "loss": 0.019, + "step": 10820 + }, + { + "epoch": 1.1731352992194275, + "grad_norm": 3.9717838764190674, + "learning_rate": 1.2223269984041784e-05, + "loss": 0.0553, + "step": 10821 + }, + { + "epoch": 1.1732437120555073, + "grad_norm": 2.419574022293091, + "learning_rate": 1.2222544610474396e-05, + "loss": 0.0123, + "step": 10822 + }, + { + "epoch": 1.1733521248915872, + "grad_norm": 8.57119369506836, + "learning_rate": 1.2221819236907008e-05, + "loss": 0.0944, + "step": 10823 + }, + { + "epoch": 1.173460537727667, + "grad_norm": 5.909200191497803, + "learning_rate": 1.2221093863339622e-05, + "loss": 0.0423, + "step": 10824 + }, + { + "epoch": 1.1735689505637468, + "grad_norm": 1.54255211353302, + "learning_rate": 1.2220368489772234e-05, + "loss": 0.0243, + "step": 10825 + }, + { + "epoch": 1.1736773633998265, + "grad_norm": 3.9132959842681885, + "learning_rate": 1.2219643116204846e-05, + "loss": 0.0203, + "step": 10826 + }, + { + "epoch": 1.1737857762359063, + "grad_norm": 5.990978240966797, + "learning_rate": 1.221891774263746e-05, + "loss": 0.059, + "step": 10827 + }, + { + "epoch": 1.173894189071986, + "grad_norm": 4.1695122718811035, + "learning_rate": 1.2218192369070072e-05, + "loss": 0.0299, + "step": 10828 + }, + { + "epoch": 1.1740026019080658, + "grad_norm": 2.2812516689300537, + "learning_rate": 1.2217466995502685e-05, + "loss": 0.0117, + "step": 10829 + }, + { + "epoch": 1.1741110147441458, + "grad_norm": 0.4937387704849243, + "learning_rate": 1.2216741621935297e-05, + "loss": 0.0012, + "step": 10830 + }, + { + "epoch": 1.1742194275802256, + "grad_norm": 3.9905002117156982, + "learning_rate": 1.221601624836791e-05, + "loss": 0.0121, + "step": 10831 + }, + { + "epoch": 1.1743278404163053, + "grad_norm": 10.538368225097656, + "learning_rate": 1.2215290874800523e-05, + "loss": 0.019, + "step": 10832 + }, + { + "epoch": 1.174436253252385, + "grad_norm": 2.127638339996338, + "learning_rate": 1.2214565501233135e-05, + "loss": 0.025, + "step": 10833 + }, + { + "epoch": 1.1745446660884649, + "grad_norm": 0.4735340178012848, + "learning_rate": 1.2213840127665749e-05, + "loss": 0.0028, + "step": 10834 + }, + { + "epoch": 1.1746530789245446, + "grad_norm": 2.128497362136841, + "learning_rate": 1.221311475409836e-05, + "loss": 0.0321, + "step": 10835 + }, + { + "epoch": 1.1747614917606244, + "grad_norm": 3.310476779937744, + "learning_rate": 1.2212389380530973e-05, + "loss": 0.0254, + "step": 10836 + }, + { + "epoch": 1.1748699045967044, + "grad_norm": 12.194673538208008, + "learning_rate": 1.2211664006963588e-05, + "loss": 0.0504, + "step": 10837 + }, + { + "epoch": 1.1749783174327841, + "grad_norm": 3.3967983722686768, + "learning_rate": 1.22109386333962e-05, + "loss": 0.025, + "step": 10838 + }, + { + "epoch": 1.175086730268864, + "grad_norm": 2.2509121894836426, + "learning_rate": 1.2210213259828814e-05, + "loss": 0.0122, + "step": 10839 + }, + { + "epoch": 1.1751951431049437, + "grad_norm": 8.519373893737793, + "learning_rate": 1.2209487886261427e-05, + "loss": 0.0871, + "step": 10840 + }, + { + "epoch": 1.1753035559410234, + "grad_norm": 0.11458663642406464, + "learning_rate": 1.2208762512694039e-05, + "loss": 0.0009, + "step": 10841 + }, + { + "epoch": 1.1754119687771032, + "grad_norm": 6.138854026794434, + "learning_rate": 1.2208037139126651e-05, + "loss": 0.0253, + "step": 10842 + }, + { + "epoch": 1.175520381613183, + "grad_norm": 2.4275152683258057, + "learning_rate": 1.2207311765559265e-05, + "loss": 0.014, + "step": 10843 + }, + { + "epoch": 1.1756287944492627, + "grad_norm": 2.7125070095062256, + "learning_rate": 1.2206586391991877e-05, + "loss": 0.0702, + "step": 10844 + }, + { + "epoch": 1.1757372072853425, + "grad_norm": 5.999517440795898, + "learning_rate": 1.2205861018424489e-05, + "loss": 0.0263, + "step": 10845 + }, + { + "epoch": 1.1758456201214225, + "grad_norm": 1.9265786409378052, + "learning_rate": 1.2205135644857103e-05, + "loss": 0.0177, + "step": 10846 + }, + { + "epoch": 1.1759540329575022, + "grad_norm": 3.6265807151794434, + "learning_rate": 1.2204410271289715e-05, + "loss": 0.0285, + "step": 10847 + }, + { + "epoch": 1.176062445793582, + "grad_norm": 1.4571236371994019, + "learning_rate": 1.2203684897722327e-05, + "loss": 0.0135, + "step": 10848 + }, + { + "epoch": 1.1761708586296618, + "grad_norm": 0.44237279891967773, + "learning_rate": 1.2202959524154941e-05, + "loss": 0.0016, + "step": 10849 + }, + { + "epoch": 1.1762792714657415, + "grad_norm": 1.366164207458496, + "learning_rate": 1.2202234150587553e-05, + "loss": 0.0158, + "step": 10850 + }, + { + "epoch": 1.1763876843018213, + "grad_norm": 0.5250595808029175, + "learning_rate": 1.2201508777020165e-05, + "loss": 0.0036, + "step": 10851 + }, + { + "epoch": 1.176496097137901, + "grad_norm": 1.2889001369476318, + "learning_rate": 1.2200783403452779e-05, + "loss": 0.0307, + "step": 10852 + }, + { + "epoch": 1.176604509973981, + "grad_norm": 2.6273434162139893, + "learning_rate": 1.2200058029885391e-05, + "loss": 0.0247, + "step": 10853 + }, + { + "epoch": 1.1767129228100608, + "grad_norm": 0.4770309329032898, + "learning_rate": 1.2199332656318003e-05, + "loss": 0.0032, + "step": 10854 + }, + { + "epoch": 1.1768213356461406, + "grad_norm": 1.7865349054336548, + "learning_rate": 1.2198607282750619e-05, + "loss": 0.0157, + "step": 10855 + }, + { + "epoch": 1.1769297484822203, + "grad_norm": 2.8710925579071045, + "learning_rate": 1.2197881909183231e-05, + "loss": 0.0219, + "step": 10856 + }, + { + "epoch": 1.1770381613183, + "grad_norm": 3.201399803161621, + "learning_rate": 1.2197156535615843e-05, + "loss": 0.0741, + "step": 10857 + }, + { + "epoch": 1.1771465741543798, + "grad_norm": 1.3911153078079224, + "learning_rate": 1.2196431162048457e-05, + "loss": 0.0093, + "step": 10858 + }, + { + "epoch": 1.1772549869904596, + "grad_norm": 0.2756260335445404, + "learning_rate": 1.219570578848107e-05, + "loss": 0.002, + "step": 10859 + }, + { + "epoch": 1.1773633998265394, + "grad_norm": 2.5862197875976562, + "learning_rate": 1.2194980414913681e-05, + "loss": 0.0249, + "step": 10860 + }, + { + "epoch": 1.1774718126626191, + "grad_norm": 1.4299325942993164, + "learning_rate": 1.2194255041346295e-05, + "loss": 0.0129, + "step": 10861 + }, + { + "epoch": 1.1775802254986991, + "grad_norm": 2.752631902694702, + "learning_rate": 1.2193529667778907e-05, + "loss": 0.0269, + "step": 10862 + }, + { + "epoch": 1.1776886383347789, + "grad_norm": 1.9775700569152832, + "learning_rate": 1.219280429421152e-05, + "loss": 0.0117, + "step": 10863 + }, + { + "epoch": 1.1777970511708586, + "grad_norm": 5.566715717315674, + "learning_rate": 1.2192078920644133e-05, + "loss": 0.0228, + "step": 10864 + }, + { + "epoch": 1.1779054640069384, + "grad_norm": 14.753198623657227, + "learning_rate": 1.2191353547076745e-05, + "loss": 0.0701, + "step": 10865 + }, + { + "epoch": 1.1780138768430182, + "grad_norm": 1.9432755708694458, + "learning_rate": 1.2190628173509358e-05, + "loss": 0.0357, + "step": 10866 + }, + { + "epoch": 1.178122289679098, + "grad_norm": 2.7080657482147217, + "learning_rate": 1.218990279994197e-05, + "loss": 0.0387, + "step": 10867 + }, + { + "epoch": 1.1782307025151777, + "grad_norm": 3.501467227935791, + "learning_rate": 1.2189177426374584e-05, + "loss": 0.0135, + "step": 10868 + }, + { + "epoch": 1.1783391153512577, + "grad_norm": 4.81029748916626, + "learning_rate": 1.2188452052807196e-05, + "loss": 0.0251, + "step": 10869 + }, + { + "epoch": 1.1784475281873374, + "grad_norm": 4.178393363952637, + "learning_rate": 1.2187726679239808e-05, + "loss": 0.0706, + "step": 10870 + }, + { + "epoch": 1.1785559410234172, + "grad_norm": 1.8883116245269775, + "learning_rate": 1.2187001305672422e-05, + "loss": 0.0441, + "step": 10871 + }, + { + "epoch": 1.178664353859497, + "grad_norm": 3.9739608764648438, + "learning_rate": 1.2186275932105036e-05, + "loss": 0.0208, + "step": 10872 + }, + { + "epoch": 1.1787727666955767, + "grad_norm": 1.7894096374511719, + "learning_rate": 1.218555055853765e-05, + "loss": 0.0133, + "step": 10873 + }, + { + "epoch": 1.1788811795316565, + "grad_norm": 2.8957815170288086, + "learning_rate": 1.2184825184970262e-05, + "loss": 0.0126, + "step": 10874 + }, + { + "epoch": 1.1789895923677363, + "grad_norm": 3.704577684402466, + "learning_rate": 1.2184099811402874e-05, + "loss": 0.0221, + "step": 10875 + }, + { + "epoch": 1.1790980052038162, + "grad_norm": 0.20105335116386414, + "learning_rate": 1.2183374437835487e-05, + "loss": 0.0026, + "step": 10876 + }, + { + "epoch": 1.179206418039896, + "grad_norm": 1.0535759925842285, + "learning_rate": 1.21826490642681e-05, + "loss": 0.0044, + "step": 10877 + }, + { + "epoch": 1.1793148308759758, + "grad_norm": 5.809727668762207, + "learning_rate": 1.2181923690700712e-05, + "loss": 0.0511, + "step": 10878 + }, + { + "epoch": 1.1794232437120555, + "grad_norm": 1.5136797428131104, + "learning_rate": 1.2181198317133326e-05, + "loss": 0.0102, + "step": 10879 + }, + { + "epoch": 1.1795316565481353, + "grad_norm": 2.7158219814300537, + "learning_rate": 1.2180472943565938e-05, + "loss": 0.0273, + "step": 10880 + }, + { + "epoch": 1.179640069384215, + "grad_norm": 0.7216088771820068, + "learning_rate": 1.217974756999855e-05, + "loss": 0.0061, + "step": 10881 + }, + { + "epoch": 1.1797484822202948, + "grad_norm": 3.517941474914551, + "learning_rate": 1.2179022196431162e-05, + "loss": 0.0248, + "step": 10882 + }, + { + "epoch": 1.1798568950563746, + "grad_norm": 2.2773125171661377, + "learning_rate": 1.2178296822863776e-05, + "loss": 0.0247, + "step": 10883 + }, + { + "epoch": 1.1799653078924544, + "grad_norm": 4.674899101257324, + "learning_rate": 1.2177571449296388e-05, + "loss": 0.0325, + "step": 10884 + }, + { + "epoch": 1.1800737207285343, + "grad_norm": 1.5613607168197632, + "learning_rate": 1.2176846075729e-05, + "loss": 0.0085, + "step": 10885 + }, + { + "epoch": 1.180182133564614, + "grad_norm": 0.37792882323265076, + "learning_rate": 1.2176120702161614e-05, + "loss": 0.0024, + "step": 10886 + }, + { + "epoch": 1.1802905464006939, + "grad_norm": 0.7925089001655579, + "learning_rate": 1.2175395328594226e-05, + "loss": 0.0045, + "step": 10887 + }, + { + "epoch": 1.1803989592367736, + "grad_norm": 0.28546956181526184, + "learning_rate": 1.2174669955026838e-05, + "loss": 0.0031, + "step": 10888 + }, + { + "epoch": 1.1805073720728534, + "grad_norm": 1.9779719114303589, + "learning_rate": 1.2173944581459454e-05, + "loss": 0.0386, + "step": 10889 + }, + { + "epoch": 1.1806157849089332, + "grad_norm": 7.749134063720703, + "learning_rate": 1.2173219207892066e-05, + "loss": 0.0333, + "step": 10890 + }, + { + "epoch": 1.180724197745013, + "grad_norm": 1.3699311017990112, + "learning_rate": 1.217249383432468e-05, + "loss": 0.0224, + "step": 10891 + }, + { + "epoch": 1.180832610581093, + "grad_norm": 2.731682777404785, + "learning_rate": 1.2171768460757292e-05, + "loss": 0.0086, + "step": 10892 + }, + { + "epoch": 1.1809410234171727, + "grad_norm": 3.268702268600464, + "learning_rate": 1.2171043087189904e-05, + "loss": 0.043, + "step": 10893 + }, + { + "epoch": 1.1810494362532524, + "grad_norm": 6.786205291748047, + "learning_rate": 1.2170317713622516e-05, + "loss": 0.0278, + "step": 10894 + }, + { + "epoch": 1.1811578490893322, + "grad_norm": 0.5602878332138062, + "learning_rate": 1.216959234005513e-05, + "loss": 0.0021, + "step": 10895 + }, + { + "epoch": 1.181266261925412, + "grad_norm": 2.629945755004883, + "learning_rate": 1.2168866966487742e-05, + "loss": 0.0452, + "step": 10896 + }, + { + "epoch": 1.1813746747614917, + "grad_norm": 3.335148572921753, + "learning_rate": 1.2168141592920354e-05, + "loss": 0.0461, + "step": 10897 + }, + { + "epoch": 1.1814830875975715, + "grad_norm": 2.1625306606292725, + "learning_rate": 1.2167416219352968e-05, + "loss": 0.0089, + "step": 10898 + }, + { + "epoch": 1.1815915004336515, + "grad_norm": 2.1364192962646484, + "learning_rate": 1.216669084578558e-05, + "loss": 0.0076, + "step": 10899 + }, + { + "epoch": 1.1816999132697312, + "grad_norm": 0.3263004422187805, + "learning_rate": 1.2165965472218192e-05, + "loss": 0.0026, + "step": 10900 + }, + { + "epoch": 1.181808326105811, + "grad_norm": 2.1587748527526855, + "learning_rate": 1.2165240098650806e-05, + "loss": 0.0295, + "step": 10901 + }, + { + "epoch": 1.1819167389418908, + "grad_norm": 4.6895341873168945, + "learning_rate": 1.2164514725083418e-05, + "loss": 0.0105, + "step": 10902 + }, + { + "epoch": 1.1820251517779705, + "grad_norm": 2.149714231491089, + "learning_rate": 1.216378935151603e-05, + "loss": 0.0095, + "step": 10903 + }, + { + "epoch": 1.1821335646140503, + "grad_norm": 4.315841197967529, + "learning_rate": 1.2163063977948644e-05, + "loss": 0.0197, + "step": 10904 + }, + { + "epoch": 1.18224197745013, + "grad_norm": 4.481909275054932, + "learning_rate": 1.2162338604381257e-05, + "loss": 0.0526, + "step": 10905 + }, + { + "epoch": 1.1823503902862098, + "grad_norm": 11.308016777038574, + "learning_rate": 1.216161323081387e-05, + "loss": 0.1344, + "step": 10906 + }, + { + "epoch": 1.1824588031222896, + "grad_norm": 6.407362937927246, + "learning_rate": 1.2160887857246484e-05, + "loss": 0.0802, + "step": 10907 + }, + { + "epoch": 1.1825672159583696, + "grad_norm": 5.533939361572266, + "learning_rate": 1.2160162483679096e-05, + "loss": 0.0179, + "step": 10908 + }, + { + "epoch": 1.1826756287944493, + "grad_norm": 3.1001107692718506, + "learning_rate": 1.2159437110111709e-05, + "loss": 0.0066, + "step": 10909 + }, + { + "epoch": 1.182784041630529, + "grad_norm": 0.5184786915779114, + "learning_rate": 1.2158711736544322e-05, + "loss": 0.0036, + "step": 10910 + }, + { + "epoch": 1.1828924544666088, + "grad_norm": 4.330779552459717, + "learning_rate": 1.2157986362976935e-05, + "loss": 0.0111, + "step": 10911 + }, + { + "epoch": 1.1830008673026886, + "grad_norm": 1.4277267456054688, + "learning_rate": 1.2157260989409547e-05, + "loss": 0.0065, + "step": 10912 + }, + { + "epoch": 1.1831092801387684, + "grad_norm": 3.059708595275879, + "learning_rate": 1.215653561584216e-05, + "loss": 0.038, + "step": 10913 + }, + { + "epoch": 1.1832176929748481, + "grad_norm": 0.1331457793712616, + "learning_rate": 1.2155810242274773e-05, + "loss": 0.0012, + "step": 10914 + }, + { + "epoch": 1.1833261058109281, + "grad_norm": 1.5870068073272705, + "learning_rate": 1.2155084868707385e-05, + "loss": 0.0039, + "step": 10915 + }, + { + "epoch": 1.1834345186470079, + "grad_norm": 4.811568260192871, + "learning_rate": 1.2154359495139999e-05, + "loss": 0.0576, + "step": 10916 + }, + { + "epoch": 1.1835429314830876, + "grad_norm": 1.442061424255371, + "learning_rate": 1.215363412157261e-05, + "loss": 0.0048, + "step": 10917 + }, + { + "epoch": 1.1836513443191674, + "grad_norm": 3.4640040397644043, + "learning_rate": 1.2152908748005223e-05, + "loss": 0.016, + "step": 10918 + }, + { + "epoch": 1.1837597571552472, + "grad_norm": 4.746277809143066, + "learning_rate": 1.2152183374437837e-05, + "loss": 0.0444, + "step": 10919 + }, + { + "epoch": 1.183868169991327, + "grad_norm": 3.570484161376953, + "learning_rate": 1.2151458000870449e-05, + "loss": 0.0302, + "step": 10920 + }, + { + "epoch": 1.1839765828274067, + "grad_norm": 1.5261626243591309, + "learning_rate": 1.2150732627303061e-05, + "loss": 0.0102, + "step": 10921 + }, + { + "epoch": 1.1840849956634865, + "grad_norm": 2.184210777282715, + "learning_rate": 1.2150007253735673e-05, + "loss": 0.0312, + "step": 10922 + }, + { + "epoch": 1.1841934084995662, + "grad_norm": 3.563316822052002, + "learning_rate": 1.2149281880168287e-05, + "loss": 0.0456, + "step": 10923 + }, + { + "epoch": 1.1843018213356462, + "grad_norm": 3.4502270221710205, + "learning_rate": 1.2148556506600901e-05, + "loss": 0.0569, + "step": 10924 + }, + { + "epoch": 1.184410234171726, + "grad_norm": 5.874255657196045, + "learning_rate": 1.2147831133033515e-05, + "loss": 0.0082, + "step": 10925 + }, + { + "epoch": 1.1845186470078057, + "grad_norm": 3.6877520084381104, + "learning_rate": 1.2147105759466127e-05, + "loss": 0.0294, + "step": 10926 + }, + { + "epoch": 1.1846270598438855, + "grad_norm": 0.08057650923728943, + "learning_rate": 1.2146380385898739e-05, + "loss": 0.0004, + "step": 10927 + }, + { + "epoch": 1.1847354726799653, + "grad_norm": 0.43983930349349976, + "learning_rate": 1.2145655012331353e-05, + "loss": 0.0039, + "step": 10928 + }, + { + "epoch": 1.184843885516045, + "grad_norm": 3.036503791809082, + "learning_rate": 1.2144929638763965e-05, + "loss": 0.0423, + "step": 10929 + }, + { + "epoch": 1.1849522983521248, + "grad_norm": 9.539541244506836, + "learning_rate": 1.2144204265196577e-05, + "loss": 0.0414, + "step": 10930 + }, + { + "epoch": 1.1850607111882048, + "grad_norm": 1.7806305885314941, + "learning_rate": 1.2143478891629191e-05, + "loss": 0.0104, + "step": 10931 + }, + { + "epoch": 1.1851691240242845, + "grad_norm": 1.8120577335357666, + "learning_rate": 1.2142753518061803e-05, + "loss": 0.0269, + "step": 10932 + }, + { + "epoch": 1.1852775368603643, + "grad_norm": 1.1882530450820923, + "learning_rate": 1.2142028144494415e-05, + "loss": 0.0359, + "step": 10933 + }, + { + "epoch": 1.185385949696444, + "grad_norm": 2.740816593170166, + "learning_rate": 1.2141302770927027e-05, + "loss": 0.0181, + "step": 10934 + }, + { + "epoch": 1.1854943625325238, + "grad_norm": 0.1968413144350052, + "learning_rate": 1.2140577397359641e-05, + "loss": 0.0014, + "step": 10935 + }, + { + "epoch": 1.1856027753686036, + "grad_norm": 3.5036449432373047, + "learning_rate": 1.2139852023792253e-05, + "loss": 0.0356, + "step": 10936 + }, + { + "epoch": 1.1857111882046834, + "grad_norm": 0.13851293921470642, + "learning_rate": 1.2139126650224866e-05, + "loss": 0.0016, + "step": 10937 + }, + { + "epoch": 1.1858196010407633, + "grad_norm": 0.1885162889957428, + "learning_rate": 1.213840127665748e-05, + "loss": 0.0022, + "step": 10938 + }, + { + "epoch": 1.185928013876843, + "grad_norm": 1.73148512840271, + "learning_rate": 1.2137675903090091e-05, + "loss": 0.0087, + "step": 10939 + }, + { + "epoch": 1.1860364267129229, + "grad_norm": 0.8917644023895264, + "learning_rate": 1.2136950529522704e-05, + "loss": 0.0139, + "step": 10940 + }, + { + "epoch": 1.1861448395490026, + "grad_norm": 2.6755521297454834, + "learning_rate": 1.213622515595532e-05, + "loss": 0.0066, + "step": 10941 + }, + { + "epoch": 1.1862532523850824, + "grad_norm": 3.0378997325897217, + "learning_rate": 1.2135499782387931e-05, + "loss": 0.0231, + "step": 10942 + }, + { + "epoch": 1.1863616652211622, + "grad_norm": 2.2261767387390137, + "learning_rate": 1.2134774408820545e-05, + "loss": 0.047, + "step": 10943 + }, + { + "epoch": 1.186470078057242, + "grad_norm": 1.6354695558547974, + "learning_rate": 1.2134049035253157e-05, + "loss": 0.0236, + "step": 10944 + }, + { + "epoch": 1.1865784908933217, + "grad_norm": 3.408337116241455, + "learning_rate": 1.213332366168577e-05, + "loss": 0.0148, + "step": 10945 + }, + { + "epoch": 1.1866869037294014, + "grad_norm": 1.3351984024047852, + "learning_rate": 1.2132598288118382e-05, + "loss": 0.0136, + "step": 10946 + }, + { + "epoch": 1.1867953165654814, + "grad_norm": 3.376216173171997, + "learning_rate": 1.2131872914550995e-05, + "loss": 0.0135, + "step": 10947 + }, + { + "epoch": 1.1869037294015612, + "grad_norm": 0.341615229845047, + "learning_rate": 1.2131147540983608e-05, + "loss": 0.0008, + "step": 10948 + }, + { + "epoch": 1.187012142237641, + "grad_norm": 7.0039544105529785, + "learning_rate": 1.213042216741622e-05, + "loss": 0.0082, + "step": 10949 + }, + { + "epoch": 1.1871205550737207, + "grad_norm": 1.5805909633636475, + "learning_rate": 1.2129696793848834e-05, + "loss": 0.0186, + "step": 10950 + }, + { + "epoch": 1.1872289679098005, + "grad_norm": 2.173403739929199, + "learning_rate": 1.2128971420281446e-05, + "loss": 0.008, + "step": 10951 + }, + { + "epoch": 1.1873373807458802, + "grad_norm": 2.0973379611968994, + "learning_rate": 1.2128246046714058e-05, + "loss": 0.0505, + "step": 10952 + }, + { + "epoch": 1.18744579358196, + "grad_norm": 1.5442882776260376, + "learning_rate": 1.2127520673146672e-05, + "loss": 0.007, + "step": 10953 + }, + { + "epoch": 1.18755420641804, + "grad_norm": 4.798357009887695, + "learning_rate": 1.2126795299579284e-05, + "loss": 0.0446, + "step": 10954 + }, + { + "epoch": 1.1876626192541198, + "grad_norm": 0.44090843200683594, + "learning_rate": 1.2126069926011896e-05, + "loss": 0.0021, + "step": 10955 + }, + { + "epoch": 1.1877710320901995, + "grad_norm": 6.641654968261719, + "learning_rate": 1.212534455244451e-05, + "loss": 0.0521, + "step": 10956 + }, + { + "epoch": 1.1878794449262793, + "grad_norm": 2.1030900478363037, + "learning_rate": 1.2124619178877122e-05, + "loss": 0.031, + "step": 10957 + }, + { + "epoch": 1.187987857762359, + "grad_norm": 3.7226951122283936, + "learning_rate": 1.2123893805309736e-05, + "loss": 0.0235, + "step": 10958 + }, + { + "epoch": 1.1880962705984388, + "grad_norm": 0.6483659744262695, + "learning_rate": 1.212316843174235e-05, + "loss": 0.003, + "step": 10959 + }, + { + "epoch": 1.1882046834345186, + "grad_norm": 5.948858261108398, + "learning_rate": 1.2122443058174962e-05, + "loss": 0.0181, + "step": 10960 + }, + { + "epoch": 1.1883130962705986, + "grad_norm": 2.608220338821411, + "learning_rate": 1.2121717684607574e-05, + "loss": 0.03, + "step": 10961 + }, + { + "epoch": 1.1884215091066783, + "grad_norm": 1.5483180284500122, + "learning_rate": 1.2120992311040188e-05, + "loss": 0.0093, + "step": 10962 + }, + { + "epoch": 1.188529921942758, + "grad_norm": 0.5115403532981873, + "learning_rate": 1.21202669374728e-05, + "loss": 0.0045, + "step": 10963 + }, + { + "epoch": 1.1886383347788378, + "grad_norm": 4.699686527252197, + "learning_rate": 1.2119541563905412e-05, + "loss": 0.0165, + "step": 10964 + }, + { + "epoch": 1.1887467476149176, + "grad_norm": 2.1565732955932617, + "learning_rate": 1.2118816190338026e-05, + "loss": 0.0437, + "step": 10965 + }, + { + "epoch": 1.1888551604509974, + "grad_norm": 3.6670122146606445, + "learning_rate": 1.2118090816770638e-05, + "loss": 0.0139, + "step": 10966 + }, + { + "epoch": 1.1889635732870771, + "grad_norm": 2.3317792415618896, + "learning_rate": 1.211736544320325e-05, + "loss": 0.0081, + "step": 10967 + }, + { + "epoch": 1.189071986123157, + "grad_norm": 4.775241851806641, + "learning_rate": 1.2116640069635864e-05, + "loss": 0.0479, + "step": 10968 + }, + { + "epoch": 1.1891803989592367, + "grad_norm": 5.6832661628723145, + "learning_rate": 1.2115914696068476e-05, + "loss": 0.0407, + "step": 10969 + }, + { + "epoch": 1.1892888117953166, + "grad_norm": 9.238544464111328, + "learning_rate": 1.2115189322501088e-05, + "loss": 0.0178, + "step": 10970 + }, + { + "epoch": 1.1893972246313964, + "grad_norm": 2.6684701442718506, + "learning_rate": 1.2114463948933702e-05, + "loss": 0.0112, + "step": 10971 + }, + { + "epoch": 1.1895056374674762, + "grad_norm": 2.858515977859497, + "learning_rate": 1.2113738575366314e-05, + "loss": 0.0378, + "step": 10972 + }, + { + "epoch": 1.189614050303556, + "grad_norm": 4.094742774963379, + "learning_rate": 1.2113013201798926e-05, + "loss": 0.0045, + "step": 10973 + }, + { + "epoch": 1.1897224631396357, + "grad_norm": 0.7916362285614014, + "learning_rate": 1.2112287828231539e-05, + "loss": 0.0229, + "step": 10974 + }, + { + "epoch": 1.1898308759757155, + "grad_norm": 0.28878727555274963, + "learning_rate": 1.2111562454664152e-05, + "loss": 0.0018, + "step": 10975 + }, + { + "epoch": 1.1899392888117952, + "grad_norm": 1.237134575843811, + "learning_rate": 1.2110837081096766e-05, + "loss": 0.0178, + "step": 10976 + }, + { + "epoch": 1.1900477016478752, + "grad_norm": 0.8384252190589905, + "learning_rate": 1.211011170752938e-05, + "loss": 0.0022, + "step": 10977 + }, + { + "epoch": 1.190156114483955, + "grad_norm": 0.7003257870674133, + "learning_rate": 1.2109386333961992e-05, + "loss": 0.0108, + "step": 10978 + }, + { + "epoch": 1.1902645273200347, + "grad_norm": 2.068892240524292, + "learning_rate": 1.2108660960394604e-05, + "loss": 0.0272, + "step": 10979 + }, + { + "epoch": 1.1903729401561145, + "grad_norm": 4.681358337402344, + "learning_rate": 1.2107935586827218e-05, + "loss": 0.0755, + "step": 10980 + }, + { + "epoch": 1.1904813529921943, + "grad_norm": 4.779005527496338, + "learning_rate": 1.210721021325983e-05, + "loss": 0.0556, + "step": 10981 + }, + { + "epoch": 1.190589765828274, + "grad_norm": 4.438960075378418, + "learning_rate": 1.2106484839692442e-05, + "loss": 0.0136, + "step": 10982 + }, + { + "epoch": 1.1906981786643538, + "grad_norm": 0.8868601322174072, + "learning_rate": 1.2105759466125056e-05, + "loss": 0.0062, + "step": 10983 + }, + { + "epoch": 1.1908065915004338, + "grad_norm": 3.7307331562042236, + "learning_rate": 1.2105034092557668e-05, + "loss": 0.033, + "step": 10984 + }, + { + "epoch": 1.1909150043365135, + "grad_norm": 0.6922180652618408, + "learning_rate": 1.210430871899028e-05, + "loss": 0.0017, + "step": 10985 + }, + { + "epoch": 1.1910234171725933, + "grad_norm": 11.358860969543457, + "learning_rate": 1.2103583345422893e-05, + "loss": 0.0413, + "step": 10986 + }, + { + "epoch": 1.191131830008673, + "grad_norm": 1.2092887163162231, + "learning_rate": 1.2102857971855507e-05, + "loss": 0.0224, + "step": 10987 + }, + { + "epoch": 1.1912402428447528, + "grad_norm": 0.7017955780029297, + "learning_rate": 1.2102132598288119e-05, + "loss": 0.0123, + "step": 10988 + }, + { + "epoch": 1.1913486556808326, + "grad_norm": 0.7413027882575989, + "learning_rate": 1.210140722472073e-05, + "loss": 0.0053, + "step": 10989 + }, + { + "epoch": 1.1914570685169124, + "grad_norm": 0.9859144687652588, + "learning_rate": 1.2100681851153345e-05, + "loss": 0.0067, + "step": 10990 + }, + { + "epoch": 1.1915654813529921, + "grad_norm": 0.9277663826942444, + "learning_rate": 1.2099956477585957e-05, + "loss": 0.0046, + "step": 10991 + }, + { + "epoch": 1.1916738941890719, + "grad_norm": 2.197538137435913, + "learning_rate": 1.2099231104018569e-05, + "loss": 0.0101, + "step": 10992 + }, + { + "epoch": 1.1917823070251519, + "grad_norm": 2.1638495922088623, + "learning_rate": 1.2098505730451184e-05, + "loss": 0.0419, + "step": 10993 + }, + { + "epoch": 1.1918907198612316, + "grad_norm": 1.9625500440597534, + "learning_rate": 1.2097780356883797e-05, + "loss": 0.0103, + "step": 10994 + }, + { + "epoch": 1.1919991326973114, + "grad_norm": 8.477128982543945, + "learning_rate": 1.209705498331641e-05, + "loss": 0.0413, + "step": 10995 + }, + { + "epoch": 1.1921075455333912, + "grad_norm": 5.043407917022705, + "learning_rate": 1.2096329609749023e-05, + "loss": 0.063, + "step": 10996 + }, + { + "epoch": 1.192215958369471, + "grad_norm": 33.75244140625, + "learning_rate": 1.2095604236181635e-05, + "loss": 0.0434, + "step": 10997 + }, + { + "epoch": 1.1923243712055507, + "grad_norm": 3.936552047729492, + "learning_rate": 1.2094878862614247e-05, + "loss": 0.0219, + "step": 10998 + }, + { + "epoch": 1.1924327840416304, + "grad_norm": 0.42516347765922546, + "learning_rate": 1.209415348904686e-05, + "loss": 0.0027, + "step": 10999 + }, + { + "epoch": 1.1925411968777104, + "grad_norm": 7.9630818367004395, + "learning_rate": 1.2093428115479473e-05, + "loss": 0.0109, + "step": 11000 + }, + { + "epoch": 1.1926496097137902, + "grad_norm": 0.026572471484541893, + "learning_rate": 1.2092702741912085e-05, + "loss": 0.0002, + "step": 11001 + }, + { + "epoch": 1.19275802254987, + "grad_norm": 4.862512111663818, + "learning_rate": 1.2091977368344699e-05, + "loss": 0.0154, + "step": 11002 + }, + { + "epoch": 1.1928664353859497, + "grad_norm": 4.613224029541016, + "learning_rate": 1.2091251994777311e-05, + "loss": 0.0504, + "step": 11003 + }, + { + "epoch": 1.1929748482220295, + "grad_norm": 5.1128950119018555, + "learning_rate": 1.2090526621209923e-05, + "loss": 0.0354, + "step": 11004 + }, + { + "epoch": 1.1930832610581092, + "grad_norm": 1.6915740966796875, + "learning_rate": 1.2089801247642537e-05, + "loss": 0.0257, + "step": 11005 + }, + { + "epoch": 1.193191673894189, + "grad_norm": 2.1655354499816895, + "learning_rate": 1.2089075874075149e-05, + "loss": 0.0062, + "step": 11006 + }, + { + "epoch": 1.1933000867302688, + "grad_norm": 10.225187301635742, + "learning_rate": 1.2088350500507761e-05, + "loss": 0.021, + "step": 11007 + }, + { + "epoch": 1.1934084995663485, + "grad_norm": 8.189502716064453, + "learning_rate": 1.2087625126940375e-05, + "loss": 0.0495, + "step": 11008 + }, + { + "epoch": 1.1935169124024285, + "grad_norm": 0.05703819543123245, + "learning_rate": 1.2086899753372987e-05, + "loss": 0.0006, + "step": 11009 + }, + { + "epoch": 1.1936253252385083, + "grad_norm": 2.6372554302215576, + "learning_rate": 1.2086174379805601e-05, + "loss": 0.0414, + "step": 11010 + }, + { + "epoch": 1.193733738074588, + "grad_norm": 0.9687036275863647, + "learning_rate": 1.2085449006238215e-05, + "loss": 0.0102, + "step": 11011 + }, + { + "epoch": 1.1938421509106678, + "grad_norm": 2.044215679168701, + "learning_rate": 1.2084723632670827e-05, + "loss": 0.0281, + "step": 11012 + }, + { + "epoch": 1.1939505637467476, + "grad_norm": 3.673590660095215, + "learning_rate": 1.208399825910344e-05, + "loss": 0.0096, + "step": 11013 + }, + { + "epoch": 1.1940589765828273, + "grad_norm": 2.0852668285369873, + "learning_rate": 1.2083272885536053e-05, + "loss": 0.0217, + "step": 11014 + }, + { + "epoch": 1.194167389418907, + "grad_norm": 0.3848084807395935, + "learning_rate": 1.2082547511968665e-05, + "loss": 0.0043, + "step": 11015 + }, + { + "epoch": 1.194275802254987, + "grad_norm": 2.9761579036712646, + "learning_rate": 1.2081822138401277e-05, + "loss": 0.0279, + "step": 11016 + }, + { + "epoch": 1.1943842150910668, + "grad_norm": 0.08191641420125961, + "learning_rate": 1.2081096764833891e-05, + "loss": 0.0005, + "step": 11017 + }, + { + "epoch": 1.1944926279271466, + "grad_norm": 1.887429118156433, + "learning_rate": 1.2080371391266503e-05, + "loss": 0.0252, + "step": 11018 + }, + { + "epoch": 1.1946010407632264, + "grad_norm": 0.06926603615283966, + "learning_rate": 1.2079646017699115e-05, + "loss": 0.0003, + "step": 11019 + }, + { + "epoch": 1.1947094535993061, + "grad_norm": 3.299795627593994, + "learning_rate": 1.207892064413173e-05, + "loss": 0.0113, + "step": 11020 + }, + { + "epoch": 1.194817866435386, + "grad_norm": 1.5235053300857544, + "learning_rate": 1.2078195270564341e-05, + "loss": 0.0125, + "step": 11021 + }, + { + "epoch": 1.1949262792714657, + "grad_norm": 4.019629001617432, + "learning_rate": 1.2077469896996954e-05, + "loss": 0.0242, + "step": 11022 + }, + { + "epoch": 1.1950346921075456, + "grad_norm": 5.7286176681518555, + "learning_rate": 1.2076744523429567e-05, + "loss": 0.0716, + "step": 11023 + }, + { + "epoch": 1.1951431049436254, + "grad_norm": 2.5808019638061523, + "learning_rate": 1.207601914986218e-05, + "loss": 0.0749, + "step": 11024 + }, + { + "epoch": 1.1952515177797052, + "grad_norm": 0.8753448724746704, + "learning_rate": 1.2075293776294792e-05, + "loss": 0.0078, + "step": 11025 + }, + { + "epoch": 1.195359930615785, + "grad_norm": 0.6383314728736877, + "learning_rate": 1.2074568402727404e-05, + "loss": 0.0049, + "step": 11026 + }, + { + "epoch": 1.1954683434518647, + "grad_norm": 0.6312943696975708, + "learning_rate": 1.207384302916002e-05, + "loss": 0.0031, + "step": 11027 + }, + { + "epoch": 1.1955767562879445, + "grad_norm": 0.37773630023002625, + "learning_rate": 1.2073117655592632e-05, + "loss": 0.0016, + "step": 11028 + }, + { + "epoch": 1.1956851691240242, + "grad_norm": 3.2220077514648438, + "learning_rate": 1.2072392282025245e-05, + "loss": 0.0638, + "step": 11029 + }, + { + "epoch": 1.195793581960104, + "grad_norm": 0.09535878151655197, + "learning_rate": 1.2071666908457858e-05, + "loss": 0.0006, + "step": 11030 + }, + { + "epoch": 1.1959019947961838, + "grad_norm": 2.452502965927124, + "learning_rate": 1.207094153489047e-05, + "loss": 0.0074, + "step": 11031 + }, + { + "epoch": 1.1960104076322637, + "grad_norm": 4.122715473175049, + "learning_rate": 1.2070216161323083e-05, + "loss": 0.0126, + "step": 11032 + }, + { + "epoch": 1.1961188204683435, + "grad_norm": 2.5843939781188965, + "learning_rate": 1.2069490787755696e-05, + "loss": 0.0389, + "step": 11033 + }, + { + "epoch": 1.1962272333044233, + "grad_norm": 8.695602416992188, + "learning_rate": 1.2068765414188308e-05, + "loss": 0.0243, + "step": 11034 + }, + { + "epoch": 1.196335646140503, + "grad_norm": 1.7342631816864014, + "learning_rate": 1.2068040040620922e-05, + "loss": 0.0248, + "step": 11035 + }, + { + "epoch": 1.1964440589765828, + "grad_norm": 6.0707244873046875, + "learning_rate": 1.2067314667053534e-05, + "loss": 0.0731, + "step": 11036 + }, + { + "epoch": 1.1965524718126626, + "grad_norm": 1.242467999458313, + "learning_rate": 1.2066589293486146e-05, + "loss": 0.0189, + "step": 11037 + }, + { + "epoch": 1.1966608846487423, + "grad_norm": 1.0693371295928955, + "learning_rate": 1.2065863919918758e-05, + "loss": 0.0117, + "step": 11038 + }, + { + "epoch": 1.1967692974848223, + "grad_norm": 2.064786434173584, + "learning_rate": 1.2065138546351372e-05, + "loss": 0.0141, + "step": 11039 + }, + { + "epoch": 1.196877710320902, + "grad_norm": 1.769300937652588, + "learning_rate": 1.2064413172783984e-05, + "loss": 0.0115, + "step": 11040 + }, + { + "epoch": 1.1969861231569818, + "grad_norm": 8.921815872192383, + "learning_rate": 1.2063687799216596e-05, + "loss": 0.101, + "step": 11041 + }, + { + "epoch": 1.1970945359930616, + "grad_norm": 2.5630364418029785, + "learning_rate": 1.206296242564921e-05, + "loss": 0.0284, + "step": 11042 + }, + { + "epoch": 1.1972029488291414, + "grad_norm": 0.6042699217796326, + "learning_rate": 1.2062237052081822e-05, + "loss": 0.003, + "step": 11043 + }, + { + "epoch": 1.1973113616652211, + "grad_norm": 2.3795650005340576, + "learning_rate": 1.2061511678514434e-05, + "loss": 0.0163, + "step": 11044 + }, + { + "epoch": 1.1974197745013009, + "grad_norm": 0.20804482698440552, + "learning_rate": 1.206078630494705e-05, + "loss": 0.0014, + "step": 11045 + }, + { + "epoch": 1.1975281873373809, + "grad_norm": 11.588069915771484, + "learning_rate": 1.2060060931379662e-05, + "loss": 0.0395, + "step": 11046 + }, + { + "epoch": 1.1976366001734606, + "grad_norm": 0.7827935218811035, + "learning_rate": 1.2059335557812276e-05, + "loss": 0.0146, + "step": 11047 + }, + { + "epoch": 1.1977450130095404, + "grad_norm": 3.070775032043457, + "learning_rate": 1.2058610184244888e-05, + "loss": 0.0276, + "step": 11048 + }, + { + "epoch": 1.1978534258456202, + "grad_norm": 2.140207529067993, + "learning_rate": 1.20578848106775e-05, + "loss": 0.005, + "step": 11049 + }, + { + "epoch": 1.1979618386817, + "grad_norm": 2.6702992916107178, + "learning_rate": 1.2057159437110112e-05, + "loss": 0.0362, + "step": 11050 + }, + { + "epoch": 1.1980702515177797, + "grad_norm": 4.772741794586182, + "learning_rate": 1.2056434063542726e-05, + "loss": 0.0252, + "step": 11051 + }, + { + "epoch": 1.1981786643538594, + "grad_norm": 17.749568939208984, + "learning_rate": 1.2055708689975338e-05, + "loss": 0.0812, + "step": 11052 + }, + { + "epoch": 1.1982870771899392, + "grad_norm": 0.21108081936836243, + "learning_rate": 1.205498331640795e-05, + "loss": 0.0026, + "step": 11053 + }, + { + "epoch": 1.198395490026019, + "grad_norm": 0.37301263213157654, + "learning_rate": 1.2054257942840564e-05, + "loss": 0.0035, + "step": 11054 + }, + { + "epoch": 1.198503902862099, + "grad_norm": 4.045657157897949, + "learning_rate": 1.2053532569273176e-05, + "loss": 0.0111, + "step": 11055 + }, + { + "epoch": 1.1986123156981787, + "grad_norm": 0.10730772465467453, + "learning_rate": 1.2052807195705788e-05, + "loss": 0.0006, + "step": 11056 + }, + { + "epoch": 1.1987207285342585, + "grad_norm": 7.633304595947266, + "learning_rate": 1.2052081822138402e-05, + "loss": 0.1206, + "step": 11057 + }, + { + "epoch": 1.1988291413703382, + "grad_norm": 2.9325220584869385, + "learning_rate": 1.2051356448571014e-05, + "loss": 0.0065, + "step": 11058 + }, + { + "epoch": 1.198937554206418, + "grad_norm": 0.26554107666015625, + "learning_rate": 1.2050631075003627e-05, + "loss": 0.0021, + "step": 11059 + }, + { + "epoch": 1.1990459670424978, + "grad_norm": 5.702574253082275, + "learning_rate": 1.204990570143624e-05, + "loss": 0.0278, + "step": 11060 + }, + { + "epoch": 1.1991543798785775, + "grad_norm": 1.1781206130981445, + "learning_rate": 1.2049180327868853e-05, + "loss": 0.0046, + "step": 11061 + }, + { + "epoch": 1.1992627927146575, + "grad_norm": 3.7797255516052246, + "learning_rate": 1.2048454954301466e-05, + "loss": 0.0471, + "step": 11062 + }, + { + "epoch": 1.1993712055507373, + "grad_norm": 7.881364345550537, + "learning_rate": 1.204772958073408e-05, + "loss": 0.0433, + "step": 11063 + }, + { + "epoch": 1.199479618386817, + "grad_norm": 3.3005967140197754, + "learning_rate": 1.2047004207166692e-05, + "loss": 0.0885, + "step": 11064 + }, + { + "epoch": 1.1995880312228968, + "grad_norm": 4.512007236480713, + "learning_rate": 1.2046278833599305e-05, + "loss": 0.0723, + "step": 11065 + }, + { + "epoch": 1.1996964440589766, + "grad_norm": 1.3954048156738281, + "learning_rate": 1.2045553460031918e-05, + "loss": 0.0129, + "step": 11066 + }, + { + "epoch": 1.1998048568950563, + "grad_norm": 2.687807321548462, + "learning_rate": 1.204482808646453e-05, + "loss": 0.044, + "step": 11067 + }, + { + "epoch": 1.199913269731136, + "grad_norm": 1.671477198600769, + "learning_rate": 1.2044102712897143e-05, + "loss": 0.015, + "step": 11068 + }, + { + "epoch": 1.2000216825672159, + "grad_norm": 4.169572830200195, + "learning_rate": 1.2043377339329757e-05, + "loss": 0.0119, + "step": 11069 + }, + { + "epoch": 1.2001300954032956, + "grad_norm": 3.431332588195801, + "learning_rate": 1.2042651965762369e-05, + "loss": 0.031, + "step": 11070 + }, + { + "epoch": 1.2002385082393756, + "grad_norm": 2.510240316390991, + "learning_rate": 1.204192659219498e-05, + "loss": 0.0203, + "step": 11071 + }, + { + "epoch": 1.2003469210754554, + "grad_norm": 2.5190863609313965, + "learning_rate": 1.2041201218627595e-05, + "loss": 0.0208, + "step": 11072 + }, + { + "epoch": 1.2004553339115351, + "grad_norm": 3.25407338142395, + "learning_rate": 1.2040475845060207e-05, + "loss": 0.0361, + "step": 11073 + }, + { + "epoch": 1.200563746747615, + "grad_norm": 5.32167387008667, + "learning_rate": 1.2039750471492819e-05, + "loss": 0.0368, + "step": 11074 + }, + { + "epoch": 1.2006721595836947, + "grad_norm": 0.9512971043586731, + "learning_rate": 1.2039025097925433e-05, + "loss": 0.0219, + "step": 11075 + }, + { + "epoch": 1.2007805724197744, + "grad_norm": 2.4337103366851807, + "learning_rate": 1.2038299724358045e-05, + "loss": 0.0404, + "step": 11076 + }, + { + "epoch": 1.2008889852558542, + "grad_norm": 2.80267596244812, + "learning_rate": 1.2037574350790657e-05, + "loss": 0.0103, + "step": 11077 + }, + { + "epoch": 1.2009973980919342, + "grad_norm": 0.1453438401222229, + "learning_rate": 1.203684897722327e-05, + "loss": 0.0012, + "step": 11078 + }, + { + "epoch": 1.201105810928014, + "grad_norm": 1.800880789756775, + "learning_rate": 1.2036123603655885e-05, + "loss": 0.0057, + "step": 11079 + }, + { + "epoch": 1.2012142237640937, + "grad_norm": 2.446547269821167, + "learning_rate": 1.2035398230088497e-05, + "loss": 0.0388, + "step": 11080 + }, + { + "epoch": 1.2013226366001735, + "grad_norm": 4.402547836303711, + "learning_rate": 1.203467285652111e-05, + "loss": 0.0358, + "step": 11081 + }, + { + "epoch": 1.2014310494362532, + "grad_norm": 0.5625886917114258, + "learning_rate": 1.2033947482953723e-05, + "loss": 0.0062, + "step": 11082 + }, + { + "epoch": 1.201539462272333, + "grad_norm": 15.055619239807129, + "learning_rate": 1.2033222109386335e-05, + "loss": 0.028, + "step": 11083 + }, + { + "epoch": 1.2016478751084128, + "grad_norm": 5.232900142669678, + "learning_rate": 1.2032496735818949e-05, + "loss": 0.0242, + "step": 11084 + }, + { + "epoch": 1.2017562879444927, + "grad_norm": 1.280436635017395, + "learning_rate": 1.2031771362251561e-05, + "loss": 0.0165, + "step": 11085 + }, + { + "epoch": 1.2018647007805725, + "grad_norm": 0.29571446776390076, + "learning_rate": 1.2031045988684173e-05, + "loss": 0.002, + "step": 11086 + }, + { + "epoch": 1.2019731136166523, + "grad_norm": 0.5894508957862854, + "learning_rate": 1.2030320615116787e-05, + "loss": 0.0043, + "step": 11087 + }, + { + "epoch": 1.202081526452732, + "grad_norm": 1.7544406652450562, + "learning_rate": 1.2029595241549399e-05, + "loss": 0.0158, + "step": 11088 + }, + { + "epoch": 1.2021899392888118, + "grad_norm": 4.959657192230225, + "learning_rate": 1.2028869867982011e-05, + "loss": 0.0296, + "step": 11089 + }, + { + "epoch": 1.2022983521248916, + "grad_norm": 3.0941991806030273, + "learning_rate": 1.2028144494414623e-05, + "loss": 0.0374, + "step": 11090 + }, + { + "epoch": 1.2024067649609713, + "grad_norm": 3.4381585121154785, + "learning_rate": 1.2027419120847237e-05, + "loss": 0.0066, + "step": 11091 + }, + { + "epoch": 1.202515177797051, + "grad_norm": 1.4115797281265259, + "learning_rate": 1.202669374727985e-05, + "loss": 0.0226, + "step": 11092 + }, + { + "epoch": 1.2026235906331308, + "grad_norm": 6.552366256713867, + "learning_rate": 1.2025968373712462e-05, + "loss": 0.0279, + "step": 11093 + }, + { + "epoch": 1.2027320034692108, + "grad_norm": 3.243685722351074, + "learning_rate": 1.2025243000145075e-05, + "loss": 0.0887, + "step": 11094 + }, + { + "epoch": 1.2028404163052906, + "grad_norm": 5.634881496429443, + "learning_rate": 1.2024517626577687e-05, + "loss": 0.0311, + "step": 11095 + }, + { + "epoch": 1.2029488291413704, + "grad_norm": 2.487607955932617, + "learning_rate": 1.2023792253010303e-05, + "loss": 0.0422, + "step": 11096 + }, + { + "epoch": 1.2030572419774501, + "grad_norm": 2.3423521518707275, + "learning_rate": 1.2023066879442915e-05, + "loss": 0.0179, + "step": 11097 + }, + { + "epoch": 1.2031656548135299, + "grad_norm": 4.123790264129639, + "learning_rate": 1.2022341505875527e-05, + "loss": 0.0271, + "step": 11098 + }, + { + "epoch": 1.2032740676496096, + "grad_norm": 0.676606297492981, + "learning_rate": 1.2021616132308141e-05, + "loss": 0.0025, + "step": 11099 + }, + { + "epoch": 1.2033824804856894, + "grad_norm": 3.8618710041046143, + "learning_rate": 1.2020890758740753e-05, + "loss": 0.0317, + "step": 11100 + }, + { + "epoch": 1.2034908933217694, + "grad_norm": 2.4194233417510986, + "learning_rate": 1.2020165385173365e-05, + "loss": 0.0343, + "step": 11101 + }, + { + "epoch": 1.2035993061578492, + "grad_norm": 4.7392659187316895, + "learning_rate": 1.2019440011605978e-05, + "loss": 0.0196, + "step": 11102 + }, + { + "epoch": 1.203707718993929, + "grad_norm": 1.8833646774291992, + "learning_rate": 1.2018714638038591e-05, + "loss": 0.0332, + "step": 11103 + }, + { + "epoch": 1.2038161318300087, + "grad_norm": 0.2417459338903427, + "learning_rate": 1.2017989264471204e-05, + "loss": 0.002, + "step": 11104 + }, + { + "epoch": 1.2039245446660884, + "grad_norm": 1.555539608001709, + "learning_rate": 1.2017263890903816e-05, + "loss": 0.0511, + "step": 11105 + }, + { + "epoch": 1.2040329575021682, + "grad_norm": 0.033252131193876266, + "learning_rate": 1.201653851733643e-05, + "loss": 0.0004, + "step": 11106 + }, + { + "epoch": 1.204141370338248, + "grad_norm": 3.022956609725952, + "learning_rate": 1.2015813143769042e-05, + "loss": 0.0083, + "step": 11107 + }, + { + "epoch": 1.204249783174328, + "grad_norm": 3.2602970600128174, + "learning_rate": 1.2015087770201654e-05, + "loss": 0.0262, + "step": 11108 + }, + { + "epoch": 1.2043581960104077, + "grad_norm": 1.9505574703216553, + "learning_rate": 1.2014362396634268e-05, + "loss": 0.0114, + "step": 11109 + }, + { + "epoch": 1.2044666088464875, + "grad_norm": 7.272888660430908, + "learning_rate": 1.201363702306688e-05, + "loss": 0.0261, + "step": 11110 + }, + { + "epoch": 1.2045750216825672, + "grad_norm": 1.624041199684143, + "learning_rate": 1.2012911649499492e-05, + "loss": 0.0153, + "step": 11111 + }, + { + "epoch": 1.204683434518647, + "grad_norm": 1.1000339984893799, + "learning_rate": 1.2012186275932106e-05, + "loss": 0.0193, + "step": 11112 + }, + { + "epoch": 1.2047918473547268, + "grad_norm": 6.73529052734375, + "learning_rate": 1.2011460902364718e-05, + "loss": 0.0623, + "step": 11113 + }, + { + "epoch": 1.2049002601908065, + "grad_norm": 1.1785117387771606, + "learning_rate": 1.2010735528797333e-05, + "loss": 0.0045, + "step": 11114 + }, + { + "epoch": 1.2050086730268863, + "grad_norm": 3.24503493309021, + "learning_rate": 1.2010010155229946e-05, + "loss": 0.0135, + "step": 11115 + }, + { + "epoch": 1.205117085862966, + "grad_norm": 0.9319902062416077, + "learning_rate": 1.2009284781662558e-05, + "loss": 0.0067, + "step": 11116 + }, + { + "epoch": 1.205225498699046, + "grad_norm": 5.5065436363220215, + "learning_rate": 1.200855940809517e-05, + "loss": 0.0584, + "step": 11117 + }, + { + "epoch": 1.2053339115351258, + "grad_norm": 1.6675093173980713, + "learning_rate": 1.2007834034527784e-05, + "loss": 0.0064, + "step": 11118 + }, + { + "epoch": 1.2054423243712056, + "grad_norm": 1.339927077293396, + "learning_rate": 1.2007108660960396e-05, + "loss": 0.0265, + "step": 11119 + }, + { + "epoch": 1.2055507372072853, + "grad_norm": 0.5763654708862305, + "learning_rate": 1.2006383287393008e-05, + "loss": 0.006, + "step": 11120 + }, + { + "epoch": 1.205659150043365, + "grad_norm": 0.5530316233634949, + "learning_rate": 1.2005657913825622e-05, + "loss": 0.0036, + "step": 11121 + }, + { + "epoch": 1.2057675628794449, + "grad_norm": 4.495781898498535, + "learning_rate": 1.2004932540258234e-05, + "loss": 0.0576, + "step": 11122 + }, + { + "epoch": 1.2058759757155246, + "grad_norm": 2.5397303104400635, + "learning_rate": 1.2004207166690846e-05, + "loss": 0.0224, + "step": 11123 + }, + { + "epoch": 1.2059843885516046, + "grad_norm": 2.335371494293213, + "learning_rate": 1.200348179312346e-05, + "loss": 0.0132, + "step": 11124 + }, + { + "epoch": 1.2060928013876844, + "grad_norm": 6.917137622833252, + "learning_rate": 1.2002756419556072e-05, + "loss": 0.0381, + "step": 11125 + }, + { + "epoch": 1.2062012142237641, + "grad_norm": 0.5747073292732239, + "learning_rate": 1.2002031045988684e-05, + "loss": 0.0072, + "step": 11126 + }, + { + "epoch": 1.206309627059844, + "grad_norm": 2.840790033340454, + "learning_rate": 1.2001305672421298e-05, + "loss": 0.0224, + "step": 11127 + }, + { + "epoch": 1.2064180398959237, + "grad_norm": 4.881425380706787, + "learning_rate": 1.200058029885391e-05, + "loss": 0.0313, + "step": 11128 + }, + { + "epoch": 1.2065264527320034, + "grad_norm": 1.184338927268982, + "learning_rate": 1.1999854925286522e-05, + "loss": 0.0048, + "step": 11129 + }, + { + "epoch": 1.2066348655680832, + "grad_norm": 1.7501369714736938, + "learning_rate": 1.1999129551719135e-05, + "loss": 0.0196, + "step": 11130 + }, + { + "epoch": 1.2067432784041632, + "grad_norm": 2.6767327785491943, + "learning_rate": 1.199840417815175e-05, + "loss": 0.0219, + "step": 11131 + }, + { + "epoch": 1.206851691240243, + "grad_norm": 1.8959102630615234, + "learning_rate": 1.1997678804584362e-05, + "loss": 0.0277, + "step": 11132 + }, + { + "epoch": 1.2069601040763227, + "grad_norm": 1.4019455909729004, + "learning_rate": 1.1996953431016976e-05, + "loss": 0.0216, + "step": 11133 + }, + { + "epoch": 1.2070685169124025, + "grad_norm": 1.7861919403076172, + "learning_rate": 1.1996228057449588e-05, + "loss": 0.014, + "step": 11134 + }, + { + "epoch": 1.2071769297484822, + "grad_norm": 13.618083953857422, + "learning_rate": 1.19955026838822e-05, + "loss": 0.0427, + "step": 11135 + }, + { + "epoch": 1.207285342584562, + "grad_norm": 3.9565682411193848, + "learning_rate": 1.1994777310314814e-05, + "loss": 0.0186, + "step": 11136 + }, + { + "epoch": 1.2073937554206418, + "grad_norm": 0.8816223740577698, + "learning_rate": 1.1994051936747426e-05, + "loss": 0.0057, + "step": 11137 + }, + { + "epoch": 1.2075021682567215, + "grad_norm": 2.0391852855682373, + "learning_rate": 1.1993326563180038e-05, + "loss": 0.034, + "step": 11138 + }, + { + "epoch": 1.2076105810928013, + "grad_norm": 1.5858179330825806, + "learning_rate": 1.1992601189612652e-05, + "loss": 0.028, + "step": 11139 + }, + { + "epoch": 1.2077189939288813, + "grad_norm": 3.079359531402588, + "learning_rate": 1.1991875816045264e-05, + "loss": 0.0786, + "step": 11140 + }, + { + "epoch": 1.207827406764961, + "grad_norm": 1.808292031288147, + "learning_rate": 1.1991150442477877e-05, + "loss": 0.0212, + "step": 11141 + }, + { + "epoch": 1.2079358196010408, + "grad_norm": 4.3533172607421875, + "learning_rate": 1.1990425068910489e-05, + "loss": 0.0319, + "step": 11142 + }, + { + "epoch": 1.2080442324371206, + "grad_norm": 2.740485429763794, + "learning_rate": 1.1989699695343103e-05, + "loss": 0.0111, + "step": 11143 + }, + { + "epoch": 1.2081526452732003, + "grad_norm": 0.5006037354469299, + "learning_rate": 1.1988974321775715e-05, + "loss": 0.0066, + "step": 11144 + }, + { + "epoch": 1.20826105810928, + "grad_norm": 3.2558369636535645, + "learning_rate": 1.1988248948208327e-05, + "loss": 0.0598, + "step": 11145 + }, + { + "epoch": 1.2083694709453598, + "grad_norm": 5.503617286682129, + "learning_rate": 1.198752357464094e-05, + "loss": 0.0274, + "step": 11146 + }, + { + "epoch": 1.2084778837814398, + "grad_norm": 0.6088030934333801, + "learning_rate": 1.1986798201073553e-05, + "loss": 0.0095, + "step": 11147 + }, + { + "epoch": 1.2085862966175196, + "grad_norm": 3.662682294845581, + "learning_rate": 1.1986072827506168e-05, + "loss": 0.0338, + "step": 11148 + }, + { + "epoch": 1.2086947094535994, + "grad_norm": 8.145100593566895, + "learning_rate": 1.198534745393878e-05, + "loss": 0.0685, + "step": 11149 + }, + { + "epoch": 1.2088031222896791, + "grad_norm": 7.0708699226379395, + "learning_rate": 1.1984622080371393e-05, + "loss": 0.0137, + "step": 11150 + }, + { + "epoch": 1.2089115351257589, + "grad_norm": 3.348189353942871, + "learning_rate": 1.1983896706804006e-05, + "loss": 0.0381, + "step": 11151 + }, + { + "epoch": 1.2090199479618386, + "grad_norm": 1.7389404773712158, + "learning_rate": 1.1983171333236619e-05, + "loss": 0.0172, + "step": 11152 + }, + { + "epoch": 1.2091283607979184, + "grad_norm": 1.1920779943466187, + "learning_rate": 1.198244595966923e-05, + "loss": 0.0057, + "step": 11153 + }, + { + "epoch": 1.2092367736339982, + "grad_norm": 0.40478765964508057, + "learning_rate": 1.1981720586101843e-05, + "loss": 0.0028, + "step": 11154 + }, + { + "epoch": 1.209345186470078, + "grad_norm": 1.7955409288406372, + "learning_rate": 1.1980995212534457e-05, + "loss": 0.0165, + "step": 11155 + }, + { + "epoch": 1.209453599306158, + "grad_norm": 1.335359811782837, + "learning_rate": 1.1980269838967069e-05, + "loss": 0.0063, + "step": 11156 + }, + { + "epoch": 1.2095620121422377, + "grad_norm": 3.3278744220733643, + "learning_rate": 1.1979544465399681e-05, + "loss": 0.0158, + "step": 11157 + }, + { + "epoch": 1.2096704249783174, + "grad_norm": 3.3914296627044678, + "learning_rate": 1.1978819091832295e-05, + "loss": 0.0499, + "step": 11158 + }, + { + "epoch": 1.2097788378143972, + "grad_norm": 0.7931989431381226, + "learning_rate": 1.1978093718264907e-05, + "loss": 0.0086, + "step": 11159 + }, + { + "epoch": 1.209887250650477, + "grad_norm": 0.28459078073501587, + "learning_rate": 1.197736834469752e-05, + "loss": 0.0023, + "step": 11160 + }, + { + "epoch": 1.2099956634865567, + "grad_norm": 1.9310837984085083, + "learning_rate": 1.1976642971130133e-05, + "loss": 0.017, + "step": 11161 + }, + { + "epoch": 1.2101040763226365, + "grad_norm": 0.9376214146614075, + "learning_rate": 1.1975917597562745e-05, + "loss": 0.0083, + "step": 11162 + }, + { + "epoch": 1.2102124891587165, + "grad_norm": 1.6671197414398193, + "learning_rate": 1.1975192223995357e-05, + "loss": 0.0577, + "step": 11163 + }, + { + "epoch": 1.2103209019947962, + "grad_norm": 1.220201015472412, + "learning_rate": 1.1974466850427971e-05, + "loss": 0.0106, + "step": 11164 + }, + { + "epoch": 1.210429314830876, + "grad_norm": 2.6694247722625732, + "learning_rate": 1.1973741476860583e-05, + "loss": 0.0458, + "step": 11165 + }, + { + "epoch": 1.2105377276669558, + "grad_norm": 10.717500686645508, + "learning_rate": 1.1973016103293199e-05, + "loss": 0.0762, + "step": 11166 + }, + { + "epoch": 1.2106461405030355, + "grad_norm": 1.1381317377090454, + "learning_rate": 1.1972290729725811e-05, + "loss": 0.0143, + "step": 11167 + }, + { + "epoch": 1.2107545533391153, + "grad_norm": 0.0716022253036499, + "learning_rate": 1.1971565356158423e-05, + "loss": 0.0006, + "step": 11168 + }, + { + "epoch": 1.210862966175195, + "grad_norm": 2.8128347396850586, + "learning_rate": 1.1970839982591035e-05, + "loss": 0.0707, + "step": 11169 + }, + { + "epoch": 1.210971379011275, + "grad_norm": 1.7533873319625854, + "learning_rate": 1.1970114609023649e-05, + "loss": 0.0153, + "step": 11170 + }, + { + "epoch": 1.2110797918473548, + "grad_norm": 2.422497510910034, + "learning_rate": 1.1969389235456261e-05, + "loss": 0.0123, + "step": 11171 + }, + { + "epoch": 1.2111882046834346, + "grad_norm": 0.21721862256526947, + "learning_rate": 1.1968663861888873e-05, + "loss": 0.0027, + "step": 11172 + }, + { + "epoch": 1.2112966175195143, + "grad_norm": 5.090118408203125, + "learning_rate": 1.1967938488321487e-05, + "loss": 0.0327, + "step": 11173 + }, + { + "epoch": 1.211405030355594, + "grad_norm": 3.0814032554626465, + "learning_rate": 1.19672131147541e-05, + "loss": 0.0206, + "step": 11174 + }, + { + "epoch": 1.2115134431916739, + "grad_norm": 1.6212868690490723, + "learning_rate": 1.1966487741186711e-05, + "loss": 0.0069, + "step": 11175 + }, + { + "epoch": 1.2116218560277536, + "grad_norm": 2.10152006149292, + "learning_rate": 1.1965762367619325e-05, + "loss": 0.048, + "step": 11176 + }, + { + "epoch": 1.2117302688638334, + "grad_norm": 1.1514545679092407, + "learning_rate": 1.1965036994051937e-05, + "loss": 0.0148, + "step": 11177 + }, + { + "epoch": 1.2118386816999132, + "grad_norm": 0.7630380392074585, + "learning_rate": 1.196431162048455e-05, + "loss": 0.0079, + "step": 11178 + }, + { + "epoch": 1.2119470945359931, + "grad_norm": 2.9982314109802246, + "learning_rate": 1.1963586246917163e-05, + "loss": 0.0381, + "step": 11179 + }, + { + "epoch": 1.212055507372073, + "grad_norm": 1.8361276388168335, + "learning_rate": 1.1962860873349776e-05, + "loss": 0.0222, + "step": 11180 + }, + { + "epoch": 1.2121639202081527, + "grad_norm": 3.7954957485198975, + "learning_rate": 1.1962135499782388e-05, + "loss": 0.0236, + "step": 11181 + }, + { + "epoch": 1.2122723330442324, + "grad_norm": 0.18400399386882782, + "learning_rate": 1.1961410126215e-05, + "loss": 0.0021, + "step": 11182 + }, + { + "epoch": 1.2123807458803122, + "grad_norm": 4.971864700317383, + "learning_rate": 1.1960684752647615e-05, + "loss": 0.034, + "step": 11183 + }, + { + "epoch": 1.212489158716392, + "grad_norm": 0.9153687953948975, + "learning_rate": 1.1959959379080228e-05, + "loss": 0.0188, + "step": 11184 + }, + { + "epoch": 1.2125975715524717, + "grad_norm": 1.0525496006011963, + "learning_rate": 1.1959234005512841e-05, + "loss": 0.0032, + "step": 11185 + }, + { + "epoch": 1.2127059843885517, + "grad_norm": 1.6508729457855225, + "learning_rate": 1.1958508631945454e-05, + "loss": 0.008, + "step": 11186 + }, + { + "epoch": 1.2128143972246315, + "grad_norm": 0.5026692152023315, + "learning_rate": 1.1957783258378066e-05, + "loss": 0.0022, + "step": 11187 + }, + { + "epoch": 1.2129228100607112, + "grad_norm": 8.919947624206543, + "learning_rate": 1.195705788481068e-05, + "loss": 0.0141, + "step": 11188 + }, + { + "epoch": 1.213031222896791, + "grad_norm": 2.133842945098877, + "learning_rate": 1.1956332511243292e-05, + "loss": 0.0509, + "step": 11189 + }, + { + "epoch": 1.2131396357328708, + "grad_norm": 1.864548683166504, + "learning_rate": 1.1955607137675904e-05, + "loss": 0.0502, + "step": 11190 + }, + { + "epoch": 1.2132480485689505, + "grad_norm": 2.637119770050049, + "learning_rate": 1.1954881764108518e-05, + "loss": 0.0491, + "step": 11191 + }, + { + "epoch": 1.2133564614050303, + "grad_norm": 1.8785662651062012, + "learning_rate": 1.195415639054113e-05, + "loss": 0.0294, + "step": 11192 + }, + { + "epoch": 1.2134648742411103, + "grad_norm": 2.629696846008301, + "learning_rate": 1.1953431016973742e-05, + "loss": 0.0349, + "step": 11193 + }, + { + "epoch": 1.21357328707719, + "grad_norm": 3.9623520374298096, + "learning_rate": 1.1952705643406354e-05, + "loss": 0.0252, + "step": 11194 + }, + { + "epoch": 1.2136816999132698, + "grad_norm": 0.6269030570983887, + "learning_rate": 1.1951980269838968e-05, + "loss": 0.0069, + "step": 11195 + }, + { + "epoch": 1.2137901127493496, + "grad_norm": 2.8848843574523926, + "learning_rate": 1.195125489627158e-05, + "loss": 0.0702, + "step": 11196 + }, + { + "epoch": 1.2138985255854293, + "grad_norm": 0.9245696663856506, + "learning_rate": 1.1950529522704192e-05, + "loss": 0.0069, + "step": 11197 + }, + { + "epoch": 1.214006938421509, + "grad_norm": 7.036011219024658, + "learning_rate": 1.1949804149136806e-05, + "loss": 0.0452, + "step": 11198 + }, + { + "epoch": 1.2141153512575888, + "grad_norm": 1.8733690977096558, + "learning_rate": 1.1949078775569418e-05, + "loss": 0.0331, + "step": 11199 + }, + { + "epoch": 1.2142237640936686, + "grad_norm": 0.8948890566825867, + "learning_rate": 1.1948353402002034e-05, + "loss": 0.0036, + "step": 11200 + }, + { + "epoch": 1.2143321769297484, + "grad_norm": 2.26029372215271, + "learning_rate": 1.1947628028434646e-05, + "loss": 0.083, + "step": 11201 + }, + { + "epoch": 1.2144405897658284, + "grad_norm": 6.196389675140381, + "learning_rate": 1.1946902654867258e-05, + "loss": 0.0226, + "step": 11202 + }, + { + "epoch": 1.2145490026019081, + "grad_norm": 1.5161644220352173, + "learning_rate": 1.1946177281299872e-05, + "loss": 0.0142, + "step": 11203 + }, + { + "epoch": 1.2146574154379879, + "grad_norm": 2.0753092765808105, + "learning_rate": 1.1945451907732484e-05, + "loss": 0.0103, + "step": 11204 + }, + { + "epoch": 1.2147658282740676, + "grad_norm": 2.593363046646118, + "learning_rate": 1.1944726534165096e-05, + "loss": 0.0442, + "step": 11205 + }, + { + "epoch": 1.2148742411101474, + "grad_norm": 0.09941025823354721, + "learning_rate": 1.1944001160597708e-05, + "loss": 0.0014, + "step": 11206 + }, + { + "epoch": 1.2149826539462272, + "grad_norm": 1.0448335409164429, + "learning_rate": 1.1943275787030322e-05, + "loss": 0.0125, + "step": 11207 + }, + { + "epoch": 1.215091066782307, + "grad_norm": 3.9412474632263184, + "learning_rate": 1.1942550413462934e-05, + "loss": 0.0159, + "step": 11208 + }, + { + "epoch": 1.215199479618387, + "grad_norm": 0.14089301228523254, + "learning_rate": 1.1941825039895546e-05, + "loss": 0.0014, + "step": 11209 + }, + { + "epoch": 1.2153078924544667, + "grad_norm": 0.40248987078666687, + "learning_rate": 1.194109966632816e-05, + "loss": 0.0048, + "step": 11210 + }, + { + "epoch": 1.2154163052905465, + "grad_norm": 8.348001480102539, + "learning_rate": 1.1940374292760772e-05, + "loss": 0.0255, + "step": 11211 + }, + { + "epoch": 1.2155247181266262, + "grad_norm": 1.753016710281372, + "learning_rate": 1.1939648919193385e-05, + "loss": 0.0159, + "step": 11212 + }, + { + "epoch": 1.215633130962706, + "grad_norm": 0.41344934701919556, + "learning_rate": 1.1938923545625998e-05, + "loss": 0.0032, + "step": 11213 + }, + { + "epoch": 1.2157415437987857, + "grad_norm": 5.086875915527344, + "learning_rate": 1.193819817205861e-05, + "loss": 0.0381, + "step": 11214 + }, + { + "epoch": 1.2158499566348655, + "grad_norm": 4.480194091796875, + "learning_rate": 1.1937472798491223e-05, + "loss": 0.062, + "step": 11215 + }, + { + "epoch": 1.2159583694709453, + "grad_norm": 3.944117546081543, + "learning_rate": 1.1936747424923836e-05, + "loss": 0.0535, + "step": 11216 + }, + { + "epoch": 1.216066782307025, + "grad_norm": 1.7223505973815918, + "learning_rate": 1.193602205135645e-05, + "loss": 0.0081, + "step": 11217 + }, + { + "epoch": 1.216175195143105, + "grad_norm": 0.5642490983009338, + "learning_rate": 1.1935296677789064e-05, + "loss": 0.0047, + "step": 11218 + }, + { + "epoch": 1.2162836079791848, + "grad_norm": 0.5877495408058167, + "learning_rate": 1.1934571304221676e-05, + "loss": 0.0037, + "step": 11219 + }, + { + "epoch": 1.2163920208152645, + "grad_norm": 3.0847854614257812, + "learning_rate": 1.1933845930654288e-05, + "loss": 0.0271, + "step": 11220 + }, + { + "epoch": 1.2165004336513443, + "grad_norm": 1.7144527435302734, + "learning_rate": 1.19331205570869e-05, + "loss": 0.0216, + "step": 11221 + }, + { + "epoch": 1.216608846487424, + "grad_norm": 4.836950778961182, + "learning_rate": 1.1932395183519514e-05, + "loss": 0.027, + "step": 11222 + }, + { + "epoch": 1.2167172593235038, + "grad_norm": 8.946845054626465, + "learning_rate": 1.1931669809952127e-05, + "loss": 0.0362, + "step": 11223 + }, + { + "epoch": 1.2168256721595836, + "grad_norm": 1.799150824546814, + "learning_rate": 1.1930944436384739e-05, + "loss": 0.0076, + "step": 11224 + }, + { + "epoch": 1.2169340849956636, + "grad_norm": 0.5938818454742432, + "learning_rate": 1.1930219062817353e-05, + "loss": 0.0055, + "step": 11225 + }, + { + "epoch": 1.2170424978317433, + "grad_norm": 6.111761569976807, + "learning_rate": 1.1929493689249965e-05, + "loss": 0.059, + "step": 11226 + }, + { + "epoch": 1.217150910667823, + "grad_norm": 2.392953634262085, + "learning_rate": 1.1928768315682577e-05, + "loss": 0.0151, + "step": 11227 + }, + { + "epoch": 1.2172593235039029, + "grad_norm": 1.9552069902420044, + "learning_rate": 1.192804294211519e-05, + "loss": 0.026, + "step": 11228 + }, + { + "epoch": 1.2173677363399826, + "grad_norm": 4.995990753173828, + "learning_rate": 1.1927317568547803e-05, + "loss": 0.0756, + "step": 11229 + }, + { + "epoch": 1.2174761491760624, + "grad_norm": 0.7888889312744141, + "learning_rate": 1.1926592194980415e-05, + "loss": 0.0068, + "step": 11230 + }, + { + "epoch": 1.2175845620121422, + "grad_norm": 3.8362841606140137, + "learning_rate": 1.1925866821413029e-05, + "loss": 0.0483, + "step": 11231 + }, + { + "epoch": 1.2176929748482221, + "grad_norm": 2.1614253520965576, + "learning_rate": 1.1925141447845641e-05, + "loss": 0.0398, + "step": 11232 + }, + { + "epoch": 1.217801387684302, + "grad_norm": 3.4920685291290283, + "learning_rate": 1.1924416074278253e-05, + "loss": 0.014, + "step": 11233 + }, + { + "epoch": 1.2179098005203817, + "grad_norm": 0.2248363494873047, + "learning_rate": 1.1923690700710865e-05, + "loss": 0.0017, + "step": 11234 + }, + { + "epoch": 1.2180182133564614, + "grad_norm": 0.1975429505109787, + "learning_rate": 1.192296532714348e-05, + "loss": 0.0016, + "step": 11235 + }, + { + "epoch": 1.2181266261925412, + "grad_norm": 0.34364673495292664, + "learning_rate": 1.1922239953576093e-05, + "loss": 0.0041, + "step": 11236 + }, + { + "epoch": 1.218235039028621, + "grad_norm": 2.5367398262023926, + "learning_rate": 1.1921514580008707e-05, + "loss": 0.0213, + "step": 11237 + }, + { + "epoch": 1.2183434518647007, + "grad_norm": 1.7826277017593384, + "learning_rate": 1.1920789206441319e-05, + "loss": 0.0325, + "step": 11238 + }, + { + "epoch": 1.2184518647007805, + "grad_norm": 4.026008605957031, + "learning_rate": 1.1920063832873931e-05, + "loss": 0.0521, + "step": 11239 + }, + { + "epoch": 1.2185602775368602, + "grad_norm": 0.054373398423194885, + "learning_rate": 1.1919338459306545e-05, + "loss": 0.0007, + "step": 11240 + }, + { + "epoch": 1.2186686903729402, + "grad_norm": 4.646673202514648, + "learning_rate": 1.1918613085739157e-05, + "loss": 0.0373, + "step": 11241 + }, + { + "epoch": 1.21877710320902, + "grad_norm": 2.2720553874969482, + "learning_rate": 1.1917887712171769e-05, + "loss": 0.0275, + "step": 11242 + }, + { + "epoch": 1.2188855160450998, + "grad_norm": 0.3742404282093048, + "learning_rate": 1.1917162338604383e-05, + "loss": 0.0021, + "step": 11243 + }, + { + "epoch": 1.2189939288811795, + "grad_norm": 5.3039116859436035, + "learning_rate": 1.1916436965036995e-05, + "loss": 0.0112, + "step": 11244 + }, + { + "epoch": 1.2191023417172593, + "grad_norm": 1.1770086288452148, + "learning_rate": 1.1915711591469607e-05, + "loss": 0.0081, + "step": 11245 + }, + { + "epoch": 1.219210754553339, + "grad_norm": 2.9826955795288086, + "learning_rate": 1.191498621790222e-05, + "loss": 0.0183, + "step": 11246 + }, + { + "epoch": 1.2193191673894188, + "grad_norm": 1.4398186206817627, + "learning_rate": 1.1914260844334833e-05, + "loss": 0.012, + "step": 11247 + }, + { + "epoch": 1.2194275802254988, + "grad_norm": 0.8123365640640259, + "learning_rate": 1.1913535470767445e-05, + "loss": 0.0013, + "step": 11248 + }, + { + "epoch": 1.2195359930615786, + "grad_norm": 1.6746959686279297, + "learning_rate": 1.1912810097200058e-05, + "loss": 0.0273, + "step": 11249 + }, + { + "epoch": 1.2196444058976583, + "grad_norm": 3.136796712875366, + "learning_rate": 1.1912084723632671e-05, + "loss": 0.0155, + "step": 11250 + }, + { + "epoch": 1.219752818733738, + "grad_norm": 2.8711435794830322, + "learning_rate": 1.1911359350065284e-05, + "loss": 0.0267, + "step": 11251 + }, + { + "epoch": 1.2198612315698178, + "grad_norm": 1.163281798362732, + "learning_rate": 1.1910633976497899e-05, + "loss": 0.016, + "step": 11252 + }, + { + "epoch": 1.2199696444058976, + "grad_norm": 0.831619381904602, + "learning_rate": 1.1909908602930511e-05, + "loss": 0.0061, + "step": 11253 + }, + { + "epoch": 1.2200780572419774, + "grad_norm": 5.875749111175537, + "learning_rate": 1.1909183229363123e-05, + "loss": 0.0837, + "step": 11254 + }, + { + "epoch": 1.2201864700780574, + "grad_norm": 5.593197822570801, + "learning_rate": 1.1908457855795737e-05, + "loss": 0.0526, + "step": 11255 + }, + { + "epoch": 1.2202948829141371, + "grad_norm": 4.1752777099609375, + "learning_rate": 1.190773248222835e-05, + "loss": 0.0367, + "step": 11256 + }, + { + "epoch": 1.2204032957502169, + "grad_norm": 0.541658878326416, + "learning_rate": 1.1907007108660961e-05, + "loss": 0.0094, + "step": 11257 + }, + { + "epoch": 1.2205117085862967, + "grad_norm": 2.385648727416992, + "learning_rate": 1.1906281735093574e-05, + "loss": 0.0293, + "step": 11258 + }, + { + "epoch": 1.2206201214223764, + "grad_norm": 3.905632972717285, + "learning_rate": 1.1905556361526187e-05, + "loss": 0.0488, + "step": 11259 + }, + { + "epoch": 1.2207285342584562, + "grad_norm": 0.918711245059967, + "learning_rate": 1.19048309879588e-05, + "loss": 0.0105, + "step": 11260 + }, + { + "epoch": 1.220836947094536, + "grad_norm": 2.188291072845459, + "learning_rate": 1.1904105614391412e-05, + "loss": 0.0391, + "step": 11261 + }, + { + "epoch": 1.2209453599306157, + "grad_norm": 1.7330571413040161, + "learning_rate": 1.1903380240824026e-05, + "loss": 0.0316, + "step": 11262 + }, + { + "epoch": 1.2210537727666955, + "grad_norm": 1.7970333099365234, + "learning_rate": 1.1902654867256638e-05, + "loss": 0.0157, + "step": 11263 + }, + { + "epoch": 1.2211621856027755, + "grad_norm": 1.6631512641906738, + "learning_rate": 1.190192949368925e-05, + "loss": 0.017, + "step": 11264 + }, + { + "epoch": 1.2212705984388552, + "grad_norm": 3.7935709953308105, + "learning_rate": 1.1901204120121864e-05, + "loss": 0.0323, + "step": 11265 + }, + { + "epoch": 1.221379011274935, + "grad_norm": 0.715265154838562, + "learning_rate": 1.1900478746554476e-05, + "loss": 0.0086, + "step": 11266 + }, + { + "epoch": 1.2214874241110147, + "grad_norm": 1.7543259859085083, + "learning_rate": 1.1899753372987088e-05, + "loss": 0.0269, + "step": 11267 + }, + { + "epoch": 1.2215958369470945, + "grad_norm": 1.880845546722412, + "learning_rate": 1.1899027999419702e-05, + "loss": 0.0217, + "step": 11268 + }, + { + "epoch": 1.2217042497831743, + "grad_norm": 4.847179412841797, + "learning_rate": 1.1898302625852316e-05, + "loss": 0.0242, + "step": 11269 + }, + { + "epoch": 1.221812662619254, + "grad_norm": 3.940474271774292, + "learning_rate": 1.189757725228493e-05, + "loss": 0.0441, + "step": 11270 + }, + { + "epoch": 1.221921075455334, + "grad_norm": 0.6923159956932068, + "learning_rate": 1.1896851878717542e-05, + "loss": 0.0063, + "step": 11271 + }, + { + "epoch": 1.2220294882914138, + "grad_norm": 2.082714796066284, + "learning_rate": 1.1896126505150154e-05, + "loss": 0.0079, + "step": 11272 + }, + { + "epoch": 1.2221379011274935, + "grad_norm": 1.0875358581542969, + "learning_rate": 1.1895401131582766e-05, + "loss": 0.011, + "step": 11273 + }, + { + "epoch": 1.2222463139635733, + "grad_norm": 3.138864755630493, + "learning_rate": 1.189467575801538e-05, + "loss": 0.0588, + "step": 11274 + }, + { + "epoch": 1.222354726799653, + "grad_norm": 0.9344825744628906, + "learning_rate": 1.1893950384447992e-05, + "loss": 0.0085, + "step": 11275 + }, + { + "epoch": 1.2224631396357328, + "grad_norm": 0.8862912654876709, + "learning_rate": 1.1893225010880604e-05, + "loss": 0.01, + "step": 11276 + }, + { + "epoch": 1.2225715524718126, + "grad_norm": 0.7794812321662903, + "learning_rate": 1.1892499637313218e-05, + "loss": 0.0164, + "step": 11277 + }, + { + "epoch": 1.2226799653078924, + "grad_norm": 4.319443225860596, + "learning_rate": 1.189177426374583e-05, + "loss": 0.1092, + "step": 11278 + }, + { + "epoch": 1.2227883781439723, + "grad_norm": 0.5903160572052002, + "learning_rate": 1.1891048890178442e-05, + "loss": 0.0058, + "step": 11279 + }, + { + "epoch": 1.222896790980052, + "grad_norm": 1.3474838733673096, + "learning_rate": 1.1890323516611056e-05, + "loss": 0.0095, + "step": 11280 + }, + { + "epoch": 1.2230052038161319, + "grad_norm": 1.7763543128967285, + "learning_rate": 1.1889598143043668e-05, + "loss": 0.04, + "step": 11281 + }, + { + "epoch": 1.2231136166522116, + "grad_norm": 5.574120998382568, + "learning_rate": 1.188887276947628e-05, + "loss": 0.0213, + "step": 11282 + }, + { + "epoch": 1.2232220294882914, + "grad_norm": 0.9984109401702881, + "learning_rate": 1.1888147395908894e-05, + "loss": 0.0113, + "step": 11283 + }, + { + "epoch": 1.2233304423243712, + "grad_norm": 0.7481037378311157, + "learning_rate": 1.1887422022341506e-05, + "loss": 0.0055, + "step": 11284 + }, + { + "epoch": 1.223438855160451, + "grad_norm": 0.865178644657135, + "learning_rate": 1.1886696648774118e-05, + "loss": 0.018, + "step": 11285 + }, + { + "epoch": 1.2235472679965307, + "grad_norm": 0.8019473552703857, + "learning_rate": 1.188597127520673e-05, + "loss": 0.0072, + "step": 11286 + }, + { + "epoch": 1.2236556808326107, + "grad_norm": 2.654423236846924, + "learning_rate": 1.1885245901639346e-05, + "loss": 0.0077, + "step": 11287 + }, + { + "epoch": 1.2237640936686904, + "grad_norm": 5.323254585266113, + "learning_rate": 1.1884520528071958e-05, + "loss": 0.0434, + "step": 11288 + }, + { + "epoch": 1.2238725065047702, + "grad_norm": 0.9922401905059814, + "learning_rate": 1.1883795154504572e-05, + "loss": 0.0111, + "step": 11289 + }, + { + "epoch": 1.22398091934085, + "grad_norm": 0.4025920629501343, + "learning_rate": 1.1883069780937184e-05, + "loss": 0.0016, + "step": 11290 + }, + { + "epoch": 1.2240893321769297, + "grad_norm": 1.07063627243042, + "learning_rate": 1.1882344407369796e-05, + "loss": 0.0155, + "step": 11291 + }, + { + "epoch": 1.2241977450130095, + "grad_norm": 1.6150892972946167, + "learning_rate": 1.188161903380241e-05, + "loss": 0.0192, + "step": 11292 + }, + { + "epoch": 1.2243061578490892, + "grad_norm": 1.3485671281814575, + "learning_rate": 1.1880893660235022e-05, + "loss": 0.0193, + "step": 11293 + }, + { + "epoch": 1.2244145706851692, + "grad_norm": 1.5580554008483887, + "learning_rate": 1.1880168286667634e-05, + "loss": 0.0108, + "step": 11294 + }, + { + "epoch": 1.224522983521249, + "grad_norm": 0.262788325548172, + "learning_rate": 1.1879442913100248e-05, + "loss": 0.0016, + "step": 11295 + }, + { + "epoch": 1.2246313963573288, + "grad_norm": 1.688020944595337, + "learning_rate": 1.187871753953286e-05, + "loss": 0.0094, + "step": 11296 + }, + { + "epoch": 1.2247398091934085, + "grad_norm": 3.1115102767944336, + "learning_rate": 1.1877992165965473e-05, + "loss": 0.0409, + "step": 11297 + }, + { + "epoch": 1.2248482220294883, + "grad_norm": 1.3680105209350586, + "learning_rate": 1.1877266792398085e-05, + "loss": 0.0131, + "step": 11298 + }, + { + "epoch": 1.224956634865568, + "grad_norm": 5.511824607849121, + "learning_rate": 1.1876541418830699e-05, + "loss": 0.0075, + "step": 11299 + }, + { + "epoch": 1.2250650477016478, + "grad_norm": 1.6344355344772339, + "learning_rate": 1.187581604526331e-05, + "loss": 0.0127, + "step": 11300 + }, + { + "epoch": 1.2251734605377276, + "grad_norm": 2.1417734622955322, + "learning_rate": 1.1875090671695923e-05, + "loss": 0.0337, + "step": 11301 + }, + { + "epoch": 1.2252818733738073, + "grad_norm": 2.1834824085235596, + "learning_rate": 1.1874365298128537e-05, + "loss": 0.04, + "step": 11302 + }, + { + "epoch": 1.2253902862098873, + "grad_norm": 2.0766234397888184, + "learning_rate": 1.1873639924561149e-05, + "loss": 0.0091, + "step": 11303 + }, + { + "epoch": 1.225498699045967, + "grad_norm": 0.9136611223220825, + "learning_rate": 1.1872914550993764e-05, + "loss": 0.0053, + "step": 11304 + }, + { + "epoch": 1.2256071118820469, + "grad_norm": 2.2830371856689453, + "learning_rate": 1.1872189177426377e-05, + "loss": 0.0305, + "step": 11305 + }, + { + "epoch": 1.2257155247181266, + "grad_norm": 1.180869460105896, + "learning_rate": 1.1871463803858989e-05, + "loss": 0.0355, + "step": 11306 + }, + { + "epoch": 1.2258239375542064, + "grad_norm": 3.168294906616211, + "learning_rate": 1.1870738430291603e-05, + "loss": 0.0054, + "step": 11307 + }, + { + "epoch": 1.2259323503902861, + "grad_norm": 3.5462534427642822, + "learning_rate": 1.1870013056724215e-05, + "loss": 0.0083, + "step": 11308 + }, + { + "epoch": 1.226040763226366, + "grad_norm": 2.513456106185913, + "learning_rate": 1.1869287683156827e-05, + "loss": 0.0441, + "step": 11309 + }, + { + "epoch": 1.226149176062446, + "grad_norm": 1.8261414766311646, + "learning_rate": 1.1868562309589439e-05, + "loss": 0.0179, + "step": 11310 + }, + { + "epoch": 1.2262575888985257, + "grad_norm": 2.652709484100342, + "learning_rate": 1.1867836936022053e-05, + "loss": 0.0284, + "step": 11311 + }, + { + "epoch": 1.2263660017346054, + "grad_norm": 1.7069838047027588, + "learning_rate": 1.1867111562454665e-05, + "loss": 0.0239, + "step": 11312 + }, + { + "epoch": 1.2264744145706852, + "grad_norm": 6.2696943283081055, + "learning_rate": 1.1866386188887277e-05, + "loss": 0.0559, + "step": 11313 + }, + { + "epoch": 1.226582827406765, + "grad_norm": 4.090828895568848, + "learning_rate": 1.1865660815319891e-05, + "loss": 0.0347, + "step": 11314 + }, + { + "epoch": 1.2266912402428447, + "grad_norm": 3.1729648113250732, + "learning_rate": 1.1864935441752503e-05, + "loss": 0.0308, + "step": 11315 + }, + { + "epoch": 1.2267996530789245, + "grad_norm": 1.0350786447525024, + "learning_rate": 1.1864210068185115e-05, + "loss": 0.0101, + "step": 11316 + }, + { + "epoch": 1.2269080659150045, + "grad_norm": 4.470774173736572, + "learning_rate": 1.1863484694617729e-05, + "loss": 0.0088, + "step": 11317 + }, + { + "epoch": 1.2270164787510842, + "grad_norm": 2.9218430519104004, + "learning_rate": 1.1862759321050341e-05, + "loss": 0.0347, + "step": 11318 + }, + { + "epoch": 1.227124891587164, + "grad_norm": 5.767371654510498, + "learning_rate": 1.1862033947482953e-05, + "loss": 0.0387, + "step": 11319 + }, + { + "epoch": 1.2272333044232437, + "grad_norm": 2.9370853900909424, + "learning_rate": 1.1861308573915567e-05, + "loss": 0.0177, + "step": 11320 + }, + { + "epoch": 1.2273417172593235, + "grad_norm": 4.033493518829346, + "learning_rate": 1.1860583200348181e-05, + "loss": 0.025, + "step": 11321 + }, + { + "epoch": 1.2274501300954033, + "grad_norm": 4.597550868988037, + "learning_rate": 1.1859857826780795e-05, + "loss": 0.0535, + "step": 11322 + }, + { + "epoch": 1.227558542931483, + "grad_norm": 0.6606891751289368, + "learning_rate": 1.1859132453213407e-05, + "loss": 0.0051, + "step": 11323 + }, + { + "epoch": 1.2276669557675628, + "grad_norm": 7.77566385269165, + "learning_rate": 1.1858407079646019e-05, + "loss": 0.0533, + "step": 11324 + }, + { + "epoch": 1.2277753686036426, + "grad_norm": 2.2703702449798584, + "learning_rate": 1.1857681706078631e-05, + "loss": 0.0186, + "step": 11325 + }, + { + "epoch": 1.2278837814397225, + "grad_norm": 0.07882317155599594, + "learning_rate": 1.1856956332511245e-05, + "loss": 0.0004, + "step": 11326 + }, + { + "epoch": 1.2279921942758023, + "grad_norm": 1.8505936861038208, + "learning_rate": 1.1856230958943857e-05, + "loss": 0.0176, + "step": 11327 + }, + { + "epoch": 1.228100607111882, + "grad_norm": 0.647451639175415, + "learning_rate": 1.185550558537647e-05, + "loss": 0.0034, + "step": 11328 + }, + { + "epoch": 1.2282090199479618, + "grad_norm": 0.6557692289352417, + "learning_rate": 1.1854780211809083e-05, + "loss": 0.0038, + "step": 11329 + }, + { + "epoch": 1.2283174327840416, + "grad_norm": 6.905179500579834, + "learning_rate": 1.1854054838241695e-05, + "loss": 0.023, + "step": 11330 + }, + { + "epoch": 1.2284258456201214, + "grad_norm": 7.0530171394348145, + "learning_rate": 1.1853329464674307e-05, + "loss": 0.0379, + "step": 11331 + }, + { + "epoch": 1.2285342584562011, + "grad_norm": 0.9074395895004272, + "learning_rate": 1.1852604091106921e-05, + "loss": 0.0077, + "step": 11332 + }, + { + "epoch": 1.228642671292281, + "grad_norm": 2.0074994564056396, + "learning_rate": 1.1851878717539533e-05, + "loss": 0.0278, + "step": 11333 + }, + { + "epoch": 1.2287510841283609, + "grad_norm": 2.228957176208496, + "learning_rate": 1.1851153343972146e-05, + "loss": 0.0275, + "step": 11334 + }, + { + "epoch": 1.2288594969644406, + "grad_norm": 2.0326461791992188, + "learning_rate": 1.185042797040476e-05, + "loss": 0.0798, + "step": 11335 + }, + { + "epoch": 1.2289679098005204, + "grad_norm": 0.4944785237312317, + "learning_rate": 1.1849702596837372e-05, + "loss": 0.0034, + "step": 11336 + }, + { + "epoch": 1.2290763226366002, + "grad_norm": 2.958433151245117, + "learning_rate": 1.1848977223269984e-05, + "loss": 0.0354, + "step": 11337 + }, + { + "epoch": 1.22918473547268, + "grad_norm": 5.84275484085083, + "learning_rate": 1.18482518497026e-05, + "loss": 0.0254, + "step": 11338 + }, + { + "epoch": 1.2292931483087597, + "grad_norm": 3.8744730949401855, + "learning_rate": 1.1847526476135211e-05, + "loss": 0.0858, + "step": 11339 + }, + { + "epoch": 1.2294015611448397, + "grad_norm": 0.15669089555740356, + "learning_rate": 1.1846801102567824e-05, + "loss": 0.0011, + "step": 11340 + }, + { + "epoch": 1.2295099739809194, + "grad_norm": 1.2993881702423096, + "learning_rate": 1.1846075729000437e-05, + "loss": 0.0147, + "step": 11341 + }, + { + "epoch": 1.2296183868169992, + "grad_norm": 4.170961856842041, + "learning_rate": 1.184535035543305e-05, + "loss": 0.0615, + "step": 11342 + }, + { + "epoch": 1.229726799653079, + "grad_norm": 2.063668966293335, + "learning_rate": 1.1844624981865662e-05, + "loss": 0.0203, + "step": 11343 + }, + { + "epoch": 1.2298352124891587, + "grad_norm": 0.3465977609157562, + "learning_rate": 1.1843899608298276e-05, + "loss": 0.0035, + "step": 11344 + }, + { + "epoch": 1.2299436253252385, + "grad_norm": 4.485915660858154, + "learning_rate": 1.1843174234730888e-05, + "loss": 0.0497, + "step": 11345 + }, + { + "epoch": 1.2300520381613183, + "grad_norm": 7.503397464752197, + "learning_rate": 1.18424488611635e-05, + "loss": 0.027, + "step": 11346 + }, + { + "epoch": 1.230160450997398, + "grad_norm": 0.5470507144927979, + "learning_rate": 1.1841723487596114e-05, + "loss": 0.0018, + "step": 11347 + }, + { + "epoch": 1.2302688638334778, + "grad_norm": 1.0539772510528564, + "learning_rate": 1.1840998114028726e-05, + "loss": 0.0045, + "step": 11348 + }, + { + "epoch": 1.2303772766695578, + "grad_norm": 7.43709659576416, + "learning_rate": 1.1840272740461338e-05, + "loss": 0.0695, + "step": 11349 + }, + { + "epoch": 1.2304856895056375, + "grad_norm": 1.0458756685256958, + "learning_rate": 1.183954736689395e-05, + "loss": 0.0077, + "step": 11350 + }, + { + "epoch": 1.2305941023417173, + "grad_norm": 1.6878207921981812, + "learning_rate": 1.1838821993326564e-05, + "loss": 0.0164, + "step": 11351 + }, + { + "epoch": 1.230702515177797, + "grad_norm": 2.925837516784668, + "learning_rate": 1.1838096619759176e-05, + "loss": 0.0283, + "step": 11352 + }, + { + "epoch": 1.2308109280138768, + "grad_norm": 3.886037588119507, + "learning_rate": 1.1837371246191788e-05, + "loss": 0.0352, + "step": 11353 + }, + { + "epoch": 1.2309193408499566, + "grad_norm": 1.246667504310608, + "learning_rate": 1.1836645872624402e-05, + "loss": 0.0149, + "step": 11354 + }, + { + "epoch": 1.2310277536860363, + "grad_norm": 0.3133998513221741, + "learning_rate": 1.1835920499057014e-05, + "loss": 0.0028, + "step": 11355 + }, + { + "epoch": 1.2311361665221163, + "grad_norm": 0.24627436697483063, + "learning_rate": 1.183519512548963e-05, + "loss": 0.0016, + "step": 11356 + }, + { + "epoch": 1.231244579358196, + "grad_norm": 0.6330596208572388, + "learning_rate": 1.1834469751922242e-05, + "loss": 0.0025, + "step": 11357 + }, + { + "epoch": 1.2313529921942759, + "grad_norm": 2.2428464889526367, + "learning_rate": 1.1833744378354854e-05, + "loss": 0.0294, + "step": 11358 + }, + { + "epoch": 1.2314614050303556, + "grad_norm": 0.9309895634651184, + "learning_rate": 1.1833019004787468e-05, + "loss": 0.0085, + "step": 11359 + }, + { + "epoch": 1.2315698178664354, + "grad_norm": 4.8980326652526855, + "learning_rate": 1.183229363122008e-05, + "loss": 0.0191, + "step": 11360 + }, + { + "epoch": 1.2316782307025151, + "grad_norm": 0.47512587904930115, + "learning_rate": 1.1831568257652692e-05, + "loss": 0.0017, + "step": 11361 + }, + { + "epoch": 1.231786643538595, + "grad_norm": 1.1783452033996582, + "learning_rate": 1.1830842884085306e-05, + "loss": 0.0049, + "step": 11362 + }, + { + "epoch": 1.2318950563746747, + "grad_norm": 3.7093591690063477, + "learning_rate": 1.1830117510517918e-05, + "loss": 0.021, + "step": 11363 + }, + { + "epoch": 1.2320034692107544, + "grad_norm": 1.536479115486145, + "learning_rate": 1.182939213695053e-05, + "loss": 0.017, + "step": 11364 + }, + { + "epoch": 1.2321118820468344, + "grad_norm": 2.701570987701416, + "learning_rate": 1.1828666763383142e-05, + "loss": 0.0092, + "step": 11365 + }, + { + "epoch": 1.2322202948829142, + "grad_norm": 3.4540464878082275, + "learning_rate": 1.1827941389815756e-05, + "loss": 0.0196, + "step": 11366 + }, + { + "epoch": 1.232328707718994, + "grad_norm": 11.577061653137207, + "learning_rate": 1.1827216016248368e-05, + "loss": 0.0358, + "step": 11367 + }, + { + "epoch": 1.2324371205550737, + "grad_norm": 2.077003002166748, + "learning_rate": 1.182649064268098e-05, + "loss": 0.0921, + "step": 11368 + }, + { + "epoch": 1.2325455333911535, + "grad_norm": 8.396980285644531, + "learning_rate": 1.1825765269113594e-05, + "loss": 0.0201, + "step": 11369 + }, + { + "epoch": 1.2326539462272332, + "grad_norm": 1.9353504180908203, + "learning_rate": 1.1825039895546206e-05, + "loss": 0.0213, + "step": 11370 + }, + { + "epoch": 1.232762359063313, + "grad_norm": 1.8930537700653076, + "learning_rate": 1.1824314521978819e-05, + "loss": 0.0072, + "step": 11371 + }, + { + "epoch": 1.232870771899393, + "grad_norm": 0.8293355107307434, + "learning_rate": 1.1823589148411432e-05, + "loss": 0.0081, + "step": 11372 + }, + { + "epoch": 1.2329791847354727, + "grad_norm": 0.5141850709915161, + "learning_rate": 1.1822863774844046e-05, + "loss": 0.0018, + "step": 11373 + }, + { + "epoch": 1.2330875975715525, + "grad_norm": 1.678640365600586, + "learning_rate": 1.182213840127666e-05, + "loss": 0.0321, + "step": 11374 + }, + { + "epoch": 1.2331960104076323, + "grad_norm": 0.6700099110603333, + "learning_rate": 1.1821413027709272e-05, + "loss": 0.0068, + "step": 11375 + }, + { + "epoch": 1.233304423243712, + "grad_norm": 3.56257700920105, + "learning_rate": 1.1820687654141884e-05, + "loss": 0.0877, + "step": 11376 + }, + { + "epoch": 1.2334128360797918, + "grad_norm": 3.645672559738159, + "learning_rate": 1.1819962280574497e-05, + "loss": 0.0186, + "step": 11377 + }, + { + "epoch": 1.2335212489158716, + "grad_norm": 0.7846016883850098, + "learning_rate": 1.181923690700711e-05, + "loss": 0.0169, + "step": 11378 + }, + { + "epoch": 1.2336296617519515, + "grad_norm": 3.8703482151031494, + "learning_rate": 1.1818511533439723e-05, + "loss": 0.0471, + "step": 11379 + }, + { + "epoch": 1.2337380745880313, + "grad_norm": 0.6158815026283264, + "learning_rate": 1.1817786159872335e-05, + "loss": 0.0027, + "step": 11380 + }, + { + "epoch": 1.233846487424111, + "grad_norm": 1.5006051063537598, + "learning_rate": 1.1817060786304949e-05, + "loss": 0.0271, + "step": 11381 + }, + { + "epoch": 1.2339549002601908, + "grad_norm": 1.9355885982513428, + "learning_rate": 1.181633541273756e-05, + "loss": 0.0189, + "step": 11382 + }, + { + "epoch": 1.2340633130962706, + "grad_norm": 1.8988962173461914, + "learning_rate": 1.1815610039170173e-05, + "loss": 0.0077, + "step": 11383 + }, + { + "epoch": 1.2341717259323504, + "grad_norm": 1.7833846807479858, + "learning_rate": 1.1814884665602787e-05, + "loss": 0.0036, + "step": 11384 + }, + { + "epoch": 1.2342801387684301, + "grad_norm": 0.9203770756721497, + "learning_rate": 1.1814159292035399e-05, + "loss": 0.0041, + "step": 11385 + }, + { + "epoch": 1.2343885516045099, + "grad_norm": 1.9775996208190918, + "learning_rate": 1.1813433918468011e-05, + "loss": 0.0255, + "step": 11386 + }, + { + "epoch": 1.2344969644405897, + "grad_norm": 5.180322647094727, + "learning_rate": 1.1812708544900625e-05, + "loss": 0.0713, + "step": 11387 + }, + { + "epoch": 1.2346053772766696, + "grad_norm": 2.619947671890259, + "learning_rate": 1.1811983171333237e-05, + "loss": 0.0205, + "step": 11388 + }, + { + "epoch": 1.2347137901127494, + "grad_norm": 2.4295949935913086, + "learning_rate": 1.1811257797765849e-05, + "loss": 0.0232, + "step": 11389 + }, + { + "epoch": 1.2348222029488292, + "grad_norm": 3.906369924545288, + "learning_rate": 1.1810532424198465e-05, + "loss": 0.0682, + "step": 11390 + }, + { + "epoch": 1.234930615784909, + "grad_norm": 5.57454776763916, + "learning_rate": 1.1809807050631077e-05, + "loss": 0.0331, + "step": 11391 + }, + { + "epoch": 1.2350390286209887, + "grad_norm": 1.997550129890442, + "learning_rate": 1.1809081677063689e-05, + "loss": 0.0105, + "step": 11392 + }, + { + "epoch": 1.2351474414570685, + "grad_norm": 6.683823108673096, + "learning_rate": 1.1808356303496303e-05, + "loss": 0.0145, + "step": 11393 + }, + { + "epoch": 1.2352558542931482, + "grad_norm": 1.882739782333374, + "learning_rate": 1.1807630929928915e-05, + "loss": 0.0096, + "step": 11394 + }, + { + "epoch": 1.2353642671292282, + "grad_norm": 0.2630617320537567, + "learning_rate": 1.1806905556361527e-05, + "loss": 0.0026, + "step": 11395 + }, + { + "epoch": 1.235472679965308, + "grad_norm": 0.6537471413612366, + "learning_rate": 1.1806180182794141e-05, + "loss": 0.0175, + "step": 11396 + }, + { + "epoch": 1.2355810928013877, + "grad_norm": 2.8787097930908203, + "learning_rate": 1.1805454809226753e-05, + "loss": 0.0575, + "step": 11397 + }, + { + "epoch": 1.2356895056374675, + "grad_norm": 0.5769741535186768, + "learning_rate": 1.1804729435659365e-05, + "loss": 0.0036, + "step": 11398 + }, + { + "epoch": 1.2357979184735473, + "grad_norm": 2.472036123275757, + "learning_rate": 1.1804004062091979e-05, + "loss": 0.0351, + "step": 11399 + }, + { + "epoch": 1.235906331309627, + "grad_norm": 4.239490032196045, + "learning_rate": 1.1803278688524591e-05, + "loss": 0.0232, + "step": 11400 + }, + { + "epoch": 1.2360147441457068, + "grad_norm": 3.7110602855682373, + "learning_rate": 1.1802553314957203e-05, + "loss": 0.0199, + "step": 11401 + }, + { + "epoch": 1.2361231569817868, + "grad_norm": 1.7259160280227661, + "learning_rate": 1.1801827941389815e-05, + "loss": 0.0214, + "step": 11402 + }, + { + "epoch": 1.2362315698178665, + "grad_norm": 1.728865146636963, + "learning_rate": 1.180110256782243e-05, + "loss": 0.0132, + "step": 11403 + }, + { + "epoch": 1.2363399826539463, + "grad_norm": 6.498849391937256, + "learning_rate": 1.1800377194255041e-05, + "loss": 0.0283, + "step": 11404 + }, + { + "epoch": 1.236448395490026, + "grad_norm": 7.09821081161499, + "learning_rate": 1.1799651820687654e-05, + "loss": 0.0328, + "step": 11405 + }, + { + "epoch": 1.2365568083261058, + "grad_norm": 1.5570188760757446, + "learning_rate": 1.1798926447120267e-05, + "loss": 0.0127, + "step": 11406 + }, + { + "epoch": 1.2366652211621856, + "grad_norm": 3.2428174018859863, + "learning_rate": 1.1798201073552881e-05, + "loss": 0.0096, + "step": 11407 + }, + { + "epoch": 1.2367736339982653, + "grad_norm": 6.09340238571167, + "learning_rate": 1.1797475699985495e-05, + "loss": 0.0914, + "step": 11408 + }, + { + "epoch": 1.236882046834345, + "grad_norm": 1.1333624124526978, + "learning_rate": 1.1796750326418107e-05, + "loss": 0.0147, + "step": 11409 + }, + { + "epoch": 1.2369904596704249, + "grad_norm": 0.19675827026367188, + "learning_rate": 1.179602495285072e-05, + "loss": 0.0022, + "step": 11410 + }, + { + "epoch": 1.2370988725065049, + "grad_norm": 0.1494835764169693, + "learning_rate": 1.1795299579283333e-05, + "loss": 0.0013, + "step": 11411 + }, + { + "epoch": 1.2372072853425846, + "grad_norm": 1.4718416929244995, + "learning_rate": 1.1794574205715945e-05, + "loss": 0.0158, + "step": 11412 + }, + { + "epoch": 1.2373156981786644, + "grad_norm": 1.6973236799240112, + "learning_rate": 1.1793848832148557e-05, + "loss": 0.019, + "step": 11413 + }, + { + "epoch": 1.2374241110147441, + "grad_norm": 3.001767635345459, + "learning_rate": 1.1793123458581171e-05, + "loss": 0.011, + "step": 11414 + }, + { + "epoch": 1.237532523850824, + "grad_norm": 1.9026620388031006, + "learning_rate": 1.1792398085013783e-05, + "loss": 0.0159, + "step": 11415 + }, + { + "epoch": 1.2376409366869037, + "grad_norm": 1.2442986965179443, + "learning_rate": 1.1791672711446396e-05, + "loss": 0.016, + "step": 11416 + }, + { + "epoch": 1.2377493495229834, + "grad_norm": 1.7361786365509033, + "learning_rate": 1.1790947337879008e-05, + "loss": 0.0217, + "step": 11417 + }, + { + "epoch": 1.2378577623590634, + "grad_norm": 12.018014907836914, + "learning_rate": 1.1790221964311622e-05, + "loss": 0.0265, + "step": 11418 + }, + { + "epoch": 1.2379661751951432, + "grad_norm": 0.38137826323509216, + "learning_rate": 1.1789496590744234e-05, + "loss": 0.0018, + "step": 11419 + }, + { + "epoch": 1.238074588031223, + "grad_norm": 2.562875747680664, + "learning_rate": 1.1788771217176846e-05, + "loss": 0.0493, + "step": 11420 + }, + { + "epoch": 1.2381830008673027, + "grad_norm": 1.6857736110687256, + "learning_rate": 1.178804584360946e-05, + "loss": 0.0163, + "step": 11421 + }, + { + "epoch": 1.2382914137033825, + "grad_norm": 7.812307834625244, + "learning_rate": 1.1787320470042072e-05, + "loss": 0.0666, + "step": 11422 + }, + { + "epoch": 1.2383998265394622, + "grad_norm": 0.6823754906654358, + "learning_rate": 1.1786595096474684e-05, + "loss": 0.0052, + "step": 11423 + }, + { + "epoch": 1.238508239375542, + "grad_norm": 3.6793034076690674, + "learning_rate": 1.1785869722907298e-05, + "loss": 0.0447, + "step": 11424 + }, + { + "epoch": 1.2386166522116218, + "grad_norm": 1.5447691679000854, + "learning_rate": 1.1785144349339912e-05, + "loss": 0.0093, + "step": 11425 + }, + { + "epoch": 1.2387250650477015, + "grad_norm": 0.10144198685884476, + "learning_rate": 1.1784418975772525e-05, + "loss": 0.001, + "step": 11426 + }, + { + "epoch": 1.2388334778837815, + "grad_norm": 2.696298837661743, + "learning_rate": 1.1783693602205138e-05, + "loss": 0.0264, + "step": 11427 + }, + { + "epoch": 1.2389418907198613, + "grad_norm": 3.481113910675049, + "learning_rate": 1.178296822863775e-05, + "loss": 0.0362, + "step": 11428 + }, + { + "epoch": 1.239050303555941, + "grad_norm": 1.802541971206665, + "learning_rate": 1.1782242855070362e-05, + "loss": 0.0109, + "step": 11429 + }, + { + "epoch": 1.2391587163920208, + "grad_norm": 1.2911251783370972, + "learning_rate": 1.1781517481502976e-05, + "loss": 0.0093, + "step": 11430 + }, + { + "epoch": 1.2392671292281006, + "grad_norm": 3.0227694511413574, + "learning_rate": 1.1780792107935588e-05, + "loss": 0.0097, + "step": 11431 + }, + { + "epoch": 1.2393755420641803, + "grad_norm": 0.9753325581550598, + "learning_rate": 1.17800667343682e-05, + "loss": 0.0146, + "step": 11432 + }, + { + "epoch": 1.23948395490026, + "grad_norm": 6.712053298950195, + "learning_rate": 1.1779341360800814e-05, + "loss": 0.0441, + "step": 11433 + }, + { + "epoch": 1.23959236773634, + "grad_norm": 3.1280133724212646, + "learning_rate": 1.1778615987233426e-05, + "loss": 0.0276, + "step": 11434 + }, + { + "epoch": 1.2397007805724198, + "grad_norm": 1.3443399667739868, + "learning_rate": 1.1777890613666038e-05, + "loss": 0.0038, + "step": 11435 + }, + { + "epoch": 1.2398091934084996, + "grad_norm": 1.7209036350250244, + "learning_rate": 1.1777165240098652e-05, + "loss": 0.0338, + "step": 11436 + }, + { + "epoch": 1.2399176062445794, + "grad_norm": 1.682844638824463, + "learning_rate": 1.1776439866531264e-05, + "loss": 0.0071, + "step": 11437 + }, + { + "epoch": 1.2400260190806591, + "grad_norm": 7.207466125488281, + "learning_rate": 1.1775714492963876e-05, + "loss": 0.051, + "step": 11438 + }, + { + "epoch": 1.2401344319167389, + "grad_norm": 2.362234592437744, + "learning_rate": 1.177498911939649e-05, + "loss": 0.0183, + "step": 11439 + }, + { + "epoch": 1.2402428447528187, + "grad_norm": 0.86563640832901, + "learning_rate": 1.1774263745829102e-05, + "loss": 0.018, + "step": 11440 + }, + { + "epoch": 1.2403512575888986, + "grad_norm": 3.196321725845337, + "learning_rate": 1.1773538372261714e-05, + "loss": 0.0247, + "step": 11441 + }, + { + "epoch": 1.2404596704249784, + "grad_norm": 1.2546815872192383, + "learning_rate": 1.177281299869433e-05, + "loss": 0.0276, + "step": 11442 + }, + { + "epoch": 1.2405680832610582, + "grad_norm": 2.87628173828125, + "learning_rate": 1.1772087625126942e-05, + "loss": 0.0226, + "step": 11443 + }, + { + "epoch": 1.240676496097138, + "grad_norm": 0.3318707048892975, + "learning_rate": 1.1771362251559554e-05, + "loss": 0.0038, + "step": 11444 + }, + { + "epoch": 1.2407849089332177, + "grad_norm": 2.43548846244812, + "learning_rate": 1.1770636877992168e-05, + "loss": 0.0232, + "step": 11445 + }, + { + "epoch": 1.2408933217692975, + "grad_norm": 5.741044044494629, + "learning_rate": 1.176991150442478e-05, + "loss": 0.0163, + "step": 11446 + }, + { + "epoch": 1.2410017346053772, + "grad_norm": 0.9741248488426208, + "learning_rate": 1.1769186130857392e-05, + "loss": 0.0043, + "step": 11447 + }, + { + "epoch": 1.241110147441457, + "grad_norm": 0.23143407702445984, + "learning_rate": 1.1768460757290006e-05, + "loss": 0.0014, + "step": 11448 + }, + { + "epoch": 1.2412185602775367, + "grad_norm": 0.9558852910995483, + "learning_rate": 1.1767735383722618e-05, + "loss": 0.0036, + "step": 11449 + }, + { + "epoch": 1.2413269731136167, + "grad_norm": 0.15362277626991272, + "learning_rate": 1.176701001015523e-05, + "loss": 0.0014, + "step": 11450 + }, + { + "epoch": 1.2414353859496965, + "grad_norm": 0.9995386004447937, + "learning_rate": 1.1766284636587844e-05, + "loss": 0.0066, + "step": 11451 + }, + { + "epoch": 1.2415437987857763, + "grad_norm": 1.4767369031906128, + "learning_rate": 1.1765559263020456e-05, + "loss": 0.0058, + "step": 11452 + }, + { + "epoch": 1.241652211621856, + "grad_norm": 4.828004837036133, + "learning_rate": 1.1764833889453069e-05, + "loss": 0.0375, + "step": 11453 + }, + { + "epoch": 1.2417606244579358, + "grad_norm": 0.3634689152240753, + "learning_rate": 1.176410851588568e-05, + "loss": 0.0039, + "step": 11454 + }, + { + "epoch": 1.2418690372940155, + "grad_norm": 0.8697302937507629, + "learning_rate": 1.1763383142318295e-05, + "loss": 0.0036, + "step": 11455 + }, + { + "epoch": 1.2419774501300953, + "grad_norm": 3.908069610595703, + "learning_rate": 1.1762657768750907e-05, + "loss": 0.0214, + "step": 11456 + }, + { + "epoch": 1.2420858629661753, + "grad_norm": 0.6763132214546204, + "learning_rate": 1.1761932395183519e-05, + "loss": 0.0133, + "step": 11457 + }, + { + "epoch": 1.242194275802255, + "grad_norm": 0.20947754383087158, + "learning_rate": 1.1761207021616133e-05, + "loss": 0.0024, + "step": 11458 + }, + { + "epoch": 1.2423026886383348, + "grad_norm": 3.2694315910339355, + "learning_rate": 1.1760481648048747e-05, + "loss": 0.0191, + "step": 11459 + }, + { + "epoch": 1.2424111014744146, + "grad_norm": 0.33844056725502014, + "learning_rate": 1.175975627448136e-05, + "loss": 0.0035, + "step": 11460 + }, + { + "epoch": 1.2425195143104943, + "grad_norm": 3.9414613246917725, + "learning_rate": 1.1759030900913973e-05, + "loss": 0.0242, + "step": 11461 + }, + { + "epoch": 1.242627927146574, + "grad_norm": 2.4020256996154785, + "learning_rate": 1.1758305527346585e-05, + "loss": 0.0371, + "step": 11462 + }, + { + "epoch": 1.2427363399826539, + "grad_norm": 2.8017306327819824, + "learning_rate": 1.1757580153779199e-05, + "loss": 0.0393, + "step": 11463 + }, + { + "epoch": 1.2428447528187339, + "grad_norm": 3.7687439918518066, + "learning_rate": 1.175685478021181e-05, + "loss": 0.006, + "step": 11464 + }, + { + "epoch": 1.2429531656548136, + "grad_norm": 1.9097293615341187, + "learning_rate": 1.1756129406644423e-05, + "loss": 0.0221, + "step": 11465 + }, + { + "epoch": 1.2430615784908934, + "grad_norm": 1.3242278099060059, + "learning_rate": 1.1755404033077037e-05, + "loss": 0.0373, + "step": 11466 + }, + { + "epoch": 1.2431699913269731, + "grad_norm": 3.3966782093048096, + "learning_rate": 1.1754678659509649e-05, + "loss": 0.0187, + "step": 11467 + }, + { + "epoch": 1.243278404163053, + "grad_norm": 0.25210481882095337, + "learning_rate": 1.1753953285942261e-05, + "loss": 0.0022, + "step": 11468 + }, + { + "epoch": 1.2433868169991327, + "grad_norm": 1.0500717163085938, + "learning_rate": 1.1753227912374873e-05, + "loss": 0.0062, + "step": 11469 + }, + { + "epoch": 1.2434952298352124, + "grad_norm": 2.5202722549438477, + "learning_rate": 1.1752502538807487e-05, + "loss": 0.0207, + "step": 11470 + }, + { + "epoch": 1.2436036426712922, + "grad_norm": 4.5414347648620605, + "learning_rate": 1.1751777165240099e-05, + "loss": 0.0219, + "step": 11471 + }, + { + "epoch": 1.243712055507372, + "grad_norm": 0.2680347263813019, + "learning_rate": 1.1751051791672711e-05, + "loss": 0.002, + "step": 11472 + }, + { + "epoch": 1.243820468343452, + "grad_norm": 2.440150499343872, + "learning_rate": 1.1750326418105325e-05, + "loss": 0.0305, + "step": 11473 + }, + { + "epoch": 1.2439288811795317, + "grad_norm": 2.465653896331787, + "learning_rate": 1.1749601044537937e-05, + "loss": 0.0526, + "step": 11474 + }, + { + "epoch": 1.2440372940156115, + "grad_norm": 0.9856359958648682, + "learning_rate": 1.174887567097055e-05, + "loss": 0.0024, + "step": 11475 + }, + { + "epoch": 1.2441457068516912, + "grad_norm": 1.5722821950912476, + "learning_rate": 1.1748150297403163e-05, + "loss": 0.01, + "step": 11476 + }, + { + "epoch": 1.244254119687771, + "grad_norm": 3.693209409713745, + "learning_rate": 1.1747424923835777e-05, + "loss": 0.0409, + "step": 11477 + }, + { + "epoch": 1.2443625325238508, + "grad_norm": 0.34112316370010376, + "learning_rate": 1.174669955026839e-05, + "loss": 0.0019, + "step": 11478 + }, + { + "epoch": 1.2444709453599305, + "grad_norm": 2.3190770149230957, + "learning_rate": 1.1745974176701003e-05, + "loss": 0.0111, + "step": 11479 + }, + { + "epoch": 1.2445793581960105, + "grad_norm": 0.3905872106552124, + "learning_rate": 1.1745248803133615e-05, + "loss": 0.0023, + "step": 11480 + }, + { + "epoch": 1.2446877710320903, + "grad_norm": 0.0971546620130539, + "learning_rate": 1.1744523429566227e-05, + "loss": 0.0007, + "step": 11481 + }, + { + "epoch": 1.24479618386817, + "grad_norm": 1.4099701642990112, + "learning_rate": 1.1743798055998841e-05, + "loss": 0.0267, + "step": 11482 + }, + { + "epoch": 1.2449045967042498, + "grad_norm": 1.5968067646026611, + "learning_rate": 1.1743072682431453e-05, + "loss": 0.0086, + "step": 11483 + }, + { + "epoch": 1.2450130095403296, + "grad_norm": 4.369510650634766, + "learning_rate": 1.1742347308864065e-05, + "loss": 0.0367, + "step": 11484 + }, + { + "epoch": 1.2451214223764093, + "grad_norm": 1.870280385017395, + "learning_rate": 1.174162193529668e-05, + "loss": 0.0084, + "step": 11485 + }, + { + "epoch": 1.245229835212489, + "grad_norm": 4.7417192459106445, + "learning_rate": 1.1740896561729291e-05, + "loss": 0.0658, + "step": 11486 + }, + { + "epoch": 1.245338248048569, + "grad_norm": 0.3867582380771637, + "learning_rate": 1.1740171188161904e-05, + "loss": 0.0029, + "step": 11487 + }, + { + "epoch": 1.2454466608846488, + "grad_norm": 4.292038917541504, + "learning_rate": 1.1739445814594517e-05, + "loss": 0.0542, + "step": 11488 + }, + { + "epoch": 1.2455550737207286, + "grad_norm": 1.057504415512085, + "learning_rate": 1.173872044102713e-05, + "loss": 0.0112, + "step": 11489 + }, + { + "epoch": 1.2456634865568084, + "grad_norm": 1.9589917659759521, + "learning_rate": 1.1737995067459742e-05, + "loss": 0.0106, + "step": 11490 + }, + { + "epoch": 1.2457718993928881, + "grad_norm": 2.7629177570343018, + "learning_rate": 1.1737269693892355e-05, + "loss": 0.0156, + "step": 11491 + }, + { + "epoch": 1.245880312228968, + "grad_norm": 2.602487325668335, + "learning_rate": 1.1736544320324968e-05, + "loss": 0.0285, + "step": 11492 + }, + { + "epoch": 1.2459887250650477, + "grad_norm": 0.2141207456588745, + "learning_rate": 1.173581894675758e-05, + "loss": 0.0014, + "step": 11493 + }, + { + "epoch": 1.2460971379011274, + "grad_norm": 6.124194622039795, + "learning_rate": 1.1735093573190195e-05, + "loss": 0.0645, + "step": 11494 + }, + { + "epoch": 1.2462055507372072, + "grad_norm": 0.2865127623081207, + "learning_rate": 1.1734368199622807e-05, + "loss": 0.0026, + "step": 11495 + }, + { + "epoch": 1.2463139635732872, + "grad_norm": 6.634365558624268, + "learning_rate": 1.173364282605542e-05, + "loss": 0.0189, + "step": 11496 + }, + { + "epoch": 1.246422376409367, + "grad_norm": 0.07518467307090759, + "learning_rate": 1.1732917452488033e-05, + "loss": 0.0007, + "step": 11497 + }, + { + "epoch": 1.2465307892454467, + "grad_norm": 5.235434055328369, + "learning_rate": 1.1732192078920646e-05, + "loss": 0.1189, + "step": 11498 + }, + { + "epoch": 1.2466392020815265, + "grad_norm": 0.5821162462234497, + "learning_rate": 1.1731466705353258e-05, + "loss": 0.0051, + "step": 11499 + }, + { + "epoch": 1.2467476149176062, + "grad_norm": 0.9143983721733093, + "learning_rate": 1.1730741331785872e-05, + "loss": 0.007, + "step": 11500 + }, + { + "epoch": 1.246856027753686, + "grad_norm": 7.737701892852783, + "learning_rate": 1.1730015958218484e-05, + "loss": 0.0106, + "step": 11501 + }, + { + "epoch": 1.2469644405897657, + "grad_norm": 0.09390152990818024, + "learning_rate": 1.1729290584651096e-05, + "loss": 0.001, + "step": 11502 + }, + { + "epoch": 1.2470728534258457, + "grad_norm": 0.39294537901878357, + "learning_rate": 1.172856521108371e-05, + "loss": 0.002, + "step": 11503 + }, + { + "epoch": 1.2471812662619255, + "grad_norm": 3.7231462001800537, + "learning_rate": 1.1727839837516322e-05, + "loss": 0.0553, + "step": 11504 + }, + { + "epoch": 1.2472896790980053, + "grad_norm": 2.0150516033172607, + "learning_rate": 1.1727114463948934e-05, + "loss": 0.0333, + "step": 11505 + }, + { + "epoch": 1.247398091934085, + "grad_norm": 0.7825204730033875, + "learning_rate": 1.1726389090381546e-05, + "loss": 0.008, + "step": 11506 + }, + { + "epoch": 1.2475065047701648, + "grad_norm": 2.1091818809509277, + "learning_rate": 1.172566371681416e-05, + "loss": 0.0147, + "step": 11507 + }, + { + "epoch": 1.2476149176062445, + "grad_norm": 0.10326322913169861, + "learning_rate": 1.1724938343246772e-05, + "loss": 0.0005, + "step": 11508 + }, + { + "epoch": 1.2477233304423243, + "grad_norm": 0.15533988177776337, + "learning_rate": 1.1724212969679384e-05, + "loss": 0.0018, + "step": 11509 + }, + { + "epoch": 1.247831743278404, + "grad_norm": 0.15782420337200165, + "learning_rate": 1.1723487596111998e-05, + "loss": 0.0012, + "step": 11510 + }, + { + "epoch": 1.2479401561144838, + "grad_norm": 0.8658953309059143, + "learning_rate": 1.1722762222544612e-05, + "loss": 0.0178, + "step": 11511 + }, + { + "epoch": 1.2480485689505638, + "grad_norm": 1.8079967498779297, + "learning_rate": 1.1722036848977226e-05, + "loss": 0.0031, + "step": 11512 + }, + { + "epoch": 1.2481569817866436, + "grad_norm": 1.5312422513961792, + "learning_rate": 1.1721311475409838e-05, + "loss": 0.0439, + "step": 11513 + }, + { + "epoch": 1.2482653946227233, + "grad_norm": 0.5243824124336243, + "learning_rate": 1.172058610184245e-05, + "loss": 0.006, + "step": 11514 + }, + { + "epoch": 1.248373807458803, + "grad_norm": 4.450527191162109, + "learning_rate": 1.1719860728275064e-05, + "loss": 0.0137, + "step": 11515 + }, + { + "epoch": 1.2484822202948829, + "grad_norm": 11.948699951171875, + "learning_rate": 1.1719135354707676e-05, + "loss": 0.0454, + "step": 11516 + }, + { + "epoch": 1.2485906331309626, + "grad_norm": 3.994791269302368, + "learning_rate": 1.1718409981140288e-05, + "loss": 0.0835, + "step": 11517 + }, + { + "epoch": 1.2486990459670424, + "grad_norm": 1.9248647689819336, + "learning_rate": 1.1717684607572902e-05, + "loss": 0.0042, + "step": 11518 + }, + { + "epoch": 1.2488074588031224, + "grad_norm": 1.8929553031921387, + "learning_rate": 1.1716959234005514e-05, + "loss": 0.0092, + "step": 11519 + }, + { + "epoch": 1.2489158716392021, + "grad_norm": 1.902626872062683, + "learning_rate": 1.1716233860438126e-05, + "loss": 0.0185, + "step": 11520 + }, + { + "epoch": 1.249024284475282, + "grad_norm": 0.2700817584991455, + "learning_rate": 1.1715508486870738e-05, + "loss": 0.0019, + "step": 11521 + }, + { + "epoch": 1.2491326973113617, + "grad_norm": 2.746325969696045, + "learning_rate": 1.1714783113303352e-05, + "loss": 0.0529, + "step": 11522 + }, + { + "epoch": 1.2492411101474414, + "grad_norm": 0.2162579447031021, + "learning_rate": 1.1714057739735964e-05, + "loss": 0.0016, + "step": 11523 + }, + { + "epoch": 1.2493495229835212, + "grad_norm": 7.17433500289917, + "learning_rate": 1.1713332366168577e-05, + "loss": 0.0247, + "step": 11524 + }, + { + "epoch": 1.249457935819601, + "grad_norm": 3.4242095947265625, + "learning_rate": 1.171260699260119e-05, + "loss": 0.0337, + "step": 11525 + }, + { + "epoch": 1.249566348655681, + "grad_norm": 6.685492515563965, + "learning_rate": 1.1711881619033803e-05, + "loss": 0.0249, + "step": 11526 + }, + { + "epoch": 1.2496747614917607, + "grad_norm": 1.820225715637207, + "learning_rate": 1.1711156245466415e-05, + "loss": 0.0213, + "step": 11527 + }, + { + "epoch": 1.2497831743278405, + "grad_norm": 0.5636224746704102, + "learning_rate": 1.171043087189903e-05, + "loss": 0.0045, + "step": 11528 + }, + { + "epoch": 1.2498915871639202, + "grad_norm": 2.7379298210144043, + "learning_rate": 1.1709705498331642e-05, + "loss": 0.0333, + "step": 11529 + }, + { + "epoch": 1.25, + "grad_norm": 1.6509596109390259, + "learning_rate": 1.1708980124764256e-05, + "loss": 0.0149, + "step": 11530 + }, + { + "epoch": 1.2501084128360798, + "grad_norm": 0.18939277529716492, + "learning_rate": 1.1708254751196868e-05, + "loss": 0.0013, + "step": 11531 + }, + { + "epoch": 1.2502168256721595, + "grad_norm": 0.32981497049331665, + "learning_rate": 1.170752937762948e-05, + "loss": 0.0017, + "step": 11532 + }, + { + "epoch": 1.2503252385082395, + "grad_norm": 2.965470552444458, + "learning_rate": 1.1706804004062093e-05, + "loss": 0.0594, + "step": 11533 + }, + { + "epoch": 1.250433651344319, + "grad_norm": 0.19015352427959442, + "learning_rate": 1.1706078630494706e-05, + "loss": 0.0014, + "step": 11534 + }, + { + "epoch": 1.250542064180399, + "grad_norm": 0.6116951107978821, + "learning_rate": 1.1705353256927319e-05, + "loss": 0.0094, + "step": 11535 + }, + { + "epoch": 1.2506504770164788, + "grad_norm": 5.783957481384277, + "learning_rate": 1.170462788335993e-05, + "loss": 0.033, + "step": 11536 + }, + { + "epoch": 1.2507588898525586, + "grad_norm": 4.225584983825684, + "learning_rate": 1.1703902509792545e-05, + "loss": 0.0457, + "step": 11537 + }, + { + "epoch": 1.2508673026886383, + "grad_norm": 0.8848621845245361, + "learning_rate": 1.1703177136225157e-05, + "loss": 0.0053, + "step": 11538 + }, + { + "epoch": 1.250975715524718, + "grad_norm": 1.2212563753128052, + "learning_rate": 1.1702451762657769e-05, + "loss": 0.0095, + "step": 11539 + }, + { + "epoch": 1.2510841283607979, + "grad_norm": 1.0062496662139893, + "learning_rate": 1.1701726389090383e-05, + "loss": 0.0106, + "step": 11540 + }, + { + "epoch": 1.2511925411968776, + "grad_norm": 0.8965194821357727, + "learning_rate": 1.1701001015522995e-05, + "loss": 0.0098, + "step": 11541 + }, + { + "epoch": 1.2513009540329576, + "grad_norm": 1.8203169107437134, + "learning_rate": 1.1700275641955607e-05, + "loss": 0.0232, + "step": 11542 + }, + { + "epoch": 1.2514093668690374, + "grad_norm": 2.0375330448150635, + "learning_rate": 1.169955026838822e-05, + "loss": 0.0238, + "step": 11543 + }, + { + "epoch": 1.2515177797051171, + "grad_norm": 4.122982025146484, + "learning_rate": 1.1698824894820833e-05, + "loss": 0.0245, + "step": 11544 + }, + { + "epoch": 1.251626192541197, + "grad_norm": 0.42924439907073975, + "learning_rate": 1.1698099521253445e-05, + "loss": 0.0035, + "step": 11545 + }, + { + "epoch": 1.2517346053772767, + "grad_norm": 5.311482906341553, + "learning_rate": 1.169737414768606e-05, + "loss": 0.0085, + "step": 11546 + }, + { + "epoch": 1.2518430182133564, + "grad_norm": 2.9532787799835205, + "learning_rate": 1.1696648774118673e-05, + "loss": 0.0328, + "step": 11547 + }, + { + "epoch": 1.2519514310494362, + "grad_norm": 0.11516541242599487, + "learning_rate": 1.1695923400551285e-05, + "loss": 0.0013, + "step": 11548 + }, + { + "epoch": 1.2520598438855162, + "grad_norm": 6.37075138092041, + "learning_rate": 1.1695198026983899e-05, + "loss": 0.0529, + "step": 11549 + }, + { + "epoch": 1.2521682567215957, + "grad_norm": 1.6374682188034058, + "learning_rate": 1.1694472653416511e-05, + "loss": 0.0481, + "step": 11550 + }, + { + "epoch": 1.2522766695576757, + "grad_norm": 8.66281795501709, + "learning_rate": 1.1693747279849123e-05, + "loss": 0.0144, + "step": 11551 + }, + { + "epoch": 1.2523850823937555, + "grad_norm": 2.021393060684204, + "learning_rate": 1.1693021906281737e-05, + "loss": 0.009, + "step": 11552 + }, + { + "epoch": 1.2524934952298352, + "grad_norm": 2.5200867652893066, + "learning_rate": 1.1692296532714349e-05, + "loss": 0.0415, + "step": 11553 + }, + { + "epoch": 1.252601908065915, + "grad_norm": 4.8826069831848145, + "learning_rate": 1.1691571159146961e-05, + "loss": 0.0222, + "step": 11554 + }, + { + "epoch": 1.2527103209019947, + "grad_norm": 1.3399536609649658, + "learning_rate": 1.1690845785579575e-05, + "loss": 0.0074, + "step": 11555 + }, + { + "epoch": 1.2528187337380745, + "grad_norm": 0.9532161951065063, + "learning_rate": 1.1690120412012187e-05, + "loss": 0.0113, + "step": 11556 + }, + { + "epoch": 1.2529271465741543, + "grad_norm": 0.10485632717609406, + "learning_rate": 1.16893950384448e-05, + "loss": 0.0009, + "step": 11557 + }, + { + "epoch": 1.2530355594102343, + "grad_norm": 3.5388150215148926, + "learning_rate": 1.1688669664877413e-05, + "loss": 0.0107, + "step": 11558 + }, + { + "epoch": 1.253143972246314, + "grad_norm": 2.080005645751953, + "learning_rate": 1.1687944291310025e-05, + "loss": 0.0046, + "step": 11559 + }, + { + "epoch": 1.2532523850823938, + "grad_norm": 0.8385708928108215, + "learning_rate": 1.1687218917742637e-05, + "loss": 0.0104, + "step": 11560 + }, + { + "epoch": 1.2533607979184735, + "grad_norm": 0.4794655740261078, + "learning_rate": 1.168649354417525e-05, + "loss": 0.0051, + "step": 11561 + }, + { + "epoch": 1.2534692107545533, + "grad_norm": 1.0247089862823486, + "learning_rate": 1.1685768170607863e-05, + "loss": 0.0223, + "step": 11562 + }, + { + "epoch": 1.253577623590633, + "grad_norm": 1.5572055578231812, + "learning_rate": 1.1685042797040477e-05, + "loss": 0.0155, + "step": 11563 + }, + { + "epoch": 1.2536860364267128, + "grad_norm": 4.302605628967285, + "learning_rate": 1.1684317423473091e-05, + "loss": 0.0365, + "step": 11564 + }, + { + "epoch": 1.2537944492627928, + "grad_norm": 0.9041655659675598, + "learning_rate": 1.1683592049905703e-05, + "loss": 0.0115, + "step": 11565 + }, + { + "epoch": 1.2539028620988726, + "grad_norm": 5.833982944488525, + "learning_rate": 1.1682866676338315e-05, + "loss": 0.0198, + "step": 11566 + }, + { + "epoch": 1.2540112749349523, + "grad_norm": 1.4616844654083252, + "learning_rate": 1.168214130277093e-05, + "loss": 0.0221, + "step": 11567 + }, + { + "epoch": 1.254119687771032, + "grad_norm": 0.7555902600288391, + "learning_rate": 1.1681415929203541e-05, + "loss": 0.0065, + "step": 11568 + }, + { + "epoch": 1.2542281006071119, + "grad_norm": 1.2411900758743286, + "learning_rate": 1.1680690555636153e-05, + "loss": 0.0157, + "step": 11569 + }, + { + "epoch": 1.2543365134431916, + "grad_norm": 2.3088865280151367, + "learning_rate": 1.1679965182068767e-05, + "loss": 0.0244, + "step": 11570 + }, + { + "epoch": 1.2544449262792714, + "grad_norm": 0.7963682413101196, + "learning_rate": 1.167923980850138e-05, + "loss": 0.0541, + "step": 11571 + }, + { + "epoch": 1.2545533391153514, + "grad_norm": 4.821978569030762, + "learning_rate": 1.1678514434933992e-05, + "loss": 0.014, + "step": 11572 + }, + { + "epoch": 1.254661751951431, + "grad_norm": 0.8020120859146118, + "learning_rate": 1.1677789061366604e-05, + "loss": 0.0086, + "step": 11573 + }, + { + "epoch": 1.254770164787511, + "grad_norm": 2.156904697418213, + "learning_rate": 1.1677063687799218e-05, + "loss": 0.0095, + "step": 11574 + }, + { + "epoch": 1.2548785776235907, + "grad_norm": 4.712786674499512, + "learning_rate": 1.167633831423183e-05, + "loss": 0.0228, + "step": 11575 + }, + { + "epoch": 1.2549869904596704, + "grad_norm": 2.435558795928955, + "learning_rate": 1.1675612940664442e-05, + "loss": 0.0092, + "step": 11576 + }, + { + "epoch": 1.2550954032957502, + "grad_norm": 2.0452144145965576, + "learning_rate": 1.1674887567097056e-05, + "loss": 0.0257, + "step": 11577 + }, + { + "epoch": 1.25520381613183, + "grad_norm": 0.1445963978767395, + "learning_rate": 1.1674162193529668e-05, + "loss": 0.0011, + "step": 11578 + }, + { + "epoch": 1.2553122289679097, + "grad_norm": 2.4845798015594482, + "learning_rate": 1.167343681996228e-05, + "loss": 0.0596, + "step": 11579 + }, + { + "epoch": 1.2554206418039895, + "grad_norm": 3.3812015056610107, + "learning_rate": 1.1672711446394896e-05, + "loss": 0.0244, + "step": 11580 + }, + { + "epoch": 1.2555290546400695, + "grad_norm": 2.6056361198425293, + "learning_rate": 1.1671986072827508e-05, + "loss": 0.0181, + "step": 11581 + }, + { + "epoch": 1.2556374674761492, + "grad_norm": 1.0399479866027832, + "learning_rate": 1.1671260699260122e-05, + "loss": 0.0138, + "step": 11582 + }, + { + "epoch": 1.255745880312229, + "grad_norm": 2.7096502780914307, + "learning_rate": 1.1670535325692734e-05, + "loss": 0.0172, + "step": 11583 + }, + { + "epoch": 1.2558542931483088, + "grad_norm": 1.4163848161697388, + "learning_rate": 1.1669809952125346e-05, + "loss": 0.0339, + "step": 11584 + }, + { + "epoch": 1.2559627059843885, + "grad_norm": 1.5552977323532104, + "learning_rate": 1.1669084578557958e-05, + "loss": 0.0381, + "step": 11585 + }, + { + "epoch": 1.2560711188204683, + "grad_norm": 4.0519609451293945, + "learning_rate": 1.1668359204990572e-05, + "loss": 0.0688, + "step": 11586 + }, + { + "epoch": 1.256179531656548, + "grad_norm": 3.077451705932617, + "learning_rate": 1.1667633831423184e-05, + "loss": 0.0301, + "step": 11587 + }, + { + "epoch": 1.256287944492628, + "grad_norm": 1.9906002283096313, + "learning_rate": 1.1666908457855796e-05, + "loss": 0.0087, + "step": 11588 + }, + { + "epoch": 1.2563963573287076, + "grad_norm": 1.7867738008499146, + "learning_rate": 1.166618308428841e-05, + "loss": 0.0176, + "step": 11589 + }, + { + "epoch": 1.2565047701647876, + "grad_norm": 3.757171154022217, + "learning_rate": 1.1665457710721022e-05, + "loss": 0.029, + "step": 11590 + }, + { + "epoch": 1.2566131830008673, + "grad_norm": 5.922281265258789, + "learning_rate": 1.1664732337153634e-05, + "loss": 0.0561, + "step": 11591 + }, + { + "epoch": 1.256721595836947, + "grad_norm": 0.27121394872665405, + "learning_rate": 1.1664006963586248e-05, + "loss": 0.004, + "step": 11592 + }, + { + "epoch": 1.2568300086730269, + "grad_norm": 1.980400562286377, + "learning_rate": 1.166328159001886e-05, + "loss": 0.0134, + "step": 11593 + }, + { + "epoch": 1.2569384215091066, + "grad_norm": 6.0631632804870605, + "learning_rate": 1.1662556216451472e-05, + "loss": 0.0237, + "step": 11594 + }, + { + "epoch": 1.2570468343451866, + "grad_norm": 4.45794677734375, + "learning_rate": 1.1661830842884086e-05, + "loss": 0.0358, + "step": 11595 + }, + { + "epoch": 1.2571552471812661, + "grad_norm": 5.42434024810791, + "learning_rate": 1.1661105469316698e-05, + "loss": 0.0483, + "step": 11596 + }, + { + "epoch": 1.2572636600173461, + "grad_norm": 0.7236917614936829, + "learning_rate": 1.1660380095749312e-05, + "loss": 0.0077, + "step": 11597 + }, + { + "epoch": 1.257372072853426, + "grad_norm": 1.5041953325271606, + "learning_rate": 1.1659654722181926e-05, + "loss": 0.0227, + "step": 11598 + }, + { + "epoch": 1.2574804856895057, + "grad_norm": 1.6879827976226807, + "learning_rate": 1.1658929348614538e-05, + "loss": 0.0106, + "step": 11599 + }, + { + "epoch": 1.2575888985255854, + "grad_norm": 1.9407639503479004, + "learning_rate": 1.165820397504715e-05, + "loss": 0.0273, + "step": 11600 + }, + { + "epoch": 1.2576973113616652, + "grad_norm": 0.4885503947734833, + "learning_rate": 1.1657478601479764e-05, + "loss": 0.0028, + "step": 11601 + }, + { + "epoch": 1.257805724197745, + "grad_norm": 0.43635207414627075, + "learning_rate": 1.1656753227912376e-05, + "loss": 0.0048, + "step": 11602 + }, + { + "epoch": 1.2579141370338247, + "grad_norm": 0.9897868037223816, + "learning_rate": 1.1656027854344988e-05, + "loss": 0.0229, + "step": 11603 + }, + { + "epoch": 1.2580225498699047, + "grad_norm": 0.42621996998786926, + "learning_rate": 1.1655302480777602e-05, + "loss": 0.0039, + "step": 11604 + }, + { + "epoch": 1.2581309627059845, + "grad_norm": 7.32119083404541, + "learning_rate": 1.1654577107210214e-05, + "loss": 0.0525, + "step": 11605 + }, + { + "epoch": 1.2582393755420642, + "grad_norm": 1.678557276725769, + "learning_rate": 1.1653851733642826e-05, + "loss": 0.0117, + "step": 11606 + }, + { + "epoch": 1.258347788378144, + "grad_norm": 1.0528289079666138, + "learning_rate": 1.165312636007544e-05, + "loss": 0.0176, + "step": 11607 + }, + { + "epoch": 1.2584562012142237, + "grad_norm": 2.0950489044189453, + "learning_rate": 1.1652400986508052e-05, + "loss": 0.0303, + "step": 11608 + }, + { + "epoch": 1.2585646140503035, + "grad_norm": 1.1500976085662842, + "learning_rate": 1.1651675612940665e-05, + "loss": 0.0181, + "step": 11609 + }, + { + "epoch": 1.2586730268863833, + "grad_norm": 2.421144723892212, + "learning_rate": 1.1650950239373278e-05, + "loss": 0.0557, + "step": 11610 + }, + { + "epoch": 1.2587814397224633, + "grad_norm": 7.132301330566406, + "learning_rate": 1.165022486580589e-05, + "loss": 0.1235, + "step": 11611 + }, + { + "epoch": 1.2588898525585428, + "grad_norm": 4.058872699737549, + "learning_rate": 1.1649499492238503e-05, + "loss": 0.0286, + "step": 11612 + }, + { + "epoch": 1.2589982653946228, + "grad_norm": 0.12242577224969864, + "learning_rate": 1.1648774118671115e-05, + "loss": 0.0006, + "step": 11613 + }, + { + "epoch": 1.2591066782307025, + "grad_norm": 1.7997901439666748, + "learning_rate": 1.1648048745103729e-05, + "loss": 0.0164, + "step": 11614 + }, + { + "epoch": 1.2592150910667823, + "grad_norm": 2.6998398303985596, + "learning_rate": 1.1647323371536343e-05, + "loss": 0.0223, + "step": 11615 + }, + { + "epoch": 1.259323503902862, + "grad_norm": 4.2627081871032715, + "learning_rate": 1.1646597997968956e-05, + "loss": 0.0557, + "step": 11616 + }, + { + "epoch": 1.2594319167389418, + "grad_norm": 0.5051974058151245, + "learning_rate": 1.1645872624401569e-05, + "loss": 0.004, + "step": 11617 + }, + { + "epoch": 1.2595403295750216, + "grad_norm": 2.1279683113098145, + "learning_rate": 1.164514725083418e-05, + "loss": 0.0354, + "step": 11618 + }, + { + "epoch": 1.2596487424111014, + "grad_norm": 3.8999440670013428, + "learning_rate": 1.1644421877266795e-05, + "loss": 0.0174, + "step": 11619 + }, + { + "epoch": 1.2597571552471813, + "grad_norm": 0.20493941009044647, + "learning_rate": 1.1643696503699407e-05, + "loss": 0.0019, + "step": 11620 + }, + { + "epoch": 1.2598655680832611, + "grad_norm": 2.050981283187866, + "learning_rate": 1.1642971130132019e-05, + "loss": 0.0079, + "step": 11621 + }, + { + "epoch": 1.2599739809193409, + "grad_norm": 1.5235964059829712, + "learning_rate": 1.1642245756564633e-05, + "loss": 0.0369, + "step": 11622 + }, + { + "epoch": 1.2600823937554206, + "grad_norm": 3.0109760761260986, + "learning_rate": 1.1641520382997245e-05, + "loss": 0.0287, + "step": 11623 + }, + { + "epoch": 1.2601908065915004, + "grad_norm": 3.510751962661743, + "learning_rate": 1.1640795009429857e-05, + "loss": 0.0269, + "step": 11624 + }, + { + "epoch": 1.2602992194275802, + "grad_norm": 0.3035518527030945, + "learning_rate": 1.1640069635862469e-05, + "loss": 0.005, + "step": 11625 + }, + { + "epoch": 1.26040763226366, + "grad_norm": 5.700677394866943, + "learning_rate": 1.1639344262295083e-05, + "loss": 0.0246, + "step": 11626 + }, + { + "epoch": 1.26051604509974, + "grad_norm": 2.2869746685028076, + "learning_rate": 1.1638618888727695e-05, + "loss": 0.0185, + "step": 11627 + }, + { + "epoch": 1.2606244579358197, + "grad_norm": 8.753729820251465, + "learning_rate": 1.1637893515160307e-05, + "loss": 0.0682, + "step": 11628 + }, + { + "epoch": 1.2607328707718994, + "grad_norm": 2.225914716720581, + "learning_rate": 1.1637168141592921e-05, + "loss": 0.0585, + "step": 11629 + }, + { + "epoch": 1.2608412836079792, + "grad_norm": 0.5041934847831726, + "learning_rate": 1.1636442768025533e-05, + "loss": 0.0021, + "step": 11630 + }, + { + "epoch": 1.260949696444059, + "grad_norm": 0.3280925452709198, + "learning_rate": 1.1635717394458145e-05, + "loss": 0.0028, + "step": 11631 + }, + { + "epoch": 1.2610581092801387, + "grad_norm": 3.200312376022339, + "learning_rate": 1.1634992020890761e-05, + "loss": 0.0088, + "step": 11632 + }, + { + "epoch": 1.2611665221162185, + "grad_norm": 0.9470556974411011, + "learning_rate": 1.1634266647323373e-05, + "loss": 0.0068, + "step": 11633 + }, + { + "epoch": 1.2612749349522985, + "grad_norm": 1.5451313257217407, + "learning_rate": 1.1633541273755987e-05, + "loss": 0.0107, + "step": 11634 + }, + { + "epoch": 1.261383347788378, + "grad_norm": 0.8289352059364319, + "learning_rate": 1.1632815900188599e-05, + "loss": 0.0072, + "step": 11635 + }, + { + "epoch": 1.261491760624458, + "grad_norm": 0.3030121922492981, + "learning_rate": 1.1632090526621211e-05, + "loss": 0.0036, + "step": 11636 + }, + { + "epoch": 1.2616001734605378, + "grad_norm": 2.5506954193115234, + "learning_rate": 1.1631365153053823e-05, + "loss": 0.0252, + "step": 11637 + }, + { + "epoch": 1.2617085862966175, + "grad_norm": 0.41480618715286255, + "learning_rate": 1.1630639779486437e-05, + "loss": 0.0056, + "step": 11638 + }, + { + "epoch": 1.2618169991326973, + "grad_norm": 2.4244959354400635, + "learning_rate": 1.162991440591905e-05, + "loss": 0.0324, + "step": 11639 + }, + { + "epoch": 1.261925411968777, + "grad_norm": 2.2215986251831055, + "learning_rate": 1.1629189032351661e-05, + "loss": 0.0078, + "step": 11640 + }, + { + "epoch": 1.2620338248048568, + "grad_norm": 0.6420688629150391, + "learning_rate": 1.1628463658784275e-05, + "loss": 0.008, + "step": 11641 + }, + { + "epoch": 1.2621422376409366, + "grad_norm": 1.7221819162368774, + "learning_rate": 1.1627738285216887e-05, + "loss": 0.0068, + "step": 11642 + }, + { + "epoch": 1.2622506504770166, + "grad_norm": 0.4470738470554352, + "learning_rate": 1.16270129116495e-05, + "loss": 0.0043, + "step": 11643 + }, + { + "epoch": 1.2623590633130963, + "grad_norm": 3.0459461212158203, + "learning_rate": 1.1626287538082113e-05, + "loss": 0.0066, + "step": 11644 + }, + { + "epoch": 1.262467476149176, + "grad_norm": 0.4625111222267151, + "learning_rate": 1.1625562164514726e-05, + "loss": 0.0035, + "step": 11645 + }, + { + "epoch": 1.2625758889852559, + "grad_norm": 2.712825298309326, + "learning_rate": 1.1624836790947338e-05, + "loss": 0.0343, + "step": 11646 + }, + { + "epoch": 1.2626843018213356, + "grad_norm": 0.3836487829685211, + "learning_rate": 1.1624111417379951e-05, + "loss": 0.0045, + "step": 11647 + }, + { + "epoch": 1.2627927146574154, + "grad_norm": 6.289482116699219, + "learning_rate": 1.1623386043812564e-05, + "loss": 0.0265, + "step": 11648 + }, + { + "epoch": 1.2629011274934951, + "grad_norm": 0.262167751789093, + "learning_rate": 1.1622660670245177e-05, + "loss": 0.002, + "step": 11649 + }, + { + "epoch": 1.2630095403295751, + "grad_norm": 1.8376318216323853, + "learning_rate": 1.1621935296677791e-05, + "loss": 0.0248, + "step": 11650 + }, + { + "epoch": 1.263117953165655, + "grad_norm": 0.8017836809158325, + "learning_rate": 1.1621209923110403e-05, + "loss": 0.0088, + "step": 11651 + }, + { + "epoch": 1.2632263660017347, + "grad_norm": 1.457308053970337, + "learning_rate": 1.1620484549543016e-05, + "loss": 0.0081, + "step": 11652 + }, + { + "epoch": 1.2633347788378144, + "grad_norm": 1.6124157905578613, + "learning_rate": 1.161975917597563e-05, + "loss": 0.0069, + "step": 11653 + }, + { + "epoch": 1.2634431916738942, + "grad_norm": 0.46229803562164307, + "learning_rate": 1.1619033802408242e-05, + "loss": 0.0053, + "step": 11654 + }, + { + "epoch": 1.263551604509974, + "grad_norm": 2.557399272918701, + "learning_rate": 1.1618308428840854e-05, + "loss": 0.0107, + "step": 11655 + }, + { + "epoch": 1.2636600173460537, + "grad_norm": 1.7603071928024292, + "learning_rate": 1.1617583055273468e-05, + "loss": 0.0025, + "step": 11656 + }, + { + "epoch": 1.2637684301821337, + "grad_norm": 4.16232967376709, + "learning_rate": 1.161685768170608e-05, + "loss": 0.0776, + "step": 11657 + }, + { + "epoch": 1.2638768430182132, + "grad_norm": 1.8286347389221191, + "learning_rate": 1.1616132308138692e-05, + "loss": 0.0178, + "step": 11658 + }, + { + "epoch": 1.2639852558542932, + "grad_norm": 0.7113137245178223, + "learning_rate": 1.1615406934571306e-05, + "loss": 0.005, + "step": 11659 + }, + { + "epoch": 1.264093668690373, + "grad_norm": 2.403437614440918, + "learning_rate": 1.1614681561003918e-05, + "loss": 0.0268, + "step": 11660 + }, + { + "epoch": 1.2642020815264527, + "grad_norm": 4.8254075050354, + "learning_rate": 1.161395618743653e-05, + "loss": 0.0265, + "step": 11661 + }, + { + "epoch": 1.2643104943625325, + "grad_norm": 3.223154067993164, + "learning_rate": 1.1613230813869144e-05, + "loss": 0.0682, + "step": 11662 + }, + { + "epoch": 1.2644189071986123, + "grad_norm": 2.2651736736297607, + "learning_rate": 1.1612505440301756e-05, + "loss": 0.0367, + "step": 11663 + }, + { + "epoch": 1.264527320034692, + "grad_norm": 3.6262664794921875, + "learning_rate": 1.1611780066734368e-05, + "loss": 0.0244, + "step": 11664 + }, + { + "epoch": 1.2646357328707718, + "grad_norm": 1.9901024103164673, + "learning_rate": 1.161105469316698e-05, + "loss": 0.0374, + "step": 11665 + }, + { + "epoch": 1.2647441457068518, + "grad_norm": 1.4044686555862427, + "learning_rate": 1.1610329319599594e-05, + "loss": 0.0111, + "step": 11666 + }, + { + "epoch": 1.2648525585429315, + "grad_norm": 4.03148889541626, + "learning_rate": 1.1609603946032208e-05, + "loss": 0.0062, + "step": 11667 + }, + { + "epoch": 1.2649609713790113, + "grad_norm": 0.6264079213142395, + "learning_rate": 1.1608878572464822e-05, + "loss": 0.0106, + "step": 11668 + }, + { + "epoch": 1.265069384215091, + "grad_norm": 0.9776129722595215, + "learning_rate": 1.1608153198897434e-05, + "loss": 0.015, + "step": 11669 + }, + { + "epoch": 1.2651777970511708, + "grad_norm": 1.703527808189392, + "learning_rate": 1.1607427825330046e-05, + "loss": 0.0204, + "step": 11670 + }, + { + "epoch": 1.2652862098872506, + "grad_norm": 2.7592391967773438, + "learning_rate": 1.160670245176266e-05, + "loss": 0.0128, + "step": 11671 + }, + { + "epoch": 1.2653946227233304, + "grad_norm": 3.4703309535980225, + "learning_rate": 1.1605977078195272e-05, + "loss": 0.0224, + "step": 11672 + }, + { + "epoch": 1.2655030355594103, + "grad_norm": 6.8111677169799805, + "learning_rate": 1.1605251704627884e-05, + "loss": 0.0602, + "step": 11673 + }, + { + "epoch": 1.26561144839549, + "grad_norm": 9.821481704711914, + "learning_rate": 1.1604526331060498e-05, + "loss": 0.0121, + "step": 11674 + }, + { + "epoch": 1.2657198612315699, + "grad_norm": 0.08060978353023529, + "learning_rate": 1.160380095749311e-05, + "loss": 0.0009, + "step": 11675 + }, + { + "epoch": 1.2658282740676496, + "grad_norm": 1.815869688987732, + "learning_rate": 1.1603075583925722e-05, + "loss": 0.0153, + "step": 11676 + }, + { + "epoch": 1.2659366869037294, + "grad_norm": 2.193861722946167, + "learning_rate": 1.1602350210358334e-05, + "loss": 0.0618, + "step": 11677 + }, + { + "epoch": 1.2660450997398092, + "grad_norm": 3.1738739013671875, + "learning_rate": 1.1601624836790948e-05, + "loss": 0.0312, + "step": 11678 + }, + { + "epoch": 1.266153512575889, + "grad_norm": 1.8410955667495728, + "learning_rate": 1.160089946322356e-05, + "loss": 0.007, + "step": 11679 + }, + { + "epoch": 1.266261925411969, + "grad_norm": 2.253429412841797, + "learning_rate": 1.1600174089656173e-05, + "loss": 0.0054, + "step": 11680 + }, + { + "epoch": 1.2663703382480485, + "grad_norm": 5.623150825500488, + "learning_rate": 1.1599448716088786e-05, + "loss": 0.0406, + "step": 11681 + }, + { + "epoch": 1.2664787510841284, + "grad_norm": 0.07678385823965073, + "learning_rate": 1.1598723342521399e-05, + "loss": 0.0006, + "step": 11682 + }, + { + "epoch": 1.2665871639202082, + "grad_norm": 1.9292323589324951, + "learning_rate": 1.159799796895401e-05, + "loss": 0.0369, + "step": 11683 + }, + { + "epoch": 1.266695576756288, + "grad_norm": 8.603964805603027, + "learning_rate": 1.1597272595386626e-05, + "loss": 0.0237, + "step": 11684 + }, + { + "epoch": 1.2668039895923677, + "grad_norm": 1.9850327968597412, + "learning_rate": 1.1596547221819238e-05, + "loss": 0.0103, + "step": 11685 + }, + { + "epoch": 1.2669124024284475, + "grad_norm": 2.511540651321411, + "learning_rate": 1.1595821848251852e-05, + "loss": 0.0032, + "step": 11686 + }, + { + "epoch": 1.2670208152645273, + "grad_norm": 7.828097343444824, + "learning_rate": 1.1595096474684464e-05, + "loss": 0.0809, + "step": 11687 + }, + { + "epoch": 1.267129228100607, + "grad_norm": 4.1477227210998535, + "learning_rate": 1.1594371101117076e-05, + "loss": 0.0202, + "step": 11688 + }, + { + "epoch": 1.267237640936687, + "grad_norm": 0.9227142333984375, + "learning_rate": 1.1593645727549689e-05, + "loss": 0.0167, + "step": 11689 + }, + { + "epoch": 1.2673460537727668, + "grad_norm": 12.3572998046875, + "learning_rate": 1.1592920353982302e-05, + "loss": 0.0412, + "step": 11690 + }, + { + "epoch": 1.2674544666088465, + "grad_norm": 2.0090088844299316, + "learning_rate": 1.1592194980414915e-05, + "loss": 0.0192, + "step": 11691 + }, + { + "epoch": 1.2675628794449263, + "grad_norm": 2.6178555488586426, + "learning_rate": 1.1591469606847527e-05, + "loss": 0.0104, + "step": 11692 + }, + { + "epoch": 1.267671292281006, + "grad_norm": 3.239149808883667, + "learning_rate": 1.159074423328014e-05, + "loss": 0.0166, + "step": 11693 + }, + { + "epoch": 1.2677797051170858, + "grad_norm": 2.665550708770752, + "learning_rate": 1.1590018859712753e-05, + "loss": 0.0681, + "step": 11694 + }, + { + "epoch": 1.2678881179531656, + "grad_norm": 0.7279052138328552, + "learning_rate": 1.1589293486145365e-05, + "loss": 0.0025, + "step": 11695 + }, + { + "epoch": 1.2679965307892456, + "grad_norm": 5.544084548950195, + "learning_rate": 1.1588568112577979e-05, + "loss": 0.0804, + "step": 11696 + }, + { + "epoch": 1.268104943625325, + "grad_norm": 4.432824611663818, + "learning_rate": 1.158784273901059e-05, + "loss": 0.0931, + "step": 11697 + }, + { + "epoch": 1.268213356461405, + "grad_norm": 3.8487863540649414, + "learning_rate": 1.1587117365443203e-05, + "loss": 0.0125, + "step": 11698 + }, + { + "epoch": 1.2683217692974849, + "grad_norm": 1.5741806030273438, + "learning_rate": 1.1586391991875817e-05, + "loss": 0.0525, + "step": 11699 + }, + { + "epoch": 1.2684301821335646, + "grad_norm": 1.4832563400268555, + "learning_rate": 1.1585666618308429e-05, + "loss": 0.0057, + "step": 11700 + }, + { + "epoch": 1.2685385949696444, + "grad_norm": 0.7836477756500244, + "learning_rate": 1.1584941244741043e-05, + "loss": 0.0072, + "step": 11701 + }, + { + "epoch": 1.2686470078057241, + "grad_norm": 2.115859270095825, + "learning_rate": 1.1584215871173657e-05, + "loss": 0.0119, + "step": 11702 + }, + { + "epoch": 1.268755420641804, + "grad_norm": 3.043753147125244, + "learning_rate": 1.1583490497606269e-05, + "loss": 0.0158, + "step": 11703 + }, + { + "epoch": 1.2688638334778837, + "grad_norm": 3.146158456802368, + "learning_rate": 1.1582765124038881e-05, + "loss": 0.036, + "step": 11704 + }, + { + "epoch": 1.2689722463139637, + "grad_norm": 3.1592941284179688, + "learning_rate": 1.1582039750471495e-05, + "loss": 0.0357, + "step": 11705 + }, + { + "epoch": 1.2690806591500434, + "grad_norm": 1.7906471490859985, + "learning_rate": 1.1581314376904107e-05, + "loss": 0.0122, + "step": 11706 + }, + { + "epoch": 1.2691890719861232, + "grad_norm": 0.8641167879104614, + "learning_rate": 1.1580589003336719e-05, + "loss": 0.0064, + "step": 11707 + }, + { + "epoch": 1.269297484822203, + "grad_norm": 1.098360300064087, + "learning_rate": 1.1579863629769333e-05, + "loss": 0.0136, + "step": 11708 + }, + { + "epoch": 1.2694058976582827, + "grad_norm": 3.7014999389648438, + "learning_rate": 1.1579138256201945e-05, + "loss": 0.0367, + "step": 11709 + }, + { + "epoch": 1.2695143104943625, + "grad_norm": 1.8034522533416748, + "learning_rate": 1.1578412882634557e-05, + "loss": 0.0216, + "step": 11710 + }, + { + "epoch": 1.2696227233304422, + "grad_norm": 3.316129684448242, + "learning_rate": 1.1577687509067171e-05, + "loss": 0.0425, + "step": 11711 + }, + { + "epoch": 1.2697311361665222, + "grad_norm": 7.926636695861816, + "learning_rate": 1.1576962135499783e-05, + "loss": 0.0887, + "step": 11712 + }, + { + "epoch": 1.269839549002602, + "grad_norm": 3.7268564701080322, + "learning_rate": 1.1576236761932395e-05, + "loss": 0.0193, + "step": 11713 + }, + { + "epoch": 1.2699479618386817, + "grad_norm": 2.87399959564209, + "learning_rate": 1.1575511388365009e-05, + "loss": 0.0099, + "step": 11714 + }, + { + "epoch": 1.2700563746747615, + "grad_norm": 4.8783979415893555, + "learning_rate": 1.1574786014797621e-05, + "loss": 0.0401, + "step": 11715 + }, + { + "epoch": 1.2701647875108413, + "grad_norm": 1.2632085084915161, + "learning_rate": 1.1574060641230233e-05, + "loss": 0.0271, + "step": 11716 + }, + { + "epoch": 1.270273200346921, + "grad_norm": 2.866255044937134, + "learning_rate": 1.1573335267662846e-05, + "loss": 0.0585, + "step": 11717 + }, + { + "epoch": 1.2703816131830008, + "grad_norm": 2.8189361095428467, + "learning_rate": 1.1572609894095461e-05, + "loss": 0.0141, + "step": 11718 + }, + { + "epoch": 1.2704900260190808, + "grad_norm": 2.573333740234375, + "learning_rate": 1.1571884520528073e-05, + "loss": 0.014, + "step": 11719 + }, + { + "epoch": 1.2705984388551603, + "grad_norm": 2.5952110290527344, + "learning_rate": 1.1571159146960687e-05, + "loss": 0.0146, + "step": 11720 + }, + { + "epoch": 1.2707068516912403, + "grad_norm": 1.9430824518203735, + "learning_rate": 1.15704337733933e-05, + "loss": 0.012, + "step": 11721 + }, + { + "epoch": 1.27081526452732, + "grad_norm": 5.671247959136963, + "learning_rate": 1.1569708399825911e-05, + "loss": 0.0361, + "step": 11722 + }, + { + "epoch": 1.2709236773633998, + "grad_norm": 1.0521832704544067, + "learning_rate": 1.1568983026258525e-05, + "loss": 0.006, + "step": 11723 + }, + { + "epoch": 1.2710320901994796, + "grad_norm": 2.3945839405059814, + "learning_rate": 1.1568257652691137e-05, + "loss": 0.0144, + "step": 11724 + }, + { + "epoch": 1.2711405030355594, + "grad_norm": 0.5622982382774353, + "learning_rate": 1.156753227912375e-05, + "loss": 0.0017, + "step": 11725 + }, + { + "epoch": 1.2712489158716391, + "grad_norm": 1.3976068496704102, + "learning_rate": 1.1566806905556363e-05, + "loss": 0.0098, + "step": 11726 + }, + { + "epoch": 1.271357328707719, + "grad_norm": 3.4666197299957275, + "learning_rate": 1.1566081531988975e-05, + "loss": 0.0272, + "step": 11727 + }, + { + "epoch": 1.2714657415437989, + "grad_norm": 1.2334108352661133, + "learning_rate": 1.1565356158421588e-05, + "loss": 0.0213, + "step": 11728 + }, + { + "epoch": 1.2715741543798786, + "grad_norm": 1.7258635759353638, + "learning_rate": 1.15646307848542e-05, + "loss": 0.0367, + "step": 11729 + }, + { + "epoch": 1.2716825672159584, + "grad_norm": 1.4160442352294922, + "learning_rate": 1.1563905411286814e-05, + "loss": 0.022, + "step": 11730 + }, + { + "epoch": 1.2717909800520382, + "grad_norm": 0.4800032377243042, + "learning_rate": 1.1563180037719426e-05, + "loss": 0.0052, + "step": 11731 + }, + { + "epoch": 1.271899392888118, + "grad_norm": 2.7456390857696533, + "learning_rate": 1.1562454664152038e-05, + "loss": 0.0158, + "step": 11732 + }, + { + "epoch": 1.2720078057241977, + "grad_norm": 0.5488936901092529, + "learning_rate": 1.1561729290584652e-05, + "loss": 0.0044, + "step": 11733 + }, + { + "epoch": 1.2721162185602775, + "grad_norm": 7.223422527313232, + "learning_rate": 1.1561003917017264e-05, + "loss": 0.0548, + "step": 11734 + }, + { + "epoch": 1.2722246313963574, + "grad_norm": 3.3052327632904053, + "learning_rate": 1.1560278543449876e-05, + "loss": 0.0079, + "step": 11735 + }, + { + "epoch": 1.272333044232437, + "grad_norm": 6.590529918670654, + "learning_rate": 1.1559553169882492e-05, + "loss": 0.069, + "step": 11736 + }, + { + "epoch": 1.272441457068517, + "grad_norm": 4.077194690704346, + "learning_rate": 1.1558827796315104e-05, + "loss": 0.029, + "step": 11737 + }, + { + "epoch": 1.2725498699045967, + "grad_norm": 1.4888213872909546, + "learning_rate": 1.1558102422747718e-05, + "loss": 0.014, + "step": 11738 + }, + { + "epoch": 1.2726582827406765, + "grad_norm": 1.0397140979766846, + "learning_rate": 1.155737704918033e-05, + "loss": 0.0089, + "step": 11739 + }, + { + "epoch": 1.2727666955767563, + "grad_norm": 4.533462047576904, + "learning_rate": 1.1556651675612942e-05, + "loss": 0.0479, + "step": 11740 + }, + { + "epoch": 1.272875108412836, + "grad_norm": 1.7095590829849243, + "learning_rate": 1.1555926302045554e-05, + "loss": 0.0174, + "step": 11741 + }, + { + "epoch": 1.272983521248916, + "grad_norm": 0.24984540045261383, + "learning_rate": 1.1555200928478168e-05, + "loss": 0.0021, + "step": 11742 + }, + { + "epoch": 1.2730919340849955, + "grad_norm": 4.4932541847229, + "learning_rate": 1.155447555491078e-05, + "loss": 0.0521, + "step": 11743 + }, + { + "epoch": 1.2732003469210755, + "grad_norm": 1.8989602327346802, + "learning_rate": 1.1553750181343392e-05, + "loss": 0.0159, + "step": 11744 + }, + { + "epoch": 1.2733087597571553, + "grad_norm": 0.4874529242515564, + "learning_rate": 1.1553024807776006e-05, + "loss": 0.0043, + "step": 11745 + }, + { + "epoch": 1.273417172593235, + "grad_norm": 3.5630996227264404, + "learning_rate": 1.1552299434208618e-05, + "loss": 0.014, + "step": 11746 + }, + { + "epoch": 1.2735255854293148, + "grad_norm": 3.4626505374908447, + "learning_rate": 1.155157406064123e-05, + "loss": 0.0277, + "step": 11747 + }, + { + "epoch": 1.2736339982653946, + "grad_norm": 0.5569185614585876, + "learning_rate": 1.1550848687073844e-05, + "loss": 0.0029, + "step": 11748 + }, + { + "epoch": 1.2737424111014743, + "grad_norm": 0.1655801236629486, + "learning_rate": 1.1550123313506456e-05, + "loss": 0.0024, + "step": 11749 + }, + { + "epoch": 1.273850823937554, + "grad_norm": 3.134687900543213, + "learning_rate": 1.1549397939939068e-05, + "loss": 0.0124, + "step": 11750 + }, + { + "epoch": 1.273959236773634, + "grad_norm": 2.1900365352630615, + "learning_rate": 1.1548672566371682e-05, + "loss": 0.063, + "step": 11751 + }, + { + "epoch": 1.2740676496097139, + "grad_norm": 2.5976762771606445, + "learning_rate": 1.1547947192804294e-05, + "loss": 0.0096, + "step": 11752 + }, + { + "epoch": 1.2741760624457936, + "grad_norm": 1.4295542240142822, + "learning_rate": 1.154722181923691e-05, + "loss": 0.0124, + "step": 11753 + }, + { + "epoch": 1.2742844752818734, + "grad_norm": 1.4430330991744995, + "learning_rate": 1.1546496445669522e-05, + "loss": 0.0138, + "step": 11754 + }, + { + "epoch": 1.2743928881179531, + "grad_norm": 6.162997245788574, + "learning_rate": 1.1545771072102134e-05, + "loss": 0.0219, + "step": 11755 + }, + { + "epoch": 1.274501300954033, + "grad_norm": 5.250458240509033, + "learning_rate": 1.1545045698534746e-05, + "loss": 0.0273, + "step": 11756 + }, + { + "epoch": 1.2746097137901127, + "grad_norm": 1.7590632438659668, + "learning_rate": 1.154432032496736e-05, + "loss": 0.0149, + "step": 11757 + }, + { + "epoch": 1.2747181266261927, + "grad_norm": 6.468204975128174, + "learning_rate": 1.1543594951399972e-05, + "loss": 0.0336, + "step": 11758 + }, + { + "epoch": 1.2748265394622722, + "grad_norm": 11.704224586486816, + "learning_rate": 1.1542869577832584e-05, + "loss": 0.0299, + "step": 11759 + }, + { + "epoch": 1.2749349522983522, + "grad_norm": 1.6990468502044678, + "learning_rate": 1.1542144204265198e-05, + "loss": 0.0083, + "step": 11760 + }, + { + "epoch": 1.275043365134432, + "grad_norm": 7.266627311706543, + "learning_rate": 1.154141883069781e-05, + "loss": 0.0288, + "step": 11761 + }, + { + "epoch": 1.2751517779705117, + "grad_norm": 1.631374716758728, + "learning_rate": 1.1540693457130423e-05, + "loss": 0.0327, + "step": 11762 + }, + { + "epoch": 1.2752601908065915, + "grad_norm": 2.981912851333618, + "learning_rate": 1.1539968083563036e-05, + "loss": 0.0339, + "step": 11763 + }, + { + "epoch": 1.2753686036426712, + "grad_norm": 0.8774703145027161, + "learning_rate": 1.1539242709995648e-05, + "loss": 0.0035, + "step": 11764 + }, + { + "epoch": 1.275477016478751, + "grad_norm": 1.8119899034500122, + "learning_rate": 1.153851733642826e-05, + "loss": 0.0078, + "step": 11765 + }, + { + "epoch": 1.2755854293148308, + "grad_norm": 2.765652656555176, + "learning_rate": 1.1537791962860874e-05, + "loss": 0.0148, + "step": 11766 + }, + { + "epoch": 1.2756938421509108, + "grad_norm": 0.3423917889595032, + "learning_rate": 1.1537066589293487e-05, + "loss": 0.0035, + "step": 11767 + }, + { + "epoch": 1.2758022549869905, + "grad_norm": 2.9870526790618896, + "learning_rate": 1.1536341215726099e-05, + "loss": 0.0254, + "step": 11768 + }, + { + "epoch": 1.2759106678230703, + "grad_norm": 1.6294255256652832, + "learning_rate": 1.1535615842158711e-05, + "loss": 0.0391, + "step": 11769 + }, + { + "epoch": 1.27601908065915, + "grad_norm": 4.5969767570495605, + "learning_rate": 1.1534890468591326e-05, + "loss": 0.0267, + "step": 11770 + }, + { + "epoch": 1.2761274934952298, + "grad_norm": 1.1914418935775757, + "learning_rate": 1.1534165095023939e-05, + "loss": 0.0213, + "step": 11771 + }, + { + "epoch": 1.2762359063313096, + "grad_norm": 0.08371999859809875, + "learning_rate": 1.1533439721456552e-05, + "loss": 0.0013, + "step": 11772 + }, + { + "epoch": 1.2763443191673893, + "grad_norm": 0.26472708582878113, + "learning_rate": 1.1532714347889165e-05, + "loss": 0.0035, + "step": 11773 + }, + { + "epoch": 1.2764527320034693, + "grad_norm": 5.10481071472168, + "learning_rate": 1.1531988974321777e-05, + "loss": 0.0329, + "step": 11774 + }, + { + "epoch": 1.276561144839549, + "grad_norm": 2.723982095718384, + "learning_rate": 1.153126360075439e-05, + "loss": 0.0228, + "step": 11775 + }, + { + "epoch": 1.2766695576756288, + "grad_norm": 5.07232666015625, + "learning_rate": 1.1530538227187003e-05, + "loss": 0.0382, + "step": 11776 + }, + { + "epoch": 1.2767779705117086, + "grad_norm": 4.714159965515137, + "learning_rate": 1.1529812853619615e-05, + "loss": 0.031, + "step": 11777 + }, + { + "epoch": 1.2768863833477884, + "grad_norm": 2.17661714553833, + "learning_rate": 1.1529087480052229e-05, + "loss": 0.0188, + "step": 11778 + }, + { + "epoch": 1.2769947961838681, + "grad_norm": 0.1396646648645401, + "learning_rate": 1.152836210648484e-05, + "loss": 0.0018, + "step": 11779 + }, + { + "epoch": 1.277103209019948, + "grad_norm": 0.30581918358802795, + "learning_rate": 1.1527636732917453e-05, + "loss": 0.002, + "step": 11780 + }, + { + "epoch": 1.2772116218560279, + "grad_norm": 0.5631721019744873, + "learning_rate": 1.1526911359350065e-05, + "loss": 0.0033, + "step": 11781 + }, + { + "epoch": 1.2773200346921074, + "grad_norm": 2.4522876739501953, + "learning_rate": 1.1526185985782679e-05, + "loss": 0.047, + "step": 11782 + }, + { + "epoch": 1.2774284475281874, + "grad_norm": 0.8434100151062012, + "learning_rate": 1.1525460612215291e-05, + "loss": 0.0066, + "step": 11783 + }, + { + "epoch": 1.2775368603642672, + "grad_norm": 0.5388967990875244, + "learning_rate": 1.1524735238647903e-05, + "loss": 0.0016, + "step": 11784 + }, + { + "epoch": 1.277645273200347, + "grad_norm": 1.5623148679733276, + "learning_rate": 1.1524009865080517e-05, + "loss": 0.0315, + "step": 11785 + }, + { + "epoch": 1.2777536860364267, + "grad_norm": 3.234025239944458, + "learning_rate": 1.152328449151313e-05, + "loss": 0.0399, + "step": 11786 + }, + { + "epoch": 1.2778620988725065, + "grad_norm": 1.081760048866272, + "learning_rate": 1.1522559117945741e-05, + "loss": 0.0169, + "step": 11787 + }, + { + "epoch": 1.2779705117085862, + "grad_norm": 19.553640365600586, + "learning_rate": 1.1521833744378357e-05, + "loss": 0.0387, + "step": 11788 + }, + { + "epoch": 1.278078924544666, + "grad_norm": 0.626275897026062, + "learning_rate": 1.1521108370810969e-05, + "loss": 0.0021, + "step": 11789 + }, + { + "epoch": 1.278187337380746, + "grad_norm": 1.1946828365325928, + "learning_rate": 1.1520382997243583e-05, + "loss": 0.0084, + "step": 11790 + }, + { + "epoch": 1.2782957502168257, + "grad_norm": 0.472512423992157, + "learning_rate": 1.1519657623676195e-05, + "loss": 0.006, + "step": 11791 + }, + { + "epoch": 1.2784041630529055, + "grad_norm": 6.344176769256592, + "learning_rate": 1.1518932250108807e-05, + "loss": 0.0623, + "step": 11792 + }, + { + "epoch": 1.2785125758889853, + "grad_norm": 2.131347179412842, + "learning_rate": 1.151820687654142e-05, + "loss": 0.0094, + "step": 11793 + }, + { + "epoch": 1.278620988725065, + "grad_norm": 1.0358364582061768, + "learning_rate": 1.1517481502974033e-05, + "loss": 0.031, + "step": 11794 + }, + { + "epoch": 1.2787294015611448, + "grad_norm": 3.8783888816833496, + "learning_rate": 1.1516756129406645e-05, + "loss": 0.0295, + "step": 11795 + }, + { + "epoch": 1.2788378143972245, + "grad_norm": 2.470076084136963, + "learning_rate": 1.1516030755839257e-05, + "loss": 0.0331, + "step": 11796 + }, + { + "epoch": 1.2789462272333045, + "grad_norm": 1.7516098022460938, + "learning_rate": 1.1515305382271871e-05, + "loss": 0.0082, + "step": 11797 + }, + { + "epoch": 1.2790546400693843, + "grad_norm": 5.6193647384643555, + "learning_rate": 1.1514580008704483e-05, + "loss": 0.0638, + "step": 11798 + }, + { + "epoch": 1.279163052905464, + "grad_norm": 4.3007330894470215, + "learning_rate": 1.1513854635137096e-05, + "loss": 0.0523, + "step": 11799 + }, + { + "epoch": 1.2792714657415438, + "grad_norm": 1.7411943674087524, + "learning_rate": 1.151312926156971e-05, + "loss": 0.0141, + "step": 11800 + }, + { + "epoch": 1.2793798785776236, + "grad_norm": 4.6030707359313965, + "learning_rate": 1.1512403888002322e-05, + "loss": 0.0258, + "step": 11801 + }, + { + "epoch": 1.2794882914137033, + "grad_norm": 1.2621424198150635, + "learning_rate": 1.1511678514434934e-05, + "loss": 0.0327, + "step": 11802 + }, + { + "epoch": 1.2795967042497831, + "grad_norm": 0.9322152137756348, + "learning_rate": 1.1510953140867547e-05, + "loss": 0.0089, + "step": 11803 + }, + { + "epoch": 1.279705117085863, + "grad_norm": 3.192678451538086, + "learning_rate": 1.151022776730016e-05, + "loss": 0.0216, + "step": 11804 + }, + { + "epoch": 1.2798135299219426, + "grad_norm": 1.6059331893920898, + "learning_rate": 1.1509502393732775e-05, + "loss": 0.0134, + "step": 11805 + }, + { + "epoch": 1.2799219427580226, + "grad_norm": 3.7484893798828125, + "learning_rate": 1.1508777020165387e-05, + "loss": 0.0415, + "step": 11806 + }, + { + "epoch": 1.2800303555941024, + "grad_norm": 0.9313836097717285, + "learning_rate": 1.1508051646598e-05, + "loss": 0.0031, + "step": 11807 + }, + { + "epoch": 1.2801387684301822, + "grad_norm": 1.1435991525650024, + "learning_rate": 1.1507326273030612e-05, + "loss": 0.0098, + "step": 11808 + }, + { + "epoch": 1.280247181266262, + "grad_norm": 1.2288728952407837, + "learning_rate": 1.1506600899463225e-05, + "loss": 0.0101, + "step": 11809 + }, + { + "epoch": 1.2803555941023417, + "grad_norm": 1.2355856895446777, + "learning_rate": 1.1505875525895838e-05, + "loss": 0.0113, + "step": 11810 + }, + { + "epoch": 1.2804640069384214, + "grad_norm": 3.7121057510375977, + "learning_rate": 1.150515015232845e-05, + "loss": 0.0347, + "step": 11811 + }, + { + "epoch": 1.2805724197745012, + "grad_norm": 0.1437765210866928, + "learning_rate": 1.1504424778761064e-05, + "loss": 0.0013, + "step": 11812 + }, + { + "epoch": 1.2806808326105812, + "grad_norm": 3.5519473552703857, + "learning_rate": 1.1503699405193676e-05, + "loss": 0.0194, + "step": 11813 + }, + { + "epoch": 1.280789245446661, + "grad_norm": 0.8939430713653564, + "learning_rate": 1.1502974031626288e-05, + "loss": 0.017, + "step": 11814 + }, + { + "epoch": 1.2808976582827407, + "grad_norm": 2.048297882080078, + "learning_rate": 1.1502248658058902e-05, + "loss": 0.0688, + "step": 11815 + }, + { + "epoch": 1.2810060711188205, + "grad_norm": 2.6086812019348145, + "learning_rate": 1.1501523284491514e-05, + "loss": 0.0434, + "step": 11816 + }, + { + "epoch": 1.2811144839549002, + "grad_norm": 0.34612828493118286, + "learning_rate": 1.1500797910924126e-05, + "loss": 0.0028, + "step": 11817 + }, + { + "epoch": 1.28122289679098, + "grad_norm": 0.5556033849716187, + "learning_rate": 1.150007253735674e-05, + "loss": 0.0043, + "step": 11818 + }, + { + "epoch": 1.2813313096270598, + "grad_norm": 5.349969863891602, + "learning_rate": 1.1499347163789352e-05, + "loss": 0.0571, + "step": 11819 + }, + { + "epoch": 1.2814397224631398, + "grad_norm": 0.8301472663879395, + "learning_rate": 1.1498621790221964e-05, + "loss": 0.0145, + "step": 11820 + }, + { + "epoch": 1.2815481352992193, + "grad_norm": 8.722193717956543, + "learning_rate": 1.1497896416654576e-05, + "loss": 0.0155, + "step": 11821 + }, + { + "epoch": 1.2816565481352993, + "grad_norm": 2.441547155380249, + "learning_rate": 1.1497171043087192e-05, + "loss": 0.0039, + "step": 11822 + }, + { + "epoch": 1.281764960971379, + "grad_norm": 3.881908893585205, + "learning_rate": 1.1496445669519804e-05, + "loss": 0.0342, + "step": 11823 + }, + { + "epoch": 1.2818733738074588, + "grad_norm": 5.693874835968018, + "learning_rate": 1.1495720295952418e-05, + "loss": 0.0177, + "step": 11824 + }, + { + "epoch": 1.2819817866435386, + "grad_norm": 1.548944115638733, + "learning_rate": 1.149499492238503e-05, + "loss": 0.0248, + "step": 11825 + }, + { + "epoch": 1.2820901994796183, + "grad_norm": 1.1840174198150635, + "learning_rate": 1.1494269548817642e-05, + "loss": 0.0034, + "step": 11826 + }, + { + "epoch": 1.2821986123156983, + "grad_norm": 1.0672974586486816, + "learning_rate": 1.1493544175250256e-05, + "loss": 0.0029, + "step": 11827 + }, + { + "epoch": 1.2823070251517779, + "grad_norm": 2.975863218307495, + "learning_rate": 1.1492818801682868e-05, + "loss": 0.0265, + "step": 11828 + }, + { + "epoch": 1.2824154379878578, + "grad_norm": 2.5600669384002686, + "learning_rate": 1.149209342811548e-05, + "loss": 0.0256, + "step": 11829 + }, + { + "epoch": 1.2825238508239376, + "grad_norm": 0.6045833230018616, + "learning_rate": 1.1491368054548094e-05, + "loss": 0.0118, + "step": 11830 + }, + { + "epoch": 1.2826322636600174, + "grad_norm": 3.413750648498535, + "learning_rate": 1.1490642680980706e-05, + "loss": 0.0342, + "step": 11831 + }, + { + "epoch": 1.2827406764960971, + "grad_norm": 3.2737040519714355, + "learning_rate": 1.1489917307413318e-05, + "loss": 0.023, + "step": 11832 + }, + { + "epoch": 1.282849089332177, + "grad_norm": 4.377079486846924, + "learning_rate": 1.148919193384593e-05, + "loss": 0.0301, + "step": 11833 + }, + { + "epoch": 1.2829575021682567, + "grad_norm": 1.9013619422912598, + "learning_rate": 1.1488466560278544e-05, + "loss": 0.0096, + "step": 11834 + }, + { + "epoch": 1.2830659150043364, + "grad_norm": 2.628532886505127, + "learning_rate": 1.1487741186711156e-05, + "loss": 0.0532, + "step": 11835 + }, + { + "epoch": 1.2831743278404164, + "grad_norm": 1.1136763095855713, + "learning_rate": 1.1487015813143769e-05, + "loss": 0.027, + "step": 11836 + }, + { + "epoch": 1.2832827406764962, + "grad_norm": 0.7663905024528503, + "learning_rate": 1.1486290439576382e-05, + "loss": 0.0107, + "step": 11837 + }, + { + "epoch": 1.283391153512576, + "grad_norm": 2.7233633995056152, + "learning_rate": 1.1485565066008995e-05, + "loss": 0.0269, + "step": 11838 + }, + { + "epoch": 1.2834995663486557, + "grad_norm": 0.9469456672668457, + "learning_rate": 1.148483969244161e-05, + "loss": 0.0138, + "step": 11839 + }, + { + "epoch": 1.2836079791847355, + "grad_norm": 1.1173419952392578, + "learning_rate": 1.1484114318874222e-05, + "loss": 0.0107, + "step": 11840 + }, + { + "epoch": 1.2837163920208152, + "grad_norm": 1.1014957427978516, + "learning_rate": 1.1483388945306834e-05, + "loss": 0.0179, + "step": 11841 + }, + { + "epoch": 1.283824804856895, + "grad_norm": 1.1695568561553955, + "learning_rate": 1.1482663571739448e-05, + "loss": 0.0036, + "step": 11842 + }, + { + "epoch": 1.283933217692975, + "grad_norm": 0.5793567895889282, + "learning_rate": 1.148193819817206e-05, + "loss": 0.0066, + "step": 11843 + }, + { + "epoch": 1.2840416305290545, + "grad_norm": 1.830100417137146, + "learning_rate": 1.1481212824604672e-05, + "loss": 0.015, + "step": 11844 + }, + { + "epoch": 1.2841500433651345, + "grad_norm": 2.5142133235931396, + "learning_rate": 1.1480487451037285e-05, + "loss": 0.0241, + "step": 11845 + }, + { + "epoch": 1.2842584562012143, + "grad_norm": 4.038412570953369, + "learning_rate": 1.1479762077469898e-05, + "loss": 0.0374, + "step": 11846 + }, + { + "epoch": 1.284366869037294, + "grad_norm": 1.8847447633743286, + "learning_rate": 1.147903670390251e-05, + "loss": 0.0026, + "step": 11847 + }, + { + "epoch": 1.2844752818733738, + "grad_norm": 2.038238286972046, + "learning_rate": 1.1478311330335123e-05, + "loss": 0.0546, + "step": 11848 + }, + { + "epoch": 1.2845836947094535, + "grad_norm": 0.5342165231704712, + "learning_rate": 1.1477585956767737e-05, + "loss": 0.0084, + "step": 11849 + }, + { + "epoch": 1.2846921075455333, + "grad_norm": 1.3990237712860107, + "learning_rate": 1.1476860583200349e-05, + "loss": 0.0193, + "step": 11850 + }, + { + "epoch": 1.284800520381613, + "grad_norm": 2.5910096168518066, + "learning_rate": 1.1476135209632961e-05, + "loss": 0.0113, + "step": 11851 + }, + { + "epoch": 1.284908933217693, + "grad_norm": 4.213073253631592, + "learning_rate": 1.1475409836065575e-05, + "loss": 0.0176, + "step": 11852 + }, + { + "epoch": 1.2850173460537728, + "grad_norm": 2.1867942810058594, + "learning_rate": 1.1474684462498187e-05, + "loss": 0.0286, + "step": 11853 + }, + { + "epoch": 1.2851257588898526, + "grad_norm": 3.5400824546813965, + "learning_rate": 1.1473959088930799e-05, + "loss": 0.0147, + "step": 11854 + }, + { + "epoch": 1.2852341717259324, + "grad_norm": 2.241475820541382, + "learning_rate": 1.1473233715363413e-05, + "loss": 0.0088, + "step": 11855 + }, + { + "epoch": 1.2853425845620121, + "grad_norm": 0.7408382892608643, + "learning_rate": 1.1472508341796025e-05, + "loss": 0.0122, + "step": 11856 + }, + { + "epoch": 1.2854509973980919, + "grad_norm": 3.131955146789551, + "learning_rate": 1.147178296822864e-05, + "loss": 0.0504, + "step": 11857 + }, + { + "epoch": 1.2855594102341716, + "grad_norm": 3.315699815750122, + "learning_rate": 1.1471057594661253e-05, + "loss": 0.0238, + "step": 11858 + }, + { + "epoch": 1.2856678230702516, + "grad_norm": 3.8639800548553467, + "learning_rate": 1.1470332221093865e-05, + "loss": 0.0153, + "step": 11859 + }, + { + "epoch": 1.2857762359063314, + "grad_norm": 1.735235333442688, + "learning_rate": 1.1469606847526477e-05, + "loss": 0.0075, + "step": 11860 + }, + { + "epoch": 1.2858846487424112, + "grad_norm": 1.6827658414840698, + "learning_rate": 1.146888147395909e-05, + "loss": 0.0231, + "step": 11861 + }, + { + "epoch": 1.285993061578491, + "grad_norm": 3.1732170581817627, + "learning_rate": 1.1468156100391703e-05, + "loss": 0.0303, + "step": 11862 + }, + { + "epoch": 1.2861014744145707, + "grad_norm": 3.956406593322754, + "learning_rate": 1.1467430726824315e-05, + "loss": 0.0599, + "step": 11863 + }, + { + "epoch": 1.2862098872506504, + "grad_norm": 9.192933082580566, + "learning_rate": 1.1466705353256929e-05, + "loss": 0.0228, + "step": 11864 + }, + { + "epoch": 1.2863183000867302, + "grad_norm": 3.539268970489502, + "learning_rate": 1.1465979979689541e-05, + "loss": 0.015, + "step": 11865 + }, + { + "epoch": 1.2864267129228102, + "grad_norm": 1.3571789264678955, + "learning_rate": 1.1465254606122153e-05, + "loss": 0.0038, + "step": 11866 + }, + { + "epoch": 1.2865351257588897, + "grad_norm": 2.5212936401367188, + "learning_rate": 1.1464529232554767e-05, + "loss": 0.0124, + "step": 11867 + }, + { + "epoch": 1.2866435385949697, + "grad_norm": 2.0168533325195312, + "learning_rate": 1.146380385898738e-05, + "loss": 0.006, + "step": 11868 + }, + { + "epoch": 1.2867519514310495, + "grad_norm": 0.4573265612125397, + "learning_rate": 1.1463078485419991e-05, + "loss": 0.0087, + "step": 11869 + }, + { + "epoch": 1.2868603642671292, + "grad_norm": 3.0693159103393555, + "learning_rate": 1.1462353111852605e-05, + "loss": 0.0113, + "step": 11870 + }, + { + "epoch": 1.286968777103209, + "grad_norm": 3.834648847579956, + "learning_rate": 1.1461627738285217e-05, + "loss": 0.0232, + "step": 11871 + }, + { + "epoch": 1.2870771899392888, + "grad_norm": 2.465183734893799, + "learning_rate": 1.146090236471783e-05, + "loss": 0.0384, + "step": 11872 + }, + { + "epoch": 1.2871856027753685, + "grad_norm": 0.36501264572143555, + "learning_rate": 1.1460176991150442e-05, + "loss": 0.0073, + "step": 11873 + }, + { + "epoch": 1.2872940156114483, + "grad_norm": 9.566673278808594, + "learning_rate": 1.1459451617583057e-05, + "loss": 0.0257, + "step": 11874 + }, + { + "epoch": 1.2874024284475283, + "grad_norm": 0.30803561210632324, + "learning_rate": 1.145872624401567e-05, + "loss": 0.0052, + "step": 11875 + }, + { + "epoch": 1.287510841283608, + "grad_norm": 1.1592166423797607, + "learning_rate": 1.1458000870448283e-05, + "loss": 0.0059, + "step": 11876 + }, + { + "epoch": 1.2876192541196878, + "grad_norm": 7.816730499267578, + "learning_rate": 1.1457275496880895e-05, + "loss": 0.0197, + "step": 11877 + }, + { + "epoch": 1.2877276669557676, + "grad_norm": 2.866760015487671, + "learning_rate": 1.1456550123313507e-05, + "loss": 0.0117, + "step": 11878 + }, + { + "epoch": 1.2878360797918473, + "grad_norm": 3.079254388809204, + "learning_rate": 1.1455824749746121e-05, + "loss": 0.0444, + "step": 11879 + }, + { + "epoch": 1.287944492627927, + "grad_norm": 2.7519803047180176, + "learning_rate": 1.1455099376178733e-05, + "loss": 0.0239, + "step": 11880 + }, + { + "epoch": 1.2880529054640069, + "grad_norm": 1.5701360702514648, + "learning_rate": 1.1454374002611346e-05, + "loss": 0.0256, + "step": 11881 + }, + { + "epoch": 1.2881613183000868, + "grad_norm": 1.2814825773239136, + "learning_rate": 1.145364862904396e-05, + "loss": 0.0074, + "step": 11882 + }, + { + "epoch": 1.2882697311361664, + "grad_norm": 2.8968796730041504, + "learning_rate": 1.1452923255476571e-05, + "loss": 0.0386, + "step": 11883 + }, + { + "epoch": 1.2883781439722464, + "grad_norm": 1.7822530269622803, + "learning_rate": 1.1452197881909184e-05, + "loss": 0.0146, + "step": 11884 + }, + { + "epoch": 1.2884865568083261, + "grad_norm": 3.4900739192962646, + "learning_rate": 1.1451472508341796e-05, + "loss": 0.0157, + "step": 11885 + }, + { + "epoch": 1.288594969644406, + "grad_norm": 0.41407111287117004, + "learning_rate": 1.145074713477441e-05, + "loss": 0.0025, + "step": 11886 + }, + { + "epoch": 1.2887033824804857, + "grad_norm": 7.358691215515137, + "learning_rate": 1.1450021761207022e-05, + "loss": 0.0531, + "step": 11887 + }, + { + "epoch": 1.2888117953165654, + "grad_norm": 2.571317434310913, + "learning_rate": 1.1449296387639634e-05, + "loss": 0.037, + "step": 11888 + }, + { + "epoch": 1.2889202081526454, + "grad_norm": 1.6539419889450073, + "learning_rate": 1.1448571014072248e-05, + "loss": 0.0199, + "step": 11889 + }, + { + "epoch": 1.289028620988725, + "grad_norm": 2.0668044090270996, + "learning_rate": 1.144784564050486e-05, + "loss": 0.0104, + "step": 11890 + }, + { + "epoch": 1.289137033824805, + "grad_norm": 3.429048776626587, + "learning_rate": 1.1447120266937475e-05, + "loss": 0.0433, + "step": 11891 + }, + { + "epoch": 1.2892454466608847, + "grad_norm": 1.4073314666748047, + "learning_rate": 1.1446394893370088e-05, + "loss": 0.0378, + "step": 11892 + }, + { + "epoch": 1.2893538594969645, + "grad_norm": 3.1240150928497314, + "learning_rate": 1.14456695198027e-05, + "loss": 0.0414, + "step": 11893 + }, + { + "epoch": 1.2894622723330442, + "grad_norm": 2.1009960174560547, + "learning_rate": 1.1444944146235314e-05, + "loss": 0.0174, + "step": 11894 + }, + { + "epoch": 1.289570685169124, + "grad_norm": 4.683756351470947, + "learning_rate": 1.1444218772667926e-05, + "loss": 0.0417, + "step": 11895 + }, + { + "epoch": 1.2896790980052038, + "grad_norm": 2.4591610431671143, + "learning_rate": 1.1443493399100538e-05, + "loss": 0.0292, + "step": 11896 + }, + { + "epoch": 1.2897875108412835, + "grad_norm": 6.3245463371276855, + "learning_rate": 1.144276802553315e-05, + "loss": 0.0354, + "step": 11897 + }, + { + "epoch": 1.2898959236773635, + "grad_norm": 4.167765140533447, + "learning_rate": 1.1442042651965764e-05, + "loss": 0.0242, + "step": 11898 + }, + { + "epoch": 1.2900043365134433, + "grad_norm": 0.17975886166095734, + "learning_rate": 1.1441317278398376e-05, + "loss": 0.0018, + "step": 11899 + }, + { + "epoch": 1.290112749349523, + "grad_norm": 3.195218086242676, + "learning_rate": 1.1440591904830988e-05, + "loss": 0.0331, + "step": 11900 + }, + { + "epoch": 1.2902211621856028, + "grad_norm": 0.0779767706990242, + "learning_rate": 1.1439866531263602e-05, + "loss": 0.0005, + "step": 11901 + }, + { + "epoch": 1.2903295750216826, + "grad_norm": 1.5669907331466675, + "learning_rate": 1.1439141157696214e-05, + "loss": 0.0246, + "step": 11902 + }, + { + "epoch": 1.2904379878577623, + "grad_norm": 8.003105163574219, + "learning_rate": 1.1438415784128826e-05, + "loss": 0.0288, + "step": 11903 + }, + { + "epoch": 1.290546400693842, + "grad_norm": 0.8405659794807434, + "learning_rate": 1.143769041056144e-05, + "loss": 0.0129, + "step": 11904 + }, + { + "epoch": 1.290654813529922, + "grad_norm": 4.532107830047607, + "learning_rate": 1.1436965036994052e-05, + "loss": 0.0678, + "step": 11905 + }, + { + "epoch": 1.2907632263660016, + "grad_norm": 2.227649450302124, + "learning_rate": 1.1436239663426664e-05, + "loss": 0.0348, + "step": 11906 + }, + { + "epoch": 1.2908716392020816, + "grad_norm": 11.408702850341797, + "learning_rate": 1.1435514289859278e-05, + "loss": 0.0313, + "step": 11907 + }, + { + "epoch": 1.2909800520381614, + "grad_norm": 0.9756157994270325, + "learning_rate": 1.1434788916291892e-05, + "loss": 0.0036, + "step": 11908 + }, + { + "epoch": 1.2910884648742411, + "grad_norm": 2.200488805770874, + "learning_rate": 1.1434063542724506e-05, + "loss": 0.0501, + "step": 11909 + }, + { + "epoch": 1.2911968777103209, + "grad_norm": 1.1152160167694092, + "learning_rate": 1.1433338169157118e-05, + "loss": 0.0136, + "step": 11910 + }, + { + "epoch": 1.2913052905464006, + "grad_norm": 1.1397026777267456, + "learning_rate": 1.143261279558973e-05, + "loss": 0.0046, + "step": 11911 + }, + { + "epoch": 1.2914137033824804, + "grad_norm": 0.16868330538272858, + "learning_rate": 1.1431887422022342e-05, + "loss": 0.0019, + "step": 11912 + }, + { + "epoch": 1.2915221162185602, + "grad_norm": 8.145560264587402, + "learning_rate": 1.1431162048454956e-05, + "loss": 0.0339, + "step": 11913 + }, + { + "epoch": 1.2916305290546402, + "grad_norm": 4.434595584869385, + "learning_rate": 1.1430436674887568e-05, + "loss": 0.0497, + "step": 11914 + }, + { + "epoch": 1.29173894189072, + "grad_norm": 3.893749713897705, + "learning_rate": 1.142971130132018e-05, + "loss": 0.0233, + "step": 11915 + }, + { + "epoch": 1.2918473547267997, + "grad_norm": 0.9293097257614136, + "learning_rate": 1.1428985927752794e-05, + "loss": 0.0101, + "step": 11916 + }, + { + "epoch": 1.2919557675628794, + "grad_norm": 3.4264304637908936, + "learning_rate": 1.1428260554185406e-05, + "loss": 0.0301, + "step": 11917 + }, + { + "epoch": 1.2920641803989592, + "grad_norm": 1.584109902381897, + "learning_rate": 1.1427535180618019e-05, + "loss": 0.0396, + "step": 11918 + }, + { + "epoch": 1.292172593235039, + "grad_norm": 1.4659603834152222, + "learning_rate": 1.1426809807050632e-05, + "loss": 0.0195, + "step": 11919 + }, + { + "epoch": 1.2922810060711187, + "grad_norm": 0.540863037109375, + "learning_rate": 1.1426084433483245e-05, + "loss": 0.0054, + "step": 11920 + }, + { + "epoch": 1.2923894189071987, + "grad_norm": 0.9589094519615173, + "learning_rate": 1.1425359059915857e-05, + "loss": 0.0173, + "step": 11921 + }, + { + "epoch": 1.2924978317432785, + "grad_norm": 0.2576214671134949, + "learning_rate": 1.142463368634847e-05, + "loss": 0.0026, + "step": 11922 + }, + { + "epoch": 1.2926062445793582, + "grad_norm": 0.989271342754364, + "learning_rate": 1.1423908312781083e-05, + "loss": 0.0077, + "step": 11923 + }, + { + "epoch": 1.292714657415438, + "grad_norm": 0.14599238336086273, + "learning_rate": 1.1423182939213695e-05, + "loss": 0.0014, + "step": 11924 + }, + { + "epoch": 1.2928230702515178, + "grad_norm": 1.9217232465744019, + "learning_rate": 1.1422457565646307e-05, + "loss": 0.0205, + "step": 11925 + }, + { + "epoch": 1.2929314830875975, + "grad_norm": 1.3083120584487915, + "learning_rate": 1.1421732192078922e-05, + "loss": 0.01, + "step": 11926 + }, + { + "epoch": 1.2930398959236773, + "grad_norm": 2.3726725578308105, + "learning_rate": 1.1421006818511535e-05, + "loss": 0.0685, + "step": 11927 + }, + { + "epoch": 1.2931483087597573, + "grad_norm": 0.36587440967559814, + "learning_rate": 1.1420281444944148e-05, + "loss": 0.0023, + "step": 11928 + }, + { + "epoch": 1.2932567215958368, + "grad_norm": 1.8346242904663086, + "learning_rate": 1.141955607137676e-05, + "loss": 0.0528, + "step": 11929 + }, + { + "epoch": 1.2933651344319168, + "grad_norm": 2.437711715698242, + "learning_rate": 1.1418830697809373e-05, + "loss": 0.0171, + "step": 11930 + }, + { + "epoch": 1.2934735472679966, + "grad_norm": 2.0134963989257812, + "learning_rate": 1.1418105324241987e-05, + "loss": 0.0179, + "step": 11931 + }, + { + "epoch": 1.2935819601040763, + "grad_norm": 9.019145965576172, + "learning_rate": 1.1417379950674599e-05, + "loss": 0.048, + "step": 11932 + }, + { + "epoch": 1.293690372940156, + "grad_norm": 1.504395604133606, + "learning_rate": 1.141665457710721e-05, + "loss": 0.0244, + "step": 11933 + }, + { + "epoch": 1.2937987857762359, + "grad_norm": 9.473291397094727, + "learning_rate": 1.1415929203539825e-05, + "loss": 0.0618, + "step": 11934 + }, + { + "epoch": 1.2939071986123156, + "grad_norm": 0.8194406032562256, + "learning_rate": 1.1415203829972437e-05, + "loss": 0.0053, + "step": 11935 + }, + { + "epoch": 1.2940156114483954, + "grad_norm": 0.8660087585449219, + "learning_rate": 1.1414478456405049e-05, + "loss": 0.0038, + "step": 11936 + }, + { + "epoch": 1.2941240242844754, + "grad_norm": 1.4617855548858643, + "learning_rate": 1.1413753082837661e-05, + "loss": 0.0209, + "step": 11937 + }, + { + "epoch": 1.2942324371205551, + "grad_norm": 1.0381817817687988, + "learning_rate": 1.1413027709270275e-05, + "loss": 0.0089, + "step": 11938 + }, + { + "epoch": 1.294340849956635, + "grad_norm": 1.2492777109146118, + "learning_rate": 1.1412302335702887e-05, + "loss": 0.0415, + "step": 11939 + }, + { + "epoch": 1.2944492627927147, + "grad_norm": 4.591025352478027, + "learning_rate": 1.14115769621355e-05, + "loss": 0.0335, + "step": 11940 + }, + { + "epoch": 1.2945576756287944, + "grad_norm": 1.5646578073501587, + "learning_rate": 1.1410851588568113e-05, + "loss": 0.0155, + "step": 11941 + }, + { + "epoch": 1.2946660884648742, + "grad_norm": 3.5059027671813965, + "learning_rate": 1.1410126215000725e-05, + "loss": 0.0302, + "step": 11942 + }, + { + "epoch": 1.294774501300954, + "grad_norm": 0.338072270154953, + "learning_rate": 1.140940084143334e-05, + "loss": 0.0045, + "step": 11943 + }, + { + "epoch": 1.294882914137034, + "grad_norm": 14.300735473632812, + "learning_rate": 1.1408675467865953e-05, + "loss": 0.0316, + "step": 11944 + }, + { + "epoch": 1.2949913269731137, + "grad_norm": 1.3863494396209717, + "learning_rate": 1.1407950094298565e-05, + "loss": 0.0125, + "step": 11945 + }, + { + "epoch": 1.2950997398091935, + "grad_norm": 0.44315478205680847, + "learning_rate": 1.1407224720731179e-05, + "loss": 0.0048, + "step": 11946 + }, + { + "epoch": 1.2952081526452732, + "grad_norm": 2.433008909225464, + "learning_rate": 1.1406499347163791e-05, + "loss": 0.0177, + "step": 11947 + }, + { + "epoch": 1.295316565481353, + "grad_norm": 3.4068427085876465, + "learning_rate": 1.1405773973596403e-05, + "loss": 0.0317, + "step": 11948 + }, + { + "epoch": 1.2954249783174328, + "grad_norm": 0.6703949570655823, + "learning_rate": 1.1405048600029017e-05, + "loss": 0.0087, + "step": 11949 + }, + { + "epoch": 1.2955333911535125, + "grad_norm": 1.7064348459243774, + "learning_rate": 1.1404323226461629e-05, + "loss": 0.0523, + "step": 11950 + }, + { + "epoch": 1.2956418039895925, + "grad_norm": 1.703890323638916, + "learning_rate": 1.1403597852894241e-05, + "loss": 0.0186, + "step": 11951 + }, + { + "epoch": 1.295750216825672, + "grad_norm": 3.3803834915161133, + "learning_rate": 1.1402872479326853e-05, + "loss": 0.0071, + "step": 11952 + }, + { + "epoch": 1.295858629661752, + "grad_norm": 0.9044457077980042, + "learning_rate": 1.1402147105759467e-05, + "loss": 0.0078, + "step": 11953 + }, + { + "epoch": 1.2959670424978318, + "grad_norm": 5.226736545562744, + "learning_rate": 1.140142173219208e-05, + "loss": 0.006, + "step": 11954 + }, + { + "epoch": 1.2960754553339116, + "grad_norm": 0.2627963125705719, + "learning_rate": 1.1400696358624692e-05, + "loss": 0.0028, + "step": 11955 + }, + { + "epoch": 1.2961838681699913, + "grad_norm": 1.0330034494400024, + "learning_rate": 1.1399970985057305e-05, + "loss": 0.0113, + "step": 11956 + }, + { + "epoch": 1.296292281006071, + "grad_norm": 1.3214483261108398, + "learning_rate": 1.1399245611489918e-05, + "loss": 0.0208, + "step": 11957 + }, + { + "epoch": 1.2964006938421508, + "grad_norm": 3.2071635723114014, + "learning_rate": 1.139852023792253e-05, + "loss": 0.0198, + "step": 11958 + }, + { + "epoch": 1.2965091066782306, + "grad_norm": 0.2287968397140503, + "learning_rate": 1.1397794864355144e-05, + "loss": 0.0022, + "step": 11959 + }, + { + "epoch": 1.2966175195143106, + "grad_norm": 1.6360363960266113, + "learning_rate": 1.1397069490787757e-05, + "loss": 0.0172, + "step": 11960 + }, + { + "epoch": 1.2967259323503904, + "grad_norm": 2.6854171752929688, + "learning_rate": 1.1396344117220371e-05, + "loss": 0.0182, + "step": 11961 + }, + { + "epoch": 1.2968343451864701, + "grad_norm": 4.694920063018799, + "learning_rate": 1.1395618743652983e-05, + "loss": 0.0522, + "step": 11962 + }, + { + "epoch": 1.2969427580225499, + "grad_norm": 4.3327484130859375, + "learning_rate": 1.1394893370085595e-05, + "loss": 0.044, + "step": 11963 + }, + { + "epoch": 1.2970511708586296, + "grad_norm": 5.451180458068848, + "learning_rate": 1.1394167996518208e-05, + "loss": 0.087, + "step": 11964 + }, + { + "epoch": 1.2971595836947094, + "grad_norm": 1.4167102575302124, + "learning_rate": 1.1393442622950821e-05, + "loss": 0.0146, + "step": 11965 + }, + { + "epoch": 1.2972679965307892, + "grad_norm": 1.4719648361206055, + "learning_rate": 1.1392717249383434e-05, + "loss": 0.0291, + "step": 11966 + }, + { + "epoch": 1.2973764093668692, + "grad_norm": 5.464376449584961, + "learning_rate": 1.1391991875816046e-05, + "loss": 0.0417, + "step": 11967 + }, + { + "epoch": 1.2974848222029487, + "grad_norm": 1.9648077487945557, + "learning_rate": 1.139126650224866e-05, + "loss": 0.0398, + "step": 11968 + }, + { + "epoch": 1.2975932350390287, + "grad_norm": 2.0524439811706543, + "learning_rate": 1.1390541128681272e-05, + "loss": 0.0568, + "step": 11969 + }, + { + "epoch": 1.2977016478751084, + "grad_norm": 3.1146152019500732, + "learning_rate": 1.1389815755113884e-05, + "loss": 0.0222, + "step": 11970 + }, + { + "epoch": 1.2978100607111882, + "grad_norm": 1.8110172748565674, + "learning_rate": 1.1389090381546498e-05, + "loss": 0.021, + "step": 11971 + }, + { + "epoch": 1.297918473547268, + "grad_norm": 3.3133480548858643, + "learning_rate": 1.138836500797911e-05, + "loss": 0.0612, + "step": 11972 + }, + { + "epoch": 1.2980268863833477, + "grad_norm": 5.653692245483398, + "learning_rate": 1.1387639634411722e-05, + "loss": 0.0676, + "step": 11973 + }, + { + "epoch": 1.2981352992194277, + "grad_norm": 0.8076189160346985, + "learning_rate": 1.1386914260844336e-05, + "loss": 0.0137, + "step": 11974 + }, + { + "epoch": 1.2982437120555073, + "grad_norm": 2.6987879276275635, + "learning_rate": 1.1386188887276948e-05, + "loss": 0.0107, + "step": 11975 + }, + { + "epoch": 1.2983521248915872, + "grad_norm": 1.8801584243774414, + "learning_rate": 1.138546351370956e-05, + "loss": 0.0431, + "step": 11976 + }, + { + "epoch": 1.298460537727667, + "grad_norm": 0.7647951245307922, + "learning_rate": 1.1384738140142172e-05, + "loss": 0.0079, + "step": 11977 + }, + { + "epoch": 1.2985689505637468, + "grad_norm": 0.92032390832901, + "learning_rate": 1.1384012766574788e-05, + "loss": 0.0096, + "step": 11978 + }, + { + "epoch": 1.2986773633998265, + "grad_norm": 5.34149694442749, + "learning_rate": 1.13832873930074e-05, + "loss": 0.0207, + "step": 11979 + }, + { + "epoch": 1.2987857762359063, + "grad_norm": 0.49675866961479187, + "learning_rate": 1.1382562019440014e-05, + "loss": 0.006, + "step": 11980 + }, + { + "epoch": 1.298894189071986, + "grad_norm": 1.6426920890808105, + "learning_rate": 1.1381836645872626e-05, + "loss": 0.0071, + "step": 11981 + }, + { + "epoch": 1.2990026019080658, + "grad_norm": 2.0121755599975586, + "learning_rate": 1.1381111272305238e-05, + "loss": 0.0196, + "step": 11982 + }, + { + "epoch": 1.2991110147441458, + "grad_norm": 2.8855278491973877, + "learning_rate": 1.1380385898737852e-05, + "loss": 0.0139, + "step": 11983 + }, + { + "epoch": 1.2992194275802256, + "grad_norm": 0.46979793906211853, + "learning_rate": 1.1379660525170464e-05, + "loss": 0.009, + "step": 11984 + }, + { + "epoch": 1.2993278404163053, + "grad_norm": 1.5382381677627563, + "learning_rate": 1.1378935151603076e-05, + "loss": 0.0338, + "step": 11985 + }, + { + "epoch": 1.299436253252385, + "grad_norm": 5.027726173400879, + "learning_rate": 1.137820977803569e-05, + "loss": 0.0276, + "step": 11986 + }, + { + "epoch": 1.2995446660884649, + "grad_norm": 2.9508557319641113, + "learning_rate": 1.1377484404468302e-05, + "loss": 0.044, + "step": 11987 + }, + { + "epoch": 1.2996530789245446, + "grad_norm": 10.191129684448242, + "learning_rate": 1.1376759030900914e-05, + "loss": 0.0136, + "step": 11988 + }, + { + "epoch": 1.2997614917606244, + "grad_norm": 0.6647191047668457, + "learning_rate": 1.1376033657333526e-05, + "loss": 0.0068, + "step": 11989 + }, + { + "epoch": 1.2998699045967044, + "grad_norm": 2.1151044368743896, + "learning_rate": 1.137530828376614e-05, + "loss": 0.0842, + "step": 11990 + }, + { + "epoch": 1.299978317432784, + "grad_norm": 0.6239014267921448, + "learning_rate": 1.1374582910198752e-05, + "loss": 0.0109, + "step": 11991 + }, + { + "epoch": 1.300086730268864, + "grad_norm": 0.1841917783021927, + "learning_rate": 1.1373857536631365e-05, + "loss": 0.0018, + "step": 11992 + }, + { + "epoch": 1.3001951431049437, + "grad_norm": 3.7772538661956787, + "learning_rate": 1.1373132163063978e-05, + "loss": 0.0164, + "step": 11993 + }, + { + "epoch": 1.3003035559410234, + "grad_norm": 2.5585720539093018, + "learning_rate": 1.137240678949659e-05, + "loss": 0.0562, + "step": 11994 + }, + { + "epoch": 1.3004119687771032, + "grad_norm": 4.220559597015381, + "learning_rate": 1.1371681415929206e-05, + "loss": 0.0264, + "step": 11995 + }, + { + "epoch": 1.300520381613183, + "grad_norm": 2.0183045864105225, + "learning_rate": 1.1370956042361818e-05, + "loss": 0.0545, + "step": 11996 + }, + { + "epoch": 1.3006287944492627, + "grad_norm": 7.610161781311035, + "learning_rate": 1.137023066879443e-05, + "loss": 0.01, + "step": 11997 + }, + { + "epoch": 1.3007372072853425, + "grad_norm": 1.191515564918518, + "learning_rate": 1.1369505295227044e-05, + "loss": 0.0205, + "step": 11998 + }, + { + "epoch": 1.3008456201214225, + "grad_norm": 1.9084415435791016, + "learning_rate": 1.1368779921659656e-05, + "loss": 0.0138, + "step": 11999 + }, + { + "epoch": 1.3009540329575022, + "grad_norm": 1.241932988166809, + "learning_rate": 1.1368054548092268e-05, + "loss": 0.0087, + "step": 12000 + }, + { + "epoch": 1.301062445793582, + "grad_norm": 2.5723068714141846, + "learning_rate": 1.1367329174524882e-05, + "loss": 0.0155, + "step": 12001 + }, + { + "epoch": 1.3011708586296618, + "grad_norm": 0.248820960521698, + "learning_rate": 1.1366603800957494e-05, + "loss": 0.0024, + "step": 12002 + }, + { + "epoch": 1.3012792714657415, + "grad_norm": 3.1384994983673096, + "learning_rate": 1.1365878427390107e-05, + "loss": 0.0348, + "step": 12003 + }, + { + "epoch": 1.3013876843018213, + "grad_norm": 0.556496262550354, + "learning_rate": 1.1365153053822719e-05, + "loss": 0.0079, + "step": 12004 + }, + { + "epoch": 1.301496097137901, + "grad_norm": 4.077425003051758, + "learning_rate": 1.1364427680255333e-05, + "loss": 0.0169, + "step": 12005 + }, + { + "epoch": 1.301604509973981, + "grad_norm": 2.1867570877075195, + "learning_rate": 1.1363702306687945e-05, + "loss": 0.0168, + "step": 12006 + }, + { + "epoch": 1.3017129228100608, + "grad_norm": 1.3824641704559326, + "learning_rate": 1.1362976933120557e-05, + "loss": 0.0367, + "step": 12007 + }, + { + "epoch": 1.3018213356461406, + "grad_norm": 1.708643913269043, + "learning_rate": 1.136225155955317e-05, + "loss": 0.0257, + "step": 12008 + }, + { + "epoch": 1.3019297484822203, + "grad_norm": 1.977794885635376, + "learning_rate": 1.1361526185985783e-05, + "loss": 0.0264, + "step": 12009 + }, + { + "epoch": 1.3020381613183, + "grad_norm": 1.1328206062316895, + "learning_rate": 1.1360800812418395e-05, + "loss": 0.0114, + "step": 12010 + }, + { + "epoch": 1.3021465741543798, + "grad_norm": 2.766172170639038, + "learning_rate": 1.1360075438851009e-05, + "loss": 0.0093, + "step": 12011 + }, + { + "epoch": 1.3022549869904596, + "grad_norm": 1.3972961902618408, + "learning_rate": 1.1359350065283623e-05, + "loss": 0.0047, + "step": 12012 + }, + { + "epoch": 1.3023633998265396, + "grad_norm": 1.9413259029388428, + "learning_rate": 1.1358624691716237e-05, + "loss": 0.0153, + "step": 12013 + }, + { + "epoch": 1.3024718126626191, + "grad_norm": 0.4873514473438263, + "learning_rate": 1.1357899318148849e-05, + "loss": 0.0051, + "step": 12014 + }, + { + "epoch": 1.3025802254986991, + "grad_norm": 0.7754082083702087, + "learning_rate": 1.135717394458146e-05, + "loss": 0.0091, + "step": 12015 + }, + { + "epoch": 1.3026886383347789, + "grad_norm": 4.392388820648193, + "learning_rate": 1.1356448571014073e-05, + "loss": 0.0587, + "step": 12016 + }, + { + "epoch": 1.3027970511708586, + "grad_norm": 3.3671607971191406, + "learning_rate": 1.1355723197446687e-05, + "loss": 0.0473, + "step": 12017 + }, + { + "epoch": 1.3029054640069384, + "grad_norm": 1.3911876678466797, + "learning_rate": 1.1354997823879299e-05, + "loss": 0.0107, + "step": 12018 + }, + { + "epoch": 1.3030138768430182, + "grad_norm": 3.050536632537842, + "learning_rate": 1.1354272450311911e-05, + "loss": 0.0227, + "step": 12019 + }, + { + "epoch": 1.303122289679098, + "grad_norm": 2.919222116470337, + "learning_rate": 1.1353547076744525e-05, + "loss": 0.0576, + "step": 12020 + }, + { + "epoch": 1.3032307025151777, + "grad_norm": 2.607517719268799, + "learning_rate": 1.1352821703177137e-05, + "loss": 0.0122, + "step": 12021 + }, + { + "epoch": 1.3033391153512577, + "grad_norm": 0.18755961954593658, + "learning_rate": 1.135209632960975e-05, + "loss": 0.002, + "step": 12022 + }, + { + "epoch": 1.3034475281873374, + "grad_norm": 1.3217015266418457, + "learning_rate": 1.1351370956042363e-05, + "loss": 0.0288, + "step": 12023 + }, + { + "epoch": 1.3035559410234172, + "grad_norm": 0.9875718951225281, + "learning_rate": 1.1350645582474975e-05, + "loss": 0.0088, + "step": 12024 + }, + { + "epoch": 1.303664353859497, + "grad_norm": 2.6831018924713135, + "learning_rate": 1.1349920208907587e-05, + "loss": 0.0683, + "step": 12025 + }, + { + "epoch": 1.3037727666955767, + "grad_norm": 0.5128638744354248, + "learning_rate": 1.1349194835340201e-05, + "loss": 0.0027, + "step": 12026 + }, + { + "epoch": 1.3038811795316565, + "grad_norm": 4.514005661010742, + "learning_rate": 1.1348469461772813e-05, + "loss": 0.0841, + "step": 12027 + }, + { + "epoch": 1.3039895923677363, + "grad_norm": 7.289529800415039, + "learning_rate": 1.1347744088205425e-05, + "loss": 0.025, + "step": 12028 + }, + { + "epoch": 1.3040980052038162, + "grad_norm": 0.42040035128593445, + "learning_rate": 1.1347018714638041e-05, + "loss": 0.002, + "step": 12029 + }, + { + "epoch": 1.3042064180398958, + "grad_norm": 2.715160846710205, + "learning_rate": 1.1346293341070653e-05, + "loss": 0.07, + "step": 12030 + }, + { + "epoch": 1.3043148308759758, + "grad_norm": 2.4273898601531982, + "learning_rate": 1.1345567967503265e-05, + "loss": 0.0163, + "step": 12031 + }, + { + "epoch": 1.3044232437120555, + "grad_norm": 2.6119232177734375, + "learning_rate": 1.1344842593935879e-05, + "loss": 0.0528, + "step": 12032 + }, + { + "epoch": 1.3045316565481353, + "grad_norm": 3.090576648712158, + "learning_rate": 1.1344117220368491e-05, + "loss": 0.0244, + "step": 12033 + }, + { + "epoch": 1.304640069384215, + "grad_norm": 4.093257427215576, + "learning_rate": 1.1343391846801103e-05, + "loss": 0.0627, + "step": 12034 + }, + { + "epoch": 1.3047484822202948, + "grad_norm": 1.7890543937683105, + "learning_rate": 1.1342666473233717e-05, + "loss": 0.0096, + "step": 12035 + }, + { + "epoch": 1.3048568950563748, + "grad_norm": 3.0267486572265625, + "learning_rate": 1.134194109966633e-05, + "loss": 0.047, + "step": 12036 + }, + { + "epoch": 1.3049653078924544, + "grad_norm": 5.45891809463501, + "learning_rate": 1.1341215726098942e-05, + "loss": 0.0576, + "step": 12037 + }, + { + "epoch": 1.3050737207285343, + "grad_norm": 0.1511220782995224, + "learning_rate": 1.1340490352531555e-05, + "loss": 0.002, + "step": 12038 + }, + { + "epoch": 1.305182133564614, + "grad_norm": 2.1353096961975098, + "learning_rate": 1.1339764978964167e-05, + "loss": 0.0297, + "step": 12039 + }, + { + "epoch": 1.3052905464006939, + "grad_norm": 5.187039852142334, + "learning_rate": 1.133903960539678e-05, + "loss": 0.0435, + "step": 12040 + }, + { + "epoch": 1.3053989592367736, + "grad_norm": 0.4237084984779358, + "learning_rate": 1.1338314231829392e-05, + "loss": 0.0056, + "step": 12041 + }, + { + "epoch": 1.3055073720728534, + "grad_norm": 2.4610400199890137, + "learning_rate": 1.1337588858262006e-05, + "loss": 0.0117, + "step": 12042 + }, + { + "epoch": 1.3056157849089332, + "grad_norm": 2.8712971210479736, + "learning_rate": 1.1336863484694618e-05, + "loss": 0.0173, + "step": 12043 + }, + { + "epoch": 1.305724197745013, + "grad_norm": 1.5158504247665405, + "learning_rate": 1.133613811112723e-05, + "loss": 0.0259, + "step": 12044 + }, + { + "epoch": 1.305832610581093, + "grad_norm": 6.475000858306885, + "learning_rate": 1.1335412737559844e-05, + "loss": 0.0818, + "step": 12045 + }, + { + "epoch": 1.3059410234171727, + "grad_norm": 1.2118406295776367, + "learning_rate": 1.1334687363992456e-05, + "loss": 0.0273, + "step": 12046 + }, + { + "epoch": 1.3060494362532524, + "grad_norm": 2.842216730117798, + "learning_rate": 1.1333961990425071e-05, + "loss": 0.026, + "step": 12047 + }, + { + "epoch": 1.3061578490893322, + "grad_norm": 3.744492530822754, + "learning_rate": 1.1333236616857684e-05, + "loss": 0.0072, + "step": 12048 + }, + { + "epoch": 1.306266261925412, + "grad_norm": 6.039683818817139, + "learning_rate": 1.1332511243290296e-05, + "loss": 0.0428, + "step": 12049 + }, + { + "epoch": 1.3063746747614917, + "grad_norm": 1.7393759489059448, + "learning_rate": 1.133178586972291e-05, + "loss": 0.0348, + "step": 12050 + }, + { + "epoch": 1.3064830875975715, + "grad_norm": 0.7560756206512451, + "learning_rate": 1.1331060496155522e-05, + "loss": 0.0039, + "step": 12051 + }, + { + "epoch": 1.3065915004336515, + "grad_norm": 2.3473782539367676, + "learning_rate": 1.1330335122588134e-05, + "loss": 0.0402, + "step": 12052 + }, + { + "epoch": 1.306699913269731, + "grad_norm": 3.2011170387268066, + "learning_rate": 1.1329609749020748e-05, + "loss": 0.018, + "step": 12053 + }, + { + "epoch": 1.306808326105811, + "grad_norm": 1.0573426485061646, + "learning_rate": 1.132888437545336e-05, + "loss": 0.0132, + "step": 12054 + }, + { + "epoch": 1.3069167389418908, + "grad_norm": 0.24900373816490173, + "learning_rate": 1.1328159001885972e-05, + "loss": 0.0035, + "step": 12055 + }, + { + "epoch": 1.3070251517779705, + "grad_norm": 0.49540477991104126, + "learning_rate": 1.1327433628318584e-05, + "loss": 0.0055, + "step": 12056 + }, + { + "epoch": 1.3071335646140503, + "grad_norm": 1.193881869316101, + "learning_rate": 1.1326708254751198e-05, + "loss": 0.0327, + "step": 12057 + }, + { + "epoch": 1.30724197745013, + "grad_norm": 0.3911750316619873, + "learning_rate": 1.132598288118381e-05, + "loss": 0.0043, + "step": 12058 + }, + { + "epoch": 1.3073503902862098, + "grad_norm": 5.559259414672852, + "learning_rate": 1.1325257507616422e-05, + "loss": 0.0202, + "step": 12059 + }, + { + "epoch": 1.3074588031222896, + "grad_norm": 0.6209067106246948, + "learning_rate": 1.1324532134049036e-05, + "loss": 0.0053, + "step": 12060 + }, + { + "epoch": 1.3075672159583696, + "grad_norm": 2.6058390140533447, + "learning_rate": 1.1323806760481648e-05, + "loss": 0.0203, + "step": 12061 + }, + { + "epoch": 1.3076756287944493, + "grad_norm": 1.5178738832473755, + "learning_rate": 1.132308138691426e-05, + "loss": 0.0416, + "step": 12062 + }, + { + "epoch": 1.307784041630529, + "grad_norm": 3.020810604095459, + "learning_rate": 1.1322356013346874e-05, + "loss": 0.0092, + "step": 12063 + }, + { + "epoch": 1.3078924544666088, + "grad_norm": 0.6314139366149902, + "learning_rate": 1.1321630639779488e-05, + "loss": 0.004, + "step": 12064 + }, + { + "epoch": 1.3080008673026886, + "grad_norm": 2.6311540603637695, + "learning_rate": 1.1320905266212102e-05, + "loss": 0.0376, + "step": 12065 + }, + { + "epoch": 1.3081092801387684, + "grad_norm": 0.8903903961181641, + "learning_rate": 1.1320179892644714e-05, + "loss": 0.0204, + "step": 12066 + }, + { + "epoch": 1.3082176929748481, + "grad_norm": 2.487748146057129, + "learning_rate": 1.1319454519077326e-05, + "loss": 0.0567, + "step": 12067 + }, + { + "epoch": 1.3083261058109281, + "grad_norm": 5.482598781585693, + "learning_rate": 1.1318729145509938e-05, + "loss": 0.0236, + "step": 12068 + }, + { + "epoch": 1.3084345186470079, + "grad_norm": 4.416208744049072, + "learning_rate": 1.1318003771942552e-05, + "loss": 0.0453, + "step": 12069 + }, + { + "epoch": 1.3085429314830876, + "grad_norm": 1.0628925561904907, + "learning_rate": 1.1317278398375164e-05, + "loss": 0.0086, + "step": 12070 + }, + { + "epoch": 1.3086513443191674, + "grad_norm": 2.6247828006744385, + "learning_rate": 1.1316553024807776e-05, + "loss": 0.0176, + "step": 12071 + }, + { + "epoch": 1.3087597571552472, + "grad_norm": 1.9179530143737793, + "learning_rate": 1.131582765124039e-05, + "loss": 0.032, + "step": 12072 + }, + { + "epoch": 1.308868169991327, + "grad_norm": 2.214423656463623, + "learning_rate": 1.1315102277673002e-05, + "loss": 0.0132, + "step": 12073 + }, + { + "epoch": 1.3089765828274067, + "grad_norm": 1.2458810806274414, + "learning_rate": 1.1314376904105615e-05, + "loss": 0.0181, + "step": 12074 + }, + { + "epoch": 1.3090849956634867, + "grad_norm": 2.286007881164551, + "learning_rate": 1.1313651530538228e-05, + "loss": 0.0161, + "step": 12075 + }, + { + "epoch": 1.3091934084995662, + "grad_norm": 1.8700382709503174, + "learning_rate": 1.131292615697084e-05, + "loss": 0.0092, + "step": 12076 + }, + { + "epoch": 1.3093018213356462, + "grad_norm": 0.2231234312057495, + "learning_rate": 1.1312200783403453e-05, + "loss": 0.0032, + "step": 12077 + }, + { + "epoch": 1.309410234171726, + "grad_norm": 2.869114637374878, + "learning_rate": 1.1311475409836066e-05, + "loss": 0.0476, + "step": 12078 + }, + { + "epoch": 1.3095186470078057, + "grad_norm": 0.3552064299583435, + "learning_rate": 1.1310750036268679e-05, + "loss": 0.0022, + "step": 12079 + }, + { + "epoch": 1.3096270598438855, + "grad_norm": 2.725510597229004, + "learning_rate": 1.131002466270129e-05, + "loss": 0.0368, + "step": 12080 + }, + { + "epoch": 1.3097354726799653, + "grad_norm": 0.7716849446296692, + "learning_rate": 1.1309299289133906e-05, + "loss": 0.0122, + "step": 12081 + }, + { + "epoch": 1.309843885516045, + "grad_norm": 3.1471145153045654, + "learning_rate": 1.1308573915566518e-05, + "loss": 0.0592, + "step": 12082 + }, + { + "epoch": 1.3099522983521248, + "grad_norm": 4.457059383392334, + "learning_rate": 1.130784854199913e-05, + "loss": 0.013, + "step": 12083 + }, + { + "epoch": 1.3100607111882048, + "grad_norm": 2.0959389209747314, + "learning_rate": 1.1307123168431744e-05, + "loss": 0.0187, + "step": 12084 + }, + { + "epoch": 1.3101691240242845, + "grad_norm": 1.7522417306900024, + "learning_rate": 1.1306397794864357e-05, + "loss": 0.0116, + "step": 12085 + }, + { + "epoch": 1.3102775368603643, + "grad_norm": 0.5337127447128296, + "learning_rate": 1.1305672421296969e-05, + "loss": 0.0018, + "step": 12086 + }, + { + "epoch": 1.310385949696444, + "grad_norm": 2.9964263439178467, + "learning_rate": 1.1304947047729583e-05, + "loss": 0.0343, + "step": 12087 + }, + { + "epoch": 1.3104943625325238, + "grad_norm": 6.77292537689209, + "learning_rate": 1.1304221674162195e-05, + "loss": 0.0779, + "step": 12088 + }, + { + "epoch": 1.3106027753686036, + "grad_norm": 2.0304205417633057, + "learning_rate": 1.1303496300594807e-05, + "loss": 0.0126, + "step": 12089 + }, + { + "epoch": 1.3107111882046834, + "grad_norm": 4.328827857971191, + "learning_rate": 1.130277092702742e-05, + "loss": 0.0615, + "step": 12090 + }, + { + "epoch": 1.3108196010407633, + "grad_norm": 2.582791805267334, + "learning_rate": 1.1302045553460033e-05, + "loss": 0.0172, + "step": 12091 + }, + { + "epoch": 1.310928013876843, + "grad_norm": 7.501500129699707, + "learning_rate": 1.1301320179892645e-05, + "loss": 0.0337, + "step": 12092 + }, + { + "epoch": 1.3110364267129229, + "grad_norm": 1.038116693496704, + "learning_rate": 1.1300594806325257e-05, + "loss": 0.0129, + "step": 12093 + }, + { + "epoch": 1.3111448395490026, + "grad_norm": 3.338987112045288, + "learning_rate": 1.1299869432757871e-05, + "loss": 0.0315, + "step": 12094 + }, + { + "epoch": 1.3112532523850824, + "grad_norm": 1.1900458335876465, + "learning_rate": 1.1299144059190483e-05, + "loss": 0.0356, + "step": 12095 + }, + { + "epoch": 1.3113616652211622, + "grad_norm": 4.720247745513916, + "learning_rate": 1.1298418685623095e-05, + "loss": 0.0324, + "step": 12096 + }, + { + "epoch": 1.311470078057242, + "grad_norm": 11.472301483154297, + "learning_rate": 1.1297693312055709e-05, + "loss": 0.0604, + "step": 12097 + }, + { + "epoch": 1.311578490893322, + "grad_norm": 1.2041586637496948, + "learning_rate": 1.1296967938488323e-05, + "loss": 0.0092, + "step": 12098 + }, + { + "epoch": 1.3116869037294014, + "grad_norm": 0.8571547269821167, + "learning_rate": 1.1296242564920937e-05, + "loss": 0.0079, + "step": 12099 + }, + { + "epoch": 1.3117953165654814, + "grad_norm": 5.939694404602051, + "learning_rate": 1.1295517191353549e-05, + "loss": 0.0402, + "step": 12100 + }, + { + "epoch": 1.3119037294015612, + "grad_norm": 0.2655602693557739, + "learning_rate": 1.1294791817786161e-05, + "loss": 0.0029, + "step": 12101 + }, + { + "epoch": 1.312012142237641, + "grad_norm": 0.826551616191864, + "learning_rate": 1.1294066444218775e-05, + "loss": 0.0055, + "step": 12102 + }, + { + "epoch": 1.3121205550737207, + "grad_norm": 3.7356958389282227, + "learning_rate": 1.1293341070651387e-05, + "loss": 0.0183, + "step": 12103 + }, + { + "epoch": 1.3122289679098005, + "grad_norm": 0.8562637567520142, + "learning_rate": 1.1292615697084e-05, + "loss": 0.0088, + "step": 12104 + }, + { + "epoch": 1.3123373807458802, + "grad_norm": 1.9971942901611328, + "learning_rate": 1.1291890323516613e-05, + "loss": 0.0269, + "step": 12105 + }, + { + "epoch": 1.31244579358196, + "grad_norm": 1.5724941492080688, + "learning_rate": 1.1291164949949225e-05, + "loss": 0.0121, + "step": 12106 + }, + { + "epoch": 1.31255420641804, + "grad_norm": 5.285279750823975, + "learning_rate": 1.1290439576381837e-05, + "loss": 0.0514, + "step": 12107 + }, + { + "epoch": 1.3126626192541198, + "grad_norm": 3.3419573307037354, + "learning_rate": 1.128971420281445e-05, + "loss": 0.0318, + "step": 12108 + }, + { + "epoch": 1.3127710320901995, + "grad_norm": 3.1752209663391113, + "learning_rate": 1.1288988829247063e-05, + "loss": 0.0232, + "step": 12109 + }, + { + "epoch": 1.3128794449262793, + "grad_norm": 0.6240624189376831, + "learning_rate": 1.1288263455679675e-05, + "loss": 0.0049, + "step": 12110 + }, + { + "epoch": 1.312987857762359, + "grad_norm": 3.199247121810913, + "learning_rate": 1.1287538082112288e-05, + "loss": 0.0411, + "step": 12111 + }, + { + "epoch": 1.3130962705984388, + "grad_norm": 5.0129547119140625, + "learning_rate": 1.1286812708544901e-05, + "loss": 0.0427, + "step": 12112 + }, + { + "epoch": 1.3132046834345186, + "grad_norm": 0.6949417591094971, + "learning_rate": 1.1286087334977514e-05, + "loss": 0.0156, + "step": 12113 + }, + { + "epoch": 1.3133130962705986, + "grad_norm": 0.9101467728614807, + "learning_rate": 1.1285361961410126e-05, + "loss": 0.0047, + "step": 12114 + }, + { + "epoch": 1.313421509106678, + "grad_norm": 1.1331963539123535, + "learning_rate": 1.128463658784274e-05, + "loss": 0.0057, + "step": 12115 + }, + { + "epoch": 1.313529921942758, + "grad_norm": 2.625223398208618, + "learning_rate": 1.1283911214275353e-05, + "loss": 0.0188, + "step": 12116 + }, + { + "epoch": 1.3136383347788378, + "grad_norm": 4.553254127502441, + "learning_rate": 1.1283185840707967e-05, + "loss": 0.0312, + "step": 12117 + }, + { + "epoch": 1.3137467476149176, + "grad_norm": 0.4382019639015198, + "learning_rate": 1.128246046714058e-05, + "loss": 0.0036, + "step": 12118 + }, + { + "epoch": 1.3138551604509974, + "grad_norm": 1.1944385766983032, + "learning_rate": 1.1281735093573191e-05, + "loss": 0.0075, + "step": 12119 + }, + { + "epoch": 1.3139635732870771, + "grad_norm": 0.30266261100769043, + "learning_rate": 1.1281009720005804e-05, + "loss": 0.002, + "step": 12120 + }, + { + "epoch": 1.314071986123157, + "grad_norm": 0.7517775297164917, + "learning_rate": 1.1280284346438417e-05, + "loss": 0.0054, + "step": 12121 + }, + { + "epoch": 1.3141803989592367, + "grad_norm": 1.39436674118042, + "learning_rate": 1.127955897287103e-05, + "loss": 0.017, + "step": 12122 + }, + { + "epoch": 1.3142888117953166, + "grad_norm": 4.137460708618164, + "learning_rate": 1.1278833599303642e-05, + "loss": 0.0885, + "step": 12123 + }, + { + "epoch": 1.3143972246313964, + "grad_norm": 0.32125794887542725, + "learning_rate": 1.1278108225736256e-05, + "loss": 0.0013, + "step": 12124 + }, + { + "epoch": 1.3145056374674762, + "grad_norm": 6.752223968505859, + "learning_rate": 1.1277382852168868e-05, + "loss": 0.0548, + "step": 12125 + }, + { + "epoch": 1.314614050303556, + "grad_norm": 4.168037414550781, + "learning_rate": 1.127665747860148e-05, + "loss": 0.035, + "step": 12126 + }, + { + "epoch": 1.3147224631396357, + "grad_norm": 1.3898190259933472, + "learning_rate": 1.1275932105034094e-05, + "loss": 0.0129, + "step": 12127 + }, + { + "epoch": 1.3148308759757155, + "grad_norm": 2.9891810417175293, + "learning_rate": 1.1275206731466706e-05, + "loss": 0.032, + "step": 12128 + }, + { + "epoch": 1.3149392888117952, + "grad_norm": 5.08235502243042, + "learning_rate": 1.1274481357899318e-05, + "loss": 0.0195, + "step": 12129 + }, + { + "epoch": 1.3150477016478752, + "grad_norm": 0.6921291351318359, + "learning_rate": 1.1273755984331932e-05, + "loss": 0.0041, + "step": 12130 + }, + { + "epoch": 1.315156114483955, + "grad_norm": 6.753121852874756, + "learning_rate": 1.1273030610764544e-05, + "loss": 0.0654, + "step": 12131 + }, + { + "epoch": 1.3152645273200347, + "grad_norm": 0.38577818870544434, + "learning_rate": 1.1272305237197156e-05, + "loss": 0.0032, + "step": 12132 + }, + { + "epoch": 1.3153729401561145, + "grad_norm": 2.549999475479126, + "learning_rate": 1.1271579863629772e-05, + "loss": 0.0611, + "step": 12133 + }, + { + "epoch": 1.3154813529921943, + "grad_norm": 1.0437268018722534, + "learning_rate": 1.1270854490062384e-05, + "loss": 0.006, + "step": 12134 + }, + { + "epoch": 1.315589765828274, + "grad_norm": 1.844895362854004, + "learning_rate": 1.1270129116494996e-05, + "loss": 0.0043, + "step": 12135 + }, + { + "epoch": 1.3156981786643538, + "grad_norm": 3.1926069259643555, + "learning_rate": 1.126940374292761e-05, + "loss": 0.0186, + "step": 12136 + }, + { + "epoch": 1.3158065915004338, + "grad_norm": 1.6764609813690186, + "learning_rate": 1.1268678369360222e-05, + "loss": 0.0061, + "step": 12137 + }, + { + "epoch": 1.3159150043365133, + "grad_norm": 1.2194044589996338, + "learning_rate": 1.1267952995792834e-05, + "loss": 0.0077, + "step": 12138 + }, + { + "epoch": 1.3160234171725933, + "grad_norm": 1.256633996963501, + "learning_rate": 1.1267227622225448e-05, + "loss": 0.0341, + "step": 12139 + }, + { + "epoch": 1.316131830008673, + "grad_norm": 1.8580896854400635, + "learning_rate": 1.126650224865806e-05, + "loss": 0.0179, + "step": 12140 + }, + { + "epoch": 1.3162402428447528, + "grad_norm": 3.0444371700286865, + "learning_rate": 1.1265776875090672e-05, + "loss": 0.0253, + "step": 12141 + }, + { + "epoch": 1.3163486556808326, + "grad_norm": 4.864675998687744, + "learning_rate": 1.1265051501523286e-05, + "loss": 0.0255, + "step": 12142 + }, + { + "epoch": 1.3164570685169124, + "grad_norm": 3.4637393951416016, + "learning_rate": 1.1264326127955898e-05, + "loss": 0.0589, + "step": 12143 + }, + { + "epoch": 1.3165654813529921, + "grad_norm": 0.849449634552002, + "learning_rate": 1.126360075438851e-05, + "loss": 0.0096, + "step": 12144 + }, + { + "epoch": 1.3166738941890719, + "grad_norm": 5.620277404785156, + "learning_rate": 1.1262875380821124e-05, + "loss": 0.0187, + "step": 12145 + }, + { + "epoch": 1.3167823070251519, + "grad_norm": 1.917995572090149, + "learning_rate": 1.1262150007253736e-05, + "loss": 0.0286, + "step": 12146 + }, + { + "epoch": 1.3168907198612316, + "grad_norm": 1.9653040170669556, + "learning_rate": 1.1261424633686348e-05, + "loss": 0.0269, + "step": 12147 + }, + { + "epoch": 1.3169991326973114, + "grad_norm": 1.315578579902649, + "learning_rate": 1.126069926011896e-05, + "loss": 0.0052, + "step": 12148 + }, + { + "epoch": 1.3171075455333912, + "grad_norm": 0.4783739745616913, + "learning_rate": 1.1259973886551574e-05, + "loss": 0.0034, + "step": 12149 + }, + { + "epoch": 1.317215958369471, + "grad_norm": 0.4038616716861725, + "learning_rate": 1.1259248512984188e-05, + "loss": 0.0037, + "step": 12150 + }, + { + "epoch": 1.3173243712055507, + "grad_norm": 0.5686510801315308, + "learning_rate": 1.1258523139416802e-05, + "loss": 0.0073, + "step": 12151 + }, + { + "epoch": 1.3174327840416304, + "grad_norm": 1.7238919734954834, + "learning_rate": 1.1257797765849414e-05, + "loss": 0.02, + "step": 12152 + }, + { + "epoch": 1.3175411968777104, + "grad_norm": 2.2348544597625732, + "learning_rate": 1.1257072392282026e-05, + "loss": 0.0085, + "step": 12153 + }, + { + "epoch": 1.3176496097137902, + "grad_norm": 1.5306317806243896, + "learning_rate": 1.125634701871464e-05, + "loss": 0.0237, + "step": 12154 + }, + { + "epoch": 1.31775802254987, + "grad_norm": 1.640206217765808, + "learning_rate": 1.1255621645147252e-05, + "loss": 0.0041, + "step": 12155 + }, + { + "epoch": 1.3178664353859497, + "grad_norm": 2.0933785438537598, + "learning_rate": 1.1254896271579865e-05, + "loss": 0.0166, + "step": 12156 + }, + { + "epoch": 1.3179748482220295, + "grad_norm": 4.048447132110596, + "learning_rate": 1.1254170898012478e-05, + "loss": 0.0414, + "step": 12157 + }, + { + "epoch": 1.3180832610581092, + "grad_norm": 0.8125308752059937, + "learning_rate": 1.125344552444509e-05, + "loss": 0.0144, + "step": 12158 + }, + { + "epoch": 1.318191673894189, + "grad_norm": 3.5950000286102295, + "learning_rate": 1.1252720150877703e-05, + "loss": 0.0145, + "step": 12159 + }, + { + "epoch": 1.318300086730269, + "grad_norm": 1.432621717453003, + "learning_rate": 1.1251994777310315e-05, + "loss": 0.0125, + "step": 12160 + }, + { + "epoch": 1.3184084995663485, + "grad_norm": 1.8391727209091187, + "learning_rate": 1.1251269403742929e-05, + "loss": 0.0204, + "step": 12161 + }, + { + "epoch": 1.3185169124024285, + "grad_norm": 2.8993446826934814, + "learning_rate": 1.125054403017554e-05, + "loss": 0.0127, + "step": 12162 + }, + { + "epoch": 1.3186253252385083, + "grad_norm": 1.4156218767166138, + "learning_rate": 1.1249818656608153e-05, + "loss": 0.0123, + "step": 12163 + }, + { + "epoch": 1.318733738074588, + "grad_norm": 3.2948591709136963, + "learning_rate": 1.1249093283040767e-05, + "loss": 0.0094, + "step": 12164 + }, + { + "epoch": 1.3188421509106678, + "grad_norm": 0.8099051117897034, + "learning_rate": 1.1248367909473379e-05, + "loss": 0.0029, + "step": 12165 + }, + { + "epoch": 1.3189505637467476, + "grad_norm": 7.2870378494262695, + "learning_rate": 1.1247642535905991e-05, + "loss": 0.0122, + "step": 12166 + }, + { + "epoch": 1.3190589765828273, + "grad_norm": 2.312861204147339, + "learning_rate": 1.1246917162338605e-05, + "loss": 0.025, + "step": 12167 + }, + { + "epoch": 1.319167389418907, + "grad_norm": 2.136519432067871, + "learning_rate": 1.1246191788771219e-05, + "loss": 0.0303, + "step": 12168 + }, + { + "epoch": 1.319275802254987, + "grad_norm": 4.6239705085754395, + "learning_rate": 1.1245466415203833e-05, + "loss": 0.0235, + "step": 12169 + }, + { + "epoch": 1.3193842150910668, + "grad_norm": 2.62667179107666, + "learning_rate": 1.1244741041636445e-05, + "loss": 0.0201, + "step": 12170 + }, + { + "epoch": 1.3194926279271466, + "grad_norm": 3.0896146297454834, + "learning_rate": 1.1244015668069057e-05, + "loss": 0.0111, + "step": 12171 + }, + { + "epoch": 1.3196010407632264, + "grad_norm": 2.4141435623168945, + "learning_rate": 1.1243290294501669e-05, + "loss": 0.0785, + "step": 12172 + }, + { + "epoch": 1.3197094535993061, + "grad_norm": 4.086180686950684, + "learning_rate": 1.1242564920934283e-05, + "loss": 0.0362, + "step": 12173 + }, + { + "epoch": 1.319817866435386, + "grad_norm": 0.17740853130817413, + "learning_rate": 1.1241839547366895e-05, + "loss": 0.0018, + "step": 12174 + }, + { + "epoch": 1.3199262792714657, + "grad_norm": 3.038224935531616, + "learning_rate": 1.1241114173799507e-05, + "loss": 0.0157, + "step": 12175 + }, + { + "epoch": 1.3200346921075456, + "grad_norm": 3.294450044631958, + "learning_rate": 1.1240388800232121e-05, + "loss": 0.0434, + "step": 12176 + }, + { + "epoch": 1.3201431049436252, + "grad_norm": 0.7408373951911926, + "learning_rate": 1.1239663426664733e-05, + "loss": 0.0072, + "step": 12177 + }, + { + "epoch": 1.3202515177797052, + "grad_norm": 5.77374792098999, + "learning_rate": 1.1238938053097345e-05, + "loss": 0.0068, + "step": 12178 + }, + { + "epoch": 1.320359930615785, + "grad_norm": 1.1567790508270264, + "learning_rate": 1.1238212679529959e-05, + "loss": 0.0126, + "step": 12179 + }, + { + "epoch": 1.3204683434518647, + "grad_norm": 2.407824993133545, + "learning_rate": 1.1237487305962571e-05, + "loss": 0.0188, + "step": 12180 + }, + { + "epoch": 1.3205767562879445, + "grad_norm": 0.72088623046875, + "learning_rate": 1.1236761932395183e-05, + "loss": 0.0086, + "step": 12181 + }, + { + "epoch": 1.3206851691240242, + "grad_norm": 2.656100034713745, + "learning_rate": 1.1236036558827797e-05, + "loss": 0.0351, + "step": 12182 + }, + { + "epoch": 1.3207935819601042, + "grad_norm": 4.948014259338379, + "learning_rate": 1.123531118526041e-05, + "loss": 0.0669, + "step": 12183 + }, + { + "epoch": 1.3209019947961838, + "grad_norm": 2.006694793701172, + "learning_rate": 1.1234585811693021e-05, + "loss": 0.0224, + "step": 12184 + }, + { + "epoch": 1.3210104076322637, + "grad_norm": 1.0286146402359009, + "learning_rate": 1.1233860438125637e-05, + "loss": 0.0278, + "step": 12185 + }, + { + "epoch": 1.3211188204683435, + "grad_norm": 3.5262246131896973, + "learning_rate": 1.1233135064558249e-05, + "loss": 0.0548, + "step": 12186 + }, + { + "epoch": 1.3212272333044233, + "grad_norm": 0.9151023626327515, + "learning_rate": 1.1232409690990861e-05, + "loss": 0.0149, + "step": 12187 + }, + { + "epoch": 1.321335646140503, + "grad_norm": 1.647917628288269, + "learning_rate": 1.1231684317423475e-05, + "loss": 0.0075, + "step": 12188 + }, + { + "epoch": 1.3214440589765828, + "grad_norm": 4.834316730499268, + "learning_rate": 1.1230958943856087e-05, + "loss": 0.0262, + "step": 12189 + }, + { + "epoch": 1.3215524718126626, + "grad_norm": 2.5862927436828613, + "learning_rate": 1.12302335702887e-05, + "loss": 0.0164, + "step": 12190 + }, + { + "epoch": 1.3216608846487423, + "grad_norm": 4.188961029052734, + "learning_rate": 1.1229508196721313e-05, + "loss": 0.0156, + "step": 12191 + }, + { + "epoch": 1.3217692974848223, + "grad_norm": 5.266927242279053, + "learning_rate": 1.1228782823153925e-05, + "loss": 0.0242, + "step": 12192 + }, + { + "epoch": 1.321877710320902, + "grad_norm": 1.3568075895309448, + "learning_rate": 1.1228057449586538e-05, + "loss": 0.0087, + "step": 12193 + }, + { + "epoch": 1.3219861231569818, + "grad_norm": 1.9512139558792114, + "learning_rate": 1.1227332076019151e-05, + "loss": 0.0116, + "step": 12194 + }, + { + "epoch": 1.3220945359930616, + "grad_norm": 3.361417531967163, + "learning_rate": 1.1226606702451764e-05, + "loss": 0.0084, + "step": 12195 + }, + { + "epoch": 1.3222029488291414, + "grad_norm": 1.5970045328140259, + "learning_rate": 1.1225881328884376e-05, + "loss": 0.0113, + "step": 12196 + }, + { + "epoch": 1.3223113616652211, + "grad_norm": 3.6414005756378174, + "learning_rate": 1.122515595531699e-05, + "loss": 0.0571, + "step": 12197 + }, + { + "epoch": 1.3224197745013009, + "grad_norm": 1.6675583124160767, + "learning_rate": 1.1224430581749602e-05, + "loss": 0.0415, + "step": 12198 + }, + { + "epoch": 1.3225281873373809, + "grad_norm": 5.435262680053711, + "learning_rate": 1.1223705208182214e-05, + "loss": 0.0245, + "step": 12199 + }, + { + "epoch": 1.3226366001734604, + "grad_norm": 2.0606586933135986, + "learning_rate": 1.1222979834614826e-05, + "loss": 0.0306, + "step": 12200 + }, + { + "epoch": 1.3227450130095404, + "grad_norm": 1.2856651544570923, + "learning_rate": 1.122225446104744e-05, + "loss": 0.0133, + "step": 12201 + }, + { + "epoch": 1.3228534258456202, + "grad_norm": 2.3349130153656006, + "learning_rate": 1.1221529087480054e-05, + "loss": 0.0161, + "step": 12202 + }, + { + "epoch": 1.3229618386817, + "grad_norm": 3.576054334640503, + "learning_rate": 1.1220803713912667e-05, + "loss": 0.0285, + "step": 12203 + }, + { + "epoch": 1.3230702515177797, + "grad_norm": 3.0102851390838623, + "learning_rate": 1.122007834034528e-05, + "loss": 0.0215, + "step": 12204 + }, + { + "epoch": 1.3231786643538594, + "grad_norm": 0.6733456254005432, + "learning_rate": 1.1219352966777892e-05, + "loss": 0.0089, + "step": 12205 + }, + { + "epoch": 1.3232870771899392, + "grad_norm": 1.7200850248336792, + "learning_rate": 1.1218627593210506e-05, + "loss": 0.0177, + "step": 12206 + }, + { + "epoch": 1.323395490026019, + "grad_norm": 1.8319121599197388, + "learning_rate": 1.1217902219643118e-05, + "loss": 0.0135, + "step": 12207 + }, + { + "epoch": 1.323503902862099, + "grad_norm": 0.7434324622154236, + "learning_rate": 1.121717684607573e-05, + "loss": 0.0111, + "step": 12208 + }, + { + "epoch": 1.3236123156981787, + "grad_norm": 3.0690810680389404, + "learning_rate": 1.1216451472508344e-05, + "loss": 0.0119, + "step": 12209 + }, + { + "epoch": 1.3237207285342585, + "grad_norm": 0.32323434948921204, + "learning_rate": 1.1215726098940956e-05, + "loss": 0.0041, + "step": 12210 + }, + { + "epoch": 1.3238291413703382, + "grad_norm": 4.580904960632324, + "learning_rate": 1.1215000725373568e-05, + "loss": 0.0169, + "step": 12211 + }, + { + "epoch": 1.323937554206418, + "grad_norm": 1.4006036520004272, + "learning_rate": 1.121427535180618e-05, + "loss": 0.0094, + "step": 12212 + }, + { + "epoch": 1.3240459670424978, + "grad_norm": 1.2164921760559082, + "learning_rate": 1.1213549978238794e-05, + "loss": 0.0078, + "step": 12213 + }, + { + "epoch": 1.3241543798785775, + "grad_norm": 2.9790046215057373, + "learning_rate": 1.1212824604671406e-05, + "loss": 0.0083, + "step": 12214 + }, + { + "epoch": 1.3242627927146575, + "grad_norm": 2.8787856101989746, + "learning_rate": 1.1212099231104018e-05, + "loss": 0.0581, + "step": 12215 + }, + { + "epoch": 1.3243712055507373, + "grad_norm": 4.053049087524414, + "learning_rate": 1.1211373857536632e-05, + "loss": 0.0323, + "step": 12216 + }, + { + "epoch": 1.324479618386817, + "grad_norm": 2.1989872455596924, + "learning_rate": 1.1210648483969244e-05, + "loss": 0.0127, + "step": 12217 + }, + { + "epoch": 1.3245880312228968, + "grad_norm": 1.1361021995544434, + "learning_rate": 1.1209923110401856e-05, + "loss": 0.0156, + "step": 12218 + }, + { + "epoch": 1.3246964440589766, + "grad_norm": 5.642844200134277, + "learning_rate": 1.1209197736834472e-05, + "loss": 0.0103, + "step": 12219 + }, + { + "epoch": 1.3248048568950563, + "grad_norm": 11.79979419708252, + "learning_rate": 1.1208472363267084e-05, + "loss": 0.0159, + "step": 12220 + }, + { + "epoch": 1.324913269731136, + "grad_norm": 0.10586083680391312, + "learning_rate": 1.1207746989699698e-05, + "loss": 0.0015, + "step": 12221 + }, + { + "epoch": 1.325021682567216, + "grad_norm": 0.5136100053787231, + "learning_rate": 1.120702161613231e-05, + "loss": 0.0038, + "step": 12222 + }, + { + "epoch": 1.3251300954032956, + "grad_norm": 0.7804862260818481, + "learning_rate": 1.1206296242564922e-05, + "loss": 0.0065, + "step": 12223 + }, + { + "epoch": 1.3252385082393756, + "grad_norm": 1.4579025506973267, + "learning_rate": 1.1205570868997534e-05, + "loss": 0.0178, + "step": 12224 + }, + { + "epoch": 1.3253469210754554, + "grad_norm": 1.4907097816467285, + "learning_rate": 1.1204845495430148e-05, + "loss": 0.0129, + "step": 12225 + }, + { + "epoch": 1.3254553339115351, + "grad_norm": 2.4541335105895996, + "learning_rate": 1.120412012186276e-05, + "loss": 0.0203, + "step": 12226 + }, + { + "epoch": 1.325563746747615, + "grad_norm": 7.148022651672363, + "learning_rate": 1.1203394748295372e-05, + "loss": 0.0189, + "step": 12227 + }, + { + "epoch": 1.3256721595836947, + "grad_norm": 0.6286354660987854, + "learning_rate": 1.1202669374727986e-05, + "loss": 0.0063, + "step": 12228 + }, + { + "epoch": 1.3257805724197744, + "grad_norm": 3.3685462474823, + "learning_rate": 1.1201944001160598e-05, + "loss": 0.0416, + "step": 12229 + }, + { + "epoch": 1.3258889852558542, + "grad_norm": 3.443103790283203, + "learning_rate": 1.120121862759321e-05, + "loss": 0.0324, + "step": 12230 + }, + { + "epoch": 1.3259973980919342, + "grad_norm": 2.3978991508483887, + "learning_rate": 1.1200493254025824e-05, + "loss": 0.0295, + "step": 12231 + }, + { + "epoch": 1.326105810928014, + "grad_norm": 4.0297441482543945, + "learning_rate": 1.1199767880458437e-05, + "loss": 0.0153, + "step": 12232 + }, + { + "epoch": 1.3262142237640937, + "grad_norm": 0.05243903398513794, + "learning_rate": 1.1199042506891049e-05, + "loss": 0.0007, + "step": 12233 + }, + { + "epoch": 1.3263226366001735, + "grad_norm": 0.4056730270385742, + "learning_rate": 1.1198317133323663e-05, + "loss": 0.0033, + "step": 12234 + }, + { + "epoch": 1.3264310494362532, + "grad_norm": 0.3570799231529236, + "learning_rate": 1.1197591759756275e-05, + "loss": 0.0022, + "step": 12235 + }, + { + "epoch": 1.326539462272333, + "grad_norm": 1.6773253679275513, + "learning_rate": 1.1196866386188887e-05, + "loss": 0.037, + "step": 12236 + }, + { + "epoch": 1.3266478751084128, + "grad_norm": 2.130418539047241, + "learning_rate": 1.1196141012621502e-05, + "loss": 0.0102, + "step": 12237 + }, + { + "epoch": 1.3267562879444927, + "grad_norm": 1.3959503173828125, + "learning_rate": 1.1195415639054114e-05, + "loss": 0.0279, + "step": 12238 + }, + { + "epoch": 1.3268647007805723, + "grad_norm": 5.417057037353516, + "learning_rate": 1.1194690265486727e-05, + "loss": 0.0078, + "step": 12239 + }, + { + "epoch": 1.3269731136166523, + "grad_norm": 13.251835823059082, + "learning_rate": 1.119396489191934e-05, + "loss": 0.0368, + "step": 12240 + }, + { + "epoch": 1.327081526452732, + "grad_norm": 4.781350135803223, + "learning_rate": 1.1193239518351953e-05, + "loss": 0.0112, + "step": 12241 + }, + { + "epoch": 1.3271899392888118, + "grad_norm": 4.82141637802124, + "learning_rate": 1.1192514144784565e-05, + "loss": 0.0233, + "step": 12242 + }, + { + "epoch": 1.3272983521248916, + "grad_norm": 1.9921824932098389, + "learning_rate": 1.1191788771217179e-05, + "loss": 0.0313, + "step": 12243 + }, + { + "epoch": 1.3274067649609713, + "grad_norm": 0.33174028992652893, + "learning_rate": 1.119106339764979e-05, + "loss": 0.0034, + "step": 12244 + }, + { + "epoch": 1.3275151777970513, + "grad_norm": 2.3581626415252686, + "learning_rate": 1.1190338024082403e-05, + "loss": 0.0576, + "step": 12245 + }, + { + "epoch": 1.3276235906331308, + "grad_norm": 1.4943889379501343, + "learning_rate": 1.1189612650515017e-05, + "loss": 0.0106, + "step": 12246 + }, + { + "epoch": 1.3277320034692108, + "grad_norm": 1.5284377336502075, + "learning_rate": 1.1188887276947629e-05, + "loss": 0.0325, + "step": 12247 + }, + { + "epoch": 1.3278404163052906, + "grad_norm": 1.7745237350463867, + "learning_rate": 1.1188161903380241e-05, + "loss": 0.0419, + "step": 12248 + }, + { + "epoch": 1.3279488291413704, + "grad_norm": 0.46844446659088135, + "learning_rate": 1.1187436529812855e-05, + "loss": 0.0024, + "step": 12249 + }, + { + "epoch": 1.3280572419774501, + "grad_norm": 0.9233652353286743, + "learning_rate": 1.1186711156245467e-05, + "loss": 0.0057, + "step": 12250 + }, + { + "epoch": 1.3281656548135299, + "grad_norm": 0.7699439525604248, + "learning_rate": 1.1185985782678079e-05, + "loss": 0.0531, + "step": 12251 + }, + { + "epoch": 1.3282740676496096, + "grad_norm": 4.420161724090576, + "learning_rate": 1.1185260409110691e-05, + "loss": 0.0673, + "step": 12252 + }, + { + "epoch": 1.3283824804856894, + "grad_norm": 5.281742095947266, + "learning_rate": 1.1184535035543305e-05, + "loss": 0.0428, + "step": 12253 + }, + { + "epoch": 1.3284908933217694, + "grad_norm": 2.3947253227233887, + "learning_rate": 1.1183809661975919e-05, + "loss": 0.009, + "step": 12254 + }, + { + "epoch": 1.3285993061578492, + "grad_norm": 5.315145015716553, + "learning_rate": 1.1183084288408533e-05, + "loss": 0.0874, + "step": 12255 + }, + { + "epoch": 1.328707718993929, + "grad_norm": 0.27057981491088867, + "learning_rate": 1.1182358914841145e-05, + "loss": 0.0022, + "step": 12256 + }, + { + "epoch": 1.3288161318300087, + "grad_norm": 0.09066613018512726, + "learning_rate": 1.1181633541273757e-05, + "loss": 0.0009, + "step": 12257 + }, + { + "epoch": 1.3289245446660884, + "grad_norm": 0.6580047011375427, + "learning_rate": 1.1180908167706371e-05, + "loss": 0.0038, + "step": 12258 + }, + { + "epoch": 1.3290329575021682, + "grad_norm": 0.19397331774234772, + "learning_rate": 1.1180182794138983e-05, + "loss": 0.002, + "step": 12259 + }, + { + "epoch": 1.329141370338248, + "grad_norm": 0.2864179015159607, + "learning_rate": 1.1179457420571595e-05, + "loss": 0.0027, + "step": 12260 + }, + { + "epoch": 1.329249783174328, + "grad_norm": 0.4476136863231659, + "learning_rate": 1.1178732047004209e-05, + "loss": 0.0023, + "step": 12261 + }, + { + "epoch": 1.3293581960104075, + "grad_norm": 3.1945013999938965, + "learning_rate": 1.1178006673436821e-05, + "loss": 0.0187, + "step": 12262 + }, + { + "epoch": 1.3294666088464875, + "grad_norm": 2.823814630508423, + "learning_rate": 1.1177281299869433e-05, + "loss": 0.0393, + "step": 12263 + }, + { + "epoch": 1.3295750216825672, + "grad_norm": 1.372666835784912, + "learning_rate": 1.1176555926302045e-05, + "loss": 0.0092, + "step": 12264 + }, + { + "epoch": 1.329683434518647, + "grad_norm": 4.258316516876221, + "learning_rate": 1.117583055273466e-05, + "loss": 0.0175, + "step": 12265 + }, + { + "epoch": 1.3297918473547268, + "grad_norm": 0.886155903339386, + "learning_rate": 1.1175105179167271e-05, + "loss": 0.0317, + "step": 12266 + }, + { + "epoch": 1.3299002601908065, + "grad_norm": 1.980998158454895, + "learning_rate": 1.1174379805599884e-05, + "loss": 0.0349, + "step": 12267 + }, + { + "epoch": 1.3300086730268863, + "grad_norm": 0.07863664627075195, + "learning_rate": 1.1173654432032497e-05, + "loss": 0.0012, + "step": 12268 + }, + { + "epoch": 1.330117085862966, + "grad_norm": 0.7911503314971924, + "learning_rate": 1.117292905846511e-05, + "loss": 0.0032, + "step": 12269 + }, + { + "epoch": 1.330225498699046, + "grad_norm": 5.846598148345947, + "learning_rate": 1.1172203684897722e-05, + "loss": 0.0475, + "step": 12270 + }, + { + "epoch": 1.3303339115351258, + "grad_norm": 2.149035930633545, + "learning_rate": 1.1171478311330337e-05, + "loss": 0.0132, + "step": 12271 + }, + { + "epoch": 1.3304423243712056, + "grad_norm": 0.9686325192451477, + "learning_rate": 1.117075293776295e-05, + "loss": 0.0054, + "step": 12272 + }, + { + "epoch": 1.3305507372072853, + "grad_norm": 7.062028884887695, + "learning_rate": 1.1170027564195563e-05, + "loss": 0.0714, + "step": 12273 + }, + { + "epoch": 1.330659150043365, + "grad_norm": 1.0100138187408447, + "learning_rate": 1.1169302190628175e-05, + "loss": 0.004, + "step": 12274 + }, + { + "epoch": 1.3307675628794449, + "grad_norm": 4.528227806091309, + "learning_rate": 1.1168576817060787e-05, + "loss": 0.1067, + "step": 12275 + }, + { + "epoch": 1.3308759757155246, + "grad_norm": 0.3829570710659027, + "learning_rate": 1.11678514434934e-05, + "loss": 0.0033, + "step": 12276 + }, + { + "epoch": 1.3309843885516046, + "grad_norm": 1.3648681640625, + "learning_rate": 1.1167126069926013e-05, + "loss": 0.0101, + "step": 12277 + }, + { + "epoch": 1.3310928013876844, + "grad_norm": 2.4496824741363525, + "learning_rate": 1.1166400696358626e-05, + "loss": 0.0177, + "step": 12278 + }, + { + "epoch": 1.3312012142237641, + "grad_norm": 1.722772479057312, + "learning_rate": 1.1165675322791238e-05, + "loss": 0.0141, + "step": 12279 + }, + { + "epoch": 1.331309627059844, + "grad_norm": 0.4767642319202423, + "learning_rate": 1.1164949949223852e-05, + "loss": 0.0065, + "step": 12280 + }, + { + "epoch": 1.3314180398959237, + "grad_norm": 2.583038568496704, + "learning_rate": 1.1164224575656464e-05, + "loss": 0.0307, + "step": 12281 + }, + { + "epoch": 1.3315264527320034, + "grad_norm": 1.3929882049560547, + "learning_rate": 1.1163499202089076e-05, + "loss": 0.0247, + "step": 12282 + }, + { + "epoch": 1.3316348655680832, + "grad_norm": 1.7931514978408813, + "learning_rate": 1.116277382852169e-05, + "loss": 0.0309, + "step": 12283 + }, + { + "epoch": 1.3317432784041632, + "grad_norm": 9.379828453063965, + "learning_rate": 1.1162048454954302e-05, + "loss": 0.0455, + "step": 12284 + }, + { + "epoch": 1.3318516912402427, + "grad_norm": 1.8203027248382568, + "learning_rate": 1.1161323081386914e-05, + "loss": 0.0075, + "step": 12285 + }, + { + "epoch": 1.3319601040763227, + "grad_norm": 4.585367202758789, + "learning_rate": 1.1160597707819528e-05, + "loss": 0.0281, + "step": 12286 + }, + { + "epoch": 1.3320685169124025, + "grad_norm": 0.8782538771629333, + "learning_rate": 1.115987233425214e-05, + "loss": 0.009, + "step": 12287 + }, + { + "epoch": 1.3321769297484822, + "grad_norm": 4.464561939239502, + "learning_rate": 1.1159146960684754e-05, + "loss": 0.0336, + "step": 12288 + }, + { + "epoch": 1.332285342584562, + "grad_norm": 0.404926598072052, + "learning_rate": 1.1158421587117368e-05, + "loss": 0.005, + "step": 12289 + }, + { + "epoch": 1.3323937554206418, + "grad_norm": 1.8619835376739502, + "learning_rate": 1.115769621354998e-05, + "loss": 0.032, + "step": 12290 + }, + { + "epoch": 1.3325021682567215, + "grad_norm": 6.758624076843262, + "learning_rate": 1.1156970839982592e-05, + "loss": 0.0631, + "step": 12291 + }, + { + "epoch": 1.3326105810928013, + "grad_norm": 0.7059540152549744, + "learning_rate": 1.1156245466415206e-05, + "loss": 0.009, + "step": 12292 + }, + { + "epoch": 1.3327189939288813, + "grad_norm": 1.909334421157837, + "learning_rate": 1.1155520092847818e-05, + "loss": 0.0414, + "step": 12293 + }, + { + "epoch": 1.332827406764961, + "grad_norm": 0.6293389201164246, + "learning_rate": 1.115479471928043e-05, + "loss": 0.0096, + "step": 12294 + }, + { + "epoch": 1.3329358196010408, + "grad_norm": 3.163323163986206, + "learning_rate": 1.1154069345713044e-05, + "loss": 0.0031, + "step": 12295 + }, + { + "epoch": 1.3330442324371206, + "grad_norm": 1.255902647972107, + "learning_rate": 1.1153343972145656e-05, + "loss": 0.0563, + "step": 12296 + }, + { + "epoch": 1.3331526452732003, + "grad_norm": 2.4491922855377197, + "learning_rate": 1.1152618598578268e-05, + "loss": 0.0241, + "step": 12297 + }, + { + "epoch": 1.33326105810928, + "grad_norm": 2.073092222213745, + "learning_rate": 1.1151893225010882e-05, + "loss": 0.0041, + "step": 12298 + }, + { + "epoch": 1.3333694709453598, + "grad_norm": 0.7849981784820557, + "learning_rate": 1.1151167851443494e-05, + "loss": 0.0121, + "step": 12299 + }, + { + "epoch": 1.3334778837814398, + "grad_norm": 1.5389331579208374, + "learning_rate": 1.1150442477876106e-05, + "loss": 0.0156, + "step": 12300 + }, + { + "epoch": 1.3335862966175196, + "grad_norm": 1.7713302373886108, + "learning_rate": 1.114971710430872e-05, + "loss": 0.0301, + "step": 12301 + }, + { + "epoch": 1.3336947094535994, + "grad_norm": 2.8171064853668213, + "learning_rate": 1.1148991730741332e-05, + "loss": 0.0247, + "step": 12302 + }, + { + "epoch": 1.3338031222896791, + "grad_norm": 2.014286994934082, + "learning_rate": 1.1148266357173944e-05, + "loss": 0.0255, + "step": 12303 + }, + { + "epoch": 1.3339115351257589, + "grad_norm": 1.1433335542678833, + "learning_rate": 1.1147540983606557e-05, + "loss": 0.0152, + "step": 12304 + }, + { + "epoch": 1.3340199479618386, + "grad_norm": 2.2020657062530518, + "learning_rate": 1.114681561003917e-05, + "loss": 0.0193, + "step": 12305 + }, + { + "epoch": 1.3341283607979184, + "grad_norm": 0.5418851375579834, + "learning_rate": 1.1146090236471784e-05, + "loss": 0.0078, + "step": 12306 + }, + { + "epoch": 1.3342367736339984, + "grad_norm": 1.9406658411026, + "learning_rate": 1.1145364862904398e-05, + "loss": 0.0263, + "step": 12307 + }, + { + "epoch": 1.334345186470078, + "grad_norm": 5.240561008453369, + "learning_rate": 1.114463948933701e-05, + "loss": 0.0632, + "step": 12308 + }, + { + "epoch": 1.334453599306158, + "grad_norm": 1.8034640550613403, + "learning_rate": 1.1143914115769622e-05, + "loss": 0.0029, + "step": 12309 + }, + { + "epoch": 1.3345620121422377, + "grad_norm": 3.702529191970825, + "learning_rate": 1.1143188742202236e-05, + "loss": 0.0271, + "step": 12310 + }, + { + "epoch": 1.3346704249783174, + "grad_norm": 3.2114853858947754, + "learning_rate": 1.1142463368634848e-05, + "loss": 0.0132, + "step": 12311 + }, + { + "epoch": 1.3347788378143972, + "grad_norm": 1.4085044860839844, + "learning_rate": 1.114173799506746e-05, + "loss": 0.0429, + "step": 12312 + }, + { + "epoch": 1.334887250650477, + "grad_norm": 3.0100722312927246, + "learning_rate": 1.1141012621500074e-05, + "loss": 0.0216, + "step": 12313 + }, + { + "epoch": 1.3349956634865567, + "grad_norm": 3.522705078125, + "learning_rate": 1.1140287247932686e-05, + "loss": 0.0196, + "step": 12314 + }, + { + "epoch": 1.3351040763226365, + "grad_norm": 1.2406071424484253, + "learning_rate": 1.1139561874365299e-05, + "loss": 0.0083, + "step": 12315 + }, + { + "epoch": 1.3352124891587165, + "grad_norm": 6.966735363006592, + "learning_rate": 1.113883650079791e-05, + "loss": 0.0341, + "step": 12316 + }, + { + "epoch": 1.3353209019947962, + "grad_norm": 0.5415646433830261, + "learning_rate": 1.1138111127230525e-05, + "loss": 0.0161, + "step": 12317 + }, + { + "epoch": 1.335429314830876, + "grad_norm": 0.8731982707977295, + "learning_rate": 1.1137385753663137e-05, + "loss": 0.0089, + "step": 12318 + }, + { + "epoch": 1.3355377276669558, + "grad_norm": 0.4509146213531494, + "learning_rate": 1.1136660380095749e-05, + "loss": 0.0066, + "step": 12319 + }, + { + "epoch": 1.3356461405030355, + "grad_norm": 0.9037414193153381, + "learning_rate": 1.1135935006528363e-05, + "loss": 0.0099, + "step": 12320 + }, + { + "epoch": 1.3357545533391153, + "grad_norm": 0.8849184513092041, + "learning_rate": 1.1135209632960975e-05, + "loss": 0.0151, + "step": 12321 + }, + { + "epoch": 1.335862966175195, + "grad_norm": 1.5883822441101074, + "learning_rate": 1.1134484259393587e-05, + "loss": 0.0313, + "step": 12322 + }, + { + "epoch": 1.335971379011275, + "grad_norm": 1.2406729459762573, + "learning_rate": 1.1133758885826203e-05, + "loss": 0.0061, + "step": 12323 + }, + { + "epoch": 1.3360797918473546, + "grad_norm": 4.934607982635498, + "learning_rate": 1.1133033512258815e-05, + "loss": 0.0451, + "step": 12324 + }, + { + "epoch": 1.3361882046834346, + "grad_norm": 1.7178843021392822, + "learning_rate": 1.1132308138691429e-05, + "loss": 0.0271, + "step": 12325 + }, + { + "epoch": 1.3362966175195143, + "grad_norm": 4.110766410827637, + "learning_rate": 1.113158276512404e-05, + "loss": 0.0243, + "step": 12326 + }, + { + "epoch": 1.336405030355594, + "grad_norm": 0.20419059693813324, + "learning_rate": 1.1130857391556653e-05, + "loss": 0.0017, + "step": 12327 + }, + { + "epoch": 1.3365134431916739, + "grad_norm": 2.6273398399353027, + "learning_rate": 1.1130132017989265e-05, + "loss": 0.0338, + "step": 12328 + }, + { + "epoch": 1.3366218560277536, + "grad_norm": 1.396575927734375, + "learning_rate": 1.1129406644421879e-05, + "loss": 0.0146, + "step": 12329 + }, + { + "epoch": 1.3367302688638336, + "grad_norm": 1.3824621438980103, + "learning_rate": 1.1128681270854491e-05, + "loss": 0.0235, + "step": 12330 + }, + { + "epoch": 1.3368386816999132, + "grad_norm": 0.11904776841402054, + "learning_rate": 1.1127955897287103e-05, + "loss": 0.0017, + "step": 12331 + }, + { + "epoch": 1.3369470945359931, + "grad_norm": 3.058628559112549, + "learning_rate": 1.1127230523719717e-05, + "loss": 0.0669, + "step": 12332 + }, + { + "epoch": 1.337055507372073, + "grad_norm": 1.585895299911499, + "learning_rate": 1.1126505150152329e-05, + "loss": 0.036, + "step": 12333 + }, + { + "epoch": 1.3371639202081527, + "grad_norm": 1.2459465265274048, + "learning_rate": 1.1125779776584941e-05, + "loss": 0.0082, + "step": 12334 + }, + { + "epoch": 1.3372723330442324, + "grad_norm": 0.5899012088775635, + "learning_rate": 1.1125054403017555e-05, + "loss": 0.0077, + "step": 12335 + }, + { + "epoch": 1.3373807458803122, + "grad_norm": 2.9055840969085693, + "learning_rate": 1.1124329029450167e-05, + "loss": 0.0391, + "step": 12336 + }, + { + "epoch": 1.337489158716392, + "grad_norm": 0.2021481692790985, + "learning_rate": 1.112360365588278e-05, + "loss": 0.0024, + "step": 12337 + }, + { + "epoch": 1.3375975715524717, + "grad_norm": 2.434872627258301, + "learning_rate": 1.1122878282315393e-05, + "loss": 0.0176, + "step": 12338 + }, + { + "epoch": 1.3377059843885517, + "grad_norm": 1.2865101099014282, + "learning_rate": 1.1122152908748005e-05, + "loss": 0.0338, + "step": 12339 + }, + { + "epoch": 1.3378143972246315, + "grad_norm": 3.6415038108825684, + "learning_rate": 1.1121427535180621e-05, + "loss": 0.0503, + "step": 12340 + }, + { + "epoch": 1.3379228100607112, + "grad_norm": 2.939521312713623, + "learning_rate": 1.1120702161613233e-05, + "loss": 0.047, + "step": 12341 + }, + { + "epoch": 1.338031222896791, + "grad_norm": 0.4959146976470947, + "learning_rate": 1.1119976788045845e-05, + "loss": 0.002, + "step": 12342 + }, + { + "epoch": 1.3381396357328708, + "grad_norm": 1.7296290397644043, + "learning_rate": 1.1119251414478457e-05, + "loss": 0.0231, + "step": 12343 + }, + { + "epoch": 1.3382480485689505, + "grad_norm": 0.8590090870857239, + "learning_rate": 1.1118526040911071e-05, + "loss": 0.0077, + "step": 12344 + }, + { + "epoch": 1.3383564614050303, + "grad_norm": 2.2305314540863037, + "learning_rate": 1.1117800667343683e-05, + "loss": 0.0207, + "step": 12345 + }, + { + "epoch": 1.3384648742411103, + "grad_norm": 1.7988662719726562, + "learning_rate": 1.1117075293776295e-05, + "loss": 0.0172, + "step": 12346 + }, + { + "epoch": 1.3385732870771898, + "grad_norm": 0.7179656028747559, + "learning_rate": 1.111634992020891e-05, + "loss": 0.0046, + "step": 12347 + }, + { + "epoch": 1.3386816999132698, + "grad_norm": 5.362243175506592, + "learning_rate": 1.1115624546641521e-05, + "loss": 0.0204, + "step": 12348 + }, + { + "epoch": 1.3387901127493496, + "grad_norm": 0.9371793866157532, + "learning_rate": 1.1114899173074134e-05, + "loss": 0.012, + "step": 12349 + }, + { + "epoch": 1.3388985255854293, + "grad_norm": 4.439329624176025, + "learning_rate": 1.1114173799506747e-05, + "loss": 0.0282, + "step": 12350 + }, + { + "epoch": 1.339006938421509, + "grad_norm": 5.557624816894531, + "learning_rate": 1.111344842593936e-05, + "loss": 0.0198, + "step": 12351 + }, + { + "epoch": 1.3391153512575888, + "grad_norm": 2.777803421020508, + "learning_rate": 1.1112723052371972e-05, + "loss": 0.0495, + "step": 12352 + }, + { + "epoch": 1.3392237640936686, + "grad_norm": 0.47729477286338806, + "learning_rate": 1.1111997678804586e-05, + "loss": 0.0063, + "step": 12353 + }, + { + "epoch": 1.3393321769297484, + "grad_norm": 2.035599708557129, + "learning_rate": 1.1111272305237198e-05, + "loss": 0.0225, + "step": 12354 + }, + { + "epoch": 1.3394405897658284, + "grad_norm": 0.9841970205307007, + "learning_rate": 1.111054693166981e-05, + "loss": 0.0182, + "step": 12355 + }, + { + "epoch": 1.3395490026019081, + "grad_norm": 3.351457118988037, + "learning_rate": 1.1109821558102422e-05, + "loss": 0.0221, + "step": 12356 + }, + { + "epoch": 1.3396574154379879, + "grad_norm": 2.887439250946045, + "learning_rate": 1.1109096184535036e-05, + "loss": 0.0276, + "step": 12357 + }, + { + "epoch": 1.3397658282740676, + "grad_norm": 7.8270955085754395, + "learning_rate": 1.110837081096765e-05, + "loss": 0.042, + "step": 12358 + }, + { + "epoch": 1.3398742411101474, + "grad_norm": 2.431936264038086, + "learning_rate": 1.1107645437400263e-05, + "loss": 0.0068, + "step": 12359 + }, + { + "epoch": 1.3399826539462272, + "grad_norm": 6.90570068359375, + "learning_rate": 1.1106920063832876e-05, + "loss": 0.0323, + "step": 12360 + }, + { + "epoch": 1.340091066782307, + "grad_norm": 0.446132630109787, + "learning_rate": 1.1106194690265488e-05, + "loss": 0.0063, + "step": 12361 + }, + { + "epoch": 1.340199479618387, + "grad_norm": 0.5111770629882812, + "learning_rate": 1.1105469316698102e-05, + "loss": 0.0059, + "step": 12362 + }, + { + "epoch": 1.3403078924544667, + "grad_norm": 1.3903884887695312, + "learning_rate": 1.1104743943130714e-05, + "loss": 0.0201, + "step": 12363 + }, + { + "epoch": 1.3404163052905465, + "grad_norm": 2.564476728439331, + "learning_rate": 1.1104018569563326e-05, + "loss": 0.0155, + "step": 12364 + }, + { + "epoch": 1.3405247181266262, + "grad_norm": 0.3078902065753937, + "learning_rate": 1.110329319599594e-05, + "loss": 0.0025, + "step": 12365 + }, + { + "epoch": 1.340633130962706, + "grad_norm": 0.3653162717819214, + "learning_rate": 1.1102567822428552e-05, + "loss": 0.0038, + "step": 12366 + }, + { + "epoch": 1.3407415437987857, + "grad_norm": 3.3840978145599365, + "learning_rate": 1.1101842448861164e-05, + "loss": 0.0216, + "step": 12367 + }, + { + "epoch": 1.3408499566348655, + "grad_norm": 3.193263292312622, + "learning_rate": 1.1101117075293776e-05, + "loss": 0.011, + "step": 12368 + }, + { + "epoch": 1.3409583694709455, + "grad_norm": 0.4972400367259979, + "learning_rate": 1.110039170172639e-05, + "loss": 0.0023, + "step": 12369 + }, + { + "epoch": 1.341066782307025, + "grad_norm": 2.411888599395752, + "learning_rate": 1.1099666328159002e-05, + "loss": 0.0174, + "step": 12370 + }, + { + "epoch": 1.341175195143105, + "grad_norm": 3.488008737564087, + "learning_rate": 1.1098940954591614e-05, + "loss": 0.0107, + "step": 12371 + }, + { + "epoch": 1.3412836079791848, + "grad_norm": 1.9955966472625732, + "learning_rate": 1.1098215581024228e-05, + "loss": 0.0156, + "step": 12372 + }, + { + "epoch": 1.3413920208152645, + "grad_norm": 1.335044503211975, + "learning_rate": 1.109749020745684e-05, + "loss": 0.015, + "step": 12373 + }, + { + "epoch": 1.3415004336513443, + "grad_norm": 1.6809965372085571, + "learning_rate": 1.1096764833889452e-05, + "loss": 0.009, + "step": 12374 + }, + { + "epoch": 1.341608846487424, + "grad_norm": 2.775258779525757, + "learning_rate": 1.1096039460322068e-05, + "loss": 0.0317, + "step": 12375 + }, + { + "epoch": 1.3417172593235038, + "grad_norm": 0.5690003037452698, + "learning_rate": 1.109531408675468e-05, + "loss": 0.0086, + "step": 12376 + }, + { + "epoch": 1.3418256721595836, + "grad_norm": 4.369827747344971, + "learning_rate": 1.1094588713187294e-05, + "loss": 0.0515, + "step": 12377 + }, + { + "epoch": 1.3419340849956636, + "grad_norm": 0.664008617401123, + "learning_rate": 1.1093863339619906e-05, + "loss": 0.005, + "step": 12378 + }, + { + "epoch": 1.3420424978317433, + "grad_norm": 1.742969274520874, + "learning_rate": 1.1093137966052518e-05, + "loss": 0.0323, + "step": 12379 + }, + { + "epoch": 1.342150910667823, + "grad_norm": 5.741184711456299, + "learning_rate": 1.109241259248513e-05, + "loss": 0.0347, + "step": 12380 + }, + { + "epoch": 1.3422593235039029, + "grad_norm": 13.93034839630127, + "learning_rate": 1.1091687218917744e-05, + "loss": 0.0451, + "step": 12381 + }, + { + "epoch": 1.3423677363399826, + "grad_norm": 3.9460151195526123, + "learning_rate": 1.1090961845350356e-05, + "loss": 0.0236, + "step": 12382 + }, + { + "epoch": 1.3424761491760624, + "grad_norm": 0.9567580223083496, + "learning_rate": 1.1090236471782968e-05, + "loss": 0.0069, + "step": 12383 + }, + { + "epoch": 1.3425845620121422, + "grad_norm": 3.6838974952697754, + "learning_rate": 1.1089511098215582e-05, + "loss": 0.008, + "step": 12384 + }, + { + "epoch": 1.3426929748482221, + "grad_norm": 1.6901443004608154, + "learning_rate": 1.1088785724648194e-05, + "loss": 0.0158, + "step": 12385 + }, + { + "epoch": 1.3428013876843017, + "grad_norm": 0.0358455553650856, + "learning_rate": 1.1088060351080807e-05, + "loss": 0.0003, + "step": 12386 + }, + { + "epoch": 1.3429098005203817, + "grad_norm": 1.0022830963134766, + "learning_rate": 1.108733497751342e-05, + "loss": 0.0073, + "step": 12387 + }, + { + "epoch": 1.3430182133564614, + "grad_norm": 6.77326774597168, + "learning_rate": 1.1086609603946033e-05, + "loss": 0.0222, + "step": 12388 + }, + { + "epoch": 1.3431266261925412, + "grad_norm": 3.922053337097168, + "learning_rate": 1.1085884230378645e-05, + "loss": 0.0433, + "step": 12389 + }, + { + "epoch": 1.343235039028621, + "grad_norm": 2.705855131149292, + "learning_rate": 1.1085158856811259e-05, + "loss": 0.0351, + "step": 12390 + }, + { + "epoch": 1.3433434518647007, + "grad_norm": 2.2176308631896973, + "learning_rate": 1.108443348324387e-05, + "loss": 0.0144, + "step": 12391 + }, + { + "epoch": 1.3434518647007807, + "grad_norm": 10.41695785522461, + "learning_rate": 1.1083708109676486e-05, + "loss": 0.0362, + "step": 12392 + }, + { + "epoch": 1.3435602775368602, + "grad_norm": 5.010303020477295, + "learning_rate": 1.1082982736109098e-05, + "loss": 0.007, + "step": 12393 + }, + { + "epoch": 1.3436686903729402, + "grad_norm": 0.5949323773384094, + "learning_rate": 1.108225736254171e-05, + "loss": 0.0032, + "step": 12394 + }, + { + "epoch": 1.34377710320902, + "grad_norm": 0.3981853425502777, + "learning_rate": 1.1081531988974323e-05, + "loss": 0.0062, + "step": 12395 + }, + { + "epoch": 1.3438855160450998, + "grad_norm": 7.584989070892334, + "learning_rate": 1.1080806615406936e-05, + "loss": 0.0246, + "step": 12396 + }, + { + "epoch": 1.3439939288811795, + "grad_norm": 0.2204221785068512, + "learning_rate": 1.1080081241839549e-05, + "loss": 0.0023, + "step": 12397 + }, + { + "epoch": 1.3441023417172593, + "grad_norm": 4.283815383911133, + "learning_rate": 1.107935586827216e-05, + "loss": 0.0155, + "step": 12398 + }, + { + "epoch": 1.344210754553339, + "grad_norm": 4.93024206161499, + "learning_rate": 1.1078630494704775e-05, + "loss": 0.0253, + "step": 12399 + }, + { + "epoch": 1.3443191673894188, + "grad_norm": 1.030413031578064, + "learning_rate": 1.1077905121137387e-05, + "loss": 0.0185, + "step": 12400 + }, + { + "epoch": 1.3444275802254988, + "grad_norm": 3.8264853954315186, + "learning_rate": 1.1077179747569999e-05, + "loss": 0.0537, + "step": 12401 + }, + { + "epoch": 1.3445359930615786, + "grad_norm": 1.5697228908538818, + "learning_rate": 1.1076454374002613e-05, + "loss": 0.0252, + "step": 12402 + }, + { + "epoch": 1.3446444058976583, + "grad_norm": 3.594534397125244, + "learning_rate": 1.1075729000435225e-05, + "loss": 0.0841, + "step": 12403 + }, + { + "epoch": 1.344752818733738, + "grad_norm": 0.7100617289543152, + "learning_rate": 1.1075003626867837e-05, + "loss": 0.0057, + "step": 12404 + }, + { + "epoch": 1.3448612315698178, + "grad_norm": 1.1111048460006714, + "learning_rate": 1.107427825330045e-05, + "loss": 0.0104, + "step": 12405 + }, + { + "epoch": 1.3449696444058976, + "grad_norm": 1.4627156257629395, + "learning_rate": 1.1073552879733063e-05, + "loss": 0.0051, + "step": 12406 + }, + { + "epoch": 1.3450780572419774, + "grad_norm": 3.574563980102539, + "learning_rate": 1.1072827506165675e-05, + "loss": 0.0144, + "step": 12407 + }, + { + "epoch": 1.3451864700780574, + "grad_norm": 2.828780174255371, + "learning_rate": 1.1072102132598287e-05, + "loss": 0.0233, + "step": 12408 + }, + { + "epoch": 1.345294882914137, + "grad_norm": 4.723928451538086, + "learning_rate": 1.1071376759030903e-05, + "loss": 0.0432, + "step": 12409 + }, + { + "epoch": 1.3454032957502169, + "grad_norm": 0.20197579264640808, + "learning_rate": 1.1070651385463515e-05, + "loss": 0.0016, + "step": 12410 + }, + { + "epoch": 1.3455117085862967, + "grad_norm": 2.4329006671905518, + "learning_rate": 1.1069926011896129e-05, + "loss": 0.0227, + "step": 12411 + }, + { + "epoch": 1.3456201214223764, + "grad_norm": 0.0710696429014206, + "learning_rate": 1.1069200638328741e-05, + "loss": 0.0009, + "step": 12412 + }, + { + "epoch": 1.3457285342584562, + "grad_norm": 2.7972216606140137, + "learning_rate": 1.1068475264761353e-05, + "loss": 0.0149, + "step": 12413 + }, + { + "epoch": 1.345836947094536, + "grad_norm": 1.7546306848526, + "learning_rate": 1.1067749891193967e-05, + "loss": 0.0275, + "step": 12414 + }, + { + "epoch": 1.3459453599306157, + "grad_norm": 1.3069254159927368, + "learning_rate": 1.1067024517626579e-05, + "loss": 0.0546, + "step": 12415 + }, + { + "epoch": 1.3460537727666955, + "grad_norm": 3.027007818222046, + "learning_rate": 1.1066299144059191e-05, + "loss": 0.0176, + "step": 12416 + }, + { + "epoch": 1.3461621856027755, + "grad_norm": 4.405303478240967, + "learning_rate": 1.1065573770491805e-05, + "loss": 0.0675, + "step": 12417 + }, + { + "epoch": 1.3462705984388552, + "grad_norm": 2.3199095726013184, + "learning_rate": 1.1064848396924417e-05, + "loss": 0.0109, + "step": 12418 + }, + { + "epoch": 1.346379011274935, + "grad_norm": 6.178552150726318, + "learning_rate": 1.106412302335703e-05, + "loss": 0.0286, + "step": 12419 + }, + { + "epoch": 1.3464874241110147, + "grad_norm": 5.085862636566162, + "learning_rate": 1.1063397649789641e-05, + "loss": 0.0318, + "step": 12420 + }, + { + "epoch": 1.3465958369470945, + "grad_norm": 1.680721402168274, + "learning_rate": 1.1062672276222255e-05, + "loss": 0.0052, + "step": 12421 + }, + { + "epoch": 1.3467042497831743, + "grad_norm": 0.5211300849914551, + "learning_rate": 1.1061946902654867e-05, + "loss": 0.0064, + "step": 12422 + }, + { + "epoch": 1.346812662619254, + "grad_norm": 2.2933287620544434, + "learning_rate": 1.106122152908748e-05, + "loss": 0.0176, + "step": 12423 + }, + { + "epoch": 1.346921075455334, + "grad_norm": 2.6488873958587646, + "learning_rate": 1.1060496155520093e-05, + "loss": 0.0183, + "step": 12424 + }, + { + "epoch": 1.3470294882914138, + "grad_norm": 3.150874137878418, + "learning_rate": 1.1059770781952706e-05, + "loss": 0.0141, + "step": 12425 + }, + { + "epoch": 1.3471379011274935, + "grad_norm": 2.076447010040283, + "learning_rate": 1.1059045408385318e-05, + "loss": 0.0194, + "step": 12426 + }, + { + "epoch": 1.3472463139635733, + "grad_norm": 0.10395865887403488, + "learning_rate": 1.1058320034817933e-05, + "loss": 0.0007, + "step": 12427 + }, + { + "epoch": 1.347354726799653, + "grad_norm": 0.5338011980056763, + "learning_rate": 1.1057594661250545e-05, + "loss": 0.0042, + "step": 12428 + }, + { + "epoch": 1.3474631396357328, + "grad_norm": 2.676051139831543, + "learning_rate": 1.105686928768316e-05, + "loss": 0.0302, + "step": 12429 + }, + { + "epoch": 1.3475715524718126, + "grad_norm": 2.615360975265503, + "learning_rate": 1.1056143914115771e-05, + "loss": 0.0106, + "step": 12430 + }, + { + "epoch": 1.3476799653078926, + "grad_norm": 1.687715768814087, + "learning_rate": 1.1055418540548384e-05, + "loss": 0.0047, + "step": 12431 + }, + { + "epoch": 1.3477883781439721, + "grad_norm": 4.175474643707275, + "learning_rate": 1.1054693166980996e-05, + "loss": 0.0361, + "step": 12432 + }, + { + "epoch": 1.347896790980052, + "grad_norm": 2.9269473552703857, + "learning_rate": 1.105396779341361e-05, + "loss": 0.0275, + "step": 12433 + }, + { + "epoch": 1.3480052038161319, + "grad_norm": 1.0419467687606812, + "learning_rate": 1.1053242419846222e-05, + "loss": 0.0152, + "step": 12434 + }, + { + "epoch": 1.3481136166522116, + "grad_norm": 1.6172393560409546, + "learning_rate": 1.1052517046278834e-05, + "loss": 0.0135, + "step": 12435 + }, + { + "epoch": 1.3482220294882914, + "grad_norm": 11.34139633178711, + "learning_rate": 1.1051791672711448e-05, + "loss": 0.0652, + "step": 12436 + }, + { + "epoch": 1.3483304423243712, + "grad_norm": 5.585667133331299, + "learning_rate": 1.105106629914406e-05, + "loss": 0.0188, + "step": 12437 + }, + { + "epoch": 1.348438855160451, + "grad_norm": 1.2185347080230713, + "learning_rate": 1.1050340925576672e-05, + "loss": 0.0054, + "step": 12438 + }, + { + "epoch": 1.3485472679965307, + "grad_norm": 3.4198410511016846, + "learning_rate": 1.1049615552009286e-05, + "loss": 0.056, + "step": 12439 + }, + { + "epoch": 1.3486556808326107, + "grad_norm": 4.463300704956055, + "learning_rate": 1.1048890178441898e-05, + "loss": 0.0604, + "step": 12440 + }, + { + "epoch": 1.3487640936686904, + "grad_norm": 1.3688669204711914, + "learning_rate": 1.104816480487451e-05, + "loss": 0.0318, + "step": 12441 + }, + { + "epoch": 1.3488725065047702, + "grad_norm": 1.1975575685501099, + "learning_rate": 1.1047439431307124e-05, + "loss": 0.0473, + "step": 12442 + }, + { + "epoch": 1.34898091934085, + "grad_norm": 2.4305994510650635, + "learning_rate": 1.1046714057739736e-05, + "loss": 0.0469, + "step": 12443 + }, + { + "epoch": 1.3490893321769297, + "grad_norm": 0.8104126453399658, + "learning_rate": 1.1045988684172352e-05, + "loss": 0.0067, + "step": 12444 + }, + { + "epoch": 1.3491977450130095, + "grad_norm": 1.8583098649978638, + "learning_rate": 1.1045263310604964e-05, + "loss": 0.0166, + "step": 12445 + }, + { + "epoch": 1.3493061578490892, + "grad_norm": 1.6029270887374878, + "learning_rate": 1.1044537937037576e-05, + "loss": 0.0046, + "step": 12446 + }, + { + "epoch": 1.3494145706851692, + "grad_norm": 3.212322473526001, + "learning_rate": 1.1043812563470188e-05, + "loss": 0.0323, + "step": 12447 + }, + { + "epoch": 1.349522983521249, + "grad_norm": 3.2699503898620605, + "learning_rate": 1.1043087189902802e-05, + "loss": 0.0199, + "step": 12448 + }, + { + "epoch": 1.3496313963573288, + "grad_norm": 2.046828269958496, + "learning_rate": 1.1042361816335414e-05, + "loss": 0.0176, + "step": 12449 + }, + { + "epoch": 1.3497398091934085, + "grad_norm": 2.4121570587158203, + "learning_rate": 1.1041636442768026e-05, + "loss": 0.009, + "step": 12450 + }, + { + "epoch": 1.3498482220294883, + "grad_norm": 0.9340944290161133, + "learning_rate": 1.104091106920064e-05, + "loss": 0.0045, + "step": 12451 + }, + { + "epoch": 1.349956634865568, + "grad_norm": 2.4837381839752197, + "learning_rate": 1.1040185695633252e-05, + "loss": 0.0103, + "step": 12452 + }, + { + "epoch": 1.3500650477016478, + "grad_norm": 2.993232011795044, + "learning_rate": 1.1039460322065864e-05, + "loss": 0.0332, + "step": 12453 + }, + { + "epoch": 1.3501734605377278, + "grad_norm": 2.3700156211853027, + "learning_rate": 1.1038734948498478e-05, + "loss": 0.0375, + "step": 12454 + }, + { + "epoch": 1.3502818733738073, + "grad_norm": 1.9785566329956055, + "learning_rate": 1.103800957493109e-05, + "loss": 0.0282, + "step": 12455 + }, + { + "epoch": 1.3503902862098873, + "grad_norm": 1.99606192111969, + "learning_rate": 1.1037284201363702e-05, + "loss": 0.0379, + "step": 12456 + }, + { + "epoch": 1.350498699045967, + "grad_norm": 4.841333389282227, + "learning_rate": 1.1036558827796316e-05, + "loss": 0.0082, + "step": 12457 + }, + { + "epoch": 1.3506071118820469, + "grad_norm": 1.5952578783035278, + "learning_rate": 1.1035833454228928e-05, + "loss": 0.0218, + "step": 12458 + }, + { + "epoch": 1.3507155247181266, + "grad_norm": 1.5945775508880615, + "learning_rate": 1.103510808066154e-05, + "loss": 0.0255, + "step": 12459 + }, + { + "epoch": 1.3508239375542064, + "grad_norm": 0.471233069896698, + "learning_rate": 1.1034382707094153e-05, + "loss": 0.0064, + "step": 12460 + }, + { + "epoch": 1.3509323503902861, + "grad_norm": 0.7758845686912537, + "learning_rate": 1.1033657333526768e-05, + "loss": 0.0057, + "step": 12461 + }, + { + "epoch": 1.351040763226366, + "grad_norm": 4.40626859664917, + "learning_rate": 1.103293195995938e-05, + "loss": 0.0214, + "step": 12462 + }, + { + "epoch": 1.351149176062446, + "grad_norm": 1.6420080661773682, + "learning_rate": 1.1032206586391994e-05, + "loss": 0.0254, + "step": 12463 + }, + { + "epoch": 1.3512575888985257, + "grad_norm": 12.138104438781738, + "learning_rate": 1.1031481212824606e-05, + "loss": 0.031, + "step": 12464 + }, + { + "epoch": 1.3513660017346054, + "grad_norm": 3.1525447368621826, + "learning_rate": 1.1030755839257218e-05, + "loss": 0.0125, + "step": 12465 + }, + { + "epoch": 1.3514744145706852, + "grad_norm": 1.7739017009735107, + "learning_rate": 1.1030030465689832e-05, + "loss": 0.0177, + "step": 12466 + }, + { + "epoch": 1.351582827406765, + "grad_norm": 0.15227940678596497, + "learning_rate": 1.1029305092122444e-05, + "loss": 0.0011, + "step": 12467 + }, + { + "epoch": 1.3516912402428447, + "grad_norm": 0.3670990467071533, + "learning_rate": 1.1028579718555057e-05, + "loss": 0.0029, + "step": 12468 + }, + { + "epoch": 1.3517996530789245, + "grad_norm": 0.9285610318183899, + "learning_rate": 1.102785434498767e-05, + "loss": 0.0098, + "step": 12469 + }, + { + "epoch": 1.3519080659150045, + "grad_norm": 0.5905293822288513, + "learning_rate": 1.1027128971420283e-05, + "loss": 0.0039, + "step": 12470 + }, + { + "epoch": 1.352016478751084, + "grad_norm": 1.1409392356872559, + "learning_rate": 1.1026403597852895e-05, + "loss": 0.0216, + "step": 12471 + }, + { + "epoch": 1.352124891587164, + "grad_norm": 0.9164707660675049, + "learning_rate": 1.1025678224285507e-05, + "loss": 0.0094, + "step": 12472 + }, + { + "epoch": 1.3522333044232437, + "grad_norm": 4.921554088592529, + "learning_rate": 1.102495285071812e-05, + "loss": 0.0619, + "step": 12473 + }, + { + "epoch": 1.3523417172593235, + "grad_norm": 0.8818501234054565, + "learning_rate": 1.1024227477150733e-05, + "loss": 0.0054, + "step": 12474 + }, + { + "epoch": 1.3524501300954033, + "grad_norm": 1.8443162441253662, + "learning_rate": 1.1023502103583345e-05, + "loss": 0.009, + "step": 12475 + }, + { + "epoch": 1.352558542931483, + "grad_norm": 0.8221901059150696, + "learning_rate": 1.1022776730015959e-05, + "loss": 0.0074, + "step": 12476 + }, + { + "epoch": 1.352666955767563, + "grad_norm": 1.594741702079773, + "learning_rate": 1.1022051356448571e-05, + "loss": 0.0108, + "step": 12477 + }, + { + "epoch": 1.3527753686036426, + "grad_norm": 0.38601818680763245, + "learning_rate": 1.1021325982881183e-05, + "loss": 0.0023, + "step": 12478 + }, + { + "epoch": 1.3528837814397225, + "grad_norm": 0.7915199995040894, + "learning_rate": 1.1020600609313799e-05, + "loss": 0.0203, + "step": 12479 + }, + { + "epoch": 1.3529921942758023, + "grad_norm": 0.9677950739860535, + "learning_rate": 1.101987523574641e-05, + "loss": 0.004, + "step": 12480 + }, + { + "epoch": 1.353100607111882, + "grad_norm": 1.3704125881195068, + "learning_rate": 1.1019149862179025e-05, + "loss": 0.0291, + "step": 12481 + }, + { + "epoch": 1.3532090199479618, + "grad_norm": 8.595775604248047, + "learning_rate": 1.1018424488611637e-05, + "loss": 0.0107, + "step": 12482 + }, + { + "epoch": 1.3533174327840416, + "grad_norm": 0.12078754603862762, + "learning_rate": 1.1017699115044249e-05, + "loss": 0.0012, + "step": 12483 + }, + { + "epoch": 1.3534258456201214, + "grad_norm": 9.488210678100586, + "learning_rate": 1.1016973741476861e-05, + "loss": 0.0148, + "step": 12484 + }, + { + "epoch": 1.3535342584562011, + "grad_norm": 1.7606220245361328, + "learning_rate": 1.1016248367909475e-05, + "loss": 0.028, + "step": 12485 + }, + { + "epoch": 1.353642671292281, + "grad_norm": 1.4303420782089233, + "learning_rate": 1.1015522994342087e-05, + "loss": 0.0254, + "step": 12486 + }, + { + "epoch": 1.3537510841283609, + "grad_norm": 0.35404983162879944, + "learning_rate": 1.1014797620774699e-05, + "loss": 0.0014, + "step": 12487 + }, + { + "epoch": 1.3538594969644406, + "grad_norm": 5.159871578216553, + "learning_rate": 1.1014072247207313e-05, + "loss": 0.0495, + "step": 12488 + }, + { + "epoch": 1.3539679098005204, + "grad_norm": 1.6853046417236328, + "learning_rate": 1.1013346873639925e-05, + "loss": 0.0392, + "step": 12489 + }, + { + "epoch": 1.3540763226366002, + "grad_norm": 8.050193786621094, + "learning_rate": 1.1012621500072537e-05, + "loss": 0.0301, + "step": 12490 + }, + { + "epoch": 1.35418473547268, + "grad_norm": 1.01979398727417, + "learning_rate": 1.1011896126505151e-05, + "loss": 0.0066, + "step": 12491 + }, + { + "epoch": 1.3542931483087597, + "grad_norm": 1.5497812032699585, + "learning_rate": 1.1011170752937763e-05, + "loss": 0.0203, + "step": 12492 + }, + { + "epoch": 1.3544015611448397, + "grad_norm": 4.635148048400879, + "learning_rate": 1.1010445379370375e-05, + "loss": 0.0593, + "step": 12493 + }, + { + "epoch": 1.3545099739809192, + "grad_norm": 0.6006371974945068, + "learning_rate": 1.100972000580299e-05, + "loss": 0.0027, + "step": 12494 + }, + { + "epoch": 1.3546183868169992, + "grad_norm": 1.0935757160186768, + "learning_rate": 1.1008994632235601e-05, + "loss": 0.0082, + "step": 12495 + }, + { + "epoch": 1.354726799653079, + "grad_norm": 2.222628355026245, + "learning_rate": 1.1008269258668217e-05, + "loss": 0.0074, + "step": 12496 + }, + { + "epoch": 1.3548352124891587, + "grad_norm": 1.6313765048980713, + "learning_rate": 1.1007543885100829e-05, + "loss": 0.0147, + "step": 12497 + }, + { + "epoch": 1.3549436253252385, + "grad_norm": 1.1145236492156982, + "learning_rate": 1.1006818511533441e-05, + "loss": 0.0304, + "step": 12498 + }, + { + "epoch": 1.3550520381613183, + "grad_norm": 0.7998908162117004, + "learning_rate": 1.1006093137966053e-05, + "loss": 0.0052, + "step": 12499 + }, + { + "epoch": 1.355160450997398, + "grad_norm": 4.836028099060059, + "learning_rate": 1.1005367764398667e-05, + "loss": 0.0297, + "step": 12500 + }, + { + "epoch": 1.3552688638334778, + "grad_norm": 2.2866764068603516, + "learning_rate": 1.100464239083128e-05, + "loss": 0.021, + "step": 12501 + }, + { + "epoch": 1.3553772766695578, + "grad_norm": 4.0136284828186035, + "learning_rate": 1.1003917017263891e-05, + "loss": 0.0374, + "step": 12502 + }, + { + "epoch": 1.3554856895056375, + "grad_norm": 2.591918468475342, + "learning_rate": 1.1003191643696505e-05, + "loss": 0.0288, + "step": 12503 + }, + { + "epoch": 1.3555941023417173, + "grad_norm": 5.2313432693481445, + "learning_rate": 1.1002466270129117e-05, + "loss": 0.036, + "step": 12504 + }, + { + "epoch": 1.355702515177797, + "grad_norm": 7.826157569885254, + "learning_rate": 1.100174089656173e-05, + "loss": 0.044, + "step": 12505 + }, + { + "epoch": 1.3558109280138768, + "grad_norm": 6.471358299255371, + "learning_rate": 1.1001015522994343e-05, + "loss": 0.0285, + "step": 12506 + }, + { + "epoch": 1.3559193408499566, + "grad_norm": 20.12605094909668, + "learning_rate": 1.1000290149426956e-05, + "loss": 0.0447, + "step": 12507 + }, + { + "epoch": 1.3560277536860363, + "grad_norm": 0.6355889439582825, + "learning_rate": 1.0999564775859568e-05, + "loss": 0.0044, + "step": 12508 + }, + { + "epoch": 1.3561361665221163, + "grad_norm": 1.3516814708709717, + "learning_rate": 1.0998839402292182e-05, + "loss": 0.0247, + "step": 12509 + }, + { + "epoch": 1.356244579358196, + "grad_norm": 0.2980460822582245, + "learning_rate": 1.0998114028724794e-05, + "loss": 0.0028, + "step": 12510 + }, + { + "epoch": 1.3563529921942759, + "grad_norm": 0.889867901802063, + "learning_rate": 1.0997388655157406e-05, + "loss": 0.0119, + "step": 12511 + }, + { + "epoch": 1.3564614050303556, + "grad_norm": 3.423039197921753, + "learning_rate": 1.0996663281590018e-05, + "loss": 0.0107, + "step": 12512 + }, + { + "epoch": 1.3565698178664354, + "grad_norm": 0.22343724966049194, + "learning_rate": 1.0995937908022633e-05, + "loss": 0.0008, + "step": 12513 + }, + { + "epoch": 1.3566782307025151, + "grad_norm": 5.6687798500061035, + "learning_rate": 1.0995212534455246e-05, + "loss": 0.0176, + "step": 12514 + }, + { + "epoch": 1.356786643538595, + "grad_norm": 0.5136627554893494, + "learning_rate": 1.099448716088786e-05, + "loss": 0.0061, + "step": 12515 + }, + { + "epoch": 1.356895056374675, + "grad_norm": 0.7176454067230225, + "learning_rate": 1.0993761787320472e-05, + "loss": 0.0058, + "step": 12516 + }, + { + "epoch": 1.3570034692107544, + "grad_norm": 2.2518458366394043, + "learning_rate": 1.0993036413753084e-05, + "loss": 0.031, + "step": 12517 + }, + { + "epoch": 1.3571118820468344, + "grad_norm": 5.867212772369385, + "learning_rate": 1.0992311040185698e-05, + "loss": 0.0653, + "step": 12518 + }, + { + "epoch": 1.3572202948829142, + "grad_norm": 4.858579158782959, + "learning_rate": 1.099158566661831e-05, + "loss": 0.0294, + "step": 12519 + }, + { + "epoch": 1.357328707718994, + "grad_norm": 0.13835762441158295, + "learning_rate": 1.0990860293050922e-05, + "loss": 0.002, + "step": 12520 + }, + { + "epoch": 1.3574371205550737, + "grad_norm": 3.1025924682617188, + "learning_rate": 1.0990134919483536e-05, + "loss": 0.0353, + "step": 12521 + }, + { + "epoch": 1.3575455333911535, + "grad_norm": 1.3218899965286255, + "learning_rate": 1.0989409545916148e-05, + "loss": 0.018, + "step": 12522 + }, + { + "epoch": 1.3576539462272332, + "grad_norm": 1.99193274974823, + "learning_rate": 1.098868417234876e-05, + "loss": 0.0108, + "step": 12523 + }, + { + "epoch": 1.357762359063313, + "grad_norm": 0.5870062708854675, + "learning_rate": 1.0987958798781372e-05, + "loss": 0.0049, + "step": 12524 + }, + { + "epoch": 1.357870771899393, + "grad_norm": 0.8779837489128113, + "learning_rate": 1.0987233425213986e-05, + "loss": 0.0039, + "step": 12525 + }, + { + "epoch": 1.3579791847354727, + "grad_norm": 6.606132507324219, + "learning_rate": 1.0986508051646598e-05, + "loss": 0.0334, + "step": 12526 + }, + { + "epoch": 1.3580875975715525, + "grad_norm": 12.179676055908203, + "learning_rate": 1.098578267807921e-05, + "loss": 0.0275, + "step": 12527 + }, + { + "epoch": 1.3581960104076323, + "grad_norm": 0.41231170296669006, + "learning_rate": 1.0985057304511824e-05, + "loss": 0.0022, + "step": 12528 + }, + { + "epoch": 1.358304423243712, + "grad_norm": 0.633420467376709, + "learning_rate": 1.0984331930944436e-05, + "loss": 0.0038, + "step": 12529 + }, + { + "epoch": 1.3584128360797918, + "grad_norm": 2.2167258262634277, + "learning_rate": 1.0983606557377052e-05, + "loss": 0.0187, + "step": 12530 + }, + { + "epoch": 1.3585212489158716, + "grad_norm": 0.4141755700111389, + "learning_rate": 1.0982881183809664e-05, + "loss": 0.0039, + "step": 12531 + }, + { + "epoch": 1.3586296617519515, + "grad_norm": 3.0545833110809326, + "learning_rate": 1.0982155810242276e-05, + "loss": 0.0613, + "step": 12532 + }, + { + "epoch": 1.358738074588031, + "grad_norm": 0.4923432767391205, + "learning_rate": 1.098143043667489e-05, + "loss": 0.0028, + "step": 12533 + }, + { + "epoch": 1.358846487424111, + "grad_norm": 3.6807055473327637, + "learning_rate": 1.0980705063107502e-05, + "loss": 0.0096, + "step": 12534 + }, + { + "epoch": 1.3589549002601908, + "grad_norm": 0.3374450206756592, + "learning_rate": 1.0979979689540114e-05, + "loss": 0.0028, + "step": 12535 + }, + { + "epoch": 1.3590633130962706, + "grad_norm": 0.3315739631652832, + "learning_rate": 1.0979254315972726e-05, + "loss": 0.0017, + "step": 12536 + }, + { + "epoch": 1.3591717259323504, + "grad_norm": 3.3872225284576416, + "learning_rate": 1.097852894240534e-05, + "loss": 0.0149, + "step": 12537 + }, + { + "epoch": 1.3592801387684301, + "grad_norm": 0.5122140049934387, + "learning_rate": 1.0977803568837952e-05, + "loss": 0.0067, + "step": 12538 + }, + { + "epoch": 1.35938855160451, + "grad_norm": 6.850526332855225, + "learning_rate": 1.0977078195270564e-05, + "loss": 0.0298, + "step": 12539 + }, + { + "epoch": 1.3594969644405897, + "grad_norm": 8.730216026306152, + "learning_rate": 1.0976352821703178e-05, + "loss": 0.0313, + "step": 12540 + }, + { + "epoch": 1.3596053772766696, + "grad_norm": 0.8126007914543152, + "learning_rate": 1.097562744813579e-05, + "loss": 0.0048, + "step": 12541 + }, + { + "epoch": 1.3597137901127494, + "grad_norm": 9.223495483398438, + "learning_rate": 1.0974902074568403e-05, + "loss": 0.0453, + "step": 12542 + }, + { + "epoch": 1.3598222029488292, + "grad_norm": 8.753787994384766, + "learning_rate": 1.0974176701001016e-05, + "loss": 0.0713, + "step": 12543 + }, + { + "epoch": 1.359930615784909, + "grad_norm": 3.6923885345458984, + "learning_rate": 1.0973451327433629e-05, + "loss": 0.0155, + "step": 12544 + }, + { + "epoch": 1.3600390286209887, + "grad_norm": 1.4631348848342896, + "learning_rate": 1.097272595386624e-05, + "loss": 0.0071, + "step": 12545 + }, + { + "epoch": 1.3601474414570685, + "grad_norm": 0.8548509478569031, + "learning_rate": 1.0972000580298855e-05, + "loss": 0.01, + "step": 12546 + }, + { + "epoch": 1.3602558542931482, + "grad_norm": 1.6687513589859009, + "learning_rate": 1.0971275206731467e-05, + "loss": 0.0117, + "step": 12547 + }, + { + "epoch": 1.3603642671292282, + "grad_norm": 0.11812341958284378, + "learning_rate": 1.0970549833164082e-05, + "loss": 0.0007, + "step": 12548 + }, + { + "epoch": 1.360472679965308, + "grad_norm": 0.4007009267807007, + "learning_rate": 1.0969824459596694e-05, + "loss": 0.0018, + "step": 12549 + }, + { + "epoch": 1.3605810928013877, + "grad_norm": 1.5772604942321777, + "learning_rate": 1.0969099086029306e-05, + "loss": 0.0053, + "step": 12550 + }, + { + "epoch": 1.3606895056374675, + "grad_norm": 0.4650269150733948, + "learning_rate": 1.0968373712461919e-05, + "loss": 0.0013, + "step": 12551 + }, + { + "epoch": 1.3607979184735473, + "grad_norm": 4.2252044677734375, + "learning_rate": 1.0967648338894532e-05, + "loss": 0.0202, + "step": 12552 + }, + { + "epoch": 1.360906331309627, + "grad_norm": 7.841843605041504, + "learning_rate": 1.0966922965327145e-05, + "loss": 0.0569, + "step": 12553 + }, + { + "epoch": 1.3610147441457068, + "grad_norm": 3.6394786834716797, + "learning_rate": 1.0966197591759757e-05, + "loss": 0.0249, + "step": 12554 + }, + { + "epoch": 1.3611231569817868, + "grad_norm": 3.07370662689209, + "learning_rate": 1.096547221819237e-05, + "loss": 0.0568, + "step": 12555 + }, + { + "epoch": 1.3612315698178663, + "grad_norm": 2.5722334384918213, + "learning_rate": 1.0964746844624983e-05, + "loss": 0.044, + "step": 12556 + }, + { + "epoch": 1.3613399826539463, + "grad_norm": 0.6655086278915405, + "learning_rate": 1.0964021471057595e-05, + "loss": 0.0074, + "step": 12557 + }, + { + "epoch": 1.361448395490026, + "grad_norm": 0.12410645931959152, + "learning_rate": 1.0963296097490209e-05, + "loss": 0.0011, + "step": 12558 + }, + { + "epoch": 1.3615568083261058, + "grad_norm": 1.8395966291427612, + "learning_rate": 1.0962570723922821e-05, + "loss": 0.0265, + "step": 12559 + }, + { + "epoch": 1.3616652211621856, + "grad_norm": 3.1905009746551514, + "learning_rate": 1.0961845350355433e-05, + "loss": 0.0403, + "step": 12560 + }, + { + "epoch": 1.3617736339982653, + "grad_norm": 1.8678408861160278, + "learning_rate": 1.0961119976788047e-05, + "loss": 0.0266, + "step": 12561 + }, + { + "epoch": 1.361882046834345, + "grad_norm": 0.3252362608909607, + "learning_rate": 1.0960394603220659e-05, + "loss": 0.001, + "step": 12562 + }, + { + "epoch": 1.3619904596704249, + "grad_norm": 0.5569558143615723, + "learning_rate": 1.0959669229653271e-05, + "loss": 0.0034, + "step": 12563 + }, + { + "epoch": 1.3620988725065049, + "grad_norm": 5.434787273406982, + "learning_rate": 1.0958943856085883e-05, + "loss": 0.0402, + "step": 12564 + }, + { + "epoch": 1.3622072853425846, + "grad_norm": 0.36683374643325806, + "learning_rate": 1.0958218482518499e-05, + "loss": 0.0033, + "step": 12565 + }, + { + "epoch": 1.3623156981786644, + "grad_norm": 2.2386696338653564, + "learning_rate": 1.0957493108951111e-05, + "loss": 0.0555, + "step": 12566 + }, + { + "epoch": 1.3624241110147441, + "grad_norm": 3.825895309448242, + "learning_rate": 1.0956767735383725e-05, + "loss": 0.0846, + "step": 12567 + }, + { + "epoch": 1.362532523850824, + "grad_norm": 4.495478630065918, + "learning_rate": 1.0956042361816337e-05, + "loss": 0.0253, + "step": 12568 + }, + { + "epoch": 1.3626409366869037, + "grad_norm": 8.314257621765137, + "learning_rate": 1.0955316988248949e-05, + "loss": 0.0699, + "step": 12569 + }, + { + "epoch": 1.3627493495229834, + "grad_norm": 2.777621269226074, + "learning_rate": 1.0954591614681563e-05, + "loss": 0.0157, + "step": 12570 + }, + { + "epoch": 1.3628577623590634, + "grad_norm": 1.8612556457519531, + "learning_rate": 1.0953866241114175e-05, + "loss": 0.0271, + "step": 12571 + }, + { + "epoch": 1.3629661751951432, + "grad_norm": 0.4669438600540161, + "learning_rate": 1.0953140867546787e-05, + "loss": 0.0027, + "step": 12572 + }, + { + "epoch": 1.363074588031223, + "grad_norm": 4.518540382385254, + "learning_rate": 1.0952415493979401e-05, + "loss": 0.0384, + "step": 12573 + }, + { + "epoch": 1.3631830008673027, + "grad_norm": 0.6222785711288452, + "learning_rate": 1.0951690120412013e-05, + "loss": 0.0125, + "step": 12574 + }, + { + "epoch": 1.3632914137033825, + "grad_norm": 3.2819273471832275, + "learning_rate": 1.0950964746844625e-05, + "loss": 0.03, + "step": 12575 + }, + { + "epoch": 1.3633998265394622, + "grad_norm": 2.5703556537628174, + "learning_rate": 1.0950239373277237e-05, + "loss": 0.0183, + "step": 12576 + }, + { + "epoch": 1.363508239375542, + "grad_norm": 1.2579959630966187, + "learning_rate": 1.0949513999709851e-05, + "loss": 0.0122, + "step": 12577 + }, + { + "epoch": 1.363616652211622, + "grad_norm": 1.6256814002990723, + "learning_rate": 1.0948788626142463e-05, + "loss": 0.02, + "step": 12578 + }, + { + "epoch": 1.3637250650477015, + "grad_norm": 2.303171396255493, + "learning_rate": 1.0948063252575076e-05, + "loss": 0.0534, + "step": 12579 + }, + { + "epoch": 1.3638334778837815, + "grad_norm": 1.0167944431304932, + "learning_rate": 1.094733787900769e-05, + "loss": 0.0073, + "step": 12580 + }, + { + "epoch": 1.3639418907198613, + "grad_norm": 7.484897136688232, + "learning_rate": 1.0946612505440302e-05, + "loss": 0.0461, + "step": 12581 + }, + { + "epoch": 1.364050303555941, + "grad_norm": 7.206745147705078, + "learning_rate": 1.0945887131872917e-05, + "loss": 0.0272, + "step": 12582 + }, + { + "epoch": 1.3641587163920208, + "grad_norm": 1.086199164390564, + "learning_rate": 1.094516175830553e-05, + "loss": 0.0056, + "step": 12583 + }, + { + "epoch": 1.3642671292281006, + "grad_norm": 1.3125766515731812, + "learning_rate": 1.0944436384738141e-05, + "loss": 0.0122, + "step": 12584 + }, + { + "epoch": 1.3643755420641803, + "grad_norm": 3.1230275630950928, + "learning_rate": 1.0943711011170755e-05, + "loss": 0.0301, + "step": 12585 + }, + { + "epoch": 1.36448395490026, + "grad_norm": 1.939231514930725, + "learning_rate": 1.0942985637603367e-05, + "loss": 0.0181, + "step": 12586 + }, + { + "epoch": 1.36459236773634, + "grad_norm": 0.9050084948539734, + "learning_rate": 1.094226026403598e-05, + "loss": 0.0081, + "step": 12587 + }, + { + "epoch": 1.3647007805724198, + "grad_norm": 4.439608573913574, + "learning_rate": 1.0941534890468593e-05, + "loss": 0.0764, + "step": 12588 + }, + { + "epoch": 1.3648091934084996, + "grad_norm": 3.151613712310791, + "learning_rate": 1.0940809516901206e-05, + "loss": 0.0532, + "step": 12589 + }, + { + "epoch": 1.3649176062445794, + "grad_norm": 1.3582040071487427, + "learning_rate": 1.0940084143333818e-05, + "loss": 0.0093, + "step": 12590 + }, + { + "epoch": 1.3650260190806591, + "grad_norm": 2.621432304382324, + "learning_rate": 1.093935876976643e-05, + "loss": 0.0202, + "step": 12591 + }, + { + "epoch": 1.3651344319167389, + "grad_norm": 9.749279975891113, + "learning_rate": 1.0938633396199044e-05, + "loss": 0.042, + "step": 12592 + }, + { + "epoch": 1.3652428447528187, + "grad_norm": 2.1604580879211426, + "learning_rate": 1.0937908022631656e-05, + "loss": 0.0301, + "step": 12593 + }, + { + "epoch": 1.3653512575888986, + "grad_norm": 0.8020502328872681, + "learning_rate": 1.0937182649064268e-05, + "loss": 0.0027, + "step": 12594 + }, + { + "epoch": 1.3654596704249784, + "grad_norm": 4.021482467651367, + "learning_rate": 1.0936457275496882e-05, + "loss": 0.021, + "step": 12595 + }, + { + "epoch": 1.3655680832610582, + "grad_norm": 2.892784595489502, + "learning_rate": 1.0935731901929494e-05, + "loss": 0.0218, + "step": 12596 + }, + { + "epoch": 1.365676496097138, + "grad_norm": 1.818537712097168, + "learning_rate": 1.0935006528362106e-05, + "loss": 0.0151, + "step": 12597 + }, + { + "epoch": 1.3657849089332177, + "grad_norm": 4.358804225921631, + "learning_rate": 1.093428115479472e-05, + "loss": 0.003, + "step": 12598 + }, + { + "epoch": 1.3658933217692975, + "grad_norm": 0.27435359358787537, + "learning_rate": 1.0933555781227334e-05, + "loss": 0.0036, + "step": 12599 + }, + { + "epoch": 1.3660017346053772, + "grad_norm": 0.8177271485328674, + "learning_rate": 1.0932830407659948e-05, + "loss": 0.0036, + "step": 12600 + }, + { + "epoch": 1.3661101474414572, + "grad_norm": 0.47807663679122925, + "learning_rate": 1.093210503409256e-05, + "loss": 0.002, + "step": 12601 + }, + { + "epoch": 1.3662185602775367, + "grad_norm": 4.5368757247924805, + "learning_rate": 1.0931379660525172e-05, + "loss": 0.0561, + "step": 12602 + }, + { + "epoch": 1.3663269731136167, + "grad_norm": 4.470602512359619, + "learning_rate": 1.0930654286957784e-05, + "loss": 0.0326, + "step": 12603 + }, + { + "epoch": 1.3664353859496965, + "grad_norm": 1.068881869316101, + "learning_rate": 1.0929928913390398e-05, + "loss": 0.0061, + "step": 12604 + }, + { + "epoch": 1.3665437987857763, + "grad_norm": 1.9257168769836426, + "learning_rate": 1.092920353982301e-05, + "loss": 0.0173, + "step": 12605 + }, + { + "epoch": 1.366652211621856, + "grad_norm": 0.8952816128730774, + "learning_rate": 1.0928478166255622e-05, + "loss": 0.0051, + "step": 12606 + }, + { + "epoch": 1.3667606244579358, + "grad_norm": 9.63854694366455, + "learning_rate": 1.0927752792688236e-05, + "loss": 0.0196, + "step": 12607 + }, + { + "epoch": 1.3668690372940155, + "grad_norm": 0.9832680225372314, + "learning_rate": 1.0927027419120848e-05, + "loss": 0.0208, + "step": 12608 + }, + { + "epoch": 1.3669774501300953, + "grad_norm": 3.0834591388702393, + "learning_rate": 1.092630204555346e-05, + "loss": 0.0418, + "step": 12609 + }, + { + "epoch": 1.3670858629661753, + "grad_norm": 1.9513918161392212, + "learning_rate": 1.0925576671986074e-05, + "loss": 0.0209, + "step": 12610 + }, + { + "epoch": 1.367194275802255, + "grad_norm": 3.866154432296753, + "learning_rate": 1.0924851298418686e-05, + "loss": 0.0399, + "step": 12611 + }, + { + "epoch": 1.3673026886383348, + "grad_norm": 0.8022381067276001, + "learning_rate": 1.0924125924851298e-05, + "loss": 0.0071, + "step": 12612 + }, + { + "epoch": 1.3674111014744146, + "grad_norm": 4.037613868713379, + "learning_rate": 1.0923400551283912e-05, + "loss": 0.0297, + "step": 12613 + }, + { + "epoch": 1.3675195143104943, + "grad_norm": 1.1415845155715942, + "learning_rate": 1.0922675177716524e-05, + "loss": 0.004, + "step": 12614 + }, + { + "epoch": 1.367627927146574, + "grad_norm": 4.9741129875183105, + "learning_rate": 1.0921949804149136e-05, + "loss": 0.0723, + "step": 12615 + }, + { + "epoch": 1.3677363399826539, + "grad_norm": 1.1749358177185059, + "learning_rate": 1.0921224430581749e-05, + "loss": 0.0028, + "step": 12616 + }, + { + "epoch": 1.3678447528187339, + "grad_norm": 4.195013523101807, + "learning_rate": 1.0920499057014364e-05, + "loss": 0.055, + "step": 12617 + }, + { + "epoch": 1.3679531656548134, + "grad_norm": 0.2702142298221588, + "learning_rate": 1.0919773683446976e-05, + "loss": 0.0029, + "step": 12618 + }, + { + "epoch": 1.3680615784908934, + "grad_norm": 3.1008105278015137, + "learning_rate": 1.091904830987959e-05, + "loss": 0.0201, + "step": 12619 + }, + { + "epoch": 1.3681699913269731, + "grad_norm": 3.490799903869629, + "learning_rate": 1.0918322936312202e-05, + "loss": 0.0087, + "step": 12620 + }, + { + "epoch": 1.368278404163053, + "grad_norm": 1.1376222372055054, + "learning_rate": 1.0917597562744814e-05, + "loss": 0.0058, + "step": 12621 + }, + { + "epoch": 1.3683868169991327, + "grad_norm": 2.1371545791625977, + "learning_rate": 1.0916872189177428e-05, + "loss": 0.019, + "step": 12622 + }, + { + "epoch": 1.3684952298352124, + "grad_norm": 0.36463049054145813, + "learning_rate": 1.091614681561004e-05, + "loss": 0.0037, + "step": 12623 + }, + { + "epoch": 1.3686036426712924, + "grad_norm": 3.1729061603546143, + "learning_rate": 1.0915421442042653e-05, + "loss": 0.0314, + "step": 12624 + }, + { + "epoch": 1.368712055507372, + "grad_norm": 4.818906784057617, + "learning_rate": 1.0914696068475266e-05, + "loss": 0.0477, + "step": 12625 + }, + { + "epoch": 1.368820468343452, + "grad_norm": 3.376342535018921, + "learning_rate": 1.0913970694907879e-05, + "loss": 0.0252, + "step": 12626 + }, + { + "epoch": 1.3689288811795317, + "grad_norm": 1.6945440769195557, + "learning_rate": 1.091324532134049e-05, + "loss": 0.0103, + "step": 12627 + }, + { + "epoch": 1.3690372940156115, + "grad_norm": 6.463050842285156, + "learning_rate": 1.0912519947773103e-05, + "loss": 0.0366, + "step": 12628 + }, + { + "epoch": 1.3691457068516912, + "grad_norm": 1.0592050552368164, + "learning_rate": 1.0911794574205717e-05, + "loss": 0.0173, + "step": 12629 + }, + { + "epoch": 1.369254119687771, + "grad_norm": 1.761641502380371, + "learning_rate": 1.0911069200638329e-05, + "loss": 0.0075, + "step": 12630 + }, + { + "epoch": 1.3693625325238508, + "grad_norm": 4.183227062225342, + "learning_rate": 1.0910343827070941e-05, + "loss": 0.0107, + "step": 12631 + }, + { + "epoch": 1.3694709453599305, + "grad_norm": 1.3689773082733154, + "learning_rate": 1.0909618453503555e-05, + "loss": 0.004, + "step": 12632 + }, + { + "epoch": 1.3695793581960105, + "grad_norm": 6.849910259246826, + "learning_rate": 1.0908893079936167e-05, + "loss": 0.0256, + "step": 12633 + }, + { + "epoch": 1.3696877710320903, + "grad_norm": 1.985892653465271, + "learning_rate": 1.0908167706368782e-05, + "loss": 0.0342, + "step": 12634 + }, + { + "epoch": 1.36979618386817, + "grad_norm": 4.003647804260254, + "learning_rate": 1.0907442332801395e-05, + "loss": 0.06, + "step": 12635 + }, + { + "epoch": 1.3699045967042498, + "grad_norm": 6.360057830810547, + "learning_rate": 1.0906716959234007e-05, + "loss": 0.034, + "step": 12636 + }, + { + "epoch": 1.3700130095403296, + "grad_norm": 8.121271133422852, + "learning_rate": 1.090599158566662e-05, + "loss": 0.0235, + "step": 12637 + }, + { + "epoch": 1.3701214223764093, + "grad_norm": 1.7716096639633179, + "learning_rate": 1.0905266212099233e-05, + "loss": 0.024, + "step": 12638 + }, + { + "epoch": 1.370229835212489, + "grad_norm": 3.1624131202697754, + "learning_rate": 1.0904540838531845e-05, + "loss": 0.0313, + "step": 12639 + }, + { + "epoch": 1.370338248048569, + "grad_norm": 1.727105975151062, + "learning_rate": 1.0903815464964459e-05, + "loss": 0.0288, + "step": 12640 + }, + { + "epoch": 1.3704466608846486, + "grad_norm": 7.657697677612305, + "learning_rate": 1.090309009139707e-05, + "loss": 0.0585, + "step": 12641 + }, + { + "epoch": 1.3705550737207286, + "grad_norm": 3.983966827392578, + "learning_rate": 1.0902364717829683e-05, + "loss": 0.0293, + "step": 12642 + }, + { + "epoch": 1.3706634865568084, + "grad_norm": 8.201592445373535, + "learning_rate": 1.0901639344262295e-05, + "loss": 0.0247, + "step": 12643 + }, + { + "epoch": 1.3707718993928881, + "grad_norm": 6.834232330322266, + "learning_rate": 1.0900913970694909e-05, + "loss": 0.0276, + "step": 12644 + }, + { + "epoch": 1.370880312228968, + "grad_norm": 0.9704095125198364, + "learning_rate": 1.0900188597127521e-05, + "loss": 0.0099, + "step": 12645 + }, + { + "epoch": 1.3709887250650477, + "grad_norm": 6.182572841644287, + "learning_rate": 1.0899463223560133e-05, + "loss": 0.0221, + "step": 12646 + }, + { + "epoch": 1.3710971379011274, + "grad_norm": 0.909884512424469, + "learning_rate": 1.0898737849992747e-05, + "loss": 0.0062, + "step": 12647 + }, + { + "epoch": 1.3712055507372072, + "grad_norm": 0.15546666085720062, + "learning_rate": 1.089801247642536e-05, + "loss": 0.0022, + "step": 12648 + }, + { + "epoch": 1.3713139635732872, + "grad_norm": 8.3597412109375, + "learning_rate": 1.0897287102857971e-05, + "loss": 0.0123, + "step": 12649 + }, + { + "epoch": 1.371422376409367, + "grad_norm": 0.9798212051391602, + "learning_rate": 1.0896561729290585e-05, + "loss": 0.0118, + "step": 12650 + }, + { + "epoch": 1.3715307892454467, + "grad_norm": 6.264994144439697, + "learning_rate": 1.0895836355723199e-05, + "loss": 0.0485, + "step": 12651 + }, + { + "epoch": 1.3716392020815265, + "grad_norm": 0.6997154355049133, + "learning_rate": 1.0895110982155813e-05, + "loss": 0.0035, + "step": 12652 + }, + { + "epoch": 1.3717476149176062, + "grad_norm": 2.0624372959136963, + "learning_rate": 1.0894385608588425e-05, + "loss": 0.0056, + "step": 12653 + }, + { + "epoch": 1.371856027753686, + "grad_norm": 2.3619070053100586, + "learning_rate": 1.0893660235021037e-05, + "loss": 0.0622, + "step": 12654 + }, + { + "epoch": 1.3719644405897657, + "grad_norm": 1.0430808067321777, + "learning_rate": 1.089293486145365e-05, + "loss": 0.0066, + "step": 12655 + }, + { + "epoch": 1.3720728534258457, + "grad_norm": 3.4700169563293457, + "learning_rate": 1.0892209487886263e-05, + "loss": 0.014, + "step": 12656 + }, + { + "epoch": 1.3721812662619255, + "grad_norm": 10.496434211730957, + "learning_rate": 1.0891484114318875e-05, + "loss": 0.0148, + "step": 12657 + }, + { + "epoch": 1.3722896790980053, + "grad_norm": 1.1209349632263184, + "learning_rate": 1.0890758740751487e-05, + "loss": 0.028, + "step": 12658 + }, + { + "epoch": 1.372398091934085, + "grad_norm": 2.2380332946777344, + "learning_rate": 1.0890033367184101e-05, + "loss": 0.031, + "step": 12659 + }, + { + "epoch": 1.3725065047701648, + "grad_norm": 4.092536926269531, + "learning_rate": 1.0889307993616713e-05, + "loss": 0.0159, + "step": 12660 + }, + { + "epoch": 1.3726149176062445, + "grad_norm": 1.03084135055542, + "learning_rate": 1.0888582620049326e-05, + "loss": 0.0228, + "step": 12661 + }, + { + "epoch": 1.3727233304423243, + "grad_norm": 4.1750264167785645, + "learning_rate": 1.088785724648194e-05, + "loss": 0.042, + "step": 12662 + }, + { + "epoch": 1.3728317432784043, + "grad_norm": 1.4110764265060425, + "learning_rate": 1.0887131872914552e-05, + "loss": 0.0383, + "step": 12663 + }, + { + "epoch": 1.3729401561144838, + "grad_norm": 2.8070731163024902, + "learning_rate": 1.0886406499347164e-05, + "loss": 0.0445, + "step": 12664 + }, + { + "epoch": 1.3730485689505638, + "grad_norm": 5.7356414794921875, + "learning_rate": 1.0885681125779778e-05, + "loss": 0.038, + "step": 12665 + }, + { + "epoch": 1.3731569817866436, + "grad_norm": 7.913786888122559, + "learning_rate": 1.088495575221239e-05, + "loss": 0.0132, + "step": 12666 + }, + { + "epoch": 1.3732653946227233, + "grad_norm": 3.9444048404693604, + "learning_rate": 1.0884230378645002e-05, + "loss": 0.0475, + "step": 12667 + }, + { + "epoch": 1.373373807458803, + "grad_norm": 3.7046239376068115, + "learning_rate": 1.0883505005077614e-05, + "loss": 0.0225, + "step": 12668 + }, + { + "epoch": 1.3734822202948829, + "grad_norm": 1.5049885511398315, + "learning_rate": 1.088277963151023e-05, + "loss": 0.0332, + "step": 12669 + }, + { + "epoch": 1.3735906331309626, + "grad_norm": 1.1721853017807007, + "learning_rate": 1.0882054257942842e-05, + "loss": 0.0117, + "step": 12670 + }, + { + "epoch": 1.3736990459670424, + "grad_norm": 0.5209397077560425, + "learning_rate": 1.0881328884375455e-05, + "loss": 0.0013, + "step": 12671 + }, + { + "epoch": 1.3738074588031224, + "grad_norm": 5.012429714202881, + "learning_rate": 1.0880603510808068e-05, + "loss": 0.0367, + "step": 12672 + }, + { + "epoch": 1.3739158716392021, + "grad_norm": 1.3886909484863281, + "learning_rate": 1.087987813724068e-05, + "loss": 0.0115, + "step": 12673 + }, + { + "epoch": 1.374024284475282, + "grad_norm": 0.2786993086338043, + "learning_rate": 1.0879152763673294e-05, + "loss": 0.0016, + "step": 12674 + }, + { + "epoch": 1.3741326973113617, + "grad_norm": 1.6654659509658813, + "learning_rate": 1.0878427390105906e-05, + "loss": 0.0111, + "step": 12675 + }, + { + "epoch": 1.3742411101474414, + "grad_norm": 0.6537554264068604, + "learning_rate": 1.0877702016538518e-05, + "loss": 0.0048, + "step": 12676 + }, + { + "epoch": 1.3743495229835212, + "grad_norm": 1.9353983402252197, + "learning_rate": 1.0876976642971132e-05, + "loss": 0.0209, + "step": 12677 + }, + { + "epoch": 1.374457935819601, + "grad_norm": 2.425570487976074, + "learning_rate": 1.0876251269403744e-05, + "loss": 0.0246, + "step": 12678 + }, + { + "epoch": 1.374566348655681, + "grad_norm": 0.8753952980041504, + "learning_rate": 1.0875525895836356e-05, + "loss": 0.006, + "step": 12679 + }, + { + "epoch": 1.3746747614917605, + "grad_norm": 1.9173853397369385, + "learning_rate": 1.0874800522268968e-05, + "loss": 0.0272, + "step": 12680 + }, + { + "epoch": 1.3747831743278405, + "grad_norm": 2.2259693145751953, + "learning_rate": 1.0874075148701582e-05, + "loss": 0.0328, + "step": 12681 + }, + { + "epoch": 1.3748915871639202, + "grad_norm": 1.5098212957382202, + "learning_rate": 1.0873349775134194e-05, + "loss": 0.0164, + "step": 12682 + }, + { + "epoch": 1.375, + "grad_norm": 1.843118667602539, + "learning_rate": 1.0872624401566806e-05, + "loss": 0.04, + "step": 12683 + }, + { + "epoch": 1.3751084128360798, + "grad_norm": 3.8681371212005615, + "learning_rate": 1.087189902799942e-05, + "loss": 0.049, + "step": 12684 + }, + { + "epoch": 1.3752168256721595, + "grad_norm": 1.927850365638733, + "learning_rate": 1.0871173654432032e-05, + "loss": 0.0067, + "step": 12685 + }, + { + "epoch": 1.3753252385082395, + "grad_norm": 0.7607128620147705, + "learning_rate": 1.0870448280864648e-05, + "loss": 0.0087, + "step": 12686 + }, + { + "epoch": 1.375433651344319, + "grad_norm": 5.377935886383057, + "learning_rate": 1.086972290729726e-05, + "loss": 0.0206, + "step": 12687 + }, + { + "epoch": 1.375542064180399, + "grad_norm": 3.645998239517212, + "learning_rate": 1.0868997533729872e-05, + "loss": 0.0553, + "step": 12688 + }, + { + "epoch": 1.3756504770164788, + "grad_norm": 1.7170124053955078, + "learning_rate": 1.0868272160162486e-05, + "loss": 0.0407, + "step": 12689 + }, + { + "epoch": 1.3757588898525586, + "grad_norm": 4.408538818359375, + "learning_rate": 1.0867546786595098e-05, + "loss": 0.0413, + "step": 12690 + }, + { + "epoch": 1.3758673026886383, + "grad_norm": 0.6444125175476074, + "learning_rate": 1.086682141302771e-05, + "loss": 0.0022, + "step": 12691 + }, + { + "epoch": 1.375975715524718, + "grad_norm": 5.396101474761963, + "learning_rate": 1.0866096039460324e-05, + "loss": 0.0755, + "step": 12692 + }, + { + "epoch": 1.3760841283607979, + "grad_norm": 6.270349979400635, + "learning_rate": 1.0865370665892936e-05, + "loss": 0.0111, + "step": 12693 + }, + { + "epoch": 1.3761925411968776, + "grad_norm": 13.584466934204102, + "learning_rate": 1.0864645292325548e-05, + "loss": 0.0339, + "step": 12694 + }, + { + "epoch": 1.3763009540329576, + "grad_norm": 2.124577760696411, + "learning_rate": 1.086391991875816e-05, + "loss": 0.0176, + "step": 12695 + }, + { + "epoch": 1.3764093668690374, + "grad_norm": 3.472637176513672, + "learning_rate": 1.0863194545190774e-05, + "loss": 0.0373, + "step": 12696 + }, + { + "epoch": 1.3765177797051171, + "grad_norm": 0.83498215675354, + "learning_rate": 1.0862469171623386e-05, + "loss": 0.0125, + "step": 12697 + }, + { + "epoch": 1.376626192541197, + "grad_norm": 1.9711858034133911, + "learning_rate": 1.0861743798055999e-05, + "loss": 0.0448, + "step": 12698 + }, + { + "epoch": 1.3767346053772767, + "grad_norm": 1.5044610500335693, + "learning_rate": 1.0861018424488612e-05, + "loss": 0.0206, + "step": 12699 + }, + { + "epoch": 1.3768430182133564, + "grad_norm": 0.3200973570346832, + "learning_rate": 1.0860293050921225e-05, + "loss": 0.0023, + "step": 12700 + }, + { + "epoch": 1.3769514310494362, + "grad_norm": 2.372769355773926, + "learning_rate": 1.0859567677353837e-05, + "loss": 0.0665, + "step": 12701 + }, + { + "epoch": 1.3770598438855162, + "grad_norm": 0.32546547055244446, + "learning_rate": 1.085884230378645e-05, + "loss": 0.0017, + "step": 12702 + }, + { + "epoch": 1.3771682567215957, + "grad_norm": 1.0154736042022705, + "learning_rate": 1.0858116930219064e-05, + "loss": 0.0096, + "step": 12703 + }, + { + "epoch": 1.3772766695576757, + "grad_norm": 7.715486526489258, + "learning_rate": 1.0857391556651678e-05, + "loss": 0.025, + "step": 12704 + }, + { + "epoch": 1.3773850823937555, + "grad_norm": 1.7699943780899048, + "learning_rate": 1.085666618308429e-05, + "loss": 0.0046, + "step": 12705 + }, + { + "epoch": 1.3774934952298352, + "grad_norm": 1.2936104536056519, + "learning_rate": 1.0855940809516903e-05, + "loss": 0.0155, + "step": 12706 + }, + { + "epoch": 1.377601908065915, + "grad_norm": 1.291724443435669, + "learning_rate": 1.0855215435949515e-05, + "loss": 0.0452, + "step": 12707 + }, + { + "epoch": 1.3777103209019947, + "grad_norm": 4.08176851272583, + "learning_rate": 1.0854490062382128e-05, + "loss": 0.0133, + "step": 12708 + }, + { + "epoch": 1.3778187337380745, + "grad_norm": 1.5702829360961914, + "learning_rate": 1.085376468881474e-05, + "loss": 0.0086, + "step": 12709 + }, + { + "epoch": 1.3779271465741543, + "grad_norm": 0.610687792301178, + "learning_rate": 1.0853039315247353e-05, + "loss": 0.0069, + "step": 12710 + }, + { + "epoch": 1.3780355594102343, + "grad_norm": 2.395592451095581, + "learning_rate": 1.0852313941679967e-05, + "loss": 0.0353, + "step": 12711 + }, + { + "epoch": 1.378143972246314, + "grad_norm": 0.6038702130317688, + "learning_rate": 1.0851588568112579e-05, + "loss": 0.0068, + "step": 12712 + }, + { + "epoch": 1.3782523850823938, + "grad_norm": 0.652614176273346, + "learning_rate": 1.0850863194545191e-05, + "loss": 0.0158, + "step": 12713 + }, + { + "epoch": 1.3783607979184735, + "grad_norm": 4.468873023986816, + "learning_rate": 1.0850137820977805e-05, + "loss": 0.0125, + "step": 12714 + }, + { + "epoch": 1.3784692107545533, + "grad_norm": 0.0718424916267395, + "learning_rate": 1.0849412447410417e-05, + "loss": 0.0007, + "step": 12715 + }, + { + "epoch": 1.378577623590633, + "grad_norm": 2.546738386154175, + "learning_rate": 1.0848687073843029e-05, + "loss": 0.0448, + "step": 12716 + }, + { + "epoch": 1.3786860364267128, + "grad_norm": 4.516765594482422, + "learning_rate": 1.0847961700275643e-05, + "loss": 0.02, + "step": 12717 + }, + { + "epoch": 1.3787944492627928, + "grad_norm": 3.889049530029297, + "learning_rate": 1.0847236326708255e-05, + "loss": 0.026, + "step": 12718 + }, + { + "epoch": 1.3789028620988726, + "grad_norm": 14.18101978302002, + "learning_rate": 1.0846510953140867e-05, + "loss": 0.0346, + "step": 12719 + }, + { + "epoch": 1.3790112749349523, + "grad_norm": 3.711709499359131, + "learning_rate": 1.0845785579573483e-05, + "loss": 0.0145, + "step": 12720 + }, + { + "epoch": 1.379119687771032, + "grad_norm": 1.3709423542022705, + "learning_rate": 1.0845060206006095e-05, + "loss": 0.0116, + "step": 12721 + }, + { + "epoch": 1.3792281006071119, + "grad_norm": 6.101949691772461, + "learning_rate": 1.0844334832438707e-05, + "loss": 0.0129, + "step": 12722 + }, + { + "epoch": 1.3793365134431916, + "grad_norm": 2.134406089782715, + "learning_rate": 1.084360945887132e-05, + "loss": 0.0122, + "step": 12723 + }, + { + "epoch": 1.3794449262792714, + "grad_norm": 1.9359017610549927, + "learning_rate": 1.0842884085303933e-05, + "loss": 0.0049, + "step": 12724 + }, + { + "epoch": 1.3795533391153514, + "grad_norm": 4.434771537780762, + "learning_rate": 1.0842158711736545e-05, + "loss": 0.0651, + "step": 12725 + }, + { + "epoch": 1.379661751951431, + "grad_norm": 1.127925992012024, + "learning_rate": 1.0841433338169159e-05, + "loss": 0.0056, + "step": 12726 + }, + { + "epoch": 1.379770164787511, + "grad_norm": 0.8207732439041138, + "learning_rate": 1.0840707964601771e-05, + "loss": 0.0059, + "step": 12727 + }, + { + "epoch": 1.3798785776235907, + "grad_norm": 1.7286900281906128, + "learning_rate": 1.0839982591034383e-05, + "loss": 0.0322, + "step": 12728 + }, + { + "epoch": 1.3799869904596704, + "grad_norm": 3.957892894744873, + "learning_rate": 1.0839257217466997e-05, + "loss": 0.0667, + "step": 12729 + }, + { + "epoch": 1.3800954032957502, + "grad_norm": 8.251474380493164, + "learning_rate": 1.083853184389961e-05, + "loss": 0.0165, + "step": 12730 + }, + { + "epoch": 1.38020381613183, + "grad_norm": 3.372379779815674, + "learning_rate": 1.0837806470332221e-05, + "loss": 0.0226, + "step": 12731 + }, + { + "epoch": 1.3803122289679097, + "grad_norm": 4.176194667816162, + "learning_rate": 1.0837081096764833e-05, + "loss": 0.036, + "step": 12732 + }, + { + "epoch": 1.3804206418039895, + "grad_norm": 0.44494062662124634, + "learning_rate": 1.0836355723197447e-05, + "loss": 0.0037, + "step": 12733 + }, + { + "epoch": 1.3805290546400695, + "grad_norm": 3.8281264305114746, + "learning_rate": 1.083563034963006e-05, + "loss": 0.0428, + "step": 12734 + }, + { + "epoch": 1.3806374674761492, + "grad_norm": 1.6021504402160645, + "learning_rate": 1.0834904976062672e-05, + "loss": 0.023, + "step": 12735 + }, + { + "epoch": 1.380745880312229, + "grad_norm": 0.31781184673309326, + "learning_rate": 1.0834179602495285e-05, + "loss": 0.0027, + "step": 12736 + }, + { + "epoch": 1.3808542931483088, + "grad_norm": 5.846031665802002, + "learning_rate": 1.0833454228927898e-05, + "loss": 0.0092, + "step": 12737 + }, + { + "epoch": 1.3809627059843885, + "grad_norm": 1.5935423374176025, + "learning_rate": 1.0832728855360513e-05, + "loss": 0.0274, + "step": 12738 + }, + { + "epoch": 1.3810711188204683, + "grad_norm": 0.8973013758659363, + "learning_rate": 1.0832003481793125e-05, + "loss": 0.0056, + "step": 12739 + }, + { + "epoch": 1.381179531656548, + "grad_norm": 2.55853009223938, + "learning_rate": 1.0831278108225737e-05, + "loss": 0.055, + "step": 12740 + }, + { + "epoch": 1.381287944492628, + "grad_norm": 0.8339183926582336, + "learning_rate": 1.0830552734658351e-05, + "loss": 0.0046, + "step": 12741 + }, + { + "epoch": 1.3813963573287076, + "grad_norm": 1.215064525604248, + "learning_rate": 1.0829827361090963e-05, + "loss": 0.0136, + "step": 12742 + }, + { + "epoch": 1.3815047701647876, + "grad_norm": 1.894760012626648, + "learning_rate": 1.0829101987523576e-05, + "loss": 0.0079, + "step": 12743 + }, + { + "epoch": 1.3816131830008673, + "grad_norm": 1.8890964984893799, + "learning_rate": 1.082837661395619e-05, + "loss": 0.0266, + "step": 12744 + }, + { + "epoch": 1.381721595836947, + "grad_norm": 3.3056769371032715, + "learning_rate": 1.0827651240388802e-05, + "loss": 0.0426, + "step": 12745 + }, + { + "epoch": 1.3818300086730269, + "grad_norm": 0.1733887791633606, + "learning_rate": 1.0826925866821414e-05, + "loss": 0.002, + "step": 12746 + }, + { + "epoch": 1.3819384215091066, + "grad_norm": 2.009261131286621, + "learning_rate": 1.0826200493254026e-05, + "loss": 0.0287, + "step": 12747 + }, + { + "epoch": 1.3820468343451866, + "grad_norm": 4.883185386657715, + "learning_rate": 1.082547511968664e-05, + "loss": 0.0538, + "step": 12748 + }, + { + "epoch": 1.3821552471812661, + "grad_norm": 0.7971482872962952, + "learning_rate": 1.0824749746119252e-05, + "loss": 0.0091, + "step": 12749 + }, + { + "epoch": 1.3822636600173461, + "grad_norm": 2.972299814224243, + "learning_rate": 1.0824024372551864e-05, + "loss": 0.0401, + "step": 12750 + }, + { + "epoch": 1.382372072853426, + "grad_norm": 1.3061957359313965, + "learning_rate": 1.0823298998984478e-05, + "loss": 0.0099, + "step": 12751 + }, + { + "epoch": 1.3824804856895057, + "grad_norm": 0.8634995818138123, + "learning_rate": 1.082257362541709e-05, + "loss": 0.0052, + "step": 12752 + }, + { + "epoch": 1.3825888985255854, + "grad_norm": 2.7707982063293457, + "learning_rate": 1.0821848251849702e-05, + "loss": 0.0203, + "step": 12753 + }, + { + "epoch": 1.3826973113616652, + "grad_norm": 1.383506178855896, + "learning_rate": 1.0821122878282316e-05, + "loss": 0.0285, + "step": 12754 + }, + { + "epoch": 1.382805724197745, + "grad_norm": 1.3393349647521973, + "learning_rate": 1.082039750471493e-05, + "loss": 0.0072, + "step": 12755 + }, + { + "epoch": 1.3829141370338247, + "grad_norm": 0.731289803981781, + "learning_rate": 1.0819672131147544e-05, + "loss": 0.0056, + "step": 12756 + }, + { + "epoch": 1.3830225498699047, + "grad_norm": 1.5042623281478882, + "learning_rate": 1.0818946757580156e-05, + "loss": 0.0444, + "step": 12757 + }, + { + "epoch": 1.3831309627059845, + "grad_norm": 3.1403701305389404, + "learning_rate": 1.0818221384012768e-05, + "loss": 0.0307, + "step": 12758 + }, + { + "epoch": 1.3832393755420642, + "grad_norm": 1.316498875617981, + "learning_rate": 1.081749601044538e-05, + "loss": 0.0418, + "step": 12759 + }, + { + "epoch": 1.383347788378144, + "grad_norm": 4.923954486846924, + "learning_rate": 1.0816770636877994e-05, + "loss": 0.0362, + "step": 12760 + }, + { + "epoch": 1.3834562012142237, + "grad_norm": 1.0482637882232666, + "learning_rate": 1.0816045263310606e-05, + "loss": 0.0093, + "step": 12761 + }, + { + "epoch": 1.3835646140503035, + "grad_norm": 1.3485935926437378, + "learning_rate": 1.0815319889743218e-05, + "loss": 0.0069, + "step": 12762 + }, + { + "epoch": 1.3836730268863833, + "grad_norm": 0.4447047710418701, + "learning_rate": 1.0814594516175832e-05, + "loss": 0.0059, + "step": 12763 + }, + { + "epoch": 1.3837814397224633, + "grad_norm": 2.3348069190979004, + "learning_rate": 1.0813869142608444e-05, + "loss": 0.0116, + "step": 12764 + }, + { + "epoch": 1.3838898525585428, + "grad_norm": 0.5135981440544128, + "learning_rate": 1.0813143769041056e-05, + "loss": 0.0044, + "step": 12765 + }, + { + "epoch": 1.3839982653946228, + "grad_norm": 1.944605827331543, + "learning_rate": 1.081241839547367e-05, + "loss": 0.0326, + "step": 12766 + }, + { + "epoch": 1.3841066782307025, + "grad_norm": 1.6740496158599854, + "learning_rate": 1.0811693021906282e-05, + "loss": 0.0342, + "step": 12767 + }, + { + "epoch": 1.3842150910667823, + "grad_norm": 5.204766750335693, + "learning_rate": 1.0810967648338894e-05, + "loss": 0.0199, + "step": 12768 + }, + { + "epoch": 1.384323503902862, + "grad_norm": 0.29862603545188904, + "learning_rate": 1.0810242274771508e-05, + "loss": 0.0038, + "step": 12769 + }, + { + "epoch": 1.3844319167389418, + "grad_norm": 3.141665458679199, + "learning_rate": 1.080951690120412e-05, + "loss": 0.0327, + "step": 12770 + }, + { + "epoch": 1.3845403295750216, + "grad_norm": 5.0076904296875, + "learning_rate": 1.0808791527636732e-05, + "loss": 0.0707, + "step": 12771 + }, + { + "epoch": 1.3846487424111014, + "grad_norm": 2.969111442565918, + "learning_rate": 1.0808066154069348e-05, + "loss": 0.0206, + "step": 12772 + }, + { + "epoch": 1.3847571552471813, + "grad_norm": 4.754726886749268, + "learning_rate": 1.080734078050196e-05, + "loss": 0.0366, + "step": 12773 + }, + { + "epoch": 1.3848655680832611, + "grad_norm": 0.8793985843658447, + "learning_rate": 1.0806615406934572e-05, + "loss": 0.0043, + "step": 12774 + }, + { + "epoch": 1.3849739809193409, + "grad_norm": 4.201662540435791, + "learning_rate": 1.0805890033367186e-05, + "loss": 0.023, + "step": 12775 + }, + { + "epoch": 1.3850823937554206, + "grad_norm": 1.0333843231201172, + "learning_rate": 1.0805164659799798e-05, + "loss": 0.0097, + "step": 12776 + }, + { + "epoch": 1.3851908065915004, + "grad_norm": 4.821164131164551, + "learning_rate": 1.080443928623241e-05, + "loss": 0.0565, + "step": 12777 + }, + { + "epoch": 1.3852992194275802, + "grad_norm": 0.47376006841659546, + "learning_rate": 1.0803713912665024e-05, + "loss": 0.0061, + "step": 12778 + }, + { + "epoch": 1.38540763226366, + "grad_norm": 2.2732346057891846, + "learning_rate": 1.0802988539097636e-05, + "loss": 0.0277, + "step": 12779 + }, + { + "epoch": 1.38551604509974, + "grad_norm": 1.5333590507507324, + "learning_rate": 1.0802263165530249e-05, + "loss": 0.0092, + "step": 12780 + }, + { + "epoch": 1.3856244579358197, + "grad_norm": 2.2523598670959473, + "learning_rate": 1.0801537791962862e-05, + "loss": 0.0535, + "step": 12781 + }, + { + "epoch": 1.3857328707718994, + "grad_norm": 3.059346914291382, + "learning_rate": 1.0800812418395475e-05, + "loss": 0.0173, + "step": 12782 + }, + { + "epoch": 1.3858412836079792, + "grad_norm": 3.6289658546447754, + "learning_rate": 1.0800087044828087e-05, + "loss": 0.0296, + "step": 12783 + }, + { + "epoch": 1.385949696444059, + "grad_norm": 1.0314507484436035, + "learning_rate": 1.07993616712607e-05, + "loss": 0.0213, + "step": 12784 + }, + { + "epoch": 1.3860581092801387, + "grad_norm": 1.370330810546875, + "learning_rate": 1.0798636297693313e-05, + "loss": 0.021, + "step": 12785 + }, + { + "epoch": 1.3861665221162185, + "grad_norm": 1.2148183584213257, + "learning_rate": 1.0797910924125925e-05, + "loss": 0.0267, + "step": 12786 + }, + { + "epoch": 1.3862749349522985, + "grad_norm": 0.24762988090515137, + "learning_rate": 1.0797185550558537e-05, + "loss": 0.0031, + "step": 12787 + }, + { + "epoch": 1.386383347788378, + "grad_norm": 0.22986868023872375, + "learning_rate": 1.079646017699115e-05, + "loss": 0.0015, + "step": 12788 + }, + { + "epoch": 1.386491760624458, + "grad_norm": 1.1030888557434082, + "learning_rate": 1.0795734803423765e-05, + "loss": 0.0263, + "step": 12789 + }, + { + "epoch": 1.3866001734605378, + "grad_norm": 4.341595649719238, + "learning_rate": 1.0795009429856378e-05, + "loss": 0.0459, + "step": 12790 + }, + { + "epoch": 1.3867085862966175, + "grad_norm": 2.353782892227173, + "learning_rate": 1.079428405628899e-05, + "loss": 0.0174, + "step": 12791 + }, + { + "epoch": 1.3868169991326973, + "grad_norm": 6.4222092628479, + "learning_rate": 1.0793558682721603e-05, + "loss": 0.0364, + "step": 12792 + }, + { + "epoch": 1.386925411968777, + "grad_norm": 2.129676580429077, + "learning_rate": 1.0792833309154217e-05, + "loss": 0.0759, + "step": 12793 + }, + { + "epoch": 1.3870338248048568, + "grad_norm": 0.6958866119384766, + "learning_rate": 1.0792107935586829e-05, + "loss": 0.0056, + "step": 12794 + }, + { + "epoch": 1.3871422376409366, + "grad_norm": 17.537918090820312, + "learning_rate": 1.0791382562019441e-05, + "loss": 0.07, + "step": 12795 + }, + { + "epoch": 1.3872506504770166, + "grad_norm": 5.633789539337158, + "learning_rate": 1.0790657188452055e-05, + "loss": 0.0212, + "step": 12796 + }, + { + "epoch": 1.3873590633130963, + "grad_norm": 3.5309805870056152, + "learning_rate": 1.0789931814884667e-05, + "loss": 0.0172, + "step": 12797 + }, + { + "epoch": 1.387467476149176, + "grad_norm": 0.3388778269290924, + "learning_rate": 1.0789206441317279e-05, + "loss": 0.0057, + "step": 12798 + }, + { + "epoch": 1.3875758889852559, + "grad_norm": 2.6688947677612305, + "learning_rate": 1.0788481067749891e-05, + "loss": 0.0448, + "step": 12799 + }, + { + "epoch": 1.3876843018213356, + "grad_norm": 2.2247960567474365, + "learning_rate": 1.0787755694182505e-05, + "loss": 0.0252, + "step": 12800 + }, + { + "epoch": 1.3877927146574154, + "grad_norm": 1.6283323764801025, + "learning_rate": 1.0787030320615117e-05, + "loss": 0.0121, + "step": 12801 + }, + { + "epoch": 1.3879011274934951, + "grad_norm": 1.5362083911895752, + "learning_rate": 1.078630494704773e-05, + "loss": 0.0328, + "step": 12802 + }, + { + "epoch": 1.3880095403295751, + "grad_norm": 0.924293041229248, + "learning_rate": 1.0785579573480343e-05, + "loss": 0.009, + "step": 12803 + }, + { + "epoch": 1.388117953165655, + "grad_norm": 1.2777436971664429, + "learning_rate": 1.0784854199912955e-05, + "loss": 0.0357, + "step": 12804 + }, + { + "epoch": 1.3882263660017347, + "grad_norm": 4.414102077484131, + "learning_rate": 1.0784128826345567e-05, + "loss": 0.07, + "step": 12805 + }, + { + "epoch": 1.3883347788378144, + "grad_norm": 0.5407480597496033, + "learning_rate": 1.0783403452778181e-05, + "loss": 0.0038, + "step": 12806 + }, + { + "epoch": 1.3884431916738942, + "grad_norm": 0.4922523498535156, + "learning_rate": 1.0782678079210795e-05, + "loss": 0.0051, + "step": 12807 + }, + { + "epoch": 1.388551604509974, + "grad_norm": 2.6492245197296143, + "learning_rate": 1.0781952705643409e-05, + "loss": 0.0301, + "step": 12808 + }, + { + "epoch": 1.3886600173460537, + "grad_norm": 4.015470027923584, + "learning_rate": 1.0781227332076021e-05, + "loss": 0.0363, + "step": 12809 + }, + { + "epoch": 1.3887684301821337, + "grad_norm": 1.6455143690109253, + "learning_rate": 1.0780501958508633e-05, + "loss": 0.03, + "step": 12810 + }, + { + "epoch": 1.3888768430182132, + "grad_norm": 0.4264014959335327, + "learning_rate": 1.0779776584941245e-05, + "loss": 0.0067, + "step": 12811 + }, + { + "epoch": 1.3889852558542932, + "grad_norm": 1.3546892404556274, + "learning_rate": 1.077905121137386e-05, + "loss": 0.0229, + "step": 12812 + }, + { + "epoch": 1.389093668690373, + "grad_norm": 0.4581690728664398, + "learning_rate": 1.0778325837806471e-05, + "loss": 0.0081, + "step": 12813 + }, + { + "epoch": 1.3892020815264527, + "grad_norm": 0.7767194509506226, + "learning_rate": 1.0777600464239083e-05, + "loss": 0.0222, + "step": 12814 + }, + { + "epoch": 1.3893104943625325, + "grad_norm": 1.5763887166976929, + "learning_rate": 1.0776875090671697e-05, + "loss": 0.0035, + "step": 12815 + }, + { + "epoch": 1.3894189071986123, + "grad_norm": 1.5336447954177856, + "learning_rate": 1.077614971710431e-05, + "loss": 0.0153, + "step": 12816 + }, + { + "epoch": 1.389527320034692, + "grad_norm": 3.199845790863037, + "learning_rate": 1.0775424343536922e-05, + "loss": 0.0477, + "step": 12817 + }, + { + "epoch": 1.3896357328707718, + "grad_norm": 3.0580174922943115, + "learning_rate": 1.0774698969969535e-05, + "loss": 0.0697, + "step": 12818 + }, + { + "epoch": 1.3897441457068518, + "grad_norm": 1.481196641921997, + "learning_rate": 1.0773973596402148e-05, + "loss": 0.0412, + "step": 12819 + }, + { + "epoch": 1.3898525585429315, + "grad_norm": 4.491064548492432, + "learning_rate": 1.077324822283476e-05, + "loss": 0.0184, + "step": 12820 + }, + { + "epoch": 1.3899609713790113, + "grad_norm": 1.0874611139297485, + "learning_rate": 1.0772522849267374e-05, + "loss": 0.0156, + "step": 12821 + }, + { + "epoch": 1.390069384215091, + "grad_norm": 3.769745111465454, + "learning_rate": 1.0771797475699986e-05, + "loss": 0.0169, + "step": 12822 + }, + { + "epoch": 1.3901777970511708, + "grad_norm": 4.8122453689575195, + "learning_rate": 1.0771072102132598e-05, + "loss": 0.0164, + "step": 12823 + }, + { + "epoch": 1.3902862098872506, + "grad_norm": 0.17664937674999237, + "learning_rate": 1.0770346728565213e-05, + "loss": 0.0017, + "step": 12824 + }, + { + "epoch": 1.3903946227233304, + "grad_norm": 0.9349498748779297, + "learning_rate": 1.0769621354997826e-05, + "loss": 0.0046, + "step": 12825 + }, + { + "epoch": 1.3905030355594103, + "grad_norm": 8.568536758422852, + "learning_rate": 1.0768895981430438e-05, + "loss": 0.0257, + "step": 12826 + }, + { + "epoch": 1.39061144839549, + "grad_norm": 4.0128045082092285, + "learning_rate": 1.0768170607863051e-05, + "loss": 0.008, + "step": 12827 + }, + { + "epoch": 1.3907198612315699, + "grad_norm": 3.023186683654785, + "learning_rate": 1.0767445234295664e-05, + "loss": 0.0134, + "step": 12828 + }, + { + "epoch": 1.3908282740676496, + "grad_norm": 0.12661610543727875, + "learning_rate": 1.0766719860728276e-05, + "loss": 0.0012, + "step": 12829 + }, + { + "epoch": 1.3909366869037294, + "grad_norm": 0.13537676632404327, + "learning_rate": 1.076599448716089e-05, + "loss": 0.0015, + "step": 12830 + }, + { + "epoch": 1.3910450997398092, + "grad_norm": 3.241459608078003, + "learning_rate": 1.0765269113593502e-05, + "loss": 0.0076, + "step": 12831 + }, + { + "epoch": 1.391153512575889, + "grad_norm": 1.7418986558914185, + "learning_rate": 1.0764543740026114e-05, + "loss": 0.0234, + "step": 12832 + }, + { + "epoch": 1.391261925411969, + "grad_norm": 3.2490034103393555, + "learning_rate": 1.0763818366458728e-05, + "loss": 0.0261, + "step": 12833 + }, + { + "epoch": 1.3913703382480485, + "grad_norm": 0.28426656126976013, + "learning_rate": 1.076309299289134e-05, + "loss": 0.0031, + "step": 12834 + }, + { + "epoch": 1.3914787510841284, + "grad_norm": 3.017298936843872, + "learning_rate": 1.0762367619323952e-05, + "loss": 0.0313, + "step": 12835 + }, + { + "epoch": 1.3915871639202082, + "grad_norm": 1.4019125699996948, + "learning_rate": 1.0761642245756566e-05, + "loss": 0.0243, + "step": 12836 + }, + { + "epoch": 1.391695576756288, + "grad_norm": 1.7702810764312744, + "learning_rate": 1.0760916872189178e-05, + "loss": 0.0192, + "step": 12837 + }, + { + "epoch": 1.3918039895923677, + "grad_norm": 9.35237979888916, + "learning_rate": 1.076019149862179e-05, + "loss": 0.0264, + "step": 12838 + }, + { + "epoch": 1.3919124024284475, + "grad_norm": 3.49100923538208, + "learning_rate": 1.0759466125054402e-05, + "loss": 0.04, + "step": 12839 + }, + { + "epoch": 1.3920208152645273, + "grad_norm": 0.9793646335601807, + "learning_rate": 1.0758740751487016e-05, + "loss": 0.0094, + "step": 12840 + }, + { + "epoch": 1.392129228100607, + "grad_norm": 0.8566296696662903, + "learning_rate": 1.075801537791963e-05, + "loss": 0.003, + "step": 12841 + }, + { + "epoch": 1.392237640936687, + "grad_norm": 2.1965930461883545, + "learning_rate": 1.0757290004352244e-05, + "loss": 0.0411, + "step": 12842 + }, + { + "epoch": 1.3923460537727668, + "grad_norm": 3.5039498805999756, + "learning_rate": 1.0756564630784856e-05, + "loss": 0.0302, + "step": 12843 + }, + { + "epoch": 1.3924544666088465, + "grad_norm": 5.494017601013184, + "learning_rate": 1.0755839257217468e-05, + "loss": 0.0251, + "step": 12844 + }, + { + "epoch": 1.3925628794449263, + "grad_norm": 8.61113166809082, + "learning_rate": 1.0755113883650082e-05, + "loss": 0.0612, + "step": 12845 + }, + { + "epoch": 1.392671292281006, + "grad_norm": 1.0346226692199707, + "learning_rate": 1.0754388510082694e-05, + "loss": 0.0299, + "step": 12846 + }, + { + "epoch": 1.3927797051170858, + "grad_norm": 1.1056392192840576, + "learning_rate": 1.0753663136515306e-05, + "loss": 0.0048, + "step": 12847 + }, + { + "epoch": 1.3928881179531656, + "grad_norm": 2.7388243675231934, + "learning_rate": 1.075293776294792e-05, + "loss": 0.0177, + "step": 12848 + }, + { + "epoch": 1.3929965307892456, + "grad_norm": 2.791400671005249, + "learning_rate": 1.0752212389380532e-05, + "loss": 0.0182, + "step": 12849 + }, + { + "epoch": 1.393104943625325, + "grad_norm": 1.494019627571106, + "learning_rate": 1.0751487015813144e-05, + "loss": 0.0101, + "step": 12850 + }, + { + "epoch": 1.393213356461405, + "grad_norm": 0.9105382561683655, + "learning_rate": 1.0750761642245756e-05, + "loss": 0.0203, + "step": 12851 + }, + { + "epoch": 1.3933217692974849, + "grad_norm": 0.31667622923851013, + "learning_rate": 1.075003626867837e-05, + "loss": 0.0038, + "step": 12852 + }, + { + "epoch": 1.3934301821335646, + "grad_norm": 4.75027322769165, + "learning_rate": 1.0749310895110982e-05, + "loss": 0.0078, + "step": 12853 + }, + { + "epoch": 1.3935385949696444, + "grad_norm": 2.2353570461273193, + "learning_rate": 1.0748585521543595e-05, + "loss": 0.0311, + "step": 12854 + }, + { + "epoch": 1.3936470078057241, + "grad_norm": 2.9218525886535645, + "learning_rate": 1.0747860147976208e-05, + "loss": 0.0191, + "step": 12855 + }, + { + "epoch": 1.393755420641804, + "grad_norm": 0.5753218531608582, + "learning_rate": 1.074713477440882e-05, + "loss": 0.0089, + "step": 12856 + }, + { + "epoch": 1.3938638334778837, + "grad_norm": 2.6046698093414307, + "learning_rate": 1.0746409400841433e-05, + "loss": 0.0273, + "step": 12857 + }, + { + "epoch": 1.3939722463139637, + "grad_norm": 4.934001445770264, + "learning_rate": 1.0745684027274047e-05, + "loss": 0.0992, + "step": 12858 + }, + { + "epoch": 1.3940806591500434, + "grad_norm": 3.8221583366394043, + "learning_rate": 1.074495865370666e-05, + "loss": 0.0206, + "step": 12859 + }, + { + "epoch": 1.3941890719861232, + "grad_norm": 0.9652352929115295, + "learning_rate": 1.0744233280139274e-05, + "loss": 0.0209, + "step": 12860 + }, + { + "epoch": 1.394297484822203, + "grad_norm": 0.808538019657135, + "learning_rate": 1.0743507906571886e-05, + "loss": 0.0129, + "step": 12861 + }, + { + "epoch": 1.3944058976582827, + "grad_norm": 1.3942681550979614, + "learning_rate": 1.0742782533004499e-05, + "loss": 0.0081, + "step": 12862 + }, + { + "epoch": 1.3945143104943625, + "grad_norm": 3.082242012023926, + "learning_rate": 1.074205715943711e-05, + "loss": 0.0679, + "step": 12863 + }, + { + "epoch": 1.3946227233304422, + "grad_norm": 0.7042055726051331, + "learning_rate": 1.0741331785869725e-05, + "loss": 0.0086, + "step": 12864 + }, + { + "epoch": 1.3947311361665222, + "grad_norm": 4.694452285766602, + "learning_rate": 1.0740606412302337e-05, + "loss": 0.0167, + "step": 12865 + }, + { + "epoch": 1.394839549002602, + "grad_norm": 1.1401594877243042, + "learning_rate": 1.0739881038734949e-05, + "loss": 0.0391, + "step": 12866 + }, + { + "epoch": 1.3949479618386817, + "grad_norm": 6.355288028717041, + "learning_rate": 1.0739155665167563e-05, + "loss": 0.0395, + "step": 12867 + }, + { + "epoch": 1.3950563746747615, + "grad_norm": 1.2677066326141357, + "learning_rate": 1.0738430291600175e-05, + "loss": 0.0114, + "step": 12868 + }, + { + "epoch": 1.3951647875108413, + "grad_norm": 5.0422186851501465, + "learning_rate": 1.0737704918032787e-05, + "loss": 0.0337, + "step": 12869 + }, + { + "epoch": 1.395273200346921, + "grad_norm": 1.2241634130477905, + "learning_rate": 1.07369795444654e-05, + "loss": 0.0248, + "step": 12870 + }, + { + "epoch": 1.3953816131830008, + "grad_norm": 3.3976874351501465, + "learning_rate": 1.0736254170898013e-05, + "loss": 0.0199, + "step": 12871 + }, + { + "epoch": 1.3954900260190808, + "grad_norm": 2.4752755165100098, + "learning_rate": 1.0735528797330625e-05, + "loss": 0.0084, + "step": 12872 + }, + { + "epoch": 1.3955984388551603, + "grad_norm": 0.43160057067871094, + "learning_rate": 1.0734803423763239e-05, + "loss": 0.0048, + "step": 12873 + }, + { + "epoch": 1.3957068516912403, + "grad_norm": 0.2998785972595215, + "learning_rate": 1.0734078050195851e-05, + "loss": 0.0051, + "step": 12874 + }, + { + "epoch": 1.39581526452732, + "grad_norm": 4.389449119567871, + "learning_rate": 1.0733352676628463e-05, + "loss": 0.0119, + "step": 12875 + }, + { + "epoch": 1.3959236773633998, + "grad_norm": 0.30808204412460327, + "learning_rate": 1.0732627303061079e-05, + "loss": 0.0025, + "step": 12876 + }, + { + "epoch": 1.3960320901994796, + "grad_norm": 1.7494529485702515, + "learning_rate": 1.073190192949369e-05, + "loss": 0.0176, + "step": 12877 + }, + { + "epoch": 1.3961405030355594, + "grad_norm": 1.3489819765090942, + "learning_rate": 1.0731176555926303e-05, + "loss": 0.0227, + "step": 12878 + }, + { + "epoch": 1.3962489158716391, + "grad_norm": 3.458550214767456, + "learning_rate": 1.0730451182358917e-05, + "loss": 0.0179, + "step": 12879 + }, + { + "epoch": 1.396357328707719, + "grad_norm": 0.8699920177459717, + "learning_rate": 1.0729725808791529e-05, + "loss": 0.0055, + "step": 12880 + }, + { + "epoch": 1.3964657415437989, + "grad_norm": 1.0720373392105103, + "learning_rate": 1.0729000435224141e-05, + "loss": 0.0082, + "step": 12881 + }, + { + "epoch": 1.3965741543798786, + "grad_norm": 1.5605696439743042, + "learning_rate": 1.0728275061656755e-05, + "loss": 0.018, + "step": 12882 + }, + { + "epoch": 1.3966825672159584, + "grad_norm": 0.4888755977153778, + "learning_rate": 1.0727549688089367e-05, + "loss": 0.0026, + "step": 12883 + }, + { + "epoch": 1.3967909800520382, + "grad_norm": 0.565929651260376, + "learning_rate": 1.072682431452198e-05, + "loss": 0.0029, + "step": 12884 + }, + { + "epoch": 1.396899392888118, + "grad_norm": 1.6358314752578735, + "learning_rate": 1.0726098940954593e-05, + "loss": 0.0315, + "step": 12885 + }, + { + "epoch": 1.3970078057241977, + "grad_norm": 0.9710747003555298, + "learning_rate": 1.0725373567387205e-05, + "loss": 0.0086, + "step": 12886 + }, + { + "epoch": 1.3971162185602775, + "grad_norm": 1.3258717060089111, + "learning_rate": 1.0724648193819817e-05, + "loss": 0.018, + "step": 12887 + }, + { + "epoch": 1.3972246313963574, + "grad_norm": 2.227959156036377, + "learning_rate": 1.0723922820252431e-05, + "loss": 0.0219, + "step": 12888 + }, + { + "epoch": 1.397333044232437, + "grad_norm": 2.9743640422821045, + "learning_rate": 1.0723197446685043e-05, + "loss": 0.0546, + "step": 12889 + }, + { + "epoch": 1.397441457068517, + "grad_norm": 1.3328113555908203, + "learning_rate": 1.0722472073117655e-05, + "loss": 0.023, + "step": 12890 + }, + { + "epoch": 1.3975498699045967, + "grad_norm": 0.497917115688324, + "learning_rate": 1.0721746699550268e-05, + "loss": 0.0041, + "step": 12891 + }, + { + "epoch": 1.3976582827406765, + "grad_norm": 0.4952050745487213, + "learning_rate": 1.0721021325982881e-05, + "loss": 0.0036, + "step": 12892 + }, + { + "epoch": 1.3977666955767563, + "grad_norm": 2.3076484203338623, + "learning_rate": 1.0720295952415495e-05, + "loss": 0.0077, + "step": 12893 + }, + { + "epoch": 1.397875108412836, + "grad_norm": 1.3116739988327026, + "learning_rate": 1.0719570578848109e-05, + "loss": 0.0044, + "step": 12894 + }, + { + "epoch": 1.397983521248916, + "grad_norm": 0.759070634841919, + "learning_rate": 1.0718845205280721e-05, + "loss": 0.0049, + "step": 12895 + }, + { + "epoch": 1.3980919340849955, + "grad_norm": 5.456062316894531, + "learning_rate": 1.0718119831713333e-05, + "loss": 0.0366, + "step": 12896 + }, + { + "epoch": 1.3982003469210755, + "grad_norm": 0.14363832771778107, + "learning_rate": 1.0717394458145947e-05, + "loss": 0.0016, + "step": 12897 + }, + { + "epoch": 1.3983087597571553, + "grad_norm": 4.481915473937988, + "learning_rate": 1.071666908457856e-05, + "loss": 0.0266, + "step": 12898 + }, + { + "epoch": 1.398417172593235, + "grad_norm": 0.7970936894416809, + "learning_rate": 1.0715943711011172e-05, + "loss": 0.0075, + "step": 12899 + }, + { + "epoch": 1.3985255854293148, + "grad_norm": 1.769010305404663, + "learning_rate": 1.0715218337443785e-05, + "loss": 0.0162, + "step": 12900 + }, + { + "epoch": 1.3986339982653946, + "grad_norm": 3.1678357124328613, + "learning_rate": 1.0714492963876398e-05, + "loss": 0.0241, + "step": 12901 + }, + { + "epoch": 1.3987424111014743, + "grad_norm": 0.5274682641029358, + "learning_rate": 1.071376759030901e-05, + "loss": 0.0036, + "step": 12902 + }, + { + "epoch": 1.398850823937554, + "grad_norm": 0.22540824115276337, + "learning_rate": 1.0713042216741622e-05, + "loss": 0.0013, + "step": 12903 + }, + { + "epoch": 1.398959236773634, + "grad_norm": 2.231076717376709, + "learning_rate": 1.0712316843174236e-05, + "loss": 0.0234, + "step": 12904 + }, + { + "epoch": 1.3990676496097139, + "grad_norm": 0.15937021374702454, + "learning_rate": 1.0711591469606848e-05, + "loss": 0.001, + "step": 12905 + }, + { + "epoch": 1.3991760624457936, + "grad_norm": 1.835520625114441, + "learning_rate": 1.071086609603946e-05, + "loss": 0.0189, + "step": 12906 + }, + { + "epoch": 1.3992844752818734, + "grad_norm": 3.7292568683624268, + "learning_rate": 1.0710140722472074e-05, + "loss": 0.0745, + "step": 12907 + }, + { + "epoch": 1.3993928881179531, + "grad_norm": 3.254782199859619, + "learning_rate": 1.0709415348904686e-05, + "loss": 0.0167, + "step": 12908 + }, + { + "epoch": 1.399501300954033, + "grad_norm": 1.5339140892028809, + "learning_rate": 1.0708689975337298e-05, + "loss": 0.0478, + "step": 12909 + }, + { + "epoch": 1.3996097137901127, + "grad_norm": 2.8734703063964844, + "learning_rate": 1.0707964601769914e-05, + "loss": 0.0197, + "step": 12910 + }, + { + "epoch": 1.3997181266261927, + "grad_norm": 2.572202682495117, + "learning_rate": 1.0707239228202526e-05, + "loss": 0.0577, + "step": 12911 + }, + { + "epoch": 1.3998265394622722, + "grad_norm": 2.1862752437591553, + "learning_rate": 1.070651385463514e-05, + "loss": 0.0138, + "step": 12912 + }, + { + "epoch": 1.3999349522983522, + "grad_norm": 3.4221031665802, + "learning_rate": 1.0705788481067752e-05, + "loss": 0.0445, + "step": 12913 + }, + { + "epoch": 1.400043365134432, + "grad_norm": 2.693599224090576, + "learning_rate": 1.0705063107500364e-05, + "loss": 0.0483, + "step": 12914 + }, + { + "epoch": 1.4001517779705117, + "grad_norm": 5.034224987030029, + "learning_rate": 1.0704337733932976e-05, + "loss": 0.0376, + "step": 12915 + }, + { + "epoch": 1.4002601908065915, + "grad_norm": 5.141780853271484, + "learning_rate": 1.070361236036559e-05, + "loss": 0.0491, + "step": 12916 + }, + { + "epoch": 1.4003686036426712, + "grad_norm": 0.506639301776886, + "learning_rate": 1.0702886986798202e-05, + "loss": 0.0036, + "step": 12917 + }, + { + "epoch": 1.400477016478751, + "grad_norm": 0.7929856181144714, + "learning_rate": 1.0702161613230814e-05, + "loss": 0.0436, + "step": 12918 + }, + { + "epoch": 1.4005854293148308, + "grad_norm": 0.2785893380641937, + "learning_rate": 1.0701436239663428e-05, + "loss": 0.0024, + "step": 12919 + }, + { + "epoch": 1.4006938421509108, + "grad_norm": 6.657963275909424, + "learning_rate": 1.070071086609604e-05, + "loss": 0.0325, + "step": 12920 + }, + { + "epoch": 1.4008022549869905, + "grad_norm": 1.3114210367202759, + "learning_rate": 1.0699985492528652e-05, + "loss": 0.0352, + "step": 12921 + }, + { + "epoch": 1.4009106678230703, + "grad_norm": 2.104513168334961, + "learning_rate": 1.0699260118961266e-05, + "loss": 0.0144, + "step": 12922 + }, + { + "epoch": 1.40101908065915, + "grad_norm": 1.0237332582473755, + "learning_rate": 1.0698534745393878e-05, + "loss": 0.0041, + "step": 12923 + }, + { + "epoch": 1.4011274934952298, + "grad_norm": 6.176941871643066, + "learning_rate": 1.069780937182649e-05, + "loss": 0.0357, + "step": 12924 + }, + { + "epoch": 1.4012359063313096, + "grad_norm": 2.2525627613067627, + "learning_rate": 1.0697083998259104e-05, + "loss": 0.0336, + "step": 12925 + }, + { + "epoch": 1.4013443191673893, + "grad_norm": 2.3854777812957764, + "learning_rate": 1.0696358624691716e-05, + "loss": 0.0358, + "step": 12926 + }, + { + "epoch": 1.4014527320034693, + "grad_norm": 5.494147777557373, + "learning_rate": 1.0695633251124329e-05, + "loss": 0.0513, + "step": 12927 + }, + { + "epoch": 1.401561144839549, + "grad_norm": 1.7758747339248657, + "learning_rate": 1.0694907877556944e-05, + "loss": 0.0111, + "step": 12928 + }, + { + "epoch": 1.4016695576756288, + "grad_norm": 5.529743671417236, + "learning_rate": 1.0694182503989556e-05, + "loss": 0.0277, + "step": 12929 + }, + { + "epoch": 1.4017779705117086, + "grad_norm": 2.6311447620391846, + "learning_rate": 1.0693457130422168e-05, + "loss": 0.0595, + "step": 12930 + }, + { + "epoch": 1.4018863833477884, + "grad_norm": 2.714804172515869, + "learning_rate": 1.0692731756854782e-05, + "loss": 0.059, + "step": 12931 + }, + { + "epoch": 1.4019947961838681, + "grad_norm": 2.436175584793091, + "learning_rate": 1.0692006383287394e-05, + "loss": 0.0282, + "step": 12932 + }, + { + "epoch": 1.402103209019948, + "grad_norm": 1.928215742111206, + "learning_rate": 1.0691281009720006e-05, + "loss": 0.0097, + "step": 12933 + }, + { + "epoch": 1.4022116218560279, + "grad_norm": 5.629776954650879, + "learning_rate": 1.069055563615262e-05, + "loss": 0.0524, + "step": 12934 + }, + { + "epoch": 1.4023200346921074, + "grad_norm": 1.0685133934020996, + "learning_rate": 1.0689830262585232e-05, + "loss": 0.0109, + "step": 12935 + }, + { + "epoch": 1.4024284475281874, + "grad_norm": 3.4914650917053223, + "learning_rate": 1.0689104889017845e-05, + "loss": 0.0329, + "step": 12936 + }, + { + "epoch": 1.4025368603642672, + "grad_norm": 0.8254522681236267, + "learning_rate": 1.0688379515450458e-05, + "loss": 0.0054, + "step": 12937 + }, + { + "epoch": 1.402645273200347, + "grad_norm": 0.9787054657936096, + "learning_rate": 1.068765414188307e-05, + "loss": 0.0066, + "step": 12938 + }, + { + "epoch": 1.4027536860364267, + "grad_norm": 1.4262588024139404, + "learning_rate": 1.0686928768315683e-05, + "loss": 0.0323, + "step": 12939 + }, + { + "epoch": 1.4028620988725065, + "grad_norm": 0.5398805141448975, + "learning_rate": 1.0686203394748297e-05, + "loss": 0.0035, + "step": 12940 + }, + { + "epoch": 1.4029705117085862, + "grad_norm": 14.34963607788086, + "learning_rate": 1.0685478021180909e-05, + "loss": 0.0464, + "step": 12941 + }, + { + "epoch": 1.403078924544666, + "grad_norm": 11.057358741760254, + "learning_rate": 1.068475264761352e-05, + "loss": 0.0339, + "step": 12942 + }, + { + "epoch": 1.403187337380746, + "grad_norm": 3.144440174102783, + "learning_rate": 1.0684027274046133e-05, + "loss": 0.0124, + "step": 12943 + }, + { + "epoch": 1.4032957502168257, + "grad_norm": 0.8507645130157471, + "learning_rate": 1.0683301900478747e-05, + "loss": 0.0077, + "step": 12944 + }, + { + "epoch": 1.4034041630529055, + "grad_norm": 1.268986701965332, + "learning_rate": 1.068257652691136e-05, + "loss": 0.0054, + "step": 12945 + }, + { + "epoch": 1.4035125758889853, + "grad_norm": 1.157460331916809, + "learning_rate": 1.0681851153343974e-05, + "loss": 0.0102, + "step": 12946 + }, + { + "epoch": 1.403620988725065, + "grad_norm": 1.5861355066299438, + "learning_rate": 1.0681125779776587e-05, + "loss": 0.0396, + "step": 12947 + }, + { + "epoch": 1.4037294015611448, + "grad_norm": 1.1196010112762451, + "learning_rate": 1.0680400406209199e-05, + "loss": 0.0156, + "step": 12948 + }, + { + "epoch": 1.4038378143972245, + "grad_norm": 7.900961399078369, + "learning_rate": 1.0679675032641813e-05, + "loss": 0.1106, + "step": 12949 + }, + { + "epoch": 1.4039462272333045, + "grad_norm": 0.7505343556404114, + "learning_rate": 1.0678949659074425e-05, + "loss": 0.0047, + "step": 12950 + }, + { + "epoch": 1.4040546400693843, + "grad_norm": 1.0119233131408691, + "learning_rate": 1.0678224285507037e-05, + "loss": 0.0141, + "step": 12951 + }, + { + "epoch": 1.404163052905464, + "grad_norm": 2.3323771953582764, + "learning_rate": 1.067749891193965e-05, + "loss": 0.0456, + "step": 12952 + }, + { + "epoch": 1.4042714657415438, + "grad_norm": 4.5879926681518555, + "learning_rate": 1.0676773538372263e-05, + "loss": 0.0399, + "step": 12953 + }, + { + "epoch": 1.4043798785776236, + "grad_norm": 5.493820667266846, + "learning_rate": 1.0676048164804875e-05, + "loss": 0.0091, + "step": 12954 + }, + { + "epoch": 1.4044882914137033, + "grad_norm": 2.1510024070739746, + "learning_rate": 1.0675322791237487e-05, + "loss": 0.0397, + "step": 12955 + }, + { + "epoch": 1.4045967042497831, + "grad_norm": 0.7275189757347107, + "learning_rate": 1.0674597417670101e-05, + "loss": 0.0046, + "step": 12956 + }, + { + "epoch": 1.404705117085863, + "grad_norm": 2.828371047973633, + "learning_rate": 1.0673872044102713e-05, + "loss": 0.03, + "step": 12957 + }, + { + "epoch": 1.4048135299219426, + "grad_norm": 3.9462199211120605, + "learning_rate": 1.0673146670535325e-05, + "loss": 0.0377, + "step": 12958 + }, + { + "epoch": 1.4049219427580226, + "grad_norm": 3.937436580657959, + "learning_rate": 1.0672421296967939e-05, + "loss": 0.0839, + "step": 12959 + }, + { + "epoch": 1.4050303555941024, + "grad_norm": 0.21015164256095886, + "learning_rate": 1.0671695923400551e-05, + "loss": 0.0033, + "step": 12960 + }, + { + "epoch": 1.4051387684301822, + "grad_norm": 2.3102657794952393, + "learning_rate": 1.0670970549833163e-05, + "loss": 0.0374, + "step": 12961 + }, + { + "epoch": 1.405247181266262, + "grad_norm": 1.3031368255615234, + "learning_rate": 1.0670245176265779e-05, + "loss": 0.0086, + "step": 12962 + }, + { + "epoch": 1.4053555941023417, + "grad_norm": 2.014681100845337, + "learning_rate": 1.0669519802698391e-05, + "loss": 0.0109, + "step": 12963 + }, + { + "epoch": 1.4054640069384214, + "grad_norm": 0.15883868932724, + "learning_rate": 1.0668794429131005e-05, + "loss": 0.0019, + "step": 12964 + }, + { + "epoch": 1.4055724197745012, + "grad_norm": 0.32955849170684814, + "learning_rate": 1.0668069055563617e-05, + "loss": 0.002, + "step": 12965 + }, + { + "epoch": 1.4056808326105812, + "grad_norm": 2.5530593395233154, + "learning_rate": 1.066734368199623e-05, + "loss": 0.0269, + "step": 12966 + }, + { + "epoch": 1.405789245446661, + "grad_norm": 7.669620513916016, + "learning_rate": 1.0666618308428841e-05, + "loss": 0.0348, + "step": 12967 + }, + { + "epoch": 1.4058976582827407, + "grad_norm": 2.718491315841675, + "learning_rate": 1.0665892934861455e-05, + "loss": 0.0179, + "step": 12968 + }, + { + "epoch": 1.4060060711188205, + "grad_norm": 0.5079123377799988, + "learning_rate": 1.0665167561294067e-05, + "loss": 0.0036, + "step": 12969 + }, + { + "epoch": 1.4061144839549002, + "grad_norm": 1.4302994012832642, + "learning_rate": 1.066444218772668e-05, + "loss": 0.0388, + "step": 12970 + }, + { + "epoch": 1.40622289679098, + "grad_norm": 1.7313241958618164, + "learning_rate": 1.0663716814159293e-05, + "loss": 0.0333, + "step": 12971 + }, + { + "epoch": 1.4063313096270598, + "grad_norm": 2.7667760848999023, + "learning_rate": 1.0662991440591905e-05, + "loss": 0.0087, + "step": 12972 + }, + { + "epoch": 1.4064397224631398, + "grad_norm": 1.042690634727478, + "learning_rate": 1.0662266067024518e-05, + "loss": 0.0151, + "step": 12973 + }, + { + "epoch": 1.4065481352992193, + "grad_norm": 2.8913824558258057, + "learning_rate": 1.0661540693457131e-05, + "loss": 0.011, + "step": 12974 + }, + { + "epoch": 1.4066565481352993, + "grad_norm": 5.041173934936523, + "learning_rate": 1.0660815319889744e-05, + "loss": 0.1559, + "step": 12975 + }, + { + "epoch": 1.406764960971379, + "grad_norm": 0.8582003116607666, + "learning_rate": 1.0660089946322356e-05, + "loss": 0.0095, + "step": 12976 + }, + { + "epoch": 1.4068733738074588, + "grad_norm": 1.944566011428833, + "learning_rate": 1.065936457275497e-05, + "loss": 0.0087, + "step": 12977 + }, + { + "epoch": 1.4069817866435386, + "grad_norm": 0.4609197974205017, + "learning_rate": 1.0658639199187582e-05, + "loss": 0.006, + "step": 12978 + }, + { + "epoch": 1.4070901994796183, + "grad_norm": 8.472318649291992, + "learning_rate": 1.0657913825620194e-05, + "loss": 0.0447, + "step": 12979 + }, + { + "epoch": 1.4071986123156983, + "grad_norm": 0.6342862844467163, + "learning_rate": 1.065718845205281e-05, + "loss": 0.007, + "step": 12980 + }, + { + "epoch": 1.4073070251517779, + "grad_norm": 2.232816219329834, + "learning_rate": 1.0656463078485422e-05, + "loss": 0.0309, + "step": 12981 + }, + { + "epoch": 1.4074154379878578, + "grad_norm": 2.390104055404663, + "learning_rate": 1.0655737704918034e-05, + "loss": 0.0188, + "step": 12982 + }, + { + "epoch": 1.4075238508239376, + "grad_norm": 2.408566951751709, + "learning_rate": 1.0655012331350647e-05, + "loss": 0.0208, + "step": 12983 + }, + { + "epoch": 1.4076322636600174, + "grad_norm": 2.4157023429870605, + "learning_rate": 1.065428695778326e-05, + "loss": 0.0584, + "step": 12984 + }, + { + "epoch": 1.4077406764960971, + "grad_norm": 0.27068859338760376, + "learning_rate": 1.0653561584215872e-05, + "loss": 0.0031, + "step": 12985 + }, + { + "epoch": 1.407849089332177, + "grad_norm": 3.6822316646575928, + "learning_rate": 1.0652836210648486e-05, + "loss": 0.0046, + "step": 12986 + }, + { + "epoch": 1.4079575021682567, + "grad_norm": 2.713517189025879, + "learning_rate": 1.0652110837081098e-05, + "loss": 0.0063, + "step": 12987 + }, + { + "epoch": 1.4080659150043364, + "grad_norm": 1.1852024793624878, + "learning_rate": 1.065138546351371e-05, + "loss": 0.0176, + "step": 12988 + }, + { + "epoch": 1.4081743278404164, + "grad_norm": 5.078118324279785, + "learning_rate": 1.0650660089946324e-05, + "loss": 0.0238, + "step": 12989 + }, + { + "epoch": 1.4082827406764962, + "grad_norm": 1.1906650066375732, + "learning_rate": 1.0649934716378936e-05, + "loss": 0.0154, + "step": 12990 + }, + { + "epoch": 1.408391153512576, + "grad_norm": 0.49151507019996643, + "learning_rate": 1.0649209342811548e-05, + "loss": 0.0069, + "step": 12991 + }, + { + "epoch": 1.4084995663486557, + "grad_norm": 0.9636626839637756, + "learning_rate": 1.0648483969244162e-05, + "loss": 0.0169, + "step": 12992 + }, + { + "epoch": 1.4086079791847355, + "grad_norm": 2.37392258644104, + "learning_rate": 1.0647758595676774e-05, + "loss": 0.0081, + "step": 12993 + }, + { + "epoch": 1.4087163920208152, + "grad_norm": 5.647275924682617, + "learning_rate": 1.0647033222109386e-05, + "loss": 0.0215, + "step": 12994 + }, + { + "epoch": 1.408824804856895, + "grad_norm": 0.8826005458831787, + "learning_rate": 1.0646307848541998e-05, + "loss": 0.0097, + "step": 12995 + }, + { + "epoch": 1.408933217692975, + "grad_norm": 3.255640745162964, + "learning_rate": 1.0645582474974612e-05, + "loss": 0.0196, + "step": 12996 + }, + { + "epoch": 1.4090416305290545, + "grad_norm": 3.296154260635376, + "learning_rate": 1.0644857101407226e-05, + "loss": 0.0318, + "step": 12997 + }, + { + "epoch": 1.4091500433651345, + "grad_norm": 1.5127372741699219, + "learning_rate": 1.064413172783984e-05, + "loss": 0.0095, + "step": 12998 + }, + { + "epoch": 1.4092584562012143, + "grad_norm": 3.925801992416382, + "learning_rate": 1.0643406354272452e-05, + "loss": 0.0736, + "step": 12999 + }, + { + "epoch": 1.409366869037294, + "grad_norm": 0.4768068194389343, + "learning_rate": 1.0642680980705064e-05, + "loss": 0.0069, + "step": 13000 + }, + { + "epoch": 1.4094752818733738, + "grad_norm": 2.8171234130859375, + "learning_rate": 1.0641955607137678e-05, + "loss": 0.0713, + "step": 13001 + }, + { + "epoch": 1.4095836947094535, + "grad_norm": 1.4024995565414429, + "learning_rate": 1.064123023357029e-05, + "loss": 0.0165, + "step": 13002 + }, + { + "epoch": 1.4096921075455333, + "grad_norm": 0.812329888343811, + "learning_rate": 1.0640504860002902e-05, + "loss": 0.0129, + "step": 13003 + }, + { + "epoch": 1.409800520381613, + "grad_norm": 1.1825816631317139, + "learning_rate": 1.0639779486435516e-05, + "loss": 0.0071, + "step": 13004 + }, + { + "epoch": 1.409908933217693, + "grad_norm": 0.36408382654190063, + "learning_rate": 1.0639054112868128e-05, + "loss": 0.0044, + "step": 13005 + }, + { + "epoch": 1.4100173460537728, + "grad_norm": 5.419857501983643, + "learning_rate": 1.063832873930074e-05, + "loss": 0.045, + "step": 13006 + }, + { + "epoch": 1.4101257588898526, + "grad_norm": 1.2507872581481934, + "learning_rate": 1.0637603365733352e-05, + "loss": 0.0116, + "step": 13007 + }, + { + "epoch": 1.4102341717259324, + "grad_norm": 2.6180851459503174, + "learning_rate": 1.0636877992165966e-05, + "loss": 0.0151, + "step": 13008 + }, + { + "epoch": 1.4103425845620121, + "grad_norm": 0.662930428981781, + "learning_rate": 1.0636152618598578e-05, + "loss": 0.0077, + "step": 13009 + }, + { + "epoch": 1.4104509973980919, + "grad_norm": 3.8168938159942627, + "learning_rate": 1.063542724503119e-05, + "loss": 0.0356, + "step": 13010 + }, + { + "epoch": 1.4105594102341716, + "grad_norm": 1.7789125442504883, + "learning_rate": 1.0634701871463804e-05, + "loss": 0.0234, + "step": 13011 + }, + { + "epoch": 1.4106678230702516, + "grad_norm": 3.7443840503692627, + "learning_rate": 1.0633976497896417e-05, + "loss": 0.0157, + "step": 13012 + }, + { + "epoch": 1.4107762359063314, + "grad_norm": 1.5138916969299316, + "learning_rate": 1.0633251124329029e-05, + "loss": 0.0209, + "step": 13013 + }, + { + "epoch": 1.4108846487424112, + "grad_norm": 3.366373300552368, + "learning_rate": 1.0632525750761644e-05, + "loss": 0.048, + "step": 13014 + }, + { + "epoch": 1.410993061578491, + "grad_norm": 4.6957106590271, + "learning_rate": 1.0631800377194256e-05, + "loss": 0.0339, + "step": 13015 + }, + { + "epoch": 1.4111014744145707, + "grad_norm": 1.274024486541748, + "learning_rate": 1.063107500362687e-05, + "loss": 0.0385, + "step": 13016 + }, + { + "epoch": 1.4112098872506504, + "grad_norm": 4.970404148101807, + "learning_rate": 1.0630349630059482e-05, + "loss": 0.012, + "step": 13017 + }, + { + "epoch": 1.4113183000867302, + "grad_norm": 4.04224967956543, + "learning_rate": 1.0629624256492095e-05, + "loss": 0.0546, + "step": 13018 + }, + { + "epoch": 1.4114267129228102, + "grad_norm": 1.685858964920044, + "learning_rate": 1.0628898882924707e-05, + "loss": 0.0255, + "step": 13019 + }, + { + "epoch": 1.4115351257588897, + "grad_norm": 0.36981815099716187, + "learning_rate": 1.062817350935732e-05, + "loss": 0.0026, + "step": 13020 + }, + { + "epoch": 1.4116435385949697, + "grad_norm": 0.09736111760139465, + "learning_rate": 1.0627448135789933e-05, + "loss": 0.0011, + "step": 13021 + }, + { + "epoch": 1.4117519514310495, + "grad_norm": 6.781869888305664, + "learning_rate": 1.0626722762222545e-05, + "loss": 0.0308, + "step": 13022 + }, + { + "epoch": 1.4118603642671292, + "grad_norm": 3.5113108158111572, + "learning_rate": 1.0625997388655159e-05, + "loss": 0.0149, + "step": 13023 + }, + { + "epoch": 1.411968777103209, + "grad_norm": 1.1939507722854614, + "learning_rate": 1.062527201508777e-05, + "loss": 0.0079, + "step": 13024 + }, + { + "epoch": 1.4120771899392888, + "grad_norm": 6.499022006988525, + "learning_rate": 1.0624546641520383e-05, + "loss": 0.0147, + "step": 13025 + }, + { + "epoch": 1.4121856027753685, + "grad_norm": 1.7014539241790771, + "learning_rate": 1.0623821267952997e-05, + "loss": 0.022, + "step": 13026 + }, + { + "epoch": 1.4122940156114483, + "grad_norm": 3.149529218673706, + "learning_rate": 1.0623095894385609e-05, + "loss": 0.0314, + "step": 13027 + }, + { + "epoch": 1.4124024284475283, + "grad_norm": 1.092481255531311, + "learning_rate": 1.0622370520818221e-05, + "loss": 0.0153, + "step": 13028 + }, + { + "epoch": 1.412510841283608, + "grad_norm": 2.6177947521209717, + "learning_rate": 1.0621645147250835e-05, + "loss": 0.0482, + "step": 13029 + }, + { + "epoch": 1.4126192541196878, + "grad_norm": 9.50150203704834, + "learning_rate": 1.0620919773683447e-05, + "loss": 0.0437, + "step": 13030 + }, + { + "epoch": 1.4127276669557676, + "grad_norm": 2.082472324371338, + "learning_rate": 1.0620194400116063e-05, + "loss": 0.0196, + "step": 13031 + }, + { + "epoch": 1.4128360797918473, + "grad_norm": 0.8853821158409119, + "learning_rate": 1.0619469026548675e-05, + "loss": 0.011, + "step": 13032 + }, + { + "epoch": 1.412944492627927, + "grad_norm": 1.7139586210250854, + "learning_rate": 1.0618743652981287e-05, + "loss": 0.0075, + "step": 13033 + }, + { + "epoch": 1.4130529054640069, + "grad_norm": 3.2406742572784424, + "learning_rate": 1.0618018279413899e-05, + "loss": 0.036, + "step": 13034 + }, + { + "epoch": 1.4131613183000868, + "grad_norm": 3.147491455078125, + "learning_rate": 1.0617292905846513e-05, + "loss": 0.0172, + "step": 13035 + }, + { + "epoch": 1.4132697311361664, + "grad_norm": 4.976644515991211, + "learning_rate": 1.0616567532279125e-05, + "loss": 0.0211, + "step": 13036 + }, + { + "epoch": 1.4133781439722464, + "grad_norm": 1.416043996810913, + "learning_rate": 1.0615842158711737e-05, + "loss": 0.0147, + "step": 13037 + }, + { + "epoch": 1.4134865568083261, + "grad_norm": 3.8085122108459473, + "learning_rate": 1.0615116785144351e-05, + "loss": 0.0279, + "step": 13038 + }, + { + "epoch": 1.413594969644406, + "grad_norm": 0.6269848942756653, + "learning_rate": 1.0614391411576963e-05, + "loss": 0.0094, + "step": 13039 + }, + { + "epoch": 1.4137033824804857, + "grad_norm": 1.9202516078948975, + "learning_rate": 1.0613666038009575e-05, + "loss": 0.0134, + "step": 13040 + }, + { + "epoch": 1.4138117953165654, + "grad_norm": 2.5531909465789795, + "learning_rate": 1.0612940664442189e-05, + "loss": 0.0299, + "step": 13041 + }, + { + "epoch": 1.4139202081526454, + "grad_norm": 1.4700945615768433, + "learning_rate": 1.0612215290874801e-05, + "loss": 0.0112, + "step": 13042 + }, + { + "epoch": 1.414028620988725, + "grad_norm": 2.2894134521484375, + "learning_rate": 1.0611489917307413e-05, + "loss": 0.0388, + "step": 13043 + }, + { + "epoch": 1.414137033824805, + "grad_norm": 0.3746837377548218, + "learning_rate": 1.0610764543740027e-05, + "loss": 0.0024, + "step": 13044 + }, + { + "epoch": 1.4142454466608847, + "grad_norm": 2.126894235610962, + "learning_rate": 1.061003917017264e-05, + "loss": 0.0428, + "step": 13045 + }, + { + "epoch": 1.4143538594969645, + "grad_norm": 1.6602269411087036, + "learning_rate": 1.0609313796605251e-05, + "loss": 0.0144, + "step": 13046 + }, + { + "epoch": 1.4144622723330442, + "grad_norm": 1.5459157228469849, + "learning_rate": 1.0608588423037864e-05, + "loss": 0.013, + "step": 13047 + }, + { + "epoch": 1.414570685169124, + "grad_norm": 1.5004198551177979, + "learning_rate": 1.0607863049470477e-05, + "loss": 0.0147, + "step": 13048 + }, + { + "epoch": 1.4146790980052038, + "grad_norm": 8.418013572692871, + "learning_rate": 1.0607137675903091e-05, + "loss": 0.0344, + "step": 13049 + }, + { + "epoch": 1.4147875108412835, + "grad_norm": 2.287452459335327, + "learning_rate": 1.0606412302335705e-05, + "loss": 0.0276, + "step": 13050 + }, + { + "epoch": 1.4148959236773635, + "grad_norm": 1.1516457796096802, + "learning_rate": 1.0605686928768317e-05, + "loss": 0.0066, + "step": 13051 + }, + { + "epoch": 1.4150043365134433, + "grad_norm": 5.2514801025390625, + "learning_rate": 1.060496155520093e-05, + "loss": 0.0699, + "step": 13052 + }, + { + "epoch": 1.415112749349523, + "grad_norm": 0.6236515045166016, + "learning_rate": 1.0604236181633543e-05, + "loss": 0.0049, + "step": 13053 + }, + { + "epoch": 1.4152211621856028, + "grad_norm": 1.1379899978637695, + "learning_rate": 1.0603510808066155e-05, + "loss": 0.012, + "step": 13054 + }, + { + "epoch": 1.4153295750216826, + "grad_norm": 3.4078738689422607, + "learning_rate": 1.0602785434498768e-05, + "loss": 0.0249, + "step": 13055 + }, + { + "epoch": 1.4154379878577623, + "grad_norm": 0.31247591972351074, + "learning_rate": 1.0602060060931381e-05, + "loss": 0.003, + "step": 13056 + }, + { + "epoch": 1.415546400693842, + "grad_norm": 3.1437063217163086, + "learning_rate": 1.0601334687363994e-05, + "loss": 0.0385, + "step": 13057 + }, + { + "epoch": 1.415654813529922, + "grad_norm": 0.5689377784729004, + "learning_rate": 1.0600609313796606e-05, + "loss": 0.0054, + "step": 13058 + }, + { + "epoch": 1.4157632263660016, + "grad_norm": 2.884932279586792, + "learning_rate": 1.0599883940229218e-05, + "loss": 0.0335, + "step": 13059 + }, + { + "epoch": 1.4158716392020816, + "grad_norm": 0.7791730165481567, + "learning_rate": 1.0599158566661832e-05, + "loss": 0.0029, + "step": 13060 + }, + { + "epoch": 1.4159800520381614, + "grad_norm": 2.318103313446045, + "learning_rate": 1.0598433193094444e-05, + "loss": 0.0297, + "step": 13061 + }, + { + "epoch": 1.4160884648742411, + "grad_norm": 1.9561169147491455, + "learning_rate": 1.0597707819527056e-05, + "loss": 0.0168, + "step": 13062 + }, + { + "epoch": 1.4161968777103209, + "grad_norm": 2.151817560195923, + "learning_rate": 1.059698244595967e-05, + "loss": 0.0174, + "step": 13063 + }, + { + "epoch": 1.4163052905464006, + "grad_norm": 1.091600775718689, + "learning_rate": 1.0596257072392282e-05, + "loss": 0.0172, + "step": 13064 + }, + { + "epoch": 1.4164137033824804, + "grad_norm": 4.11770486831665, + "learning_rate": 1.0595531698824894e-05, + "loss": 0.0047, + "step": 13065 + }, + { + "epoch": 1.4165221162185602, + "grad_norm": 1.9969336986541748, + "learning_rate": 1.059480632525751e-05, + "loss": 0.0222, + "step": 13066 + }, + { + "epoch": 1.4166305290546402, + "grad_norm": 1.1641541719436646, + "learning_rate": 1.0594080951690122e-05, + "loss": 0.0163, + "step": 13067 + }, + { + "epoch": 1.41673894189072, + "grad_norm": 6.585204601287842, + "learning_rate": 1.0593355578122736e-05, + "loss": 0.052, + "step": 13068 + }, + { + "epoch": 1.4168473547267997, + "grad_norm": 3.7138559818267822, + "learning_rate": 1.0592630204555348e-05, + "loss": 0.0477, + "step": 13069 + }, + { + "epoch": 1.4169557675628794, + "grad_norm": 1.477765440940857, + "learning_rate": 1.059190483098796e-05, + "loss": 0.0089, + "step": 13070 + }, + { + "epoch": 1.4170641803989592, + "grad_norm": 1.5093212127685547, + "learning_rate": 1.0591179457420572e-05, + "loss": 0.0255, + "step": 13071 + }, + { + "epoch": 1.417172593235039, + "grad_norm": 1.3220490217208862, + "learning_rate": 1.0590454083853186e-05, + "loss": 0.008, + "step": 13072 + }, + { + "epoch": 1.4172810060711187, + "grad_norm": 0.7055206894874573, + "learning_rate": 1.0589728710285798e-05, + "loss": 0.002, + "step": 13073 + }, + { + "epoch": 1.4173894189071987, + "grad_norm": 5.411159992218018, + "learning_rate": 1.058900333671841e-05, + "loss": 0.041, + "step": 13074 + }, + { + "epoch": 1.4174978317432785, + "grad_norm": 1.9598627090454102, + "learning_rate": 1.0588277963151024e-05, + "loss": 0.0234, + "step": 13075 + }, + { + "epoch": 1.4176062445793582, + "grad_norm": 1.0563875436782837, + "learning_rate": 1.0587552589583636e-05, + "loss": 0.0107, + "step": 13076 + }, + { + "epoch": 1.417714657415438, + "grad_norm": 0.21556898951530457, + "learning_rate": 1.0586827216016248e-05, + "loss": 0.0023, + "step": 13077 + }, + { + "epoch": 1.4178230702515178, + "grad_norm": 1.8837190866470337, + "learning_rate": 1.0586101842448862e-05, + "loss": 0.0293, + "step": 13078 + }, + { + "epoch": 1.4179314830875975, + "grad_norm": 3.194291353225708, + "learning_rate": 1.0585376468881474e-05, + "loss": 0.028, + "step": 13079 + }, + { + "epoch": 1.4180398959236773, + "grad_norm": 6.227427005767822, + "learning_rate": 1.0584651095314086e-05, + "loss": 0.0357, + "step": 13080 + }, + { + "epoch": 1.4181483087597573, + "grad_norm": 5.570324897766113, + "learning_rate": 1.05839257217467e-05, + "loss": 0.0509, + "step": 13081 + }, + { + "epoch": 1.4182567215958368, + "grad_norm": 2.335049867630005, + "learning_rate": 1.0583200348179312e-05, + "loss": 0.0229, + "step": 13082 + }, + { + "epoch": 1.4183651344319168, + "grad_norm": 3.4547317028045654, + "learning_rate": 1.0582474974611928e-05, + "loss": 0.0252, + "step": 13083 + }, + { + "epoch": 1.4184735472679966, + "grad_norm": 14.795618057250977, + "learning_rate": 1.058174960104454e-05, + "loss": 0.0228, + "step": 13084 + }, + { + "epoch": 1.4185819601040763, + "grad_norm": 1.4037235975265503, + "learning_rate": 1.0581024227477152e-05, + "loss": 0.0255, + "step": 13085 + }, + { + "epoch": 1.418690372940156, + "grad_norm": 9.012496948242188, + "learning_rate": 1.0580298853909764e-05, + "loss": 0.0142, + "step": 13086 + }, + { + "epoch": 1.4187987857762359, + "grad_norm": 1.9923722743988037, + "learning_rate": 1.0579573480342378e-05, + "loss": 0.014, + "step": 13087 + }, + { + "epoch": 1.4189071986123156, + "grad_norm": 2.846360445022583, + "learning_rate": 1.057884810677499e-05, + "loss": 0.0224, + "step": 13088 + }, + { + "epoch": 1.4190156114483954, + "grad_norm": 3.6040871143341064, + "learning_rate": 1.0578122733207602e-05, + "loss": 0.0374, + "step": 13089 + }, + { + "epoch": 1.4191240242844754, + "grad_norm": 4.311765193939209, + "learning_rate": 1.0577397359640216e-05, + "loss": 0.0083, + "step": 13090 + }, + { + "epoch": 1.4192324371205551, + "grad_norm": 4.131398677825928, + "learning_rate": 1.0576671986072828e-05, + "loss": 0.0377, + "step": 13091 + }, + { + "epoch": 1.419340849956635, + "grad_norm": 2.131211519241333, + "learning_rate": 1.057594661250544e-05, + "loss": 0.0123, + "step": 13092 + }, + { + "epoch": 1.4194492627927147, + "grad_norm": 1.8067481517791748, + "learning_rate": 1.0575221238938054e-05, + "loss": 0.0165, + "step": 13093 + }, + { + "epoch": 1.4195576756287944, + "grad_norm": 0.2604126036167145, + "learning_rate": 1.0574495865370667e-05, + "loss": 0.0022, + "step": 13094 + }, + { + "epoch": 1.4196660884648742, + "grad_norm": 0.014509792439639568, + "learning_rate": 1.0573770491803279e-05, + "loss": 0.0002, + "step": 13095 + }, + { + "epoch": 1.419774501300954, + "grad_norm": 1.1659553050994873, + "learning_rate": 1.0573045118235893e-05, + "loss": 0.0244, + "step": 13096 + }, + { + "epoch": 1.419882914137034, + "grad_norm": 2.6532254219055176, + "learning_rate": 1.0572319744668505e-05, + "loss": 0.0152, + "step": 13097 + }, + { + "epoch": 1.4199913269731137, + "grad_norm": 1.1493768692016602, + "learning_rate": 1.0571594371101117e-05, + "loss": 0.0216, + "step": 13098 + }, + { + "epoch": 1.4200997398091935, + "grad_norm": 5.183714866638184, + "learning_rate": 1.0570868997533729e-05, + "loss": 0.0318, + "step": 13099 + }, + { + "epoch": 1.4202081526452732, + "grad_norm": 1.1041074991226196, + "learning_rate": 1.0570143623966345e-05, + "loss": 0.0152, + "step": 13100 + }, + { + "epoch": 1.420316565481353, + "grad_norm": 1.3752155303955078, + "learning_rate": 1.0569418250398957e-05, + "loss": 0.0146, + "step": 13101 + }, + { + "epoch": 1.4204249783174328, + "grad_norm": 5.619792938232422, + "learning_rate": 1.056869287683157e-05, + "loss": 0.0212, + "step": 13102 + }, + { + "epoch": 1.4205333911535125, + "grad_norm": 0.5040199160575867, + "learning_rate": 1.0567967503264183e-05, + "loss": 0.0017, + "step": 13103 + }, + { + "epoch": 1.4206418039895925, + "grad_norm": 0.5096346139907837, + "learning_rate": 1.0567242129696795e-05, + "loss": 0.0101, + "step": 13104 + }, + { + "epoch": 1.420750216825672, + "grad_norm": 2.128218173980713, + "learning_rate": 1.0566516756129409e-05, + "loss": 0.0072, + "step": 13105 + }, + { + "epoch": 1.420858629661752, + "grad_norm": 4.102447986602783, + "learning_rate": 1.056579138256202e-05, + "loss": 0.0251, + "step": 13106 + }, + { + "epoch": 1.4209670424978318, + "grad_norm": 0.6041157841682434, + "learning_rate": 1.0565066008994633e-05, + "loss": 0.0068, + "step": 13107 + }, + { + "epoch": 1.4210754553339116, + "grad_norm": 0.5991945862770081, + "learning_rate": 1.0564340635427247e-05, + "loss": 0.0039, + "step": 13108 + }, + { + "epoch": 1.4211838681699913, + "grad_norm": 0.2613890767097473, + "learning_rate": 1.0563615261859859e-05, + "loss": 0.002, + "step": 13109 + }, + { + "epoch": 1.421292281006071, + "grad_norm": 1.5503824949264526, + "learning_rate": 1.0562889888292471e-05, + "loss": 0.0251, + "step": 13110 + }, + { + "epoch": 1.4214006938421508, + "grad_norm": 0.08866018056869507, + "learning_rate": 1.0562164514725083e-05, + "loss": 0.0007, + "step": 13111 + }, + { + "epoch": 1.4215091066782306, + "grad_norm": 3.7088677883148193, + "learning_rate": 1.0561439141157697e-05, + "loss": 0.0472, + "step": 13112 + }, + { + "epoch": 1.4216175195143106, + "grad_norm": 2.4870738983154297, + "learning_rate": 1.0560713767590309e-05, + "loss": 0.0299, + "step": 13113 + }, + { + "epoch": 1.4217259323503904, + "grad_norm": 3.122107982635498, + "learning_rate": 1.0559988394022921e-05, + "loss": 0.0458, + "step": 13114 + }, + { + "epoch": 1.4218343451864701, + "grad_norm": 0.15808379650115967, + "learning_rate": 1.0559263020455535e-05, + "loss": 0.0017, + "step": 13115 + }, + { + "epoch": 1.4219427580225499, + "grad_norm": 3.4505326747894287, + "learning_rate": 1.0558537646888147e-05, + "loss": 0.0269, + "step": 13116 + }, + { + "epoch": 1.4220511708586296, + "grad_norm": 0.15044936537742615, + "learning_rate": 1.055781227332076e-05, + "loss": 0.0013, + "step": 13117 + }, + { + "epoch": 1.4221595836947094, + "grad_norm": 1.808715581893921, + "learning_rate": 1.0557086899753375e-05, + "loss": 0.007, + "step": 13118 + }, + { + "epoch": 1.4222679965307892, + "grad_norm": 0.39233484864234924, + "learning_rate": 1.0556361526185987e-05, + "loss": 0.0023, + "step": 13119 + }, + { + "epoch": 1.4223764093668692, + "grad_norm": 15.306288719177246, + "learning_rate": 1.0555636152618601e-05, + "loss": 0.0483, + "step": 13120 + }, + { + "epoch": 1.4224848222029487, + "grad_norm": 9.85411262512207, + "learning_rate": 1.0554910779051213e-05, + "loss": 0.0372, + "step": 13121 + }, + { + "epoch": 1.4225932350390287, + "grad_norm": 1.1986079216003418, + "learning_rate": 1.0554185405483825e-05, + "loss": 0.0061, + "step": 13122 + }, + { + "epoch": 1.4227016478751084, + "grad_norm": 5.071477890014648, + "learning_rate": 1.0553460031916437e-05, + "loss": 0.0085, + "step": 13123 + }, + { + "epoch": 1.4228100607111882, + "grad_norm": 6.151615142822266, + "learning_rate": 1.0552734658349051e-05, + "loss": 0.079, + "step": 13124 + }, + { + "epoch": 1.422918473547268, + "grad_norm": 3.183593988418579, + "learning_rate": 1.0552009284781663e-05, + "loss": 0.0393, + "step": 13125 + }, + { + "epoch": 1.4230268863833477, + "grad_norm": 1.3286080360412598, + "learning_rate": 1.0551283911214275e-05, + "loss": 0.0153, + "step": 13126 + }, + { + "epoch": 1.4231352992194277, + "grad_norm": 0.12281150370836258, + "learning_rate": 1.055055853764689e-05, + "loss": 0.0018, + "step": 13127 + }, + { + "epoch": 1.4232437120555073, + "grad_norm": 0.587715208530426, + "learning_rate": 1.0549833164079501e-05, + "loss": 0.0058, + "step": 13128 + }, + { + "epoch": 1.4233521248915872, + "grad_norm": 3.1769285202026367, + "learning_rate": 1.0549107790512114e-05, + "loss": 0.0604, + "step": 13129 + }, + { + "epoch": 1.423460537727667, + "grad_norm": 1.0398368835449219, + "learning_rate": 1.0548382416944727e-05, + "loss": 0.0173, + "step": 13130 + }, + { + "epoch": 1.4235689505637468, + "grad_norm": 2.812129020690918, + "learning_rate": 1.054765704337734e-05, + "loss": 0.0381, + "step": 13131 + }, + { + "epoch": 1.4236773633998265, + "grad_norm": 1.5143234729766846, + "learning_rate": 1.0546931669809952e-05, + "loss": 0.0205, + "step": 13132 + }, + { + "epoch": 1.4237857762359063, + "grad_norm": 0.3874683678150177, + "learning_rate": 1.0546206296242566e-05, + "loss": 0.004, + "step": 13133 + }, + { + "epoch": 1.423894189071986, + "grad_norm": 0.22980783879756927, + "learning_rate": 1.0545480922675178e-05, + "loss": 0.0026, + "step": 13134 + }, + { + "epoch": 1.4240026019080658, + "grad_norm": 7.72109842300415, + "learning_rate": 1.0544755549107793e-05, + "loss": 0.0296, + "step": 13135 + }, + { + "epoch": 1.4241110147441458, + "grad_norm": 4.675834655761719, + "learning_rate": 1.0544030175540405e-05, + "loss": 0.0502, + "step": 13136 + }, + { + "epoch": 1.4242194275802256, + "grad_norm": 1.1277961730957031, + "learning_rate": 1.0543304801973018e-05, + "loss": 0.0184, + "step": 13137 + }, + { + "epoch": 1.4243278404163053, + "grad_norm": 0.18765303492546082, + "learning_rate": 1.054257942840563e-05, + "loss": 0.002, + "step": 13138 + }, + { + "epoch": 1.424436253252385, + "grad_norm": 0.4669662117958069, + "learning_rate": 1.0541854054838244e-05, + "loss": 0.0028, + "step": 13139 + }, + { + "epoch": 1.4245446660884649, + "grad_norm": 2.9972362518310547, + "learning_rate": 1.0541128681270856e-05, + "loss": 0.046, + "step": 13140 + }, + { + "epoch": 1.4246530789245446, + "grad_norm": 3.4690847396850586, + "learning_rate": 1.0540403307703468e-05, + "loss": 0.0154, + "step": 13141 + }, + { + "epoch": 1.4247614917606244, + "grad_norm": 2.810991048812866, + "learning_rate": 1.0539677934136082e-05, + "loss": 0.0292, + "step": 13142 + }, + { + "epoch": 1.4248699045967044, + "grad_norm": 2.161797046661377, + "learning_rate": 1.0538952560568694e-05, + "loss": 0.01, + "step": 13143 + }, + { + "epoch": 1.424978317432784, + "grad_norm": 3.884153127670288, + "learning_rate": 1.0538227187001306e-05, + "loss": 0.0227, + "step": 13144 + }, + { + "epoch": 1.425086730268864, + "grad_norm": 2.6933295726776123, + "learning_rate": 1.053750181343392e-05, + "loss": 0.0108, + "step": 13145 + }, + { + "epoch": 1.4251951431049437, + "grad_norm": 0.21486429870128632, + "learning_rate": 1.0536776439866532e-05, + "loss": 0.0017, + "step": 13146 + }, + { + "epoch": 1.4253035559410234, + "grad_norm": 1.6798666715621948, + "learning_rate": 1.0536051066299144e-05, + "loss": 0.0221, + "step": 13147 + }, + { + "epoch": 1.4254119687771032, + "grad_norm": 5.784639358520508, + "learning_rate": 1.0535325692731758e-05, + "loss": 0.0296, + "step": 13148 + }, + { + "epoch": 1.425520381613183, + "grad_norm": 9.742291450500488, + "learning_rate": 1.053460031916437e-05, + "loss": 0.0515, + "step": 13149 + }, + { + "epoch": 1.4256287944492627, + "grad_norm": 0.46521008014678955, + "learning_rate": 1.0533874945596982e-05, + "loss": 0.0035, + "step": 13150 + }, + { + "epoch": 1.4257372072853425, + "grad_norm": 0.5779054760932922, + "learning_rate": 1.0533149572029594e-05, + "loss": 0.0017, + "step": 13151 + }, + { + "epoch": 1.4258456201214225, + "grad_norm": 4.8416595458984375, + "learning_rate": 1.053242419846221e-05, + "loss": 0.0652, + "step": 13152 + }, + { + "epoch": 1.4259540329575022, + "grad_norm": 8.223062515258789, + "learning_rate": 1.0531698824894822e-05, + "loss": 0.0453, + "step": 13153 + }, + { + "epoch": 1.426062445793582, + "grad_norm": 2.0463809967041016, + "learning_rate": 1.0530973451327436e-05, + "loss": 0.0423, + "step": 13154 + }, + { + "epoch": 1.4261708586296618, + "grad_norm": 2.4377694129943848, + "learning_rate": 1.0530248077760048e-05, + "loss": 0.0248, + "step": 13155 + }, + { + "epoch": 1.4262792714657415, + "grad_norm": 5.30227518081665, + "learning_rate": 1.052952270419266e-05, + "loss": 0.0084, + "step": 13156 + }, + { + "epoch": 1.4263876843018213, + "grad_norm": 0.5261715054512024, + "learning_rate": 1.0528797330625274e-05, + "loss": 0.0064, + "step": 13157 + }, + { + "epoch": 1.426496097137901, + "grad_norm": 0.7586100101470947, + "learning_rate": 1.0528071957057886e-05, + "loss": 0.0071, + "step": 13158 + }, + { + "epoch": 1.426604509973981, + "grad_norm": 3.2048492431640625, + "learning_rate": 1.0527346583490498e-05, + "loss": 0.0101, + "step": 13159 + }, + { + "epoch": 1.4267129228100608, + "grad_norm": 4.4781599044799805, + "learning_rate": 1.0526621209923112e-05, + "loss": 0.1081, + "step": 13160 + }, + { + "epoch": 1.4268213356461406, + "grad_norm": 5.144163131713867, + "learning_rate": 1.0525895836355724e-05, + "loss": 0.0447, + "step": 13161 + }, + { + "epoch": 1.4269297484822203, + "grad_norm": 0.5274400115013123, + "learning_rate": 1.0525170462788336e-05, + "loss": 0.0048, + "step": 13162 + }, + { + "epoch": 1.4270381613183, + "grad_norm": 5.7359795570373535, + "learning_rate": 1.0524445089220949e-05, + "loss": 0.0152, + "step": 13163 + }, + { + "epoch": 1.4271465741543798, + "grad_norm": 1.1246014833450317, + "learning_rate": 1.0523719715653562e-05, + "loss": 0.0059, + "step": 13164 + }, + { + "epoch": 1.4272549869904596, + "grad_norm": 3.7928178310394287, + "learning_rate": 1.0522994342086174e-05, + "loss": 0.028, + "step": 13165 + }, + { + "epoch": 1.4273633998265396, + "grad_norm": 3.05709171295166, + "learning_rate": 1.0522268968518787e-05, + "loss": 0.0295, + "step": 13166 + }, + { + "epoch": 1.4274718126626191, + "grad_norm": 1.2671394348144531, + "learning_rate": 1.05215435949514e-05, + "loss": 0.0163, + "step": 13167 + }, + { + "epoch": 1.4275802254986991, + "grad_norm": 0.5028981566429138, + "learning_rate": 1.0520818221384013e-05, + "loss": 0.0023, + "step": 13168 + }, + { + "epoch": 1.4276886383347789, + "grad_norm": 1.3568053245544434, + "learning_rate": 1.0520092847816625e-05, + "loss": 0.0154, + "step": 13169 + }, + { + "epoch": 1.4277970511708586, + "grad_norm": 2.4510111808776855, + "learning_rate": 1.051936747424924e-05, + "loss": 0.0182, + "step": 13170 + }, + { + "epoch": 1.4279054640069384, + "grad_norm": 4.062160491943359, + "learning_rate": 1.0518642100681852e-05, + "loss": 0.0538, + "step": 13171 + }, + { + "epoch": 1.4280138768430182, + "grad_norm": 2.3960933685302734, + "learning_rate": 1.0517916727114466e-05, + "loss": 0.0445, + "step": 13172 + }, + { + "epoch": 1.428122289679098, + "grad_norm": 1.5305583477020264, + "learning_rate": 1.0517191353547078e-05, + "loss": 0.0108, + "step": 13173 + }, + { + "epoch": 1.4282307025151777, + "grad_norm": 2.119398593902588, + "learning_rate": 1.051646597997969e-05, + "loss": 0.0485, + "step": 13174 + }, + { + "epoch": 1.4283391153512577, + "grad_norm": 2.4225287437438965, + "learning_rate": 1.0515740606412304e-05, + "loss": 0.0435, + "step": 13175 + }, + { + "epoch": 1.4284475281873374, + "grad_norm": 5.613282680511475, + "learning_rate": 1.0515015232844917e-05, + "loss": 0.0425, + "step": 13176 + }, + { + "epoch": 1.4285559410234172, + "grad_norm": 3.035191297531128, + "learning_rate": 1.0514289859277529e-05, + "loss": 0.0335, + "step": 13177 + }, + { + "epoch": 1.428664353859497, + "grad_norm": 2.345165967941284, + "learning_rate": 1.051356448571014e-05, + "loss": 0.0143, + "step": 13178 + }, + { + "epoch": 1.4287727666955767, + "grad_norm": 0.48523738980293274, + "learning_rate": 1.0512839112142755e-05, + "loss": 0.0038, + "step": 13179 + }, + { + "epoch": 1.4288811795316565, + "grad_norm": 1.8091551065444946, + "learning_rate": 1.0512113738575367e-05, + "loss": 0.0211, + "step": 13180 + }, + { + "epoch": 1.4289895923677363, + "grad_norm": 1.453395128250122, + "learning_rate": 1.0511388365007979e-05, + "loss": 0.0121, + "step": 13181 + }, + { + "epoch": 1.4290980052038162, + "grad_norm": 1.1728134155273438, + "learning_rate": 1.0510662991440593e-05, + "loss": 0.0334, + "step": 13182 + }, + { + "epoch": 1.4292064180398958, + "grad_norm": 3.16866397857666, + "learning_rate": 1.0509937617873205e-05, + "loss": 0.0134, + "step": 13183 + }, + { + "epoch": 1.4293148308759758, + "grad_norm": 3.0153415203094482, + "learning_rate": 1.0509212244305817e-05, + "loss": 0.0265, + "step": 13184 + }, + { + "epoch": 1.4294232437120555, + "grad_norm": 2.14906907081604, + "learning_rate": 1.0508486870738431e-05, + "loss": 0.0347, + "step": 13185 + }, + { + "epoch": 1.4295316565481353, + "grad_norm": 1.107893943786621, + "learning_rate": 1.0507761497171043e-05, + "loss": 0.0072, + "step": 13186 + }, + { + "epoch": 1.429640069384215, + "grad_norm": 1.760419487953186, + "learning_rate": 1.0507036123603659e-05, + "loss": 0.0206, + "step": 13187 + }, + { + "epoch": 1.4297484822202948, + "grad_norm": 3.87030029296875, + "learning_rate": 1.050631075003627e-05, + "loss": 0.0351, + "step": 13188 + }, + { + "epoch": 1.4298568950563748, + "grad_norm": 8.023579597473145, + "learning_rate": 1.0505585376468883e-05, + "loss": 0.0656, + "step": 13189 + }, + { + "epoch": 1.4299653078924544, + "grad_norm": 1.1583189964294434, + "learning_rate": 1.0504860002901495e-05, + "loss": 0.009, + "step": 13190 + }, + { + "epoch": 1.4300737207285343, + "grad_norm": 1.5876221656799316, + "learning_rate": 1.0504134629334109e-05, + "loss": 0.0259, + "step": 13191 + }, + { + "epoch": 1.430182133564614, + "grad_norm": 2.8265926837921143, + "learning_rate": 1.0503409255766721e-05, + "loss": 0.0222, + "step": 13192 + }, + { + "epoch": 1.4302905464006939, + "grad_norm": 4.288464069366455, + "learning_rate": 1.0502683882199333e-05, + "loss": 0.0831, + "step": 13193 + }, + { + "epoch": 1.4303989592367736, + "grad_norm": 3.4667317867279053, + "learning_rate": 1.0501958508631947e-05, + "loss": 0.0203, + "step": 13194 + }, + { + "epoch": 1.4305073720728534, + "grad_norm": 8.87198257446289, + "learning_rate": 1.0501233135064559e-05, + "loss": 0.0186, + "step": 13195 + }, + { + "epoch": 1.4306157849089332, + "grad_norm": 3.4399538040161133, + "learning_rate": 1.0500507761497171e-05, + "loss": 0.0574, + "step": 13196 + }, + { + "epoch": 1.430724197745013, + "grad_norm": 3.5601906776428223, + "learning_rate": 1.0499782387929785e-05, + "loss": 0.0053, + "step": 13197 + }, + { + "epoch": 1.430832610581093, + "grad_norm": 1.9509692192077637, + "learning_rate": 1.0499057014362397e-05, + "loss": 0.0183, + "step": 13198 + }, + { + "epoch": 1.4309410234171727, + "grad_norm": 0.5209412574768066, + "learning_rate": 1.049833164079501e-05, + "loss": 0.0086, + "step": 13199 + }, + { + "epoch": 1.4310494362532524, + "grad_norm": 10.659388542175293, + "learning_rate": 1.0497606267227623e-05, + "loss": 0.0277, + "step": 13200 + }, + { + "epoch": 1.4311578490893322, + "grad_norm": 2.930453300476074, + "learning_rate": 1.0496880893660235e-05, + "loss": 0.036, + "step": 13201 + }, + { + "epoch": 1.431266261925412, + "grad_norm": 1.3265796899795532, + "learning_rate": 1.0496155520092848e-05, + "loss": 0.0213, + "step": 13202 + }, + { + "epoch": 1.4313746747614917, + "grad_norm": 7.130378723144531, + "learning_rate": 1.049543014652546e-05, + "loss": 0.0329, + "step": 13203 + }, + { + "epoch": 1.4314830875975715, + "grad_norm": 1.0822190046310425, + "learning_rate": 1.0494704772958075e-05, + "loss": 0.0055, + "step": 13204 + }, + { + "epoch": 1.4315915004336515, + "grad_norm": 1.8361765146255493, + "learning_rate": 1.0493979399390687e-05, + "loss": 0.0224, + "step": 13205 + }, + { + "epoch": 1.431699913269731, + "grad_norm": 3.149667978286743, + "learning_rate": 1.0493254025823301e-05, + "loss": 0.0205, + "step": 13206 + }, + { + "epoch": 1.431808326105811, + "grad_norm": 1.006217122077942, + "learning_rate": 1.0492528652255913e-05, + "loss": 0.016, + "step": 13207 + }, + { + "epoch": 1.4319167389418908, + "grad_norm": 1.7525615692138672, + "learning_rate": 1.0491803278688525e-05, + "loss": 0.0141, + "step": 13208 + }, + { + "epoch": 1.4320251517779705, + "grad_norm": 0.7284812331199646, + "learning_rate": 1.049107790512114e-05, + "loss": 0.0073, + "step": 13209 + }, + { + "epoch": 1.4321335646140503, + "grad_norm": 2.849973201751709, + "learning_rate": 1.0490352531553751e-05, + "loss": 0.0564, + "step": 13210 + }, + { + "epoch": 1.43224197745013, + "grad_norm": 4.919819355010986, + "learning_rate": 1.0489627157986364e-05, + "loss": 0.0326, + "step": 13211 + }, + { + "epoch": 1.4323503902862098, + "grad_norm": 4.483032703399658, + "learning_rate": 1.0488901784418977e-05, + "loss": 0.0486, + "step": 13212 + }, + { + "epoch": 1.4324588031222896, + "grad_norm": 4.263299942016602, + "learning_rate": 1.048817641085159e-05, + "loss": 0.034, + "step": 13213 + }, + { + "epoch": 1.4325672159583696, + "grad_norm": 6.415048122406006, + "learning_rate": 1.0487451037284202e-05, + "loss": 0.0478, + "step": 13214 + }, + { + "epoch": 1.4326756287944493, + "grad_norm": 0.7445715069770813, + "learning_rate": 1.0486725663716814e-05, + "loss": 0.0086, + "step": 13215 + }, + { + "epoch": 1.432784041630529, + "grad_norm": 0.8067322969436646, + "learning_rate": 1.0486000290149428e-05, + "loss": 0.0162, + "step": 13216 + }, + { + "epoch": 1.4328924544666088, + "grad_norm": 4.596406936645508, + "learning_rate": 1.048527491658204e-05, + "loss": 0.051, + "step": 13217 + }, + { + "epoch": 1.4330008673026886, + "grad_norm": 0.54643714427948, + "learning_rate": 1.0484549543014652e-05, + "loss": 0.0084, + "step": 13218 + }, + { + "epoch": 1.4331092801387684, + "grad_norm": 4.620246410369873, + "learning_rate": 1.0483824169447266e-05, + "loss": 0.0221, + "step": 13219 + }, + { + "epoch": 1.4332176929748481, + "grad_norm": 2.012425422668457, + "learning_rate": 1.0483098795879878e-05, + "loss": 0.0338, + "step": 13220 + }, + { + "epoch": 1.4333261058109281, + "grad_norm": 3.0407350063323975, + "learning_rate": 1.0482373422312493e-05, + "loss": 0.029, + "step": 13221 + }, + { + "epoch": 1.4334345186470079, + "grad_norm": 2.014235734939575, + "learning_rate": 1.0481648048745106e-05, + "loss": 0.007, + "step": 13222 + }, + { + "epoch": 1.4335429314830876, + "grad_norm": 0.34601274132728577, + "learning_rate": 1.0480922675177718e-05, + "loss": 0.0027, + "step": 13223 + }, + { + "epoch": 1.4336513443191674, + "grad_norm": 1.6482696533203125, + "learning_rate": 1.0480197301610332e-05, + "loss": 0.0157, + "step": 13224 + }, + { + "epoch": 1.4337597571552472, + "grad_norm": 2.08119535446167, + "learning_rate": 1.0479471928042944e-05, + "loss": 0.0187, + "step": 13225 + }, + { + "epoch": 1.433868169991327, + "grad_norm": 2.151407241821289, + "learning_rate": 1.0478746554475556e-05, + "loss": 0.0542, + "step": 13226 + }, + { + "epoch": 1.4339765828274067, + "grad_norm": 3.3273985385894775, + "learning_rate": 1.047802118090817e-05, + "loss": 0.0271, + "step": 13227 + }, + { + "epoch": 1.4340849956634867, + "grad_norm": 0.9249669909477234, + "learning_rate": 1.0477295807340782e-05, + "loss": 0.0077, + "step": 13228 + }, + { + "epoch": 1.4341934084995662, + "grad_norm": 1.6626310348510742, + "learning_rate": 1.0476570433773394e-05, + "loss": 0.0193, + "step": 13229 + }, + { + "epoch": 1.4343018213356462, + "grad_norm": 0.48182883858680725, + "learning_rate": 1.0475845060206006e-05, + "loss": 0.0064, + "step": 13230 + }, + { + "epoch": 1.434410234171726, + "grad_norm": 7.243904113769531, + "learning_rate": 1.047511968663862e-05, + "loss": 0.0398, + "step": 13231 + }, + { + "epoch": 1.4345186470078057, + "grad_norm": 1.3469388484954834, + "learning_rate": 1.0474394313071232e-05, + "loss": 0.0111, + "step": 13232 + }, + { + "epoch": 1.4346270598438855, + "grad_norm": 0.7934983968734741, + "learning_rate": 1.0473668939503844e-05, + "loss": 0.0057, + "step": 13233 + }, + { + "epoch": 1.4347354726799653, + "grad_norm": 0.24887876212596893, + "learning_rate": 1.0472943565936458e-05, + "loss": 0.0027, + "step": 13234 + }, + { + "epoch": 1.434843885516045, + "grad_norm": 2.750706672668457, + "learning_rate": 1.047221819236907e-05, + "loss": 0.0141, + "step": 13235 + }, + { + "epoch": 1.4349522983521248, + "grad_norm": 4.25213623046875, + "learning_rate": 1.0471492818801682e-05, + "loss": 0.0651, + "step": 13236 + }, + { + "epoch": 1.4350607111882048, + "grad_norm": 3.1263513565063477, + "learning_rate": 1.0470767445234296e-05, + "loss": 0.0347, + "step": 13237 + }, + { + "epoch": 1.4351691240242845, + "grad_norm": 2.4726791381835938, + "learning_rate": 1.0470042071666908e-05, + "loss": 0.0362, + "step": 13238 + }, + { + "epoch": 1.4352775368603643, + "grad_norm": 2.6384875774383545, + "learning_rate": 1.0469316698099524e-05, + "loss": 0.0319, + "step": 13239 + }, + { + "epoch": 1.435385949696444, + "grad_norm": 0.49662336707115173, + "learning_rate": 1.0468591324532136e-05, + "loss": 0.0051, + "step": 13240 + }, + { + "epoch": 1.4354943625325238, + "grad_norm": 0.5432146191596985, + "learning_rate": 1.0467865950964748e-05, + "loss": 0.0021, + "step": 13241 + }, + { + "epoch": 1.4356027753686036, + "grad_norm": 3.395495653152466, + "learning_rate": 1.046714057739736e-05, + "loss": 0.0189, + "step": 13242 + }, + { + "epoch": 1.4357111882046834, + "grad_norm": 1.2862961292266846, + "learning_rate": 1.0466415203829974e-05, + "loss": 0.0081, + "step": 13243 + }, + { + "epoch": 1.4358196010407633, + "grad_norm": 2.386077880859375, + "learning_rate": 1.0465689830262586e-05, + "loss": 0.034, + "step": 13244 + }, + { + "epoch": 1.435928013876843, + "grad_norm": 1.254560947418213, + "learning_rate": 1.0464964456695198e-05, + "loss": 0.0048, + "step": 13245 + }, + { + "epoch": 1.4360364267129229, + "grad_norm": 0.3680093288421631, + "learning_rate": 1.0464239083127812e-05, + "loss": 0.004, + "step": 13246 + }, + { + "epoch": 1.4361448395490026, + "grad_norm": 2.961237668991089, + "learning_rate": 1.0463513709560424e-05, + "loss": 0.0609, + "step": 13247 + }, + { + "epoch": 1.4362532523850824, + "grad_norm": 2.5651392936706543, + "learning_rate": 1.0462788335993037e-05, + "loss": 0.0157, + "step": 13248 + }, + { + "epoch": 1.4363616652211622, + "grad_norm": 0.24679873883724213, + "learning_rate": 1.046206296242565e-05, + "loss": 0.0028, + "step": 13249 + }, + { + "epoch": 1.436470078057242, + "grad_norm": 2.5701138973236084, + "learning_rate": 1.0461337588858263e-05, + "loss": 0.0507, + "step": 13250 + }, + { + "epoch": 1.436578490893322, + "grad_norm": 3.134051561355591, + "learning_rate": 1.0460612215290875e-05, + "loss": 0.0248, + "step": 13251 + }, + { + "epoch": 1.4366869037294014, + "grad_norm": 3.3657584190368652, + "learning_rate": 1.0459886841723489e-05, + "loss": 0.0352, + "step": 13252 + }, + { + "epoch": 1.4367953165654814, + "grad_norm": 0.7344945073127747, + "learning_rate": 1.04591614681561e-05, + "loss": 0.008, + "step": 13253 + }, + { + "epoch": 1.4369037294015612, + "grad_norm": 0.3947541117668152, + "learning_rate": 1.0458436094588713e-05, + "loss": 0.0044, + "step": 13254 + }, + { + "epoch": 1.437012142237641, + "grad_norm": 0.21220625936985016, + "learning_rate": 1.0457710721021325e-05, + "loss": 0.0024, + "step": 13255 + }, + { + "epoch": 1.4371205550737207, + "grad_norm": 3.4692273139953613, + "learning_rate": 1.045698534745394e-05, + "loss": 0.0086, + "step": 13256 + }, + { + "epoch": 1.4372289679098005, + "grad_norm": 4.259817123413086, + "learning_rate": 1.0456259973886553e-05, + "loss": 0.0235, + "step": 13257 + }, + { + "epoch": 1.4373373807458802, + "grad_norm": 5.957659721374512, + "learning_rate": 1.0455534600319166e-05, + "loss": 0.0891, + "step": 13258 + }, + { + "epoch": 1.43744579358196, + "grad_norm": 3.4056620597839355, + "learning_rate": 1.0454809226751779e-05, + "loss": 0.0128, + "step": 13259 + }, + { + "epoch": 1.43755420641804, + "grad_norm": 2.0218658447265625, + "learning_rate": 1.045408385318439e-05, + "loss": 0.0301, + "step": 13260 + }, + { + "epoch": 1.4376626192541198, + "grad_norm": 4.748575687408447, + "learning_rate": 1.0453358479617005e-05, + "loss": 0.0331, + "step": 13261 + }, + { + "epoch": 1.4377710320901995, + "grad_norm": 1.935227394104004, + "learning_rate": 1.0452633106049617e-05, + "loss": 0.023, + "step": 13262 + }, + { + "epoch": 1.4378794449262793, + "grad_norm": 3.932683229446411, + "learning_rate": 1.0451907732482229e-05, + "loss": 0.0339, + "step": 13263 + }, + { + "epoch": 1.437987857762359, + "grad_norm": 8.657546043395996, + "learning_rate": 1.0451182358914843e-05, + "loss": 0.0462, + "step": 13264 + }, + { + "epoch": 1.4380962705984388, + "grad_norm": 0.9134162664413452, + "learning_rate": 1.0450456985347455e-05, + "loss": 0.007, + "step": 13265 + }, + { + "epoch": 1.4382046834345186, + "grad_norm": 9.963912010192871, + "learning_rate": 1.0449731611780067e-05, + "loss": 0.0276, + "step": 13266 + }, + { + "epoch": 1.4383130962705986, + "grad_norm": 0.3074791133403778, + "learning_rate": 1.044900623821268e-05, + "loss": 0.0034, + "step": 13267 + }, + { + "epoch": 1.438421509106678, + "grad_norm": 1.9452850818634033, + "learning_rate": 1.0448280864645293e-05, + "loss": 0.0139, + "step": 13268 + }, + { + "epoch": 1.438529921942758, + "grad_norm": 1.4993027448654175, + "learning_rate": 1.0447555491077905e-05, + "loss": 0.0198, + "step": 13269 + }, + { + "epoch": 1.4386383347788378, + "grad_norm": 0.21366673707962036, + "learning_rate": 1.0446830117510517e-05, + "loss": 0.0017, + "step": 13270 + }, + { + "epoch": 1.4387467476149176, + "grad_norm": 0.8510813117027283, + "learning_rate": 1.0446104743943131e-05, + "loss": 0.0128, + "step": 13271 + }, + { + "epoch": 1.4388551604509974, + "grad_norm": 4.599380970001221, + "learning_rate": 1.0445379370375743e-05, + "loss": 0.0725, + "step": 13272 + }, + { + "epoch": 1.4389635732870771, + "grad_norm": 1.7283732891082764, + "learning_rate": 1.0444653996808359e-05, + "loss": 0.0138, + "step": 13273 + }, + { + "epoch": 1.439071986123157, + "grad_norm": 0.5637298226356506, + "learning_rate": 1.0443928623240971e-05, + "loss": 0.0113, + "step": 13274 + }, + { + "epoch": 1.4391803989592367, + "grad_norm": 2.2102155685424805, + "learning_rate": 1.0443203249673583e-05, + "loss": 0.0118, + "step": 13275 + }, + { + "epoch": 1.4392888117953166, + "grad_norm": 1.8776849508285522, + "learning_rate": 1.0442477876106197e-05, + "loss": 0.0068, + "step": 13276 + }, + { + "epoch": 1.4393972246313964, + "grad_norm": 3.1472902297973633, + "learning_rate": 1.0441752502538809e-05, + "loss": 0.0242, + "step": 13277 + }, + { + "epoch": 1.4395056374674762, + "grad_norm": 2.092175245285034, + "learning_rate": 1.0441027128971421e-05, + "loss": 0.0363, + "step": 13278 + }, + { + "epoch": 1.439614050303556, + "grad_norm": 2.177035093307495, + "learning_rate": 1.0440301755404035e-05, + "loss": 0.0275, + "step": 13279 + }, + { + "epoch": 1.4397224631396357, + "grad_norm": 1.386497974395752, + "learning_rate": 1.0439576381836647e-05, + "loss": 0.033, + "step": 13280 + }, + { + "epoch": 1.4398308759757155, + "grad_norm": 1.0789151191711426, + "learning_rate": 1.043885100826926e-05, + "loss": 0.0153, + "step": 13281 + }, + { + "epoch": 1.4399392888117952, + "grad_norm": 0.491199254989624, + "learning_rate": 1.0438125634701871e-05, + "loss": 0.0043, + "step": 13282 + }, + { + "epoch": 1.4400477016478752, + "grad_norm": 2.6540067195892334, + "learning_rate": 1.0437400261134485e-05, + "loss": 0.0112, + "step": 13283 + }, + { + "epoch": 1.440156114483955, + "grad_norm": 1.3355913162231445, + "learning_rate": 1.0436674887567097e-05, + "loss": 0.0048, + "step": 13284 + }, + { + "epoch": 1.4402645273200347, + "grad_norm": 2.036362886428833, + "learning_rate": 1.043594951399971e-05, + "loss": 0.0511, + "step": 13285 + }, + { + "epoch": 1.4403729401561145, + "grad_norm": 1.4771783351898193, + "learning_rate": 1.0435224140432323e-05, + "loss": 0.013, + "step": 13286 + }, + { + "epoch": 1.4404813529921943, + "grad_norm": 0.5813188552856445, + "learning_rate": 1.0434498766864936e-05, + "loss": 0.008, + "step": 13287 + }, + { + "epoch": 1.440589765828274, + "grad_norm": 3.923888683319092, + "learning_rate": 1.0433773393297548e-05, + "loss": 0.0271, + "step": 13288 + }, + { + "epoch": 1.4406981786643538, + "grad_norm": 3.335798501968384, + "learning_rate": 1.0433048019730162e-05, + "loss": 0.0345, + "step": 13289 + }, + { + "epoch": 1.4408065915004338, + "grad_norm": 1.1800220012664795, + "learning_rate": 1.0432322646162775e-05, + "loss": 0.0102, + "step": 13290 + }, + { + "epoch": 1.4409150043365133, + "grad_norm": 1.1035100221633911, + "learning_rate": 1.043159727259539e-05, + "loss": 0.0056, + "step": 13291 + }, + { + "epoch": 1.4410234171725933, + "grad_norm": 2.5369935035705566, + "learning_rate": 1.0430871899028001e-05, + "loss": 0.0143, + "step": 13292 + }, + { + "epoch": 1.441131830008673, + "grad_norm": 0.3107411861419678, + "learning_rate": 1.0430146525460614e-05, + "loss": 0.0028, + "step": 13293 + }, + { + "epoch": 1.4412402428447528, + "grad_norm": 6.525786876678467, + "learning_rate": 1.0429421151893226e-05, + "loss": 0.0113, + "step": 13294 + }, + { + "epoch": 1.4413486556808326, + "grad_norm": 0.9396243691444397, + "learning_rate": 1.042869577832584e-05, + "loss": 0.0163, + "step": 13295 + }, + { + "epoch": 1.4414570685169124, + "grad_norm": 2.6286635398864746, + "learning_rate": 1.0427970404758452e-05, + "loss": 0.0168, + "step": 13296 + }, + { + "epoch": 1.4415654813529921, + "grad_norm": 0.36121368408203125, + "learning_rate": 1.0427245031191064e-05, + "loss": 0.0017, + "step": 13297 + }, + { + "epoch": 1.4416738941890719, + "grad_norm": 5.0764031410217285, + "learning_rate": 1.0426519657623678e-05, + "loss": 0.0143, + "step": 13298 + }, + { + "epoch": 1.4417823070251519, + "grad_norm": 7.001651763916016, + "learning_rate": 1.042579428405629e-05, + "loss": 0.0437, + "step": 13299 + }, + { + "epoch": 1.4418907198612316, + "grad_norm": 2.6496081352233887, + "learning_rate": 1.0425068910488902e-05, + "loss": 0.0096, + "step": 13300 + }, + { + "epoch": 1.4419991326973114, + "grad_norm": 1.540818452835083, + "learning_rate": 1.0424343536921516e-05, + "loss": 0.0628, + "step": 13301 + }, + { + "epoch": 1.4421075455333912, + "grad_norm": 1.6658222675323486, + "learning_rate": 1.0423618163354128e-05, + "loss": 0.0241, + "step": 13302 + }, + { + "epoch": 1.442215958369471, + "grad_norm": 2.367523193359375, + "learning_rate": 1.042289278978674e-05, + "loss": 0.0585, + "step": 13303 + }, + { + "epoch": 1.4423243712055507, + "grad_norm": 2.301872730255127, + "learning_rate": 1.0422167416219354e-05, + "loss": 0.0616, + "step": 13304 + }, + { + "epoch": 1.4424327840416304, + "grad_norm": 2.758660316467285, + "learning_rate": 1.0421442042651966e-05, + "loss": 0.0192, + "step": 13305 + }, + { + "epoch": 1.4425411968777104, + "grad_norm": 3.4026803970336914, + "learning_rate": 1.0420716669084578e-05, + "loss": 0.0612, + "step": 13306 + }, + { + "epoch": 1.4426496097137902, + "grad_norm": 0.45740488171577454, + "learning_rate": 1.041999129551719e-05, + "loss": 0.0049, + "step": 13307 + }, + { + "epoch": 1.44275802254987, + "grad_norm": 3.0613796710968018, + "learning_rate": 1.0419265921949806e-05, + "loss": 0.011, + "step": 13308 + }, + { + "epoch": 1.4428664353859497, + "grad_norm": 1.9631109237670898, + "learning_rate": 1.0418540548382418e-05, + "loss": 0.0151, + "step": 13309 + }, + { + "epoch": 1.4429748482220295, + "grad_norm": 0.7413866519927979, + "learning_rate": 1.0417815174815032e-05, + "loss": 0.0055, + "step": 13310 + }, + { + "epoch": 1.4430832610581092, + "grad_norm": 3.732710599899292, + "learning_rate": 1.0417089801247644e-05, + "loss": 0.0163, + "step": 13311 + }, + { + "epoch": 1.443191673894189, + "grad_norm": 4.891359329223633, + "learning_rate": 1.0416364427680256e-05, + "loss": 0.0311, + "step": 13312 + }, + { + "epoch": 1.443300086730269, + "grad_norm": 2.3776183128356934, + "learning_rate": 1.041563905411287e-05, + "loss": 0.0372, + "step": 13313 + }, + { + "epoch": 1.4434084995663485, + "grad_norm": 7.084397315979004, + "learning_rate": 1.0414913680545482e-05, + "loss": 0.0322, + "step": 13314 + }, + { + "epoch": 1.4435169124024285, + "grad_norm": 2.407865285873413, + "learning_rate": 1.0414188306978094e-05, + "loss": 0.0071, + "step": 13315 + }, + { + "epoch": 1.4436253252385083, + "grad_norm": 1.1012078523635864, + "learning_rate": 1.0413462933410708e-05, + "loss": 0.0208, + "step": 13316 + }, + { + "epoch": 1.443733738074588, + "grad_norm": 1.8904839754104614, + "learning_rate": 1.041273755984332e-05, + "loss": 0.0189, + "step": 13317 + }, + { + "epoch": 1.4438421509106678, + "grad_norm": 10.284232139587402, + "learning_rate": 1.0412012186275932e-05, + "loss": 0.0422, + "step": 13318 + }, + { + "epoch": 1.4439505637467476, + "grad_norm": 0.2711775004863739, + "learning_rate": 1.0411286812708545e-05, + "loss": 0.0034, + "step": 13319 + }, + { + "epoch": 1.4440589765828273, + "grad_norm": 1.7686103582382202, + "learning_rate": 1.0410561439141158e-05, + "loss": 0.017, + "step": 13320 + }, + { + "epoch": 1.444167389418907, + "grad_norm": 1.7123041152954102, + "learning_rate": 1.040983606557377e-05, + "loss": 0.0068, + "step": 13321 + }, + { + "epoch": 1.444275802254987, + "grad_norm": 8.993258476257324, + "learning_rate": 1.0409110692006383e-05, + "loss": 0.0327, + "step": 13322 + }, + { + "epoch": 1.4443842150910668, + "grad_norm": 1.4158549308776855, + "learning_rate": 1.0408385318438996e-05, + "loss": 0.0273, + "step": 13323 + }, + { + "epoch": 1.4444926279271466, + "grad_norm": 3.6778833866119385, + "learning_rate": 1.0407659944871609e-05, + "loss": 0.0387, + "step": 13324 + }, + { + "epoch": 1.4446010407632264, + "grad_norm": 2.6977334022521973, + "learning_rate": 1.0406934571304224e-05, + "loss": 0.021, + "step": 13325 + }, + { + "epoch": 1.4447094535993061, + "grad_norm": 3.972500801086426, + "learning_rate": 1.0406209197736836e-05, + "loss": 0.037, + "step": 13326 + }, + { + "epoch": 1.444817866435386, + "grad_norm": 1.6884757280349731, + "learning_rate": 1.0405483824169448e-05, + "loss": 0.0171, + "step": 13327 + }, + { + "epoch": 1.4449262792714657, + "grad_norm": 2.8198578357696533, + "learning_rate": 1.0404758450602062e-05, + "loss": 0.0238, + "step": 13328 + }, + { + "epoch": 1.4450346921075456, + "grad_norm": 0.26386210322380066, + "learning_rate": 1.0404033077034674e-05, + "loss": 0.0014, + "step": 13329 + }, + { + "epoch": 1.4451431049436252, + "grad_norm": 6.833298206329346, + "learning_rate": 1.0403307703467287e-05, + "loss": 0.0447, + "step": 13330 + }, + { + "epoch": 1.4452515177797052, + "grad_norm": 4.046966075897217, + "learning_rate": 1.04025823298999e-05, + "loss": 0.0577, + "step": 13331 + }, + { + "epoch": 1.445359930615785, + "grad_norm": 1.1686865091323853, + "learning_rate": 1.0401856956332513e-05, + "loss": 0.0109, + "step": 13332 + }, + { + "epoch": 1.4454683434518647, + "grad_norm": 0.8529655933380127, + "learning_rate": 1.0401131582765125e-05, + "loss": 0.0151, + "step": 13333 + }, + { + "epoch": 1.4455767562879445, + "grad_norm": 4.684746742248535, + "learning_rate": 1.0400406209197737e-05, + "loss": 0.0887, + "step": 13334 + }, + { + "epoch": 1.4456851691240242, + "grad_norm": 1.5837562084197998, + "learning_rate": 1.039968083563035e-05, + "loss": 0.0123, + "step": 13335 + }, + { + "epoch": 1.4457935819601042, + "grad_norm": 4.834225654602051, + "learning_rate": 1.0398955462062963e-05, + "loss": 0.0227, + "step": 13336 + }, + { + "epoch": 1.4459019947961838, + "grad_norm": 1.065320372581482, + "learning_rate": 1.0398230088495575e-05, + "loss": 0.0048, + "step": 13337 + }, + { + "epoch": 1.4460104076322637, + "grad_norm": 2.183051109313965, + "learning_rate": 1.0397504714928189e-05, + "loss": 0.0073, + "step": 13338 + }, + { + "epoch": 1.4461188204683435, + "grad_norm": 0.5378861427307129, + "learning_rate": 1.0396779341360801e-05, + "loss": 0.0049, + "step": 13339 + }, + { + "epoch": 1.4462272333044233, + "grad_norm": 5.703693389892578, + "learning_rate": 1.0396053967793413e-05, + "loss": 0.0289, + "step": 13340 + }, + { + "epoch": 1.446335646140503, + "grad_norm": 2.2838258743286133, + "learning_rate": 1.0395328594226027e-05, + "loss": 0.0282, + "step": 13341 + }, + { + "epoch": 1.4464440589765828, + "grad_norm": 9.913546562194824, + "learning_rate": 1.039460322065864e-05, + "loss": 0.0178, + "step": 13342 + }, + { + "epoch": 1.4465524718126626, + "grad_norm": 2.172369956970215, + "learning_rate": 1.0393877847091255e-05, + "loss": 0.0237, + "step": 13343 + }, + { + "epoch": 1.4466608846487423, + "grad_norm": 4.14707088470459, + "learning_rate": 1.0393152473523867e-05, + "loss": 0.0148, + "step": 13344 + }, + { + "epoch": 1.4467692974848223, + "grad_norm": 2.5386147499084473, + "learning_rate": 1.0392427099956479e-05, + "loss": 0.0177, + "step": 13345 + }, + { + "epoch": 1.446877710320902, + "grad_norm": 0.31711333990097046, + "learning_rate": 1.0391701726389091e-05, + "loss": 0.0055, + "step": 13346 + }, + { + "epoch": 1.4469861231569818, + "grad_norm": 2.642122983932495, + "learning_rate": 1.0390976352821705e-05, + "loss": 0.0152, + "step": 13347 + }, + { + "epoch": 1.4470945359930616, + "grad_norm": 6.296942234039307, + "learning_rate": 1.0390250979254317e-05, + "loss": 0.0512, + "step": 13348 + }, + { + "epoch": 1.4472029488291414, + "grad_norm": 0.7749602198600769, + "learning_rate": 1.0389525605686929e-05, + "loss": 0.0101, + "step": 13349 + }, + { + "epoch": 1.4473113616652211, + "grad_norm": 4.187204360961914, + "learning_rate": 1.0388800232119543e-05, + "loss": 0.0176, + "step": 13350 + }, + { + "epoch": 1.4474197745013009, + "grad_norm": 0.8319427967071533, + "learning_rate": 1.0388074858552155e-05, + "loss": 0.009, + "step": 13351 + }, + { + "epoch": 1.4475281873373809, + "grad_norm": 1.190983533859253, + "learning_rate": 1.0387349484984767e-05, + "loss": 0.0131, + "step": 13352 + }, + { + "epoch": 1.4476366001734604, + "grad_norm": 4.981357574462891, + "learning_rate": 1.0386624111417381e-05, + "loss": 0.0809, + "step": 13353 + }, + { + "epoch": 1.4477450130095404, + "grad_norm": 3.6585659980773926, + "learning_rate": 1.0385898737849993e-05, + "loss": 0.0569, + "step": 13354 + }, + { + "epoch": 1.4478534258456202, + "grad_norm": 0.3332812190055847, + "learning_rate": 1.0385173364282605e-05, + "loss": 0.0081, + "step": 13355 + }, + { + "epoch": 1.4479618386817, + "grad_norm": 0.5132730603218079, + "learning_rate": 1.038444799071522e-05, + "loss": 0.0035, + "step": 13356 + }, + { + "epoch": 1.4480702515177797, + "grad_norm": 3.0084383487701416, + "learning_rate": 1.0383722617147831e-05, + "loss": 0.0104, + "step": 13357 + }, + { + "epoch": 1.4481786643538594, + "grad_norm": 0.6810177564620972, + "learning_rate": 1.0382997243580444e-05, + "loss": 0.0042, + "step": 13358 + }, + { + "epoch": 1.4482870771899392, + "grad_norm": 0.43651193380355835, + "learning_rate": 1.0382271870013056e-05, + "loss": 0.0029, + "step": 13359 + }, + { + "epoch": 1.448395490026019, + "grad_norm": 0.9095348119735718, + "learning_rate": 1.0381546496445671e-05, + "loss": 0.0029, + "step": 13360 + }, + { + "epoch": 1.448503902862099, + "grad_norm": 1.9092094898223877, + "learning_rate": 1.0380821122878283e-05, + "loss": 0.0155, + "step": 13361 + }, + { + "epoch": 1.4486123156981787, + "grad_norm": 10.554791450500488, + "learning_rate": 1.0380095749310897e-05, + "loss": 0.0506, + "step": 13362 + }, + { + "epoch": 1.4487207285342585, + "grad_norm": 1.1489535570144653, + "learning_rate": 1.037937037574351e-05, + "loss": 0.0103, + "step": 13363 + }, + { + "epoch": 1.4488291413703382, + "grad_norm": 0.1417556256055832, + "learning_rate": 1.0378645002176121e-05, + "loss": 0.0011, + "step": 13364 + }, + { + "epoch": 1.448937554206418, + "grad_norm": 3.048945426940918, + "learning_rate": 1.0377919628608735e-05, + "loss": 0.0354, + "step": 13365 + }, + { + "epoch": 1.4490459670424978, + "grad_norm": 0.8703384399414062, + "learning_rate": 1.0377194255041347e-05, + "loss": 0.0036, + "step": 13366 + }, + { + "epoch": 1.4491543798785775, + "grad_norm": 0.5071571469306946, + "learning_rate": 1.037646888147396e-05, + "loss": 0.0026, + "step": 13367 + }, + { + "epoch": 1.4492627927146575, + "grad_norm": 2.9860169887542725, + "learning_rate": 1.0375743507906573e-05, + "loss": 0.0386, + "step": 13368 + }, + { + "epoch": 1.4493712055507373, + "grad_norm": 3.502014636993408, + "learning_rate": 1.0375018134339186e-05, + "loss": 0.0318, + "step": 13369 + }, + { + "epoch": 1.449479618386817, + "grad_norm": 0.6995792388916016, + "learning_rate": 1.0374292760771798e-05, + "loss": 0.0045, + "step": 13370 + }, + { + "epoch": 1.4495880312228968, + "grad_norm": 5.497830867767334, + "learning_rate": 1.037356738720441e-05, + "loss": 0.0317, + "step": 13371 + }, + { + "epoch": 1.4496964440589766, + "grad_norm": 0.09778746962547302, + "learning_rate": 1.0372842013637024e-05, + "loss": 0.0008, + "step": 13372 + }, + { + "epoch": 1.4498048568950563, + "grad_norm": 0.14365151524543762, + "learning_rate": 1.0372116640069636e-05, + "loss": 0.0014, + "step": 13373 + }, + { + "epoch": 1.449913269731136, + "grad_norm": 2.0663955211639404, + "learning_rate": 1.0371391266502248e-05, + "loss": 0.012, + "step": 13374 + }, + { + "epoch": 1.450021682567216, + "grad_norm": 2.768495559692383, + "learning_rate": 1.0370665892934862e-05, + "loss": 0.0147, + "step": 13375 + }, + { + "epoch": 1.4501300954032956, + "grad_norm": 4.12230110168457, + "learning_rate": 1.0369940519367474e-05, + "loss": 0.025, + "step": 13376 + }, + { + "epoch": 1.4502385082393756, + "grad_norm": 1.2648285627365112, + "learning_rate": 1.036921514580009e-05, + "loss": 0.0145, + "step": 13377 + }, + { + "epoch": 1.4503469210754554, + "grad_norm": 4.676119327545166, + "learning_rate": 1.0368489772232702e-05, + "loss": 0.0366, + "step": 13378 + }, + { + "epoch": 1.4504553339115351, + "grad_norm": 0.530485212802887, + "learning_rate": 1.0367764398665314e-05, + "loss": 0.0048, + "step": 13379 + }, + { + "epoch": 1.450563746747615, + "grad_norm": 2.0759787559509277, + "learning_rate": 1.0367039025097928e-05, + "loss": 0.0443, + "step": 13380 + }, + { + "epoch": 1.4506721595836947, + "grad_norm": 2.1541495323181152, + "learning_rate": 1.036631365153054e-05, + "loss": 0.0067, + "step": 13381 + }, + { + "epoch": 1.4507805724197744, + "grad_norm": 9.127037048339844, + "learning_rate": 1.0365588277963152e-05, + "loss": 0.0362, + "step": 13382 + }, + { + "epoch": 1.4508889852558542, + "grad_norm": 1.0534318685531616, + "learning_rate": 1.0364862904395766e-05, + "loss": 0.0122, + "step": 13383 + }, + { + "epoch": 1.4509973980919342, + "grad_norm": 0.026007385924458504, + "learning_rate": 1.0364137530828378e-05, + "loss": 0.0001, + "step": 13384 + }, + { + "epoch": 1.451105810928014, + "grad_norm": 2.57802677154541, + "learning_rate": 1.036341215726099e-05, + "loss": 0.0284, + "step": 13385 + }, + { + "epoch": 1.4512142237640937, + "grad_norm": 3.896221160888672, + "learning_rate": 1.0362686783693602e-05, + "loss": 0.0233, + "step": 13386 + }, + { + "epoch": 1.4513226366001735, + "grad_norm": 5.29439640045166, + "learning_rate": 1.0361961410126216e-05, + "loss": 0.005, + "step": 13387 + }, + { + "epoch": 1.4514310494362532, + "grad_norm": 3.1093151569366455, + "learning_rate": 1.0361236036558828e-05, + "loss": 0.0078, + "step": 13388 + }, + { + "epoch": 1.451539462272333, + "grad_norm": 1.678244709968567, + "learning_rate": 1.036051066299144e-05, + "loss": 0.0431, + "step": 13389 + }, + { + "epoch": 1.4516478751084128, + "grad_norm": 3.9804396629333496, + "learning_rate": 1.0359785289424054e-05, + "loss": 0.0387, + "step": 13390 + }, + { + "epoch": 1.4517562879444927, + "grad_norm": 1.367019534111023, + "learning_rate": 1.0359059915856666e-05, + "loss": 0.0364, + "step": 13391 + }, + { + "epoch": 1.4518647007805723, + "grad_norm": 1.7650898694992065, + "learning_rate": 1.0358334542289278e-05, + "loss": 0.0208, + "step": 13392 + }, + { + "epoch": 1.4519731136166523, + "grad_norm": 2.0098235607147217, + "learning_rate": 1.0357609168721892e-05, + "loss": 0.0279, + "step": 13393 + }, + { + "epoch": 1.452081526452732, + "grad_norm": 0.8237881660461426, + "learning_rate": 1.0356883795154506e-05, + "loss": 0.0168, + "step": 13394 + }, + { + "epoch": 1.4521899392888118, + "grad_norm": 1.3929933309555054, + "learning_rate": 1.035615842158712e-05, + "loss": 0.0119, + "step": 13395 + }, + { + "epoch": 1.4522983521248916, + "grad_norm": 1.06464684009552, + "learning_rate": 1.0355433048019732e-05, + "loss": 0.0132, + "step": 13396 + }, + { + "epoch": 1.4524067649609713, + "grad_norm": 3.980579137802124, + "learning_rate": 1.0354707674452344e-05, + "loss": 0.0313, + "step": 13397 + }, + { + "epoch": 1.4525151777970513, + "grad_norm": 2.7141127586364746, + "learning_rate": 1.0353982300884956e-05, + "loss": 0.0078, + "step": 13398 + }, + { + "epoch": 1.4526235906331308, + "grad_norm": 4.5867228507995605, + "learning_rate": 1.035325692731757e-05, + "loss": 0.03, + "step": 13399 + }, + { + "epoch": 1.4527320034692108, + "grad_norm": 5.7028117179870605, + "learning_rate": 1.0352531553750182e-05, + "loss": 0.0136, + "step": 13400 + }, + { + "epoch": 1.4528404163052906, + "grad_norm": 3.814584732055664, + "learning_rate": 1.0351806180182794e-05, + "loss": 0.0594, + "step": 13401 + }, + { + "epoch": 1.4529488291413704, + "grad_norm": 2.9624457359313965, + "learning_rate": 1.0351080806615408e-05, + "loss": 0.0163, + "step": 13402 + }, + { + "epoch": 1.4530572419774501, + "grad_norm": 5.8720316886901855, + "learning_rate": 1.035035543304802e-05, + "loss": 0.0514, + "step": 13403 + }, + { + "epoch": 1.4531656548135299, + "grad_norm": 0.28131264448165894, + "learning_rate": 1.0349630059480633e-05, + "loss": 0.0014, + "step": 13404 + }, + { + "epoch": 1.4532740676496096, + "grad_norm": 0.7257330417633057, + "learning_rate": 1.0348904685913246e-05, + "loss": 0.0062, + "step": 13405 + }, + { + "epoch": 1.4533824804856894, + "grad_norm": 0.22882212698459625, + "learning_rate": 1.0348179312345859e-05, + "loss": 0.0018, + "step": 13406 + }, + { + "epoch": 1.4534908933217694, + "grad_norm": 0.6152474284172058, + "learning_rate": 1.034745393877847e-05, + "loss": 0.0023, + "step": 13407 + }, + { + "epoch": 1.4535993061578492, + "grad_norm": 0.6604790687561035, + "learning_rate": 1.0346728565211085e-05, + "loss": 0.0042, + "step": 13408 + }, + { + "epoch": 1.453707718993929, + "grad_norm": 9.356086730957031, + "learning_rate": 1.0346003191643697e-05, + "loss": 0.0395, + "step": 13409 + }, + { + "epoch": 1.4538161318300087, + "grad_norm": 3.223620891571045, + "learning_rate": 1.0345277818076309e-05, + "loss": 0.0541, + "step": 13410 + }, + { + "epoch": 1.4539245446660884, + "grad_norm": 2.874915838241577, + "learning_rate": 1.0344552444508924e-05, + "loss": 0.0255, + "step": 13411 + }, + { + "epoch": 1.4540329575021682, + "grad_norm": 3.542708396911621, + "learning_rate": 1.0343827070941537e-05, + "loss": 0.0184, + "step": 13412 + }, + { + "epoch": 1.454141370338248, + "grad_norm": 0.9175704121589661, + "learning_rate": 1.0343101697374149e-05, + "loss": 0.006, + "step": 13413 + }, + { + "epoch": 1.454249783174328, + "grad_norm": 0.7119910717010498, + "learning_rate": 1.0342376323806763e-05, + "loss": 0.0026, + "step": 13414 + }, + { + "epoch": 1.4543581960104075, + "grad_norm": 0.5091844797134399, + "learning_rate": 1.0341650950239375e-05, + "loss": 0.0039, + "step": 13415 + }, + { + "epoch": 1.4544666088464875, + "grad_norm": 0.35023701190948486, + "learning_rate": 1.0340925576671987e-05, + "loss": 0.0034, + "step": 13416 + }, + { + "epoch": 1.4545750216825672, + "grad_norm": 2.5116052627563477, + "learning_rate": 1.03402002031046e-05, + "loss": 0.0155, + "step": 13417 + }, + { + "epoch": 1.454683434518647, + "grad_norm": 0.305433452129364, + "learning_rate": 1.0339474829537213e-05, + "loss": 0.0021, + "step": 13418 + }, + { + "epoch": 1.4547918473547268, + "grad_norm": 3.826442241668701, + "learning_rate": 1.0338749455969825e-05, + "loss": 0.0133, + "step": 13419 + }, + { + "epoch": 1.4549002601908065, + "grad_norm": 1.701577067375183, + "learning_rate": 1.0338024082402439e-05, + "loss": 0.0256, + "step": 13420 + }, + { + "epoch": 1.4550086730268863, + "grad_norm": 2.4897677898406982, + "learning_rate": 1.0337298708835051e-05, + "loss": 0.0364, + "step": 13421 + }, + { + "epoch": 1.455117085862966, + "grad_norm": 0.15196001529693604, + "learning_rate": 1.0336573335267663e-05, + "loss": 0.0003, + "step": 13422 + }, + { + "epoch": 1.455225498699046, + "grad_norm": 0.7596310973167419, + "learning_rate": 1.0335847961700277e-05, + "loss": 0.0081, + "step": 13423 + }, + { + "epoch": 1.4553339115351258, + "grad_norm": 0.5403118133544922, + "learning_rate": 1.0335122588132889e-05, + "loss": 0.0083, + "step": 13424 + }, + { + "epoch": 1.4554423243712056, + "grad_norm": 1.891972303390503, + "learning_rate": 1.0334397214565501e-05, + "loss": 0.0075, + "step": 13425 + }, + { + "epoch": 1.4555507372072853, + "grad_norm": 0.9168118834495544, + "learning_rate": 1.0333671840998113e-05, + "loss": 0.005, + "step": 13426 + }, + { + "epoch": 1.455659150043365, + "grad_norm": 2.027693271636963, + "learning_rate": 1.0332946467430727e-05, + "loss": 0.0171, + "step": 13427 + }, + { + "epoch": 1.4557675628794449, + "grad_norm": 3.0096590518951416, + "learning_rate": 1.033222109386334e-05, + "loss": 0.0475, + "step": 13428 + }, + { + "epoch": 1.4558759757155246, + "grad_norm": 2.007378339767456, + "learning_rate": 1.0331495720295955e-05, + "loss": 0.0383, + "step": 13429 + }, + { + "epoch": 1.4559843885516046, + "grad_norm": 1.2053903341293335, + "learning_rate": 1.0330770346728567e-05, + "loss": 0.0121, + "step": 13430 + }, + { + "epoch": 1.4560928013876844, + "grad_norm": 0.22666433453559875, + "learning_rate": 1.0330044973161179e-05, + "loss": 0.0021, + "step": 13431 + }, + { + "epoch": 1.4562012142237641, + "grad_norm": 0.4000251293182373, + "learning_rate": 1.0329319599593793e-05, + "loss": 0.0017, + "step": 13432 + }, + { + "epoch": 1.456309627059844, + "grad_norm": 2.369323253631592, + "learning_rate": 1.0328594226026405e-05, + "loss": 0.0241, + "step": 13433 + }, + { + "epoch": 1.4564180398959237, + "grad_norm": 1.8697985410690308, + "learning_rate": 1.0327868852459017e-05, + "loss": 0.0189, + "step": 13434 + }, + { + "epoch": 1.4565264527320034, + "grad_norm": 0.09618739783763885, + "learning_rate": 1.0327143478891631e-05, + "loss": 0.0015, + "step": 13435 + }, + { + "epoch": 1.4566348655680832, + "grad_norm": 3.4392037391662598, + "learning_rate": 1.0326418105324243e-05, + "loss": 0.0275, + "step": 13436 + }, + { + "epoch": 1.4567432784041632, + "grad_norm": 2.1070797443389893, + "learning_rate": 1.0325692731756855e-05, + "loss": 0.022, + "step": 13437 + }, + { + "epoch": 1.4568516912402427, + "grad_norm": 4.384252071380615, + "learning_rate": 1.0324967358189468e-05, + "loss": 0.0152, + "step": 13438 + }, + { + "epoch": 1.4569601040763227, + "grad_norm": 5.938276290893555, + "learning_rate": 1.0324241984622081e-05, + "loss": 0.0191, + "step": 13439 + }, + { + "epoch": 1.4570685169124025, + "grad_norm": 2.244091749191284, + "learning_rate": 1.0323516611054693e-05, + "loss": 0.0202, + "step": 13440 + }, + { + "epoch": 1.4571769297484822, + "grad_norm": 1.3153778314590454, + "learning_rate": 1.0322791237487306e-05, + "loss": 0.0046, + "step": 13441 + }, + { + "epoch": 1.457285342584562, + "grad_norm": 4.604784965515137, + "learning_rate": 1.032206586391992e-05, + "loss": 0.0567, + "step": 13442 + }, + { + "epoch": 1.4573937554206418, + "grad_norm": 1.203913688659668, + "learning_rate": 1.0321340490352532e-05, + "loss": 0.018, + "step": 13443 + }, + { + "epoch": 1.4575021682567215, + "grad_norm": 2.474358558654785, + "learning_rate": 1.0320615116785144e-05, + "loss": 0.005, + "step": 13444 + }, + { + "epoch": 1.4576105810928013, + "grad_norm": 7.5012359619140625, + "learning_rate": 1.0319889743217758e-05, + "loss": 0.0426, + "step": 13445 + }, + { + "epoch": 1.4577189939288813, + "grad_norm": 0.13020889461040497, + "learning_rate": 1.0319164369650371e-05, + "loss": 0.0015, + "step": 13446 + }, + { + "epoch": 1.457827406764961, + "grad_norm": 2.053496837615967, + "learning_rate": 1.0318438996082985e-05, + "loss": 0.0011, + "step": 13447 + }, + { + "epoch": 1.4579358196010408, + "grad_norm": 0.037665218114852905, + "learning_rate": 1.0317713622515597e-05, + "loss": 0.0003, + "step": 13448 + }, + { + "epoch": 1.4580442324371206, + "grad_norm": 1.9856663942337036, + "learning_rate": 1.031698824894821e-05, + "loss": 0.0124, + "step": 13449 + }, + { + "epoch": 1.4581526452732003, + "grad_norm": 2.254204034805298, + "learning_rate": 1.0316262875380822e-05, + "loss": 0.0307, + "step": 13450 + }, + { + "epoch": 1.45826105810928, + "grad_norm": 5.4092183113098145, + "learning_rate": 1.0315537501813436e-05, + "loss": 0.0751, + "step": 13451 + }, + { + "epoch": 1.4583694709453598, + "grad_norm": 1.7256851196289062, + "learning_rate": 1.0314812128246048e-05, + "loss": 0.0281, + "step": 13452 + }, + { + "epoch": 1.4584778837814398, + "grad_norm": 2.2596352100372314, + "learning_rate": 1.031408675467866e-05, + "loss": 0.036, + "step": 13453 + }, + { + "epoch": 1.4585862966175196, + "grad_norm": 0.3105604648590088, + "learning_rate": 1.0313361381111274e-05, + "loss": 0.004, + "step": 13454 + }, + { + "epoch": 1.4586947094535994, + "grad_norm": 5.930051803588867, + "learning_rate": 1.0312636007543886e-05, + "loss": 0.0172, + "step": 13455 + }, + { + "epoch": 1.4588031222896791, + "grad_norm": 4.215670585632324, + "learning_rate": 1.0311910633976498e-05, + "loss": 0.0588, + "step": 13456 + }, + { + "epoch": 1.4589115351257589, + "grad_norm": 6.163724899291992, + "learning_rate": 1.0311185260409112e-05, + "loss": 0.0376, + "step": 13457 + }, + { + "epoch": 1.4590199479618386, + "grad_norm": 1.4593415260314941, + "learning_rate": 1.0310459886841724e-05, + "loss": 0.0189, + "step": 13458 + }, + { + "epoch": 1.4591283607979184, + "grad_norm": 2.033503293991089, + "learning_rate": 1.0309734513274336e-05, + "loss": 0.0086, + "step": 13459 + }, + { + "epoch": 1.4592367736339984, + "grad_norm": 3.37739896774292, + "learning_rate": 1.030900913970695e-05, + "loss": 0.0117, + "step": 13460 + }, + { + "epoch": 1.459345186470078, + "grad_norm": 3.0045289993286133, + "learning_rate": 1.0308283766139562e-05, + "loss": 0.0231, + "step": 13461 + }, + { + "epoch": 1.459453599306158, + "grad_norm": 7.523885726928711, + "learning_rate": 1.0307558392572174e-05, + "loss": 0.0821, + "step": 13462 + }, + { + "epoch": 1.4595620121422377, + "grad_norm": 2.0882985591888428, + "learning_rate": 1.030683301900479e-05, + "loss": 0.0378, + "step": 13463 + }, + { + "epoch": 1.4596704249783174, + "grad_norm": 4.837644577026367, + "learning_rate": 1.0306107645437402e-05, + "loss": 0.018, + "step": 13464 + }, + { + "epoch": 1.4597788378143972, + "grad_norm": 4.189456939697266, + "learning_rate": 1.0305382271870014e-05, + "loss": 0.0345, + "step": 13465 + }, + { + "epoch": 1.459887250650477, + "grad_norm": 1.6830470561981201, + "learning_rate": 1.0304656898302628e-05, + "loss": 0.0186, + "step": 13466 + }, + { + "epoch": 1.4599956634865567, + "grad_norm": 1.8700215816497803, + "learning_rate": 1.030393152473524e-05, + "loss": 0.0136, + "step": 13467 + }, + { + "epoch": 1.4601040763226365, + "grad_norm": 2.5988192558288574, + "learning_rate": 1.0303206151167852e-05, + "loss": 0.0088, + "step": 13468 + }, + { + "epoch": 1.4602124891587165, + "grad_norm": 1.1859524250030518, + "learning_rate": 1.0302480777600466e-05, + "loss": 0.0109, + "step": 13469 + }, + { + "epoch": 1.4603209019947962, + "grad_norm": 3.995414972305298, + "learning_rate": 1.0301755404033078e-05, + "loss": 0.0823, + "step": 13470 + }, + { + "epoch": 1.460429314830876, + "grad_norm": 3.996455430984497, + "learning_rate": 1.030103003046569e-05, + "loss": 0.0217, + "step": 13471 + }, + { + "epoch": 1.4605377276669558, + "grad_norm": 0.9844216704368591, + "learning_rate": 1.0300304656898304e-05, + "loss": 0.0036, + "step": 13472 + }, + { + "epoch": 1.4606461405030355, + "grad_norm": 1.7082908153533936, + "learning_rate": 1.0299579283330916e-05, + "loss": 0.0114, + "step": 13473 + }, + { + "epoch": 1.4607545533391153, + "grad_norm": 1.3717284202575684, + "learning_rate": 1.0298853909763528e-05, + "loss": 0.0163, + "step": 13474 + }, + { + "epoch": 1.460862966175195, + "grad_norm": 0.3752720057964325, + "learning_rate": 1.0298128536196142e-05, + "loss": 0.004, + "step": 13475 + }, + { + "epoch": 1.460971379011275, + "grad_norm": 3.5370709896087646, + "learning_rate": 1.0297403162628754e-05, + "loss": 0.028, + "step": 13476 + }, + { + "epoch": 1.4610797918473546, + "grad_norm": 0.28830018639564514, + "learning_rate": 1.0296677789061367e-05, + "loss": 0.0037, + "step": 13477 + }, + { + "epoch": 1.4611882046834346, + "grad_norm": 1.7735182046890259, + "learning_rate": 1.0295952415493979e-05, + "loss": 0.0273, + "step": 13478 + }, + { + "epoch": 1.4612966175195143, + "grad_norm": 0.8053417205810547, + "learning_rate": 1.0295227041926592e-05, + "loss": 0.0182, + "step": 13479 + }, + { + "epoch": 1.461405030355594, + "grad_norm": 1.7865036725997925, + "learning_rate": 1.0294501668359205e-05, + "loss": 0.0118, + "step": 13480 + }, + { + "epoch": 1.4615134431916739, + "grad_norm": 2.349545955657959, + "learning_rate": 1.029377629479182e-05, + "loss": 0.0323, + "step": 13481 + }, + { + "epoch": 1.4616218560277536, + "grad_norm": 0.9132789373397827, + "learning_rate": 1.0293050921224432e-05, + "loss": 0.0045, + "step": 13482 + }, + { + "epoch": 1.4617302688638336, + "grad_norm": 2.898679733276367, + "learning_rate": 1.0292325547657044e-05, + "loss": 0.0483, + "step": 13483 + }, + { + "epoch": 1.4618386816999132, + "grad_norm": 2.379481554031372, + "learning_rate": 1.0291600174089658e-05, + "loss": 0.0115, + "step": 13484 + }, + { + "epoch": 1.4619470945359931, + "grad_norm": 1.7973824739456177, + "learning_rate": 1.029087480052227e-05, + "loss": 0.0071, + "step": 13485 + }, + { + "epoch": 1.462055507372073, + "grad_norm": 4.199108123779297, + "learning_rate": 1.0290149426954883e-05, + "loss": 0.0257, + "step": 13486 + }, + { + "epoch": 1.4621639202081527, + "grad_norm": 0.1924719363451004, + "learning_rate": 1.0289424053387496e-05, + "loss": 0.0021, + "step": 13487 + }, + { + "epoch": 1.4622723330442324, + "grad_norm": 0.9019542932510376, + "learning_rate": 1.0288698679820109e-05, + "loss": 0.0085, + "step": 13488 + }, + { + "epoch": 1.4623807458803122, + "grad_norm": 2.362276554107666, + "learning_rate": 1.028797330625272e-05, + "loss": 0.0423, + "step": 13489 + }, + { + "epoch": 1.462489158716392, + "grad_norm": 0.6831929683685303, + "learning_rate": 1.0287247932685333e-05, + "loss": 0.0057, + "step": 13490 + }, + { + "epoch": 1.4625975715524717, + "grad_norm": 1.5299454927444458, + "learning_rate": 1.0286522559117947e-05, + "loss": 0.0046, + "step": 13491 + }, + { + "epoch": 1.4627059843885517, + "grad_norm": 4.607886791229248, + "learning_rate": 1.0285797185550559e-05, + "loss": 0.0215, + "step": 13492 + }, + { + "epoch": 1.4628143972246315, + "grad_norm": 0.23729518055915833, + "learning_rate": 1.0285071811983171e-05, + "loss": 0.0014, + "step": 13493 + }, + { + "epoch": 1.4629228100607112, + "grad_norm": 0.5839195847511292, + "learning_rate": 1.0284346438415785e-05, + "loss": 0.0071, + "step": 13494 + }, + { + "epoch": 1.463031222896791, + "grad_norm": 2.9581151008605957, + "learning_rate": 1.0283621064848397e-05, + "loss": 0.0395, + "step": 13495 + }, + { + "epoch": 1.4631396357328708, + "grad_norm": 2.9813950061798096, + "learning_rate": 1.0282895691281009e-05, + "loss": 0.0077, + "step": 13496 + }, + { + "epoch": 1.4632480485689505, + "grad_norm": 0.9535032510757446, + "learning_rate": 1.0282170317713623e-05, + "loss": 0.0051, + "step": 13497 + }, + { + "epoch": 1.4633564614050303, + "grad_norm": 2.9226088523864746, + "learning_rate": 1.0281444944146237e-05, + "loss": 0.0131, + "step": 13498 + }, + { + "epoch": 1.4634648742411103, + "grad_norm": 2.6512832641601562, + "learning_rate": 1.028071957057885e-05, + "loss": 0.0158, + "step": 13499 + }, + { + "epoch": 1.4635732870771898, + "grad_norm": 5.167930603027344, + "learning_rate": 1.0279994197011463e-05, + "loss": 0.0356, + "step": 13500 + }, + { + "epoch": 1.4636816999132698, + "grad_norm": 2.72117280960083, + "learning_rate": 1.0279268823444075e-05, + "loss": 0.0192, + "step": 13501 + }, + { + "epoch": 1.4637901127493496, + "grad_norm": 0.976584792137146, + "learning_rate": 1.0278543449876687e-05, + "loss": 0.0065, + "step": 13502 + }, + { + "epoch": 1.4638985255854293, + "grad_norm": 1.4132272005081177, + "learning_rate": 1.0277818076309301e-05, + "loss": 0.029, + "step": 13503 + }, + { + "epoch": 1.464006938421509, + "grad_norm": 0.39123737812042236, + "learning_rate": 1.0277092702741913e-05, + "loss": 0.0065, + "step": 13504 + }, + { + "epoch": 1.4641153512575888, + "grad_norm": 3.4945321083068848, + "learning_rate": 1.0276367329174525e-05, + "loss": 0.0679, + "step": 13505 + }, + { + "epoch": 1.4642237640936686, + "grad_norm": 6.296879768371582, + "learning_rate": 1.0275641955607139e-05, + "loss": 0.0836, + "step": 13506 + }, + { + "epoch": 1.4643321769297484, + "grad_norm": 3.3311350345611572, + "learning_rate": 1.0274916582039751e-05, + "loss": 0.0283, + "step": 13507 + }, + { + "epoch": 1.4644405897658284, + "grad_norm": 4.142587661743164, + "learning_rate": 1.0274191208472363e-05, + "loss": 0.0248, + "step": 13508 + }, + { + "epoch": 1.4645490026019081, + "grad_norm": 2.0252187252044678, + "learning_rate": 1.0273465834904977e-05, + "loss": 0.0254, + "step": 13509 + }, + { + "epoch": 1.4646574154379879, + "grad_norm": 2.7416701316833496, + "learning_rate": 1.027274046133759e-05, + "loss": 0.0184, + "step": 13510 + }, + { + "epoch": 1.4647658282740676, + "grad_norm": 3.072291612625122, + "learning_rate": 1.0272015087770201e-05, + "loss": 0.0551, + "step": 13511 + }, + { + "epoch": 1.4648742411101474, + "grad_norm": 0.16819769144058228, + "learning_rate": 1.0271289714202815e-05, + "loss": 0.0009, + "step": 13512 + }, + { + "epoch": 1.4649826539462272, + "grad_norm": 1.023672342300415, + "learning_rate": 1.0270564340635427e-05, + "loss": 0.0045, + "step": 13513 + }, + { + "epoch": 1.465091066782307, + "grad_norm": 2.081380605697632, + "learning_rate": 1.026983896706804e-05, + "loss": 0.0288, + "step": 13514 + }, + { + "epoch": 1.465199479618387, + "grad_norm": 0.7970161437988281, + "learning_rate": 1.0269113593500655e-05, + "loss": 0.0105, + "step": 13515 + }, + { + "epoch": 1.4653078924544667, + "grad_norm": 0.651565670967102, + "learning_rate": 1.0268388219933267e-05, + "loss": 0.0028, + "step": 13516 + }, + { + "epoch": 1.4654163052905465, + "grad_norm": 0.4111004173755646, + "learning_rate": 1.026766284636588e-05, + "loss": 0.0039, + "step": 13517 + }, + { + "epoch": 1.4655247181266262, + "grad_norm": 0.9942089319229126, + "learning_rate": 1.0266937472798493e-05, + "loss": 0.0173, + "step": 13518 + }, + { + "epoch": 1.465633130962706, + "grad_norm": 1.6259623765945435, + "learning_rate": 1.0266212099231105e-05, + "loss": 0.0328, + "step": 13519 + }, + { + "epoch": 1.4657415437987857, + "grad_norm": 0.9491989016532898, + "learning_rate": 1.0265486725663717e-05, + "loss": 0.0043, + "step": 13520 + }, + { + "epoch": 1.4658499566348655, + "grad_norm": 0.319254606962204, + "learning_rate": 1.0264761352096331e-05, + "loss": 0.0017, + "step": 13521 + }, + { + "epoch": 1.4659583694709455, + "grad_norm": 3.2264513969421387, + "learning_rate": 1.0264035978528943e-05, + "loss": 0.0641, + "step": 13522 + }, + { + "epoch": 1.466066782307025, + "grad_norm": 1.5751227140426636, + "learning_rate": 1.0263310604961556e-05, + "loss": 0.0107, + "step": 13523 + }, + { + "epoch": 1.466175195143105, + "grad_norm": 1.4336589574813843, + "learning_rate": 1.026258523139417e-05, + "loss": 0.0229, + "step": 13524 + }, + { + "epoch": 1.4662836079791848, + "grad_norm": 2.3483898639678955, + "learning_rate": 1.0261859857826782e-05, + "loss": 0.0192, + "step": 13525 + }, + { + "epoch": 1.4663920208152645, + "grad_norm": 2.1680681705474854, + "learning_rate": 1.0261134484259394e-05, + "loss": 0.0047, + "step": 13526 + }, + { + "epoch": 1.4665004336513443, + "grad_norm": 2.8806841373443604, + "learning_rate": 1.0260409110692008e-05, + "loss": 0.0232, + "step": 13527 + }, + { + "epoch": 1.466608846487424, + "grad_norm": 0.2074345499277115, + "learning_rate": 1.025968373712462e-05, + "loss": 0.0011, + "step": 13528 + }, + { + "epoch": 1.4667172593235038, + "grad_norm": 2.6241888999938965, + "learning_rate": 1.0258958363557232e-05, + "loss": 0.0433, + "step": 13529 + }, + { + "epoch": 1.4668256721595836, + "grad_norm": 0.5244432091712952, + "learning_rate": 1.0258232989989844e-05, + "loss": 0.0023, + "step": 13530 + }, + { + "epoch": 1.4669340849956636, + "grad_norm": 1.9843438863754272, + "learning_rate": 1.0257507616422458e-05, + "loss": 0.007, + "step": 13531 + }, + { + "epoch": 1.4670424978317433, + "grad_norm": 0.5890781879425049, + "learning_rate": 1.0256782242855072e-05, + "loss": 0.0042, + "step": 13532 + }, + { + "epoch": 1.467150910667823, + "grad_norm": 7.328338623046875, + "learning_rate": 1.0256056869287686e-05, + "loss": 0.0354, + "step": 13533 + }, + { + "epoch": 1.4672593235039029, + "grad_norm": 2.939567804336548, + "learning_rate": 1.0255331495720298e-05, + "loss": 0.0084, + "step": 13534 + }, + { + "epoch": 1.4673677363399826, + "grad_norm": 4.051928997039795, + "learning_rate": 1.025460612215291e-05, + "loss": 0.0216, + "step": 13535 + }, + { + "epoch": 1.4674761491760624, + "grad_norm": 0.5892722606658936, + "learning_rate": 1.0253880748585524e-05, + "loss": 0.0025, + "step": 13536 + }, + { + "epoch": 1.4675845620121422, + "grad_norm": 1.7503561973571777, + "learning_rate": 1.0253155375018136e-05, + "loss": 0.0064, + "step": 13537 + }, + { + "epoch": 1.4676929748482221, + "grad_norm": 1.7861547470092773, + "learning_rate": 1.0252430001450748e-05, + "loss": 0.0442, + "step": 13538 + }, + { + "epoch": 1.4678013876843017, + "grad_norm": 2.6046552658081055, + "learning_rate": 1.0251704627883362e-05, + "loss": 0.0082, + "step": 13539 + }, + { + "epoch": 1.4679098005203817, + "grad_norm": 3.736717462539673, + "learning_rate": 1.0250979254315974e-05, + "loss": 0.0391, + "step": 13540 + }, + { + "epoch": 1.4680182133564614, + "grad_norm": 0.8254467248916626, + "learning_rate": 1.0250253880748586e-05, + "loss": 0.0086, + "step": 13541 + }, + { + "epoch": 1.4681266261925412, + "grad_norm": 0.14475834369659424, + "learning_rate": 1.0249528507181198e-05, + "loss": 0.0012, + "step": 13542 + }, + { + "epoch": 1.468235039028621, + "grad_norm": 1.0163251161575317, + "learning_rate": 1.0248803133613812e-05, + "loss": 0.0076, + "step": 13543 + }, + { + "epoch": 1.4683434518647007, + "grad_norm": 3.3935387134552, + "learning_rate": 1.0248077760046424e-05, + "loss": 0.0283, + "step": 13544 + }, + { + "epoch": 1.4684518647007807, + "grad_norm": 1.4632326364517212, + "learning_rate": 1.0247352386479036e-05, + "loss": 0.0172, + "step": 13545 + }, + { + "epoch": 1.4685602775368602, + "grad_norm": 1.4470657110214233, + "learning_rate": 1.024662701291165e-05, + "loss": 0.0089, + "step": 13546 + }, + { + "epoch": 1.4686686903729402, + "grad_norm": 0.14983370900154114, + "learning_rate": 1.0245901639344262e-05, + "loss": 0.002, + "step": 13547 + }, + { + "epoch": 1.46877710320902, + "grad_norm": 3.2065978050231934, + "learning_rate": 1.0245176265776874e-05, + "loss": 0.0248, + "step": 13548 + }, + { + "epoch": 1.4688855160450998, + "grad_norm": 1.7741509675979614, + "learning_rate": 1.0244450892209488e-05, + "loss": 0.0056, + "step": 13549 + }, + { + "epoch": 1.4689939288811795, + "grad_norm": 3.532919406890869, + "learning_rate": 1.0243725518642102e-05, + "loss": 0.0128, + "step": 13550 + }, + { + "epoch": 1.4691023417172593, + "grad_norm": 6.018467903137207, + "learning_rate": 1.0243000145074716e-05, + "loss": 0.0181, + "step": 13551 + }, + { + "epoch": 1.469210754553339, + "grad_norm": 12.177389144897461, + "learning_rate": 1.0242274771507328e-05, + "loss": 0.0501, + "step": 13552 + }, + { + "epoch": 1.4693191673894188, + "grad_norm": 2.543306589126587, + "learning_rate": 1.024154939793994e-05, + "loss": 0.0425, + "step": 13553 + }, + { + "epoch": 1.4694275802254988, + "grad_norm": 1.9469597339630127, + "learning_rate": 1.0240824024372552e-05, + "loss": 0.0053, + "step": 13554 + }, + { + "epoch": 1.4695359930615786, + "grad_norm": 2.9130215644836426, + "learning_rate": 1.0240098650805166e-05, + "loss": 0.0226, + "step": 13555 + }, + { + "epoch": 1.4696444058976583, + "grad_norm": 2.090080738067627, + "learning_rate": 1.0239373277237778e-05, + "loss": 0.0319, + "step": 13556 + }, + { + "epoch": 1.469752818733738, + "grad_norm": 3.794750690460205, + "learning_rate": 1.023864790367039e-05, + "loss": 0.009, + "step": 13557 + }, + { + "epoch": 1.4698612315698178, + "grad_norm": 0.7066375017166138, + "learning_rate": 1.0237922530103004e-05, + "loss": 0.0073, + "step": 13558 + }, + { + "epoch": 1.4699696444058976, + "grad_norm": 0.1260513961315155, + "learning_rate": 1.0237197156535616e-05, + "loss": 0.0016, + "step": 13559 + }, + { + "epoch": 1.4700780572419774, + "grad_norm": 2.474067449569702, + "learning_rate": 1.0236471782968229e-05, + "loss": 0.0122, + "step": 13560 + }, + { + "epoch": 1.4701864700780574, + "grad_norm": 0.329587459564209, + "learning_rate": 1.0235746409400842e-05, + "loss": 0.0023, + "step": 13561 + }, + { + "epoch": 1.470294882914137, + "grad_norm": 4.0952839851379395, + "learning_rate": 1.0235021035833455e-05, + "loss": 0.0515, + "step": 13562 + }, + { + "epoch": 1.4704032957502169, + "grad_norm": 1.6598520278930664, + "learning_rate": 1.0234295662266067e-05, + "loss": 0.0117, + "step": 13563 + }, + { + "epoch": 1.4705117085862967, + "grad_norm": 0.4519343376159668, + "learning_rate": 1.023357028869868e-05, + "loss": 0.0042, + "step": 13564 + }, + { + "epoch": 1.4706201214223764, + "grad_norm": 5.434100151062012, + "learning_rate": 1.0232844915131293e-05, + "loss": 0.0435, + "step": 13565 + }, + { + "epoch": 1.4707285342584562, + "grad_norm": 1.744571328163147, + "learning_rate": 1.0232119541563905e-05, + "loss": 0.0155, + "step": 13566 + }, + { + "epoch": 1.470836947094536, + "grad_norm": 1.2636388540267944, + "learning_rate": 1.023139416799652e-05, + "loss": 0.012, + "step": 13567 + }, + { + "epoch": 1.4709453599306157, + "grad_norm": 6.002057075500488, + "learning_rate": 1.0230668794429133e-05, + "loss": 0.049, + "step": 13568 + }, + { + "epoch": 1.4710537727666955, + "grad_norm": 2.8964757919311523, + "learning_rate": 1.0229943420861745e-05, + "loss": 0.0527, + "step": 13569 + }, + { + "epoch": 1.4711621856027755, + "grad_norm": 0.6173309087753296, + "learning_rate": 1.0229218047294359e-05, + "loss": 0.0041, + "step": 13570 + }, + { + "epoch": 1.4712705984388552, + "grad_norm": 5.657412052154541, + "learning_rate": 1.022849267372697e-05, + "loss": 0.0125, + "step": 13571 + }, + { + "epoch": 1.471379011274935, + "grad_norm": 3.02897047996521, + "learning_rate": 1.0227767300159583e-05, + "loss": 0.0169, + "step": 13572 + }, + { + "epoch": 1.4714874241110147, + "grad_norm": 1.116289734840393, + "learning_rate": 1.0227041926592197e-05, + "loss": 0.006, + "step": 13573 + }, + { + "epoch": 1.4715958369470945, + "grad_norm": 8.033984184265137, + "learning_rate": 1.0226316553024809e-05, + "loss": 0.0123, + "step": 13574 + }, + { + "epoch": 1.4717042497831743, + "grad_norm": 3.1044814586639404, + "learning_rate": 1.0225591179457421e-05, + "loss": 0.0115, + "step": 13575 + }, + { + "epoch": 1.471812662619254, + "grad_norm": 0.05780436471104622, + "learning_rate": 1.0224865805890035e-05, + "loss": 0.0005, + "step": 13576 + }, + { + "epoch": 1.471921075455334, + "grad_norm": 2.187190532684326, + "learning_rate": 1.0224140432322647e-05, + "loss": 0.0313, + "step": 13577 + }, + { + "epoch": 1.4720294882914138, + "grad_norm": 7.426743030548096, + "learning_rate": 1.0223415058755259e-05, + "loss": 0.0521, + "step": 13578 + }, + { + "epoch": 1.4721379011274935, + "grad_norm": 4.6033525466918945, + "learning_rate": 1.0222689685187873e-05, + "loss": 0.0347, + "step": 13579 + }, + { + "epoch": 1.4722463139635733, + "grad_norm": 2.751215696334839, + "learning_rate": 1.0221964311620485e-05, + "loss": 0.0451, + "step": 13580 + }, + { + "epoch": 1.472354726799653, + "grad_norm": 4.993922710418701, + "learning_rate": 1.0221238938053097e-05, + "loss": 0.0663, + "step": 13581 + }, + { + "epoch": 1.4724631396357328, + "grad_norm": 2.3545594215393066, + "learning_rate": 1.022051356448571e-05, + "loss": 0.0294, + "step": 13582 + }, + { + "epoch": 1.4725715524718126, + "grad_norm": 3.0548160076141357, + "learning_rate": 1.0219788190918323e-05, + "loss": 0.0101, + "step": 13583 + }, + { + "epoch": 1.4726799653078926, + "grad_norm": 2.3989078998565674, + "learning_rate": 1.0219062817350937e-05, + "loss": 0.0172, + "step": 13584 + }, + { + "epoch": 1.4727883781439721, + "grad_norm": 2.9845218658447266, + "learning_rate": 1.021833744378355e-05, + "loss": 0.0378, + "step": 13585 + }, + { + "epoch": 1.472896790980052, + "grad_norm": 1.2128167152404785, + "learning_rate": 1.0217612070216163e-05, + "loss": 0.0098, + "step": 13586 + }, + { + "epoch": 1.4730052038161319, + "grad_norm": 0.21968793869018555, + "learning_rate": 1.0216886696648775e-05, + "loss": 0.0021, + "step": 13587 + }, + { + "epoch": 1.4731136166522116, + "grad_norm": 5.135882377624512, + "learning_rate": 1.0216161323081389e-05, + "loss": 0.018, + "step": 13588 + }, + { + "epoch": 1.4732220294882914, + "grad_norm": 0.9813191890716553, + "learning_rate": 1.0215435949514001e-05, + "loss": 0.0184, + "step": 13589 + }, + { + "epoch": 1.4733304423243712, + "grad_norm": 0.5804511308670044, + "learning_rate": 1.0214710575946613e-05, + "loss": 0.0018, + "step": 13590 + }, + { + "epoch": 1.473438855160451, + "grad_norm": 0.7196137309074402, + "learning_rate": 1.0213985202379227e-05, + "loss": 0.0033, + "step": 13591 + }, + { + "epoch": 1.4735472679965307, + "grad_norm": 11.10470199584961, + "learning_rate": 1.021325982881184e-05, + "loss": 0.0294, + "step": 13592 + }, + { + "epoch": 1.4736556808326107, + "grad_norm": 1.2018837928771973, + "learning_rate": 1.0212534455244451e-05, + "loss": 0.0123, + "step": 13593 + }, + { + "epoch": 1.4737640936686904, + "grad_norm": 1.2308785915374756, + "learning_rate": 1.0211809081677064e-05, + "loss": 0.0054, + "step": 13594 + }, + { + "epoch": 1.4738725065047702, + "grad_norm": 1.9269294738769531, + "learning_rate": 1.0211083708109677e-05, + "loss": 0.0284, + "step": 13595 + }, + { + "epoch": 1.47398091934085, + "grad_norm": 1.5317022800445557, + "learning_rate": 1.021035833454229e-05, + "loss": 0.0147, + "step": 13596 + }, + { + "epoch": 1.4740893321769297, + "grad_norm": 1.2309178113937378, + "learning_rate": 1.0209632960974902e-05, + "loss": 0.0058, + "step": 13597 + }, + { + "epoch": 1.4741977450130095, + "grad_norm": 4.338597297668457, + "learning_rate": 1.0208907587407515e-05, + "loss": 0.0102, + "step": 13598 + }, + { + "epoch": 1.4743061578490892, + "grad_norm": 2.515564441680908, + "learning_rate": 1.0208182213840128e-05, + "loss": 0.0386, + "step": 13599 + }, + { + "epoch": 1.4744145706851692, + "grad_norm": 2.9309821128845215, + "learning_rate": 1.020745684027274e-05, + "loss": 0.0137, + "step": 13600 + }, + { + "epoch": 1.474522983521249, + "grad_norm": 2.7422735691070557, + "learning_rate": 1.0206731466705355e-05, + "loss": 0.0138, + "step": 13601 + }, + { + "epoch": 1.4746313963573288, + "grad_norm": 0.6014840006828308, + "learning_rate": 1.0206006093137967e-05, + "loss": 0.0034, + "step": 13602 + }, + { + "epoch": 1.4747398091934085, + "grad_norm": 8.339134216308594, + "learning_rate": 1.0205280719570581e-05, + "loss": 0.055, + "step": 13603 + }, + { + "epoch": 1.4748482220294883, + "grad_norm": 0.9311899542808533, + "learning_rate": 1.0204555346003193e-05, + "loss": 0.0517, + "step": 13604 + }, + { + "epoch": 1.474956634865568, + "grad_norm": 3.02911639213562, + "learning_rate": 1.0203829972435806e-05, + "loss": 0.0444, + "step": 13605 + }, + { + "epoch": 1.4750650477016478, + "grad_norm": 3.6373233795166016, + "learning_rate": 1.0203104598868418e-05, + "loss": 0.0209, + "step": 13606 + }, + { + "epoch": 1.4751734605377278, + "grad_norm": 3.515009641647339, + "learning_rate": 1.0202379225301032e-05, + "loss": 0.0495, + "step": 13607 + }, + { + "epoch": 1.4752818733738073, + "grad_norm": 4.464715957641602, + "learning_rate": 1.0201653851733644e-05, + "loss": 0.0577, + "step": 13608 + }, + { + "epoch": 1.4753902862098873, + "grad_norm": 0.5016193389892578, + "learning_rate": 1.0200928478166256e-05, + "loss": 0.0025, + "step": 13609 + }, + { + "epoch": 1.475498699045967, + "grad_norm": 8.993647575378418, + "learning_rate": 1.020020310459887e-05, + "loss": 0.0703, + "step": 13610 + }, + { + "epoch": 1.4756071118820469, + "grad_norm": 1.137766718864441, + "learning_rate": 1.0199477731031482e-05, + "loss": 0.0064, + "step": 13611 + }, + { + "epoch": 1.4757155247181266, + "grad_norm": 3.1534688472747803, + "learning_rate": 1.0198752357464094e-05, + "loss": 0.0228, + "step": 13612 + }, + { + "epoch": 1.4758239375542064, + "grad_norm": 1.4792165756225586, + "learning_rate": 1.0198026983896708e-05, + "loss": 0.0048, + "step": 13613 + }, + { + "epoch": 1.4759323503902861, + "grad_norm": 3.7940781116485596, + "learning_rate": 1.019730161032932e-05, + "loss": 0.0269, + "step": 13614 + }, + { + "epoch": 1.476040763226366, + "grad_norm": 0.2230808585882187, + "learning_rate": 1.0196576236761932e-05, + "loss": 0.0015, + "step": 13615 + }, + { + "epoch": 1.476149176062446, + "grad_norm": 5.0990309715271, + "learning_rate": 1.0195850863194546e-05, + "loss": 0.0589, + "step": 13616 + }, + { + "epoch": 1.4762575888985257, + "grad_norm": 2.191037178039551, + "learning_rate": 1.0195125489627158e-05, + "loss": 0.0375, + "step": 13617 + }, + { + "epoch": 1.4763660017346054, + "grad_norm": 6.839838027954102, + "learning_rate": 1.019440011605977e-05, + "loss": 0.0259, + "step": 13618 + }, + { + "epoch": 1.4764744145706852, + "grad_norm": 1.1966454982757568, + "learning_rate": 1.0193674742492386e-05, + "loss": 0.0088, + "step": 13619 + }, + { + "epoch": 1.476582827406765, + "grad_norm": 0.5885257124900818, + "learning_rate": 1.0192949368924998e-05, + "loss": 0.0057, + "step": 13620 + }, + { + "epoch": 1.4766912402428447, + "grad_norm": 8.847613334655762, + "learning_rate": 1.019222399535761e-05, + "loss": 0.0056, + "step": 13621 + }, + { + "epoch": 1.4767996530789245, + "grad_norm": 4.290712833404541, + "learning_rate": 1.0191498621790224e-05, + "loss": 0.0197, + "step": 13622 + }, + { + "epoch": 1.4769080659150045, + "grad_norm": 2.316007137298584, + "learning_rate": 1.0190773248222836e-05, + "loss": 0.0349, + "step": 13623 + }, + { + "epoch": 1.477016478751084, + "grad_norm": 1.9815387725830078, + "learning_rate": 1.0190047874655448e-05, + "loss": 0.0089, + "step": 13624 + }, + { + "epoch": 1.477124891587164, + "grad_norm": 0.32743382453918457, + "learning_rate": 1.0189322501088062e-05, + "loss": 0.0047, + "step": 13625 + }, + { + "epoch": 1.4772333044232437, + "grad_norm": 1.5513460636138916, + "learning_rate": 1.0188597127520674e-05, + "loss": 0.0095, + "step": 13626 + }, + { + "epoch": 1.4773417172593235, + "grad_norm": 5.253073215484619, + "learning_rate": 1.0187871753953286e-05, + "loss": 0.0144, + "step": 13627 + }, + { + "epoch": 1.4774501300954033, + "grad_norm": 1.5009653568267822, + "learning_rate": 1.01871463803859e-05, + "loss": 0.0052, + "step": 13628 + }, + { + "epoch": 1.477558542931483, + "grad_norm": 5.301366806030273, + "learning_rate": 1.0186421006818512e-05, + "loss": 0.0785, + "step": 13629 + }, + { + "epoch": 1.477666955767563, + "grad_norm": 2.357386350631714, + "learning_rate": 1.0185695633251124e-05, + "loss": 0.0427, + "step": 13630 + }, + { + "epoch": 1.4777753686036426, + "grad_norm": 4.11829948425293, + "learning_rate": 1.0184970259683738e-05, + "loss": 0.0662, + "step": 13631 + }, + { + "epoch": 1.4778837814397225, + "grad_norm": 2.7873170375823975, + "learning_rate": 1.018424488611635e-05, + "loss": 0.0405, + "step": 13632 + }, + { + "epoch": 1.4779921942758023, + "grad_norm": 3.652772903442383, + "learning_rate": 1.0183519512548963e-05, + "loss": 0.0224, + "step": 13633 + }, + { + "epoch": 1.478100607111882, + "grad_norm": 0.39595621824264526, + "learning_rate": 1.0182794138981575e-05, + "loss": 0.0041, + "step": 13634 + }, + { + "epoch": 1.4782090199479618, + "grad_norm": 5.84077262878418, + "learning_rate": 1.0182068765414189e-05, + "loss": 0.0557, + "step": 13635 + }, + { + "epoch": 1.4783174327840416, + "grad_norm": 2.483686923980713, + "learning_rate": 1.0181343391846802e-05, + "loss": 0.0324, + "step": 13636 + }, + { + "epoch": 1.4784258456201214, + "grad_norm": 5.200314998626709, + "learning_rate": 1.0180618018279416e-05, + "loss": 0.034, + "step": 13637 + }, + { + "epoch": 1.4785342584562011, + "grad_norm": 2.9112467765808105, + "learning_rate": 1.0179892644712028e-05, + "loss": 0.0448, + "step": 13638 + }, + { + "epoch": 1.478642671292281, + "grad_norm": 5.2380547523498535, + "learning_rate": 1.017916727114464e-05, + "loss": 0.0159, + "step": 13639 + }, + { + "epoch": 1.4787510841283609, + "grad_norm": 2.435338020324707, + "learning_rate": 1.0178441897577254e-05, + "loss": 0.0318, + "step": 13640 + }, + { + "epoch": 1.4788594969644406, + "grad_norm": 2.0534322261810303, + "learning_rate": 1.0177716524009866e-05, + "loss": 0.0423, + "step": 13641 + }, + { + "epoch": 1.4789679098005204, + "grad_norm": 2.170768976211548, + "learning_rate": 1.0176991150442479e-05, + "loss": 0.0185, + "step": 13642 + }, + { + "epoch": 1.4790763226366002, + "grad_norm": 5.942361831665039, + "learning_rate": 1.0176265776875092e-05, + "loss": 0.0607, + "step": 13643 + }, + { + "epoch": 1.47918473547268, + "grad_norm": 2.863933563232422, + "learning_rate": 1.0175540403307705e-05, + "loss": 0.0288, + "step": 13644 + }, + { + "epoch": 1.4792931483087597, + "grad_norm": 5.190810680389404, + "learning_rate": 1.0174815029740317e-05, + "loss": 0.0456, + "step": 13645 + }, + { + "epoch": 1.4794015611448397, + "grad_norm": 3.25700306892395, + "learning_rate": 1.0174089656172929e-05, + "loss": 0.0201, + "step": 13646 + }, + { + "epoch": 1.4795099739809192, + "grad_norm": 3.064462184906006, + "learning_rate": 1.0173364282605543e-05, + "loss": 0.0465, + "step": 13647 + }, + { + "epoch": 1.4796183868169992, + "grad_norm": 6.02052116394043, + "learning_rate": 1.0172638909038155e-05, + "loss": 0.0184, + "step": 13648 + }, + { + "epoch": 1.479726799653079, + "grad_norm": 2.6876258850097656, + "learning_rate": 1.0171913535470767e-05, + "loss": 0.0257, + "step": 13649 + }, + { + "epoch": 1.4798352124891587, + "grad_norm": 2.3392415046691895, + "learning_rate": 1.017118816190338e-05, + "loss": 0.0103, + "step": 13650 + }, + { + "epoch": 1.4799436253252385, + "grad_norm": 1.3228895664215088, + "learning_rate": 1.0170462788335993e-05, + "loss": 0.0103, + "step": 13651 + }, + { + "epoch": 1.4800520381613183, + "grad_norm": 1.916831374168396, + "learning_rate": 1.0169737414768605e-05, + "loss": 0.0254, + "step": 13652 + }, + { + "epoch": 1.480160450997398, + "grad_norm": 2.2468855381011963, + "learning_rate": 1.016901204120122e-05, + "loss": 0.023, + "step": 13653 + }, + { + "epoch": 1.4802688638334778, + "grad_norm": 1.4079384803771973, + "learning_rate": 1.0168286667633833e-05, + "loss": 0.003, + "step": 13654 + }, + { + "epoch": 1.4803772766695578, + "grad_norm": 2.489466905593872, + "learning_rate": 1.0167561294066447e-05, + "loss": 0.0167, + "step": 13655 + }, + { + "epoch": 1.4804856895056375, + "grad_norm": 7.946707725524902, + "learning_rate": 1.0166835920499059e-05, + "loss": 0.0467, + "step": 13656 + }, + { + "epoch": 1.4805941023417173, + "grad_norm": 2.7630515098571777, + "learning_rate": 1.0166110546931671e-05, + "loss": 0.0303, + "step": 13657 + }, + { + "epoch": 1.480702515177797, + "grad_norm": 2.146308183670044, + "learning_rate": 1.0165385173364283e-05, + "loss": 0.0223, + "step": 13658 + }, + { + "epoch": 1.4808109280138768, + "grad_norm": 2.686035633087158, + "learning_rate": 1.0164659799796897e-05, + "loss": 0.0141, + "step": 13659 + }, + { + "epoch": 1.4809193408499566, + "grad_norm": 0.9613189101219177, + "learning_rate": 1.0163934426229509e-05, + "loss": 0.0076, + "step": 13660 + }, + { + "epoch": 1.4810277536860363, + "grad_norm": 2.099543333053589, + "learning_rate": 1.0163209052662121e-05, + "loss": 0.0313, + "step": 13661 + }, + { + "epoch": 1.4811361665221163, + "grad_norm": 4.57460880279541, + "learning_rate": 1.0162483679094735e-05, + "loss": 0.0218, + "step": 13662 + }, + { + "epoch": 1.481244579358196, + "grad_norm": 0.22841444611549377, + "learning_rate": 1.0161758305527347e-05, + "loss": 0.0025, + "step": 13663 + }, + { + "epoch": 1.4813529921942759, + "grad_norm": 1.6640527248382568, + "learning_rate": 1.016103293195996e-05, + "loss": 0.0299, + "step": 13664 + }, + { + "epoch": 1.4814614050303556, + "grad_norm": 1.2041454315185547, + "learning_rate": 1.0160307558392573e-05, + "loss": 0.0305, + "step": 13665 + }, + { + "epoch": 1.4815698178664354, + "grad_norm": 0.918700098991394, + "learning_rate": 1.0159582184825185e-05, + "loss": 0.005, + "step": 13666 + }, + { + "epoch": 1.4816782307025151, + "grad_norm": 9.576589584350586, + "learning_rate": 1.0158856811257797e-05, + "loss": 0.0166, + "step": 13667 + }, + { + "epoch": 1.481786643538595, + "grad_norm": 2.9314839839935303, + "learning_rate": 1.0158131437690411e-05, + "loss": 0.0207, + "step": 13668 + }, + { + "epoch": 1.481895056374675, + "grad_norm": 5.9460954666137695, + "learning_rate": 1.0157406064123023e-05, + "loss": 0.0413, + "step": 13669 + }, + { + "epoch": 1.4820034692107544, + "grad_norm": 1.4711183309555054, + "learning_rate": 1.0156680690555636e-05, + "loss": 0.0098, + "step": 13670 + }, + { + "epoch": 1.4821118820468344, + "grad_norm": 5.2272233963012695, + "learning_rate": 1.0155955316988251e-05, + "loss": 0.0305, + "step": 13671 + }, + { + "epoch": 1.4822202948829142, + "grad_norm": 0.8527733683586121, + "learning_rate": 1.0155229943420863e-05, + "loss": 0.0044, + "step": 13672 + }, + { + "epoch": 1.482328707718994, + "grad_norm": 4.481319427490234, + "learning_rate": 1.0154504569853475e-05, + "loss": 0.0324, + "step": 13673 + }, + { + "epoch": 1.4824371205550737, + "grad_norm": 0.68211430311203, + "learning_rate": 1.015377919628609e-05, + "loss": 0.003, + "step": 13674 + }, + { + "epoch": 1.4825455333911535, + "grad_norm": 2.160254716873169, + "learning_rate": 1.0153053822718701e-05, + "loss": 0.0197, + "step": 13675 + }, + { + "epoch": 1.4826539462272332, + "grad_norm": 0.5877444744110107, + "learning_rate": 1.0152328449151313e-05, + "loss": 0.0031, + "step": 13676 + }, + { + "epoch": 1.482762359063313, + "grad_norm": 2.135477066040039, + "learning_rate": 1.0151603075583927e-05, + "loss": 0.0187, + "step": 13677 + }, + { + "epoch": 1.482870771899393, + "grad_norm": 0.40766099095344543, + "learning_rate": 1.015087770201654e-05, + "loss": 0.0037, + "step": 13678 + }, + { + "epoch": 1.4829791847354727, + "grad_norm": 0.2837033271789551, + "learning_rate": 1.0150152328449152e-05, + "loss": 0.003, + "step": 13679 + }, + { + "epoch": 1.4830875975715525, + "grad_norm": 2.1341114044189453, + "learning_rate": 1.0149426954881765e-05, + "loss": 0.0098, + "step": 13680 + }, + { + "epoch": 1.4831960104076323, + "grad_norm": 7.145084857940674, + "learning_rate": 1.0148701581314378e-05, + "loss": 0.0327, + "step": 13681 + }, + { + "epoch": 1.483304423243712, + "grad_norm": 4.974673271179199, + "learning_rate": 1.014797620774699e-05, + "loss": 0.0302, + "step": 13682 + }, + { + "epoch": 1.4834128360797918, + "grad_norm": 1.0049614906311035, + "learning_rate": 1.0147250834179604e-05, + "loss": 0.0057, + "step": 13683 + }, + { + "epoch": 1.4835212489158716, + "grad_norm": 2.401648998260498, + "learning_rate": 1.0146525460612216e-05, + "loss": 0.0244, + "step": 13684 + }, + { + "epoch": 1.4836296617519515, + "grad_norm": 1.9643371105194092, + "learning_rate": 1.0145800087044828e-05, + "loss": 0.0216, + "step": 13685 + }, + { + "epoch": 1.483738074588031, + "grad_norm": 1.0366817712783813, + "learning_rate": 1.014507471347744e-05, + "loss": 0.0126, + "step": 13686 + }, + { + "epoch": 1.483846487424111, + "grad_norm": 0.0753197968006134, + "learning_rate": 1.0144349339910054e-05, + "loss": 0.0009, + "step": 13687 + }, + { + "epoch": 1.4839549002601908, + "grad_norm": 1.6921138763427734, + "learning_rate": 1.0143623966342668e-05, + "loss": 0.0367, + "step": 13688 + }, + { + "epoch": 1.4840633130962706, + "grad_norm": 0.7662966847419739, + "learning_rate": 1.0142898592775282e-05, + "loss": 0.0041, + "step": 13689 + }, + { + "epoch": 1.4841717259323504, + "grad_norm": 2.207954168319702, + "learning_rate": 1.0142173219207894e-05, + "loss": 0.0179, + "step": 13690 + }, + { + "epoch": 1.4842801387684301, + "grad_norm": 0.7690500020980835, + "learning_rate": 1.0141447845640506e-05, + "loss": 0.0054, + "step": 13691 + }, + { + "epoch": 1.48438855160451, + "grad_norm": 0.05776800587773323, + "learning_rate": 1.014072247207312e-05, + "loss": 0.0007, + "step": 13692 + }, + { + "epoch": 1.4844969644405897, + "grad_norm": 3.136356830596924, + "learning_rate": 1.0139997098505732e-05, + "loss": 0.0364, + "step": 13693 + }, + { + "epoch": 1.4846053772766696, + "grad_norm": 1.778024673461914, + "learning_rate": 1.0139271724938344e-05, + "loss": 0.0124, + "step": 13694 + }, + { + "epoch": 1.4847137901127494, + "grad_norm": 3.0271012783050537, + "learning_rate": 1.0138546351370958e-05, + "loss": 0.0419, + "step": 13695 + }, + { + "epoch": 1.4848222029488292, + "grad_norm": 3.6123239994049072, + "learning_rate": 1.013782097780357e-05, + "loss": 0.0298, + "step": 13696 + }, + { + "epoch": 1.484930615784909, + "grad_norm": 2.627802610397339, + "learning_rate": 1.0137095604236182e-05, + "loss": 0.0144, + "step": 13697 + }, + { + "epoch": 1.4850390286209887, + "grad_norm": 3.6461024284362793, + "learning_rate": 1.0136370230668794e-05, + "loss": 0.0475, + "step": 13698 + }, + { + "epoch": 1.4851474414570685, + "grad_norm": 1.6912957429885864, + "learning_rate": 1.0135644857101408e-05, + "loss": 0.0075, + "step": 13699 + }, + { + "epoch": 1.4852558542931482, + "grad_norm": 3.453763723373413, + "learning_rate": 1.013491948353402e-05, + "loss": 0.021, + "step": 13700 + }, + { + "epoch": 1.4853642671292282, + "grad_norm": 2.0090491771698, + "learning_rate": 1.0134194109966632e-05, + "loss": 0.0278, + "step": 13701 + }, + { + "epoch": 1.485472679965308, + "grad_norm": 1.503150224685669, + "learning_rate": 1.0133468736399246e-05, + "loss": 0.0031, + "step": 13702 + }, + { + "epoch": 1.4855810928013877, + "grad_norm": 1.697040319442749, + "learning_rate": 1.0132743362831858e-05, + "loss": 0.0106, + "step": 13703 + }, + { + "epoch": 1.4856895056374675, + "grad_norm": 2.747363805770874, + "learning_rate": 1.013201798926447e-05, + "loss": 0.0337, + "step": 13704 + }, + { + "epoch": 1.4857979184735473, + "grad_norm": 1.6492645740509033, + "learning_rate": 1.0131292615697086e-05, + "loss": 0.0262, + "step": 13705 + }, + { + "epoch": 1.485906331309627, + "grad_norm": 1.5054494142532349, + "learning_rate": 1.0130567242129698e-05, + "loss": 0.0125, + "step": 13706 + }, + { + "epoch": 1.4860147441457068, + "grad_norm": 0.9131182432174683, + "learning_rate": 1.0129841868562312e-05, + "loss": 0.0271, + "step": 13707 + }, + { + "epoch": 1.4861231569817868, + "grad_norm": 0.33409714698791504, + "learning_rate": 1.0129116494994924e-05, + "loss": 0.0035, + "step": 13708 + }, + { + "epoch": 1.4862315698178663, + "grad_norm": 3.1811108589172363, + "learning_rate": 1.0128391121427536e-05, + "loss": 0.0181, + "step": 13709 + }, + { + "epoch": 1.4863399826539463, + "grad_norm": 2.7369303703308105, + "learning_rate": 1.0127665747860148e-05, + "loss": 0.1103, + "step": 13710 + }, + { + "epoch": 1.486448395490026, + "grad_norm": 3.5381107330322266, + "learning_rate": 1.0126940374292762e-05, + "loss": 0.0161, + "step": 13711 + }, + { + "epoch": 1.4865568083261058, + "grad_norm": 0.3571731150150299, + "learning_rate": 1.0126215000725374e-05, + "loss": 0.0027, + "step": 13712 + }, + { + "epoch": 1.4866652211621856, + "grad_norm": 6.4459943771362305, + "learning_rate": 1.0125489627157987e-05, + "loss": 0.029, + "step": 13713 + }, + { + "epoch": 1.4867736339982653, + "grad_norm": 3.8079895973205566, + "learning_rate": 1.01247642535906e-05, + "loss": 0.0567, + "step": 13714 + }, + { + "epoch": 1.486882046834345, + "grad_norm": 2.5958380699157715, + "learning_rate": 1.0124038880023212e-05, + "loss": 0.0405, + "step": 13715 + }, + { + "epoch": 1.4869904596704249, + "grad_norm": 4.435421943664551, + "learning_rate": 1.0123313506455825e-05, + "loss": 0.0203, + "step": 13716 + }, + { + "epoch": 1.4870988725065049, + "grad_norm": 6.550109386444092, + "learning_rate": 1.0122588132888438e-05, + "loss": 0.0505, + "step": 13717 + }, + { + "epoch": 1.4872072853425846, + "grad_norm": 0.30197370052337646, + "learning_rate": 1.012186275932105e-05, + "loss": 0.0027, + "step": 13718 + }, + { + "epoch": 1.4873156981786644, + "grad_norm": 0.1298479288816452, + "learning_rate": 1.0121137385753663e-05, + "loss": 0.001, + "step": 13719 + }, + { + "epoch": 1.4874241110147441, + "grad_norm": 0.3191663920879364, + "learning_rate": 1.0120412012186277e-05, + "loss": 0.0029, + "step": 13720 + }, + { + "epoch": 1.487532523850824, + "grad_norm": 2.094675064086914, + "learning_rate": 1.0119686638618889e-05, + "loss": 0.0095, + "step": 13721 + }, + { + "epoch": 1.4876409366869037, + "grad_norm": 1.3953464031219482, + "learning_rate": 1.0118961265051504e-05, + "loss": 0.0143, + "step": 13722 + }, + { + "epoch": 1.4877493495229834, + "grad_norm": 2.294440984725952, + "learning_rate": 1.0118235891484116e-05, + "loss": 0.0328, + "step": 13723 + }, + { + "epoch": 1.4878577623590634, + "grad_norm": 0.6626320481300354, + "learning_rate": 1.0117510517916729e-05, + "loss": 0.0022, + "step": 13724 + }, + { + "epoch": 1.4879661751951432, + "grad_norm": 3.6306746006011963, + "learning_rate": 1.011678514434934e-05, + "loss": 0.0212, + "step": 13725 + }, + { + "epoch": 1.488074588031223, + "grad_norm": 1.5240780115127563, + "learning_rate": 1.0116059770781955e-05, + "loss": 0.007, + "step": 13726 + }, + { + "epoch": 1.4881830008673027, + "grad_norm": 0.23070187866687775, + "learning_rate": 1.0115334397214567e-05, + "loss": 0.0015, + "step": 13727 + }, + { + "epoch": 1.4882914137033825, + "grad_norm": 2.3579323291778564, + "learning_rate": 1.0114609023647179e-05, + "loss": 0.0706, + "step": 13728 + }, + { + "epoch": 1.4883998265394622, + "grad_norm": 0.47314146161079407, + "learning_rate": 1.0113883650079793e-05, + "loss": 0.0026, + "step": 13729 + }, + { + "epoch": 1.488508239375542, + "grad_norm": 6.494227409362793, + "learning_rate": 1.0113158276512405e-05, + "loss": 0.0259, + "step": 13730 + }, + { + "epoch": 1.488616652211622, + "grad_norm": 0.33039212226867676, + "learning_rate": 1.0112432902945017e-05, + "loss": 0.0019, + "step": 13731 + }, + { + "epoch": 1.4887250650477015, + "grad_norm": 6.1426615715026855, + "learning_rate": 1.011170752937763e-05, + "loss": 0.0216, + "step": 13732 + }, + { + "epoch": 1.4888334778837815, + "grad_norm": 0.8303987383842468, + "learning_rate": 1.0110982155810243e-05, + "loss": 0.0046, + "step": 13733 + }, + { + "epoch": 1.4889418907198613, + "grad_norm": 4.2013750076293945, + "learning_rate": 1.0110256782242855e-05, + "loss": 0.0508, + "step": 13734 + }, + { + "epoch": 1.489050303555941, + "grad_norm": 1.3223754167556763, + "learning_rate": 1.0109531408675469e-05, + "loss": 0.0186, + "step": 13735 + }, + { + "epoch": 1.4891587163920208, + "grad_norm": 2.7364397048950195, + "learning_rate": 1.0108806035108081e-05, + "loss": 0.0106, + "step": 13736 + }, + { + "epoch": 1.4892671292281006, + "grad_norm": 2.9760730266571045, + "learning_rate": 1.0108080661540693e-05, + "loss": 0.0382, + "step": 13737 + }, + { + "epoch": 1.4893755420641803, + "grad_norm": 0.09321416914463043, + "learning_rate": 1.0107355287973305e-05, + "loss": 0.0008, + "step": 13738 + }, + { + "epoch": 1.48948395490026, + "grad_norm": 3.84789776802063, + "learning_rate": 1.010662991440592e-05, + "loss": 0.0479, + "step": 13739 + }, + { + "epoch": 1.48959236773634, + "grad_norm": 1.9202221632003784, + "learning_rate": 1.0105904540838533e-05, + "loss": 0.0454, + "step": 13740 + }, + { + "epoch": 1.4897007805724198, + "grad_norm": 2.5787363052368164, + "learning_rate": 1.0105179167271147e-05, + "loss": 0.0096, + "step": 13741 + }, + { + "epoch": 1.4898091934084996, + "grad_norm": 0.9766278266906738, + "learning_rate": 1.0104453793703759e-05, + "loss": 0.0072, + "step": 13742 + }, + { + "epoch": 1.4899176062445794, + "grad_norm": 2.167609691619873, + "learning_rate": 1.0103728420136371e-05, + "loss": 0.0271, + "step": 13743 + }, + { + "epoch": 1.4900260190806591, + "grad_norm": 3.848961114883423, + "learning_rate": 1.0103003046568985e-05, + "loss": 0.0495, + "step": 13744 + }, + { + "epoch": 1.4901344319167389, + "grad_norm": 0.9975814819335938, + "learning_rate": 1.0102277673001597e-05, + "loss": 0.003, + "step": 13745 + }, + { + "epoch": 1.4902428447528187, + "grad_norm": 0.8493456840515137, + "learning_rate": 1.010155229943421e-05, + "loss": 0.0068, + "step": 13746 + }, + { + "epoch": 1.4903512575888986, + "grad_norm": 2.6017444133758545, + "learning_rate": 1.0100826925866823e-05, + "loss": 0.0184, + "step": 13747 + }, + { + "epoch": 1.4904596704249784, + "grad_norm": 4.125046730041504, + "learning_rate": 1.0100101552299435e-05, + "loss": 0.0716, + "step": 13748 + }, + { + "epoch": 1.4905680832610582, + "grad_norm": 3.8744585514068604, + "learning_rate": 1.0099376178732047e-05, + "loss": 0.0428, + "step": 13749 + }, + { + "epoch": 1.490676496097138, + "grad_norm": 2.145167589187622, + "learning_rate": 1.009865080516466e-05, + "loss": 0.0176, + "step": 13750 + }, + { + "epoch": 1.4907849089332177, + "grad_norm": 1.8756284713745117, + "learning_rate": 1.0097925431597273e-05, + "loss": 0.0288, + "step": 13751 + }, + { + "epoch": 1.4908933217692975, + "grad_norm": 2.4407761096954346, + "learning_rate": 1.0097200058029886e-05, + "loss": 0.0451, + "step": 13752 + }, + { + "epoch": 1.4910017346053772, + "grad_norm": 1.0526413917541504, + "learning_rate": 1.0096474684462498e-05, + "loss": 0.0104, + "step": 13753 + }, + { + "epoch": 1.4911101474414572, + "grad_norm": 0.7474502325057983, + "learning_rate": 1.0095749310895111e-05, + "loss": 0.0173, + "step": 13754 + }, + { + "epoch": 1.4912185602775367, + "grad_norm": 0.28733524680137634, + "learning_rate": 1.0095023937327724e-05, + "loss": 0.0024, + "step": 13755 + }, + { + "epoch": 1.4913269731136167, + "grad_norm": 2.8476693630218506, + "learning_rate": 1.0094298563760336e-05, + "loss": 0.0449, + "step": 13756 + }, + { + "epoch": 1.4914353859496965, + "grad_norm": 2.9259064197540283, + "learning_rate": 1.0093573190192951e-05, + "loss": 0.0215, + "step": 13757 + }, + { + "epoch": 1.4915437987857763, + "grad_norm": 2.377124547958374, + "learning_rate": 1.0092847816625563e-05, + "loss": 0.0187, + "step": 13758 + }, + { + "epoch": 1.491652211621856, + "grad_norm": 2.757261037826538, + "learning_rate": 1.0092122443058177e-05, + "loss": 0.0416, + "step": 13759 + }, + { + "epoch": 1.4917606244579358, + "grad_norm": 2.729388475418091, + "learning_rate": 1.009139706949079e-05, + "loss": 0.0469, + "step": 13760 + }, + { + "epoch": 1.4918690372940155, + "grad_norm": 1.410288691520691, + "learning_rate": 1.0090671695923402e-05, + "loss": 0.0049, + "step": 13761 + }, + { + "epoch": 1.4919774501300953, + "grad_norm": 1.2583200931549072, + "learning_rate": 1.0089946322356014e-05, + "loss": 0.0078, + "step": 13762 + }, + { + "epoch": 1.4920858629661753, + "grad_norm": 1.6680697202682495, + "learning_rate": 1.0089220948788628e-05, + "loss": 0.0182, + "step": 13763 + }, + { + "epoch": 1.492194275802255, + "grad_norm": 9.702043533325195, + "learning_rate": 1.008849557522124e-05, + "loss": 0.0314, + "step": 13764 + }, + { + "epoch": 1.4923026886383348, + "grad_norm": 0.5015084743499756, + "learning_rate": 1.0087770201653852e-05, + "loss": 0.0038, + "step": 13765 + }, + { + "epoch": 1.4924111014744146, + "grad_norm": 3.887892007827759, + "learning_rate": 1.0087044828086466e-05, + "loss": 0.0506, + "step": 13766 + }, + { + "epoch": 1.4925195143104943, + "grad_norm": 5.115818023681641, + "learning_rate": 1.0086319454519078e-05, + "loss": 0.0342, + "step": 13767 + }, + { + "epoch": 1.492627927146574, + "grad_norm": 2.099214792251587, + "learning_rate": 1.008559408095169e-05, + "loss": 0.0358, + "step": 13768 + }, + { + "epoch": 1.4927363399826539, + "grad_norm": 0.23913134634494781, + "learning_rate": 1.0084868707384304e-05, + "loss": 0.0024, + "step": 13769 + }, + { + "epoch": 1.4928447528187339, + "grad_norm": 2.2838778495788574, + "learning_rate": 1.0084143333816916e-05, + "loss": 0.0179, + "step": 13770 + }, + { + "epoch": 1.4929531656548134, + "grad_norm": 1.0603588819503784, + "learning_rate": 1.0083417960249528e-05, + "loss": 0.008, + "step": 13771 + }, + { + "epoch": 1.4930615784908934, + "grad_norm": 0.667354941368103, + "learning_rate": 1.0082692586682142e-05, + "loss": 0.0048, + "step": 13772 + }, + { + "epoch": 1.4931699913269731, + "grad_norm": 0.7345873117446899, + "learning_rate": 1.0081967213114754e-05, + "loss": 0.0068, + "step": 13773 + }, + { + "epoch": 1.493278404163053, + "grad_norm": 1.5481489896774292, + "learning_rate": 1.008124183954737e-05, + "loss": 0.0134, + "step": 13774 + }, + { + "epoch": 1.4933868169991327, + "grad_norm": 5.158048152923584, + "learning_rate": 1.0080516465979982e-05, + "loss": 0.0448, + "step": 13775 + }, + { + "epoch": 1.4934952298352124, + "grad_norm": 2.855713367462158, + "learning_rate": 1.0079791092412594e-05, + "loss": 0.0057, + "step": 13776 + }, + { + "epoch": 1.4936036426712924, + "grad_norm": 4.784635543823242, + "learning_rate": 1.0079065718845206e-05, + "loss": 0.0266, + "step": 13777 + }, + { + "epoch": 1.493712055507372, + "grad_norm": 1.0657793283462524, + "learning_rate": 1.007834034527782e-05, + "loss": 0.0123, + "step": 13778 + }, + { + "epoch": 1.493820468343452, + "grad_norm": 8.833027839660645, + "learning_rate": 1.0077614971710432e-05, + "loss": 0.0678, + "step": 13779 + }, + { + "epoch": 1.4939288811795317, + "grad_norm": 2.864414691925049, + "learning_rate": 1.0076889598143044e-05, + "loss": 0.0278, + "step": 13780 + }, + { + "epoch": 1.4940372940156115, + "grad_norm": 3.3246448040008545, + "learning_rate": 1.0076164224575658e-05, + "loss": 0.0309, + "step": 13781 + }, + { + "epoch": 1.4941457068516912, + "grad_norm": 5.994459629058838, + "learning_rate": 1.007543885100827e-05, + "loss": 0.0461, + "step": 13782 + }, + { + "epoch": 1.494254119687771, + "grad_norm": 1.582712173461914, + "learning_rate": 1.0074713477440882e-05, + "loss": 0.0129, + "step": 13783 + }, + { + "epoch": 1.4943625325238508, + "grad_norm": 0.3368752598762512, + "learning_rate": 1.0073988103873496e-05, + "loss": 0.0041, + "step": 13784 + }, + { + "epoch": 1.4944709453599305, + "grad_norm": 1.349961280822754, + "learning_rate": 1.0073262730306108e-05, + "loss": 0.0187, + "step": 13785 + }, + { + "epoch": 1.4945793581960105, + "grad_norm": 0.1979830116033554, + "learning_rate": 1.007253735673872e-05, + "loss": 0.0012, + "step": 13786 + }, + { + "epoch": 1.4946877710320903, + "grad_norm": 2.597350597381592, + "learning_rate": 1.0071811983171334e-05, + "loss": 0.0219, + "step": 13787 + }, + { + "epoch": 1.49479618386817, + "grad_norm": 3.7835631370544434, + "learning_rate": 1.0071086609603946e-05, + "loss": 0.0262, + "step": 13788 + }, + { + "epoch": 1.4949045967042498, + "grad_norm": 0.41736626625061035, + "learning_rate": 1.0070361236036559e-05, + "loss": 0.0031, + "step": 13789 + }, + { + "epoch": 1.4950130095403296, + "grad_norm": 0.6210877299308777, + "learning_rate": 1.006963586246917e-05, + "loss": 0.0065, + "step": 13790 + }, + { + "epoch": 1.4951214223764093, + "grad_norm": 1.1219110488891602, + "learning_rate": 1.0068910488901786e-05, + "loss": 0.0071, + "step": 13791 + }, + { + "epoch": 1.495229835212489, + "grad_norm": 2.3569984436035156, + "learning_rate": 1.0068185115334398e-05, + "loss": 0.0118, + "step": 13792 + }, + { + "epoch": 1.495338248048569, + "grad_norm": 0.1672455370426178, + "learning_rate": 1.0067459741767012e-05, + "loss": 0.0021, + "step": 13793 + }, + { + "epoch": 1.4954466608846486, + "grad_norm": 1.2663707733154297, + "learning_rate": 1.0066734368199624e-05, + "loss": 0.0139, + "step": 13794 + }, + { + "epoch": 1.4955550737207286, + "grad_norm": 0.2389967441558838, + "learning_rate": 1.0066008994632236e-05, + "loss": 0.0032, + "step": 13795 + }, + { + "epoch": 1.4956634865568084, + "grad_norm": 0.7805501222610474, + "learning_rate": 1.006528362106485e-05, + "loss": 0.0123, + "step": 13796 + }, + { + "epoch": 1.4957718993928881, + "grad_norm": 0.3971271812915802, + "learning_rate": 1.0064558247497462e-05, + "loss": 0.0034, + "step": 13797 + }, + { + "epoch": 1.495880312228968, + "grad_norm": 1.2755136489868164, + "learning_rate": 1.0063832873930075e-05, + "loss": 0.0173, + "step": 13798 + }, + { + "epoch": 1.4959887250650477, + "grad_norm": 0.6300585865974426, + "learning_rate": 1.0063107500362688e-05, + "loss": 0.0044, + "step": 13799 + }, + { + "epoch": 1.4960971379011274, + "grad_norm": 1.4975688457489014, + "learning_rate": 1.00623821267953e-05, + "loss": 0.0153, + "step": 13800 + }, + { + "epoch": 1.4962055507372072, + "grad_norm": 1.0559481382369995, + "learning_rate": 1.0061656753227913e-05, + "loss": 0.0068, + "step": 13801 + }, + { + "epoch": 1.4963139635732872, + "grad_norm": 5.910882949829102, + "learning_rate": 1.0060931379660525e-05, + "loss": 0.0241, + "step": 13802 + }, + { + "epoch": 1.496422376409367, + "grad_norm": 3.617215633392334, + "learning_rate": 1.0060206006093139e-05, + "loss": 0.081, + "step": 13803 + }, + { + "epoch": 1.4965307892454467, + "grad_norm": 2.980299472808838, + "learning_rate": 1.0059480632525751e-05, + "loss": 0.0155, + "step": 13804 + }, + { + "epoch": 1.4966392020815265, + "grad_norm": 0.6876992583274841, + "learning_rate": 1.0058755258958363e-05, + "loss": 0.0026, + "step": 13805 + }, + { + "epoch": 1.4967476149176062, + "grad_norm": 6.2382354736328125, + "learning_rate": 1.0058029885390977e-05, + "loss": 0.1096, + "step": 13806 + }, + { + "epoch": 1.496856027753686, + "grad_norm": 0.8156270384788513, + "learning_rate": 1.0057304511823589e-05, + "loss": 0.0036, + "step": 13807 + }, + { + "epoch": 1.4969644405897657, + "grad_norm": 2.5977137088775635, + "learning_rate": 1.0056579138256201e-05, + "loss": 0.032, + "step": 13808 + }, + { + "epoch": 1.4970728534258457, + "grad_norm": 1.8210538625717163, + "learning_rate": 1.0055853764688817e-05, + "loss": 0.0117, + "step": 13809 + }, + { + "epoch": 1.4971812662619255, + "grad_norm": 2.4406638145446777, + "learning_rate": 1.0055128391121429e-05, + "loss": 0.0681, + "step": 13810 + }, + { + "epoch": 1.4972896790980053, + "grad_norm": 1.255943775177002, + "learning_rate": 1.0054403017554043e-05, + "loss": 0.0085, + "step": 13811 + }, + { + "epoch": 1.497398091934085, + "grad_norm": 0.3521621525287628, + "learning_rate": 1.0053677643986655e-05, + "loss": 0.0016, + "step": 13812 + }, + { + "epoch": 1.4975065047701648, + "grad_norm": 3.519860029220581, + "learning_rate": 1.0052952270419267e-05, + "loss": 0.0293, + "step": 13813 + }, + { + "epoch": 1.4976149176062445, + "grad_norm": 2.42350435256958, + "learning_rate": 1.005222689685188e-05, + "loss": 0.0101, + "step": 13814 + }, + { + "epoch": 1.4977233304423243, + "grad_norm": 2.992737293243408, + "learning_rate": 1.0051501523284493e-05, + "loss": 0.0093, + "step": 13815 + }, + { + "epoch": 1.4978317432784043, + "grad_norm": 5.025697708129883, + "learning_rate": 1.0050776149717105e-05, + "loss": 0.092, + "step": 13816 + }, + { + "epoch": 1.4979401561144838, + "grad_norm": 0.23526275157928467, + "learning_rate": 1.0050050776149717e-05, + "loss": 0.0017, + "step": 13817 + }, + { + "epoch": 1.4980485689505638, + "grad_norm": 3.5477185249328613, + "learning_rate": 1.0049325402582331e-05, + "loss": 0.0164, + "step": 13818 + }, + { + "epoch": 1.4981569817866436, + "grad_norm": 2.062041759490967, + "learning_rate": 1.0048600029014943e-05, + "loss": 0.0094, + "step": 13819 + }, + { + "epoch": 1.4982653946227233, + "grad_norm": 1.5823451280593872, + "learning_rate": 1.0047874655447555e-05, + "loss": 0.0124, + "step": 13820 + }, + { + "epoch": 1.498373807458803, + "grad_norm": 0.37717321515083313, + "learning_rate": 1.0047149281880169e-05, + "loss": 0.0015, + "step": 13821 + }, + { + "epoch": 1.4984822202948829, + "grad_norm": 3.1663613319396973, + "learning_rate": 1.0046423908312781e-05, + "loss": 0.0648, + "step": 13822 + }, + { + "epoch": 1.4985906331309626, + "grad_norm": 1.7521663904190063, + "learning_rate": 1.0045698534745393e-05, + "loss": 0.0294, + "step": 13823 + }, + { + "epoch": 1.4986990459670424, + "grad_norm": 0.16964472830295563, + "learning_rate": 1.0044973161178007e-05, + "loss": 0.0014, + "step": 13824 + }, + { + "epoch": 1.4988074588031224, + "grad_norm": 4.695245265960693, + "learning_rate": 1.004424778761062e-05, + "loss": 0.0654, + "step": 13825 + }, + { + "epoch": 1.4989158716392021, + "grad_norm": 1.706323504447937, + "learning_rate": 1.0043522414043235e-05, + "loss": 0.0221, + "step": 13826 + }, + { + "epoch": 1.499024284475282, + "grad_norm": 0.4726199507713318, + "learning_rate": 1.0042797040475847e-05, + "loss": 0.0024, + "step": 13827 + }, + { + "epoch": 1.4991326973113617, + "grad_norm": 3.6176388263702393, + "learning_rate": 1.004207166690846e-05, + "loss": 0.0625, + "step": 13828 + }, + { + "epoch": 1.4992411101474414, + "grad_norm": 5.913489818572998, + "learning_rate": 1.0041346293341071e-05, + "loss": 0.0654, + "step": 13829 + }, + { + "epoch": 1.4993495229835212, + "grad_norm": 7.810354709625244, + "learning_rate": 1.0040620919773685e-05, + "loss": 0.0851, + "step": 13830 + }, + { + "epoch": 1.499457935819601, + "grad_norm": 5.925777912139893, + "learning_rate": 1.0039895546206297e-05, + "loss": 0.0389, + "step": 13831 + }, + { + "epoch": 1.499566348655681, + "grad_norm": 4.764732837677002, + "learning_rate": 1.003917017263891e-05, + "loss": 0.0467, + "step": 13832 + }, + { + "epoch": 1.4996747614917605, + "grad_norm": 2.7887308597564697, + "learning_rate": 1.0038444799071523e-05, + "loss": 0.0292, + "step": 13833 + }, + { + "epoch": 1.4997831743278405, + "grad_norm": 0.2300749272108078, + "learning_rate": 1.0037719425504135e-05, + "loss": 0.0022, + "step": 13834 + }, + { + "epoch": 1.4998915871639202, + "grad_norm": 4.006282329559326, + "learning_rate": 1.0036994051936748e-05, + "loss": 0.0705, + "step": 13835 + }, + { + "epoch": 1.5, + "grad_norm": 4.531085968017578, + "learning_rate": 1.0036268678369361e-05, + "loss": 0.026, + "step": 13836 + }, + { + "epoch": 1.5001084128360798, + "grad_norm": 4.4877119064331055, + "learning_rate": 1.0035543304801974e-05, + "loss": 0.0264, + "step": 13837 + }, + { + "epoch": 1.5002168256721595, + "grad_norm": 1.7553181648254395, + "learning_rate": 1.0034817931234586e-05, + "loss": 0.0445, + "step": 13838 + }, + { + "epoch": 1.5003252385082395, + "grad_norm": 2.470978021621704, + "learning_rate": 1.00340925576672e-05, + "loss": 0.0191, + "step": 13839 + }, + { + "epoch": 1.500433651344319, + "grad_norm": 5.807684898376465, + "learning_rate": 1.0033367184099812e-05, + "loss": 0.0747, + "step": 13840 + }, + { + "epoch": 1.500542064180399, + "grad_norm": 1.2715721130371094, + "learning_rate": 1.0032641810532424e-05, + "loss": 0.0207, + "step": 13841 + }, + { + "epoch": 1.5006504770164788, + "grad_norm": 1.0779861211776733, + "learning_rate": 1.0031916436965036e-05, + "loss": 0.0167, + "step": 13842 + }, + { + "epoch": 1.5007588898525586, + "grad_norm": 0.28257328271865845, + "learning_rate": 1.0031191063397652e-05, + "loss": 0.0031, + "step": 13843 + }, + { + "epoch": 1.5008673026886383, + "grad_norm": 8.428487777709961, + "learning_rate": 1.0030465689830264e-05, + "loss": 0.047, + "step": 13844 + }, + { + "epoch": 1.500975715524718, + "grad_norm": 1.6525845527648926, + "learning_rate": 1.0029740316262878e-05, + "loss": 0.0236, + "step": 13845 + }, + { + "epoch": 1.501084128360798, + "grad_norm": 0.9106578230857849, + "learning_rate": 1.002901494269549e-05, + "loss": 0.0168, + "step": 13846 + }, + { + "epoch": 1.5011925411968776, + "grad_norm": 6.953898906707764, + "learning_rate": 1.0028289569128102e-05, + "loss": 0.0126, + "step": 13847 + }, + { + "epoch": 1.5013009540329576, + "grad_norm": 3.4105069637298584, + "learning_rate": 1.0027564195560716e-05, + "loss": 0.0197, + "step": 13848 + }, + { + "epoch": 1.5014093668690371, + "grad_norm": 4.818270206451416, + "learning_rate": 1.0026838821993328e-05, + "loss": 0.0148, + "step": 13849 + }, + { + "epoch": 1.5015177797051171, + "grad_norm": 5.896425724029541, + "learning_rate": 1.002611344842594e-05, + "loss": 0.0214, + "step": 13850 + }, + { + "epoch": 1.501626192541197, + "grad_norm": 0.20602072775363922, + "learning_rate": 1.0025388074858554e-05, + "loss": 0.0012, + "step": 13851 + }, + { + "epoch": 1.5017346053772767, + "grad_norm": 2.6388936042785645, + "learning_rate": 1.0024662701291166e-05, + "loss": 0.0337, + "step": 13852 + }, + { + "epoch": 1.5018430182133564, + "grad_norm": 0.6080498099327087, + "learning_rate": 1.0023937327723778e-05, + "loss": 0.0046, + "step": 13853 + }, + { + "epoch": 1.5019514310494362, + "grad_norm": 3.2436769008636475, + "learning_rate": 1.002321195415639e-05, + "loss": 0.0167, + "step": 13854 + }, + { + "epoch": 1.5020598438855162, + "grad_norm": 1.777742624282837, + "learning_rate": 1.0022486580589004e-05, + "loss": 0.0126, + "step": 13855 + }, + { + "epoch": 1.5021682567215957, + "grad_norm": 2.103224515914917, + "learning_rate": 1.0021761207021616e-05, + "loss": 0.0293, + "step": 13856 + }, + { + "epoch": 1.5022766695576757, + "grad_norm": 0.5598563551902771, + "learning_rate": 1.0021035833454228e-05, + "loss": 0.0048, + "step": 13857 + }, + { + "epoch": 1.5023850823937555, + "grad_norm": 2.705043315887451, + "learning_rate": 1.0020310459886842e-05, + "loss": 0.0384, + "step": 13858 + }, + { + "epoch": 1.5024934952298352, + "grad_norm": 1.1304950714111328, + "learning_rate": 1.0019585086319454e-05, + "loss": 0.0198, + "step": 13859 + }, + { + "epoch": 1.502601908065915, + "grad_norm": 0.45111358165740967, + "learning_rate": 1.0018859712752066e-05, + "loss": 0.0059, + "step": 13860 + }, + { + "epoch": 1.5027103209019947, + "grad_norm": 0.5551831126213074, + "learning_rate": 1.0018134339184682e-05, + "loss": 0.0035, + "step": 13861 + }, + { + "epoch": 1.5028187337380747, + "grad_norm": 0.6525649428367615, + "learning_rate": 1.0017408965617294e-05, + "loss": 0.0066, + "step": 13862 + }, + { + "epoch": 1.5029271465741543, + "grad_norm": 4.572358131408691, + "learning_rate": 1.0016683592049908e-05, + "loss": 0.0102, + "step": 13863 + }, + { + "epoch": 1.5030355594102343, + "grad_norm": 0.11104831844568253, + "learning_rate": 1.001595821848252e-05, + "loss": 0.0014, + "step": 13864 + }, + { + "epoch": 1.5031439722463138, + "grad_norm": 1.7160013914108276, + "learning_rate": 1.0015232844915132e-05, + "loss": 0.0105, + "step": 13865 + }, + { + "epoch": 1.5032523850823938, + "grad_norm": 1.887343168258667, + "learning_rate": 1.0014507471347746e-05, + "loss": 0.0128, + "step": 13866 + }, + { + "epoch": 1.5033607979184735, + "grad_norm": 1.1680232286453247, + "learning_rate": 1.0013782097780358e-05, + "loss": 0.0018, + "step": 13867 + }, + { + "epoch": 1.5034692107545533, + "grad_norm": 0.960898756980896, + "learning_rate": 1.001305672421297e-05, + "loss": 0.0078, + "step": 13868 + }, + { + "epoch": 1.5035776235906333, + "grad_norm": 0.20384858548641205, + "learning_rate": 1.0012331350645583e-05, + "loss": 0.0012, + "step": 13869 + }, + { + "epoch": 1.5036860364267128, + "grad_norm": 6.634060382843018, + "learning_rate": 1.0011605977078196e-05, + "loss": 0.0467, + "step": 13870 + }, + { + "epoch": 1.5037944492627928, + "grad_norm": 0.3513566553592682, + "learning_rate": 1.0010880603510809e-05, + "loss": 0.0019, + "step": 13871 + }, + { + "epoch": 1.5039028620988724, + "grad_norm": 0.8164077401161194, + "learning_rate": 1.001015522994342e-05, + "loss": 0.0051, + "step": 13872 + }, + { + "epoch": 1.5040112749349523, + "grad_norm": 3.945807933807373, + "learning_rate": 1.0009429856376034e-05, + "loss": 0.0124, + "step": 13873 + }, + { + "epoch": 1.504119687771032, + "grad_norm": 2.261516571044922, + "learning_rate": 1.0008704482808647e-05, + "loss": 0.0192, + "step": 13874 + }, + { + "epoch": 1.5042281006071119, + "grad_norm": 1.8509562015533447, + "learning_rate": 1.0007979109241259e-05, + "loss": 0.0558, + "step": 13875 + }, + { + "epoch": 1.5043365134431916, + "grad_norm": 1.6213277578353882, + "learning_rate": 1.0007253735673873e-05, + "loss": 0.0336, + "step": 13876 + }, + { + "epoch": 1.5044449262792714, + "grad_norm": 3.304058790206909, + "learning_rate": 1.0006528362106485e-05, + "loss": 0.0374, + "step": 13877 + }, + { + "epoch": 1.5045533391153514, + "grad_norm": 0.1287023276090622, + "learning_rate": 1.00058029885391e-05, + "loss": 0.0007, + "step": 13878 + }, + { + "epoch": 1.504661751951431, + "grad_norm": 0.8313965201377869, + "learning_rate": 1.0005077614971712e-05, + "loss": 0.0049, + "step": 13879 + }, + { + "epoch": 1.504770164787511, + "grad_norm": 12.751513481140137, + "learning_rate": 1.0004352241404325e-05, + "loss": 0.0604, + "step": 13880 + }, + { + "epoch": 1.5048785776235907, + "grad_norm": 1.8777190446853638, + "learning_rate": 1.0003626867836937e-05, + "loss": 0.0323, + "step": 13881 + }, + { + "epoch": 1.5049869904596704, + "grad_norm": 3.251375436782837, + "learning_rate": 1.000290149426955e-05, + "loss": 0.0191, + "step": 13882 + }, + { + "epoch": 1.5050954032957502, + "grad_norm": 3.3134043216705322, + "learning_rate": 1.0002176120702163e-05, + "loss": 0.0137, + "step": 13883 + }, + { + "epoch": 1.50520381613183, + "grad_norm": 3.3566222190856934, + "learning_rate": 1.0001450747134775e-05, + "loss": 0.0104, + "step": 13884 + }, + { + "epoch": 1.50531222896791, + "grad_norm": 6.650158405303955, + "learning_rate": 1.0000725373567389e-05, + "loss": 0.0151, + "step": 13885 + }, + { + "epoch": 1.5054206418039895, + "grad_norm": 0.5109790563583374, + "learning_rate": 1e-05, + "loss": 0.0038, + "step": 13886 + }, + { + "epoch": 1.5055290546400695, + "grad_norm": 2.8372573852539062, + "learning_rate": 9.999274626432613e-06, + "loss": 0.1169, + "step": 13887 + }, + { + "epoch": 1.505637467476149, + "grad_norm": 0.3923371434211731, + "learning_rate": 9.998549252865227e-06, + "loss": 0.004, + "step": 13888 + }, + { + "epoch": 1.505745880312229, + "grad_norm": 2.8637282848358154, + "learning_rate": 9.997823879297839e-06, + "loss": 0.056, + "step": 13889 + }, + { + "epoch": 1.5058542931483088, + "grad_norm": 1.162330985069275, + "learning_rate": 9.997098505730451e-06, + "loss": 0.0051, + "step": 13890 + }, + { + "epoch": 1.5059627059843885, + "grad_norm": 2.934067964553833, + "learning_rate": 9.996373132163065e-06, + "loss": 0.0988, + "step": 13891 + }, + { + "epoch": 1.5060711188204683, + "grad_norm": 1.2807573080062866, + "learning_rate": 9.995647758595679e-06, + "loss": 0.0213, + "step": 13892 + }, + { + "epoch": 1.506179531656548, + "grad_norm": 4.850294589996338, + "learning_rate": 9.994922385028291e-06, + "loss": 0.0195, + "step": 13893 + }, + { + "epoch": 1.506287944492628, + "grad_norm": 1.2368836402893066, + "learning_rate": 9.994197011460903e-06, + "loss": 0.0188, + "step": 13894 + }, + { + "epoch": 1.5063963573287076, + "grad_norm": 4.764686584472656, + "learning_rate": 9.993471637893515e-06, + "loss": 0.0807, + "step": 13895 + }, + { + "epoch": 1.5065047701647876, + "grad_norm": 0.2163659930229187, + "learning_rate": 9.992746264326129e-06, + "loss": 0.0015, + "step": 13896 + }, + { + "epoch": 1.5066131830008673, + "grad_norm": 5.2471184730529785, + "learning_rate": 9.992020890758741e-06, + "loss": 0.0308, + "step": 13897 + }, + { + "epoch": 1.506721595836947, + "grad_norm": 2.999235153198242, + "learning_rate": 9.991295517191353e-06, + "loss": 0.0085, + "step": 13898 + }, + { + "epoch": 1.5068300086730269, + "grad_norm": 2.626810073852539, + "learning_rate": 9.990570143623967e-06, + "loss": 0.0327, + "step": 13899 + }, + { + "epoch": 1.5069384215091066, + "grad_norm": 4.109116077423096, + "learning_rate": 9.989844770056581e-06, + "loss": 0.0636, + "step": 13900 + }, + { + "epoch": 1.5070468343451866, + "grad_norm": 2.4420955181121826, + "learning_rate": 9.989119396489193e-06, + "loss": 0.0291, + "step": 13901 + }, + { + "epoch": 1.5071552471812661, + "grad_norm": 4.076071739196777, + "learning_rate": 9.988394022921805e-06, + "loss": 0.0093, + "step": 13902 + }, + { + "epoch": 1.5072636600173461, + "grad_norm": 1.5168925523757935, + "learning_rate": 9.987668649354419e-06, + "loss": 0.0348, + "step": 13903 + }, + { + "epoch": 1.507372072853426, + "grad_norm": 8.315774917602539, + "learning_rate": 9.986943275787031e-06, + "loss": 0.0607, + "step": 13904 + }, + { + "epoch": 1.5074804856895057, + "grad_norm": 4.1890869140625, + "learning_rate": 9.986217902219643e-06, + "loss": 0.0237, + "step": 13905 + }, + { + "epoch": 1.5075888985255854, + "grad_norm": 2.5234930515289307, + "learning_rate": 9.985492528652256e-06, + "loss": 0.0353, + "step": 13906 + }, + { + "epoch": 1.5076973113616652, + "grad_norm": 2.2482588291168213, + "learning_rate": 9.98476715508487e-06, + "loss": 0.0379, + "step": 13907 + }, + { + "epoch": 1.5078057241977452, + "grad_norm": 2.767578601837158, + "learning_rate": 9.984041781517483e-06, + "loss": 0.044, + "step": 13908 + }, + { + "epoch": 1.5079141370338247, + "grad_norm": 2.816660165786743, + "learning_rate": 9.983316407950095e-06, + "loss": 0.0144, + "step": 13909 + }, + { + "epoch": 1.5080225498699047, + "grad_norm": 0.2759004235267639, + "learning_rate": 9.982591034382708e-06, + "loss": 0.0012, + "step": 13910 + }, + { + "epoch": 1.5081309627059842, + "grad_norm": 2.3703346252441406, + "learning_rate": 9.981865660815321e-06, + "loss": 0.0406, + "step": 13911 + }, + { + "epoch": 1.5082393755420642, + "grad_norm": 4.48095178604126, + "learning_rate": 9.981140287247933e-06, + "loss": 0.0285, + "step": 13912 + }, + { + "epoch": 1.508347788378144, + "grad_norm": 7.42864990234375, + "learning_rate": 9.980414913680546e-06, + "loss": 0.0095, + "step": 13913 + }, + { + "epoch": 1.5084562012142237, + "grad_norm": 1.0067260265350342, + "learning_rate": 9.97968954011316e-06, + "loss": 0.0035, + "step": 13914 + }, + { + "epoch": 1.5085646140503035, + "grad_norm": 1.1796541213989258, + "learning_rate": 9.978964166545772e-06, + "loss": 0.0328, + "step": 13915 + }, + { + "epoch": 1.5086730268863833, + "grad_norm": 3.0226218700408936, + "learning_rate": 9.978238792978384e-06, + "loss": 0.0107, + "step": 13916 + }, + { + "epoch": 1.5087814397224633, + "grad_norm": 3.924161195755005, + "learning_rate": 9.977513419410998e-06, + "loss": 0.0444, + "step": 13917 + }, + { + "epoch": 1.5088898525585428, + "grad_norm": 3.6168274879455566, + "learning_rate": 9.976788045843611e-06, + "loss": 0.0055, + "step": 13918 + }, + { + "epoch": 1.5089982653946228, + "grad_norm": 3.1662003993988037, + "learning_rate": 9.976062672276224e-06, + "loss": 0.0195, + "step": 13919 + }, + { + "epoch": 1.5091066782307025, + "grad_norm": 2.463879108428955, + "learning_rate": 9.975337298708836e-06, + "loss": 0.009, + "step": 13920 + }, + { + "epoch": 1.5092150910667823, + "grad_norm": 3.814218759536743, + "learning_rate": 9.974611925141448e-06, + "loss": 0.0311, + "step": 13921 + }, + { + "epoch": 1.509323503902862, + "grad_norm": 0.9333966374397278, + "learning_rate": 9.973886551574062e-06, + "loss": 0.0199, + "step": 13922 + }, + { + "epoch": 1.5094319167389418, + "grad_norm": 1.273962378501892, + "learning_rate": 9.973161178006674e-06, + "loss": 0.0473, + "step": 13923 + }, + { + "epoch": 1.5095403295750218, + "grad_norm": 2.090754985809326, + "learning_rate": 9.972435804439286e-06, + "loss": 0.0078, + "step": 13924 + }, + { + "epoch": 1.5096487424111014, + "grad_norm": 1.1424965858459473, + "learning_rate": 9.9717104308719e-06, + "loss": 0.0464, + "step": 13925 + }, + { + "epoch": 1.5097571552471813, + "grad_norm": 1.7465158700942993, + "learning_rate": 9.970985057304514e-06, + "loss": 0.0188, + "step": 13926 + }, + { + "epoch": 1.5098655680832611, + "grad_norm": 0.182902991771698, + "learning_rate": 9.970259683737126e-06, + "loss": 0.0016, + "step": 13927 + }, + { + "epoch": 1.5099739809193409, + "grad_norm": 1.8096611499786377, + "learning_rate": 9.969534310169738e-06, + "loss": 0.0246, + "step": 13928 + }, + { + "epoch": 1.5100823937554206, + "grad_norm": 2.020535707473755, + "learning_rate": 9.968808936602352e-06, + "loss": 0.0131, + "step": 13929 + }, + { + "epoch": 1.5101908065915004, + "grad_norm": 1.8808367252349854, + "learning_rate": 9.968083563034964e-06, + "loss": 0.0361, + "step": 13930 + }, + { + "epoch": 1.5102992194275804, + "grad_norm": 1.406698226928711, + "learning_rate": 9.967358189467576e-06, + "loss": 0.0273, + "step": 13931 + }, + { + "epoch": 1.51040763226366, + "grad_norm": 5.254589080810547, + "learning_rate": 9.966632815900188e-06, + "loss": 0.039, + "step": 13932 + }, + { + "epoch": 1.51051604509974, + "grad_norm": 1.4533284902572632, + "learning_rate": 9.965907442332802e-06, + "loss": 0.0347, + "step": 13933 + }, + { + "epoch": 1.5106244579358195, + "grad_norm": 0.4910569489002228, + "learning_rate": 9.965182068765416e-06, + "loss": 0.0031, + "step": 13934 + }, + { + "epoch": 1.5107328707718994, + "grad_norm": 1.6701463460922241, + "learning_rate": 9.964456695198028e-06, + "loss": 0.0094, + "step": 13935 + }, + { + "epoch": 1.5108412836079792, + "grad_norm": 6.373041152954102, + "learning_rate": 9.96373132163064e-06, + "loss": 0.0227, + "step": 13936 + }, + { + "epoch": 1.510949696444059, + "grad_norm": 1.5298887491226196, + "learning_rate": 9.963005948063254e-06, + "loss": 0.0092, + "step": 13937 + }, + { + "epoch": 1.5110581092801387, + "grad_norm": 2.6552650928497314, + "learning_rate": 9.962280574495866e-06, + "loss": 0.0328, + "step": 13938 + }, + { + "epoch": 1.5111665221162185, + "grad_norm": 0.9556626081466675, + "learning_rate": 9.961555200928478e-06, + "loss": 0.0087, + "step": 13939 + }, + { + "epoch": 1.5112749349522985, + "grad_norm": 0.8440970182418823, + "learning_rate": 9.960829827361092e-06, + "loss": 0.0105, + "step": 13940 + }, + { + "epoch": 1.511383347788378, + "grad_norm": 2.2964420318603516, + "learning_rate": 9.960104453793704e-06, + "loss": 0.0548, + "step": 13941 + }, + { + "epoch": 1.511491760624458, + "grad_norm": 2.9740848541259766, + "learning_rate": 9.959379080226316e-06, + "loss": 0.08, + "step": 13942 + }, + { + "epoch": 1.5116001734605378, + "grad_norm": 2.04079532623291, + "learning_rate": 9.95865370665893e-06, + "loss": 0.0299, + "step": 13943 + }, + { + "epoch": 1.5117085862966175, + "grad_norm": 0.375931978225708, + "learning_rate": 9.957928333091544e-06, + "loss": 0.0037, + "step": 13944 + }, + { + "epoch": 1.5118169991326973, + "grad_norm": 3.0304338932037354, + "learning_rate": 9.957202959524156e-06, + "loss": 0.0248, + "step": 13945 + }, + { + "epoch": 1.511925411968777, + "grad_norm": 1.6332931518554688, + "learning_rate": 9.956477585956768e-06, + "loss": 0.0226, + "step": 13946 + }, + { + "epoch": 1.512033824804857, + "grad_norm": 2.842068672180176, + "learning_rate": 9.95575221238938e-06, + "loss": 0.0586, + "step": 13947 + }, + { + "epoch": 1.5121422376409366, + "grad_norm": 0.42921674251556396, + "learning_rate": 9.955026838821994e-06, + "loss": 0.0073, + "step": 13948 + }, + { + "epoch": 1.5122506504770166, + "grad_norm": 2.43505597114563, + "learning_rate": 9.954301465254607e-06, + "loss": 0.0167, + "step": 13949 + }, + { + "epoch": 1.512359063313096, + "grad_norm": 1.2995874881744385, + "learning_rate": 9.953576091687219e-06, + "loss": 0.012, + "step": 13950 + }, + { + "epoch": 1.512467476149176, + "grad_norm": 0.47378429770469666, + "learning_rate": 9.952850718119832e-06, + "loss": 0.0058, + "step": 13951 + }, + { + "epoch": 1.5125758889852559, + "grad_norm": 3.4838647842407227, + "learning_rate": 9.952125344552446e-06, + "loss": 0.0373, + "step": 13952 + }, + { + "epoch": 1.5126843018213356, + "grad_norm": 1.3980464935302734, + "learning_rate": 9.951399970985058e-06, + "loss": 0.0049, + "step": 13953 + }, + { + "epoch": 1.5127927146574154, + "grad_norm": 0.6418552398681641, + "learning_rate": 9.95067459741767e-06, + "loss": 0.004, + "step": 13954 + }, + { + "epoch": 1.5129011274934951, + "grad_norm": 1.986188530921936, + "learning_rate": 9.949949223850284e-06, + "loss": 0.0176, + "step": 13955 + }, + { + "epoch": 1.5130095403295751, + "grad_norm": 2.523327589035034, + "learning_rate": 9.949223850282897e-06, + "loss": 0.0219, + "step": 13956 + }, + { + "epoch": 1.5131179531656547, + "grad_norm": 2.1131858825683594, + "learning_rate": 9.948498476715509e-06, + "loss": 0.0454, + "step": 13957 + }, + { + "epoch": 1.5132263660017347, + "grad_norm": 1.1269099712371826, + "learning_rate": 9.947773103148121e-06, + "loss": 0.0148, + "step": 13958 + }, + { + "epoch": 1.5133347788378144, + "grad_norm": 3.4144647121429443, + "learning_rate": 9.947047729580735e-06, + "loss": 0.0493, + "step": 13959 + }, + { + "epoch": 1.5134431916738942, + "grad_norm": 0.9635962843894958, + "learning_rate": 9.946322356013349e-06, + "loss": 0.0265, + "step": 13960 + }, + { + "epoch": 1.513551604509974, + "grad_norm": 4.363738536834717, + "learning_rate": 9.94559698244596e-06, + "loss": 0.0096, + "step": 13961 + }, + { + "epoch": 1.5136600173460537, + "grad_norm": 3.9042632579803467, + "learning_rate": 9.944871608878573e-06, + "loss": 0.0239, + "step": 13962 + }, + { + "epoch": 1.5137684301821337, + "grad_norm": 1.0448763370513916, + "learning_rate": 9.944146235311187e-06, + "loss": 0.0049, + "step": 13963 + }, + { + "epoch": 1.5138768430182132, + "grad_norm": 1.4771935939788818, + "learning_rate": 9.943420861743799e-06, + "loss": 0.0307, + "step": 13964 + }, + { + "epoch": 1.5139852558542932, + "grad_norm": 2.73420786857605, + "learning_rate": 9.942695488176411e-06, + "loss": 0.0438, + "step": 13965 + }, + { + "epoch": 1.514093668690373, + "grad_norm": 0.9628405570983887, + "learning_rate": 9.941970114609025e-06, + "loss": 0.0201, + "step": 13966 + }, + { + "epoch": 1.5142020815264527, + "grad_norm": 2.103389263153076, + "learning_rate": 9.941244741041637e-06, + "loss": 0.0249, + "step": 13967 + }, + { + "epoch": 1.5143104943625325, + "grad_norm": 0.35258471965789795, + "learning_rate": 9.940519367474249e-06, + "loss": 0.0049, + "step": 13968 + }, + { + "epoch": 1.5144189071986123, + "grad_norm": 6.166494369506836, + "learning_rate": 9.939793993906863e-06, + "loss": 0.0099, + "step": 13969 + }, + { + "epoch": 1.5145273200346923, + "grad_norm": 0.30146634578704834, + "learning_rate": 9.939068620339477e-06, + "loss": 0.0033, + "step": 13970 + }, + { + "epoch": 1.5146357328707718, + "grad_norm": 0.662224531173706, + "learning_rate": 9.938343246772089e-06, + "loss": 0.0052, + "step": 13971 + }, + { + "epoch": 1.5147441457068518, + "grad_norm": 4.371638298034668, + "learning_rate": 9.937617873204701e-06, + "loss": 0.0312, + "step": 13972 + }, + { + "epoch": 1.5148525585429313, + "grad_norm": 1.477210521697998, + "learning_rate": 9.936892499637313e-06, + "loss": 0.0092, + "step": 13973 + }, + { + "epoch": 1.5149609713790113, + "grad_norm": 4.81694221496582, + "learning_rate": 9.936167126069927e-06, + "loss": 0.0247, + "step": 13974 + }, + { + "epoch": 1.515069384215091, + "grad_norm": 2.3963680267333984, + "learning_rate": 9.93544175250254e-06, + "loss": 0.0358, + "step": 13975 + }, + { + "epoch": 1.5151777970511708, + "grad_norm": 2.2161614894866943, + "learning_rate": 9.934716378935151e-06, + "loss": 0.0149, + "step": 13976 + }, + { + "epoch": 1.5152862098872506, + "grad_norm": 1.7417397499084473, + "learning_rate": 9.933991005367765e-06, + "loss": 0.0104, + "step": 13977 + }, + { + "epoch": 1.5153946227233304, + "grad_norm": 0.6247486472129822, + "learning_rate": 9.933265631800379e-06, + "loss": 0.0065, + "step": 13978 + }, + { + "epoch": 1.5155030355594103, + "grad_norm": 1.4594827890396118, + "learning_rate": 9.932540258232991e-06, + "loss": 0.0066, + "step": 13979 + }, + { + "epoch": 1.51561144839549, + "grad_norm": 0.3852705657482147, + "learning_rate": 9.931814884665603e-06, + "loss": 0.005, + "step": 13980 + }, + { + "epoch": 1.5157198612315699, + "grad_norm": 3.0931615829467773, + "learning_rate": 9.931089511098217e-06, + "loss": 0.0603, + "step": 13981 + }, + { + "epoch": 1.5158282740676496, + "grad_norm": 0.07153080403804779, + "learning_rate": 9.93036413753083e-06, + "loss": 0.0006, + "step": 13982 + }, + { + "epoch": 1.5159366869037294, + "grad_norm": 1.5297037363052368, + "learning_rate": 9.929638763963441e-06, + "loss": 0.0132, + "step": 13983 + }, + { + "epoch": 1.5160450997398092, + "grad_norm": 0.6852720379829407, + "learning_rate": 9.928913390396054e-06, + "loss": 0.0074, + "step": 13984 + }, + { + "epoch": 1.516153512575889, + "grad_norm": 1.1042768955230713, + "learning_rate": 9.928188016828667e-06, + "loss": 0.0029, + "step": 13985 + }, + { + "epoch": 1.516261925411969, + "grad_norm": 1.2084819078445435, + "learning_rate": 9.927462643261281e-06, + "loss": 0.0022, + "step": 13986 + }, + { + "epoch": 1.5163703382480485, + "grad_norm": 3.708897113800049, + "learning_rate": 9.926737269693893e-06, + "loss": 0.0282, + "step": 13987 + }, + { + "epoch": 1.5164787510841284, + "grad_norm": 0.2710830271244049, + "learning_rate": 9.926011896126506e-06, + "loss": 0.0043, + "step": 13988 + }, + { + "epoch": 1.5165871639202082, + "grad_norm": 3.7659289836883545, + "learning_rate": 9.92528652255912e-06, + "loss": 0.0273, + "step": 13989 + }, + { + "epoch": 1.516695576756288, + "grad_norm": 4.9855637550354, + "learning_rate": 9.924561148991731e-06, + "loss": 0.06, + "step": 13990 + }, + { + "epoch": 1.5168039895923677, + "grad_norm": 2.537416696548462, + "learning_rate": 9.923835775424344e-06, + "loss": 0.012, + "step": 13991 + }, + { + "epoch": 1.5169124024284475, + "grad_norm": 0.24756549298763275, + "learning_rate": 9.923110401856957e-06, + "loss": 0.0023, + "step": 13992 + }, + { + "epoch": 1.5170208152645275, + "grad_norm": 0.6794537901878357, + "learning_rate": 9.92238502828957e-06, + "loss": 0.0098, + "step": 13993 + }, + { + "epoch": 1.517129228100607, + "grad_norm": 1.491883397102356, + "learning_rate": 9.921659654722183e-06, + "loss": 0.0456, + "step": 13994 + }, + { + "epoch": 1.517237640936687, + "grad_norm": 1.0168448686599731, + "learning_rate": 9.920934281154796e-06, + "loss": 0.0104, + "step": 13995 + }, + { + "epoch": 1.5173460537727665, + "grad_norm": 1.578659176826477, + "learning_rate": 9.92020890758741e-06, + "loss": 0.0145, + "step": 13996 + }, + { + "epoch": 1.5174544666088465, + "grad_norm": 4.149772644042969, + "learning_rate": 9.919483534020022e-06, + "loss": 0.0208, + "step": 13997 + }, + { + "epoch": 1.5175628794449263, + "grad_norm": 1.2004985809326172, + "learning_rate": 9.918758160452634e-06, + "loss": 0.0362, + "step": 13998 + }, + { + "epoch": 1.517671292281006, + "grad_norm": 1.3008053302764893, + "learning_rate": 9.918032786885246e-06, + "loss": 0.0264, + "step": 13999 + }, + { + "epoch": 1.5177797051170858, + "grad_norm": 2.990203380584717, + "learning_rate": 9.91730741331786e-06, + "loss": 0.0543, + "step": 14000 + }, + { + "epoch": 1.5178881179531656, + "grad_norm": 2.6498444080352783, + "learning_rate": 9.916582039750472e-06, + "loss": 0.0124, + "step": 14001 + }, + { + "epoch": 1.5179965307892456, + "grad_norm": 1.5325249433517456, + "learning_rate": 9.915856666183084e-06, + "loss": 0.0134, + "step": 14002 + }, + { + "epoch": 1.518104943625325, + "grad_norm": 2.091153383255005, + "learning_rate": 9.915131292615698e-06, + "loss": 0.0404, + "step": 14003 + }, + { + "epoch": 1.518213356461405, + "grad_norm": 4.767205715179443, + "learning_rate": 9.914405919048312e-06, + "loss": 0.0364, + "step": 14004 + }, + { + "epoch": 1.5183217692974849, + "grad_norm": 2.7364659309387207, + "learning_rate": 9.913680545480924e-06, + "loss": 0.0211, + "step": 14005 + }, + { + "epoch": 1.5184301821335646, + "grad_norm": 6.178999900817871, + "learning_rate": 9.912955171913536e-06, + "loss": 0.0633, + "step": 14006 + }, + { + "epoch": 1.5185385949696444, + "grad_norm": 6.516798496246338, + "learning_rate": 9.91222979834615e-06, + "loss": 0.0252, + "step": 14007 + }, + { + "epoch": 1.5186470078057241, + "grad_norm": 2.6374130249023438, + "learning_rate": 9.911504424778762e-06, + "loss": 0.043, + "step": 14008 + }, + { + "epoch": 1.5187554206418041, + "grad_norm": 2.5271637439727783, + "learning_rate": 9.910779051211374e-06, + "loss": 0.0289, + "step": 14009 + }, + { + "epoch": 1.5188638334778837, + "grad_norm": 2.6208415031433105, + "learning_rate": 9.910053677643988e-06, + "loss": 0.0121, + "step": 14010 + }, + { + "epoch": 1.5189722463139637, + "grad_norm": 2.5116279125213623, + "learning_rate": 9.9093283040766e-06, + "loss": 0.0487, + "step": 14011 + }, + { + "epoch": 1.5190806591500432, + "grad_norm": 0.24711686372756958, + "learning_rate": 9.908602930509214e-06, + "loss": 0.0023, + "step": 14012 + }, + { + "epoch": 1.5191890719861232, + "grad_norm": 0.3968289792537689, + "learning_rate": 9.907877556941826e-06, + "loss": 0.0033, + "step": 14013 + }, + { + "epoch": 1.519297484822203, + "grad_norm": 1.9720889329910278, + "learning_rate": 9.907152183374438e-06, + "loss": 0.0294, + "step": 14014 + }, + { + "epoch": 1.5194058976582827, + "grad_norm": 2.3841540813446045, + "learning_rate": 9.906426809807052e-06, + "loss": 0.023, + "step": 14015 + }, + { + "epoch": 1.5195143104943627, + "grad_norm": 2.9905781745910645, + "learning_rate": 9.905701436239664e-06, + "loss": 0.0259, + "step": 14016 + }, + { + "epoch": 1.5196227233304422, + "grad_norm": 1.228893756866455, + "learning_rate": 9.904976062672276e-06, + "loss": 0.0184, + "step": 14017 + }, + { + "epoch": 1.5197311361665222, + "grad_norm": 1.058542013168335, + "learning_rate": 9.90425068910489e-06, + "loss": 0.0074, + "step": 14018 + }, + { + "epoch": 1.5198395490026018, + "grad_norm": 0.4617595374584198, + "learning_rate": 9.903525315537502e-06, + "loss": 0.0045, + "step": 14019 + }, + { + "epoch": 1.5199479618386817, + "grad_norm": 1.244447112083435, + "learning_rate": 9.902799941970116e-06, + "loss": 0.0101, + "step": 14020 + }, + { + "epoch": 1.5200563746747615, + "grad_norm": 2.1469993591308594, + "learning_rate": 9.902074568402728e-06, + "loss": 0.0148, + "step": 14021 + }, + { + "epoch": 1.5201647875108413, + "grad_norm": 2.5933749675750732, + "learning_rate": 9.901349194835342e-06, + "loss": 0.0213, + "step": 14022 + }, + { + "epoch": 1.520273200346921, + "grad_norm": 0.8985424637794495, + "learning_rate": 9.900623821267954e-06, + "loss": 0.0242, + "step": 14023 + }, + { + "epoch": 1.5203816131830008, + "grad_norm": 0.7901719212532043, + "learning_rate": 9.899898447700566e-06, + "loss": 0.0094, + "step": 14024 + }, + { + "epoch": 1.5204900260190808, + "grad_norm": 1.369860053062439, + "learning_rate": 9.899173074133179e-06, + "loss": 0.0057, + "step": 14025 + }, + { + "epoch": 1.5205984388551603, + "grad_norm": 3.505465269088745, + "learning_rate": 9.898447700565792e-06, + "loss": 0.0364, + "step": 14026 + }, + { + "epoch": 1.5207068516912403, + "grad_norm": 1.3722079992294312, + "learning_rate": 9.897722326998405e-06, + "loss": 0.0252, + "step": 14027 + }, + { + "epoch": 1.52081526452732, + "grad_norm": 0.9975358843803406, + "learning_rate": 9.896996953431017e-06, + "loss": 0.0077, + "step": 14028 + }, + { + "epoch": 1.5209236773633998, + "grad_norm": 1.87330961227417, + "learning_rate": 9.89627157986363e-06, + "loss": 0.01, + "step": 14029 + }, + { + "epoch": 1.5210320901994796, + "grad_norm": 4.131607532501221, + "learning_rate": 9.895546206296244e-06, + "loss": 0.0091, + "step": 14030 + }, + { + "epoch": 1.5211405030355594, + "grad_norm": 3.119283437728882, + "learning_rate": 9.894820832728856e-06, + "loss": 0.0286, + "step": 14031 + }, + { + "epoch": 1.5212489158716394, + "grad_norm": 1.5668996572494507, + "learning_rate": 9.894095459161469e-06, + "loss": 0.0193, + "step": 14032 + }, + { + "epoch": 1.521357328707719, + "grad_norm": 3.297116279602051, + "learning_rate": 9.893370085594082e-06, + "loss": 0.0414, + "step": 14033 + }, + { + "epoch": 1.5214657415437989, + "grad_norm": 3.5463411808013916, + "learning_rate": 9.892644712026695e-06, + "loss": 0.0198, + "step": 14034 + }, + { + "epoch": 1.5215741543798784, + "grad_norm": 3.7682082653045654, + "learning_rate": 9.891919338459307e-06, + "loss": 0.011, + "step": 14035 + }, + { + "epoch": 1.5216825672159584, + "grad_norm": 0.9827446341514587, + "learning_rate": 9.89119396489192e-06, + "loss": 0.0084, + "step": 14036 + }, + { + "epoch": 1.5217909800520382, + "grad_norm": 1.452207088470459, + "learning_rate": 9.890468591324533e-06, + "loss": 0.0171, + "step": 14037 + }, + { + "epoch": 1.521899392888118, + "grad_norm": 1.2627846002578735, + "learning_rate": 9.889743217757147e-06, + "loss": 0.0489, + "step": 14038 + }, + { + "epoch": 1.5220078057241977, + "grad_norm": 0.10594231635332108, + "learning_rate": 9.889017844189759e-06, + "loss": 0.0016, + "step": 14039 + }, + { + "epoch": 1.5221162185602775, + "grad_norm": 0.09566507488489151, + "learning_rate": 9.888292470622371e-06, + "loss": 0.0015, + "step": 14040 + }, + { + "epoch": 1.5222246313963574, + "grad_norm": 2.4390201568603516, + "learning_rate": 9.887567097054985e-06, + "loss": 0.0217, + "step": 14041 + }, + { + "epoch": 1.522333044232437, + "grad_norm": 2.0587711334228516, + "learning_rate": 9.886841723487597e-06, + "loss": 0.0213, + "step": 14042 + }, + { + "epoch": 1.522441457068517, + "grad_norm": 3.4868063926696777, + "learning_rate": 9.886116349920209e-06, + "loss": 0.0086, + "step": 14043 + }, + { + "epoch": 1.5225498699045967, + "grad_norm": 2.5139353275299072, + "learning_rate": 9.885390976352823e-06, + "loss": 0.0143, + "step": 14044 + }, + { + "epoch": 1.5226582827406765, + "grad_norm": 1.6745270490646362, + "learning_rate": 9.884665602785435e-06, + "loss": 0.009, + "step": 14045 + }, + { + "epoch": 1.5227666955767563, + "grad_norm": 0.43865591287612915, + "learning_rate": 9.883940229218049e-06, + "loss": 0.0016, + "step": 14046 + }, + { + "epoch": 1.522875108412836, + "grad_norm": 0.7417949438095093, + "learning_rate": 9.883214855650661e-06, + "loss": 0.0023, + "step": 14047 + }, + { + "epoch": 1.522983521248916, + "grad_norm": 3.995868444442749, + "learning_rate": 9.882489482083275e-06, + "loss": 0.034, + "step": 14048 + }, + { + "epoch": 1.5230919340849955, + "grad_norm": 7.417863368988037, + "learning_rate": 9.881764108515887e-06, + "loss": 0.0222, + "step": 14049 + }, + { + "epoch": 1.5232003469210755, + "grad_norm": 0.6727468967437744, + "learning_rate": 9.881038734948499e-06, + "loss": 0.0025, + "step": 14050 + }, + { + "epoch": 1.5233087597571553, + "grad_norm": 2.4592537879943848, + "learning_rate": 9.880313361381111e-06, + "loss": 0.0047, + "step": 14051 + }, + { + "epoch": 1.523417172593235, + "grad_norm": 1.3505412340164185, + "learning_rate": 9.879587987813725e-06, + "loss": 0.0411, + "step": 14052 + }, + { + "epoch": 1.5235255854293148, + "grad_norm": 0.57817143201828, + "learning_rate": 9.878862614246337e-06, + "loss": 0.006, + "step": 14053 + }, + { + "epoch": 1.5236339982653946, + "grad_norm": 2.539484977722168, + "learning_rate": 9.87813724067895e-06, + "loss": 0.0414, + "step": 14054 + }, + { + "epoch": 1.5237424111014746, + "grad_norm": 2.061537981033325, + "learning_rate": 9.877411867111563e-06, + "loss": 0.0123, + "step": 14055 + }, + { + "epoch": 1.523850823937554, + "grad_norm": 0.644153892993927, + "learning_rate": 9.876686493544177e-06, + "loss": 0.0037, + "step": 14056 + }, + { + "epoch": 1.523959236773634, + "grad_norm": 0.604145348072052, + "learning_rate": 9.875961119976789e-06, + "loss": 0.0096, + "step": 14057 + }, + { + "epoch": 1.5240676496097136, + "grad_norm": 0.2902953326702118, + "learning_rate": 9.875235746409401e-06, + "loss": 0.0024, + "step": 14058 + }, + { + "epoch": 1.5241760624457936, + "grad_norm": 0.45382580161094666, + "learning_rate": 9.874510372842015e-06, + "loss": 0.0021, + "step": 14059 + }, + { + "epoch": 1.5242844752818734, + "grad_norm": 2.6542773246765137, + "learning_rate": 9.873784999274627e-06, + "loss": 0.026, + "step": 14060 + }, + { + "epoch": 1.5243928881179531, + "grad_norm": 0.1622001826763153, + "learning_rate": 9.87305962570724e-06, + "loss": 0.0018, + "step": 14061 + }, + { + "epoch": 1.524501300954033, + "grad_norm": 1.3665422201156616, + "learning_rate": 9.872334252139853e-06, + "loss": 0.0093, + "step": 14062 + }, + { + "epoch": 1.5246097137901127, + "grad_norm": 6.486844062805176, + "learning_rate": 9.871608878572465e-06, + "loss": 0.0123, + "step": 14063 + }, + { + "epoch": 1.5247181266261927, + "grad_norm": 3.174276113510132, + "learning_rate": 9.87088350500508e-06, + "loss": 0.0121, + "step": 14064 + }, + { + "epoch": 1.5248265394622722, + "grad_norm": 2.887566566467285, + "learning_rate": 9.870158131437691e-06, + "loss": 0.0461, + "step": 14065 + }, + { + "epoch": 1.5249349522983522, + "grad_norm": 4.323582172393799, + "learning_rate": 9.869432757870304e-06, + "loss": 0.0532, + "step": 14066 + }, + { + "epoch": 1.525043365134432, + "grad_norm": 0.5923543572425842, + "learning_rate": 9.868707384302917e-06, + "loss": 0.0019, + "step": 14067 + }, + { + "epoch": 1.5251517779705117, + "grad_norm": 0.5888389348983765, + "learning_rate": 9.86798201073553e-06, + "loss": 0.0042, + "step": 14068 + }, + { + "epoch": 1.5252601908065915, + "grad_norm": 2.003638744354248, + "learning_rate": 9.867256637168142e-06, + "loss": 0.0111, + "step": 14069 + }, + { + "epoch": 1.5253686036426712, + "grad_norm": 1.7433897256851196, + "learning_rate": 9.866531263600755e-06, + "loss": 0.0091, + "step": 14070 + }, + { + "epoch": 1.5254770164787512, + "grad_norm": 6.997538089752197, + "learning_rate": 9.865805890033368e-06, + "loss": 0.0102, + "step": 14071 + }, + { + "epoch": 1.5255854293148308, + "grad_norm": 6.695999622344971, + "learning_rate": 9.865080516465981e-06, + "loss": 0.0266, + "step": 14072 + }, + { + "epoch": 1.5256938421509108, + "grad_norm": 1.8321563005447388, + "learning_rate": 9.864355142898594e-06, + "loss": 0.0128, + "step": 14073 + }, + { + "epoch": 1.5258022549869905, + "grad_norm": 13.93993854522705, + "learning_rate": 9.863629769331207e-06, + "loss": 0.028, + "step": 14074 + }, + { + "epoch": 1.5259106678230703, + "grad_norm": 6.68711519241333, + "learning_rate": 9.86290439576382e-06, + "loss": 0.0367, + "step": 14075 + }, + { + "epoch": 1.52601908065915, + "grad_norm": 5.311349391937256, + "learning_rate": 9.862179022196432e-06, + "loss": 0.0105, + "step": 14076 + }, + { + "epoch": 1.5261274934952298, + "grad_norm": 3.6914403438568115, + "learning_rate": 9.861453648629044e-06, + "loss": 0.0354, + "step": 14077 + }, + { + "epoch": 1.5262359063313098, + "grad_norm": 6.145002365112305, + "learning_rate": 9.860728275061658e-06, + "loss": 0.0421, + "step": 14078 + }, + { + "epoch": 1.5263443191673893, + "grad_norm": 1.9433177709579468, + "learning_rate": 9.86000290149427e-06, + "loss": 0.0185, + "step": 14079 + }, + { + "epoch": 1.5264527320034693, + "grad_norm": 4.6939473152160645, + "learning_rate": 9.859277527926882e-06, + "loss": 0.01, + "step": 14080 + }, + { + "epoch": 1.5265611448395489, + "grad_norm": 2.1065399646759033, + "learning_rate": 9.858552154359496e-06, + "loss": 0.002, + "step": 14081 + }, + { + "epoch": 1.5266695576756288, + "grad_norm": 4.450342655181885, + "learning_rate": 9.85782678079211e-06, + "loss": 0.0341, + "step": 14082 + }, + { + "epoch": 1.5267779705117086, + "grad_norm": 1.9643434286117554, + "learning_rate": 9.857101407224722e-06, + "loss": 0.0102, + "step": 14083 + }, + { + "epoch": 1.5268863833477884, + "grad_norm": 2.1544015407562256, + "learning_rate": 9.856376033657334e-06, + "loss": 0.0082, + "step": 14084 + }, + { + "epoch": 1.5269947961838681, + "grad_norm": 1.3403794765472412, + "learning_rate": 9.855650660089948e-06, + "loss": 0.0086, + "step": 14085 + }, + { + "epoch": 1.527103209019948, + "grad_norm": 6.869955062866211, + "learning_rate": 9.85492528652256e-06, + "loss": 0.0355, + "step": 14086 + }, + { + "epoch": 1.5272116218560279, + "grad_norm": 0.4606999456882477, + "learning_rate": 9.854199912955172e-06, + "loss": 0.0023, + "step": 14087 + }, + { + "epoch": 1.5273200346921074, + "grad_norm": 1.332453966140747, + "learning_rate": 9.853474539387786e-06, + "loss": 0.0173, + "step": 14088 + }, + { + "epoch": 1.5274284475281874, + "grad_norm": 1.6449025869369507, + "learning_rate": 9.852749165820398e-06, + "loss": 0.0346, + "step": 14089 + }, + { + "epoch": 1.5275368603642672, + "grad_norm": 4.633398056030273, + "learning_rate": 9.852023792253012e-06, + "loss": 0.0154, + "step": 14090 + }, + { + "epoch": 1.527645273200347, + "grad_norm": 4.902040481567383, + "learning_rate": 9.851298418685624e-06, + "loss": 0.0982, + "step": 14091 + }, + { + "epoch": 1.5277536860364267, + "grad_norm": 9.909154891967773, + "learning_rate": 9.850573045118236e-06, + "loss": 0.0363, + "step": 14092 + }, + { + "epoch": 1.5278620988725065, + "grad_norm": 0.13486796617507935, + "learning_rate": 9.84984767155085e-06, + "loss": 0.0014, + "step": 14093 + }, + { + "epoch": 1.5279705117085864, + "grad_norm": 0.22946980595588684, + "learning_rate": 9.849122297983462e-06, + "loss": 0.002, + "step": 14094 + }, + { + "epoch": 1.528078924544666, + "grad_norm": 3.586630344390869, + "learning_rate": 9.848396924416074e-06, + "loss": 0.0546, + "step": 14095 + }, + { + "epoch": 1.528187337380746, + "grad_norm": 3.01174259185791, + "learning_rate": 9.847671550848688e-06, + "loss": 0.0328, + "step": 14096 + }, + { + "epoch": 1.5282957502168255, + "grad_norm": 6.25925350189209, + "learning_rate": 9.8469461772813e-06, + "loss": 0.0916, + "step": 14097 + }, + { + "epoch": 1.5284041630529055, + "grad_norm": 1.6466819047927856, + "learning_rate": 9.846220803713914e-06, + "loss": 0.0095, + "step": 14098 + }, + { + "epoch": 1.5285125758889853, + "grad_norm": 1.4652173519134521, + "learning_rate": 9.845495430146526e-06, + "loss": 0.0431, + "step": 14099 + }, + { + "epoch": 1.528620988725065, + "grad_norm": 10.617856979370117, + "learning_rate": 9.84477005657914e-06, + "loss": 0.0296, + "step": 14100 + }, + { + "epoch": 1.5287294015611448, + "grad_norm": 4.748671054840088, + "learning_rate": 9.844044683011752e-06, + "loss": 0.0156, + "step": 14101 + }, + { + "epoch": 1.5288378143972245, + "grad_norm": 0.36598527431488037, + "learning_rate": 9.843319309444364e-06, + "loss": 0.0014, + "step": 14102 + }, + { + "epoch": 1.5289462272333045, + "grad_norm": 1.956302523612976, + "learning_rate": 9.842593935876977e-06, + "loss": 0.0215, + "step": 14103 + }, + { + "epoch": 1.529054640069384, + "grad_norm": 8.412330627441406, + "learning_rate": 9.84186856230959e-06, + "loss": 0.0356, + "step": 14104 + }, + { + "epoch": 1.529163052905464, + "grad_norm": 8.088939666748047, + "learning_rate": 9.841143188742203e-06, + "loss": 0.0281, + "step": 14105 + }, + { + "epoch": 1.5292714657415438, + "grad_norm": 2.7289228439331055, + "learning_rate": 9.840417815174815e-06, + "loss": 0.0583, + "step": 14106 + }, + { + "epoch": 1.5293798785776236, + "grad_norm": 1.9011589288711548, + "learning_rate": 9.839692441607429e-06, + "loss": 0.013, + "step": 14107 + }, + { + "epoch": 1.5294882914137033, + "grad_norm": 0.8603560328483582, + "learning_rate": 9.838967068040042e-06, + "loss": 0.0069, + "step": 14108 + }, + { + "epoch": 1.5295967042497831, + "grad_norm": 1.681671380996704, + "learning_rate": 9.838241694472654e-06, + "loss": 0.0085, + "step": 14109 + }, + { + "epoch": 1.529705117085863, + "grad_norm": 1.458336353302002, + "learning_rate": 9.837516320905267e-06, + "loss": 0.0267, + "step": 14110 + }, + { + "epoch": 1.5298135299219426, + "grad_norm": 2.526883363723755, + "learning_rate": 9.83679094733788e-06, + "loss": 0.0141, + "step": 14111 + }, + { + "epoch": 1.5299219427580226, + "grad_norm": 1.4448267221450806, + "learning_rate": 9.836065573770493e-06, + "loss": 0.0046, + "step": 14112 + }, + { + "epoch": 1.5300303555941024, + "grad_norm": 2.4647490978240967, + "learning_rate": 9.835340200203105e-06, + "loss": 0.0416, + "step": 14113 + }, + { + "epoch": 1.5301387684301822, + "grad_norm": 2.6959831714630127, + "learning_rate": 9.834614826635719e-06, + "loss": 0.0509, + "step": 14114 + }, + { + "epoch": 1.530247181266262, + "grad_norm": 0.07427176088094711, + "learning_rate": 9.83388945306833e-06, + "loss": 0.0008, + "step": 14115 + }, + { + "epoch": 1.5303555941023417, + "grad_norm": 0.09362878650426865, + "learning_rate": 9.833164079500945e-06, + "loss": 0.001, + "step": 14116 + }, + { + "epoch": 1.5304640069384217, + "grad_norm": 2.0964131355285645, + "learning_rate": 9.832438705933557e-06, + "loss": 0.0427, + "step": 14117 + }, + { + "epoch": 1.5305724197745012, + "grad_norm": 0.0970926359295845, + "learning_rate": 9.831713332366169e-06, + "loss": 0.0009, + "step": 14118 + }, + { + "epoch": 1.5306808326105812, + "grad_norm": 5.867969512939453, + "learning_rate": 9.830987958798783e-06, + "loss": 0.044, + "step": 14119 + }, + { + "epoch": 1.5307892454466607, + "grad_norm": 3.1638851165771484, + "learning_rate": 9.830262585231395e-06, + "loss": 0.0123, + "step": 14120 + }, + { + "epoch": 1.5308976582827407, + "grad_norm": 3.618678092956543, + "learning_rate": 9.829537211664007e-06, + "loss": 0.083, + "step": 14121 + }, + { + "epoch": 1.5310060711188205, + "grad_norm": 5.672842979431152, + "learning_rate": 9.82881183809662e-06, + "loss": 0.0488, + "step": 14122 + }, + { + "epoch": 1.5311144839549002, + "grad_norm": 1.1107620000839233, + "learning_rate": 9.828086464529233e-06, + "loss": 0.0132, + "step": 14123 + }, + { + "epoch": 1.53122289679098, + "grad_norm": 2.217930793762207, + "learning_rate": 9.827361090961847e-06, + "loss": 0.0355, + "step": 14124 + }, + { + "epoch": 1.5313313096270598, + "grad_norm": 2.8035829067230225, + "learning_rate": 9.826635717394459e-06, + "loss": 0.0334, + "step": 14125 + }, + { + "epoch": 1.5314397224631398, + "grad_norm": 3.563169240951538, + "learning_rate": 9.825910343827073e-06, + "loss": 0.0623, + "step": 14126 + }, + { + "epoch": 1.5315481352992193, + "grad_norm": 3.2273337841033936, + "learning_rate": 9.825184970259685e-06, + "loss": 0.0216, + "step": 14127 + }, + { + "epoch": 1.5316565481352993, + "grad_norm": 2.2758262157440186, + "learning_rate": 9.824459596692297e-06, + "loss": 0.0231, + "step": 14128 + }, + { + "epoch": 1.531764960971379, + "grad_norm": 6.866157531738281, + "learning_rate": 9.82373422312491e-06, + "loss": 0.0508, + "step": 14129 + }, + { + "epoch": 1.5318733738074588, + "grad_norm": 1.0050873756408691, + "learning_rate": 9.823008849557523e-06, + "loss": 0.0087, + "step": 14130 + }, + { + "epoch": 1.5319817866435386, + "grad_norm": 1.7223819494247437, + "learning_rate": 9.822283475990135e-06, + "loss": 0.0132, + "step": 14131 + }, + { + "epoch": 1.5320901994796183, + "grad_norm": 2.613110303878784, + "learning_rate": 9.821558102422747e-06, + "loss": 0.0138, + "step": 14132 + }, + { + "epoch": 1.5321986123156983, + "grad_norm": 2.715790033340454, + "learning_rate": 9.820832728855361e-06, + "loss": 0.0202, + "step": 14133 + }, + { + "epoch": 1.5323070251517779, + "grad_norm": 1.6310425996780396, + "learning_rate": 9.820107355287975e-06, + "loss": 0.0412, + "step": 14134 + }, + { + "epoch": 1.5324154379878578, + "grad_norm": 2.8094186782836914, + "learning_rate": 9.819381981720587e-06, + "loss": 0.0249, + "step": 14135 + }, + { + "epoch": 1.5325238508239376, + "grad_norm": 3.3099684715270996, + "learning_rate": 9.8186566081532e-06, + "loss": 0.0204, + "step": 14136 + }, + { + "epoch": 1.5326322636600174, + "grad_norm": 2.727898120880127, + "learning_rate": 9.817931234585813e-06, + "loss": 0.057, + "step": 14137 + }, + { + "epoch": 1.5327406764960971, + "grad_norm": 2.56968092918396, + "learning_rate": 9.817205861018425e-06, + "loss": 0.0373, + "step": 14138 + }, + { + "epoch": 1.532849089332177, + "grad_norm": 1.8970388174057007, + "learning_rate": 9.816480487451037e-06, + "loss": 0.0496, + "step": 14139 + }, + { + "epoch": 1.5329575021682569, + "grad_norm": 0.26610302925109863, + "learning_rate": 9.815755113883651e-06, + "loss": 0.0031, + "step": 14140 + }, + { + "epoch": 1.5330659150043364, + "grad_norm": 3.9455409049987793, + "learning_rate": 9.815029740316263e-06, + "loss": 0.0527, + "step": 14141 + }, + { + "epoch": 1.5331743278404164, + "grad_norm": 5.015367031097412, + "learning_rate": 9.814304366748877e-06, + "loss": 0.0523, + "step": 14142 + }, + { + "epoch": 1.533282740676496, + "grad_norm": 0.3049769997596741, + "learning_rate": 9.81357899318149e-06, + "loss": 0.0024, + "step": 14143 + }, + { + "epoch": 1.533391153512576, + "grad_norm": 4.742360591888428, + "learning_rate": 9.812853619614102e-06, + "loss": 0.0634, + "step": 14144 + }, + { + "epoch": 1.5334995663486557, + "grad_norm": 0.6612123250961304, + "learning_rate": 9.812128246046715e-06, + "loss": 0.0067, + "step": 14145 + }, + { + "epoch": 1.5336079791847355, + "grad_norm": 3.465564250946045, + "learning_rate": 9.811402872479328e-06, + "loss": 0.0129, + "step": 14146 + }, + { + "epoch": 1.5337163920208152, + "grad_norm": 0.7368857860565186, + "learning_rate": 9.81067749891194e-06, + "loss": 0.0116, + "step": 14147 + }, + { + "epoch": 1.533824804856895, + "grad_norm": 2.8608593940734863, + "learning_rate": 9.809952125344553e-06, + "loss": 0.0269, + "step": 14148 + }, + { + "epoch": 1.533933217692975, + "grad_norm": 1.2601054906845093, + "learning_rate": 9.809226751777166e-06, + "loss": 0.0051, + "step": 14149 + }, + { + "epoch": 1.5340416305290545, + "grad_norm": 1.535739541053772, + "learning_rate": 9.80850137820978e-06, + "loss": 0.0118, + "step": 14150 + }, + { + "epoch": 1.5341500433651345, + "grad_norm": 2.957486867904663, + "learning_rate": 9.807776004642392e-06, + "loss": 0.0479, + "step": 14151 + }, + { + "epoch": 1.5342584562012143, + "grad_norm": 1.9060684442520142, + "learning_rate": 9.807050631075005e-06, + "loss": 0.0385, + "step": 14152 + }, + { + "epoch": 1.534366869037294, + "grad_norm": 1.3735133409500122, + "learning_rate": 9.806325257507618e-06, + "loss": 0.0159, + "step": 14153 + }, + { + "epoch": 1.5344752818733738, + "grad_norm": 3.0842623710632324, + "learning_rate": 9.80559988394023e-06, + "loss": 0.0088, + "step": 14154 + }, + { + "epoch": 1.5345836947094535, + "grad_norm": 1.0807602405548096, + "learning_rate": 9.804874510372842e-06, + "loss": 0.0309, + "step": 14155 + }, + { + "epoch": 1.5346921075455335, + "grad_norm": 5.627696990966797, + "learning_rate": 9.804149136805456e-06, + "loss": 0.0587, + "step": 14156 + }, + { + "epoch": 1.534800520381613, + "grad_norm": 2.835247755050659, + "learning_rate": 9.803423763238068e-06, + "loss": 0.0202, + "step": 14157 + }, + { + "epoch": 1.534908933217693, + "grad_norm": 2.1559841632843018, + "learning_rate": 9.80269838967068e-06, + "loss": 0.0068, + "step": 14158 + }, + { + "epoch": 1.5350173460537726, + "grad_norm": 0.07350517064332962, + "learning_rate": 9.801973016103294e-06, + "loss": 0.0009, + "step": 14159 + }, + { + "epoch": 1.5351257588898526, + "grad_norm": 2.6768057346343994, + "learning_rate": 9.801247642535908e-06, + "loss": 0.0436, + "step": 14160 + }, + { + "epoch": 1.5352341717259324, + "grad_norm": 2.8855221271514893, + "learning_rate": 9.80052226896852e-06, + "loss": 0.0236, + "step": 14161 + }, + { + "epoch": 1.5353425845620121, + "grad_norm": 0.543907642364502, + "learning_rate": 9.799796895401132e-06, + "loss": 0.0101, + "step": 14162 + }, + { + "epoch": 1.535450997398092, + "grad_norm": 0.994752049446106, + "learning_rate": 9.799071521833746e-06, + "loss": 0.0117, + "step": 14163 + }, + { + "epoch": 1.5355594102341716, + "grad_norm": 1.5199084281921387, + "learning_rate": 9.798346148266358e-06, + "loss": 0.0322, + "step": 14164 + }, + { + "epoch": 1.5356678230702516, + "grad_norm": 1.452463984489441, + "learning_rate": 9.79762077469897e-06, + "loss": 0.0064, + "step": 14165 + }, + { + "epoch": 1.5357762359063312, + "grad_norm": 1.0409315824508667, + "learning_rate": 9.796895401131584e-06, + "loss": 0.0134, + "step": 14166 + }, + { + "epoch": 1.5358846487424112, + "grad_norm": 0.26484212279319763, + "learning_rate": 9.796170027564196e-06, + "loss": 0.0028, + "step": 14167 + }, + { + "epoch": 1.535993061578491, + "grad_norm": 2.5725769996643066, + "learning_rate": 9.79544465399681e-06, + "loss": 0.0212, + "step": 14168 + }, + { + "epoch": 1.5361014744145707, + "grad_norm": 2.228388786315918, + "learning_rate": 9.794719280429422e-06, + "loss": 0.0188, + "step": 14169 + }, + { + "epoch": 1.5362098872506504, + "grad_norm": 0.3401258885860443, + "learning_rate": 9.793993906862034e-06, + "loss": 0.0074, + "step": 14170 + }, + { + "epoch": 1.5363183000867302, + "grad_norm": 1.2853251695632935, + "learning_rate": 9.793268533294648e-06, + "loss": 0.0326, + "step": 14171 + }, + { + "epoch": 1.5364267129228102, + "grad_norm": 0.7428716421127319, + "learning_rate": 9.79254315972726e-06, + "loss": 0.024, + "step": 14172 + }, + { + "epoch": 1.5365351257588897, + "grad_norm": 0.7077968716621399, + "learning_rate": 9.791817786159872e-06, + "loss": 0.0079, + "step": 14173 + }, + { + "epoch": 1.5366435385949697, + "grad_norm": 0.5773529410362244, + "learning_rate": 9.791092412592486e-06, + "loss": 0.0047, + "step": 14174 + }, + { + "epoch": 1.5367519514310495, + "grad_norm": 2.0607032775878906, + "learning_rate": 9.790367039025098e-06, + "loss": 0.0117, + "step": 14175 + }, + { + "epoch": 1.5368603642671292, + "grad_norm": 1.3286405801773071, + "learning_rate": 9.789641665457712e-06, + "loss": 0.0188, + "step": 14176 + }, + { + "epoch": 1.536968777103209, + "grad_norm": 1.261918067932129, + "learning_rate": 9.788916291890324e-06, + "loss": 0.0148, + "step": 14177 + }, + { + "epoch": 1.5370771899392888, + "grad_norm": 3.155747890472412, + "learning_rate": 9.788190918322938e-06, + "loss": 0.0286, + "step": 14178 + }, + { + "epoch": 1.5371856027753688, + "grad_norm": 0.7608159780502319, + "learning_rate": 9.78746554475555e-06, + "loss": 0.0112, + "step": 14179 + }, + { + "epoch": 1.5372940156114483, + "grad_norm": 2.5939717292785645, + "learning_rate": 9.786740171188162e-06, + "loss": 0.0194, + "step": 14180 + }, + { + "epoch": 1.5374024284475283, + "grad_norm": 1.8722691535949707, + "learning_rate": 9.786014797620775e-06, + "loss": 0.0589, + "step": 14181 + }, + { + "epoch": 1.5375108412836078, + "grad_norm": 0.31554535031318665, + "learning_rate": 9.785289424053388e-06, + "loss": 0.0039, + "step": 14182 + }, + { + "epoch": 1.5376192541196878, + "grad_norm": 3.970388174057007, + "learning_rate": 9.784564050486e-06, + "loss": 0.0145, + "step": 14183 + }, + { + "epoch": 1.5377276669557676, + "grad_norm": 2.9841527938842773, + "learning_rate": 9.783838676918613e-06, + "loss": 0.0133, + "step": 14184 + }, + { + "epoch": 1.5378360797918473, + "grad_norm": 2.0445196628570557, + "learning_rate": 9.783113303351227e-06, + "loss": 0.0253, + "step": 14185 + }, + { + "epoch": 1.537944492627927, + "grad_norm": 0.5085863471031189, + "learning_rate": 9.78238792978384e-06, + "loss": 0.0063, + "step": 14186 + }, + { + "epoch": 1.5380529054640069, + "grad_norm": 3.9838883876800537, + "learning_rate": 9.781662556216452e-06, + "loss": 0.0084, + "step": 14187 + }, + { + "epoch": 1.5381613183000868, + "grad_norm": 2.0939016342163086, + "learning_rate": 9.780937182649065e-06, + "loss": 0.0341, + "step": 14188 + }, + { + "epoch": 1.5382697311361664, + "grad_norm": 2.084411859512329, + "learning_rate": 9.780211809081678e-06, + "loss": 0.0127, + "step": 14189 + }, + { + "epoch": 1.5383781439722464, + "grad_norm": 0.9427832365036011, + "learning_rate": 9.77948643551429e-06, + "loss": 0.0074, + "step": 14190 + }, + { + "epoch": 1.5384865568083261, + "grad_norm": 6.8647141456604, + "learning_rate": 9.778761061946903e-06, + "loss": 0.026, + "step": 14191 + }, + { + "epoch": 1.538594969644406, + "grad_norm": 2.72749662399292, + "learning_rate": 9.778035688379517e-06, + "loss": 0.0103, + "step": 14192 + }, + { + "epoch": 1.5387033824804857, + "grad_norm": 0.7859935164451599, + "learning_rate": 9.777310314812129e-06, + "loss": 0.0143, + "step": 14193 + }, + { + "epoch": 1.5388117953165654, + "grad_norm": 1.5164083242416382, + "learning_rate": 9.776584941244743e-06, + "loss": 0.0072, + "step": 14194 + }, + { + "epoch": 1.5389202081526454, + "grad_norm": 2.165003538131714, + "learning_rate": 9.775859567677355e-06, + "loss": 0.043, + "step": 14195 + }, + { + "epoch": 1.539028620988725, + "grad_norm": 2.038480520248413, + "learning_rate": 9.775134194109967e-06, + "loss": 0.0305, + "step": 14196 + }, + { + "epoch": 1.539137033824805, + "grad_norm": 2.104343891143799, + "learning_rate": 9.77440882054258e-06, + "loss": 0.0101, + "step": 14197 + }, + { + "epoch": 1.5392454466608847, + "grad_norm": 4.874423503875732, + "learning_rate": 9.773683446975193e-06, + "loss": 0.0332, + "step": 14198 + }, + { + "epoch": 1.5393538594969645, + "grad_norm": 3.241950035095215, + "learning_rate": 9.772958073407805e-06, + "loss": 0.0224, + "step": 14199 + }, + { + "epoch": 1.5394622723330442, + "grad_norm": 3.8982656002044678, + "learning_rate": 9.772232699840419e-06, + "loss": 0.0342, + "step": 14200 + }, + { + "epoch": 1.539570685169124, + "grad_norm": 1.0014952421188354, + "learning_rate": 9.771507326273031e-06, + "loss": 0.0096, + "step": 14201 + }, + { + "epoch": 1.539679098005204, + "grad_norm": 1.5035127401351929, + "learning_rate": 9.770781952705645e-06, + "loss": 0.0031, + "step": 14202 + }, + { + "epoch": 1.5397875108412835, + "grad_norm": 1.4644412994384766, + "learning_rate": 9.770056579138257e-06, + "loss": 0.0194, + "step": 14203 + }, + { + "epoch": 1.5398959236773635, + "grad_norm": 3.865276575088501, + "learning_rate": 9.76933120557087e-06, + "loss": 0.0318, + "step": 14204 + }, + { + "epoch": 1.540004336513443, + "grad_norm": 3.9081931114196777, + "learning_rate": 9.768605832003483e-06, + "loss": 0.0538, + "step": 14205 + }, + { + "epoch": 1.540112749349523, + "grad_norm": 2.004295587539673, + "learning_rate": 9.767880458436095e-06, + "loss": 0.0419, + "step": 14206 + }, + { + "epoch": 1.5402211621856028, + "grad_norm": 0.7420017719268799, + "learning_rate": 9.767155084868707e-06, + "loss": 0.0172, + "step": 14207 + }, + { + "epoch": 1.5403295750216826, + "grad_norm": 2.5332791805267334, + "learning_rate": 9.766429711301321e-06, + "loss": 0.0179, + "step": 14208 + }, + { + "epoch": 1.5404379878577623, + "grad_norm": 4.336400032043457, + "learning_rate": 9.765704337733933e-06, + "loss": 0.0266, + "step": 14209 + }, + { + "epoch": 1.540546400693842, + "grad_norm": 5.164210319519043, + "learning_rate": 9.764978964166547e-06, + "loss": 0.0114, + "step": 14210 + }, + { + "epoch": 1.540654813529922, + "grad_norm": 0.5041899681091309, + "learning_rate": 9.76425359059916e-06, + "loss": 0.0061, + "step": 14211 + }, + { + "epoch": 1.5407632263660016, + "grad_norm": 1.7782189846038818, + "learning_rate": 9.763528217031773e-06, + "loss": 0.0188, + "step": 14212 + }, + { + "epoch": 1.5408716392020816, + "grad_norm": 1.7106209993362427, + "learning_rate": 9.762802843464385e-06, + "loss": 0.019, + "step": 14213 + }, + { + "epoch": 1.5409800520381614, + "grad_norm": 0.4700433015823364, + "learning_rate": 9.762077469896997e-06, + "loss": 0.0024, + "step": 14214 + }, + { + "epoch": 1.5410884648742411, + "grad_norm": 3.6477091312408447, + "learning_rate": 9.761352096329611e-06, + "loss": 0.0325, + "step": 14215 + }, + { + "epoch": 1.5411968777103209, + "grad_norm": 0.5490761995315552, + "learning_rate": 9.760626722762223e-06, + "loss": 0.0064, + "step": 14216 + }, + { + "epoch": 1.5413052905464006, + "grad_norm": 2.8497700691223145, + "learning_rate": 9.759901349194835e-06, + "loss": 0.0134, + "step": 14217 + }, + { + "epoch": 1.5414137033824806, + "grad_norm": 0.42292627692222595, + "learning_rate": 9.75917597562745e-06, + "loss": 0.0029, + "step": 14218 + }, + { + "epoch": 1.5415221162185602, + "grad_norm": 2.6240406036376953, + "learning_rate": 9.758450602060061e-06, + "loss": 0.0074, + "step": 14219 + }, + { + "epoch": 1.5416305290546402, + "grad_norm": 1.416573405265808, + "learning_rate": 9.757725228492675e-06, + "loss": 0.0154, + "step": 14220 + }, + { + "epoch": 1.54173894189072, + "grad_norm": 1.560394048690796, + "learning_rate": 9.756999854925287e-06, + "loss": 0.0317, + "step": 14221 + }, + { + "epoch": 1.5418473547267997, + "grad_norm": 1.6167362928390503, + "learning_rate": 9.7562744813579e-06, + "loss": 0.0432, + "step": 14222 + }, + { + "epoch": 1.5419557675628794, + "grad_norm": 0.6468316912651062, + "learning_rate": 9.755549107790513e-06, + "loss": 0.0036, + "step": 14223 + }, + { + "epoch": 1.5420641803989592, + "grad_norm": 1.4294852018356323, + "learning_rate": 9.754823734223126e-06, + "loss": 0.0153, + "step": 14224 + }, + { + "epoch": 1.5421725932350392, + "grad_norm": 0.5414129495620728, + "learning_rate": 9.754098360655738e-06, + "loss": 0.0045, + "step": 14225 + }, + { + "epoch": 1.5422810060711187, + "grad_norm": 0.37428951263427734, + "learning_rate": 9.753372987088351e-06, + "loss": 0.0028, + "step": 14226 + }, + { + "epoch": 1.5423894189071987, + "grad_norm": 12.490592956542969, + "learning_rate": 9.752647613520964e-06, + "loss": 0.0819, + "step": 14227 + }, + { + "epoch": 1.5424978317432783, + "grad_norm": 1.7938698530197144, + "learning_rate": 9.751922239953577e-06, + "loss": 0.0197, + "step": 14228 + }, + { + "epoch": 1.5426062445793582, + "grad_norm": 1.0662322044372559, + "learning_rate": 9.75119686638619e-06, + "loss": 0.0056, + "step": 14229 + }, + { + "epoch": 1.542714657415438, + "grad_norm": 3.179255962371826, + "learning_rate": 9.750471492818803e-06, + "loss": 0.0155, + "step": 14230 + }, + { + "epoch": 1.5428230702515178, + "grad_norm": 2.775617837905884, + "learning_rate": 9.749746119251416e-06, + "loss": 0.0296, + "step": 14231 + }, + { + "epoch": 1.5429314830875975, + "grad_norm": 5.3937177658081055, + "learning_rate": 9.749020745684028e-06, + "loss": 0.0151, + "step": 14232 + }, + { + "epoch": 1.5430398959236773, + "grad_norm": 2.7500367164611816, + "learning_rate": 9.74829537211664e-06, + "loss": 0.0263, + "step": 14233 + }, + { + "epoch": 1.5431483087597573, + "grad_norm": 3.0976803302764893, + "learning_rate": 9.747569998549254e-06, + "loss": 0.0492, + "step": 14234 + }, + { + "epoch": 1.5432567215958368, + "grad_norm": 4.491341590881348, + "learning_rate": 9.746844624981866e-06, + "loss": 0.0591, + "step": 14235 + }, + { + "epoch": 1.5433651344319168, + "grad_norm": 6.7587995529174805, + "learning_rate": 9.74611925141448e-06, + "loss": 0.0058, + "step": 14236 + }, + { + "epoch": 1.5434735472679966, + "grad_norm": 0.6564464569091797, + "learning_rate": 9.745393877847092e-06, + "loss": 0.0078, + "step": 14237 + }, + { + "epoch": 1.5435819601040763, + "grad_norm": 0.39010223746299744, + "learning_rate": 9.744668504279706e-06, + "loss": 0.0028, + "step": 14238 + }, + { + "epoch": 1.543690372940156, + "grad_norm": 8.095755577087402, + "learning_rate": 9.743943130712318e-06, + "loss": 0.0242, + "step": 14239 + }, + { + "epoch": 1.5437987857762359, + "grad_norm": 2.180830240249634, + "learning_rate": 9.74321775714493e-06, + "loss": 0.0193, + "step": 14240 + }, + { + "epoch": 1.5439071986123158, + "grad_norm": 1.305379033088684, + "learning_rate": 9.742492383577544e-06, + "loss": 0.007, + "step": 14241 + }, + { + "epoch": 1.5440156114483954, + "grad_norm": 0.25747475028038025, + "learning_rate": 9.741767010010156e-06, + "loss": 0.0018, + "step": 14242 + }, + { + "epoch": 1.5441240242844754, + "grad_norm": 4.082169055938721, + "learning_rate": 9.741041636442768e-06, + "loss": 0.082, + "step": 14243 + }, + { + "epoch": 1.544232437120555, + "grad_norm": 2.613121509552002, + "learning_rate": 9.740316262875382e-06, + "loss": 0.0467, + "step": 14244 + }, + { + "epoch": 1.544340849956635, + "grad_norm": 3.0515551567077637, + "learning_rate": 9.739590889307994e-06, + "loss": 0.0323, + "step": 14245 + }, + { + "epoch": 1.5444492627927147, + "grad_norm": 0.9478558897972107, + "learning_rate": 9.738865515740608e-06, + "loss": 0.0027, + "step": 14246 + }, + { + "epoch": 1.5445576756287944, + "grad_norm": 0.8532777428627014, + "learning_rate": 9.73814014217322e-06, + "loss": 0.0108, + "step": 14247 + }, + { + "epoch": 1.5446660884648742, + "grad_norm": 5.128138065338135, + "learning_rate": 9.737414768605832e-06, + "loss": 0.035, + "step": 14248 + }, + { + "epoch": 1.544774501300954, + "grad_norm": 12.374074935913086, + "learning_rate": 9.736689395038446e-06, + "loss": 0.0251, + "step": 14249 + }, + { + "epoch": 1.544882914137034, + "grad_norm": 0.07577485591173172, + "learning_rate": 9.735964021471058e-06, + "loss": 0.0006, + "step": 14250 + }, + { + "epoch": 1.5449913269731135, + "grad_norm": 0.854146420955658, + "learning_rate": 9.73523864790367e-06, + "loss": 0.0092, + "step": 14251 + }, + { + "epoch": 1.5450997398091935, + "grad_norm": 0.10824564844369888, + "learning_rate": 9.734513274336284e-06, + "loss": 0.0016, + "step": 14252 + }, + { + "epoch": 1.5452081526452732, + "grad_norm": 5.395183563232422, + "learning_rate": 9.733787900768896e-06, + "loss": 0.0477, + "step": 14253 + }, + { + "epoch": 1.545316565481353, + "grad_norm": 5.538419246673584, + "learning_rate": 9.73306252720151e-06, + "loss": 0.076, + "step": 14254 + }, + { + "epoch": 1.5454249783174328, + "grad_norm": 2.593752384185791, + "learning_rate": 9.732337153634122e-06, + "loss": 0.0309, + "step": 14255 + }, + { + "epoch": 1.5455333911535125, + "grad_norm": 3.0388846397399902, + "learning_rate": 9.731611780066736e-06, + "loss": 0.0163, + "step": 14256 + }, + { + "epoch": 1.5456418039895925, + "grad_norm": 0.8527621030807495, + "learning_rate": 9.730886406499348e-06, + "loss": 0.0078, + "step": 14257 + }, + { + "epoch": 1.545750216825672, + "grad_norm": 4.821852207183838, + "learning_rate": 9.73016103293196e-06, + "loss": 0.02, + "step": 14258 + }, + { + "epoch": 1.545858629661752, + "grad_norm": 4.045724391937256, + "learning_rate": 9.729435659364573e-06, + "loss": 0.0681, + "step": 14259 + }, + { + "epoch": 1.5459670424978318, + "grad_norm": 1.8223336935043335, + "learning_rate": 9.728710285797186e-06, + "loss": 0.0118, + "step": 14260 + }, + { + "epoch": 1.5460754553339116, + "grad_norm": 1.3054159879684448, + "learning_rate": 9.727984912229799e-06, + "loss": 0.0146, + "step": 14261 + }, + { + "epoch": 1.5461838681699913, + "grad_norm": 0.6320802569389343, + "learning_rate": 9.727259538662412e-06, + "loss": 0.0026, + "step": 14262 + }, + { + "epoch": 1.546292281006071, + "grad_norm": 3.2912561893463135, + "learning_rate": 9.726534165095025e-06, + "loss": 0.0203, + "step": 14263 + }, + { + "epoch": 1.546400693842151, + "grad_norm": 5.754668712615967, + "learning_rate": 9.725808791527638e-06, + "loss": 0.0361, + "step": 14264 + }, + { + "epoch": 1.5465091066782306, + "grad_norm": 2.9443464279174805, + "learning_rate": 9.72508341796025e-06, + "loss": 0.0355, + "step": 14265 + }, + { + "epoch": 1.5466175195143106, + "grad_norm": 1.0152440071105957, + "learning_rate": 9.724358044392863e-06, + "loss": 0.0133, + "step": 14266 + }, + { + "epoch": 1.5467259323503901, + "grad_norm": 0.5754757523536682, + "learning_rate": 9.723632670825476e-06, + "loss": 0.0052, + "step": 14267 + }, + { + "epoch": 1.5468343451864701, + "grad_norm": 4.293646335601807, + "learning_rate": 9.722907297258089e-06, + "loss": 0.0059, + "step": 14268 + }, + { + "epoch": 1.5469427580225499, + "grad_norm": 8.051301002502441, + "learning_rate": 9.7221819236907e-06, + "loss": 0.0291, + "step": 14269 + }, + { + "epoch": 1.5470511708586296, + "grad_norm": 0.6996867060661316, + "learning_rate": 9.721456550123315e-06, + "loss": 0.0103, + "step": 14270 + }, + { + "epoch": 1.5471595836947094, + "grad_norm": 0.9862777590751648, + "learning_rate": 9.720731176555927e-06, + "loss": 0.0141, + "step": 14271 + }, + { + "epoch": 1.5472679965307892, + "grad_norm": 3.4558050632476807, + "learning_rate": 9.72000580298854e-06, + "loss": 0.0334, + "step": 14272 + }, + { + "epoch": 1.5473764093668692, + "grad_norm": 0.4179292321205139, + "learning_rate": 9.719280429421153e-06, + "loss": 0.0021, + "step": 14273 + }, + { + "epoch": 1.5474848222029487, + "grad_norm": 1.515329122543335, + "learning_rate": 9.718555055853765e-06, + "loss": 0.0099, + "step": 14274 + }, + { + "epoch": 1.5475932350390287, + "grad_norm": 2.1155967712402344, + "learning_rate": 9.717829682286379e-06, + "loss": 0.0225, + "step": 14275 + }, + { + "epoch": 1.5477016478751084, + "grad_norm": 10.146742820739746, + "learning_rate": 9.717104308718991e-06, + "loss": 0.051, + "step": 14276 + }, + { + "epoch": 1.5478100607111882, + "grad_norm": 7.056594371795654, + "learning_rate": 9.716378935151603e-06, + "loss": 0.0556, + "step": 14277 + }, + { + "epoch": 1.547918473547268, + "grad_norm": 3.4470374584198, + "learning_rate": 9.715653561584217e-06, + "loss": 0.0071, + "step": 14278 + }, + { + "epoch": 1.5480268863833477, + "grad_norm": 2.565157890319824, + "learning_rate": 9.714928188016829e-06, + "loss": 0.0488, + "step": 14279 + }, + { + "epoch": 1.5481352992194277, + "grad_norm": 4.094224452972412, + "learning_rate": 9.714202814449443e-06, + "loss": 0.0458, + "step": 14280 + }, + { + "epoch": 1.5482437120555073, + "grad_norm": 5.8140974044799805, + "learning_rate": 9.713477440882055e-06, + "loss": 0.0232, + "step": 14281 + }, + { + "epoch": 1.5483521248915872, + "grad_norm": 7.0509934425354, + "learning_rate": 9.712752067314669e-06, + "loss": 0.0285, + "step": 14282 + }, + { + "epoch": 1.548460537727667, + "grad_norm": 2.958479881286621, + "learning_rate": 9.712026693747281e-06, + "loss": 0.0446, + "step": 14283 + }, + { + "epoch": 1.5485689505637468, + "grad_norm": 2.0191409587860107, + "learning_rate": 9.711301320179893e-06, + "loss": 0.0347, + "step": 14284 + }, + { + "epoch": 1.5486773633998265, + "grad_norm": 2.4225964546203613, + "learning_rate": 9.710575946612505e-06, + "loss": 0.0099, + "step": 14285 + }, + { + "epoch": 1.5487857762359063, + "grad_norm": 1.8643814325332642, + "learning_rate": 9.709850573045119e-06, + "loss": 0.0284, + "step": 14286 + }, + { + "epoch": 1.5488941890719863, + "grad_norm": 4.3075714111328125, + "learning_rate": 9.709125199477731e-06, + "loss": 0.0358, + "step": 14287 + }, + { + "epoch": 1.5490026019080658, + "grad_norm": 4.894271373748779, + "learning_rate": 9.708399825910345e-06, + "loss": 0.0291, + "step": 14288 + }, + { + "epoch": 1.5491110147441458, + "grad_norm": 0.6145961284637451, + "learning_rate": 9.707674452342957e-06, + "loss": 0.0034, + "step": 14289 + }, + { + "epoch": 1.5492194275802254, + "grad_norm": 3.3119406700134277, + "learning_rate": 9.706949078775571e-06, + "loss": 0.0526, + "step": 14290 + }, + { + "epoch": 1.5493278404163053, + "grad_norm": 1.923689365386963, + "learning_rate": 9.706223705208183e-06, + "loss": 0.0265, + "step": 14291 + }, + { + "epoch": 1.549436253252385, + "grad_norm": 5.908219337463379, + "learning_rate": 9.705498331640795e-06, + "loss": 0.0586, + "step": 14292 + }, + { + "epoch": 1.5495446660884649, + "grad_norm": 2.431368350982666, + "learning_rate": 9.704772958073409e-06, + "loss": 0.0179, + "step": 14293 + }, + { + "epoch": 1.5496530789245446, + "grad_norm": 3.6114163398742676, + "learning_rate": 9.704047584506021e-06, + "loss": 0.0534, + "step": 14294 + }, + { + "epoch": 1.5497614917606244, + "grad_norm": 3.9445877075195312, + "learning_rate": 9.703322210938633e-06, + "loss": 0.0429, + "step": 14295 + }, + { + "epoch": 1.5498699045967044, + "grad_norm": 0.6286389827728271, + "learning_rate": 9.702596837371247e-06, + "loss": 0.0037, + "step": 14296 + }, + { + "epoch": 1.549978317432784, + "grad_norm": 1.3412048816680908, + "learning_rate": 9.70187146380386e-06, + "loss": 0.0248, + "step": 14297 + }, + { + "epoch": 1.550086730268864, + "grad_norm": 0.5248458981513977, + "learning_rate": 9.701146090236473e-06, + "loss": 0.0011, + "step": 14298 + }, + { + "epoch": 1.5501951431049437, + "grad_norm": 2.7597930431365967, + "learning_rate": 9.700420716669085e-06, + "loss": 0.0136, + "step": 14299 + }, + { + "epoch": 1.5503035559410234, + "grad_norm": 0.7529614567756653, + "learning_rate": 9.699695343101698e-06, + "loss": 0.0084, + "step": 14300 + }, + { + "epoch": 1.5504119687771032, + "grad_norm": 1.761410117149353, + "learning_rate": 9.698969969534311e-06, + "loss": 0.0182, + "step": 14301 + }, + { + "epoch": 1.550520381613183, + "grad_norm": 5.0083184242248535, + "learning_rate": 9.698244595966924e-06, + "loss": 0.0256, + "step": 14302 + }, + { + "epoch": 1.550628794449263, + "grad_norm": 2.921295166015625, + "learning_rate": 9.697519222399536e-06, + "loss": 0.0092, + "step": 14303 + }, + { + "epoch": 1.5507372072853425, + "grad_norm": 5.677841663360596, + "learning_rate": 9.69679384883215e-06, + "loss": 0.0321, + "step": 14304 + }, + { + "epoch": 1.5508456201214225, + "grad_norm": 0.06005515903234482, + "learning_rate": 9.696068475264763e-06, + "loss": 0.0007, + "step": 14305 + }, + { + "epoch": 1.550954032957502, + "grad_norm": 3.3423590660095215, + "learning_rate": 9.695343101697375e-06, + "loss": 0.058, + "step": 14306 + }, + { + "epoch": 1.551062445793582, + "grad_norm": 0.07310450822114944, + "learning_rate": 9.694617728129988e-06, + "loss": 0.0008, + "step": 14307 + }, + { + "epoch": 1.5511708586296618, + "grad_norm": 2.810248374938965, + "learning_rate": 9.693892354562601e-06, + "loss": 0.0277, + "step": 14308 + }, + { + "epoch": 1.5512792714657415, + "grad_norm": 1.5185132026672363, + "learning_rate": 9.693166980995214e-06, + "loss": 0.0131, + "step": 14309 + }, + { + "epoch": 1.5513876843018215, + "grad_norm": 0.9228154420852661, + "learning_rate": 9.692441607427826e-06, + "loss": 0.01, + "step": 14310 + }, + { + "epoch": 1.551496097137901, + "grad_norm": 3.6220993995666504, + "learning_rate": 9.691716233860438e-06, + "loss": 0.0161, + "step": 14311 + }, + { + "epoch": 1.551604509973981, + "grad_norm": 1.3856229782104492, + "learning_rate": 9.690990860293052e-06, + "loss": 0.0099, + "step": 14312 + }, + { + "epoch": 1.5517129228100606, + "grad_norm": 1.1292009353637695, + "learning_rate": 9.690265486725664e-06, + "loss": 0.016, + "step": 14313 + }, + { + "epoch": 1.5518213356461406, + "grad_norm": 2.0602169036865234, + "learning_rate": 9.689540113158278e-06, + "loss": 0.006, + "step": 14314 + }, + { + "epoch": 1.5519297484822203, + "grad_norm": 2.556767463684082, + "learning_rate": 9.68881473959089e-06, + "loss": 0.0329, + "step": 14315 + }, + { + "epoch": 1.5520381613183, + "grad_norm": 2.70023250579834, + "learning_rate": 9.688089366023504e-06, + "loss": 0.0201, + "step": 14316 + }, + { + "epoch": 1.5521465741543798, + "grad_norm": 7.1558146476745605, + "learning_rate": 9.687363992456116e-06, + "loss": 0.0315, + "step": 14317 + }, + { + "epoch": 1.5522549869904596, + "grad_norm": 3.7624897956848145, + "learning_rate": 9.686638618888728e-06, + "loss": 0.0575, + "step": 14318 + }, + { + "epoch": 1.5523633998265396, + "grad_norm": 4.143671035766602, + "learning_rate": 9.685913245321342e-06, + "loss": 0.0385, + "step": 14319 + }, + { + "epoch": 1.5524718126626191, + "grad_norm": 0.9759340882301331, + "learning_rate": 9.685187871753954e-06, + "loss": 0.0137, + "step": 14320 + }, + { + "epoch": 1.5525802254986991, + "grad_norm": 2.5630874633789062, + "learning_rate": 9.684462498186566e-06, + "loss": 0.0258, + "step": 14321 + }, + { + "epoch": 1.5526886383347789, + "grad_norm": 1.3747836351394653, + "learning_rate": 9.68373712461918e-06, + "loss": 0.0106, + "step": 14322 + }, + { + "epoch": 1.5527970511708586, + "grad_norm": 3.428826332092285, + "learning_rate": 9.683011751051792e-06, + "loss": 0.0323, + "step": 14323 + }, + { + "epoch": 1.5529054640069384, + "grad_norm": 0.15342745184898376, + "learning_rate": 9.682286377484406e-06, + "loss": 0.0011, + "step": 14324 + }, + { + "epoch": 1.5530138768430182, + "grad_norm": 2.2528936862945557, + "learning_rate": 9.681561003917018e-06, + "loss": 0.0215, + "step": 14325 + }, + { + "epoch": 1.5531222896790982, + "grad_norm": 1.8256901502609253, + "learning_rate": 9.68083563034963e-06, + "loss": 0.0118, + "step": 14326 + }, + { + "epoch": 1.5532307025151777, + "grad_norm": 1.0617984533309937, + "learning_rate": 9.680110256782244e-06, + "loss": 0.015, + "step": 14327 + }, + { + "epoch": 1.5533391153512577, + "grad_norm": 1.6571235656738281, + "learning_rate": 9.679384883214856e-06, + "loss": 0.0113, + "step": 14328 + }, + { + "epoch": 1.5534475281873372, + "grad_norm": 4.246592998504639, + "learning_rate": 9.678659509647468e-06, + "loss": 0.0181, + "step": 14329 + }, + { + "epoch": 1.5535559410234172, + "grad_norm": 2.4252142906188965, + "learning_rate": 9.677934136080082e-06, + "loss": 0.028, + "step": 14330 + }, + { + "epoch": 1.553664353859497, + "grad_norm": 6.754024982452393, + "learning_rate": 9.677208762512696e-06, + "loss": 0.0541, + "step": 14331 + }, + { + "epoch": 1.5537727666955767, + "grad_norm": 0.22424758970737457, + "learning_rate": 9.676483388945308e-06, + "loss": 0.0018, + "step": 14332 + }, + { + "epoch": 1.5538811795316565, + "grad_norm": 2.949390172958374, + "learning_rate": 9.67575801537792e-06, + "loss": 0.0121, + "step": 14333 + }, + { + "epoch": 1.5539895923677363, + "grad_norm": 3.504917860031128, + "learning_rate": 9.675032641810534e-06, + "loss": 0.053, + "step": 14334 + }, + { + "epoch": 1.5540980052038162, + "grad_norm": 2.7778704166412354, + "learning_rate": 9.674307268243146e-06, + "loss": 0.0381, + "step": 14335 + }, + { + "epoch": 1.5542064180398958, + "grad_norm": 10.507735252380371, + "learning_rate": 9.673581894675758e-06, + "loss": 0.078, + "step": 14336 + }, + { + "epoch": 1.5543148308759758, + "grad_norm": 1.4577254056930542, + "learning_rate": 9.67285652110837e-06, + "loss": 0.0161, + "step": 14337 + }, + { + "epoch": 1.5544232437120555, + "grad_norm": 1.8186511993408203, + "learning_rate": 9.672131147540984e-06, + "loss": 0.0098, + "step": 14338 + }, + { + "epoch": 1.5545316565481353, + "grad_norm": 0.3055398166179657, + "learning_rate": 9.671405773973597e-06, + "loss": 0.0049, + "step": 14339 + }, + { + "epoch": 1.554640069384215, + "grad_norm": 3.733910083770752, + "learning_rate": 9.67068040040621e-06, + "loss": 0.0446, + "step": 14340 + }, + { + "epoch": 1.5547484822202948, + "grad_norm": 4.681849479675293, + "learning_rate": 9.669955026838823e-06, + "loss": 0.0199, + "step": 14341 + }, + { + "epoch": 1.5548568950563748, + "grad_norm": 1.467466950416565, + "learning_rate": 9.669229653271436e-06, + "loss": 0.0301, + "step": 14342 + }, + { + "epoch": 1.5549653078924544, + "grad_norm": 0.34614667296409607, + "learning_rate": 9.668504279704049e-06, + "loss": 0.0015, + "step": 14343 + }, + { + "epoch": 1.5550737207285343, + "grad_norm": 6.967965602874756, + "learning_rate": 9.66777890613666e-06, + "loss": 0.0496, + "step": 14344 + }, + { + "epoch": 1.555182133564614, + "grad_norm": 0.39702218770980835, + "learning_rate": 9.667053532569274e-06, + "loss": 0.0027, + "step": 14345 + }, + { + "epoch": 1.5552905464006939, + "grad_norm": 6.148124694824219, + "learning_rate": 9.666328159001887e-06, + "loss": 0.0217, + "step": 14346 + }, + { + "epoch": 1.5553989592367736, + "grad_norm": 1.8516809940338135, + "learning_rate": 9.665602785434499e-06, + "loss": 0.0214, + "step": 14347 + }, + { + "epoch": 1.5555073720728534, + "grad_norm": 2.067577838897705, + "learning_rate": 9.664877411867113e-06, + "loss": 0.0223, + "step": 14348 + }, + { + "epoch": 1.5556157849089334, + "grad_norm": 0.3794756531715393, + "learning_rate": 9.664152038299725e-06, + "loss": 0.002, + "step": 14349 + }, + { + "epoch": 1.555724197745013, + "grad_norm": 2.3578267097473145, + "learning_rate": 9.663426664732339e-06, + "loss": 0.024, + "step": 14350 + }, + { + "epoch": 1.555832610581093, + "grad_norm": 4.52170991897583, + "learning_rate": 9.66270129116495e-06, + "loss": 0.0219, + "step": 14351 + }, + { + "epoch": 1.5559410234171724, + "grad_norm": 1.7458595037460327, + "learning_rate": 9.661975917597563e-06, + "loss": 0.0316, + "step": 14352 + }, + { + "epoch": 1.5560494362532524, + "grad_norm": 4.763866424560547, + "learning_rate": 9.661250544030177e-06, + "loss": 0.025, + "step": 14353 + }, + { + "epoch": 1.5561578490893322, + "grad_norm": 2.122694730758667, + "learning_rate": 9.660525170462789e-06, + "loss": 0.0344, + "step": 14354 + }, + { + "epoch": 1.556266261925412, + "grad_norm": 2.6252663135528564, + "learning_rate": 9.659799796895401e-06, + "loss": 0.0206, + "step": 14355 + }, + { + "epoch": 1.5563746747614917, + "grad_norm": 3.7175703048706055, + "learning_rate": 9.659074423328015e-06, + "loss": 0.0755, + "step": 14356 + }, + { + "epoch": 1.5564830875975715, + "grad_norm": 1.2769174575805664, + "learning_rate": 9.658349049760629e-06, + "loss": 0.0104, + "step": 14357 + }, + { + "epoch": 1.5565915004336515, + "grad_norm": 3.778883218765259, + "learning_rate": 9.65762367619324e-06, + "loss": 0.0163, + "step": 14358 + }, + { + "epoch": 1.556699913269731, + "grad_norm": 0.4061899483203888, + "learning_rate": 9.656898302625853e-06, + "loss": 0.0041, + "step": 14359 + }, + { + "epoch": 1.556808326105811, + "grad_norm": 1.0340112447738647, + "learning_rate": 9.656172929058467e-06, + "loss": 0.006, + "step": 14360 + }, + { + "epoch": 1.5569167389418908, + "grad_norm": 2.662832260131836, + "learning_rate": 9.655447555491079e-06, + "loss": 0.0117, + "step": 14361 + }, + { + "epoch": 1.5570251517779705, + "grad_norm": 1.663055658340454, + "learning_rate": 9.654722181923691e-06, + "loss": 0.0252, + "step": 14362 + }, + { + "epoch": 1.5571335646140503, + "grad_norm": 3.1844589710235596, + "learning_rate": 9.653996808356303e-06, + "loss": 0.0442, + "step": 14363 + }, + { + "epoch": 1.55724197745013, + "grad_norm": 1.6799335479736328, + "learning_rate": 9.653271434788917e-06, + "loss": 0.0142, + "step": 14364 + }, + { + "epoch": 1.55735039028621, + "grad_norm": 4.973076820373535, + "learning_rate": 9.65254606122153e-06, + "loss": 0.0477, + "step": 14365 + }, + { + "epoch": 1.5574588031222896, + "grad_norm": 10.087081909179688, + "learning_rate": 9.651820687654143e-06, + "loss": 0.0823, + "step": 14366 + }, + { + "epoch": 1.5575672159583696, + "grad_norm": 2.4530766010284424, + "learning_rate": 9.651095314086755e-06, + "loss": 0.0097, + "step": 14367 + }, + { + "epoch": 1.557675628794449, + "grad_norm": 2.1080822944641113, + "learning_rate": 9.650369940519369e-06, + "loss": 0.0448, + "step": 14368 + }, + { + "epoch": 1.557784041630529, + "grad_norm": 0.5639559030532837, + "learning_rate": 9.649644566951981e-06, + "loss": 0.0024, + "step": 14369 + }, + { + "epoch": 1.5578924544666088, + "grad_norm": 1.3951555490493774, + "learning_rate": 9.648919193384593e-06, + "loss": 0.0156, + "step": 14370 + }, + { + "epoch": 1.5580008673026886, + "grad_norm": 7.692814350128174, + "learning_rate": 9.648193819817207e-06, + "loss": 0.0604, + "step": 14371 + }, + { + "epoch": 1.5581092801387686, + "grad_norm": 0.8651379346847534, + "learning_rate": 9.64746844624982e-06, + "loss": 0.0234, + "step": 14372 + }, + { + "epoch": 1.5582176929748481, + "grad_norm": 4.27868127822876, + "learning_rate": 9.646743072682431e-06, + "loss": 0.022, + "step": 14373 + }, + { + "epoch": 1.5583261058109281, + "grad_norm": 5.865891456604004, + "learning_rate": 9.646017699115045e-06, + "loss": 0.0819, + "step": 14374 + }, + { + "epoch": 1.5584345186470077, + "grad_norm": 2.730555534362793, + "learning_rate": 9.645292325547657e-06, + "loss": 0.0905, + "step": 14375 + }, + { + "epoch": 1.5585429314830876, + "grad_norm": 1.2246925830841064, + "learning_rate": 9.644566951980271e-06, + "loss": 0.0119, + "step": 14376 + }, + { + "epoch": 1.5586513443191674, + "grad_norm": 0.8823467493057251, + "learning_rate": 9.643841578412883e-06, + "loss": 0.0136, + "step": 14377 + }, + { + "epoch": 1.5587597571552472, + "grad_norm": 2.7346954345703125, + "learning_rate": 9.643116204845496e-06, + "loss": 0.022, + "step": 14378 + }, + { + "epoch": 1.558868169991327, + "grad_norm": 3.1141016483306885, + "learning_rate": 9.64239083127811e-06, + "loss": 0.0364, + "step": 14379 + }, + { + "epoch": 1.5589765828274067, + "grad_norm": 2.80971622467041, + "learning_rate": 9.641665457710722e-06, + "loss": 0.0059, + "step": 14380 + }, + { + "epoch": 1.5590849956634867, + "grad_norm": 2.2356669902801514, + "learning_rate": 9.640940084143334e-06, + "loss": 0.037, + "step": 14381 + }, + { + "epoch": 1.5591934084995662, + "grad_norm": 2.5523643493652344, + "learning_rate": 9.640214710575948e-06, + "loss": 0.0316, + "step": 14382 + }, + { + "epoch": 1.5593018213356462, + "grad_norm": 2.292649030685425, + "learning_rate": 9.639489337008561e-06, + "loss": 0.039, + "step": 14383 + }, + { + "epoch": 1.559410234171726, + "grad_norm": 1.5554243326187134, + "learning_rate": 9.638763963441173e-06, + "loss": 0.0046, + "step": 14384 + }, + { + "epoch": 1.5595186470078057, + "grad_norm": 4.759666919708252, + "learning_rate": 9.638038589873786e-06, + "loss": 0.047, + "step": 14385 + }, + { + "epoch": 1.5596270598438855, + "grad_norm": 1.4486908912658691, + "learning_rate": 9.6373132163064e-06, + "loss": 0.0362, + "step": 14386 + }, + { + "epoch": 1.5597354726799653, + "grad_norm": 2.64290189743042, + "learning_rate": 9.636587842739012e-06, + "loss": 0.0156, + "step": 14387 + }, + { + "epoch": 1.5598438855160452, + "grad_norm": 3.1934633255004883, + "learning_rate": 9.635862469171624e-06, + "loss": 0.0322, + "step": 14388 + }, + { + "epoch": 1.5599522983521248, + "grad_norm": 0.41971802711486816, + "learning_rate": 9.635137095604236e-06, + "loss": 0.0046, + "step": 14389 + }, + { + "epoch": 1.5600607111882048, + "grad_norm": 0.1836715042591095, + "learning_rate": 9.63441172203685e-06, + "loss": 0.0007, + "step": 14390 + }, + { + "epoch": 1.5601691240242843, + "grad_norm": 4.972509384155273, + "learning_rate": 9.633686348469462e-06, + "loss": 0.0275, + "step": 14391 + }, + { + "epoch": 1.5602775368603643, + "grad_norm": 0.6445940136909485, + "learning_rate": 9.632960974902076e-06, + "loss": 0.0073, + "step": 14392 + }, + { + "epoch": 1.560385949696444, + "grad_norm": 3.1793363094329834, + "learning_rate": 9.632235601334688e-06, + "loss": 0.0593, + "step": 14393 + }, + { + "epoch": 1.5604943625325238, + "grad_norm": 0.3871646225452423, + "learning_rate": 9.631510227767302e-06, + "loss": 0.008, + "step": 14394 + }, + { + "epoch": 1.5606027753686036, + "grad_norm": 1.1733012199401855, + "learning_rate": 9.630784854199914e-06, + "loss": 0.0127, + "step": 14395 + }, + { + "epoch": 1.5607111882046834, + "grad_norm": 7.94036865234375, + "learning_rate": 9.630059480632526e-06, + "loss": 0.0448, + "step": 14396 + }, + { + "epoch": 1.5608196010407633, + "grad_norm": 0.8136045336723328, + "learning_rate": 9.62933410706514e-06, + "loss": 0.0032, + "step": 14397 + }, + { + "epoch": 1.5609280138768429, + "grad_norm": 1.7327996492385864, + "learning_rate": 9.628608733497752e-06, + "loss": 0.0153, + "step": 14398 + }, + { + "epoch": 1.5610364267129229, + "grad_norm": 2.695162534713745, + "learning_rate": 9.627883359930364e-06, + "loss": 0.0855, + "step": 14399 + }, + { + "epoch": 1.5611448395490026, + "grad_norm": 5.625560760498047, + "learning_rate": 9.627157986362978e-06, + "loss": 0.063, + "step": 14400 + }, + { + "epoch": 1.5612532523850824, + "grad_norm": 5.751926898956299, + "learning_rate": 9.626432612795592e-06, + "loss": 0.0181, + "step": 14401 + }, + { + "epoch": 1.5613616652211622, + "grad_norm": 0.8973345160484314, + "learning_rate": 9.625707239228204e-06, + "loss": 0.0081, + "step": 14402 + }, + { + "epoch": 1.561470078057242, + "grad_norm": 1.4477366209030151, + "learning_rate": 9.624981865660816e-06, + "loss": 0.0268, + "step": 14403 + }, + { + "epoch": 1.561578490893322, + "grad_norm": 0.3901292085647583, + "learning_rate": 9.624256492093428e-06, + "loss": 0.0032, + "step": 14404 + }, + { + "epoch": 1.5616869037294014, + "grad_norm": 2.795372486114502, + "learning_rate": 9.623531118526042e-06, + "loss": 0.0178, + "step": 14405 + }, + { + "epoch": 1.5617953165654814, + "grad_norm": 1.295261025428772, + "learning_rate": 9.622805744958654e-06, + "loss": 0.0164, + "step": 14406 + }, + { + "epoch": 1.5619037294015612, + "grad_norm": 1.2381346225738525, + "learning_rate": 9.622080371391266e-06, + "loss": 0.0187, + "step": 14407 + }, + { + "epoch": 1.562012142237641, + "grad_norm": 1.5724059343338013, + "learning_rate": 9.62135499782388e-06, + "loss": 0.0057, + "step": 14408 + }, + { + "epoch": 1.5621205550737207, + "grad_norm": 0.2867882251739502, + "learning_rate": 9.620629624256494e-06, + "loss": 0.0043, + "step": 14409 + }, + { + "epoch": 1.5622289679098005, + "grad_norm": 1.638141393661499, + "learning_rate": 9.619904250689106e-06, + "loss": 0.0352, + "step": 14410 + }, + { + "epoch": 1.5623373807458805, + "grad_norm": 1.811963677406311, + "learning_rate": 9.619178877121718e-06, + "loss": 0.0213, + "step": 14411 + }, + { + "epoch": 1.56244579358196, + "grad_norm": 0.354973703622818, + "learning_rate": 9.618453503554332e-06, + "loss": 0.004, + "step": 14412 + }, + { + "epoch": 1.56255420641804, + "grad_norm": 0.2182742804288864, + "learning_rate": 9.617728129986944e-06, + "loss": 0.0024, + "step": 14413 + }, + { + "epoch": 1.5626626192541195, + "grad_norm": 0.48913124203681946, + "learning_rate": 9.617002756419556e-06, + "loss": 0.0063, + "step": 14414 + }, + { + "epoch": 1.5627710320901995, + "grad_norm": 0.8031596541404724, + "learning_rate": 9.616277382852169e-06, + "loss": 0.0063, + "step": 14415 + }, + { + "epoch": 1.5628794449262793, + "grad_norm": 8.134342193603516, + "learning_rate": 9.615552009284782e-06, + "loss": 0.0321, + "step": 14416 + }, + { + "epoch": 1.562987857762359, + "grad_norm": 2.54020094871521, + "learning_rate": 9.614826635717395e-06, + "loss": 0.026, + "step": 14417 + }, + { + "epoch": 1.5630962705984388, + "grad_norm": 1.166176676750183, + "learning_rate": 9.614101262150008e-06, + "loss": 0.0104, + "step": 14418 + }, + { + "epoch": 1.5632046834345186, + "grad_norm": 1.5311298370361328, + "learning_rate": 9.61337588858262e-06, + "loss": 0.0242, + "step": 14419 + }, + { + "epoch": 1.5633130962705986, + "grad_norm": 1.6998827457427979, + "learning_rate": 9.612650515015234e-06, + "loss": 0.0244, + "step": 14420 + }, + { + "epoch": 1.563421509106678, + "grad_norm": 1.4670865535736084, + "learning_rate": 9.611925141447847e-06, + "loss": 0.0274, + "step": 14421 + }, + { + "epoch": 1.563529921942758, + "grad_norm": 0.5289870500564575, + "learning_rate": 9.611199767880459e-06, + "loss": 0.0037, + "step": 14422 + }, + { + "epoch": 1.5636383347788378, + "grad_norm": 0.6499745845794678, + "learning_rate": 9.610474394313072e-06, + "loss": 0.005, + "step": 14423 + }, + { + "epoch": 1.5637467476149176, + "grad_norm": 1.5763322114944458, + "learning_rate": 9.609749020745685e-06, + "loss": 0.0075, + "step": 14424 + }, + { + "epoch": 1.5638551604509974, + "grad_norm": 2.6111719608306885, + "learning_rate": 9.609023647178297e-06, + "loss": 0.035, + "step": 14425 + }, + { + "epoch": 1.5639635732870771, + "grad_norm": 1.5179404020309448, + "learning_rate": 9.60829827361091e-06, + "loss": 0.0171, + "step": 14426 + }, + { + "epoch": 1.5640719861231571, + "grad_norm": 1.7066452503204346, + "learning_rate": 9.607572900043524e-06, + "loss": 0.0183, + "step": 14427 + }, + { + "epoch": 1.5641803989592367, + "grad_norm": 1.350555181503296, + "learning_rate": 9.606847526476137e-06, + "loss": 0.0134, + "step": 14428 + }, + { + "epoch": 1.5642888117953166, + "grad_norm": 1.662276029586792, + "learning_rate": 9.606122152908749e-06, + "loss": 0.0071, + "step": 14429 + }, + { + "epoch": 1.5643972246313964, + "grad_norm": 1.2678124904632568, + "learning_rate": 9.605396779341361e-06, + "loss": 0.0212, + "step": 14430 + }, + { + "epoch": 1.5645056374674762, + "grad_norm": 2.5159897804260254, + "learning_rate": 9.604671405773975e-06, + "loss": 0.0556, + "step": 14431 + }, + { + "epoch": 1.564614050303556, + "grad_norm": 0.7075349688529968, + "learning_rate": 9.603946032206587e-06, + "loss": 0.0034, + "step": 14432 + }, + { + "epoch": 1.5647224631396357, + "grad_norm": 1.0739189386367798, + "learning_rate": 9.603220658639199e-06, + "loss": 0.0088, + "step": 14433 + }, + { + "epoch": 1.5648308759757157, + "grad_norm": 5.44141149520874, + "learning_rate": 9.602495285071813e-06, + "loss": 0.0206, + "step": 14434 + }, + { + "epoch": 1.5649392888117952, + "grad_norm": 0.1853155940771103, + "learning_rate": 9.601769911504427e-06, + "loss": 0.0009, + "step": 14435 + }, + { + "epoch": 1.5650477016478752, + "grad_norm": 1.095808506011963, + "learning_rate": 9.601044537937039e-06, + "loss": 0.0141, + "step": 14436 + }, + { + "epoch": 1.5651561144839548, + "grad_norm": 0.7570106387138367, + "learning_rate": 9.600319164369651e-06, + "loss": 0.005, + "step": 14437 + }, + { + "epoch": 1.5652645273200347, + "grad_norm": 9.14699935913086, + "learning_rate": 9.599593790802265e-06, + "loss": 0.0393, + "step": 14438 + }, + { + "epoch": 1.5653729401561145, + "grad_norm": 1.0914056301116943, + "learning_rate": 9.598868417234877e-06, + "loss": 0.0085, + "step": 14439 + }, + { + "epoch": 1.5654813529921943, + "grad_norm": 4.171805381774902, + "learning_rate": 9.598143043667489e-06, + "loss": 0.017, + "step": 14440 + }, + { + "epoch": 1.565589765828274, + "grad_norm": 0.8719685077667236, + "learning_rate": 9.597417670100101e-06, + "loss": 0.0098, + "step": 14441 + }, + { + "epoch": 1.5656981786643538, + "grad_norm": 2.062650680541992, + "learning_rate": 9.596692296532715e-06, + "loss": 0.0241, + "step": 14442 + }, + { + "epoch": 1.5658065915004338, + "grad_norm": 2.13566255569458, + "learning_rate": 9.595966922965327e-06, + "loss": 0.0273, + "step": 14443 + }, + { + "epoch": 1.5659150043365133, + "grad_norm": 0.19598279893398285, + "learning_rate": 9.595241549397941e-06, + "loss": 0.0018, + "step": 14444 + }, + { + "epoch": 1.5660234171725933, + "grad_norm": 3.332667350769043, + "learning_rate": 9.594516175830553e-06, + "loss": 0.0115, + "step": 14445 + }, + { + "epoch": 1.566131830008673, + "grad_norm": 4.430901527404785, + "learning_rate": 9.593790802263167e-06, + "loss": 0.0278, + "step": 14446 + }, + { + "epoch": 1.5662402428447528, + "grad_norm": 2.4270055294036865, + "learning_rate": 9.59306542869578e-06, + "loss": 0.0251, + "step": 14447 + }, + { + "epoch": 1.5663486556808326, + "grad_norm": 8.705775260925293, + "learning_rate": 9.592340055128391e-06, + "loss": 0.0373, + "step": 14448 + }, + { + "epoch": 1.5664570685169124, + "grad_norm": 2.2236156463623047, + "learning_rate": 9.591614681561005e-06, + "loss": 0.003, + "step": 14449 + }, + { + "epoch": 1.5665654813529923, + "grad_norm": 0.921607494354248, + "learning_rate": 9.590889307993617e-06, + "loss": 0.0063, + "step": 14450 + }, + { + "epoch": 1.5666738941890719, + "grad_norm": 2.6485350131988525, + "learning_rate": 9.59016393442623e-06, + "loss": 0.0115, + "step": 14451 + }, + { + "epoch": 1.5667823070251519, + "grad_norm": 2.433309316635132, + "learning_rate": 9.589438560858843e-06, + "loss": 0.0212, + "step": 14452 + }, + { + "epoch": 1.5668907198612314, + "grad_norm": 0.13400502502918243, + "learning_rate": 9.588713187291457e-06, + "loss": 0.0007, + "step": 14453 + }, + { + "epoch": 1.5669991326973114, + "grad_norm": 11.080498695373535, + "learning_rate": 9.58798781372407e-06, + "loss": 0.0675, + "step": 14454 + }, + { + "epoch": 1.5671075455333912, + "grad_norm": 1.6908742189407349, + "learning_rate": 9.587262440156681e-06, + "loss": 0.0066, + "step": 14455 + }, + { + "epoch": 1.567215958369471, + "grad_norm": 2.316802740097046, + "learning_rate": 9.586537066589294e-06, + "loss": 0.0505, + "step": 14456 + }, + { + "epoch": 1.567324371205551, + "grad_norm": 0.433262437582016, + "learning_rate": 9.585811693021907e-06, + "loss": 0.0024, + "step": 14457 + }, + { + "epoch": 1.5674327840416304, + "grad_norm": 0.16641823947429657, + "learning_rate": 9.58508631945452e-06, + "loss": 0.0016, + "step": 14458 + }, + { + "epoch": 1.5675411968777104, + "grad_norm": 1.7683402299880981, + "learning_rate": 9.584360945887132e-06, + "loss": 0.0204, + "step": 14459 + }, + { + "epoch": 1.56764960971379, + "grad_norm": 0.2045791745185852, + "learning_rate": 9.583635572319746e-06, + "loss": 0.0021, + "step": 14460 + }, + { + "epoch": 1.56775802254987, + "grad_norm": 1.2716200351715088, + "learning_rate": 9.58291019875236e-06, + "loss": 0.0118, + "step": 14461 + }, + { + "epoch": 1.5678664353859497, + "grad_norm": 1.5318316221237183, + "learning_rate": 9.582184825184971e-06, + "loss": 0.009, + "step": 14462 + }, + { + "epoch": 1.5679748482220295, + "grad_norm": 5.020758628845215, + "learning_rate": 9.581459451617584e-06, + "loss": 0.0433, + "step": 14463 + }, + { + "epoch": 1.5680832610581092, + "grad_norm": 2.74118709564209, + "learning_rate": 9.580734078050197e-06, + "loss": 0.0576, + "step": 14464 + }, + { + "epoch": 1.568191673894189, + "grad_norm": 1.3703736066818237, + "learning_rate": 9.58000870448281e-06, + "loss": 0.0211, + "step": 14465 + }, + { + "epoch": 1.568300086730269, + "grad_norm": 0.44957736134529114, + "learning_rate": 9.579283330915422e-06, + "loss": 0.0012, + "step": 14466 + }, + { + "epoch": 1.5684084995663485, + "grad_norm": 0.044563740491867065, + "learning_rate": 9.578557957348034e-06, + "loss": 0.0003, + "step": 14467 + }, + { + "epoch": 1.5685169124024285, + "grad_norm": 7.258132457733154, + "learning_rate": 9.577832583780648e-06, + "loss": 0.0343, + "step": 14468 + }, + { + "epoch": 1.5686253252385083, + "grad_norm": 2.936361312866211, + "learning_rate": 9.57710721021326e-06, + "loss": 0.0495, + "step": 14469 + }, + { + "epoch": 1.568733738074588, + "grad_norm": 1.9007495641708374, + "learning_rate": 9.576381836645874e-06, + "loss": 0.02, + "step": 14470 + }, + { + "epoch": 1.5688421509106678, + "grad_norm": 0.9648280739784241, + "learning_rate": 9.575656463078486e-06, + "loss": 0.011, + "step": 14471 + }, + { + "epoch": 1.5689505637467476, + "grad_norm": 4.591507434844971, + "learning_rate": 9.5749310895111e-06, + "loss": 0.0792, + "step": 14472 + }, + { + "epoch": 1.5690589765828276, + "grad_norm": 1.6000218391418457, + "learning_rate": 9.574205715943712e-06, + "loss": 0.0169, + "step": 14473 + }, + { + "epoch": 1.569167389418907, + "grad_norm": 3.2762837409973145, + "learning_rate": 9.573480342376324e-06, + "loss": 0.012, + "step": 14474 + }, + { + "epoch": 1.569275802254987, + "grad_norm": 1.5930861234664917, + "learning_rate": 9.572754968808938e-06, + "loss": 0.0285, + "step": 14475 + }, + { + "epoch": 1.5693842150910666, + "grad_norm": 3.546323537826538, + "learning_rate": 9.57202959524155e-06, + "loss": 0.0261, + "step": 14476 + }, + { + "epoch": 1.5694926279271466, + "grad_norm": 0.32206588983535767, + "learning_rate": 9.571304221674162e-06, + "loss": 0.002, + "step": 14477 + }, + { + "epoch": 1.5696010407632264, + "grad_norm": 2.458566427230835, + "learning_rate": 9.570578848106776e-06, + "loss": 0.0089, + "step": 14478 + }, + { + "epoch": 1.5697094535993061, + "grad_norm": 0.4398423135280609, + "learning_rate": 9.56985347453939e-06, + "loss": 0.0027, + "step": 14479 + }, + { + "epoch": 1.569817866435386, + "grad_norm": 1.69191575050354, + "learning_rate": 9.569128100972002e-06, + "loss": 0.0065, + "step": 14480 + }, + { + "epoch": 1.5699262792714657, + "grad_norm": 3.376540422439575, + "learning_rate": 9.568402727404614e-06, + "loss": 0.0802, + "step": 14481 + }, + { + "epoch": 1.5700346921075456, + "grad_norm": 0.48374924063682556, + "learning_rate": 9.567677353837226e-06, + "loss": 0.0015, + "step": 14482 + }, + { + "epoch": 1.5701431049436252, + "grad_norm": 8.270515441894531, + "learning_rate": 9.56695198026984e-06, + "loss": 0.0135, + "step": 14483 + }, + { + "epoch": 1.5702515177797052, + "grad_norm": 0.37119579315185547, + "learning_rate": 9.566226606702452e-06, + "loss": 0.0018, + "step": 14484 + }, + { + "epoch": 1.570359930615785, + "grad_norm": 0.594193696975708, + "learning_rate": 9.565501233135064e-06, + "loss": 0.0034, + "step": 14485 + }, + { + "epoch": 1.5704683434518647, + "grad_norm": 6.4753289222717285, + "learning_rate": 9.564775859567678e-06, + "loss": 0.1, + "step": 14486 + }, + { + "epoch": 1.5705767562879445, + "grad_norm": 1.9673264026641846, + "learning_rate": 9.564050486000292e-06, + "loss": 0.0114, + "step": 14487 + }, + { + "epoch": 1.5706851691240242, + "grad_norm": 2.7590513229370117, + "learning_rate": 9.563325112432904e-06, + "loss": 0.0493, + "step": 14488 + }, + { + "epoch": 1.5707935819601042, + "grad_norm": 0.1331629902124405, + "learning_rate": 9.562599738865516e-06, + "loss": 0.0015, + "step": 14489 + }, + { + "epoch": 1.5709019947961838, + "grad_norm": 0.14655211567878723, + "learning_rate": 9.56187436529813e-06, + "loss": 0.0014, + "step": 14490 + }, + { + "epoch": 1.5710104076322637, + "grad_norm": 0.8996156454086304, + "learning_rate": 9.561148991730742e-06, + "loss": 0.0137, + "step": 14491 + }, + { + "epoch": 1.5711188204683435, + "grad_norm": 4.505303859710693, + "learning_rate": 9.560423618163354e-06, + "loss": 0.0331, + "step": 14492 + }, + { + "epoch": 1.5712272333044233, + "grad_norm": 0.5206652879714966, + "learning_rate": 9.559698244595967e-06, + "loss": 0.0068, + "step": 14493 + }, + { + "epoch": 1.571335646140503, + "grad_norm": 0.6591737270355225, + "learning_rate": 9.55897287102858e-06, + "loss": 0.0087, + "step": 14494 + }, + { + "epoch": 1.5714440589765828, + "grad_norm": 3.7197105884552, + "learning_rate": 9.558247497461194e-06, + "loss": 0.039, + "step": 14495 + }, + { + "epoch": 1.5715524718126628, + "grad_norm": 5.790780067443848, + "learning_rate": 9.557522123893806e-06, + "loss": 0.0198, + "step": 14496 + }, + { + "epoch": 1.5716608846487423, + "grad_norm": 0.13958552479743958, + "learning_rate": 9.556796750326419e-06, + "loss": 0.0012, + "step": 14497 + }, + { + "epoch": 1.5717692974848223, + "grad_norm": 4.787703037261963, + "learning_rate": 9.556071376759032e-06, + "loss": 0.0191, + "step": 14498 + }, + { + "epoch": 1.5718777103209018, + "grad_norm": 1.723335862159729, + "learning_rate": 9.555346003191645e-06, + "loss": 0.0272, + "step": 14499 + }, + { + "epoch": 1.5719861231569818, + "grad_norm": 4.463651657104492, + "learning_rate": 9.554620629624257e-06, + "loss": 0.0565, + "step": 14500 + }, + { + "epoch": 1.5720945359930616, + "grad_norm": 5.23942232131958, + "learning_rate": 9.55389525605687e-06, + "loss": 0.0616, + "step": 14501 + }, + { + "epoch": 1.5722029488291414, + "grad_norm": 1.3402955532073975, + "learning_rate": 9.553169882489483e-06, + "loss": 0.012, + "step": 14502 + }, + { + "epoch": 1.5723113616652211, + "grad_norm": 0.2857479453086853, + "learning_rate": 9.552444508922095e-06, + "loss": 0.003, + "step": 14503 + }, + { + "epoch": 1.5724197745013009, + "grad_norm": 4.10928201675415, + "learning_rate": 9.551719135354709e-06, + "loss": 0.041, + "step": 14504 + }, + { + "epoch": 1.5725281873373809, + "grad_norm": 4.241065979003906, + "learning_rate": 9.550993761787322e-06, + "loss": 0.0296, + "step": 14505 + }, + { + "epoch": 1.5726366001734604, + "grad_norm": 0.2658098340034485, + "learning_rate": 9.550268388219935e-06, + "loss": 0.0028, + "step": 14506 + }, + { + "epoch": 1.5727450130095404, + "grad_norm": 14.34565258026123, + "learning_rate": 9.549543014652547e-06, + "loss": 0.0139, + "step": 14507 + }, + { + "epoch": 1.5728534258456202, + "grad_norm": 6.257487773895264, + "learning_rate": 9.548817641085159e-06, + "loss": 0.0239, + "step": 14508 + }, + { + "epoch": 1.5729618386817, + "grad_norm": 0.21417759358882904, + "learning_rate": 9.548092267517773e-06, + "loss": 0.0027, + "step": 14509 + }, + { + "epoch": 1.5730702515177797, + "grad_norm": 3.2118079662323, + "learning_rate": 9.547366893950385e-06, + "loss": 0.0486, + "step": 14510 + }, + { + "epoch": 1.5731786643538594, + "grad_norm": 1.1421009302139282, + "learning_rate": 9.546641520382997e-06, + "loss": 0.0028, + "step": 14511 + }, + { + "epoch": 1.5732870771899394, + "grad_norm": 0.2655344009399414, + "learning_rate": 9.545916146815611e-06, + "loss": 0.0013, + "step": 14512 + }, + { + "epoch": 1.573395490026019, + "grad_norm": 0.22099047899246216, + "learning_rate": 9.545190773248225e-06, + "loss": 0.0032, + "step": 14513 + }, + { + "epoch": 1.573503902862099, + "grad_norm": 1.1188397407531738, + "learning_rate": 9.544465399680837e-06, + "loss": 0.0025, + "step": 14514 + }, + { + "epoch": 1.5736123156981785, + "grad_norm": 2.746340751647949, + "learning_rate": 9.543740026113449e-06, + "loss": 0.0068, + "step": 14515 + }, + { + "epoch": 1.5737207285342585, + "grad_norm": 3.4036643505096436, + "learning_rate": 9.543014652546063e-06, + "loss": 0.0124, + "step": 14516 + }, + { + "epoch": 1.5738291413703382, + "grad_norm": 1.1340970993041992, + "learning_rate": 9.542289278978675e-06, + "loss": 0.0094, + "step": 14517 + }, + { + "epoch": 1.573937554206418, + "grad_norm": 0.5753812789916992, + "learning_rate": 9.541563905411287e-06, + "loss": 0.0032, + "step": 14518 + }, + { + "epoch": 1.574045967042498, + "grad_norm": 4.1804375648498535, + "learning_rate": 9.5408385318439e-06, + "loss": 0.0456, + "step": 14519 + }, + { + "epoch": 1.5741543798785775, + "grad_norm": 5.272585391998291, + "learning_rate": 9.540113158276513e-06, + "loss": 0.0137, + "step": 14520 + }, + { + "epoch": 1.5742627927146575, + "grad_norm": 5.194829940795898, + "learning_rate": 9.539387784709127e-06, + "loss": 0.0398, + "step": 14521 + }, + { + "epoch": 1.574371205550737, + "grad_norm": 0.9636722803115845, + "learning_rate": 9.538662411141739e-06, + "loss": 0.0019, + "step": 14522 + }, + { + "epoch": 1.574479618386817, + "grad_norm": 4.280312538146973, + "learning_rate": 9.537937037574351e-06, + "loss": 0.0139, + "step": 14523 + }, + { + "epoch": 1.5745880312228968, + "grad_norm": 0.7873165607452393, + "learning_rate": 9.537211664006965e-06, + "loss": 0.0025, + "step": 14524 + }, + { + "epoch": 1.5746964440589766, + "grad_norm": 3.5056281089782715, + "learning_rate": 9.536486290439577e-06, + "loss": 0.0264, + "step": 14525 + }, + { + "epoch": 1.5748048568950563, + "grad_norm": 16.575130462646484, + "learning_rate": 9.53576091687219e-06, + "loss": 0.0726, + "step": 14526 + }, + { + "epoch": 1.574913269731136, + "grad_norm": 2.4092090129852295, + "learning_rate": 9.535035543304803e-06, + "loss": 0.0249, + "step": 14527 + }, + { + "epoch": 1.575021682567216, + "grad_norm": 2.639835834503174, + "learning_rate": 9.534310169737415e-06, + "loss": 0.0092, + "step": 14528 + }, + { + "epoch": 1.5751300954032956, + "grad_norm": 1.1490963697433472, + "learning_rate": 9.533584796170027e-06, + "loss": 0.0285, + "step": 14529 + }, + { + "epoch": 1.5752385082393756, + "grad_norm": 0.508091151714325, + "learning_rate": 9.532859422602641e-06, + "loss": 0.003, + "step": 14530 + }, + { + "epoch": 1.5753469210754554, + "grad_norm": 1.8406394720077515, + "learning_rate": 9.532134049035255e-06, + "loss": 0.0805, + "step": 14531 + }, + { + "epoch": 1.5754553339115351, + "grad_norm": 0.23549580574035645, + "learning_rate": 9.531408675467867e-06, + "loss": 0.0009, + "step": 14532 + }, + { + "epoch": 1.575563746747615, + "grad_norm": 7.154655456542969, + "learning_rate": 9.53068330190048e-06, + "loss": 0.0354, + "step": 14533 + }, + { + "epoch": 1.5756721595836947, + "grad_norm": 0.895782470703125, + "learning_rate": 9.529957928333092e-06, + "loss": 0.0083, + "step": 14534 + }, + { + "epoch": 1.5757805724197746, + "grad_norm": 0.28923299908638, + "learning_rate": 9.529232554765705e-06, + "loss": 0.0036, + "step": 14535 + }, + { + "epoch": 1.5758889852558542, + "grad_norm": 2.3425378799438477, + "learning_rate": 9.528507181198318e-06, + "loss": 0.0291, + "step": 14536 + }, + { + "epoch": 1.5759973980919342, + "grad_norm": 0.3183790445327759, + "learning_rate": 9.52778180763093e-06, + "loss": 0.0023, + "step": 14537 + }, + { + "epoch": 1.5761058109280137, + "grad_norm": 0.11509977281093597, + "learning_rate": 9.527056434063544e-06, + "loss": 0.001, + "step": 14538 + }, + { + "epoch": 1.5762142237640937, + "grad_norm": 5.219701766967773, + "learning_rate": 9.526331060496157e-06, + "loss": 0.0394, + "step": 14539 + }, + { + "epoch": 1.5763226366001735, + "grad_norm": 0.8343836069107056, + "learning_rate": 9.52560568692877e-06, + "loss": 0.0032, + "step": 14540 + }, + { + "epoch": 1.5764310494362532, + "grad_norm": 6.7560906410217285, + "learning_rate": 9.524880313361382e-06, + "loss": 0.0439, + "step": 14541 + }, + { + "epoch": 1.576539462272333, + "grad_norm": 3.264488458633423, + "learning_rate": 9.524154939793995e-06, + "loss": 0.0105, + "step": 14542 + }, + { + "epoch": 1.5766478751084128, + "grad_norm": 0.12847256660461426, + "learning_rate": 9.523429566226608e-06, + "loss": 0.0006, + "step": 14543 + }, + { + "epoch": 1.5767562879444927, + "grad_norm": 4.800922870635986, + "learning_rate": 9.52270419265922e-06, + "loss": 0.0336, + "step": 14544 + }, + { + "epoch": 1.5768647007805723, + "grad_norm": 2.546067953109741, + "learning_rate": 9.521978819091832e-06, + "loss": 0.0255, + "step": 14545 + }, + { + "epoch": 1.5769731136166523, + "grad_norm": 1.5612870454788208, + "learning_rate": 9.521253445524446e-06, + "loss": 0.0341, + "step": 14546 + }, + { + "epoch": 1.577081526452732, + "grad_norm": 0.565229058265686, + "learning_rate": 9.52052807195706e-06, + "loss": 0.003, + "step": 14547 + }, + { + "epoch": 1.5771899392888118, + "grad_norm": 3.962115526199341, + "learning_rate": 9.519802698389672e-06, + "loss": 0.0311, + "step": 14548 + }, + { + "epoch": 1.5772983521248916, + "grad_norm": 0.2634356617927551, + "learning_rate": 9.519077324822284e-06, + "loss": 0.0021, + "step": 14549 + }, + { + "epoch": 1.5774067649609713, + "grad_norm": 0.10006627440452576, + "learning_rate": 9.518351951254898e-06, + "loss": 0.0004, + "step": 14550 + }, + { + "epoch": 1.5775151777970513, + "grad_norm": 3.300541639328003, + "learning_rate": 9.51762657768751e-06, + "loss": 0.0088, + "step": 14551 + }, + { + "epoch": 1.5776235906331308, + "grad_norm": 2.082815408706665, + "learning_rate": 9.516901204120122e-06, + "loss": 0.0509, + "step": 14552 + }, + { + "epoch": 1.5777320034692108, + "grad_norm": 1.6594223976135254, + "learning_rate": 9.516175830552736e-06, + "loss": 0.0203, + "step": 14553 + }, + { + "epoch": 1.5778404163052906, + "grad_norm": 2.2990190982818604, + "learning_rate": 9.515450456985348e-06, + "loss": 0.0448, + "step": 14554 + }, + { + "epoch": 1.5779488291413704, + "grad_norm": 5.977524280548096, + "learning_rate": 9.51472508341796e-06, + "loss": 0.0239, + "step": 14555 + }, + { + "epoch": 1.5780572419774501, + "grad_norm": 3.6733312606811523, + "learning_rate": 9.513999709850574e-06, + "loss": 0.0198, + "step": 14556 + }, + { + "epoch": 1.5781656548135299, + "grad_norm": 0.16347065567970276, + "learning_rate": 9.513274336283188e-06, + "loss": 0.0013, + "step": 14557 + }, + { + "epoch": 1.5782740676496099, + "grad_norm": 2.3173272609710693, + "learning_rate": 9.5125489627158e-06, + "loss": 0.0138, + "step": 14558 + }, + { + "epoch": 1.5783824804856894, + "grad_norm": 2.0564324855804443, + "learning_rate": 9.511823589148412e-06, + "loss": 0.0414, + "step": 14559 + }, + { + "epoch": 1.5784908933217694, + "grad_norm": 0.3163565695285797, + "learning_rate": 9.511098215581024e-06, + "loss": 0.0038, + "step": 14560 + }, + { + "epoch": 1.578599306157849, + "grad_norm": 2.1828601360321045, + "learning_rate": 9.510372842013638e-06, + "loss": 0.003, + "step": 14561 + }, + { + "epoch": 1.578707718993929, + "grad_norm": 3.0824050903320312, + "learning_rate": 9.50964746844625e-06, + "loss": 0.0259, + "step": 14562 + }, + { + "epoch": 1.5788161318300087, + "grad_norm": 4.783581733703613, + "learning_rate": 9.508922094878862e-06, + "loss": 0.0079, + "step": 14563 + }, + { + "epoch": 1.5789245446660884, + "grad_norm": 2.949031114578247, + "learning_rate": 9.508196721311476e-06, + "loss": 0.0188, + "step": 14564 + }, + { + "epoch": 1.5790329575021682, + "grad_norm": 3.430037260055542, + "learning_rate": 9.50747134774409e-06, + "loss": 0.0318, + "step": 14565 + }, + { + "epoch": 1.579141370338248, + "grad_norm": 0.5980525612831116, + "learning_rate": 9.506745974176702e-06, + "loss": 0.0053, + "step": 14566 + }, + { + "epoch": 1.579249783174328, + "grad_norm": 0.5991058349609375, + "learning_rate": 9.506020600609314e-06, + "loss": 0.0029, + "step": 14567 + }, + { + "epoch": 1.5793581960104075, + "grad_norm": 1.26600980758667, + "learning_rate": 9.505295227041928e-06, + "loss": 0.0164, + "step": 14568 + }, + { + "epoch": 1.5794666088464875, + "grad_norm": 0.06766470521688461, + "learning_rate": 9.50456985347454e-06, + "loss": 0.0007, + "step": 14569 + }, + { + "epoch": 1.5795750216825672, + "grad_norm": 2.8285915851593018, + "learning_rate": 9.503844479907152e-06, + "loss": 0.0268, + "step": 14570 + }, + { + "epoch": 1.579683434518647, + "grad_norm": 1.525744915008545, + "learning_rate": 9.503119106339765e-06, + "loss": 0.0184, + "step": 14571 + }, + { + "epoch": 1.5797918473547268, + "grad_norm": 2.880523920059204, + "learning_rate": 9.502393732772378e-06, + "loss": 0.015, + "step": 14572 + }, + { + "epoch": 1.5799002601908065, + "grad_norm": 9.299827575683594, + "learning_rate": 9.501668359204992e-06, + "loss": 0.0262, + "step": 14573 + }, + { + "epoch": 1.5800086730268865, + "grad_norm": 0.23263221979141235, + "learning_rate": 9.500942985637604e-06, + "loss": 0.0017, + "step": 14574 + }, + { + "epoch": 1.580117085862966, + "grad_norm": 3.0698256492614746, + "learning_rate": 9.500217612070217e-06, + "loss": 0.0156, + "step": 14575 + }, + { + "epoch": 1.580225498699046, + "grad_norm": 1.2282633781433105, + "learning_rate": 9.49949223850283e-06, + "loss": 0.0035, + "step": 14576 + }, + { + "epoch": 1.5803339115351258, + "grad_norm": 4.766844749450684, + "learning_rate": 9.498766864935443e-06, + "loss": 0.0267, + "step": 14577 + }, + { + "epoch": 1.5804423243712056, + "grad_norm": 2.3589534759521484, + "learning_rate": 9.498041491368055e-06, + "loss": 0.045, + "step": 14578 + }, + { + "epoch": 1.5805507372072853, + "grad_norm": 3.408717393875122, + "learning_rate": 9.497316117800669e-06, + "loss": 0.0175, + "step": 14579 + }, + { + "epoch": 1.580659150043365, + "grad_norm": 0.3137449324131012, + "learning_rate": 9.49659074423328e-06, + "loss": 0.0024, + "step": 14580 + }, + { + "epoch": 1.580767562879445, + "grad_norm": 3.0813779830932617, + "learning_rate": 9.495865370665893e-06, + "loss": 0.0365, + "step": 14581 + }, + { + "epoch": 1.5808759757155246, + "grad_norm": 0.4747352600097656, + "learning_rate": 9.495139997098507e-06, + "loss": 0.0051, + "step": 14582 + }, + { + "epoch": 1.5809843885516046, + "grad_norm": 4.686669826507568, + "learning_rate": 9.49441462353112e-06, + "loss": 0.033, + "step": 14583 + }, + { + "epoch": 1.5810928013876842, + "grad_norm": 3.597675085067749, + "learning_rate": 9.493689249963733e-06, + "loss": 0.0323, + "step": 14584 + }, + { + "epoch": 1.5812012142237641, + "grad_norm": 2.533912181854248, + "learning_rate": 9.492963876396345e-06, + "loss": 0.026, + "step": 14585 + }, + { + "epoch": 1.581309627059844, + "grad_norm": 1.4313910007476807, + "learning_rate": 9.492238502828957e-06, + "loss": 0.0212, + "step": 14586 + }, + { + "epoch": 1.5814180398959237, + "grad_norm": 1.3335609436035156, + "learning_rate": 9.49151312926157e-06, + "loss": 0.0152, + "step": 14587 + }, + { + "epoch": 1.5815264527320034, + "grad_norm": 10.84238052368164, + "learning_rate": 9.490787755694183e-06, + "loss": 0.0214, + "step": 14588 + }, + { + "epoch": 1.5816348655680832, + "grad_norm": 0.6618934869766235, + "learning_rate": 9.490062382126795e-06, + "loss": 0.0037, + "step": 14589 + }, + { + "epoch": 1.5817432784041632, + "grad_norm": 0.2406158149242401, + "learning_rate": 9.489337008559409e-06, + "loss": 0.0012, + "step": 14590 + }, + { + "epoch": 1.5818516912402427, + "grad_norm": 4.241623401641846, + "learning_rate": 9.488611634992023e-06, + "loss": 0.0212, + "step": 14591 + }, + { + "epoch": 1.5819601040763227, + "grad_norm": 4.670975685119629, + "learning_rate": 9.487886261424635e-06, + "loss": 0.0338, + "step": 14592 + }, + { + "epoch": 1.5820685169124025, + "grad_norm": 5.93398904800415, + "learning_rate": 9.487160887857247e-06, + "loss": 0.0273, + "step": 14593 + }, + { + "epoch": 1.5821769297484822, + "grad_norm": 1.1370583772659302, + "learning_rate": 9.48643551428986e-06, + "loss": 0.0235, + "step": 14594 + }, + { + "epoch": 1.582285342584562, + "grad_norm": 0.3026212155818939, + "learning_rate": 9.485710140722473e-06, + "loss": 0.0011, + "step": 14595 + }, + { + "epoch": 1.5823937554206418, + "grad_norm": 8.544779777526855, + "learning_rate": 9.484984767155085e-06, + "loss": 0.0257, + "step": 14596 + }, + { + "epoch": 1.5825021682567217, + "grad_norm": 4.548305988311768, + "learning_rate": 9.484259393587697e-06, + "loss": 0.013, + "step": 14597 + }, + { + "epoch": 1.5826105810928013, + "grad_norm": 2.1775035858154297, + "learning_rate": 9.483534020020311e-06, + "loss": 0.0369, + "step": 14598 + }, + { + "epoch": 1.5827189939288813, + "grad_norm": 4.614365577697754, + "learning_rate": 9.482808646452925e-06, + "loss": 0.0132, + "step": 14599 + }, + { + "epoch": 1.5828274067649608, + "grad_norm": 0.13536624610424042, + "learning_rate": 9.482083272885537e-06, + "loss": 0.0005, + "step": 14600 + }, + { + "epoch": 1.5829358196010408, + "grad_norm": 3.893207311630249, + "learning_rate": 9.48135789931815e-06, + "loss": 0.0164, + "step": 14601 + }, + { + "epoch": 1.5830442324371206, + "grad_norm": 12.166499137878418, + "learning_rate": 9.480632525750763e-06, + "loss": 0.0575, + "step": 14602 + }, + { + "epoch": 1.5831526452732003, + "grad_norm": 13.013813972473145, + "learning_rate": 9.479907152183375e-06, + "loss": 0.0716, + "step": 14603 + }, + { + "epoch": 1.58326105810928, + "grad_norm": 0.27794045209884644, + "learning_rate": 9.479181778615987e-06, + "loss": 0.0014, + "step": 14604 + }, + { + "epoch": 1.5833694709453598, + "grad_norm": 0.3964592218399048, + "learning_rate": 9.478456405048601e-06, + "loss": 0.0033, + "step": 14605 + }, + { + "epoch": 1.5834778837814398, + "grad_norm": 3.1000843048095703, + "learning_rate": 9.477731031481213e-06, + "loss": 0.033, + "step": 14606 + }, + { + "epoch": 1.5835862966175194, + "grad_norm": 5.236990928649902, + "learning_rate": 9.477005657913825e-06, + "loss": 0.0163, + "step": 14607 + }, + { + "epoch": 1.5836947094535994, + "grad_norm": 2.6009695529937744, + "learning_rate": 9.47628028434644e-06, + "loss": 0.0083, + "step": 14608 + }, + { + "epoch": 1.5838031222896791, + "grad_norm": 0.33155158162117004, + "learning_rate": 9.475554910779053e-06, + "loss": 0.0012, + "step": 14609 + }, + { + "epoch": 1.5839115351257589, + "grad_norm": 0.5641118288040161, + "learning_rate": 9.474829537211665e-06, + "loss": 0.0028, + "step": 14610 + }, + { + "epoch": 1.5840199479618386, + "grad_norm": 2.1038713455200195, + "learning_rate": 9.474104163644277e-06, + "loss": 0.023, + "step": 14611 + }, + { + "epoch": 1.5841283607979184, + "grad_norm": 0.3486868143081665, + "learning_rate": 9.47337879007689e-06, + "loss": 0.0016, + "step": 14612 + }, + { + "epoch": 1.5842367736339984, + "grad_norm": 0.9406729340553284, + "learning_rate": 9.472653416509503e-06, + "loss": 0.0078, + "step": 14613 + }, + { + "epoch": 1.584345186470078, + "grad_norm": 2.3407607078552246, + "learning_rate": 9.471928042942116e-06, + "loss": 0.0129, + "step": 14614 + }, + { + "epoch": 1.584453599306158, + "grad_norm": 0.6506680846214294, + "learning_rate": 9.471202669374728e-06, + "loss": 0.0025, + "step": 14615 + }, + { + "epoch": 1.5845620121422377, + "grad_norm": 1.6474387645721436, + "learning_rate": 9.470477295807342e-06, + "loss": 0.0041, + "step": 14616 + }, + { + "epoch": 1.5846704249783174, + "grad_norm": 0.8728874325752258, + "learning_rate": 9.469751922239955e-06, + "loss": 0.0041, + "step": 14617 + }, + { + "epoch": 1.5847788378143972, + "grad_norm": 1.3332992792129517, + "learning_rate": 9.469026548672568e-06, + "loss": 0.0161, + "step": 14618 + }, + { + "epoch": 1.584887250650477, + "grad_norm": 0.688831627368927, + "learning_rate": 9.46830117510518e-06, + "loss": 0.0028, + "step": 14619 + }, + { + "epoch": 1.584995663486557, + "grad_norm": 1.6602528095245361, + "learning_rate": 9.467575801537793e-06, + "loss": 0.0142, + "step": 14620 + }, + { + "epoch": 1.5851040763226365, + "grad_norm": 0.4302518665790558, + "learning_rate": 9.466850427970406e-06, + "loss": 0.0032, + "step": 14621 + }, + { + "epoch": 1.5852124891587165, + "grad_norm": 5.446366310119629, + "learning_rate": 9.466125054403018e-06, + "loss": 0.0237, + "step": 14622 + }, + { + "epoch": 1.585320901994796, + "grad_norm": 1.3642239570617676, + "learning_rate": 9.465399680835632e-06, + "loss": 0.0239, + "step": 14623 + }, + { + "epoch": 1.585429314830876, + "grad_norm": 0.7335996627807617, + "learning_rate": 9.464674307268244e-06, + "loss": 0.0033, + "step": 14624 + }, + { + "epoch": 1.5855377276669558, + "grad_norm": 2.0995140075683594, + "learning_rate": 9.463948933700858e-06, + "loss": 0.0272, + "step": 14625 + }, + { + "epoch": 1.5856461405030355, + "grad_norm": 1.084208607673645, + "learning_rate": 9.46322356013347e-06, + "loss": 0.0083, + "step": 14626 + }, + { + "epoch": 1.5857545533391153, + "grad_norm": 2.688612699508667, + "learning_rate": 9.462498186566082e-06, + "loss": 0.0219, + "step": 14627 + }, + { + "epoch": 1.585862966175195, + "grad_norm": 3.746901035308838, + "learning_rate": 9.461772812998696e-06, + "loss": 0.0339, + "step": 14628 + }, + { + "epoch": 1.585971379011275, + "grad_norm": 4.087541580200195, + "learning_rate": 9.461047439431308e-06, + "loss": 0.0059, + "step": 14629 + }, + { + "epoch": 1.5860797918473546, + "grad_norm": 2.534181833267212, + "learning_rate": 9.46032206586392e-06, + "loss": 0.0237, + "step": 14630 + }, + { + "epoch": 1.5861882046834346, + "grad_norm": 9.16415786743164, + "learning_rate": 9.459596692296534e-06, + "loss": 0.0347, + "step": 14631 + }, + { + "epoch": 1.5862966175195143, + "grad_norm": 0.033297453075647354, + "learning_rate": 9.458871318729146e-06, + "loss": 0.0002, + "step": 14632 + }, + { + "epoch": 1.586405030355594, + "grad_norm": 8.215816497802734, + "learning_rate": 9.458145945161758e-06, + "loss": 0.0382, + "step": 14633 + }, + { + "epoch": 1.5865134431916739, + "grad_norm": 2.6937096118927, + "learning_rate": 9.457420571594372e-06, + "loss": 0.0432, + "step": 14634 + }, + { + "epoch": 1.5866218560277536, + "grad_norm": 4.700187683105469, + "learning_rate": 9.456695198026986e-06, + "loss": 0.0079, + "step": 14635 + }, + { + "epoch": 1.5867302688638336, + "grad_norm": 8.425793647766113, + "learning_rate": 9.455969824459598e-06, + "loss": 0.0369, + "step": 14636 + }, + { + "epoch": 1.5868386816999132, + "grad_norm": 2.1432294845581055, + "learning_rate": 9.45524445089221e-06, + "loss": 0.0214, + "step": 14637 + }, + { + "epoch": 1.5869470945359931, + "grad_norm": 2.4670450687408447, + "learning_rate": 9.454519077324822e-06, + "loss": 0.0181, + "step": 14638 + }, + { + "epoch": 1.587055507372073, + "grad_norm": 8.559819221496582, + "learning_rate": 9.453793703757436e-06, + "loss": 0.0181, + "step": 14639 + }, + { + "epoch": 1.5871639202081527, + "grad_norm": 4.8027215003967285, + "learning_rate": 9.453068330190048e-06, + "loss": 0.0465, + "step": 14640 + }, + { + "epoch": 1.5872723330442324, + "grad_norm": 20.286670684814453, + "learning_rate": 9.45234295662266e-06, + "loss": 0.0513, + "step": 14641 + }, + { + "epoch": 1.5873807458803122, + "grad_norm": 2.152989625930786, + "learning_rate": 9.451617583055274e-06, + "loss": 0.0026, + "step": 14642 + }, + { + "epoch": 1.5874891587163922, + "grad_norm": 4.818892002105713, + "learning_rate": 9.450892209487888e-06, + "loss": 0.0431, + "step": 14643 + }, + { + "epoch": 1.5875975715524717, + "grad_norm": 0.5342472791671753, + "learning_rate": 9.4501668359205e-06, + "loss": 0.0047, + "step": 14644 + }, + { + "epoch": 1.5877059843885517, + "grad_norm": 1.0106335878372192, + "learning_rate": 9.449441462353112e-06, + "loss": 0.0034, + "step": 14645 + }, + { + "epoch": 1.5878143972246312, + "grad_norm": 0.6190052032470703, + "learning_rate": 9.448716088785726e-06, + "loss": 0.0019, + "step": 14646 + }, + { + "epoch": 1.5879228100607112, + "grad_norm": 0.923407256603241, + "learning_rate": 9.447990715218338e-06, + "loss": 0.0407, + "step": 14647 + }, + { + "epoch": 1.588031222896791, + "grad_norm": 23.887401580810547, + "learning_rate": 9.44726534165095e-06, + "loss": 0.0194, + "step": 14648 + }, + { + "epoch": 1.5881396357328708, + "grad_norm": 0.9363796710968018, + "learning_rate": 9.446539968083564e-06, + "loss": 0.0187, + "step": 14649 + }, + { + "epoch": 1.5882480485689505, + "grad_norm": 6.089346408843994, + "learning_rate": 9.445814594516176e-06, + "loss": 0.0483, + "step": 14650 + }, + { + "epoch": 1.5883564614050303, + "grad_norm": 2.126939058303833, + "learning_rate": 9.44508922094879e-06, + "loss": 0.0043, + "step": 14651 + }, + { + "epoch": 1.5884648742411103, + "grad_norm": 3.5218334197998047, + "learning_rate": 9.444363847381402e-06, + "loss": 0.0196, + "step": 14652 + }, + { + "epoch": 1.5885732870771898, + "grad_norm": 0.3688805401325226, + "learning_rate": 9.443638473814015e-06, + "loss": 0.0022, + "step": 14653 + }, + { + "epoch": 1.5886816999132698, + "grad_norm": 13.237468719482422, + "learning_rate": 9.442913100246628e-06, + "loss": 0.016, + "step": 14654 + }, + { + "epoch": 1.5887901127493496, + "grad_norm": 1.4763675928115845, + "learning_rate": 9.44218772667924e-06, + "loss": 0.0516, + "step": 14655 + }, + { + "epoch": 1.5888985255854293, + "grad_norm": 11.366007804870605, + "learning_rate": 9.441462353111853e-06, + "loss": 0.0169, + "step": 14656 + }, + { + "epoch": 1.589006938421509, + "grad_norm": 2.2267746925354004, + "learning_rate": 9.440736979544467e-06, + "loss": 0.0115, + "step": 14657 + }, + { + "epoch": 1.5891153512575888, + "grad_norm": 0.9983415007591248, + "learning_rate": 9.440011605977079e-06, + "loss": 0.0072, + "step": 14658 + }, + { + "epoch": 1.5892237640936688, + "grad_norm": 2.5969362258911133, + "learning_rate": 9.43928623240969e-06, + "loss": 0.0179, + "step": 14659 + }, + { + "epoch": 1.5893321769297484, + "grad_norm": 0.7600421905517578, + "learning_rate": 9.438560858842305e-06, + "loss": 0.0067, + "step": 14660 + }, + { + "epoch": 1.5894405897658284, + "grad_norm": 12.508604049682617, + "learning_rate": 9.437835485274918e-06, + "loss": 0.0231, + "step": 14661 + }, + { + "epoch": 1.589549002601908, + "grad_norm": 7.660856246948242, + "learning_rate": 9.43711011170753e-06, + "loss": 0.0501, + "step": 14662 + }, + { + "epoch": 1.5896574154379879, + "grad_norm": 4.772892951965332, + "learning_rate": 9.436384738140143e-06, + "loss": 0.0239, + "step": 14663 + }, + { + "epoch": 1.5897658282740676, + "grad_norm": 0.15383361279964447, + "learning_rate": 9.435659364572755e-06, + "loss": 0.0017, + "step": 14664 + }, + { + "epoch": 1.5898742411101474, + "grad_norm": 3.321254253387451, + "learning_rate": 9.434933991005369e-06, + "loss": 0.0159, + "step": 14665 + }, + { + "epoch": 1.5899826539462274, + "grad_norm": 0.7878085374832153, + "learning_rate": 9.434208617437981e-06, + "loss": 0.0078, + "step": 14666 + }, + { + "epoch": 1.590091066782307, + "grad_norm": 0.5518500208854675, + "learning_rate": 9.433483243870593e-06, + "loss": 0.0044, + "step": 14667 + }, + { + "epoch": 1.590199479618387, + "grad_norm": 5.354935169219971, + "learning_rate": 9.432757870303207e-06, + "loss": 0.0543, + "step": 14668 + }, + { + "epoch": 1.5903078924544665, + "grad_norm": 1.707360863685608, + "learning_rate": 9.43203249673582e-06, + "loss": 0.0079, + "step": 14669 + }, + { + "epoch": 1.5904163052905465, + "grad_norm": 0.687987208366394, + "learning_rate": 9.431307123168433e-06, + "loss": 0.0119, + "step": 14670 + }, + { + "epoch": 1.5905247181266262, + "grad_norm": 1.1094714403152466, + "learning_rate": 9.430581749601045e-06, + "loss": 0.0263, + "step": 14671 + }, + { + "epoch": 1.590633130962706, + "grad_norm": 0.048851512372493744, + "learning_rate": 9.429856376033659e-06, + "loss": 0.0007, + "step": 14672 + }, + { + "epoch": 1.5907415437987857, + "grad_norm": 3.501941442489624, + "learning_rate": 9.429131002466271e-06, + "loss": 0.0253, + "step": 14673 + }, + { + "epoch": 1.5908499566348655, + "grad_norm": 0.40793436765670776, + "learning_rate": 9.428405628898883e-06, + "loss": 0.0038, + "step": 14674 + }, + { + "epoch": 1.5909583694709455, + "grad_norm": 1.2962453365325928, + "learning_rate": 9.427680255331497e-06, + "loss": 0.005, + "step": 14675 + }, + { + "epoch": 1.591066782307025, + "grad_norm": 1.4594165086746216, + "learning_rate": 9.426954881764109e-06, + "loss": 0.0139, + "step": 14676 + }, + { + "epoch": 1.591175195143105, + "grad_norm": 6.831852912902832, + "learning_rate": 9.426229508196723e-06, + "loss": 0.0088, + "step": 14677 + }, + { + "epoch": 1.5912836079791848, + "grad_norm": 5.684202671051025, + "learning_rate": 9.425504134629335e-06, + "loss": 0.0335, + "step": 14678 + }, + { + "epoch": 1.5913920208152645, + "grad_norm": 1.9725476503372192, + "learning_rate": 9.424778761061947e-06, + "loss": 0.0215, + "step": 14679 + }, + { + "epoch": 1.5915004336513443, + "grad_norm": 0.7396995425224304, + "learning_rate": 9.424053387494561e-06, + "loss": 0.0051, + "step": 14680 + }, + { + "epoch": 1.591608846487424, + "grad_norm": 2.489438772201538, + "learning_rate": 9.423328013927173e-06, + "loss": 0.007, + "step": 14681 + }, + { + "epoch": 1.591717259323504, + "grad_norm": 3.701439380645752, + "learning_rate": 9.422602640359785e-06, + "loss": 0.011, + "step": 14682 + }, + { + "epoch": 1.5918256721595836, + "grad_norm": 5.965895652770996, + "learning_rate": 9.4218772667924e-06, + "loss": 0.0399, + "step": 14683 + }, + { + "epoch": 1.5919340849956636, + "grad_norm": 5.328514099121094, + "learning_rate": 9.421151893225011e-06, + "loss": 0.0299, + "step": 14684 + }, + { + "epoch": 1.5920424978317431, + "grad_norm": 1.0018463134765625, + "learning_rate": 9.420426519657623e-06, + "loss": 0.003, + "step": 14685 + }, + { + "epoch": 1.592150910667823, + "grad_norm": 6.556206226348877, + "learning_rate": 9.419701146090237e-06, + "loss": 0.0377, + "step": 14686 + }, + { + "epoch": 1.5922593235039029, + "grad_norm": 1.8415881395339966, + "learning_rate": 9.418975772522851e-06, + "loss": 0.0626, + "step": 14687 + }, + { + "epoch": 1.5923677363399826, + "grad_norm": 1.1375446319580078, + "learning_rate": 9.418250398955463e-06, + "loss": 0.0071, + "step": 14688 + }, + { + "epoch": 1.5924761491760624, + "grad_norm": 5.990142822265625, + "learning_rate": 9.417525025388075e-06, + "loss": 0.0223, + "step": 14689 + }, + { + "epoch": 1.5925845620121422, + "grad_norm": 1.6366322040557861, + "learning_rate": 9.416799651820688e-06, + "loss": 0.0215, + "step": 14690 + }, + { + "epoch": 1.5926929748482221, + "grad_norm": 1.0520448684692383, + "learning_rate": 9.416074278253301e-06, + "loss": 0.0456, + "step": 14691 + }, + { + "epoch": 1.5928013876843017, + "grad_norm": 3.637686014175415, + "learning_rate": 9.415348904685914e-06, + "loss": 0.0127, + "step": 14692 + }, + { + "epoch": 1.5929098005203817, + "grad_norm": 1.5685813426971436, + "learning_rate": 9.414623531118526e-06, + "loss": 0.0137, + "step": 14693 + }, + { + "epoch": 1.5930182133564614, + "grad_norm": 4.001931667327881, + "learning_rate": 9.41389815755114e-06, + "loss": 0.0072, + "step": 14694 + }, + { + "epoch": 1.5931266261925412, + "grad_norm": 1.4603869915008545, + "learning_rate": 9.413172783983753e-06, + "loss": 0.0103, + "step": 14695 + }, + { + "epoch": 1.593235039028621, + "grad_norm": 0.21795028448104858, + "learning_rate": 9.412447410416366e-06, + "loss": 0.0036, + "step": 14696 + }, + { + "epoch": 1.5933434518647007, + "grad_norm": 3.088460683822632, + "learning_rate": 9.411722036848978e-06, + "loss": 0.0647, + "step": 14697 + }, + { + "epoch": 1.5934518647007807, + "grad_norm": 1.719025731086731, + "learning_rate": 9.410996663281591e-06, + "loss": 0.0065, + "step": 14698 + }, + { + "epoch": 1.5935602775368602, + "grad_norm": 4.430068016052246, + "learning_rate": 9.410271289714204e-06, + "loss": 0.0148, + "step": 14699 + }, + { + "epoch": 1.5936686903729402, + "grad_norm": 2.1863253116607666, + "learning_rate": 9.409545916146816e-06, + "loss": 0.0116, + "step": 14700 + }, + { + "epoch": 1.59377710320902, + "grad_norm": 0.8278784155845642, + "learning_rate": 9.40882054257943e-06, + "loss": 0.006, + "step": 14701 + }, + { + "epoch": 1.5938855160450998, + "grad_norm": 0.4612109661102295, + "learning_rate": 9.408095169012042e-06, + "loss": 0.0019, + "step": 14702 + }, + { + "epoch": 1.5939939288811795, + "grad_norm": 1.482728123664856, + "learning_rate": 9.407369795444656e-06, + "loss": 0.0061, + "step": 14703 + }, + { + "epoch": 1.5941023417172593, + "grad_norm": 1.2970808744430542, + "learning_rate": 9.406644421877268e-06, + "loss": 0.0108, + "step": 14704 + }, + { + "epoch": 1.5942107545533393, + "grad_norm": 1.9601426124572754, + "learning_rate": 9.40591904830988e-06, + "loss": 0.0052, + "step": 14705 + }, + { + "epoch": 1.5943191673894188, + "grad_norm": 1.3904856443405151, + "learning_rate": 9.405193674742494e-06, + "loss": 0.0192, + "step": 14706 + }, + { + "epoch": 1.5944275802254988, + "grad_norm": 0.708419680595398, + "learning_rate": 9.404468301175106e-06, + "loss": 0.007, + "step": 14707 + }, + { + "epoch": 1.5945359930615783, + "grad_norm": 0.5119029879570007, + "learning_rate": 9.403742927607718e-06, + "loss": 0.0038, + "step": 14708 + }, + { + "epoch": 1.5946444058976583, + "grad_norm": 4.802976608276367, + "learning_rate": 9.403017554040332e-06, + "loss": 0.0798, + "step": 14709 + }, + { + "epoch": 1.594752818733738, + "grad_norm": 3.5178613662719727, + "learning_rate": 9.402292180472944e-06, + "loss": 0.0969, + "step": 14710 + }, + { + "epoch": 1.5948612315698178, + "grad_norm": 0.41129758954048157, + "learning_rate": 9.401566806905558e-06, + "loss": 0.0023, + "step": 14711 + }, + { + "epoch": 1.5949696444058976, + "grad_norm": 1.1904945373535156, + "learning_rate": 9.40084143333817e-06, + "loss": 0.0126, + "step": 14712 + }, + { + "epoch": 1.5950780572419774, + "grad_norm": 0.05303378775715828, + "learning_rate": 9.400116059770784e-06, + "loss": 0.0006, + "step": 14713 + }, + { + "epoch": 1.5951864700780574, + "grad_norm": 3.672483444213867, + "learning_rate": 9.399390686203396e-06, + "loss": 0.0187, + "step": 14714 + }, + { + "epoch": 1.595294882914137, + "grad_norm": 9.796804428100586, + "learning_rate": 9.398665312636008e-06, + "loss": 0.0289, + "step": 14715 + }, + { + "epoch": 1.5954032957502169, + "grad_norm": 0.7007302641868591, + "learning_rate": 9.39793993906862e-06, + "loss": 0.0078, + "step": 14716 + }, + { + "epoch": 1.5955117085862967, + "grad_norm": 0.19859108328819275, + "learning_rate": 9.397214565501234e-06, + "loss": 0.001, + "step": 14717 + }, + { + "epoch": 1.5956201214223764, + "grad_norm": 3.8678760528564453, + "learning_rate": 9.396489191933846e-06, + "loss": 0.007, + "step": 14718 + }, + { + "epoch": 1.5957285342584562, + "grad_norm": 3.096461296081543, + "learning_rate": 9.395763818366458e-06, + "loss": 0.0199, + "step": 14719 + }, + { + "epoch": 1.595836947094536, + "grad_norm": 3.402552604675293, + "learning_rate": 9.395038444799072e-06, + "loss": 0.0448, + "step": 14720 + }, + { + "epoch": 1.595945359930616, + "grad_norm": 0.7141858339309692, + "learning_rate": 9.394313071231686e-06, + "loss": 0.0119, + "step": 14721 + }, + { + "epoch": 1.5960537727666955, + "grad_norm": 4.477746486663818, + "learning_rate": 9.393587697664298e-06, + "loss": 0.019, + "step": 14722 + }, + { + "epoch": 1.5961621856027755, + "grad_norm": 2.4738550186157227, + "learning_rate": 9.39286232409691e-06, + "loss": 0.0257, + "step": 14723 + }, + { + "epoch": 1.5962705984388552, + "grad_norm": 5.602283477783203, + "learning_rate": 9.392136950529524e-06, + "loss": 0.0159, + "step": 14724 + }, + { + "epoch": 1.596379011274935, + "grad_norm": 2.501607656478882, + "learning_rate": 9.391411576962136e-06, + "loss": 0.0178, + "step": 14725 + }, + { + "epoch": 1.5964874241110147, + "grad_norm": 4.531422138214111, + "learning_rate": 9.390686203394748e-06, + "loss": 0.0141, + "step": 14726 + }, + { + "epoch": 1.5965958369470945, + "grad_norm": 2.8051342964172363, + "learning_rate": 9.389960829827362e-06, + "loss": 0.0387, + "step": 14727 + }, + { + "epoch": 1.5967042497831745, + "grad_norm": 0.6132987141609192, + "learning_rate": 9.389235456259974e-06, + "loss": 0.0078, + "step": 14728 + }, + { + "epoch": 1.596812662619254, + "grad_norm": 1.1183208227157593, + "learning_rate": 9.388510082692588e-06, + "loss": 0.0082, + "step": 14729 + }, + { + "epoch": 1.596921075455334, + "grad_norm": 2.0170950889587402, + "learning_rate": 9.3877847091252e-06, + "loss": 0.0608, + "step": 14730 + }, + { + "epoch": 1.5970294882914136, + "grad_norm": 1.0198677778244019, + "learning_rate": 9.387059335557813e-06, + "loss": 0.0032, + "step": 14731 + }, + { + "epoch": 1.5971379011274935, + "grad_norm": 1.2601426839828491, + "learning_rate": 9.386333961990426e-06, + "loss": 0.0178, + "step": 14732 + }, + { + "epoch": 1.5972463139635733, + "grad_norm": 0.5600699782371521, + "learning_rate": 9.385608588423039e-06, + "loss": 0.0052, + "step": 14733 + }, + { + "epoch": 1.597354726799653, + "grad_norm": 3.5915911197662354, + "learning_rate": 9.38488321485565e-06, + "loss": 0.0273, + "step": 14734 + }, + { + "epoch": 1.5974631396357328, + "grad_norm": 7.852248191833496, + "learning_rate": 9.384157841288265e-06, + "loss": 0.0743, + "step": 14735 + }, + { + "epoch": 1.5975715524718126, + "grad_norm": 1.9270087480545044, + "learning_rate": 9.383432467720877e-06, + "loss": 0.0037, + "step": 14736 + }, + { + "epoch": 1.5976799653078926, + "grad_norm": 0.315461665391922, + "learning_rate": 9.38270709415349e-06, + "loss": 0.0014, + "step": 14737 + }, + { + "epoch": 1.5977883781439721, + "grad_norm": 2.406597137451172, + "learning_rate": 9.381981720586103e-06, + "loss": 0.0072, + "step": 14738 + }, + { + "epoch": 1.597896790980052, + "grad_norm": 6.293064594268799, + "learning_rate": 9.381256347018716e-06, + "loss": 0.0445, + "step": 14739 + }, + { + "epoch": 1.5980052038161319, + "grad_norm": 0.23088951408863068, + "learning_rate": 9.380530973451329e-06, + "loss": 0.0033, + "step": 14740 + }, + { + "epoch": 1.5981136166522116, + "grad_norm": 10.611806869506836, + "learning_rate": 9.37980559988394e-06, + "loss": 0.0331, + "step": 14741 + }, + { + "epoch": 1.5982220294882914, + "grad_norm": 1.4293994903564453, + "learning_rate": 9.379080226316553e-06, + "loss": 0.025, + "step": 14742 + }, + { + "epoch": 1.5983304423243712, + "grad_norm": 5.221975803375244, + "learning_rate": 9.378354852749167e-06, + "loss": 0.0508, + "step": 14743 + }, + { + "epoch": 1.5984388551604511, + "grad_norm": 2.952913522720337, + "learning_rate": 9.377629479181779e-06, + "loss": 0.0801, + "step": 14744 + }, + { + "epoch": 1.5985472679965307, + "grad_norm": 4.861104965209961, + "learning_rate": 9.376904105614391e-06, + "loss": 0.0461, + "step": 14745 + }, + { + "epoch": 1.5986556808326107, + "grad_norm": 2.7464804649353027, + "learning_rate": 9.376178732047005e-06, + "loss": 0.0339, + "step": 14746 + }, + { + "epoch": 1.5987640936686902, + "grad_norm": 6.453039169311523, + "learning_rate": 9.375453358479619e-06, + "loss": 0.0198, + "step": 14747 + }, + { + "epoch": 1.5988725065047702, + "grad_norm": 1.5486620664596558, + "learning_rate": 9.374727984912231e-06, + "loss": 0.0059, + "step": 14748 + }, + { + "epoch": 1.59898091934085, + "grad_norm": 5.934993267059326, + "learning_rate": 9.374002611344843e-06, + "loss": 0.0192, + "step": 14749 + }, + { + "epoch": 1.5990893321769297, + "grad_norm": 5.03770112991333, + "learning_rate": 9.373277237777457e-06, + "loss": 0.0307, + "step": 14750 + }, + { + "epoch": 1.5991977450130095, + "grad_norm": 2.8567872047424316, + "learning_rate": 9.372551864210069e-06, + "loss": 0.0309, + "step": 14751 + }, + { + "epoch": 1.5993061578490892, + "grad_norm": 0.22777962684631348, + "learning_rate": 9.371826490642681e-06, + "loss": 0.0026, + "step": 14752 + }, + { + "epoch": 1.5994145706851692, + "grad_norm": 2.232069730758667, + "learning_rate": 9.371101117075295e-06, + "loss": 0.028, + "step": 14753 + }, + { + "epoch": 1.5995229835212488, + "grad_norm": 1.6493818759918213, + "learning_rate": 9.370375743507907e-06, + "loss": 0.0289, + "step": 14754 + }, + { + "epoch": 1.5996313963573288, + "grad_norm": 0.28386178612709045, + "learning_rate": 9.369650369940521e-06, + "loss": 0.0023, + "step": 14755 + }, + { + "epoch": 1.5997398091934085, + "grad_norm": 3.4049689769744873, + "learning_rate": 9.368924996373133e-06, + "loss": 0.046, + "step": 14756 + }, + { + "epoch": 1.5998482220294883, + "grad_norm": 2.506194829940796, + "learning_rate": 9.368199622805745e-06, + "loss": 0.032, + "step": 14757 + }, + { + "epoch": 1.599956634865568, + "grad_norm": 3.8525829315185547, + "learning_rate": 9.367474249238359e-06, + "loss": 0.045, + "step": 14758 + }, + { + "epoch": 1.6000650477016478, + "grad_norm": 3.5261220932006836, + "learning_rate": 9.366748875670971e-06, + "loss": 0.0094, + "step": 14759 + }, + { + "epoch": 1.6001734605377278, + "grad_norm": 0.35831403732299805, + "learning_rate": 9.366023502103583e-06, + "loss": 0.0032, + "step": 14760 + }, + { + "epoch": 1.6002818733738073, + "grad_norm": 5.186562538146973, + "learning_rate": 9.365298128536197e-06, + "loss": 0.0266, + "step": 14761 + }, + { + "epoch": 1.6003902862098873, + "grad_norm": 0.8228137493133545, + "learning_rate": 9.36457275496881e-06, + "loss": 0.0049, + "step": 14762 + }, + { + "epoch": 1.600498699045967, + "grad_norm": 0.6562598943710327, + "learning_rate": 9.363847381401423e-06, + "loss": 0.013, + "step": 14763 + }, + { + "epoch": 1.6006071118820469, + "grad_norm": 2.5841455459594727, + "learning_rate": 9.363122007834035e-06, + "loss": 0.0088, + "step": 14764 + }, + { + "epoch": 1.6007155247181266, + "grad_norm": 3.060535430908203, + "learning_rate": 9.362396634266649e-06, + "loss": 0.0384, + "step": 14765 + }, + { + "epoch": 1.6008239375542064, + "grad_norm": 0.7824493646621704, + "learning_rate": 9.361671260699261e-06, + "loss": 0.0229, + "step": 14766 + }, + { + "epoch": 1.6009323503902864, + "grad_norm": 2.9294097423553467, + "learning_rate": 9.360945887131873e-06, + "loss": 0.0471, + "step": 14767 + }, + { + "epoch": 1.601040763226366, + "grad_norm": 3.4778544902801514, + "learning_rate": 9.360220513564486e-06, + "loss": 0.0365, + "step": 14768 + }, + { + "epoch": 1.601149176062446, + "grad_norm": 0.34128883481025696, + "learning_rate": 9.3594951399971e-06, + "loss": 0.0036, + "step": 14769 + }, + { + "epoch": 1.6012575888985254, + "grad_norm": 2.3207433223724365, + "learning_rate": 9.358769766429712e-06, + "loss": 0.0072, + "step": 14770 + }, + { + "epoch": 1.6013660017346054, + "grad_norm": 2.8976705074310303, + "learning_rate": 9.358044392862324e-06, + "loss": 0.0248, + "step": 14771 + }, + { + "epoch": 1.6014744145706852, + "grad_norm": 3.0038106441497803, + "learning_rate": 9.357319019294938e-06, + "loss": 0.025, + "step": 14772 + }, + { + "epoch": 1.601582827406765, + "grad_norm": 4.790515422821045, + "learning_rate": 9.356593645727551e-06, + "loss": 0.0708, + "step": 14773 + }, + { + "epoch": 1.6016912402428447, + "grad_norm": 2.5017435550689697, + "learning_rate": 9.355868272160164e-06, + "loss": 0.0177, + "step": 14774 + }, + { + "epoch": 1.6017996530789245, + "grad_norm": 2.5165085792541504, + "learning_rate": 9.355142898592776e-06, + "loss": 0.0096, + "step": 14775 + }, + { + "epoch": 1.6019080659150045, + "grad_norm": 0.11264500766992569, + "learning_rate": 9.35441752502539e-06, + "loss": 0.001, + "step": 14776 + }, + { + "epoch": 1.602016478751084, + "grad_norm": 1.2443264722824097, + "learning_rate": 9.353692151458002e-06, + "loss": 0.0054, + "step": 14777 + }, + { + "epoch": 1.602124891587164, + "grad_norm": 2.746511220932007, + "learning_rate": 9.352966777890614e-06, + "loss": 0.0388, + "step": 14778 + }, + { + "epoch": 1.6022333044232437, + "grad_norm": 3.0207839012145996, + "learning_rate": 9.352241404323228e-06, + "loss": 0.012, + "step": 14779 + }, + { + "epoch": 1.6023417172593235, + "grad_norm": 1.7331448793411255, + "learning_rate": 9.35151603075584e-06, + "loss": 0.0066, + "step": 14780 + }, + { + "epoch": 1.6024501300954033, + "grad_norm": 2.3668935298919678, + "learning_rate": 9.350790657188454e-06, + "loss": 0.0105, + "step": 14781 + }, + { + "epoch": 1.602558542931483, + "grad_norm": 6.4238128662109375, + "learning_rate": 9.350065283621066e-06, + "loss": 0.009, + "step": 14782 + }, + { + "epoch": 1.602666955767563, + "grad_norm": 4.3598127365112305, + "learning_rate": 9.349339910053678e-06, + "loss": 0.0748, + "step": 14783 + }, + { + "epoch": 1.6027753686036426, + "grad_norm": 2.7100112438201904, + "learning_rate": 9.348614536486292e-06, + "loss": 0.0296, + "step": 14784 + }, + { + "epoch": 1.6028837814397225, + "grad_norm": 0.8473667502403259, + "learning_rate": 9.347889162918904e-06, + "loss": 0.0145, + "step": 14785 + }, + { + "epoch": 1.6029921942758023, + "grad_norm": 2.9353621006011963, + "learning_rate": 9.347163789351516e-06, + "loss": 0.0094, + "step": 14786 + }, + { + "epoch": 1.603100607111882, + "grad_norm": 1.9455091953277588, + "learning_rate": 9.34643841578413e-06, + "loss": 0.0299, + "step": 14787 + }, + { + "epoch": 1.6032090199479618, + "grad_norm": 2.1052849292755127, + "learning_rate": 9.345713042216742e-06, + "loss": 0.0395, + "step": 14788 + }, + { + "epoch": 1.6033174327840416, + "grad_norm": 9.619465827941895, + "learning_rate": 9.344987668649356e-06, + "loss": 0.0589, + "step": 14789 + }, + { + "epoch": 1.6034258456201216, + "grad_norm": 3.2032828330993652, + "learning_rate": 9.344262295081968e-06, + "loss": 0.0274, + "step": 14790 + }, + { + "epoch": 1.6035342584562011, + "grad_norm": 4.6165289878845215, + "learning_rate": 9.343536921514582e-06, + "loss": 0.0266, + "step": 14791 + }, + { + "epoch": 1.603642671292281, + "grad_norm": 4.504141330718994, + "learning_rate": 9.342811547947194e-06, + "loss": 0.0418, + "step": 14792 + }, + { + "epoch": 1.6037510841283606, + "grad_norm": 2.207275867462158, + "learning_rate": 9.342086174379806e-06, + "loss": 0.0083, + "step": 14793 + }, + { + "epoch": 1.6038594969644406, + "grad_norm": 1.561065435409546, + "learning_rate": 9.341360800812418e-06, + "loss": 0.0308, + "step": 14794 + }, + { + "epoch": 1.6039679098005204, + "grad_norm": 5.215311527252197, + "learning_rate": 9.340635427245032e-06, + "loss": 0.0331, + "step": 14795 + }, + { + "epoch": 1.6040763226366002, + "grad_norm": 3.259934663772583, + "learning_rate": 9.339910053677644e-06, + "loss": 0.0077, + "step": 14796 + }, + { + "epoch": 1.60418473547268, + "grad_norm": 4.165036678314209, + "learning_rate": 9.339184680110256e-06, + "loss": 0.0171, + "step": 14797 + }, + { + "epoch": 1.6042931483087597, + "grad_norm": 0.33736127614974976, + "learning_rate": 9.33845930654287e-06, + "loss": 0.0027, + "step": 14798 + }, + { + "epoch": 1.6044015611448397, + "grad_norm": 1.6762769222259521, + "learning_rate": 9.337733932975484e-06, + "loss": 0.0178, + "step": 14799 + }, + { + "epoch": 1.6045099739809192, + "grad_norm": 2.910247564315796, + "learning_rate": 9.337008559408096e-06, + "loss": 0.033, + "step": 14800 + }, + { + "epoch": 1.6046183868169992, + "grad_norm": 5.006712436676025, + "learning_rate": 9.336283185840708e-06, + "loss": 0.0374, + "step": 14801 + }, + { + "epoch": 1.604726799653079, + "grad_norm": 6.984081745147705, + "learning_rate": 9.335557812273322e-06, + "loss": 0.053, + "step": 14802 + }, + { + "epoch": 1.6048352124891587, + "grad_norm": 2.273949146270752, + "learning_rate": 9.334832438705934e-06, + "loss": 0.0331, + "step": 14803 + }, + { + "epoch": 1.6049436253252385, + "grad_norm": 10.971552848815918, + "learning_rate": 9.334107065138546e-06, + "loss": 0.0616, + "step": 14804 + }, + { + "epoch": 1.6050520381613183, + "grad_norm": 8.615446090698242, + "learning_rate": 9.33338169157116e-06, + "loss": 0.0258, + "step": 14805 + }, + { + "epoch": 1.6051604509973982, + "grad_norm": 3.2036168575286865, + "learning_rate": 9.332656318003772e-06, + "loss": 0.0391, + "step": 14806 + }, + { + "epoch": 1.6052688638334778, + "grad_norm": 2.5365936756134033, + "learning_rate": 9.331930944436386e-06, + "loss": 0.0483, + "step": 14807 + }, + { + "epoch": 1.6053772766695578, + "grad_norm": 1.5429877042770386, + "learning_rate": 9.331205570868998e-06, + "loss": 0.025, + "step": 14808 + }, + { + "epoch": 1.6054856895056373, + "grad_norm": 7.270644664764404, + "learning_rate": 9.33048019730161e-06, + "loss": 0.0479, + "step": 14809 + }, + { + "epoch": 1.6055941023417173, + "grad_norm": 1.2217859029769897, + "learning_rate": 9.329754823734224e-06, + "loss": 0.0146, + "step": 14810 + }, + { + "epoch": 1.605702515177797, + "grad_norm": 0.38934648036956787, + "learning_rate": 9.329029450166837e-06, + "loss": 0.0065, + "step": 14811 + }, + { + "epoch": 1.6058109280138768, + "grad_norm": 8.258853912353516, + "learning_rate": 9.328304076599449e-06, + "loss": 0.0231, + "step": 14812 + }, + { + "epoch": 1.6059193408499568, + "grad_norm": 2.6386377811431885, + "learning_rate": 9.327578703032063e-06, + "loss": 0.0505, + "step": 14813 + }, + { + "epoch": 1.6060277536860363, + "grad_norm": 0.4190223813056946, + "learning_rate": 9.326853329464675e-06, + "loss": 0.0039, + "step": 14814 + }, + { + "epoch": 1.6061361665221163, + "grad_norm": 2.3940110206604004, + "learning_rate": 9.326127955897289e-06, + "loss": 0.0426, + "step": 14815 + }, + { + "epoch": 1.6062445793581959, + "grad_norm": 0.5936149954795837, + "learning_rate": 9.3254025823299e-06, + "loss": 0.007, + "step": 14816 + }, + { + "epoch": 1.6063529921942759, + "grad_norm": 2.3810505867004395, + "learning_rate": 9.324677208762514e-06, + "loss": 0.0409, + "step": 14817 + }, + { + "epoch": 1.6064614050303556, + "grad_norm": 0.5734612345695496, + "learning_rate": 9.323951835195127e-06, + "loss": 0.0017, + "step": 14818 + }, + { + "epoch": 1.6065698178664354, + "grad_norm": 3.3754148483276367, + "learning_rate": 9.323226461627739e-06, + "loss": 0.0474, + "step": 14819 + }, + { + "epoch": 1.6066782307025151, + "grad_norm": 7.119283676147461, + "learning_rate": 9.322501088060351e-06, + "loss": 0.0137, + "step": 14820 + }, + { + "epoch": 1.606786643538595, + "grad_norm": 2.6052961349487305, + "learning_rate": 9.321775714492965e-06, + "loss": 0.0129, + "step": 14821 + }, + { + "epoch": 1.606895056374675, + "grad_norm": 3.5148942470550537, + "learning_rate": 9.321050340925577e-06, + "loss": 0.0415, + "step": 14822 + }, + { + "epoch": 1.6070034692107544, + "grad_norm": 1.7405242919921875, + "learning_rate": 9.320324967358189e-06, + "loss": 0.0208, + "step": 14823 + }, + { + "epoch": 1.6071118820468344, + "grad_norm": 4.409396171569824, + "learning_rate": 9.319599593790803e-06, + "loss": 0.0531, + "step": 14824 + }, + { + "epoch": 1.6072202948829142, + "grad_norm": 6.986933708190918, + "learning_rate": 9.318874220223417e-06, + "loss": 0.025, + "step": 14825 + }, + { + "epoch": 1.607328707718994, + "grad_norm": 2.509627342224121, + "learning_rate": 9.318148846656029e-06, + "loss": 0.0181, + "step": 14826 + }, + { + "epoch": 1.6074371205550737, + "grad_norm": 6.715106010437012, + "learning_rate": 9.317423473088641e-06, + "loss": 0.0407, + "step": 14827 + }, + { + "epoch": 1.6075455333911535, + "grad_norm": 0.9851181507110596, + "learning_rate": 9.316698099521255e-06, + "loss": 0.004, + "step": 14828 + }, + { + "epoch": 1.6076539462272335, + "grad_norm": 4.125412940979004, + "learning_rate": 9.315972725953867e-06, + "loss": 0.0574, + "step": 14829 + }, + { + "epoch": 1.607762359063313, + "grad_norm": 7.272389888763428, + "learning_rate": 9.315247352386479e-06, + "loss": 0.0666, + "step": 14830 + }, + { + "epoch": 1.607870771899393, + "grad_norm": 0.33037546277046204, + "learning_rate": 9.314521978819093e-06, + "loss": 0.0036, + "step": 14831 + }, + { + "epoch": 1.6079791847354725, + "grad_norm": 4.401565074920654, + "learning_rate": 9.313796605251705e-06, + "loss": 0.0444, + "step": 14832 + }, + { + "epoch": 1.6080875975715525, + "grad_norm": 0.1540597826242447, + "learning_rate": 9.313071231684319e-06, + "loss": 0.002, + "step": 14833 + }, + { + "epoch": 1.6081960104076323, + "grad_norm": 1.5158089399337769, + "learning_rate": 9.312345858116931e-06, + "loss": 0.0212, + "step": 14834 + }, + { + "epoch": 1.608304423243712, + "grad_norm": 4.746935844421387, + "learning_rate": 9.311620484549543e-06, + "loss": 0.0583, + "step": 14835 + }, + { + "epoch": 1.6084128360797918, + "grad_norm": 2.758153200149536, + "learning_rate": 9.310895110982157e-06, + "loss": 0.025, + "step": 14836 + }, + { + "epoch": 1.6085212489158716, + "grad_norm": 0.4374426603317261, + "learning_rate": 9.31016973741477e-06, + "loss": 0.0072, + "step": 14837 + }, + { + "epoch": 1.6086296617519515, + "grad_norm": 1.1661275625228882, + "learning_rate": 9.309444363847381e-06, + "loss": 0.0061, + "step": 14838 + }, + { + "epoch": 1.608738074588031, + "grad_norm": 1.342272400856018, + "learning_rate": 9.308718990279995e-06, + "loss": 0.006, + "step": 14839 + }, + { + "epoch": 1.608846487424111, + "grad_norm": 2.674816608428955, + "learning_rate": 9.307993616712607e-06, + "loss": 0.0076, + "step": 14840 + }, + { + "epoch": 1.6089549002601908, + "grad_norm": 2.944392442703247, + "learning_rate": 9.307268243145221e-06, + "loss": 0.0418, + "step": 14841 + }, + { + "epoch": 1.6090633130962706, + "grad_norm": 2.2836647033691406, + "learning_rate": 9.306542869577833e-06, + "loss": 0.0101, + "step": 14842 + }, + { + "epoch": 1.6091717259323504, + "grad_norm": 2.782036542892456, + "learning_rate": 9.305817496010447e-06, + "loss": 0.0445, + "step": 14843 + }, + { + "epoch": 1.6092801387684301, + "grad_norm": 1.5213350057601929, + "learning_rate": 9.30509212244306e-06, + "loss": 0.0185, + "step": 14844 + }, + { + "epoch": 1.60938855160451, + "grad_norm": 3.042219638824463, + "learning_rate": 9.304366748875671e-06, + "loss": 0.0378, + "step": 14845 + }, + { + "epoch": 1.6094969644405897, + "grad_norm": 0.7097801566123962, + "learning_rate": 9.303641375308284e-06, + "loss": 0.0079, + "step": 14846 + }, + { + "epoch": 1.6096053772766696, + "grad_norm": 0.7971309423446655, + "learning_rate": 9.302916001740897e-06, + "loss": 0.0038, + "step": 14847 + }, + { + "epoch": 1.6097137901127494, + "grad_norm": 0.14239010214805603, + "learning_rate": 9.30219062817351e-06, + "loss": 0.0014, + "step": 14848 + }, + { + "epoch": 1.6098222029488292, + "grad_norm": 3.892735481262207, + "learning_rate": 9.301465254606122e-06, + "loss": 0.0329, + "step": 14849 + }, + { + "epoch": 1.609930615784909, + "grad_norm": 5.821685791015625, + "learning_rate": 9.300739881038736e-06, + "loss": 0.0309, + "step": 14850 + }, + { + "epoch": 1.6100390286209887, + "grad_norm": 1.2328855991363525, + "learning_rate": 9.30001450747135e-06, + "loss": 0.015, + "step": 14851 + }, + { + "epoch": 1.6101474414570687, + "grad_norm": 2.414625406265259, + "learning_rate": 9.299289133903962e-06, + "loss": 0.038, + "step": 14852 + }, + { + "epoch": 1.6102558542931482, + "grad_norm": 0.6284454464912415, + "learning_rate": 9.298563760336574e-06, + "loss": 0.0045, + "step": 14853 + }, + { + "epoch": 1.6103642671292282, + "grad_norm": 0.7834033966064453, + "learning_rate": 9.297838386769188e-06, + "loss": 0.0034, + "step": 14854 + }, + { + "epoch": 1.6104726799653077, + "grad_norm": 3.1868739128112793, + "learning_rate": 9.2971130132018e-06, + "loss": 0.0103, + "step": 14855 + }, + { + "epoch": 1.6105810928013877, + "grad_norm": 3.0718846321105957, + "learning_rate": 9.296387639634412e-06, + "loss": 0.0288, + "step": 14856 + }, + { + "epoch": 1.6106895056374675, + "grad_norm": 1.6799342632293701, + "learning_rate": 9.295662266067026e-06, + "loss": 0.0368, + "step": 14857 + }, + { + "epoch": 1.6107979184735473, + "grad_norm": 3.490402936935425, + "learning_rate": 9.294936892499638e-06, + "loss": 0.0279, + "step": 14858 + }, + { + "epoch": 1.610906331309627, + "grad_norm": 0.3178503215312958, + "learning_rate": 9.294211518932252e-06, + "loss": 0.0017, + "step": 14859 + }, + { + "epoch": 1.6110147441457068, + "grad_norm": 6.478792667388916, + "learning_rate": 9.293486145364864e-06, + "loss": 0.0389, + "step": 14860 + }, + { + "epoch": 1.6111231569817868, + "grad_norm": 0.23293104767799377, + "learning_rate": 9.292760771797476e-06, + "loss": 0.0032, + "step": 14861 + }, + { + "epoch": 1.6112315698178663, + "grad_norm": 0.6174004077911377, + "learning_rate": 9.29203539823009e-06, + "loss": 0.0034, + "step": 14862 + }, + { + "epoch": 1.6113399826539463, + "grad_norm": 2.3752329349517822, + "learning_rate": 9.291310024662702e-06, + "loss": 0.0134, + "step": 14863 + }, + { + "epoch": 1.611448395490026, + "grad_norm": 5.342087268829346, + "learning_rate": 9.290584651095314e-06, + "loss": 0.0167, + "step": 14864 + }, + { + "epoch": 1.6115568083261058, + "grad_norm": 2.2682762145996094, + "learning_rate": 9.289859277527928e-06, + "loss": 0.0063, + "step": 14865 + }, + { + "epoch": 1.6116652211621856, + "grad_norm": 5.0044684410095215, + "learning_rate": 9.28913390396054e-06, + "loss": 0.0185, + "step": 14866 + }, + { + "epoch": 1.6117736339982653, + "grad_norm": 1.5474010705947876, + "learning_rate": 9.288408530393154e-06, + "loss": 0.018, + "step": 14867 + }, + { + "epoch": 1.6118820468343453, + "grad_norm": 0.2972962260246277, + "learning_rate": 9.287683156825766e-06, + "loss": 0.005, + "step": 14868 + }, + { + "epoch": 1.6119904596704249, + "grad_norm": 3.1228957176208496, + "learning_rate": 9.28695778325838e-06, + "loss": 0.0244, + "step": 14869 + }, + { + "epoch": 1.6120988725065049, + "grad_norm": 1.0001132488250732, + "learning_rate": 9.286232409690992e-06, + "loss": 0.0107, + "step": 14870 + }, + { + "epoch": 1.6122072853425846, + "grad_norm": 2.6822574138641357, + "learning_rate": 9.285507036123604e-06, + "loss": 0.0411, + "step": 14871 + }, + { + "epoch": 1.6123156981786644, + "grad_norm": 1.78688645362854, + "learning_rate": 9.284781662556216e-06, + "loss": 0.0357, + "step": 14872 + }, + { + "epoch": 1.6124241110147441, + "grad_norm": 7.010186672210693, + "learning_rate": 9.28405628898883e-06, + "loss": 0.0439, + "step": 14873 + }, + { + "epoch": 1.612532523850824, + "grad_norm": 0.4676472842693329, + "learning_rate": 9.283330915421442e-06, + "loss": 0.0038, + "step": 14874 + }, + { + "epoch": 1.612640936686904, + "grad_norm": 0.87269127368927, + "learning_rate": 9.282605541854054e-06, + "loss": 0.0034, + "step": 14875 + }, + { + "epoch": 1.6127493495229834, + "grad_norm": 1.5969784259796143, + "learning_rate": 9.281880168286668e-06, + "loss": 0.0207, + "step": 14876 + }, + { + "epoch": 1.6128577623590634, + "grad_norm": 1.100502371788025, + "learning_rate": 9.281154794719282e-06, + "loss": 0.0063, + "step": 14877 + }, + { + "epoch": 1.612966175195143, + "grad_norm": 0.026317687705159187, + "learning_rate": 9.280429421151894e-06, + "loss": 0.0004, + "step": 14878 + }, + { + "epoch": 1.613074588031223, + "grad_norm": 6.1632513999938965, + "learning_rate": 9.279704047584506e-06, + "loss": 0.0185, + "step": 14879 + }, + { + "epoch": 1.6131830008673027, + "grad_norm": 1.1436762809753418, + "learning_rate": 9.27897867401712e-06, + "loss": 0.0126, + "step": 14880 + }, + { + "epoch": 1.6132914137033825, + "grad_norm": 1.8553893566131592, + "learning_rate": 9.278253300449732e-06, + "loss": 0.0185, + "step": 14881 + }, + { + "epoch": 1.6133998265394622, + "grad_norm": 2.536510944366455, + "learning_rate": 9.277527926882344e-06, + "loss": 0.0081, + "step": 14882 + }, + { + "epoch": 1.613508239375542, + "grad_norm": 1.9371095895767212, + "learning_rate": 9.276802553314958e-06, + "loss": 0.0168, + "step": 14883 + }, + { + "epoch": 1.613616652211622, + "grad_norm": 2.359833240509033, + "learning_rate": 9.27607717974757e-06, + "loss": 0.0245, + "step": 14884 + }, + { + "epoch": 1.6137250650477015, + "grad_norm": 3.223142147064209, + "learning_rate": 9.275351806180184e-06, + "loss": 0.0919, + "step": 14885 + }, + { + "epoch": 1.6138334778837815, + "grad_norm": 1.5179420709609985, + "learning_rate": 9.274626432612796e-06, + "loss": 0.0271, + "step": 14886 + }, + { + "epoch": 1.6139418907198613, + "grad_norm": 2.8260083198547363, + "learning_rate": 9.273901059045409e-06, + "loss": 0.0268, + "step": 14887 + }, + { + "epoch": 1.614050303555941, + "grad_norm": 0.479798823595047, + "learning_rate": 9.273175685478022e-06, + "loss": 0.0033, + "step": 14888 + }, + { + "epoch": 1.6141587163920208, + "grad_norm": 2.638624429702759, + "learning_rate": 9.272450311910635e-06, + "loss": 0.0196, + "step": 14889 + }, + { + "epoch": 1.6142671292281006, + "grad_norm": 1.9030306339263916, + "learning_rate": 9.271724938343247e-06, + "loss": 0.028, + "step": 14890 + }, + { + "epoch": 1.6143755420641805, + "grad_norm": 1.2386666536331177, + "learning_rate": 9.27099956477586e-06, + "loss": 0.0174, + "step": 14891 + }, + { + "epoch": 1.61448395490026, + "grad_norm": 0.2637467682361603, + "learning_rate": 9.270274191208473e-06, + "loss": 0.0044, + "step": 14892 + }, + { + "epoch": 1.61459236773634, + "grad_norm": 1.1859129667282104, + "learning_rate": 9.269548817641087e-06, + "loss": 0.0026, + "step": 14893 + }, + { + "epoch": 1.6147007805724196, + "grad_norm": 3.136669874191284, + "learning_rate": 9.268823444073699e-06, + "loss": 0.0339, + "step": 14894 + }, + { + "epoch": 1.6148091934084996, + "grad_norm": 1.6481494903564453, + "learning_rate": 9.268098070506312e-06, + "loss": 0.0504, + "step": 14895 + }, + { + "epoch": 1.6149176062445794, + "grad_norm": 2.5970237255096436, + "learning_rate": 9.267372696938925e-06, + "loss": 0.0099, + "step": 14896 + }, + { + "epoch": 1.6150260190806591, + "grad_norm": 1.5235086679458618, + "learning_rate": 9.266647323371537e-06, + "loss": 0.0189, + "step": 14897 + }, + { + "epoch": 1.6151344319167389, + "grad_norm": 2.4292240142822266, + "learning_rate": 9.265921949804149e-06, + "loss": 0.0219, + "step": 14898 + }, + { + "epoch": 1.6152428447528187, + "grad_norm": 0.4410060942173004, + "learning_rate": 9.265196576236763e-06, + "loss": 0.0029, + "step": 14899 + }, + { + "epoch": 1.6153512575888986, + "grad_norm": 1.5582332611083984, + "learning_rate": 9.264471202669375e-06, + "loss": 0.0147, + "step": 14900 + }, + { + "epoch": 1.6154596704249782, + "grad_norm": 1.1147260665893555, + "learning_rate": 9.263745829101989e-06, + "loss": 0.0274, + "step": 14901 + }, + { + "epoch": 1.6155680832610582, + "grad_norm": 0.1970207393169403, + "learning_rate": 9.263020455534601e-06, + "loss": 0.001, + "step": 14902 + }, + { + "epoch": 1.615676496097138, + "grad_norm": 5.274741172790527, + "learning_rate": 9.262295081967215e-06, + "loss": 0.0122, + "step": 14903 + }, + { + "epoch": 1.6157849089332177, + "grad_norm": 1.4185986518859863, + "learning_rate": 9.261569708399827e-06, + "loss": 0.0388, + "step": 14904 + }, + { + "epoch": 1.6158933217692975, + "grad_norm": 1.3981454372406006, + "learning_rate": 9.260844334832439e-06, + "loss": 0.0062, + "step": 14905 + }, + { + "epoch": 1.6160017346053772, + "grad_norm": 2.7208101749420166, + "learning_rate": 9.260118961265053e-06, + "loss": 0.0492, + "step": 14906 + }, + { + "epoch": 1.6161101474414572, + "grad_norm": 5.0677490234375, + "learning_rate": 9.259393587697665e-06, + "loss": 0.0203, + "step": 14907 + }, + { + "epoch": 1.6162185602775367, + "grad_norm": 2.446002721786499, + "learning_rate": 9.258668214130277e-06, + "loss": 0.0081, + "step": 14908 + }, + { + "epoch": 1.6163269731136167, + "grad_norm": 1.423140048980713, + "learning_rate": 9.257942840562891e-06, + "loss": 0.0156, + "step": 14909 + }, + { + "epoch": 1.6164353859496965, + "grad_norm": 3.861133337020874, + "learning_rate": 9.257217466995503e-06, + "loss": 0.0657, + "step": 14910 + }, + { + "epoch": 1.6165437987857763, + "grad_norm": 0.3100114166736603, + "learning_rate": 9.256492093428117e-06, + "loss": 0.0027, + "step": 14911 + }, + { + "epoch": 1.616652211621856, + "grad_norm": 2.9661388397216797, + "learning_rate": 9.255766719860729e-06, + "loss": 0.0368, + "step": 14912 + }, + { + "epoch": 1.6167606244579358, + "grad_norm": 0.4513704180717468, + "learning_rate": 9.255041346293341e-06, + "loss": 0.0022, + "step": 14913 + }, + { + "epoch": 1.6168690372940158, + "grad_norm": 1.6289019584655762, + "learning_rate": 9.254315972725955e-06, + "loss": 0.023, + "step": 14914 + }, + { + "epoch": 1.6169774501300953, + "grad_norm": 1.4164310693740845, + "learning_rate": 9.253590599158567e-06, + "loss": 0.0107, + "step": 14915 + }, + { + "epoch": 1.6170858629661753, + "grad_norm": 2.38551664352417, + "learning_rate": 9.25286522559118e-06, + "loss": 0.0301, + "step": 14916 + }, + { + "epoch": 1.6171942758022548, + "grad_norm": 0.21537308394908905, + "learning_rate": 9.252139852023793e-06, + "loss": 0.0023, + "step": 14917 + }, + { + "epoch": 1.6173026886383348, + "grad_norm": 0.26656922698020935, + "learning_rate": 9.251414478456405e-06, + "loss": 0.0026, + "step": 14918 + }, + { + "epoch": 1.6174111014744146, + "grad_norm": 1.3130590915679932, + "learning_rate": 9.25068910488902e-06, + "loss": 0.0204, + "step": 14919 + }, + { + "epoch": 1.6175195143104943, + "grad_norm": 1.3633689880371094, + "learning_rate": 9.249963731321631e-06, + "loss": 0.0159, + "step": 14920 + }, + { + "epoch": 1.617627927146574, + "grad_norm": 2.754244327545166, + "learning_rate": 9.249238357754245e-06, + "loss": 0.0473, + "step": 14921 + }, + { + "epoch": 1.6177363399826539, + "grad_norm": 2.7332615852355957, + "learning_rate": 9.248512984186857e-06, + "loss": 0.012, + "step": 14922 + }, + { + "epoch": 1.6178447528187339, + "grad_norm": 3.874734401702881, + "learning_rate": 9.24778761061947e-06, + "loss": 0.0294, + "step": 14923 + }, + { + "epoch": 1.6179531656548134, + "grad_norm": 3.508448600769043, + "learning_rate": 9.247062237052082e-06, + "loss": 0.0152, + "step": 14924 + }, + { + "epoch": 1.6180615784908934, + "grad_norm": 0.26551738381385803, + "learning_rate": 9.246336863484695e-06, + "loss": 0.0024, + "step": 14925 + }, + { + "epoch": 1.6181699913269731, + "grad_norm": 2.4212522506713867, + "learning_rate": 9.245611489917308e-06, + "loss": 0.0088, + "step": 14926 + }, + { + "epoch": 1.618278404163053, + "grad_norm": 4.066838264465332, + "learning_rate": 9.244886116349921e-06, + "loss": 0.0644, + "step": 14927 + }, + { + "epoch": 1.6183868169991327, + "grad_norm": 2.732771396636963, + "learning_rate": 9.244160742782534e-06, + "loss": 0.0187, + "step": 14928 + }, + { + "epoch": 1.6184952298352124, + "grad_norm": 0.1114267036318779, + "learning_rate": 9.243435369215147e-06, + "loss": 0.0014, + "step": 14929 + }, + { + "epoch": 1.6186036426712924, + "grad_norm": 1.2716273069381714, + "learning_rate": 9.24270999564776e-06, + "loss": 0.0081, + "step": 14930 + }, + { + "epoch": 1.618712055507372, + "grad_norm": 1.3370023965835571, + "learning_rate": 9.241984622080372e-06, + "loss": 0.0207, + "step": 14931 + }, + { + "epoch": 1.618820468343452, + "grad_norm": 1.9891058206558228, + "learning_rate": 9.241259248512986e-06, + "loss": 0.0152, + "step": 14932 + }, + { + "epoch": 1.6189288811795317, + "grad_norm": 3.1162197589874268, + "learning_rate": 9.240533874945598e-06, + "loss": 0.1339, + "step": 14933 + }, + { + "epoch": 1.6190372940156115, + "grad_norm": 4.455197334289551, + "learning_rate": 9.23980850137821e-06, + "loss": 0.0168, + "step": 14934 + }, + { + "epoch": 1.6191457068516912, + "grad_norm": 1.1939151287078857, + "learning_rate": 9.239083127810824e-06, + "loss": 0.0142, + "step": 14935 + }, + { + "epoch": 1.619254119687771, + "grad_norm": 1.6836150884628296, + "learning_rate": 9.238357754243436e-06, + "loss": 0.0066, + "step": 14936 + }, + { + "epoch": 1.619362532523851, + "grad_norm": 3.3279895782470703, + "learning_rate": 9.23763238067605e-06, + "loss": 0.0128, + "step": 14937 + }, + { + "epoch": 1.6194709453599305, + "grad_norm": 1.2585610151290894, + "learning_rate": 9.236907007108662e-06, + "loss": 0.0131, + "step": 14938 + }, + { + "epoch": 1.6195793581960105, + "grad_norm": 1.9406107664108276, + "learning_rate": 9.236181633541274e-06, + "loss": 0.0134, + "step": 14939 + }, + { + "epoch": 1.61968777103209, + "grad_norm": 2.653735399246216, + "learning_rate": 9.235456259973888e-06, + "loss": 0.0168, + "step": 14940 + }, + { + "epoch": 1.61979618386817, + "grad_norm": 3.1751229763031006, + "learning_rate": 9.2347308864065e-06, + "loss": 0.0264, + "step": 14941 + }, + { + "epoch": 1.6199045967042498, + "grad_norm": 2.7123820781707764, + "learning_rate": 9.234005512839112e-06, + "loss": 0.0521, + "step": 14942 + }, + { + "epoch": 1.6200130095403296, + "grad_norm": 1.0017452239990234, + "learning_rate": 9.233280139271726e-06, + "loss": 0.0079, + "step": 14943 + }, + { + "epoch": 1.6201214223764093, + "grad_norm": 3.228579521179199, + "learning_rate": 9.232554765704338e-06, + "loss": 0.0079, + "step": 14944 + }, + { + "epoch": 1.620229835212489, + "grad_norm": 0.15828923881053925, + "learning_rate": 9.231829392136952e-06, + "loss": 0.0011, + "step": 14945 + }, + { + "epoch": 1.620338248048569, + "grad_norm": 0.06122790277004242, + "learning_rate": 9.231104018569564e-06, + "loss": 0.0007, + "step": 14946 + }, + { + "epoch": 1.6204466608846486, + "grad_norm": 2.87374210357666, + "learning_rate": 9.230378645002178e-06, + "loss": 0.0164, + "step": 14947 + }, + { + "epoch": 1.6205550737207286, + "grad_norm": 4.527087688446045, + "learning_rate": 9.22965327143479e-06, + "loss": 0.0539, + "step": 14948 + }, + { + "epoch": 1.6206634865568084, + "grad_norm": 2.630368709564209, + "learning_rate": 9.228927897867402e-06, + "loss": 0.0284, + "step": 14949 + }, + { + "epoch": 1.6207718993928881, + "grad_norm": 1.5814659595489502, + "learning_rate": 9.228202524300014e-06, + "loss": 0.0339, + "step": 14950 + }, + { + "epoch": 1.620880312228968, + "grad_norm": 0.2933655381202698, + "learning_rate": 9.227477150732628e-06, + "loss": 0.0031, + "step": 14951 + }, + { + "epoch": 1.6209887250650477, + "grad_norm": 0.5984373092651367, + "learning_rate": 9.22675177716524e-06, + "loss": 0.0103, + "step": 14952 + }, + { + "epoch": 1.6210971379011276, + "grad_norm": 1.6154136657714844, + "learning_rate": 9.226026403597854e-06, + "loss": 0.0283, + "step": 14953 + }, + { + "epoch": 1.6212055507372072, + "grad_norm": 2.1973721981048584, + "learning_rate": 9.225301030030466e-06, + "loss": 0.0164, + "step": 14954 + }, + { + "epoch": 1.6213139635732872, + "grad_norm": 2.1431241035461426, + "learning_rate": 9.22457565646308e-06, + "loss": 0.0408, + "step": 14955 + }, + { + "epoch": 1.6214223764093667, + "grad_norm": 2.742690086364746, + "learning_rate": 9.223850282895692e-06, + "loss": 0.0189, + "step": 14956 + }, + { + "epoch": 1.6215307892454467, + "grad_norm": 0.9786505699157715, + "learning_rate": 9.223124909328304e-06, + "loss": 0.0183, + "step": 14957 + }, + { + "epoch": 1.6216392020815265, + "grad_norm": 0.34138551354408264, + "learning_rate": 9.222399535760918e-06, + "loss": 0.0031, + "step": 14958 + }, + { + "epoch": 1.6217476149176062, + "grad_norm": 3.805453062057495, + "learning_rate": 9.22167416219353e-06, + "loss": 0.0248, + "step": 14959 + }, + { + "epoch": 1.6218560277536862, + "grad_norm": 5.054800510406494, + "learning_rate": 9.220948788626142e-06, + "loss": 0.0239, + "step": 14960 + }, + { + "epoch": 1.6219644405897657, + "grad_norm": 3.4996232986450195, + "learning_rate": 9.220223415058756e-06, + "loss": 0.0182, + "step": 14961 + }, + { + "epoch": 1.6220728534258457, + "grad_norm": 1.9390510320663452, + "learning_rate": 9.219498041491368e-06, + "loss": 0.0339, + "step": 14962 + }, + { + "epoch": 1.6221812662619253, + "grad_norm": 0.7754453420639038, + "learning_rate": 9.218772667923982e-06, + "loss": 0.0071, + "step": 14963 + }, + { + "epoch": 1.6222896790980053, + "grad_norm": 3.232431411743164, + "learning_rate": 9.218047294356594e-06, + "loss": 0.0544, + "step": 14964 + }, + { + "epoch": 1.622398091934085, + "grad_norm": 3.081167459487915, + "learning_rate": 9.217321920789207e-06, + "loss": 0.0286, + "step": 14965 + }, + { + "epoch": 1.6225065047701648, + "grad_norm": 4.725155353546143, + "learning_rate": 9.21659654722182e-06, + "loss": 0.0517, + "step": 14966 + }, + { + "epoch": 1.6226149176062445, + "grad_norm": 2.961284637451172, + "learning_rate": 9.215871173654433e-06, + "loss": 0.015, + "step": 14967 + }, + { + "epoch": 1.6227233304423243, + "grad_norm": 2.0470149517059326, + "learning_rate": 9.215145800087045e-06, + "loss": 0.0219, + "step": 14968 + }, + { + "epoch": 1.6228317432784043, + "grad_norm": 1.16858971118927, + "learning_rate": 9.214420426519659e-06, + "loss": 0.0247, + "step": 14969 + }, + { + "epoch": 1.6229401561144838, + "grad_norm": 3.113975763320923, + "learning_rate": 9.21369505295227e-06, + "loss": 0.0518, + "step": 14970 + }, + { + "epoch": 1.6230485689505638, + "grad_norm": 5.130013465881348, + "learning_rate": 9.212969679384885e-06, + "loss": 0.0561, + "step": 14971 + }, + { + "epoch": 1.6231569817866436, + "grad_norm": 1.7530452013015747, + "learning_rate": 9.212244305817497e-06, + "loss": 0.0085, + "step": 14972 + }, + { + "epoch": 1.6232653946227233, + "grad_norm": 1.9348236322402954, + "learning_rate": 9.21151893225011e-06, + "loss": 0.0279, + "step": 14973 + }, + { + "epoch": 1.623373807458803, + "grad_norm": 0.10632398724555969, + "learning_rate": 9.210793558682723e-06, + "loss": 0.0013, + "step": 14974 + }, + { + "epoch": 1.6234822202948829, + "grad_norm": 1.7892241477966309, + "learning_rate": 9.210068185115335e-06, + "loss": 0.0094, + "step": 14975 + }, + { + "epoch": 1.6235906331309629, + "grad_norm": 0.47092604637145996, + "learning_rate": 9.209342811547947e-06, + "loss": 0.006, + "step": 14976 + }, + { + "epoch": 1.6236990459670424, + "grad_norm": 1.750895380973816, + "learning_rate": 9.20861743798056e-06, + "loss": 0.0113, + "step": 14977 + }, + { + "epoch": 1.6238074588031224, + "grad_norm": 1.34165358543396, + "learning_rate": 9.207892064413173e-06, + "loss": 0.023, + "step": 14978 + }, + { + "epoch": 1.623915871639202, + "grad_norm": 1.7434542179107666, + "learning_rate": 9.207166690845787e-06, + "loss": 0.0332, + "step": 14979 + }, + { + "epoch": 1.624024284475282, + "grad_norm": 1.8351025581359863, + "learning_rate": 9.206441317278399e-06, + "loss": 0.0248, + "step": 14980 + }, + { + "epoch": 1.6241326973113617, + "grad_norm": 4.55523681640625, + "learning_rate": 9.205715943711013e-06, + "loss": 0.023, + "step": 14981 + }, + { + "epoch": 1.6242411101474414, + "grad_norm": 0.7139286994934082, + "learning_rate": 9.204990570143625e-06, + "loss": 0.0154, + "step": 14982 + }, + { + "epoch": 1.6243495229835212, + "grad_norm": 0.15102842450141907, + "learning_rate": 9.204265196576237e-06, + "loss": 0.0014, + "step": 14983 + }, + { + "epoch": 1.624457935819601, + "grad_norm": 0.8770061731338501, + "learning_rate": 9.203539823008851e-06, + "loss": 0.0183, + "step": 14984 + }, + { + "epoch": 1.624566348655681, + "grad_norm": 2.698756694793701, + "learning_rate": 9.202814449441463e-06, + "loss": 0.0558, + "step": 14985 + }, + { + "epoch": 1.6246747614917605, + "grad_norm": 2.297858715057373, + "learning_rate": 9.202089075874075e-06, + "loss": 0.0154, + "step": 14986 + }, + { + "epoch": 1.6247831743278405, + "grad_norm": 0.6573435068130493, + "learning_rate": 9.201363702306689e-06, + "loss": 0.0051, + "step": 14987 + }, + { + "epoch": 1.6248915871639202, + "grad_norm": 5.464795112609863, + "learning_rate": 9.200638328739301e-06, + "loss": 0.0428, + "step": 14988 + }, + { + "epoch": 1.625, + "grad_norm": 1.2676451206207275, + "learning_rate": 9.199912955171915e-06, + "loss": 0.0068, + "step": 14989 + }, + { + "epoch": 1.6251084128360798, + "grad_norm": 0.7770810127258301, + "learning_rate": 9.199187581604527e-06, + "loss": 0.0087, + "step": 14990 + }, + { + "epoch": 1.6252168256721595, + "grad_norm": 1.8237894773483276, + "learning_rate": 9.19846220803714e-06, + "loss": 0.024, + "step": 14991 + }, + { + "epoch": 1.6253252385082395, + "grad_norm": 3.253490686416626, + "learning_rate": 9.197736834469753e-06, + "loss": 0.0418, + "step": 14992 + }, + { + "epoch": 1.625433651344319, + "grad_norm": 3.048492431640625, + "learning_rate": 9.197011460902365e-06, + "loss": 0.0123, + "step": 14993 + }, + { + "epoch": 1.625542064180399, + "grad_norm": 2.8859782218933105, + "learning_rate": 9.196286087334977e-06, + "loss": 0.0245, + "step": 14994 + }, + { + "epoch": 1.6256504770164788, + "grad_norm": 4.756761074066162, + "learning_rate": 9.195560713767591e-06, + "loss": 0.0306, + "step": 14995 + }, + { + "epoch": 1.6257588898525586, + "grad_norm": 0.8150362968444824, + "learning_rate": 9.194835340200205e-06, + "loss": 0.0061, + "step": 14996 + }, + { + "epoch": 1.6258673026886383, + "grad_norm": 0.8126845359802246, + "learning_rate": 9.194109966632817e-06, + "loss": 0.0073, + "step": 14997 + }, + { + "epoch": 1.625975715524718, + "grad_norm": 1.2128932476043701, + "learning_rate": 9.19338459306543e-06, + "loss": 0.0215, + "step": 14998 + }, + { + "epoch": 1.626084128360798, + "grad_norm": 1.4830033779144287, + "learning_rate": 9.192659219498043e-06, + "loss": 0.0193, + "step": 14999 + }, + { + "epoch": 1.6261925411968776, + "grad_norm": 0.5940045118331909, + "learning_rate": 9.191933845930655e-06, + "loss": 0.0048, + "step": 15000 + }, + { + "epoch": 1.6263009540329576, + "grad_norm": 3.120070457458496, + "learning_rate": 9.191208472363267e-06, + "loss": 0.0067, + "step": 15001 + }, + { + "epoch": 1.6264093668690371, + "grad_norm": 0.4811709225177765, + "learning_rate": 9.19048309879588e-06, + "loss": 0.009, + "step": 15002 + }, + { + "epoch": 1.6265177797051171, + "grad_norm": 1.7797220945358276, + "learning_rate": 9.189757725228493e-06, + "loss": 0.0264, + "step": 15003 + }, + { + "epoch": 1.626626192541197, + "grad_norm": 4.170923709869385, + "learning_rate": 9.189032351661106e-06, + "loss": 0.0181, + "step": 15004 + }, + { + "epoch": 1.6267346053772767, + "grad_norm": 0.0498419888317585, + "learning_rate": 9.18830697809372e-06, + "loss": 0.0006, + "step": 15005 + }, + { + "epoch": 1.6268430182133564, + "grad_norm": 0.21520179510116577, + "learning_rate": 9.187581604526332e-06, + "loss": 0.0011, + "step": 15006 + }, + { + "epoch": 1.6269514310494362, + "grad_norm": 0.07816744595766068, + "learning_rate": 9.186856230958945e-06, + "loss": 0.0008, + "step": 15007 + }, + { + "epoch": 1.6270598438855162, + "grad_norm": 3.561126470565796, + "learning_rate": 9.186130857391558e-06, + "loss": 0.0057, + "step": 15008 + }, + { + "epoch": 1.6271682567215957, + "grad_norm": 2.819026231765747, + "learning_rate": 9.18540548382417e-06, + "loss": 0.0335, + "step": 15009 + }, + { + "epoch": 1.6272766695576757, + "grad_norm": 1.3106744289398193, + "learning_rate": 9.184680110256784e-06, + "loss": 0.0083, + "step": 15010 + }, + { + "epoch": 1.6273850823937555, + "grad_norm": 14.440179824829102, + "learning_rate": 9.183954736689396e-06, + "loss": 0.0192, + "step": 15011 + }, + { + "epoch": 1.6274934952298352, + "grad_norm": 1.5611648559570312, + "learning_rate": 9.183229363122008e-06, + "loss": 0.0115, + "step": 15012 + }, + { + "epoch": 1.627601908065915, + "grad_norm": 5.597171306610107, + "learning_rate": 9.182503989554622e-06, + "loss": 0.0442, + "step": 15013 + }, + { + "epoch": 1.6277103209019947, + "grad_norm": 4.265579700469971, + "learning_rate": 9.181778615987235e-06, + "loss": 0.0703, + "step": 15014 + }, + { + "epoch": 1.6278187337380747, + "grad_norm": 0.6274530291557312, + "learning_rate": 9.181053242419848e-06, + "loss": 0.0064, + "step": 15015 + }, + { + "epoch": 1.6279271465741543, + "grad_norm": 0.30883318185806274, + "learning_rate": 9.18032786885246e-06, + "loss": 0.0032, + "step": 15016 + }, + { + "epoch": 1.6280355594102343, + "grad_norm": 4.040515422821045, + "learning_rate": 9.179602495285072e-06, + "loss": 0.0219, + "step": 15017 + }, + { + "epoch": 1.6281439722463138, + "grad_norm": 0.34900298714637756, + "learning_rate": 9.178877121717686e-06, + "loss": 0.0023, + "step": 15018 + }, + { + "epoch": 1.6282523850823938, + "grad_norm": 0.40953782200813293, + "learning_rate": 9.178151748150298e-06, + "loss": 0.0019, + "step": 15019 + }, + { + "epoch": 1.6283607979184735, + "grad_norm": 0.22388191521167755, + "learning_rate": 9.17742637458291e-06, + "loss": 0.0021, + "step": 15020 + }, + { + "epoch": 1.6284692107545533, + "grad_norm": 3.9781274795532227, + "learning_rate": 9.176701001015524e-06, + "loss": 0.0288, + "step": 15021 + }, + { + "epoch": 1.6285776235906333, + "grad_norm": 0.8640159964561462, + "learning_rate": 9.175975627448138e-06, + "loss": 0.0035, + "step": 15022 + }, + { + "epoch": 1.6286860364267128, + "grad_norm": 3.601971387863159, + "learning_rate": 9.17525025388075e-06, + "loss": 0.0448, + "step": 15023 + }, + { + "epoch": 1.6287944492627928, + "grad_norm": 0.7582892775535583, + "learning_rate": 9.174524880313362e-06, + "loss": 0.0073, + "step": 15024 + }, + { + "epoch": 1.6289028620988724, + "grad_norm": 4.88848352432251, + "learning_rate": 9.173799506745976e-06, + "loss": 0.0311, + "step": 15025 + }, + { + "epoch": 1.6290112749349523, + "grad_norm": 1.9968647956848145, + "learning_rate": 9.173074133178588e-06, + "loss": 0.0466, + "step": 15026 + }, + { + "epoch": 1.629119687771032, + "grad_norm": 3.673017978668213, + "learning_rate": 9.1723487596112e-06, + "loss": 0.0123, + "step": 15027 + }, + { + "epoch": 1.6292281006071119, + "grad_norm": 4.9514946937561035, + "learning_rate": 9.171623386043812e-06, + "loss": 0.0985, + "step": 15028 + }, + { + "epoch": 1.6293365134431916, + "grad_norm": 16.922378540039062, + "learning_rate": 9.170898012476426e-06, + "loss": 0.0532, + "step": 15029 + }, + { + "epoch": 1.6294449262792714, + "grad_norm": 9.318710327148438, + "learning_rate": 9.170172638909038e-06, + "loss": 0.0375, + "step": 15030 + }, + { + "epoch": 1.6295533391153514, + "grad_norm": 2.756115198135376, + "learning_rate": 9.169447265341652e-06, + "loss": 0.0182, + "step": 15031 + }, + { + "epoch": 1.629661751951431, + "grad_norm": 3.0381252765655518, + "learning_rate": 9.168721891774264e-06, + "loss": 0.0432, + "step": 15032 + }, + { + "epoch": 1.629770164787511, + "grad_norm": 2.2886476516723633, + "learning_rate": 9.167996518206878e-06, + "loss": 0.0342, + "step": 15033 + }, + { + "epoch": 1.6298785776235907, + "grad_norm": 1.3757579326629639, + "learning_rate": 9.16727114463949e-06, + "loss": 0.0139, + "step": 15034 + }, + { + "epoch": 1.6299869904596704, + "grad_norm": 1.3604207038879395, + "learning_rate": 9.166545771072102e-06, + "loss": 0.0086, + "step": 15035 + }, + { + "epoch": 1.6300954032957502, + "grad_norm": 1.9568043947219849, + "learning_rate": 9.165820397504716e-06, + "loss": 0.0247, + "step": 15036 + }, + { + "epoch": 1.63020381613183, + "grad_norm": 0.34481239318847656, + "learning_rate": 9.165095023937328e-06, + "loss": 0.0019, + "step": 15037 + }, + { + "epoch": 1.63031222896791, + "grad_norm": 0.7006718516349792, + "learning_rate": 9.16436965036994e-06, + "loss": 0.0042, + "step": 15038 + }, + { + "epoch": 1.6304206418039895, + "grad_norm": 2.279468059539795, + "learning_rate": 9.163644276802554e-06, + "loss": 0.0161, + "step": 15039 + }, + { + "epoch": 1.6305290546400695, + "grad_norm": 0.3831808567047119, + "learning_rate": 9.162918903235168e-06, + "loss": 0.0026, + "step": 15040 + }, + { + "epoch": 1.630637467476149, + "grad_norm": 0.4595387578010559, + "learning_rate": 9.16219352966778e-06, + "loss": 0.0017, + "step": 15041 + }, + { + "epoch": 1.630745880312229, + "grad_norm": 1.791296124458313, + "learning_rate": 9.161468156100392e-06, + "loss": 0.034, + "step": 15042 + }, + { + "epoch": 1.6308542931483088, + "grad_norm": 2.910297155380249, + "learning_rate": 9.160742782533005e-06, + "loss": 0.0128, + "step": 15043 + }, + { + "epoch": 1.6309627059843885, + "grad_norm": 2.8781659603118896, + "learning_rate": 9.160017408965618e-06, + "loss": 0.0077, + "step": 15044 + }, + { + "epoch": 1.6310711188204683, + "grad_norm": 10.36015510559082, + "learning_rate": 9.15929203539823e-06, + "loss": 0.0479, + "step": 15045 + }, + { + "epoch": 1.631179531656548, + "grad_norm": 0.4782431125640869, + "learning_rate": 9.158566661830843e-06, + "loss": 0.0025, + "step": 15046 + }, + { + "epoch": 1.631287944492628, + "grad_norm": 4.580526351928711, + "learning_rate": 9.157841288263457e-06, + "loss": 0.0501, + "step": 15047 + }, + { + "epoch": 1.6313963573287076, + "grad_norm": 2.9283554553985596, + "learning_rate": 9.15711591469607e-06, + "loss": 0.0452, + "step": 15048 + }, + { + "epoch": 1.6315047701647876, + "grad_norm": 4.112139701843262, + "learning_rate": 9.156390541128683e-06, + "loss": 0.0257, + "step": 15049 + }, + { + "epoch": 1.6316131830008673, + "grad_norm": 1.1178699731826782, + "learning_rate": 9.155665167561295e-06, + "loss": 0.0061, + "step": 15050 + }, + { + "epoch": 1.631721595836947, + "grad_norm": 1.2005096673965454, + "learning_rate": 9.154939793993909e-06, + "loss": 0.0214, + "step": 15051 + }, + { + "epoch": 1.6318300086730269, + "grad_norm": 2.911655902862549, + "learning_rate": 9.15421442042652e-06, + "loss": 0.0118, + "step": 15052 + }, + { + "epoch": 1.6319384215091066, + "grad_norm": 3.4195432662963867, + "learning_rate": 9.153489046859133e-06, + "loss": 0.0809, + "step": 15053 + }, + { + "epoch": 1.6320468343451866, + "grad_norm": 1.0413109064102173, + "learning_rate": 9.152763673291745e-06, + "loss": 0.0218, + "step": 15054 + }, + { + "epoch": 1.6321552471812661, + "grad_norm": 2.018096923828125, + "learning_rate": 9.152038299724359e-06, + "loss": 0.0111, + "step": 15055 + }, + { + "epoch": 1.6322636600173461, + "grad_norm": 9.901660919189453, + "learning_rate": 9.151312926156971e-06, + "loss": 0.0483, + "step": 15056 + }, + { + "epoch": 1.632372072853426, + "grad_norm": 0.8261013627052307, + "learning_rate": 9.150587552589585e-06, + "loss": 0.0038, + "step": 15057 + }, + { + "epoch": 1.6324804856895057, + "grad_norm": 0.7611111402511597, + "learning_rate": 9.149862179022197e-06, + "loss": 0.0055, + "step": 15058 + }, + { + "epoch": 1.6325888985255854, + "grad_norm": 10.83676815032959, + "learning_rate": 9.14913680545481e-06, + "loss": 0.1582, + "step": 15059 + }, + { + "epoch": 1.6326973113616652, + "grad_norm": 2.973228931427002, + "learning_rate": 9.148411431887423e-06, + "loss": 0.0254, + "step": 15060 + }, + { + "epoch": 1.6328057241977452, + "grad_norm": 2.6747870445251465, + "learning_rate": 9.147686058320035e-06, + "loss": 0.0089, + "step": 15061 + }, + { + "epoch": 1.6329141370338247, + "grad_norm": 3.431469440460205, + "learning_rate": 9.146960684752649e-06, + "loss": 0.0373, + "step": 15062 + }, + { + "epoch": 1.6330225498699047, + "grad_norm": 2.0020527839660645, + "learning_rate": 9.146235311185261e-06, + "loss": 0.0274, + "step": 15063 + }, + { + "epoch": 1.6331309627059842, + "grad_norm": 1.0997352600097656, + "learning_rate": 9.145509937617873e-06, + "loss": 0.0279, + "step": 15064 + }, + { + "epoch": 1.6332393755420642, + "grad_norm": 16.802268981933594, + "learning_rate": 9.144784564050487e-06, + "loss": 0.0152, + "step": 15065 + }, + { + "epoch": 1.633347788378144, + "grad_norm": 6.618717193603516, + "learning_rate": 9.1440591904831e-06, + "loss": 0.0276, + "step": 15066 + }, + { + "epoch": 1.6334562012142237, + "grad_norm": 3.1559360027313232, + "learning_rate": 9.143333816915713e-06, + "loss": 0.0229, + "step": 15067 + }, + { + "epoch": 1.6335646140503035, + "grad_norm": 5.29664421081543, + "learning_rate": 9.142608443348325e-06, + "loss": 0.018, + "step": 15068 + }, + { + "epoch": 1.6336730268863833, + "grad_norm": 5.386829376220703, + "learning_rate": 9.141883069780937e-06, + "loss": 0.0915, + "step": 15069 + }, + { + "epoch": 1.6337814397224633, + "grad_norm": 0.8639638423919678, + "learning_rate": 9.141157696213551e-06, + "loss": 0.0061, + "step": 15070 + }, + { + "epoch": 1.6338898525585428, + "grad_norm": 6.6624860763549805, + "learning_rate": 9.140432322646163e-06, + "loss": 0.041, + "step": 15071 + }, + { + "epoch": 1.6339982653946228, + "grad_norm": 5.809532165527344, + "learning_rate": 9.139706949078775e-06, + "loss": 0.0626, + "step": 15072 + }, + { + "epoch": 1.6341066782307025, + "grad_norm": 4.2279953956604, + "learning_rate": 9.13898157551139e-06, + "loss": 0.0097, + "step": 15073 + }, + { + "epoch": 1.6342150910667823, + "grad_norm": 5.18166446685791, + "learning_rate": 9.138256201944003e-06, + "loss": 0.0183, + "step": 15074 + }, + { + "epoch": 1.634323503902862, + "grad_norm": 3.4521801471710205, + "learning_rate": 9.137530828376615e-06, + "loss": 0.0075, + "step": 15075 + }, + { + "epoch": 1.6344319167389418, + "grad_norm": 3.4448082447052, + "learning_rate": 9.136805454809227e-06, + "loss": 0.0672, + "step": 15076 + }, + { + "epoch": 1.6345403295750218, + "grad_norm": 0.07333329319953918, + "learning_rate": 9.136080081241841e-06, + "loss": 0.0006, + "step": 15077 + }, + { + "epoch": 1.6346487424111014, + "grad_norm": 3.416919708251953, + "learning_rate": 9.135354707674453e-06, + "loss": 0.0461, + "step": 15078 + }, + { + "epoch": 1.6347571552471813, + "grad_norm": 1.029934287071228, + "learning_rate": 9.134629334107065e-06, + "loss": 0.0211, + "step": 15079 + }, + { + "epoch": 1.6348655680832611, + "grad_norm": 4.30455207824707, + "learning_rate": 9.133903960539678e-06, + "loss": 0.0345, + "step": 15080 + }, + { + "epoch": 1.6349739809193409, + "grad_norm": 0.6114381551742554, + "learning_rate": 9.133178586972291e-06, + "loss": 0.0037, + "step": 15081 + }, + { + "epoch": 1.6350823937554206, + "grad_norm": 2.2654306888580322, + "learning_rate": 9.132453213404904e-06, + "loss": 0.0095, + "step": 15082 + }, + { + "epoch": 1.6351908065915004, + "grad_norm": 1.8573918342590332, + "learning_rate": 9.131727839837517e-06, + "loss": 0.032, + "step": 15083 + }, + { + "epoch": 1.6352992194275804, + "grad_norm": 0.15959972143173218, + "learning_rate": 9.13100246627013e-06, + "loss": 0.0021, + "step": 15084 + }, + { + "epoch": 1.63540763226366, + "grad_norm": 0.6195523738861084, + "learning_rate": 9.130277092702743e-06, + "loss": 0.002, + "step": 15085 + }, + { + "epoch": 1.63551604509974, + "grad_norm": 1.7004053592681885, + "learning_rate": 9.129551719135356e-06, + "loss": 0.0271, + "step": 15086 + }, + { + "epoch": 1.6356244579358195, + "grad_norm": 1.3161334991455078, + "learning_rate": 9.128826345567968e-06, + "loss": 0.0193, + "step": 15087 + }, + { + "epoch": 1.6357328707718994, + "grad_norm": 3.146623373031616, + "learning_rate": 9.128100972000582e-06, + "loss": 0.0339, + "step": 15088 + }, + { + "epoch": 1.6358412836079792, + "grad_norm": 2.6597037315368652, + "learning_rate": 9.127375598433194e-06, + "loss": 0.0221, + "step": 15089 + }, + { + "epoch": 1.635949696444059, + "grad_norm": 2.497279644012451, + "learning_rate": 9.126650224865806e-06, + "loss": 0.0462, + "step": 15090 + }, + { + "epoch": 1.6360581092801387, + "grad_norm": 1.8871581554412842, + "learning_rate": 9.12592485129842e-06, + "loss": 0.0447, + "step": 15091 + }, + { + "epoch": 1.6361665221162185, + "grad_norm": 0.2928205132484436, + "learning_rate": 9.125199477731033e-06, + "loss": 0.0037, + "step": 15092 + }, + { + "epoch": 1.6362749349522985, + "grad_norm": 4.755125522613525, + "learning_rate": 9.124474104163646e-06, + "loss": 0.0081, + "step": 15093 + }, + { + "epoch": 1.636383347788378, + "grad_norm": 2.205085039138794, + "learning_rate": 9.123748730596258e-06, + "loss": 0.0098, + "step": 15094 + }, + { + "epoch": 1.636491760624458, + "grad_norm": 1.1984814405441284, + "learning_rate": 9.12302335702887e-06, + "loss": 0.0096, + "step": 15095 + }, + { + "epoch": 1.6366001734605378, + "grad_norm": 2.5926053524017334, + "learning_rate": 9.122297983461484e-06, + "loss": 0.0333, + "step": 15096 + }, + { + "epoch": 1.6367085862966175, + "grad_norm": 2.1576786041259766, + "learning_rate": 9.121572609894096e-06, + "loss": 0.0154, + "step": 15097 + }, + { + "epoch": 1.6368169991326973, + "grad_norm": 1.585178017616272, + "learning_rate": 9.120847236326708e-06, + "loss": 0.0082, + "step": 15098 + }, + { + "epoch": 1.636925411968777, + "grad_norm": 0.15688389539718628, + "learning_rate": 9.120121862759322e-06, + "loss": 0.0019, + "step": 15099 + }, + { + "epoch": 1.637033824804857, + "grad_norm": 0.22421041131019592, + "learning_rate": 9.119396489191936e-06, + "loss": 0.0011, + "step": 15100 + }, + { + "epoch": 1.6371422376409366, + "grad_norm": 2.2419469356536865, + "learning_rate": 9.118671115624548e-06, + "loss": 0.0223, + "step": 15101 + }, + { + "epoch": 1.6372506504770166, + "grad_norm": 5.815739154815674, + "learning_rate": 9.11794574205716e-06, + "loss": 0.0446, + "step": 15102 + }, + { + "epoch": 1.637359063313096, + "grad_norm": 0.5124808549880981, + "learning_rate": 9.117220368489774e-06, + "loss": 0.0041, + "step": 15103 + }, + { + "epoch": 1.637467476149176, + "grad_norm": 2.5934560298919678, + "learning_rate": 9.116494994922386e-06, + "loss": 0.0203, + "step": 15104 + }, + { + "epoch": 1.6375758889852559, + "grad_norm": 0.6967928409576416, + "learning_rate": 9.115769621354998e-06, + "loss": 0.0074, + "step": 15105 + }, + { + "epoch": 1.6376843018213356, + "grad_norm": 1.6457650661468506, + "learning_rate": 9.11504424778761e-06, + "loss": 0.0123, + "step": 15106 + }, + { + "epoch": 1.6377927146574154, + "grad_norm": 0.9183767437934875, + "learning_rate": 9.114318874220224e-06, + "loss": 0.0058, + "step": 15107 + }, + { + "epoch": 1.6379011274934951, + "grad_norm": 1.9182523488998413, + "learning_rate": 9.113593500652836e-06, + "loss": 0.0135, + "step": 15108 + }, + { + "epoch": 1.6380095403295751, + "grad_norm": 1.175776481628418, + "learning_rate": 9.11286812708545e-06, + "loss": 0.0096, + "step": 15109 + }, + { + "epoch": 1.6381179531656547, + "grad_norm": 0.30173537135124207, + "learning_rate": 9.112142753518062e-06, + "loss": 0.0014, + "step": 15110 + }, + { + "epoch": 1.6382263660017347, + "grad_norm": 2.5809929370880127, + "learning_rate": 9.111417379950676e-06, + "loss": 0.0754, + "step": 15111 + }, + { + "epoch": 1.6383347788378144, + "grad_norm": 0.7678766250610352, + "learning_rate": 9.110692006383288e-06, + "loss": 0.0077, + "step": 15112 + }, + { + "epoch": 1.6384431916738942, + "grad_norm": 4.103789806365967, + "learning_rate": 9.1099666328159e-06, + "loss": 0.0161, + "step": 15113 + }, + { + "epoch": 1.638551604509974, + "grad_norm": 0.6761186122894287, + "learning_rate": 9.109241259248514e-06, + "loss": 0.0055, + "step": 15114 + }, + { + "epoch": 1.6386600173460537, + "grad_norm": 0.7412194609642029, + "learning_rate": 9.108515885681126e-06, + "loss": 0.0053, + "step": 15115 + }, + { + "epoch": 1.6387684301821337, + "grad_norm": 3.004833459854126, + "learning_rate": 9.107790512113738e-06, + "loss": 0.0154, + "step": 15116 + }, + { + "epoch": 1.6388768430182132, + "grad_norm": 1.799424409866333, + "learning_rate": 9.107065138546352e-06, + "loss": 0.0083, + "step": 15117 + }, + { + "epoch": 1.6389852558542932, + "grad_norm": 2.6293399333953857, + "learning_rate": 9.106339764978966e-06, + "loss": 0.0251, + "step": 15118 + }, + { + "epoch": 1.639093668690373, + "grad_norm": 2.408799409866333, + "learning_rate": 9.105614391411578e-06, + "loss": 0.0108, + "step": 15119 + }, + { + "epoch": 1.6392020815264527, + "grad_norm": 4.233521461486816, + "learning_rate": 9.10488901784419e-06, + "loss": 0.063, + "step": 15120 + }, + { + "epoch": 1.6393104943625325, + "grad_norm": 2.191493511199951, + "learning_rate": 9.104163644276803e-06, + "loss": 0.0376, + "step": 15121 + }, + { + "epoch": 1.6394189071986123, + "grad_norm": 3.284055233001709, + "learning_rate": 9.103438270709416e-06, + "loss": 0.0249, + "step": 15122 + }, + { + "epoch": 1.6395273200346923, + "grad_norm": 3.202569007873535, + "learning_rate": 9.102712897142029e-06, + "loss": 0.0449, + "step": 15123 + }, + { + "epoch": 1.6396357328707718, + "grad_norm": 2.96220064163208, + "learning_rate": 9.10198752357464e-06, + "loss": 0.01, + "step": 15124 + }, + { + "epoch": 1.6397441457068518, + "grad_norm": 3.4101955890655518, + "learning_rate": 9.101262150007255e-06, + "loss": 0.0546, + "step": 15125 + }, + { + "epoch": 1.6398525585429313, + "grad_norm": 0.359641432762146, + "learning_rate": 9.100536776439868e-06, + "loss": 0.0026, + "step": 15126 + }, + { + "epoch": 1.6399609713790113, + "grad_norm": 3.0367655754089355, + "learning_rate": 9.09981140287248e-06, + "loss": 0.0708, + "step": 15127 + }, + { + "epoch": 1.640069384215091, + "grad_norm": 0.9641121029853821, + "learning_rate": 9.099086029305093e-06, + "loss": 0.0357, + "step": 15128 + }, + { + "epoch": 1.6401777970511708, + "grad_norm": 2.1878268718719482, + "learning_rate": 9.098360655737707e-06, + "loss": 0.0291, + "step": 15129 + }, + { + "epoch": 1.6402862098872506, + "grad_norm": 0.6336737871170044, + "learning_rate": 9.097635282170319e-06, + "loss": 0.0064, + "step": 15130 + }, + { + "epoch": 1.6403946227233304, + "grad_norm": 0.5660673975944519, + "learning_rate": 9.09690990860293e-06, + "loss": 0.0035, + "step": 15131 + }, + { + "epoch": 1.6405030355594103, + "grad_norm": 1.742060661315918, + "learning_rate": 9.096184535035543e-06, + "loss": 0.0594, + "step": 15132 + }, + { + "epoch": 1.64061144839549, + "grad_norm": 5.379342555999756, + "learning_rate": 9.095459161468157e-06, + "loss": 0.0161, + "step": 15133 + }, + { + "epoch": 1.6407198612315699, + "grad_norm": 0.09764901548624039, + "learning_rate": 9.094733787900769e-06, + "loss": 0.0012, + "step": 15134 + }, + { + "epoch": 1.6408282740676496, + "grad_norm": 1.385701060295105, + "learning_rate": 9.094008414333383e-06, + "loss": 0.0194, + "step": 15135 + }, + { + "epoch": 1.6409366869037294, + "grad_norm": 3.1719040870666504, + "learning_rate": 9.093283040765995e-06, + "loss": 0.0372, + "step": 15136 + }, + { + "epoch": 1.6410450997398092, + "grad_norm": 0.08460439741611481, + "learning_rate": 9.092557667198609e-06, + "loss": 0.0007, + "step": 15137 + }, + { + "epoch": 1.641153512575889, + "grad_norm": 2.151921033859253, + "learning_rate": 9.091832293631221e-06, + "loss": 0.0399, + "step": 15138 + }, + { + "epoch": 1.641261925411969, + "grad_norm": 3.2118866443634033, + "learning_rate": 9.091106920063833e-06, + "loss": 0.0202, + "step": 15139 + }, + { + "epoch": 1.6413703382480485, + "grad_norm": 2.1346213817596436, + "learning_rate": 9.090381546496447e-06, + "loss": 0.0242, + "step": 15140 + }, + { + "epoch": 1.6414787510841284, + "grad_norm": 0.7145130038261414, + "learning_rate": 9.089656172929059e-06, + "loss": 0.0056, + "step": 15141 + }, + { + "epoch": 1.6415871639202082, + "grad_norm": 4.639373779296875, + "learning_rate": 9.088930799361671e-06, + "loss": 0.0146, + "step": 15142 + }, + { + "epoch": 1.641695576756288, + "grad_norm": 4.049060821533203, + "learning_rate": 9.088205425794285e-06, + "loss": 0.0072, + "step": 15143 + }, + { + "epoch": 1.6418039895923677, + "grad_norm": 2.096519947052002, + "learning_rate": 9.087480052226899e-06, + "loss": 0.0152, + "step": 15144 + }, + { + "epoch": 1.6419124024284475, + "grad_norm": 3.384599208831787, + "learning_rate": 9.086754678659511e-06, + "loss": 0.0322, + "step": 15145 + }, + { + "epoch": 1.6420208152645275, + "grad_norm": 9.199766159057617, + "learning_rate": 9.086029305092123e-06, + "loss": 0.0486, + "step": 15146 + }, + { + "epoch": 1.642129228100607, + "grad_norm": 5.301544666290283, + "learning_rate": 9.085303931524735e-06, + "loss": 0.0919, + "step": 15147 + }, + { + "epoch": 1.642237640936687, + "grad_norm": 0.9186327457427979, + "learning_rate": 9.084578557957349e-06, + "loss": 0.0105, + "step": 15148 + }, + { + "epoch": 1.6423460537727665, + "grad_norm": 6.41105842590332, + "learning_rate": 9.083853184389961e-06, + "loss": 0.0753, + "step": 15149 + }, + { + "epoch": 1.6424544666088465, + "grad_norm": 1.5843231678009033, + "learning_rate": 9.083127810822573e-06, + "loss": 0.0123, + "step": 15150 + }, + { + "epoch": 1.6425628794449263, + "grad_norm": 1.8787437677383423, + "learning_rate": 9.082402437255187e-06, + "loss": 0.0048, + "step": 15151 + }, + { + "epoch": 1.642671292281006, + "grad_norm": 1.7385185956954956, + "learning_rate": 9.081677063687801e-06, + "loss": 0.0113, + "step": 15152 + }, + { + "epoch": 1.6427797051170858, + "grad_norm": 1.7829746007919312, + "learning_rate": 9.080951690120413e-06, + "loss": 0.0257, + "step": 15153 + }, + { + "epoch": 1.6428881179531656, + "grad_norm": 4.595656871795654, + "learning_rate": 9.080226316553025e-06, + "loss": 0.0094, + "step": 15154 + }, + { + "epoch": 1.6429965307892456, + "grad_norm": 1.7707258462905884, + "learning_rate": 9.07950094298564e-06, + "loss": 0.0144, + "step": 15155 + }, + { + "epoch": 1.643104943625325, + "grad_norm": 1.351307988166809, + "learning_rate": 9.078775569418251e-06, + "loss": 0.0059, + "step": 15156 + }, + { + "epoch": 1.643213356461405, + "grad_norm": 0.5211616158485413, + "learning_rate": 9.078050195850863e-06, + "loss": 0.0028, + "step": 15157 + }, + { + "epoch": 1.6433217692974849, + "grad_norm": 3.408897876739502, + "learning_rate": 9.077324822283476e-06, + "loss": 0.0612, + "step": 15158 + }, + { + "epoch": 1.6434301821335646, + "grad_norm": 0.8481748700141907, + "learning_rate": 9.07659944871609e-06, + "loss": 0.0133, + "step": 15159 + }, + { + "epoch": 1.6435385949696444, + "grad_norm": 7.3813796043396, + "learning_rate": 9.075874075148702e-06, + "loss": 0.0267, + "step": 15160 + }, + { + "epoch": 1.6436470078057241, + "grad_norm": 2.286006450653076, + "learning_rate": 9.075148701581315e-06, + "loss": 0.0079, + "step": 15161 + }, + { + "epoch": 1.6437554206418041, + "grad_norm": 0.9662683010101318, + "learning_rate": 9.074423328013928e-06, + "loss": 0.0107, + "step": 15162 + }, + { + "epoch": 1.6438638334778837, + "grad_norm": 6.755492687225342, + "learning_rate": 9.073697954446541e-06, + "loss": 0.0221, + "step": 15163 + }, + { + "epoch": 1.6439722463139637, + "grad_norm": 1.21043860912323, + "learning_rate": 9.072972580879154e-06, + "loss": 0.0052, + "step": 15164 + }, + { + "epoch": 1.6440806591500432, + "grad_norm": 5.246947765350342, + "learning_rate": 9.072247207311766e-06, + "loss": 0.04, + "step": 15165 + }, + { + "epoch": 1.6441890719861232, + "grad_norm": 0.6335392594337463, + "learning_rate": 9.07152183374438e-06, + "loss": 0.0061, + "step": 15166 + }, + { + "epoch": 1.644297484822203, + "grad_norm": 4.632327556610107, + "learning_rate": 9.070796460176992e-06, + "loss": 0.0567, + "step": 15167 + }, + { + "epoch": 1.6444058976582827, + "grad_norm": 3.637938976287842, + "learning_rate": 9.070071086609604e-06, + "loss": 0.043, + "step": 15168 + }, + { + "epoch": 1.6445143104943627, + "grad_norm": 4.170497894287109, + "learning_rate": 9.069345713042218e-06, + "loss": 0.0552, + "step": 15169 + }, + { + "epoch": 1.6446227233304422, + "grad_norm": 2.6395065784454346, + "learning_rate": 9.068620339474831e-06, + "loss": 0.0274, + "step": 15170 + }, + { + "epoch": 1.6447311361665222, + "grad_norm": 8.513704299926758, + "learning_rate": 9.067894965907444e-06, + "loss": 0.0504, + "step": 15171 + }, + { + "epoch": 1.6448395490026018, + "grad_norm": 2.979362726211548, + "learning_rate": 9.067169592340056e-06, + "loss": 0.0531, + "step": 15172 + }, + { + "epoch": 1.6449479618386817, + "grad_norm": 4.04323148727417, + "learning_rate": 9.066444218772668e-06, + "loss": 0.0455, + "step": 15173 + }, + { + "epoch": 1.6450563746747615, + "grad_norm": 1.513139247894287, + "learning_rate": 9.065718845205282e-06, + "loss": 0.013, + "step": 15174 + }, + { + "epoch": 1.6451647875108413, + "grad_norm": 2.51208233833313, + "learning_rate": 9.064993471637894e-06, + "loss": 0.0294, + "step": 15175 + }, + { + "epoch": 1.645273200346921, + "grad_norm": 0.14568549394607544, + "learning_rate": 9.064268098070506e-06, + "loss": 0.002, + "step": 15176 + }, + { + "epoch": 1.6453816131830008, + "grad_norm": 10.58383846282959, + "learning_rate": 9.06354272450312e-06, + "loss": 0.0194, + "step": 15177 + }, + { + "epoch": 1.6454900260190808, + "grad_norm": 3.2892262935638428, + "learning_rate": 9.062817350935734e-06, + "loss": 0.0691, + "step": 15178 + }, + { + "epoch": 1.6455984388551603, + "grad_norm": 2.1746203899383545, + "learning_rate": 9.062091977368346e-06, + "loss": 0.0442, + "step": 15179 + }, + { + "epoch": 1.6457068516912403, + "grad_norm": 1.0952889919281006, + "learning_rate": 9.061366603800958e-06, + "loss": 0.0078, + "step": 15180 + }, + { + "epoch": 1.64581526452732, + "grad_norm": 0.7363060712814331, + "learning_rate": 9.060641230233572e-06, + "loss": 0.0239, + "step": 15181 + }, + { + "epoch": 1.6459236773633998, + "grad_norm": 2.42158579826355, + "learning_rate": 9.059915856666184e-06, + "loss": 0.0189, + "step": 15182 + }, + { + "epoch": 1.6460320901994796, + "grad_norm": 5.932661056518555, + "learning_rate": 9.059190483098796e-06, + "loss": 0.0398, + "step": 15183 + }, + { + "epoch": 1.6461405030355594, + "grad_norm": 2.8384175300598145, + "learning_rate": 9.058465109531408e-06, + "loss": 0.0351, + "step": 15184 + }, + { + "epoch": 1.6462489158716394, + "grad_norm": 2.8317244052886963, + "learning_rate": 9.057739735964022e-06, + "loss": 0.0459, + "step": 15185 + }, + { + "epoch": 1.646357328707719, + "grad_norm": 0.3818444609642029, + "learning_rate": 9.057014362396636e-06, + "loss": 0.0056, + "step": 15186 + }, + { + "epoch": 1.6464657415437989, + "grad_norm": 1.914783000946045, + "learning_rate": 9.056288988829248e-06, + "loss": 0.0205, + "step": 15187 + }, + { + "epoch": 1.6465741543798784, + "grad_norm": 0.9120813012123108, + "learning_rate": 9.05556361526186e-06, + "loss": 0.0086, + "step": 15188 + }, + { + "epoch": 1.6466825672159584, + "grad_norm": 2.7358386516571045, + "learning_rate": 9.054838241694474e-06, + "loss": 0.0134, + "step": 15189 + }, + { + "epoch": 1.6467909800520382, + "grad_norm": 0.17251130938529968, + "learning_rate": 9.054112868127086e-06, + "loss": 0.0024, + "step": 15190 + }, + { + "epoch": 1.646899392888118, + "grad_norm": 0.8544442057609558, + "learning_rate": 9.053387494559698e-06, + "loss": 0.0048, + "step": 15191 + }, + { + "epoch": 1.6470078057241977, + "grad_norm": 0.4739527106285095, + "learning_rate": 9.052662120992312e-06, + "loss": 0.0036, + "step": 15192 + }, + { + "epoch": 1.6471162185602775, + "grad_norm": 6.114862442016602, + "learning_rate": 9.051936747424924e-06, + "loss": 0.0476, + "step": 15193 + }, + { + "epoch": 1.6472246313963574, + "grad_norm": 0.9074092507362366, + "learning_rate": 9.051211373857536e-06, + "loss": 0.0059, + "step": 15194 + }, + { + "epoch": 1.647333044232437, + "grad_norm": 3.387421131134033, + "learning_rate": 9.05048600029015e-06, + "loss": 0.0157, + "step": 15195 + }, + { + "epoch": 1.647441457068517, + "grad_norm": 2.429966688156128, + "learning_rate": 9.049760626722764e-06, + "loss": 0.0274, + "step": 15196 + }, + { + "epoch": 1.6475498699045967, + "grad_norm": 4.204904079437256, + "learning_rate": 9.049035253155376e-06, + "loss": 0.0077, + "step": 15197 + }, + { + "epoch": 1.6476582827406765, + "grad_norm": 1.249934434890747, + "learning_rate": 9.048309879587988e-06, + "loss": 0.0168, + "step": 15198 + }, + { + "epoch": 1.6477666955767563, + "grad_norm": 10.826662063598633, + "learning_rate": 9.0475845060206e-06, + "loss": 0.0561, + "step": 15199 + }, + { + "epoch": 1.647875108412836, + "grad_norm": 0.1603817492723465, + "learning_rate": 9.046859132453214e-06, + "loss": 0.0023, + "step": 15200 + }, + { + "epoch": 1.647983521248916, + "grad_norm": 2.0891640186309814, + "learning_rate": 9.046133758885827e-06, + "loss": 0.0234, + "step": 15201 + }, + { + "epoch": 1.6480919340849955, + "grad_norm": 4.998790264129639, + "learning_rate": 9.045408385318439e-06, + "loss": 0.0459, + "step": 15202 + }, + { + "epoch": 1.6482003469210755, + "grad_norm": 2.3793163299560547, + "learning_rate": 9.044683011751053e-06, + "loss": 0.0264, + "step": 15203 + }, + { + "epoch": 1.6483087597571553, + "grad_norm": 0.6744216084480286, + "learning_rate": 9.043957638183666e-06, + "loss": 0.0038, + "step": 15204 + }, + { + "epoch": 1.648417172593235, + "grad_norm": 0.7240985631942749, + "learning_rate": 9.043232264616279e-06, + "loss": 0.0087, + "step": 15205 + }, + { + "epoch": 1.6485255854293148, + "grad_norm": 0.8411048054695129, + "learning_rate": 9.04250689104889e-06, + "loss": 0.0077, + "step": 15206 + }, + { + "epoch": 1.6486339982653946, + "grad_norm": 2.500577211380005, + "learning_rate": 9.041781517481505e-06, + "loss": 0.0109, + "step": 15207 + }, + { + "epoch": 1.6487424111014746, + "grad_norm": 9.377569198608398, + "learning_rate": 9.041056143914117e-06, + "loss": 0.0199, + "step": 15208 + }, + { + "epoch": 1.648850823937554, + "grad_norm": 5.412203311920166, + "learning_rate": 9.040330770346729e-06, + "loss": 0.0351, + "step": 15209 + }, + { + "epoch": 1.648959236773634, + "grad_norm": 6.351574897766113, + "learning_rate": 9.039605396779341e-06, + "loss": 0.0327, + "step": 15210 + }, + { + "epoch": 1.6490676496097136, + "grad_norm": 3.414163589477539, + "learning_rate": 9.038880023211955e-06, + "loss": 0.0123, + "step": 15211 + }, + { + "epoch": 1.6491760624457936, + "grad_norm": 0.4142901599407196, + "learning_rate": 9.038154649644569e-06, + "loss": 0.0028, + "step": 15212 + }, + { + "epoch": 1.6492844752818734, + "grad_norm": 0.5716197490692139, + "learning_rate": 9.03742927607718e-06, + "loss": 0.0181, + "step": 15213 + }, + { + "epoch": 1.6493928881179531, + "grad_norm": 4.806109428405762, + "learning_rate": 9.036703902509793e-06, + "loss": 0.0216, + "step": 15214 + }, + { + "epoch": 1.649501300954033, + "grad_norm": 0.4645998775959015, + "learning_rate": 9.035978528942407e-06, + "loss": 0.0064, + "step": 15215 + }, + { + "epoch": 1.6496097137901127, + "grad_norm": 0.4346325397491455, + "learning_rate": 9.035253155375019e-06, + "loss": 0.0036, + "step": 15216 + }, + { + "epoch": 1.6497181266261927, + "grad_norm": 1.3776805400848389, + "learning_rate": 9.034527781807631e-06, + "loss": 0.0101, + "step": 15217 + }, + { + "epoch": 1.6498265394622722, + "grad_norm": 0.9945998787879944, + "learning_rate": 9.033802408240245e-06, + "loss": 0.0044, + "step": 15218 + }, + { + "epoch": 1.6499349522983522, + "grad_norm": 0.061900146305561066, + "learning_rate": 9.033077034672857e-06, + "loss": 0.0007, + "step": 15219 + }, + { + "epoch": 1.650043365134432, + "grad_norm": 3.2881104946136475, + "learning_rate": 9.03235166110547e-06, + "loss": 0.0239, + "step": 15220 + }, + { + "epoch": 1.6501517779705117, + "grad_norm": 0.40724050998687744, + "learning_rate": 9.031626287538083e-06, + "loss": 0.0039, + "step": 15221 + }, + { + "epoch": 1.6502601908065915, + "grad_norm": 5.8708600997924805, + "learning_rate": 9.030900913970697e-06, + "loss": 0.0265, + "step": 15222 + }, + { + "epoch": 1.6503686036426712, + "grad_norm": 0.4629899859428406, + "learning_rate": 9.030175540403309e-06, + "loss": 0.002, + "step": 15223 + }, + { + "epoch": 1.6504770164787512, + "grad_norm": 0.5286461710929871, + "learning_rate": 9.029450166835921e-06, + "loss": 0.0055, + "step": 15224 + }, + { + "epoch": 1.6505854293148308, + "grad_norm": 1.8261916637420654, + "learning_rate": 9.028724793268533e-06, + "loss": 0.0165, + "step": 15225 + }, + { + "epoch": 1.6506938421509108, + "grad_norm": 1.4119880199432373, + "learning_rate": 9.027999419701147e-06, + "loss": 0.0088, + "step": 15226 + }, + { + "epoch": 1.6508022549869905, + "grad_norm": 2.412759780883789, + "learning_rate": 9.02727404613376e-06, + "loss": 0.023, + "step": 15227 + }, + { + "epoch": 1.6509106678230703, + "grad_norm": 3.8170483112335205, + "learning_rate": 9.026548672566371e-06, + "loss": 0.0297, + "step": 15228 + }, + { + "epoch": 1.65101908065915, + "grad_norm": 0.43522340059280396, + "learning_rate": 9.025823298998985e-06, + "loss": 0.0031, + "step": 15229 + }, + { + "epoch": 1.6511274934952298, + "grad_norm": 0.5490811467170715, + "learning_rate": 9.025097925431599e-06, + "loss": 0.0051, + "step": 15230 + }, + { + "epoch": 1.6512359063313098, + "grad_norm": 1.6868668794631958, + "learning_rate": 9.024372551864211e-06, + "loss": 0.0045, + "step": 15231 + }, + { + "epoch": 1.6513443191673893, + "grad_norm": 20.65224266052246, + "learning_rate": 9.023647178296823e-06, + "loss": 0.0418, + "step": 15232 + }, + { + "epoch": 1.6514527320034693, + "grad_norm": 2.819880723953247, + "learning_rate": 9.022921804729437e-06, + "loss": 0.0576, + "step": 15233 + }, + { + "epoch": 1.6515611448395489, + "grad_norm": 4.552738666534424, + "learning_rate": 9.02219643116205e-06, + "loss": 0.0139, + "step": 15234 + }, + { + "epoch": 1.6516695576756288, + "grad_norm": 1.0433270931243896, + "learning_rate": 9.021471057594661e-06, + "loss": 0.0022, + "step": 15235 + }, + { + "epoch": 1.6517779705117086, + "grad_norm": 3.407832384109497, + "learning_rate": 9.020745684027275e-06, + "loss": 0.0445, + "step": 15236 + }, + { + "epoch": 1.6518863833477884, + "grad_norm": 3.3478047847747803, + "learning_rate": 9.020020310459887e-06, + "loss": 0.0412, + "step": 15237 + }, + { + "epoch": 1.6519947961838681, + "grad_norm": 2.8801004886627197, + "learning_rate": 9.019294936892501e-06, + "loss": 0.0068, + "step": 15238 + }, + { + "epoch": 1.652103209019948, + "grad_norm": 0.3381327986717224, + "learning_rate": 9.018569563325113e-06, + "loss": 0.0013, + "step": 15239 + }, + { + "epoch": 1.6522116218560279, + "grad_norm": 2.5226356983184814, + "learning_rate": 9.017844189757726e-06, + "loss": 0.0125, + "step": 15240 + }, + { + "epoch": 1.6523200346921074, + "grad_norm": 2.3583438396453857, + "learning_rate": 9.01711881619034e-06, + "loss": 0.0116, + "step": 15241 + }, + { + "epoch": 1.6524284475281874, + "grad_norm": 3.1075055599212646, + "learning_rate": 9.016393442622952e-06, + "loss": 0.0257, + "step": 15242 + }, + { + "epoch": 1.6525368603642672, + "grad_norm": 0.5741642117500305, + "learning_rate": 9.015668069055564e-06, + "loss": 0.0089, + "step": 15243 + }, + { + "epoch": 1.652645273200347, + "grad_norm": 1.9263300895690918, + "learning_rate": 9.014942695488178e-06, + "loss": 0.054, + "step": 15244 + }, + { + "epoch": 1.6527536860364267, + "grad_norm": 0.9971808195114136, + "learning_rate": 9.01421732192079e-06, + "loss": 0.0054, + "step": 15245 + }, + { + "epoch": 1.6528620988725065, + "grad_norm": 3.346095085144043, + "learning_rate": 9.013491948353402e-06, + "loss": 0.0803, + "step": 15246 + }, + { + "epoch": 1.6529705117085864, + "grad_norm": 2.8511555194854736, + "learning_rate": 9.012766574786016e-06, + "loss": 0.0143, + "step": 15247 + }, + { + "epoch": 1.653078924544666, + "grad_norm": 5.474279403686523, + "learning_rate": 9.01204120121863e-06, + "loss": 0.0509, + "step": 15248 + }, + { + "epoch": 1.653187337380746, + "grad_norm": 0.931442141532898, + "learning_rate": 9.011315827651242e-06, + "loss": 0.0034, + "step": 15249 + }, + { + "epoch": 1.6532957502168255, + "grad_norm": 0.2390918731689453, + "learning_rate": 9.010590454083854e-06, + "loss": 0.0024, + "step": 15250 + }, + { + "epoch": 1.6534041630529055, + "grad_norm": 1.6490943431854248, + "learning_rate": 9.009865080516466e-06, + "loss": 0.042, + "step": 15251 + }, + { + "epoch": 1.6535125758889853, + "grad_norm": 13.747880935668945, + "learning_rate": 9.00913970694908e-06, + "loss": 0.0685, + "step": 15252 + }, + { + "epoch": 1.653620988725065, + "grad_norm": 5.462089538574219, + "learning_rate": 9.008414333381692e-06, + "loss": 0.0262, + "step": 15253 + }, + { + "epoch": 1.6537294015611448, + "grad_norm": 3.4514384269714355, + "learning_rate": 9.007688959814304e-06, + "loss": 0.0493, + "step": 15254 + }, + { + "epoch": 1.6538378143972245, + "grad_norm": 3.8442959785461426, + "learning_rate": 9.006963586246918e-06, + "loss": 0.0239, + "step": 15255 + }, + { + "epoch": 1.6539462272333045, + "grad_norm": 0.14760857820510864, + "learning_rate": 9.006238212679532e-06, + "loss": 0.0013, + "step": 15256 + }, + { + "epoch": 1.654054640069384, + "grad_norm": 2.425424337387085, + "learning_rate": 9.005512839112144e-06, + "loss": 0.0068, + "step": 15257 + }, + { + "epoch": 1.654163052905464, + "grad_norm": 2.492689609527588, + "learning_rate": 9.004787465544756e-06, + "loss": 0.0327, + "step": 15258 + }, + { + "epoch": 1.6542714657415438, + "grad_norm": 7.911708354949951, + "learning_rate": 9.00406209197737e-06, + "loss": 0.0514, + "step": 15259 + }, + { + "epoch": 1.6543798785776236, + "grad_norm": 1.6999422311782837, + "learning_rate": 9.003336718409982e-06, + "loss": 0.0174, + "step": 15260 + }, + { + "epoch": 1.6544882914137033, + "grad_norm": 6.789894104003906, + "learning_rate": 9.002611344842594e-06, + "loss": 0.0169, + "step": 15261 + }, + { + "epoch": 1.6545967042497831, + "grad_norm": 4.151337146759033, + "learning_rate": 9.001885971275208e-06, + "loss": 0.0187, + "step": 15262 + }, + { + "epoch": 1.654705117085863, + "grad_norm": 1.931434154510498, + "learning_rate": 9.00116059770782e-06, + "loss": 0.0041, + "step": 15263 + }, + { + "epoch": 1.6548135299219426, + "grad_norm": 4.784007549285889, + "learning_rate": 9.000435224140434e-06, + "loss": 0.011, + "step": 15264 + }, + { + "epoch": 1.6549219427580226, + "grad_norm": 6.346812725067139, + "learning_rate": 8.999709850573046e-06, + "loss": 0.0311, + "step": 15265 + }, + { + "epoch": 1.6550303555941024, + "grad_norm": 1.5421512126922607, + "learning_rate": 8.998984477005658e-06, + "loss": 0.056, + "step": 15266 + }, + { + "epoch": 1.6551387684301822, + "grad_norm": 1.347989559173584, + "learning_rate": 8.998259103438272e-06, + "loss": 0.0062, + "step": 15267 + }, + { + "epoch": 1.655247181266262, + "grad_norm": 4.05648946762085, + "learning_rate": 8.997533729870884e-06, + "loss": 0.0564, + "step": 15268 + }, + { + "epoch": 1.6553555941023417, + "grad_norm": 2.030794143676758, + "learning_rate": 8.996808356303496e-06, + "loss": 0.0094, + "step": 15269 + }, + { + "epoch": 1.6554640069384217, + "grad_norm": 8.822395324707031, + "learning_rate": 8.99608298273611e-06, + "loss": 0.0295, + "step": 15270 + }, + { + "epoch": 1.6555724197745012, + "grad_norm": 0.46637555956840515, + "learning_rate": 8.995357609168722e-06, + "loss": 0.0041, + "step": 15271 + }, + { + "epoch": 1.6556808326105812, + "grad_norm": 14.434494018554688, + "learning_rate": 8.994632235601334e-06, + "loss": 0.0548, + "step": 15272 + }, + { + "epoch": 1.6557892454466607, + "grad_norm": 2.6172986030578613, + "learning_rate": 8.993906862033948e-06, + "loss": 0.0262, + "step": 15273 + }, + { + "epoch": 1.6558976582827407, + "grad_norm": 2.337202787399292, + "learning_rate": 8.993181488466562e-06, + "loss": 0.0199, + "step": 15274 + }, + { + "epoch": 1.6560060711188205, + "grad_norm": 0.9593358635902405, + "learning_rate": 8.992456114899174e-06, + "loss": 0.0074, + "step": 15275 + }, + { + "epoch": 1.6561144839549002, + "grad_norm": 2.4350295066833496, + "learning_rate": 8.991730741331786e-06, + "loss": 0.0511, + "step": 15276 + }, + { + "epoch": 1.65622289679098, + "grad_norm": 1.7930923700332642, + "learning_rate": 8.991005367764399e-06, + "loss": 0.0361, + "step": 15277 + }, + { + "epoch": 1.6563313096270598, + "grad_norm": 0.18731532990932465, + "learning_rate": 8.990279994197012e-06, + "loss": 0.0023, + "step": 15278 + }, + { + "epoch": 1.6564397224631398, + "grad_norm": 0.706030011177063, + "learning_rate": 8.989554620629625e-06, + "loss": 0.0224, + "step": 15279 + }, + { + "epoch": 1.6565481352992193, + "grad_norm": 1.8321858644485474, + "learning_rate": 8.988829247062237e-06, + "loss": 0.0355, + "step": 15280 + }, + { + "epoch": 1.6566565481352993, + "grad_norm": 2.6248631477355957, + "learning_rate": 8.98810387349485e-06, + "loss": 0.0363, + "step": 15281 + }, + { + "epoch": 1.656764960971379, + "grad_norm": 4.4939727783203125, + "learning_rate": 8.987378499927464e-06, + "loss": 0.0115, + "step": 15282 + }, + { + "epoch": 1.6568733738074588, + "grad_norm": 0.4942265450954437, + "learning_rate": 8.986653126360077e-06, + "loss": 0.0045, + "step": 15283 + }, + { + "epoch": 1.6569817866435386, + "grad_norm": 0.5015109777450562, + "learning_rate": 8.985927752792689e-06, + "loss": 0.0046, + "step": 15284 + }, + { + "epoch": 1.6570901994796183, + "grad_norm": 9.134953498840332, + "learning_rate": 8.985202379225303e-06, + "loss": 0.0037, + "step": 15285 + }, + { + "epoch": 1.6571986123156983, + "grad_norm": 1.593599557876587, + "learning_rate": 8.984477005657915e-06, + "loss": 0.0238, + "step": 15286 + }, + { + "epoch": 1.6573070251517779, + "grad_norm": 2.5993237495422363, + "learning_rate": 8.983751632090527e-06, + "loss": 0.006, + "step": 15287 + }, + { + "epoch": 1.6574154379878578, + "grad_norm": 2.214791774749756, + "learning_rate": 8.98302625852314e-06, + "loss": 0.0172, + "step": 15288 + }, + { + "epoch": 1.6575238508239376, + "grad_norm": 4.301880836486816, + "learning_rate": 8.982300884955753e-06, + "loss": 0.0278, + "step": 15289 + }, + { + "epoch": 1.6576322636600174, + "grad_norm": 3.5939176082611084, + "learning_rate": 8.981575511388367e-06, + "loss": 0.0099, + "step": 15290 + }, + { + "epoch": 1.6577406764960971, + "grad_norm": 2.0319736003875732, + "learning_rate": 8.980850137820979e-06, + "loss": 0.0518, + "step": 15291 + }, + { + "epoch": 1.657849089332177, + "grad_norm": 2.1076741218566895, + "learning_rate": 8.980124764253591e-06, + "loss": 0.0453, + "step": 15292 + }, + { + "epoch": 1.6579575021682569, + "grad_norm": 2.655526638031006, + "learning_rate": 8.979399390686205e-06, + "loss": 0.0067, + "step": 15293 + }, + { + "epoch": 1.6580659150043364, + "grad_norm": 1.7988383769989014, + "learning_rate": 8.978674017118817e-06, + "loss": 0.0055, + "step": 15294 + }, + { + "epoch": 1.6581743278404164, + "grad_norm": 0.6978223919868469, + "learning_rate": 8.977948643551429e-06, + "loss": 0.004, + "step": 15295 + }, + { + "epoch": 1.658282740676496, + "grad_norm": 0.4294644296169281, + "learning_rate": 8.977223269984043e-06, + "loss": 0.0021, + "step": 15296 + }, + { + "epoch": 1.658391153512576, + "grad_norm": 1.2274912595748901, + "learning_rate": 8.976497896416655e-06, + "loss": 0.0087, + "step": 15297 + }, + { + "epoch": 1.6584995663486557, + "grad_norm": 0.5657258033752441, + "learning_rate": 8.975772522849267e-06, + "loss": 0.0074, + "step": 15298 + }, + { + "epoch": 1.6586079791847355, + "grad_norm": 3.9567923545837402, + "learning_rate": 8.975047149281881e-06, + "loss": 0.0251, + "step": 15299 + }, + { + "epoch": 1.6587163920208152, + "grad_norm": 2.368483066558838, + "learning_rate": 8.974321775714495e-06, + "loss": 0.0239, + "step": 15300 + }, + { + "epoch": 1.658824804856895, + "grad_norm": 3.6191799640655518, + "learning_rate": 8.973596402147107e-06, + "loss": 0.0278, + "step": 15301 + }, + { + "epoch": 1.658933217692975, + "grad_norm": 0.8514920473098755, + "learning_rate": 8.972871028579719e-06, + "loss": 0.0028, + "step": 15302 + }, + { + "epoch": 1.6590416305290545, + "grad_norm": 1.6652626991271973, + "learning_rate": 8.972145655012331e-06, + "loss": 0.0088, + "step": 15303 + }, + { + "epoch": 1.6591500433651345, + "grad_norm": 0.313130259513855, + "learning_rate": 8.971420281444945e-06, + "loss": 0.0034, + "step": 15304 + }, + { + "epoch": 1.6592584562012143, + "grad_norm": 0.2577938437461853, + "learning_rate": 8.970694907877557e-06, + "loss": 0.0027, + "step": 15305 + }, + { + "epoch": 1.659366869037294, + "grad_norm": 9.557720184326172, + "learning_rate": 8.96996953431017e-06, + "loss": 0.0354, + "step": 15306 + }, + { + "epoch": 1.6594752818733738, + "grad_norm": 0.4763588309288025, + "learning_rate": 8.969244160742783e-06, + "loss": 0.0021, + "step": 15307 + }, + { + "epoch": 1.6595836947094535, + "grad_norm": 0.8259401321411133, + "learning_rate": 8.968518787175397e-06, + "loss": 0.0047, + "step": 15308 + }, + { + "epoch": 1.6596921075455335, + "grad_norm": 0.1219048798084259, + "learning_rate": 8.96779341360801e-06, + "loss": 0.001, + "step": 15309 + }, + { + "epoch": 1.659800520381613, + "grad_norm": 3.348257064819336, + "learning_rate": 8.967068040040621e-06, + "loss": 0.0156, + "step": 15310 + }, + { + "epoch": 1.659908933217693, + "grad_norm": 3.3328816890716553, + "learning_rate": 8.966342666473235e-06, + "loss": 0.0073, + "step": 15311 + }, + { + "epoch": 1.6600173460537726, + "grad_norm": 1.664862871170044, + "learning_rate": 8.965617292905847e-06, + "loss": 0.0235, + "step": 15312 + }, + { + "epoch": 1.6601257588898526, + "grad_norm": 1.421972393989563, + "learning_rate": 8.96489191933846e-06, + "loss": 0.0119, + "step": 15313 + }, + { + "epoch": 1.6602341717259324, + "grad_norm": 0.24418599903583527, + "learning_rate": 8.964166545771073e-06, + "loss": 0.0018, + "step": 15314 + }, + { + "epoch": 1.6603425845620121, + "grad_norm": 3.549732208251953, + "learning_rate": 8.963441172203685e-06, + "loss": 0.0161, + "step": 15315 + }, + { + "epoch": 1.660450997398092, + "grad_norm": 5.876729965209961, + "learning_rate": 8.9627157986363e-06, + "loss": 0.0426, + "step": 15316 + }, + { + "epoch": 1.6605594102341716, + "grad_norm": 7.901429653167725, + "learning_rate": 8.961990425068911e-06, + "loss": 0.0623, + "step": 15317 + }, + { + "epoch": 1.6606678230702516, + "grad_norm": 3.7705838680267334, + "learning_rate": 8.961265051501524e-06, + "loss": 0.0431, + "step": 15318 + }, + { + "epoch": 1.6607762359063312, + "grad_norm": 1.8853012323379517, + "learning_rate": 8.960539677934137e-06, + "loss": 0.0149, + "step": 15319 + }, + { + "epoch": 1.6608846487424112, + "grad_norm": 1.1837165355682373, + "learning_rate": 8.95981430436675e-06, + "loss": 0.0153, + "step": 15320 + }, + { + "epoch": 1.660993061578491, + "grad_norm": 1.3985929489135742, + "learning_rate": 8.959088930799362e-06, + "loss": 0.0052, + "step": 15321 + }, + { + "epoch": 1.6611014744145707, + "grad_norm": 4.23245096206665, + "learning_rate": 8.958363557231976e-06, + "loss": 0.0069, + "step": 15322 + }, + { + "epoch": 1.6612098872506504, + "grad_norm": 4.062020778656006, + "learning_rate": 8.957638183664588e-06, + "loss": 0.0157, + "step": 15323 + }, + { + "epoch": 1.6613183000867302, + "grad_norm": 4.6165618896484375, + "learning_rate": 8.9569128100972e-06, + "loss": 0.0702, + "step": 15324 + }, + { + "epoch": 1.6614267129228102, + "grad_norm": 3.261040687561035, + "learning_rate": 8.956187436529814e-06, + "loss": 0.044, + "step": 15325 + }, + { + "epoch": 1.6615351257588897, + "grad_norm": 0.16609078645706177, + "learning_rate": 8.955462062962428e-06, + "loss": 0.001, + "step": 15326 + }, + { + "epoch": 1.6616435385949697, + "grad_norm": 1.4511768817901611, + "learning_rate": 8.95473668939504e-06, + "loss": 0.0358, + "step": 15327 + }, + { + "epoch": 1.6617519514310495, + "grad_norm": 5.714128017425537, + "learning_rate": 8.954011315827652e-06, + "loss": 0.0755, + "step": 15328 + }, + { + "epoch": 1.6618603642671292, + "grad_norm": 3.1608147621154785, + "learning_rate": 8.953285942260264e-06, + "loss": 0.03, + "step": 15329 + }, + { + "epoch": 1.661968777103209, + "grad_norm": 2.7674219608306885, + "learning_rate": 8.952560568692878e-06, + "loss": 0.0448, + "step": 15330 + }, + { + "epoch": 1.6620771899392888, + "grad_norm": 1.1557180881500244, + "learning_rate": 8.95183519512549e-06, + "loss": 0.0064, + "step": 15331 + }, + { + "epoch": 1.6621856027753688, + "grad_norm": 1.6322546005249023, + "learning_rate": 8.951109821558102e-06, + "loss": 0.034, + "step": 15332 + }, + { + "epoch": 1.6622940156114483, + "grad_norm": 0.42907920479774475, + "learning_rate": 8.950384447990716e-06, + "loss": 0.0019, + "step": 15333 + }, + { + "epoch": 1.6624024284475283, + "grad_norm": 0.620978593826294, + "learning_rate": 8.94965907442333e-06, + "loss": 0.0035, + "step": 15334 + }, + { + "epoch": 1.6625108412836078, + "grad_norm": 1.088890790939331, + "learning_rate": 8.948933700855942e-06, + "loss": 0.0065, + "step": 15335 + }, + { + "epoch": 1.6626192541196878, + "grad_norm": 1.8295366764068604, + "learning_rate": 8.948208327288554e-06, + "loss": 0.0188, + "step": 15336 + }, + { + "epoch": 1.6627276669557676, + "grad_norm": 1.8207467794418335, + "learning_rate": 8.947482953721168e-06, + "loss": 0.0254, + "step": 15337 + }, + { + "epoch": 1.6628360797918473, + "grad_norm": 0.11489665508270264, + "learning_rate": 8.94675758015378e-06, + "loss": 0.001, + "step": 15338 + }, + { + "epoch": 1.662944492627927, + "grad_norm": 5.880344867706299, + "learning_rate": 8.946032206586392e-06, + "loss": 0.024, + "step": 15339 + }, + { + "epoch": 1.6630529054640069, + "grad_norm": 2.9526679515838623, + "learning_rate": 8.945306833019006e-06, + "loss": 0.0279, + "step": 15340 + }, + { + "epoch": 1.6631613183000868, + "grad_norm": 6.564674377441406, + "learning_rate": 8.944581459451618e-06, + "loss": 0.043, + "step": 15341 + }, + { + "epoch": 1.6632697311361664, + "grad_norm": 0.3754734992980957, + "learning_rate": 8.943856085884232e-06, + "loss": 0.0018, + "step": 15342 + }, + { + "epoch": 1.6633781439722464, + "grad_norm": 2.438176393508911, + "learning_rate": 8.943130712316844e-06, + "loss": 0.0195, + "step": 15343 + }, + { + "epoch": 1.6634865568083261, + "grad_norm": 0.03752409294247627, + "learning_rate": 8.942405338749456e-06, + "loss": 0.0003, + "step": 15344 + }, + { + "epoch": 1.663594969644406, + "grad_norm": 1.2881066799163818, + "learning_rate": 8.94167996518207e-06, + "loss": 0.0164, + "step": 15345 + }, + { + "epoch": 1.6637033824804857, + "grad_norm": 0.9250105619430542, + "learning_rate": 8.940954591614682e-06, + "loss": 0.0055, + "step": 15346 + }, + { + "epoch": 1.6638117953165654, + "grad_norm": 0.797249972820282, + "learning_rate": 8.940229218047294e-06, + "loss": 0.0026, + "step": 15347 + }, + { + "epoch": 1.6639202081526454, + "grad_norm": 2.1250417232513428, + "learning_rate": 8.939503844479908e-06, + "loss": 0.0036, + "step": 15348 + }, + { + "epoch": 1.664028620988725, + "grad_norm": 1.4425252676010132, + "learning_rate": 8.93877847091252e-06, + "loss": 0.0042, + "step": 15349 + }, + { + "epoch": 1.664137033824805, + "grad_norm": 0.7296822667121887, + "learning_rate": 8.938053097345133e-06, + "loss": 0.0061, + "step": 15350 + }, + { + "epoch": 1.6642454466608847, + "grad_norm": 0.2788282036781311, + "learning_rate": 8.937327723777746e-06, + "loss": 0.0012, + "step": 15351 + }, + { + "epoch": 1.6643538594969645, + "grad_norm": 5.291530609130859, + "learning_rate": 8.93660235021036e-06, + "loss": 0.03, + "step": 15352 + }, + { + "epoch": 1.6644622723330442, + "grad_norm": 0.17528846859931946, + "learning_rate": 8.935876976642972e-06, + "loss": 0.0027, + "step": 15353 + }, + { + "epoch": 1.664570685169124, + "grad_norm": 0.7403987050056458, + "learning_rate": 8.935151603075584e-06, + "loss": 0.0037, + "step": 15354 + }, + { + "epoch": 1.664679098005204, + "grad_norm": 1.359751582145691, + "learning_rate": 8.934426229508197e-06, + "loss": 0.0219, + "step": 15355 + }, + { + "epoch": 1.6647875108412835, + "grad_norm": 1.3544318675994873, + "learning_rate": 8.93370085594081e-06, + "loss": 0.0315, + "step": 15356 + }, + { + "epoch": 1.6648959236773635, + "grad_norm": 7.133899211883545, + "learning_rate": 8.932975482373423e-06, + "loss": 0.0507, + "step": 15357 + }, + { + "epoch": 1.665004336513443, + "grad_norm": 6.469229221343994, + "learning_rate": 8.932250108806035e-06, + "loss": 0.0169, + "step": 15358 + }, + { + "epoch": 1.665112749349523, + "grad_norm": 1.6684094667434692, + "learning_rate": 8.931524735238649e-06, + "loss": 0.0083, + "step": 15359 + }, + { + "epoch": 1.6652211621856028, + "grad_norm": 11.229959487915039, + "learning_rate": 8.930799361671262e-06, + "loss": 0.0111, + "step": 15360 + }, + { + "epoch": 1.6653295750216826, + "grad_norm": 2.6556804180145264, + "learning_rate": 8.930073988103875e-06, + "loss": 0.0267, + "step": 15361 + }, + { + "epoch": 1.6654379878577623, + "grad_norm": 2.189997673034668, + "learning_rate": 8.929348614536487e-06, + "loss": 0.0049, + "step": 15362 + }, + { + "epoch": 1.665546400693842, + "grad_norm": 5.6049933433532715, + "learning_rate": 8.9286232409691e-06, + "loss": 0.0571, + "step": 15363 + }, + { + "epoch": 1.665654813529922, + "grad_norm": 5.482438087463379, + "learning_rate": 8.927897867401713e-06, + "loss": 0.0225, + "step": 15364 + }, + { + "epoch": 1.6657632263660016, + "grad_norm": 1.0037217140197754, + "learning_rate": 8.927172493834325e-06, + "loss": 0.0053, + "step": 15365 + }, + { + "epoch": 1.6658716392020816, + "grad_norm": 5.04372501373291, + "learning_rate": 8.926447120266939e-06, + "loss": 0.0695, + "step": 15366 + }, + { + "epoch": 1.6659800520381614, + "grad_norm": 2.959502935409546, + "learning_rate": 8.92572174669955e-06, + "loss": 0.0279, + "step": 15367 + }, + { + "epoch": 1.6660884648742411, + "grad_norm": 3.71817684173584, + "learning_rate": 8.924996373132165e-06, + "loss": 0.0205, + "step": 15368 + }, + { + "epoch": 1.6661968777103209, + "grad_norm": 0.6701454520225525, + "learning_rate": 8.924270999564777e-06, + "loss": 0.0069, + "step": 15369 + }, + { + "epoch": 1.6663052905464006, + "grad_norm": 0.6639268398284912, + "learning_rate": 8.923545625997389e-06, + "loss": 0.0041, + "step": 15370 + }, + { + "epoch": 1.6664137033824806, + "grad_norm": 0.7200480699539185, + "learning_rate": 8.922820252430003e-06, + "loss": 0.0022, + "step": 15371 + }, + { + "epoch": 1.6665221162185602, + "grad_norm": 2.3662145137786865, + "learning_rate": 8.922094878862615e-06, + "loss": 0.0156, + "step": 15372 + }, + { + "epoch": 1.6666305290546402, + "grad_norm": 0.8614923357963562, + "learning_rate": 8.921369505295227e-06, + "loss": 0.0045, + "step": 15373 + }, + { + "epoch": 1.66673894189072, + "grad_norm": 5.795076847076416, + "learning_rate": 8.920644131727841e-06, + "loss": 0.0261, + "step": 15374 + }, + { + "epoch": 1.6668473547267997, + "grad_norm": 0.5917428135871887, + "learning_rate": 8.919918758160453e-06, + "loss": 0.003, + "step": 15375 + }, + { + "epoch": 1.6669557675628794, + "grad_norm": 7.046903610229492, + "learning_rate": 8.919193384593065e-06, + "loss": 0.0666, + "step": 15376 + }, + { + "epoch": 1.6670641803989592, + "grad_norm": 1.0543233156204224, + "learning_rate": 8.918468011025679e-06, + "loss": 0.018, + "step": 15377 + }, + { + "epoch": 1.6671725932350392, + "grad_norm": 2.320920467376709, + "learning_rate": 8.917742637458293e-06, + "loss": 0.032, + "step": 15378 + }, + { + "epoch": 1.6672810060711187, + "grad_norm": 3.119009256362915, + "learning_rate": 8.917017263890905e-06, + "loss": 0.0338, + "step": 15379 + }, + { + "epoch": 1.6673894189071987, + "grad_norm": 0.41496390104293823, + "learning_rate": 8.916291890323517e-06, + "loss": 0.0028, + "step": 15380 + }, + { + "epoch": 1.6674978317432783, + "grad_norm": 7.866201877593994, + "learning_rate": 8.91556651675613e-06, + "loss": 0.0726, + "step": 15381 + }, + { + "epoch": 1.6676062445793582, + "grad_norm": 0.8125846982002258, + "learning_rate": 8.914841143188743e-06, + "loss": 0.0042, + "step": 15382 + }, + { + "epoch": 1.667714657415438, + "grad_norm": 4.934633255004883, + "learning_rate": 8.914115769621355e-06, + "loss": 0.0156, + "step": 15383 + }, + { + "epoch": 1.6678230702515178, + "grad_norm": 1.9602383375167847, + "learning_rate": 8.913390396053967e-06, + "loss": 0.0227, + "step": 15384 + }, + { + "epoch": 1.6679314830875975, + "grad_norm": 0.2071702480316162, + "learning_rate": 8.912665022486581e-06, + "loss": 0.0022, + "step": 15385 + }, + { + "epoch": 1.6680398959236773, + "grad_norm": 4.017977237701416, + "learning_rate": 8.911939648919195e-06, + "loss": 0.0382, + "step": 15386 + }, + { + "epoch": 1.6681483087597573, + "grad_norm": 7.471292495727539, + "learning_rate": 8.911214275351807e-06, + "loss": 0.0262, + "step": 15387 + }, + { + "epoch": 1.6682567215958368, + "grad_norm": 3.2331340312957764, + "learning_rate": 8.91048890178442e-06, + "loss": 0.0493, + "step": 15388 + }, + { + "epoch": 1.6683651344319168, + "grad_norm": 0.5087749361991882, + "learning_rate": 8.909763528217033e-06, + "loss": 0.0042, + "step": 15389 + }, + { + "epoch": 1.6684735472679966, + "grad_norm": 2.278710126876831, + "learning_rate": 8.909038154649645e-06, + "loss": 0.0277, + "step": 15390 + }, + { + "epoch": 1.6685819601040763, + "grad_norm": 0.5571638941764832, + "learning_rate": 8.908312781082257e-06, + "loss": 0.0038, + "step": 15391 + }, + { + "epoch": 1.668690372940156, + "grad_norm": 1.3156323432922363, + "learning_rate": 8.907587407514871e-06, + "loss": 0.01, + "step": 15392 + }, + { + "epoch": 1.6687987857762359, + "grad_norm": 2.602548599243164, + "learning_rate": 8.906862033947483e-06, + "loss": 0.009, + "step": 15393 + }, + { + "epoch": 1.6689071986123158, + "grad_norm": 0.41082051396369934, + "learning_rate": 8.906136660380097e-06, + "loss": 0.0013, + "step": 15394 + }, + { + "epoch": 1.6690156114483954, + "grad_norm": 2.7011168003082275, + "learning_rate": 8.90541128681271e-06, + "loss": 0.0133, + "step": 15395 + }, + { + "epoch": 1.6691240242844754, + "grad_norm": 2.314711570739746, + "learning_rate": 8.904685913245322e-06, + "loss": 0.0179, + "step": 15396 + }, + { + "epoch": 1.669232437120555, + "grad_norm": 2.7672278881073, + "learning_rate": 8.903960539677935e-06, + "loss": 0.0404, + "step": 15397 + }, + { + "epoch": 1.669340849956635, + "grad_norm": 0.7615553736686707, + "learning_rate": 8.903235166110548e-06, + "loss": 0.0027, + "step": 15398 + }, + { + "epoch": 1.6694492627927147, + "grad_norm": 4.178858280181885, + "learning_rate": 8.90250979254316e-06, + "loss": 0.0095, + "step": 15399 + }, + { + "epoch": 1.6695576756287944, + "grad_norm": 3.4014689922332764, + "learning_rate": 8.901784418975774e-06, + "loss": 0.0571, + "step": 15400 + }, + { + "epoch": 1.6696660884648742, + "grad_norm": 2.4537556171417236, + "learning_rate": 8.901059045408386e-06, + "loss": 0.016, + "step": 15401 + }, + { + "epoch": 1.669774501300954, + "grad_norm": 0.6545433402061462, + "learning_rate": 8.900333671841e-06, + "loss": 0.0063, + "step": 15402 + }, + { + "epoch": 1.669882914137034, + "grad_norm": 0.4634394645690918, + "learning_rate": 8.899608298273612e-06, + "loss": 0.0038, + "step": 15403 + }, + { + "epoch": 1.6699913269731135, + "grad_norm": 0.42008861899375916, + "learning_rate": 8.898882924706226e-06, + "loss": 0.0026, + "step": 15404 + }, + { + "epoch": 1.6700997398091935, + "grad_norm": 13.027745246887207, + "learning_rate": 8.898157551138838e-06, + "loss": 0.0714, + "step": 15405 + }, + { + "epoch": 1.6702081526452732, + "grad_norm": 1.8029990196228027, + "learning_rate": 8.89743217757145e-06, + "loss": 0.0398, + "step": 15406 + }, + { + "epoch": 1.670316565481353, + "grad_norm": 2.306535243988037, + "learning_rate": 8.896706804004062e-06, + "loss": 0.0091, + "step": 15407 + }, + { + "epoch": 1.6704249783174328, + "grad_norm": 5.742320537567139, + "learning_rate": 8.895981430436676e-06, + "loss": 0.0604, + "step": 15408 + }, + { + "epoch": 1.6705333911535125, + "grad_norm": 8.128108024597168, + "learning_rate": 8.895256056869288e-06, + "loss": 0.0845, + "step": 15409 + }, + { + "epoch": 1.6706418039895925, + "grad_norm": 1.157228946685791, + "learning_rate": 8.8945306833019e-06, + "loss": 0.0234, + "step": 15410 + }, + { + "epoch": 1.670750216825672, + "grad_norm": 0.9129895567893982, + "learning_rate": 8.893805309734514e-06, + "loss": 0.0019, + "step": 15411 + }, + { + "epoch": 1.670858629661752, + "grad_norm": 3.260225534439087, + "learning_rate": 8.893079936167128e-06, + "loss": 0.0197, + "step": 15412 + }, + { + "epoch": 1.6709670424978318, + "grad_norm": 0.20418913662433624, + "learning_rate": 8.89235456259974e-06, + "loss": 0.0012, + "step": 15413 + }, + { + "epoch": 1.6710754553339116, + "grad_norm": 0.9737815260887146, + "learning_rate": 8.891629189032352e-06, + "loss": 0.0219, + "step": 15414 + }, + { + "epoch": 1.6711838681699913, + "grad_norm": 2.5984208583831787, + "learning_rate": 8.890903815464966e-06, + "loss": 0.0162, + "step": 15415 + }, + { + "epoch": 1.671292281006071, + "grad_norm": 2.0942883491516113, + "learning_rate": 8.890178441897578e-06, + "loss": 0.0256, + "step": 15416 + }, + { + "epoch": 1.671400693842151, + "grad_norm": 5.940995216369629, + "learning_rate": 8.88945306833019e-06, + "loss": 0.0173, + "step": 15417 + }, + { + "epoch": 1.6715091066782306, + "grad_norm": 1.3836450576782227, + "learning_rate": 8.888727694762804e-06, + "loss": 0.0067, + "step": 15418 + }, + { + "epoch": 1.6716175195143106, + "grad_norm": 1.5892279148101807, + "learning_rate": 8.888002321195416e-06, + "loss": 0.0092, + "step": 15419 + }, + { + "epoch": 1.6717259323503901, + "grad_norm": 1.5125235319137573, + "learning_rate": 8.88727694762803e-06, + "loss": 0.0186, + "step": 15420 + }, + { + "epoch": 1.6718343451864701, + "grad_norm": 2.367610216140747, + "learning_rate": 8.886551574060642e-06, + "loss": 0.0219, + "step": 15421 + }, + { + "epoch": 1.6719427580225499, + "grad_norm": 3.893404483795166, + "learning_rate": 8.885826200493254e-06, + "loss": 0.0254, + "step": 15422 + }, + { + "epoch": 1.6720511708586296, + "grad_norm": 3.9167661666870117, + "learning_rate": 8.885100826925868e-06, + "loss": 0.045, + "step": 15423 + }, + { + "epoch": 1.6721595836947094, + "grad_norm": 0.986513614654541, + "learning_rate": 8.88437545335848e-06, + "loss": 0.0099, + "step": 15424 + }, + { + "epoch": 1.6722679965307892, + "grad_norm": 1.4118820428848267, + "learning_rate": 8.883650079791092e-06, + "loss": 0.0066, + "step": 15425 + }, + { + "epoch": 1.6723764093668692, + "grad_norm": 4.784975051879883, + "learning_rate": 8.882924706223706e-06, + "loss": 0.0598, + "step": 15426 + }, + { + "epoch": 1.6724848222029487, + "grad_norm": 0.49465131759643555, + "learning_rate": 8.882199332656318e-06, + "loss": 0.002, + "step": 15427 + }, + { + "epoch": 1.6725932350390287, + "grad_norm": 1.395472526550293, + "learning_rate": 8.881473959088932e-06, + "loss": 0.0142, + "step": 15428 + }, + { + "epoch": 1.6727016478751084, + "grad_norm": 3.791137456893921, + "learning_rate": 8.880748585521544e-06, + "loss": 0.0663, + "step": 15429 + }, + { + "epoch": 1.6728100607111882, + "grad_norm": 2.093730926513672, + "learning_rate": 8.880023211954158e-06, + "loss": 0.0177, + "step": 15430 + }, + { + "epoch": 1.672918473547268, + "grad_norm": 1.9725375175476074, + "learning_rate": 8.87929783838677e-06, + "loss": 0.0201, + "step": 15431 + }, + { + "epoch": 1.6730268863833477, + "grad_norm": 1.46651029586792, + "learning_rate": 8.878572464819382e-06, + "loss": 0.0174, + "step": 15432 + }, + { + "epoch": 1.6731352992194277, + "grad_norm": 1.3138881921768188, + "learning_rate": 8.877847091251995e-06, + "loss": 0.0108, + "step": 15433 + }, + { + "epoch": 1.6732437120555073, + "grad_norm": 4.151981353759766, + "learning_rate": 8.877121717684608e-06, + "loss": 0.0232, + "step": 15434 + }, + { + "epoch": 1.6733521248915872, + "grad_norm": 4.007788181304932, + "learning_rate": 8.87639634411722e-06, + "loss": 0.0318, + "step": 15435 + }, + { + "epoch": 1.673460537727667, + "grad_norm": 0.6171141266822815, + "learning_rate": 8.875670970549833e-06, + "loss": 0.0045, + "step": 15436 + }, + { + "epoch": 1.6735689505637468, + "grad_norm": 8.96163558959961, + "learning_rate": 8.874945596982447e-06, + "loss": 0.083, + "step": 15437 + }, + { + "epoch": 1.6736773633998265, + "grad_norm": 1.3596320152282715, + "learning_rate": 8.87422022341506e-06, + "loss": 0.006, + "step": 15438 + }, + { + "epoch": 1.6737857762359063, + "grad_norm": 0.14574827253818512, + "learning_rate": 8.873494849847673e-06, + "loss": 0.0011, + "step": 15439 + }, + { + "epoch": 1.6738941890719863, + "grad_norm": 2.6158359050750732, + "learning_rate": 8.872769476280285e-06, + "loss": 0.0208, + "step": 15440 + }, + { + "epoch": 1.6740026019080658, + "grad_norm": 3.3447601795196533, + "learning_rate": 8.872044102712899e-06, + "loss": 0.0375, + "step": 15441 + }, + { + "epoch": 1.6741110147441458, + "grad_norm": 0.305620402097702, + "learning_rate": 8.87131872914551e-06, + "loss": 0.0034, + "step": 15442 + }, + { + "epoch": 1.6742194275802254, + "grad_norm": 2.1045026779174805, + "learning_rate": 8.870593355578123e-06, + "loss": 0.0248, + "step": 15443 + }, + { + "epoch": 1.6743278404163053, + "grad_norm": 0.6994004845619202, + "learning_rate": 8.869867982010737e-06, + "loss": 0.003, + "step": 15444 + }, + { + "epoch": 1.674436253252385, + "grad_norm": 0.13659918308258057, + "learning_rate": 8.869142608443349e-06, + "loss": 0.0009, + "step": 15445 + }, + { + "epoch": 1.6745446660884649, + "grad_norm": 13.297310829162598, + "learning_rate": 8.868417234875963e-06, + "loss": 0.0271, + "step": 15446 + }, + { + "epoch": 1.6746530789245446, + "grad_norm": 3.3647663593292236, + "learning_rate": 8.867691861308575e-06, + "loss": 0.0189, + "step": 15447 + }, + { + "epoch": 1.6747614917606244, + "grad_norm": 1.2084249258041382, + "learning_rate": 8.866966487741187e-06, + "loss": 0.0033, + "step": 15448 + }, + { + "epoch": 1.6748699045967044, + "grad_norm": 3.935784101486206, + "learning_rate": 8.8662411141738e-06, + "loss": 0.0813, + "step": 15449 + }, + { + "epoch": 1.674978317432784, + "grad_norm": 4.289231777191162, + "learning_rate": 8.865515740606413e-06, + "loss": 0.0167, + "step": 15450 + }, + { + "epoch": 1.675086730268864, + "grad_norm": 3.3272976875305176, + "learning_rate": 8.864790367039025e-06, + "loss": 0.0163, + "step": 15451 + }, + { + "epoch": 1.6751951431049437, + "grad_norm": 1.0643595457077026, + "learning_rate": 8.864064993471639e-06, + "loss": 0.0202, + "step": 15452 + }, + { + "epoch": 1.6753035559410234, + "grad_norm": 0.6639652252197266, + "learning_rate": 8.863339619904251e-06, + "loss": 0.0191, + "step": 15453 + }, + { + "epoch": 1.6754119687771032, + "grad_norm": 2.1933305263519287, + "learning_rate": 8.862614246336865e-06, + "loss": 0.0372, + "step": 15454 + }, + { + "epoch": 1.675520381613183, + "grad_norm": 4.829825401306152, + "learning_rate": 8.861888872769477e-06, + "loss": 0.0676, + "step": 15455 + }, + { + "epoch": 1.675628794449263, + "grad_norm": 4.576359748840332, + "learning_rate": 8.861163499202091e-06, + "loss": 0.0108, + "step": 15456 + }, + { + "epoch": 1.6757372072853425, + "grad_norm": 0.7397621870040894, + "learning_rate": 8.860438125634703e-06, + "loss": 0.0036, + "step": 15457 + }, + { + "epoch": 1.6758456201214225, + "grad_norm": 0.5885685682296753, + "learning_rate": 8.859712752067315e-06, + "loss": 0.0038, + "step": 15458 + }, + { + "epoch": 1.675954032957502, + "grad_norm": 0.9196416735649109, + "learning_rate": 8.858987378499927e-06, + "loss": 0.0031, + "step": 15459 + }, + { + "epoch": 1.676062445793582, + "grad_norm": 1.7963752746582031, + "learning_rate": 8.858262004932541e-06, + "loss": 0.0054, + "step": 15460 + }, + { + "epoch": 1.6761708586296618, + "grad_norm": 0.2663765549659729, + "learning_rate": 8.857536631365153e-06, + "loss": 0.0013, + "step": 15461 + }, + { + "epoch": 1.6762792714657415, + "grad_norm": 4.086231231689453, + "learning_rate": 8.856811257797765e-06, + "loss": 0.0486, + "step": 15462 + }, + { + "epoch": 1.6763876843018215, + "grad_norm": 1.0267244577407837, + "learning_rate": 8.85608588423038e-06, + "loss": 0.0043, + "step": 15463 + }, + { + "epoch": 1.676496097137901, + "grad_norm": 3.480170965194702, + "learning_rate": 8.855360510662993e-06, + "loss": 0.0547, + "step": 15464 + }, + { + "epoch": 1.676604509973981, + "grad_norm": 0.7449094653129578, + "learning_rate": 8.854635137095605e-06, + "loss": 0.0032, + "step": 15465 + }, + { + "epoch": 1.6767129228100606, + "grad_norm": 1.5629545450210571, + "learning_rate": 8.853909763528217e-06, + "loss": 0.0113, + "step": 15466 + }, + { + "epoch": 1.6768213356461406, + "grad_norm": 0.832977294921875, + "learning_rate": 8.853184389960831e-06, + "loss": 0.0032, + "step": 15467 + }, + { + "epoch": 1.6769297484822203, + "grad_norm": 1.610453724861145, + "learning_rate": 8.852459016393443e-06, + "loss": 0.0096, + "step": 15468 + }, + { + "epoch": 1.6770381613183, + "grad_norm": 0.9617730379104614, + "learning_rate": 8.851733642826055e-06, + "loss": 0.0163, + "step": 15469 + }, + { + "epoch": 1.6771465741543798, + "grad_norm": 8.145901679992676, + "learning_rate": 8.85100826925867e-06, + "loss": 0.0114, + "step": 15470 + }, + { + "epoch": 1.6772549869904596, + "grad_norm": 1.5541967153549194, + "learning_rate": 8.850282895691281e-06, + "loss": 0.0288, + "step": 15471 + }, + { + "epoch": 1.6773633998265396, + "grad_norm": 0.28274092078208923, + "learning_rate": 8.849557522123895e-06, + "loss": 0.003, + "step": 15472 + }, + { + "epoch": 1.6774718126626191, + "grad_norm": 3.000492572784424, + "learning_rate": 8.848832148556507e-06, + "loss": 0.0468, + "step": 15473 + }, + { + "epoch": 1.6775802254986991, + "grad_norm": 1.3231571912765503, + "learning_rate": 8.84810677498912e-06, + "loss": 0.0151, + "step": 15474 + }, + { + "epoch": 1.6776886383347789, + "grad_norm": 1.05660879611969, + "learning_rate": 8.847381401421733e-06, + "loss": 0.0073, + "step": 15475 + }, + { + "epoch": 1.6777970511708586, + "grad_norm": 6.505809307098389, + "learning_rate": 8.846656027854346e-06, + "loss": 0.0621, + "step": 15476 + }, + { + "epoch": 1.6779054640069384, + "grad_norm": 5.436246871948242, + "learning_rate": 8.845930654286958e-06, + "loss": 0.1013, + "step": 15477 + }, + { + "epoch": 1.6780138768430182, + "grad_norm": 3.5894687175750732, + "learning_rate": 8.845205280719572e-06, + "loss": 0.0185, + "step": 15478 + }, + { + "epoch": 1.6781222896790982, + "grad_norm": 12.282081604003906, + "learning_rate": 8.844479907152184e-06, + "loss": 0.0688, + "step": 15479 + }, + { + "epoch": 1.6782307025151777, + "grad_norm": 5.6775994300842285, + "learning_rate": 8.843754533584798e-06, + "loss": 0.015, + "step": 15480 + }, + { + "epoch": 1.6783391153512577, + "grad_norm": 3.295562267303467, + "learning_rate": 8.84302916001741e-06, + "loss": 0.0377, + "step": 15481 + }, + { + "epoch": 1.6784475281873372, + "grad_norm": 6.743276119232178, + "learning_rate": 8.842303786450024e-06, + "loss": 0.0106, + "step": 15482 + }, + { + "epoch": 1.6785559410234172, + "grad_norm": 20.4223690032959, + "learning_rate": 8.841578412882636e-06, + "loss": 0.026, + "step": 15483 + }, + { + "epoch": 1.678664353859497, + "grad_norm": 2.530205488204956, + "learning_rate": 8.840853039315248e-06, + "loss": 0.0441, + "step": 15484 + }, + { + "epoch": 1.6787727666955767, + "grad_norm": 1.4298434257507324, + "learning_rate": 8.84012766574786e-06, + "loss": 0.0017, + "step": 15485 + }, + { + "epoch": 1.6788811795316565, + "grad_norm": 9.082891464233398, + "learning_rate": 8.839402292180474e-06, + "loss": 0.0356, + "step": 15486 + }, + { + "epoch": 1.6789895923677363, + "grad_norm": 1.2749276161193848, + "learning_rate": 8.838676918613086e-06, + "loss": 0.0071, + "step": 15487 + }, + { + "epoch": 1.6790980052038162, + "grad_norm": 2.349663257598877, + "learning_rate": 8.837951545045698e-06, + "loss": 0.0296, + "step": 15488 + }, + { + "epoch": 1.6792064180398958, + "grad_norm": 1.377553105354309, + "learning_rate": 8.837226171478312e-06, + "loss": 0.0159, + "step": 15489 + }, + { + "epoch": 1.6793148308759758, + "grad_norm": 2.0177223682403564, + "learning_rate": 8.836500797910926e-06, + "loss": 0.0143, + "step": 15490 + }, + { + "epoch": 1.6794232437120555, + "grad_norm": 1.079602599143982, + "learning_rate": 8.835775424343538e-06, + "loss": 0.0043, + "step": 15491 + }, + { + "epoch": 1.6795316565481353, + "grad_norm": 1.2633068561553955, + "learning_rate": 8.83505005077615e-06, + "loss": 0.0356, + "step": 15492 + }, + { + "epoch": 1.679640069384215, + "grad_norm": 1.8405168056488037, + "learning_rate": 8.834324677208764e-06, + "loss": 0.0098, + "step": 15493 + }, + { + "epoch": 1.6797484822202948, + "grad_norm": 2.87974214553833, + "learning_rate": 8.833599303641376e-06, + "loss": 0.0112, + "step": 15494 + }, + { + "epoch": 1.6798568950563748, + "grad_norm": 4.625563621520996, + "learning_rate": 8.832873930073988e-06, + "loss": 0.0167, + "step": 15495 + }, + { + "epoch": 1.6799653078924544, + "grad_norm": 2.912895679473877, + "learning_rate": 8.832148556506602e-06, + "loss": 0.014, + "step": 15496 + }, + { + "epoch": 1.6800737207285343, + "grad_norm": 4.861792087554932, + "learning_rate": 8.831423182939214e-06, + "loss": 0.0373, + "step": 15497 + }, + { + "epoch": 1.680182133564614, + "grad_norm": 3.531287431716919, + "learning_rate": 8.830697809371828e-06, + "loss": 0.0185, + "step": 15498 + }, + { + "epoch": 1.6802905464006939, + "grad_norm": 6.649513244628906, + "learning_rate": 8.82997243580444e-06, + "loss": 0.0458, + "step": 15499 + }, + { + "epoch": 1.6803989592367736, + "grad_norm": 0.2708386480808258, + "learning_rate": 8.829247062237052e-06, + "loss": 0.0022, + "step": 15500 + }, + { + "epoch": 1.6805073720728534, + "grad_norm": 0.7374222278594971, + "learning_rate": 8.828521688669666e-06, + "loss": 0.0072, + "step": 15501 + }, + { + "epoch": 1.6806157849089334, + "grad_norm": 3.059950351715088, + "learning_rate": 8.827796315102278e-06, + "loss": 0.0129, + "step": 15502 + }, + { + "epoch": 1.680724197745013, + "grad_norm": 4.1782355308532715, + "learning_rate": 8.82707094153489e-06, + "loss": 0.008, + "step": 15503 + }, + { + "epoch": 1.680832610581093, + "grad_norm": 5.65698766708374, + "learning_rate": 8.826345567967504e-06, + "loss": 0.0439, + "step": 15504 + }, + { + "epoch": 1.6809410234171724, + "grad_norm": 0.978539764881134, + "learning_rate": 8.825620194400116e-06, + "loss": 0.0315, + "step": 15505 + }, + { + "epoch": 1.6810494362532524, + "grad_norm": 2.3826448917388916, + "learning_rate": 8.82489482083273e-06, + "loss": 0.0135, + "step": 15506 + }, + { + "epoch": 1.6811578490893322, + "grad_norm": 2.687741279602051, + "learning_rate": 8.824169447265342e-06, + "loss": 0.0322, + "step": 15507 + }, + { + "epoch": 1.681266261925412, + "grad_norm": 2.16768217086792, + "learning_rate": 8.823444073697956e-06, + "loss": 0.0189, + "step": 15508 + }, + { + "epoch": 1.6813746747614917, + "grad_norm": 0.6715871691703796, + "learning_rate": 8.822718700130568e-06, + "loss": 0.0058, + "step": 15509 + }, + { + "epoch": 1.6814830875975715, + "grad_norm": 0.6717238426208496, + "learning_rate": 8.82199332656318e-06, + "loss": 0.0049, + "step": 15510 + }, + { + "epoch": 1.6815915004336515, + "grad_norm": 0.7686122059822083, + "learning_rate": 8.821267952995793e-06, + "loss": 0.0029, + "step": 15511 + }, + { + "epoch": 1.681699913269731, + "grad_norm": 8.956548690795898, + "learning_rate": 8.820542579428406e-06, + "loss": 0.0538, + "step": 15512 + }, + { + "epoch": 1.681808326105811, + "grad_norm": 1.6817317008972168, + "learning_rate": 8.819817205861019e-06, + "loss": 0.0269, + "step": 15513 + }, + { + "epoch": 1.6819167389418908, + "grad_norm": 3.7324373722076416, + "learning_rate": 8.81909183229363e-06, + "loss": 0.0466, + "step": 15514 + }, + { + "epoch": 1.6820251517779705, + "grad_norm": 0.7445021867752075, + "learning_rate": 8.818366458726245e-06, + "loss": 0.004, + "step": 15515 + }, + { + "epoch": 1.6821335646140503, + "grad_norm": 1.104790210723877, + "learning_rate": 8.817641085158858e-06, + "loss": 0.0128, + "step": 15516 + }, + { + "epoch": 1.68224197745013, + "grad_norm": 3.0304369926452637, + "learning_rate": 8.81691571159147e-06, + "loss": 0.0213, + "step": 15517 + }, + { + "epoch": 1.68235039028621, + "grad_norm": 3.9720654487609863, + "learning_rate": 8.816190338024083e-06, + "loss": 0.0544, + "step": 15518 + }, + { + "epoch": 1.6824588031222896, + "grad_norm": 0.4141884744167328, + "learning_rate": 8.815464964456697e-06, + "loss": 0.0038, + "step": 15519 + }, + { + "epoch": 1.6825672159583696, + "grad_norm": 5.487281799316406, + "learning_rate": 8.814739590889309e-06, + "loss": 0.0105, + "step": 15520 + }, + { + "epoch": 1.682675628794449, + "grad_norm": 5.502126216888428, + "learning_rate": 8.81401421732192e-06, + "loss": 0.0317, + "step": 15521 + }, + { + "epoch": 1.682784041630529, + "grad_norm": 0.6654363870620728, + "learning_rate": 8.813288843754535e-06, + "loss": 0.0106, + "step": 15522 + }, + { + "epoch": 1.6828924544666088, + "grad_norm": 6.817348957061768, + "learning_rate": 8.812563470187147e-06, + "loss": 0.0368, + "step": 15523 + }, + { + "epoch": 1.6830008673026886, + "grad_norm": 2.5240776538848877, + "learning_rate": 8.81183809661976e-06, + "loss": 0.0258, + "step": 15524 + }, + { + "epoch": 1.6831092801387686, + "grad_norm": 0.7309066653251648, + "learning_rate": 8.811112723052373e-06, + "loss": 0.0056, + "step": 15525 + }, + { + "epoch": 1.6832176929748481, + "grad_norm": 4.6631574630737305, + "learning_rate": 8.810387349484985e-06, + "loss": 0.0297, + "step": 15526 + }, + { + "epoch": 1.6833261058109281, + "grad_norm": 0.12353529036045074, + "learning_rate": 8.809661975917599e-06, + "loss": 0.0012, + "step": 15527 + }, + { + "epoch": 1.6834345186470077, + "grad_norm": 5.111289024353027, + "learning_rate": 8.808936602350211e-06, + "loss": 0.0397, + "step": 15528 + }, + { + "epoch": 1.6835429314830876, + "grad_norm": 11.331680297851562, + "learning_rate": 8.808211228782823e-06, + "loss": 0.032, + "step": 15529 + }, + { + "epoch": 1.6836513443191674, + "grad_norm": 0.3088926374912262, + "learning_rate": 8.807485855215437e-06, + "loss": 0.001, + "step": 15530 + }, + { + "epoch": 1.6837597571552472, + "grad_norm": 5.480447292327881, + "learning_rate": 8.806760481648049e-06, + "loss": 0.0317, + "step": 15531 + }, + { + "epoch": 1.683868169991327, + "grad_norm": 4.327746391296387, + "learning_rate": 8.806035108080663e-06, + "loss": 0.0081, + "step": 15532 + }, + { + "epoch": 1.6839765828274067, + "grad_norm": 2.6294844150543213, + "learning_rate": 8.805309734513275e-06, + "loss": 0.0106, + "step": 15533 + }, + { + "epoch": 1.6840849956634867, + "grad_norm": 0.3104427456855774, + "learning_rate": 8.804584360945889e-06, + "loss": 0.0033, + "step": 15534 + }, + { + "epoch": 1.6841934084995662, + "grad_norm": 3.1296677589416504, + "learning_rate": 8.803858987378501e-06, + "loss": 0.0145, + "step": 15535 + }, + { + "epoch": 1.6843018213356462, + "grad_norm": 3.387570858001709, + "learning_rate": 8.803133613811113e-06, + "loss": 0.0129, + "step": 15536 + }, + { + "epoch": 1.684410234171726, + "grad_norm": 4.260382175445557, + "learning_rate": 8.802408240243725e-06, + "loss": 0.0307, + "step": 15537 + }, + { + "epoch": 1.6845186470078057, + "grad_norm": 2.073101758956909, + "learning_rate": 8.801682866676339e-06, + "loss": 0.0118, + "step": 15538 + }, + { + "epoch": 1.6846270598438855, + "grad_norm": 3.797626495361328, + "learning_rate": 8.800957493108951e-06, + "loss": 0.0201, + "step": 15539 + }, + { + "epoch": 1.6847354726799653, + "grad_norm": 3.24562931060791, + "learning_rate": 8.800232119541563e-06, + "loss": 0.0219, + "step": 15540 + }, + { + "epoch": 1.6848438855160452, + "grad_norm": 2.70296311378479, + "learning_rate": 8.799506745974177e-06, + "loss": 0.0514, + "step": 15541 + }, + { + "epoch": 1.6849522983521248, + "grad_norm": 3.1093409061431885, + "learning_rate": 8.798781372406791e-06, + "loss": 0.0312, + "step": 15542 + }, + { + "epoch": 1.6850607111882048, + "grad_norm": 8.856345176696777, + "learning_rate": 8.798055998839403e-06, + "loss": 0.0462, + "step": 15543 + }, + { + "epoch": 1.6851691240242843, + "grad_norm": 0.22314205765724182, + "learning_rate": 8.797330625272015e-06, + "loss": 0.0022, + "step": 15544 + }, + { + "epoch": 1.6852775368603643, + "grad_norm": 1.200954556465149, + "learning_rate": 8.79660525170463e-06, + "loss": 0.0044, + "step": 15545 + }, + { + "epoch": 1.685385949696444, + "grad_norm": 3.441282272338867, + "learning_rate": 8.795879878137241e-06, + "loss": 0.0546, + "step": 15546 + }, + { + "epoch": 1.6854943625325238, + "grad_norm": 0.4328484535217285, + "learning_rate": 8.795154504569853e-06, + "loss": 0.0049, + "step": 15547 + }, + { + "epoch": 1.6856027753686036, + "grad_norm": 6.489717483520508, + "learning_rate": 8.794429131002467e-06, + "loss": 0.0132, + "step": 15548 + }, + { + "epoch": 1.6857111882046834, + "grad_norm": 8.4205904006958, + "learning_rate": 8.79370375743508e-06, + "loss": 0.0305, + "step": 15549 + }, + { + "epoch": 1.6858196010407633, + "grad_norm": 1.4990519285202026, + "learning_rate": 8.792978383867693e-06, + "loss": 0.0178, + "step": 15550 + }, + { + "epoch": 1.6859280138768429, + "grad_norm": 1.8212509155273438, + "learning_rate": 8.792253010300305e-06, + "loss": 0.0216, + "step": 15551 + }, + { + "epoch": 1.6860364267129229, + "grad_norm": 0.44185683131217957, + "learning_rate": 8.791527636732918e-06, + "loss": 0.0027, + "step": 15552 + }, + { + "epoch": 1.6861448395490026, + "grad_norm": 1.5622800588607788, + "learning_rate": 8.790802263165531e-06, + "loss": 0.0168, + "step": 15553 + }, + { + "epoch": 1.6862532523850824, + "grad_norm": 0.657909095287323, + "learning_rate": 8.790076889598144e-06, + "loss": 0.0058, + "step": 15554 + }, + { + "epoch": 1.6863616652211622, + "grad_norm": 3.394519329071045, + "learning_rate": 8.789351516030756e-06, + "loss": 0.0478, + "step": 15555 + }, + { + "epoch": 1.686470078057242, + "grad_norm": 4.072753429412842, + "learning_rate": 8.78862614246337e-06, + "loss": 0.0278, + "step": 15556 + }, + { + "epoch": 1.686578490893322, + "grad_norm": 2.6008591651916504, + "learning_rate": 8.787900768895982e-06, + "loss": 0.0282, + "step": 15557 + }, + { + "epoch": 1.6866869037294014, + "grad_norm": 3.1585919857025146, + "learning_rate": 8.787175395328596e-06, + "loss": 0.0379, + "step": 15558 + }, + { + "epoch": 1.6867953165654814, + "grad_norm": 1.5124486684799194, + "learning_rate": 8.786450021761208e-06, + "loss": 0.0124, + "step": 15559 + }, + { + "epoch": 1.6869037294015612, + "grad_norm": 2.138033866882324, + "learning_rate": 8.785724648193822e-06, + "loss": 0.0304, + "step": 15560 + }, + { + "epoch": 1.687012142237641, + "grad_norm": 0.33075717091560364, + "learning_rate": 8.784999274626434e-06, + "loss": 0.003, + "step": 15561 + }, + { + "epoch": 1.6871205550737207, + "grad_norm": 3.2156314849853516, + "learning_rate": 8.784273901059046e-06, + "loss": 0.0183, + "step": 15562 + }, + { + "epoch": 1.6872289679098005, + "grad_norm": 2.247358560562134, + "learning_rate": 8.783548527491658e-06, + "loss": 0.0075, + "step": 15563 + }, + { + "epoch": 1.6873373807458805, + "grad_norm": 0.15764684975147247, + "learning_rate": 8.782823153924272e-06, + "loss": 0.0011, + "step": 15564 + }, + { + "epoch": 1.68744579358196, + "grad_norm": 12.42734432220459, + "learning_rate": 8.782097780356884e-06, + "loss": 0.0699, + "step": 15565 + }, + { + "epoch": 1.68755420641804, + "grad_norm": 0.40651997923851013, + "learning_rate": 8.781372406789496e-06, + "loss": 0.0016, + "step": 15566 + }, + { + "epoch": 1.6876626192541195, + "grad_norm": 1.1888434886932373, + "learning_rate": 8.78064703322211e-06, + "loss": 0.0313, + "step": 15567 + }, + { + "epoch": 1.6877710320901995, + "grad_norm": 4.292230606079102, + "learning_rate": 8.779921659654724e-06, + "loss": 0.041, + "step": 15568 + }, + { + "epoch": 1.6878794449262793, + "grad_norm": 1.9205992221832275, + "learning_rate": 8.779196286087336e-06, + "loss": 0.0109, + "step": 15569 + }, + { + "epoch": 1.687987857762359, + "grad_norm": 3.1593055725097656, + "learning_rate": 8.778470912519948e-06, + "loss": 0.0208, + "step": 15570 + }, + { + "epoch": 1.6880962705984388, + "grad_norm": 7.305316925048828, + "learning_rate": 8.777745538952562e-06, + "loss": 0.022, + "step": 15571 + }, + { + "epoch": 1.6882046834345186, + "grad_norm": 4.239538669586182, + "learning_rate": 8.777020165385174e-06, + "loss": 0.0586, + "step": 15572 + }, + { + "epoch": 1.6883130962705986, + "grad_norm": 1.7543758153915405, + "learning_rate": 8.776294791817786e-06, + "loss": 0.0083, + "step": 15573 + }, + { + "epoch": 1.688421509106678, + "grad_norm": 2.853052854537964, + "learning_rate": 8.7755694182504e-06, + "loss": 0.0573, + "step": 15574 + }, + { + "epoch": 1.688529921942758, + "grad_norm": 1.7940597534179688, + "learning_rate": 8.774844044683012e-06, + "loss": 0.0212, + "step": 15575 + }, + { + "epoch": 1.6886383347788378, + "grad_norm": 2.031275510787964, + "learning_rate": 8.774118671115626e-06, + "loss": 0.0531, + "step": 15576 + }, + { + "epoch": 1.6887467476149176, + "grad_norm": 5.540437698364258, + "learning_rate": 8.773393297548238e-06, + "loss": 0.0137, + "step": 15577 + }, + { + "epoch": 1.6888551604509974, + "grad_norm": 2.3078773021698, + "learning_rate": 8.77266792398085e-06, + "loss": 0.016, + "step": 15578 + }, + { + "epoch": 1.6889635732870771, + "grad_norm": 3.324838876724243, + "learning_rate": 8.771942550413464e-06, + "loss": 0.0498, + "step": 15579 + }, + { + "epoch": 1.6890719861231571, + "grad_norm": 3.46002459526062, + "learning_rate": 8.771217176846076e-06, + "loss": 0.0357, + "step": 15580 + }, + { + "epoch": 1.6891803989592367, + "grad_norm": 1.3696280717849731, + "learning_rate": 8.770491803278688e-06, + "loss": 0.0143, + "step": 15581 + }, + { + "epoch": 1.6892888117953166, + "grad_norm": 3.2596471309661865, + "learning_rate": 8.769766429711302e-06, + "loss": 0.0312, + "step": 15582 + }, + { + "epoch": 1.6893972246313964, + "grad_norm": 2.6565144062042236, + "learning_rate": 8.769041056143914e-06, + "loss": 0.0126, + "step": 15583 + }, + { + "epoch": 1.6895056374674762, + "grad_norm": 0.8980741500854492, + "learning_rate": 8.768315682576528e-06, + "loss": 0.0153, + "step": 15584 + }, + { + "epoch": 1.689614050303556, + "grad_norm": 0.8690634369850159, + "learning_rate": 8.76759030900914e-06, + "loss": 0.0043, + "step": 15585 + }, + { + "epoch": 1.6897224631396357, + "grad_norm": 2.636803388595581, + "learning_rate": 8.766864935441754e-06, + "loss": 0.0117, + "step": 15586 + }, + { + "epoch": 1.6898308759757157, + "grad_norm": 0.7793638706207275, + "learning_rate": 8.766139561874366e-06, + "loss": 0.0057, + "step": 15587 + }, + { + "epoch": 1.6899392888117952, + "grad_norm": 4.9159321784973145, + "learning_rate": 8.765414188306978e-06, + "loss": 0.0246, + "step": 15588 + }, + { + "epoch": 1.6900477016478752, + "grad_norm": 4.760615348815918, + "learning_rate": 8.76468881473959e-06, + "loss": 0.0461, + "step": 15589 + }, + { + "epoch": 1.6901561144839548, + "grad_norm": 3.3072502613067627, + "learning_rate": 8.763963441172204e-06, + "loss": 0.0225, + "step": 15590 + }, + { + "epoch": 1.6902645273200347, + "grad_norm": 0.9472976922988892, + "learning_rate": 8.763238067604817e-06, + "loss": 0.0084, + "step": 15591 + }, + { + "epoch": 1.6903729401561145, + "grad_norm": 4.241979122161865, + "learning_rate": 8.76251269403743e-06, + "loss": 0.0268, + "step": 15592 + }, + { + "epoch": 1.6904813529921943, + "grad_norm": 0.7728321552276611, + "learning_rate": 8.761787320470043e-06, + "loss": 0.0056, + "step": 15593 + }, + { + "epoch": 1.690589765828274, + "grad_norm": 3.349033832550049, + "learning_rate": 8.761061946902656e-06, + "loss": 0.0174, + "step": 15594 + }, + { + "epoch": 1.6906981786643538, + "grad_norm": 1.0087850093841553, + "learning_rate": 8.760336573335269e-06, + "loss": 0.0196, + "step": 15595 + }, + { + "epoch": 1.6908065915004338, + "grad_norm": 2.1174519062042236, + "learning_rate": 8.75961119976788e-06, + "loss": 0.0141, + "step": 15596 + }, + { + "epoch": 1.6909150043365133, + "grad_norm": 2.5253961086273193, + "learning_rate": 8.758885826200495e-06, + "loss": 0.0059, + "step": 15597 + }, + { + "epoch": 1.6910234171725933, + "grad_norm": 4.511455059051514, + "learning_rate": 8.758160452633107e-06, + "loss": 0.0412, + "step": 15598 + }, + { + "epoch": 1.691131830008673, + "grad_norm": 1.6444261074066162, + "learning_rate": 8.757435079065719e-06, + "loss": 0.0254, + "step": 15599 + }, + { + "epoch": 1.6912402428447528, + "grad_norm": 1.5182005167007446, + "learning_rate": 8.756709705498333e-06, + "loss": 0.0204, + "step": 15600 + }, + { + "epoch": 1.6913486556808326, + "grad_norm": 0.8683890700340271, + "learning_rate": 8.755984331930945e-06, + "loss": 0.0072, + "step": 15601 + }, + { + "epoch": 1.6914570685169124, + "grad_norm": 1.6785327196121216, + "learning_rate": 8.755258958363559e-06, + "loss": 0.002, + "step": 15602 + }, + { + "epoch": 1.6915654813529923, + "grad_norm": 1.2786344289779663, + "learning_rate": 8.75453358479617e-06, + "loss": 0.0208, + "step": 15603 + }, + { + "epoch": 1.6916738941890719, + "grad_norm": 0.5390468835830688, + "learning_rate": 8.753808211228783e-06, + "loss": 0.0043, + "step": 15604 + }, + { + "epoch": 1.6917823070251519, + "grad_norm": 1.0042214393615723, + "learning_rate": 8.753082837661397e-06, + "loss": 0.0066, + "step": 15605 + }, + { + "epoch": 1.6918907198612314, + "grad_norm": 6.012113094329834, + "learning_rate": 8.752357464094009e-06, + "loss": 0.0556, + "step": 15606 + }, + { + "epoch": 1.6919991326973114, + "grad_norm": 0.6843678951263428, + "learning_rate": 8.751632090526621e-06, + "loss": 0.0051, + "step": 15607 + }, + { + "epoch": 1.6921075455333912, + "grad_norm": 0.2648216784000397, + "learning_rate": 8.750906716959235e-06, + "loss": 0.0018, + "step": 15608 + }, + { + "epoch": 1.692215958369471, + "grad_norm": 3.528803586959839, + "learning_rate": 8.750181343391847e-06, + "loss": 0.0406, + "step": 15609 + }, + { + "epoch": 1.692324371205551, + "grad_norm": 1.3237707614898682, + "learning_rate": 8.749455969824461e-06, + "loss": 0.0118, + "step": 15610 + }, + { + "epoch": 1.6924327840416304, + "grad_norm": 2.902308940887451, + "learning_rate": 8.748730596257073e-06, + "loss": 0.0389, + "step": 15611 + }, + { + "epoch": 1.6925411968777104, + "grad_norm": 2.0732414722442627, + "learning_rate": 8.748005222689687e-06, + "loss": 0.0124, + "step": 15612 + }, + { + "epoch": 1.69264960971379, + "grad_norm": 2.05570387840271, + "learning_rate": 8.747279849122299e-06, + "loss": 0.0294, + "step": 15613 + }, + { + "epoch": 1.69275802254987, + "grad_norm": 1.5104700326919556, + "learning_rate": 8.746554475554911e-06, + "loss": 0.0159, + "step": 15614 + }, + { + "epoch": 1.6928664353859497, + "grad_norm": 0.6079983711242676, + "learning_rate": 8.745829101987523e-06, + "loss": 0.0054, + "step": 15615 + }, + { + "epoch": 1.6929748482220295, + "grad_norm": 0.7042751312255859, + "learning_rate": 8.745103728420137e-06, + "loss": 0.0082, + "step": 15616 + }, + { + "epoch": 1.6930832610581092, + "grad_norm": 6.809743404388428, + "learning_rate": 8.74437835485275e-06, + "loss": 0.0375, + "step": 15617 + }, + { + "epoch": 1.693191673894189, + "grad_norm": 2.165705919265747, + "learning_rate": 8.743652981285363e-06, + "loss": 0.0232, + "step": 15618 + }, + { + "epoch": 1.693300086730269, + "grad_norm": 0.8209806680679321, + "learning_rate": 8.742927607717975e-06, + "loss": 0.0039, + "step": 15619 + }, + { + "epoch": 1.6934084995663485, + "grad_norm": 5.1153788566589355, + "learning_rate": 8.742202234150589e-06, + "loss": 0.0331, + "step": 15620 + }, + { + "epoch": 1.6935169124024285, + "grad_norm": 1.863853096961975, + "learning_rate": 8.741476860583201e-06, + "loss": 0.0107, + "step": 15621 + }, + { + "epoch": 1.6936253252385083, + "grad_norm": 1.066510558128357, + "learning_rate": 8.740751487015813e-06, + "loss": 0.0071, + "step": 15622 + }, + { + "epoch": 1.693733738074588, + "grad_norm": 1.35600745677948, + "learning_rate": 8.740026113448427e-06, + "loss": 0.0222, + "step": 15623 + }, + { + "epoch": 1.6938421509106678, + "grad_norm": 4.313692092895508, + "learning_rate": 8.73930073988104e-06, + "loss": 0.0269, + "step": 15624 + }, + { + "epoch": 1.6939505637467476, + "grad_norm": 1.8757492303848267, + "learning_rate": 8.738575366313652e-06, + "loss": 0.0144, + "step": 15625 + }, + { + "epoch": 1.6940589765828276, + "grad_norm": 4.920413970947266, + "learning_rate": 8.737849992746265e-06, + "loss": 0.032, + "step": 15626 + }, + { + "epoch": 1.694167389418907, + "grad_norm": 0.12749633193016052, + "learning_rate": 8.737124619178877e-06, + "loss": 0.0011, + "step": 15627 + }, + { + "epoch": 1.694275802254987, + "grad_norm": 1.2418537139892578, + "learning_rate": 8.736399245611491e-06, + "loss": 0.0217, + "step": 15628 + }, + { + "epoch": 1.6943842150910666, + "grad_norm": 0.26805588603019714, + "learning_rate": 8.735673872044103e-06, + "loss": 0.0018, + "step": 15629 + }, + { + "epoch": 1.6944926279271466, + "grad_norm": 0.41745638847351074, + "learning_rate": 8.734948498476716e-06, + "loss": 0.0022, + "step": 15630 + }, + { + "epoch": 1.6946010407632264, + "grad_norm": 6.803143501281738, + "learning_rate": 8.73422312490933e-06, + "loss": 0.0188, + "step": 15631 + }, + { + "epoch": 1.6947094535993061, + "grad_norm": 2.280240297317505, + "learning_rate": 8.733497751341942e-06, + "loss": 0.0049, + "step": 15632 + }, + { + "epoch": 1.694817866435386, + "grad_norm": 9.32176399230957, + "learning_rate": 8.732772377774554e-06, + "loss": 0.0204, + "step": 15633 + }, + { + "epoch": 1.6949262792714657, + "grad_norm": 2.1401987075805664, + "learning_rate": 8.732047004207168e-06, + "loss": 0.0192, + "step": 15634 + }, + { + "epoch": 1.6950346921075456, + "grad_norm": 1.8891217708587646, + "learning_rate": 8.73132163063978e-06, + "loss": 0.0117, + "step": 15635 + }, + { + "epoch": 1.6951431049436252, + "grad_norm": 3.7736408710479736, + "learning_rate": 8.730596257072394e-06, + "loss": 0.0426, + "step": 15636 + }, + { + "epoch": 1.6952515177797052, + "grad_norm": 1.148655652999878, + "learning_rate": 8.729870883505006e-06, + "loss": 0.0339, + "step": 15637 + }, + { + "epoch": 1.695359930615785, + "grad_norm": 1.1269605159759521, + "learning_rate": 8.72914550993762e-06, + "loss": 0.0207, + "step": 15638 + }, + { + "epoch": 1.6954683434518647, + "grad_norm": 1.8990767002105713, + "learning_rate": 8.728420136370232e-06, + "loss": 0.0247, + "step": 15639 + }, + { + "epoch": 1.6955767562879445, + "grad_norm": 2.172161102294922, + "learning_rate": 8.727694762802844e-06, + "loss": 0.0073, + "step": 15640 + }, + { + "epoch": 1.6956851691240242, + "grad_norm": 1.9575151205062866, + "learning_rate": 8.726969389235456e-06, + "loss": 0.0153, + "step": 15641 + }, + { + "epoch": 1.6957935819601042, + "grad_norm": 3.730341672897339, + "learning_rate": 8.72624401566807e-06, + "loss": 0.0348, + "step": 15642 + }, + { + "epoch": 1.6959019947961838, + "grad_norm": 7.688621520996094, + "learning_rate": 8.725518642100682e-06, + "loss": 0.0471, + "step": 15643 + }, + { + "epoch": 1.6960104076322637, + "grad_norm": 0.9872711896896362, + "learning_rate": 8.724793268533296e-06, + "loss": 0.0055, + "step": 15644 + }, + { + "epoch": 1.6961188204683435, + "grad_norm": 1.9318244457244873, + "learning_rate": 8.724067894965908e-06, + "loss": 0.0261, + "step": 15645 + }, + { + "epoch": 1.6962272333044233, + "grad_norm": 3.858081579208374, + "learning_rate": 8.723342521398522e-06, + "loss": 0.0562, + "step": 15646 + }, + { + "epoch": 1.696335646140503, + "grad_norm": 21.70769691467285, + "learning_rate": 8.722617147831134e-06, + "loss": 0.0653, + "step": 15647 + }, + { + "epoch": 1.6964440589765828, + "grad_norm": 2.4676153659820557, + "learning_rate": 8.721891774263746e-06, + "loss": 0.0464, + "step": 15648 + }, + { + "epoch": 1.6965524718126628, + "grad_norm": 5.60715913772583, + "learning_rate": 8.72116640069636e-06, + "loss": 0.0333, + "step": 15649 + }, + { + "epoch": 1.6966608846487423, + "grad_norm": 2.6856067180633545, + "learning_rate": 8.720441027128972e-06, + "loss": 0.0408, + "step": 15650 + }, + { + "epoch": 1.6967692974848223, + "grad_norm": 3.1620240211486816, + "learning_rate": 8.719715653561584e-06, + "loss": 0.0681, + "step": 15651 + }, + { + "epoch": 1.6968777103209018, + "grad_norm": 1.1542418003082275, + "learning_rate": 8.718990279994198e-06, + "loss": 0.0143, + "step": 15652 + }, + { + "epoch": 1.6969861231569818, + "grad_norm": 1.7700148820877075, + "learning_rate": 8.718264906426812e-06, + "loss": 0.0302, + "step": 15653 + }, + { + "epoch": 1.6970945359930616, + "grad_norm": 0.1639808565378189, + "learning_rate": 8.717539532859424e-06, + "loss": 0.0023, + "step": 15654 + }, + { + "epoch": 1.6972029488291414, + "grad_norm": 1.130430817604065, + "learning_rate": 8.716814159292036e-06, + "loss": 0.0184, + "step": 15655 + }, + { + "epoch": 1.6973113616652211, + "grad_norm": 0.4961355924606323, + "learning_rate": 8.716088785724648e-06, + "loss": 0.0048, + "step": 15656 + }, + { + "epoch": 1.6974197745013009, + "grad_norm": 0.4074576199054718, + "learning_rate": 8.715363412157262e-06, + "loss": 0.0021, + "step": 15657 + }, + { + "epoch": 1.6975281873373809, + "grad_norm": 2.0320680141448975, + "learning_rate": 8.714638038589874e-06, + "loss": 0.0213, + "step": 15658 + }, + { + "epoch": 1.6976366001734604, + "grad_norm": 1.9440503120422363, + "learning_rate": 8.713912665022486e-06, + "loss": 0.0224, + "step": 15659 + }, + { + "epoch": 1.6977450130095404, + "grad_norm": 1.2587950229644775, + "learning_rate": 8.7131872914551e-06, + "loss": 0.0075, + "step": 15660 + }, + { + "epoch": 1.6978534258456202, + "grad_norm": 4.088563442230225, + "learning_rate": 8.712461917887712e-06, + "loss": 0.0848, + "step": 15661 + }, + { + "epoch": 1.6979618386817, + "grad_norm": 1.8111491203308105, + "learning_rate": 8.711736544320326e-06, + "loss": 0.0134, + "step": 15662 + }, + { + "epoch": 1.6980702515177797, + "grad_norm": 3.2255098819732666, + "learning_rate": 8.711011170752938e-06, + "loss": 0.0397, + "step": 15663 + }, + { + "epoch": 1.6981786643538594, + "grad_norm": 2.252246379852295, + "learning_rate": 8.710285797185552e-06, + "loss": 0.0446, + "step": 15664 + }, + { + "epoch": 1.6982870771899394, + "grad_norm": 0.4583914279937744, + "learning_rate": 8.709560423618164e-06, + "loss": 0.0044, + "step": 15665 + }, + { + "epoch": 1.698395490026019, + "grad_norm": 5.796005725860596, + "learning_rate": 8.708835050050776e-06, + "loss": 0.0132, + "step": 15666 + }, + { + "epoch": 1.698503902862099, + "grad_norm": 6.935263156890869, + "learning_rate": 8.708109676483389e-06, + "loss": 0.0277, + "step": 15667 + }, + { + "epoch": 1.6986123156981785, + "grad_norm": 8.786852836608887, + "learning_rate": 8.707384302916002e-06, + "loss": 0.0421, + "step": 15668 + }, + { + "epoch": 1.6987207285342585, + "grad_norm": 3.739163637161255, + "learning_rate": 8.706658929348615e-06, + "loss": 0.0212, + "step": 15669 + }, + { + "epoch": 1.6988291413703382, + "grad_norm": 1.1086286306381226, + "learning_rate": 8.705933555781228e-06, + "loss": 0.0059, + "step": 15670 + }, + { + "epoch": 1.698937554206418, + "grad_norm": 1.7058180570602417, + "learning_rate": 8.70520818221384e-06, + "loss": 0.0422, + "step": 15671 + }, + { + "epoch": 1.699045967042498, + "grad_norm": 2.3266172409057617, + "learning_rate": 8.704482808646454e-06, + "loss": 0.0375, + "step": 15672 + }, + { + "epoch": 1.6991543798785775, + "grad_norm": 3.2309963703155518, + "learning_rate": 8.703757435079067e-06, + "loss": 0.0353, + "step": 15673 + }, + { + "epoch": 1.6992627927146575, + "grad_norm": 1.2848824262619019, + "learning_rate": 8.703032061511679e-06, + "loss": 0.0259, + "step": 15674 + }, + { + "epoch": 1.699371205550737, + "grad_norm": 2.3875341415405273, + "learning_rate": 8.702306687944293e-06, + "loss": 0.0395, + "step": 15675 + }, + { + "epoch": 1.699479618386817, + "grad_norm": 0.0723206102848053, + "learning_rate": 8.701581314376905e-06, + "loss": 0.0011, + "step": 15676 + }, + { + "epoch": 1.6995880312228968, + "grad_norm": 6.345171928405762, + "learning_rate": 8.700855940809517e-06, + "loss": 0.0144, + "step": 15677 + }, + { + "epoch": 1.6996964440589766, + "grad_norm": 1.813197374343872, + "learning_rate": 8.70013056724213e-06, + "loss": 0.0132, + "step": 15678 + }, + { + "epoch": 1.6998048568950563, + "grad_norm": 1.7217738628387451, + "learning_rate": 8.699405193674745e-06, + "loss": 0.0471, + "step": 15679 + }, + { + "epoch": 1.699913269731136, + "grad_norm": 12.445157051086426, + "learning_rate": 8.698679820107357e-06, + "loss": 0.0223, + "step": 15680 + }, + { + "epoch": 1.700021682567216, + "grad_norm": 3.845507860183716, + "learning_rate": 8.697954446539969e-06, + "loss": 0.0147, + "step": 15681 + }, + { + "epoch": 1.7001300954032956, + "grad_norm": 33.76466369628906, + "learning_rate": 8.697229072972581e-06, + "loss": 0.0763, + "step": 15682 + }, + { + "epoch": 1.7002385082393756, + "grad_norm": 0.9981821775436401, + "learning_rate": 8.696503699405195e-06, + "loss": 0.0049, + "step": 15683 + }, + { + "epoch": 1.7003469210754554, + "grad_norm": 3.1919503211975098, + "learning_rate": 8.695778325837807e-06, + "loss": 0.0268, + "step": 15684 + }, + { + "epoch": 1.7004553339115351, + "grad_norm": 0.4763335883617401, + "learning_rate": 8.695052952270419e-06, + "loss": 0.0062, + "step": 15685 + }, + { + "epoch": 1.700563746747615, + "grad_norm": 1.555385947227478, + "learning_rate": 8.694327578703033e-06, + "loss": 0.012, + "step": 15686 + }, + { + "epoch": 1.7006721595836947, + "grad_norm": 1.8254330158233643, + "learning_rate": 8.693602205135647e-06, + "loss": 0.0323, + "step": 15687 + }, + { + "epoch": 1.7007805724197746, + "grad_norm": 1.0627174377441406, + "learning_rate": 8.692876831568259e-06, + "loss": 0.0032, + "step": 15688 + }, + { + "epoch": 1.7008889852558542, + "grad_norm": 1.939450740814209, + "learning_rate": 8.692151458000871e-06, + "loss": 0.0198, + "step": 15689 + }, + { + "epoch": 1.7009973980919342, + "grad_norm": 7.165807247161865, + "learning_rate": 8.691426084433485e-06, + "loss": 0.073, + "step": 15690 + }, + { + "epoch": 1.7011058109280137, + "grad_norm": 4.440041542053223, + "learning_rate": 8.690700710866097e-06, + "loss": 0.0356, + "step": 15691 + }, + { + "epoch": 1.7012142237640937, + "grad_norm": 0.3483026325702667, + "learning_rate": 8.68997533729871e-06, + "loss": 0.0033, + "step": 15692 + }, + { + "epoch": 1.7013226366001735, + "grad_norm": 5.112198829650879, + "learning_rate": 8.689249963731321e-06, + "loss": 0.0688, + "step": 15693 + }, + { + "epoch": 1.7014310494362532, + "grad_norm": 1.9655145406723022, + "learning_rate": 8.688524590163935e-06, + "loss": 0.0118, + "step": 15694 + }, + { + "epoch": 1.701539462272333, + "grad_norm": 1.5915627479553223, + "learning_rate": 8.687799216596547e-06, + "loss": 0.0328, + "step": 15695 + }, + { + "epoch": 1.7016478751084128, + "grad_norm": 2.185164213180542, + "learning_rate": 8.687073843029161e-06, + "loss": 0.0137, + "step": 15696 + }, + { + "epoch": 1.7017562879444927, + "grad_norm": 3.921577215194702, + "learning_rate": 8.686348469461773e-06, + "loss": 0.0336, + "step": 15697 + }, + { + "epoch": 1.7018647007805723, + "grad_norm": 0.6156889200210571, + "learning_rate": 8.685623095894387e-06, + "loss": 0.0041, + "step": 15698 + }, + { + "epoch": 1.7019731136166523, + "grad_norm": 0.36685121059417725, + "learning_rate": 8.684897722327e-06, + "loss": 0.0067, + "step": 15699 + }, + { + "epoch": 1.702081526452732, + "grad_norm": 1.1877385377883911, + "learning_rate": 8.684172348759611e-06, + "loss": 0.01, + "step": 15700 + }, + { + "epoch": 1.7021899392888118, + "grad_norm": 2.8164491653442383, + "learning_rate": 8.683446975192225e-06, + "loss": 0.0462, + "step": 15701 + }, + { + "epoch": 1.7022983521248916, + "grad_norm": 2.2262487411499023, + "learning_rate": 8.682721601624837e-06, + "loss": 0.0215, + "step": 15702 + }, + { + "epoch": 1.7024067649609713, + "grad_norm": 2.883009195327759, + "learning_rate": 8.68199622805745e-06, + "loss": 0.0632, + "step": 15703 + }, + { + "epoch": 1.7025151777970513, + "grad_norm": 1.8392388820648193, + "learning_rate": 8.681270854490063e-06, + "loss": 0.0262, + "step": 15704 + }, + { + "epoch": 1.7026235906331308, + "grad_norm": 0.375680536031723, + "learning_rate": 8.680545480922677e-06, + "loss": 0.0034, + "step": 15705 + }, + { + "epoch": 1.7027320034692108, + "grad_norm": 5.810029983520508, + "learning_rate": 8.67982010735529e-06, + "loss": 0.0423, + "step": 15706 + }, + { + "epoch": 1.7028404163052906, + "grad_norm": 6.39081335067749, + "learning_rate": 8.679094733787901e-06, + "loss": 0.0368, + "step": 15707 + }, + { + "epoch": 1.7029488291413704, + "grad_norm": 0.682942807674408, + "learning_rate": 8.678369360220514e-06, + "loss": 0.0071, + "step": 15708 + }, + { + "epoch": 1.7030572419774501, + "grad_norm": 0.4897747337818146, + "learning_rate": 8.677643986653127e-06, + "loss": 0.005, + "step": 15709 + }, + { + "epoch": 1.7031656548135299, + "grad_norm": 1.8175495862960815, + "learning_rate": 8.67691861308574e-06, + "loss": 0.0226, + "step": 15710 + }, + { + "epoch": 1.7032740676496099, + "grad_norm": 2.6934471130371094, + "learning_rate": 8.676193239518352e-06, + "loss": 0.0244, + "step": 15711 + }, + { + "epoch": 1.7033824804856894, + "grad_norm": 0.6105297803878784, + "learning_rate": 8.675467865950966e-06, + "loss": 0.0054, + "step": 15712 + }, + { + "epoch": 1.7034908933217694, + "grad_norm": 0.058420926332473755, + "learning_rate": 8.67474249238358e-06, + "loss": 0.0007, + "step": 15713 + }, + { + "epoch": 1.703599306157849, + "grad_norm": 6.557863235473633, + "learning_rate": 8.674017118816192e-06, + "loss": 0.0477, + "step": 15714 + }, + { + "epoch": 1.703707718993929, + "grad_norm": 1.8656736612319946, + "learning_rate": 8.673291745248804e-06, + "loss": 0.0095, + "step": 15715 + }, + { + "epoch": 1.7038161318300087, + "grad_norm": 1.883419394493103, + "learning_rate": 8.672566371681418e-06, + "loss": 0.0201, + "step": 15716 + }, + { + "epoch": 1.7039245446660884, + "grad_norm": 4.959756851196289, + "learning_rate": 8.67184099811403e-06, + "loss": 0.0376, + "step": 15717 + }, + { + "epoch": 1.7040329575021682, + "grad_norm": 3.0852301120758057, + "learning_rate": 8.671115624546642e-06, + "loss": 0.0509, + "step": 15718 + }, + { + "epoch": 1.704141370338248, + "grad_norm": 1.9050781726837158, + "learning_rate": 8.670390250979254e-06, + "loss": 0.0064, + "step": 15719 + }, + { + "epoch": 1.704249783174328, + "grad_norm": 1.1331278085708618, + "learning_rate": 8.669664877411868e-06, + "loss": 0.0089, + "step": 15720 + }, + { + "epoch": 1.7043581960104075, + "grad_norm": 4.862257480621338, + "learning_rate": 8.66893950384448e-06, + "loss": 0.0436, + "step": 15721 + }, + { + "epoch": 1.7044666088464875, + "grad_norm": 12.96123218536377, + "learning_rate": 8.668214130277094e-06, + "loss": 0.0382, + "step": 15722 + }, + { + "epoch": 1.7045750216825672, + "grad_norm": 2.567573070526123, + "learning_rate": 8.667488756709706e-06, + "loss": 0.0365, + "step": 15723 + }, + { + "epoch": 1.704683434518647, + "grad_norm": 0.5436980724334717, + "learning_rate": 8.66676338314232e-06, + "loss": 0.0037, + "step": 15724 + }, + { + "epoch": 1.7047918473547268, + "grad_norm": 1.8187484741210938, + "learning_rate": 8.666038009574932e-06, + "loss": 0.0313, + "step": 15725 + }, + { + "epoch": 1.7049002601908065, + "grad_norm": 2.4705164432525635, + "learning_rate": 8.665312636007544e-06, + "loss": 0.0133, + "step": 15726 + }, + { + "epoch": 1.7050086730268865, + "grad_norm": 1.3428136110305786, + "learning_rate": 8.664587262440158e-06, + "loss": 0.0127, + "step": 15727 + }, + { + "epoch": 1.705117085862966, + "grad_norm": 3.5027217864990234, + "learning_rate": 8.66386188887277e-06, + "loss": 0.0441, + "step": 15728 + }, + { + "epoch": 1.705225498699046, + "grad_norm": 2.515709400177002, + "learning_rate": 8.663136515305382e-06, + "loss": 0.0123, + "step": 15729 + }, + { + "epoch": 1.7053339115351258, + "grad_norm": 7.146026134490967, + "learning_rate": 8.662411141737996e-06, + "loss": 0.0312, + "step": 15730 + }, + { + "epoch": 1.7054423243712056, + "grad_norm": 6.916526794433594, + "learning_rate": 8.66168576817061e-06, + "loss": 0.0108, + "step": 15731 + }, + { + "epoch": 1.7055507372072853, + "grad_norm": 1.5122565031051636, + "learning_rate": 8.660960394603222e-06, + "loss": 0.0076, + "step": 15732 + }, + { + "epoch": 1.705659150043365, + "grad_norm": 6.131049633026123, + "learning_rate": 8.660235021035834e-06, + "loss": 0.0189, + "step": 15733 + }, + { + "epoch": 1.705767562879445, + "grad_norm": 1.918668270111084, + "learning_rate": 8.659509647468446e-06, + "loss": 0.0216, + "step": 15734 + }, + { + "epoch": 1.7058759757155246, + "grad_norm": 0.5416647791862488, + "learning_rate": 8.65878427390106e-06, + "loss": 0.0041, + "step": 15735 + }, + { + "epoch": 1.7059843885516046, + "grad_norm": 1.3495819568634033, + "learning_rate": 8.658058900333672e-06, + "loss": 0.0277, + "step": 15736 + }, + { + "epoch": 1.7060928013876842, + "grad_norm": 1.5771161317825317, + "learning_rate": 8.657333526766284e-06, + "loss": 0.0354, + "step": 15737 + }, + { + "epoch": 1.7062012142237641, + "grad_norm": 1.3025964498519897, + "learning_rate": 8.656608153198898e-06, + "loss": 0.0157, + "step": 15738 + }, + { + "epoch": 1.706309627059844, + "grad_norm": 1.5127851963043213, + "learning_rate": 8.655882779631512e-06, + "loss": 0.0081, + "step": 15739 + }, + { + "epoch": 1.7064180398959237, + "grad_norm": 2.2722859382629395, + "learning_rate": 8.655157406064124e-06, + "loss": 0.0256, + "step": 15740 + }, + { + "epoch": 1.7065264527320034, + "grad_norm": 2.846975088119507, + "learning_rate": 8.654432032496736e-06, + "loss": 0.0394, + "step": 15741 + }, + { + "epoch": 1.7066348655680832, + "grad_norm": 2.1544103622436523, + "learning_rate": 8.65370665892935e-06, + "loss": 0.0338, + "step": 15742 + }, + { + "epoch": 1.7067432784041632, + "grad_norm": 0.37065860629081726, + "learning_rate": 8.652981285361962e-06, + "loss": 0.0025, + "step": 15743 + }, + { + "epoch": 1.7068516912402427, + "grad_norm": 9.298243522644043, + "learning_rate": 8.652255911794574e-06, + "loss": 0.0814, + "step": 15744 + }, + { + "epoch": 1.7069601040763227, + "grad_norm": 0.6065229773521423, + "learning_rate": 8.651530538227187e-06, + "loss": 0.0147, + "step": 15745 + }, + { + "epoch": 1.7070685169124025, + "grad_norm": 4.547708511352539, + "learning_rate": 8.6508051646598e-06, + "loss": 0.02, + "step": 15746 + }, + { + "epoch": 1.7071769297484822, + "grad_norm": 7.666015148162842, + "learning_rate": 8.650079791092413e-06, + "loss": 0.0227, + "step": 15747 + }, + { + "epoch": 1.707285342584562, + "grad_norm": 0.17455418407917023, + "learning_rate": 8.649354417525026e-06, + "loss": 0.0029, + "step": 15748 + }, + { + "epoch": 1.7073937554206418, + "grad_norm": 0.5801404118537903, + "learning_rate": 8.648629043957639e-06, + "loss": 0.0107, + "step": 15749 + }, + { + "epoch": 1.7075021682567217, + "grad_norm": 8.006098747253418, + "learning_rate": 8.647903670390252e-06, + "loss": 0.0351, + "step": 15750 + }, + { + "epoch": 1.7076105810928013, + "grad_norm": 0.3777465224266052, + "learning_rate": 8.647178296822865e-06, + "loss": 0.0029, + "step": 15751 + }, + { + "epoch": 1.7077189939288813, + "grad_norm": 1.0397440195083618, + "learning_rate": 8.646452923255477e-06, + "loss": 0.015, + "step": 15752 + }, + { + "epoch": 1.7078274067649608, + "grad_norm": 0.443929523229599, + "learning_rate": 8.64572754968809e-06, + "loss": 0.0025, + "step": 15753 + }, + { + "epoch": 1.7079358196010408, + "grad_norm": 5.2439656257629395, + "learning_rate": 8.645002176120703e-06, + "loss": 0.0771, + "step": 15754 + }, + { + "epoch": 1.7080442324371206, + "grad_norm": 2.9786264896392822, + "learning_rate": 8.644276802553315e-06, + "loss": 0.0496, + "step": 15755 + }, + { + "epoch": 1.7081526452732003, + "grad_norm": 1.1777373552322388, + "learning_rate": 8.643551428985929e-06, + "loss": 0.0179, + "step": 15756 + }, + { + "epoch": 1.70826105810928, + "grad_norm": 0.24816006422042847, + "learning_rate": 8.642826055418543e-06, + "loss": 0.0019, + "step": 15757 + }, + { + "epoch": 1.7083694709453598, + "grad_norm": 0.4631592929363251, + "learning_rate": 8.642100681851155e-06, + "loss": 0.005, + "step": 15758 + }, + { + "epoch": 1.7084778837814398, + "grad_norm": 2.236940622329712, + "learning_rate": 8.641375308283767e-06, + "loss": 0.0271, + "step": 15759 + }, + { + "epoch": 1.7085862966175194, + "grad_norm": 5.326993465423584, + "learning_rate": 8.640649934716379e-06, + "loss": 0.0294, + "step": 15760 + }, + { + "epoch": 1.7086947094535994, + "grad_norm": 3.8832850456237793, + "learning_rate": 8.639924561148993e-06, + "loss": 0.0139, + "step": 15761 + }, + { + "epoch": 1.7088031222896791, + "grad_norm": 1.0714091062545776, + "learning_rate": 8.639199187581605e-06, + "loss": 0.0224, + "step": 15762 + }, + { + "epoch": 1.7089115351257589, + "grad_norm": 0.22728711366653442, + "learning_rate": 8.638473814014217e-06, + "loss": 0.0026, + "step": 15763 + }, + { + "epoch": 1.7090199479618386, + "grad_norm": 1.744515061378479, + "learning_rate": 8.637748440446831e-06, + "loss": 0.0182, + "step": 15764 + }, + { + "epoch": 1.7091283607979184, + "grad_norm": 1.3605915307998657, + "learning_rate": 8.637023066879445e-06, + "loss": 0.0509, + "step": 15765 + }, + { + "epoch": 1.7092367736339984, + "grad_norm": 2.587247371673584, + "learning_rate": 8.636297693312057e-06, + "loss": 0.0341, + "step": 15766 + }, + { + "epoch": 1.709345186470078, + "grad_norm": 2.3137683868408203, + "learning_rate": 8.635572319744669e-06, + "loss": 0.0139, + "step": 15767 + }, + { + "epoch": 1.709453599306158, + "grad_norm": 0.37126776576042175, + "learning_rate": 8.634846946177283e-06, + "loss": 0.0029, + "step": 15768 + }, + { + "epoch": 1.7095620121422377, + "grad_norm": 1.1102845668792725, + "learning_rate": 8.634121572609895e-06, + "loss": 0.0123, + "step": 15769 + }, + { + "epoch": 1.7096704249783174, + "grad_norm": 3.367042303085327, + "learning_rate": 8.633396199042507e-06, + "loss": 0.0226, + "step": 15770 + }, + { + "epoch": 1.7097788378143972, + "grad_norm": 0.64400315284729, + "learning_rate": 8.63267082547512e-06, + "loss": 0.0105, + "step": 15771 + }, + { + "epoch": 1.709887250650477, + "grad_norm": 1.2962874174118042, + "learning_rate": 8.631945451907733e-06, + "loss": 0.0231, + "step": 15772 + }, + { + "epoch": 1.709995663486557, + "grad_norm": 1.0326874256134033, + "learning_rate": 8.631220078340345e-06, + "loss": 0.0135, + "step": 15773 + }, + { + "epoch": 1.7101040763226365, + "grad_norm": 1.6588565111160278, + "learning_rate": 8.630494704772959e-06, + "loss": 0.015, + "step": 15774 + }, + { + "epoch": 1.7102124891587165, + "grad_norm": 0.8280923366546631, + "learning_rate": 8.629769331205571e-06, + "loss": 0.0052, + "step": 15775 + }, + { + "epoch": 1.710320901994796, + "grad_norm": 0.595763087272644, + "learning_rate": 8.629043957638185e-06, + "loss": 0.0046, + "step": 15776 + }, + { + "epoch": 1.710429314830876, + "grad_norm": 2.7343637943267822, + "learning_rate": 8.628318584070797e-06, + "loss": 0.0076, + "step": 15777 + }, + { + "epoch": 1.7105377276669558, + "grad_norm": 1.6968094110488892, + "learning_rate": 8.62759321050341e-06, + "loss": 0.0273, + "step": 15778 + }, + { + "epoch": 1.7106461405030355, + "grad_norm": 3.059284210205078, + "learning_rate": 8.626867836936023e-06, + "loss": 0.0186, + "step": 15779 + }, + { + "epoch": 1.7107545533391153, + "grad_norm": 2.9791948795318604, + "learning_rate": 8.626142463368635e-06, + "loss": 0.0086, + "step": 15780 + }, + { + "epoch": 1.710862966175195, + "grad_norm": 0.7415115237236023, + "learning_rate": 8.625417089801248e-06, + "loss": 0.026, + "step": 15781 + }, + { + "epoch": 1.710971379011275, + "grad_norm": 0.8152691721916199, + "learning_rate": 8.624691716233861e-06, + "loss": 0.0091, + "step": 15782 + }, + { + "epoch": 1.7110797918473546, + "grad_norm": 3.763838529586792, + "learning_rate": 8.623966342666475e-06, + "loss": 0.032, + "step": 15783 + }, + { + "epoch": 1.7111882046834346, + "grad_norm": 1.1899422407150269, + "learning_rate": 8.623240969099087e-06, + "loss": 0.0037, + "step": 15784 + }, + { + "epoch": 1.7112966175195143, + "grad_norm": 5.484167575836182, + "learning_rate": 8.6225155955317e-06, + "loss": 0.0242, + "step": 15785 + }, + { + "epoch": 1.711405030355594, + "grad_norm": 2.7613930702209473, + "learning_rate": 8.621790221964312e-06, + "loss": 0.0104, + "step": 15786 + }, + { + "epoch": 1.7115134431916739, + "grad_norm": 4.238261699676514, + "learning_rate": 8.621064848396925e-06, + "loss": 0.042, + "step": 15787 + }, + { + "epoch": 1.7116218560277536, + "grad_norm": 0.17294499278068542, + "learning_rate": 8.620339474829538e-06, + "loss": 0.0017, + "step": 15788 + }, + { + "epoch": 1.7117302688638336, + "grad_norm": 1.5093348026275635, + "learning_rate": 8.61961410126215e-06, + "loss": 0.0092, + "step": 15789 + }, + { + "epoch": 1.7118386816999132, + "grad_norm": 2.1591827869415283, + "learning_rate": 8.618888727694764e-06, + "loss": 0.0045, + "step": 15790 + }, + { + "epoch": 1.7119470945359931, + "grad_norm": 1.2541403770446777, + "learning_rate": 8.618163354127377e-06, + "loss": 0.0102, + "step": 15791 + }, + { + "epoch": 1.712055507372073, + "grad_norm": 0.7523918747901917, + "learning_rate": 8.61743798055999e-06, + "loss": 0.006, + "step": 15792 + }, + { + "epoch": 1.7121639202081527, + "grad_norm": 0.41331541538238525, + "learning_rate": 8.616712606992602e-06, + "loss": 0.0014, + "step": 15793 + }, + { + "epoch": 1.7122723330442324, + "grad_norm": 1.6225792169570923, + "learning_rate": 8.615987233425216e-06, + "loss": 0.0251, + "step": 15794 + }, + { + "epoch": 1.7123807458803122, + "grad_norm": 3.0809760093688965, + "learning_rate": 8.615261859857828e-06, + "loss": 0.0246, + "step": 15795 + }, + { + "epoch": 1.7124891587163922, + "grad_norm": 2.1982669830322266, + "learning_rate": 8.61453648629044e-06, + "loss": 0.015, + "step": 15796 + }, + { + "epoch": 1.7125975715524717, + "grad_norm": 2.0313470363616943, + "learning_rate": 8.613811112723052e-06, + "loss": 0.0186, + "step": 15797 + }, + { + "epoch": 1.7127059843885517, + "grad_norm": 0.22737452387809753, + "learning_rate": 8.613085739155666e-06, + "loss": 0.0008, + "step": 15798 + }, + { + "epoch": 1.7128143972246312, + "grad_norm": 0.3919694721698761, + "learning_rate": 8.612360365588278e-06, + "loss": 0.0018, + "step": 15799 + }, + { + "epoch": 1.7129228100607112, + "grad_norm": 2.677621603012085, + "learning_rate": 8.611634992020892e-06, + "loss": 0.0574, + "step": 15800 + }, + { + "epoch": 1.713031222896791, + "grad_norm": 0.6101173758506775, + "learning_rate": 8.610909618453504e-06, + "loss": 0.0033, + "step": 15801 + }, + { + "epoch": 1.7131396357328708, + "grad_norm": 0.6284573078155518, + "learning_rate": 8.610184244886118e-06, + "loss": 0.0028, + "step": 15802 + }, + { + "epoch": 1.7132480485689505, + "grad_norm": 0.20557673275470734, + "learning_rate": 8.60945887131873e-06, + "loss": 0.0021, + "step": 15803 + }, + { + "epoch": 1.7133564614050303, + "grad_norm": 3.3968000411987305, + "learning_rate": 8.608733497751342e-06, + "loss": 0.0402, + "step": 15804 + }, + { + "epoch": 1.7134648742411103, + "grad_norm": 8.29111099243164, + "learning_rate": 8.608008124183956e-06, + "loss": 0.0174, + "step": 15805 + }, + { + "epoch": 1.7135732870771898, + "grad_norm": 3.1240453720092773, + "learning_rate": 8.607282750616568e-06, + "loss": 0.0088, + "step": 15806 + }, + { + "epoch": 1.7136816999132698, + "grad_norm": 4.572210311889648, + "learning_rate": 8.60655737704918e-06, + "loss": 0.0136, + "step": 15807 + }, + { + "epoch": 1.7137901127493496, + "grad_norm": 7.259304046630859, + "learning_rate": 8.605832003481794e-06, + "loss": 0.0407, + "step": 15808 + }, + { + "epoch": 1.7138985255854293, + "grad_norm": 0.06824268400669098, + "learning_rate": 8.605106629914408e-06, + "loss": 0.0006, + "step": 15809 + }, + { + "epoch": 1.714006938421509, + "grad_norm": 2.7016818523406982, + "learning_rate": 8.60438125634702e-06, + "loss": 0.0457, + "step": 15810 + }, + { + "epoch": 1.7141153512575888, + "grad_norm": 5.622878551483154, + "learning_rate": 8.603655882779632e-06, + "loss": 0.0603, + "step": 15811 + }, + { + "epoch": 1.7142237640936688, + "grad_norm": 2.4673593044281006, + "learning_rate": 8.602930509212244e-06, + "loss": 0.0304, + "step": 15812 + }, + { + "epoch": 1.7143321769297484, + "grad_norm": 6.4906392097473145, + "learning_rate": 8.602205135644858e-06, + "loss": 0.0191, + "step": 15813 + }, + { + "epoch": 1.7144405897658284, + "grad_norm": 6.352865219116211, + "learning_rate": 8.60147976207747e-06, + "loss": 0.0353, + "step": 15814 + }, + { + "epoch": 1.714549002601908, + "grad_norm": 2.4683539867401123, + "learning_rate": 8.600754388510082e-06, + "loss": 0.0279, + "step": 15815 + }, + { + "epoch": 1.7146574154379879, + "grad_norm": 1.7899541854858398, + "learning_rate": 8.600029014942696e-06, + "loss": 0.0533, + "step": 15816 + }, + { + "epoch": 1.7147658282740676, + "grad_norm": 2.2052175998687744, + "learning_rate": 8.59930364137531e-06, + "loss": 0.0154, + "step": 15817 + }, + { + "epoch": 1.7148742411101474, + "grad_norm": 0.8146502375602722, + "learning_rate": 8.598578267807922e-06, + "loss": 0.0109, + "step": 15818 + }, + { + "epoch": 1.7149826539462274, + "grad_norm": 10.031349182128906, + "learning_rate": 8.597852894240534e-06, + "loss": 0.0352, + "step": 15819 + }, + { + "epoch": 1.715091066782307, + "grad_norm": 7.484759330749512, + "learning_rate": 8.597127520673148e-06, + "loss": 0.0524, + "step": 15820 + }, + { + "epoch": 1.715199479618387, + "grad_norm": 3.610250473022461, + "learning_rate": 8.59640214710576e-06, + "loss": 0.02, + "step": 15821 + }, + { + "epoch": 1.7153078924544665, + "grad_norm": 0.46307334303855896, + "learning_rate": 8.595676773538373e-06, + "loss": 0.0022, + "step": 15822 + }, + { + "epoch": 1.7154163052905465, + "grad_norm": 1.216667652130127, + "learning_rate": 8.594951399970985e-06, + "loss": 0.0071, + "step": 15823 + }, + { + "epoch": 1.7155247181266262, + "grad_norm": 2.9766852855682373, + "learning_rate": 8.594226026403598e-06, + "loss": 0.0892, + "step": 15824 + }, + { + "epoch": 1.715633130962706, + "grad_norm": 7.887444019317627, + "learning_rate": 8.59350065283621e-06, + "loss": 0.0408, + "step": 15825 + }, + { + "epoch": 1.7157415437987857, + "grad_norm": 1.0490587949752808, + "learning_rate": 8.592775279268824e-06, + "loss": 0.0036, + "step": 15826 + }, + { + "epoch": 1.7158499566348655, + "grad_norm": 0.015571482479572296, + "learning_rate": 8.592049905701437e-06, + "loss": 0.0002, + "step": 15827 + }, + { + "epoch": 1.7159583694709455, + "grad_norm": 0.8123306632041931, + "learning_rate": 8.59132453213405e-06, + "loss": 0.0044, + "step": 15828 + }, + { + "epoch": 1.716066782307025, + "grad_norm": 1.7269291877746582, + "learning_rate": 8.590599158566663e-06, + "loss": 0.0102, + "step": 15829 + }, + { + "epoch": 1.716175195143105, + "grad_norm": 0.7888323664665222, + "learning_rate": 8.589873784999275e-06, + "loss": 0.0067, + "step": 15830 + }, + { + "epoch": 1.7162836079791848, + "grad_norm": 0.7630396485328674, + "learning_rate": 8.589148411431889e-06, + "loss": 0.0055, + "step": 15831 + }, + { + "epoch": 1.7163920208152645, + "grad_norm": 16.245237350463867, + "learning_rate": 8.5884230378645e-06, + "loss": 0.0086, + "step": 15832 + }, + { + "epoch": 1.7165004336513443, + "grad_norm": 4.026881217956543, + "learning_rate": 8.587697664297113e-06, + "loss": 0.0306, + "step": 15833 + }, + { + "epoch": 1.716608846487424, + "grad_norm": 3.1427812576293945, + "learning_rate": 8.586972290729727e-06, + "loss": 0.0573, + "step": 15834 + }, + { + "epoch": 1.716717259323504, + "grad_norm": 2.987919569015503, + "learning_rate": 8.58624691716234e-06, + "loss": 0.0167, + "step": 15835 + }, + { + "epoch": 1.7168256721595836, + "grad_norm": 1.0022931098937988, + "learning_rate": 8.585521543594953e-06, + "loss": 0.0127, + "step": 15836 + }, + { + "epoch": 1.7169340849956636, + "grad_norm": 2.019001007080078, + "learning_rate": 8.584796170027565e-06, + "loss": 0.005, + "step": 15837 + }, + { + "epoch": 1.7170424978317431, + "grad_norm": 2.5714659690856934, + "learning_rate": 8.584070796460177e-06, + "loss": 0.0177, + "step": 15838 + }, + { + "epoch": 1.717150910667823, + "grad_norm": 3.589599132537842, + "learning_rate": 8.58334542289279e-06, + "loss": 0.0602, + "step": 15839 + }, + { + "epoch": 1.7172593235039029, + "grad_norm": 0.2708865702152252, + "learning_rate": 8.582620049325403e-06, + "loss": 0.0032, + "step": 15840 + }, + { + "epoch": 1.7173677363399826, + "grad_norm": 0.9285496473312378, + "learning_rate": 8.581894675758015e-06, + "loss": 0.0172, + "step": 15841 + }, + { + "epoch": 1.7174761491760624, + "grad_norm": 7.83437442779541, + "learning_rate": 8.581169302190629e-06, + "loss": 0.0495, + "step": 15842 + }, + { + "epoch": 1.7175845620121422, + "grad_norm": 7.61989164352417, + "learning_rate": 8.580443928623243e-06, + "loss": 0.0389, + "step": 15843 + }, + { + "epoch": 1.7176929748482221, + "grad_norm": 0.6415976881980896, + "learning_rate": 8.579718555055855e-06, + "loss": 0.0043, + "step": 15844 + }, + { + "epoch": 1.7178013876843017, + "grad_norm": 1.3117856979370117, + "learning_rate": 8.578993181488467e-06, + "loss": 0.0062, + "step": 15845 + }, + { + "epoch": 1.7179098005203817, + "grad_norm": 2.6452739238739014, + "learning_rate": 8.578267807921081e-06, + "loss": 0.0529, + "step": 15846 + }, + { + "epoch": 1.7180182133564614, + "grad_norm": 0.18526190519332886, + "learning_rate": 8.577542434353693e-06, + "loss": 0.0019, + "step": 15847 + }, + { + "epoch": 1.7181266261925412, + "grad_norm": 2.399566650390625, + "learning_rate": 8.576817060786305e-06, + "loss": 0.0096, + "step": 15848 + }, + { + "epoch": 1.718235039028621, + "grad_norm": 11.425283432006836, + "learning_rate": 8.576091687218919e-06, + "loss": 0.0384, + "step": 15849 + }, + { + "epoch": 1.7183434518647007, + "grad_norm": 6.983121395111084, + "learning_rate": 8.575366313651531e-06, + "loss": 0.0256, + "step": 15850 + }, + { + "epoch": 1.7184518647007807, + "grad_norm": 3.745596170425415, + "learning_rate": 8.574640940084143e-06, + "loss": 0.0364, + "step": 15851 + }, + { + "epoch": 1.7185602775368602, + "grad_norm": 1.070593237876892, + "learning_rate": 8.573915566516757e-06, + "loss": 0.0098, + "step": 15852 + }, + { + "epoch": 1.7186686903729402, + "grad_norm": 0.6382126808166504, + "learning_rate": 8.57319019294937e-06, + "loss": 0.0058, + "step": 15853 + }, + { + "epoch": 1.71877710320902, + "grad_norm": 3.476198673248291, + "learning_rate": 8.572464819381983e-06, + "loss": 0.0314, + "step": 15854 + }, + { + "epoch": 1.7188855160450998, + "grad_norm": 0.8855233192443848, + "learning_rate": 8.571739445814595e-06, + "loss": 0.0049, + "step": 15855 + }, + { + "epoch": 1.7189939288811795, + "grad_norm": 3.1502676010131836, + "learning_rate": 8.571014072247207e-06, + "loss": 0.0352, + "step": 15856 + }, + { + "epoch": 1.7191023417172593, + "grad_norm": 0.46114686131477356, + "learning_rate": 8.570288698679821e-06, + "loss": 0.0022, + "step": 15857 + }, + { + "epoch": 1.7192107545533393, + "grad_norm": 6.607454776763916, + "learning_rate": 8.569563325112433e-06, + "loss": 0.0807, + "step": 15858 + }, + { + "epoch": 1.7193191673894188, + "grad_norm": 2.36627459526062, + "learning_rate": 8.568837951545046e-06, + "loss": 0.0259, + "step": 15859 + }, + { + "epoch": 1.7194275802254988, + "grad_norm": 2.8239986896514893, + "learning_rate": 8.56811257797766e-06, + "loss": 0.0227, + "step": 15860 + }, + { + "epoch": 1.7195359930615783, + "grad_norm": 1.1753041744232178, + "learning_rate": 8.567387204410273e-06, + "loss": 0.0294, + "step": 15861 + }, + { + "epoch": 1.7196444058976583, + "grad_norm": 0.10826528072357178, + "learning_rate": 8.566661830842885e-06, + "loss": 0.001, + "step": 15862 + }, + { + "epoch": 1.719752818733738, + "grad_norm": 0.22989198565483093, + "learning_rate": 8.565936457275497e-06, + "loss": 0.0005, + "step": 15863 + }, + { + "epoch": 1.7198612315698178, + "grad_norm": 8.407045364379883, + "learning_rate": 8.56521108370811e-06, + "loss": 0.0674, + "step": 15864 + }, + { + "epoch": 1.7199696444058976, + "grad_norm": 9.151525497436523, + "learning_rate": 8.564485710140723e-06, + "loss": 0.0783, + "step": 15865 + }, + { + "epoch": 1.7200780572419774, + "grad_norm": 2.5598835945129395, + "learning_rate": 8.563760336573336e-06, + "loss": 0.0307, + "step": 15866 + }, + { + "epoch": 1.7201864700780574, + "grad_norm": 0.5169001817703247, + "learning_rate": 8.563034963005948e-06, + "loss": 0.0041, + "step": 15867 + }, + { + "epoch": 1.720294882914137, + "grad_norm": 1.9723060131072998, + "learning_rate": 8.562309589438562e-06, + "loss": 0.0076, + "step": 15868 + }, + { + "epoch": 1.7204032957502169, + "grad_norm": 0.130199134349823, + "learning_rate": 8.561584215871175e-06, + "loss": 0.0019, + "step": 15869 + }, + { + "epoch": 1.7205117085862967, + "grad_norm": 4.749673366546631, + "learning_rate": 8.560858842303788e-06, + "loss": 0.0144, + "step": 15870 + }, + { + "epoch": 1.7206201214223764, + "grad_norm": 11.728070259094238, + "learning_rate": 8.5601334687364e-06, + "loss": 0.0538, + "step": 15871 + }, + { + "epoch": 1.7207285342584562, + "grad_norm": 0.09596271812915802, + "learning_rate": 8.559408095169014e-06, + "loss": 0.0006, + "step": 15872 + }, + { + "epoch": 1.720836947094536, + "grad_norm": 1.9072248935699463, + "learning_rate": 8.558682721601626e-06, + "loss": 0.0053, + "step": 15873 + }, + { + "epoch": 1.720945359930616, + "grad_norm": 2.558788776397705, + "learning_rate": 8.557957348034238e-06, + "loss": 0.0358, + "step": 15874 + }, + { + "epoch": 1.7210537727666955, + "grad_norm": 1.4504578113555908, + "learning_rate": 8.557231974466852e-06, + "loss": 0.0081, + "step": 15875 + }, + { + "epoch": 1.7211621856027755, + "grad_norm": 1.1506555080413818, + "learning_rate": 8.556506600899464e-06, + "loss": 0.0168, + "step": 15876 + }, + { + "epoch": 1.7212705984388552, + "grad_norm": 3.2423603534698486, + "learning_rate": 8.555781227332076e-06, + "loss": 0.0288, + "step": 15877 + }, + { + "epoch": 1.721379011274935, + "grad_norm": 2.551732301712036, + "learning_rate": 8.55505585376469e-06, + "loss": 0.0506, + "step": 15878 + }, + { + "epoch": 1.7214874241110147, + "grad_norm": 5.725917339324951, + "learning_rate": 8.554330480197302e-06, + "loss": 0.0355, + "step": 15879 + }, + { + "epoch": 1.7215958369470945, + "grad_norm": 4.558155059814453, + "learning_rate": 8.553605106629916e-06, + "loss": 0.0155, + "step": 15880 + }, + { + "epoch": 1.7217042497831745, + "grad_norm": 0.6514669060707092, + "learning_rate": 8.552879733062528e-06, + "loss": 0.0077, + "step": 15881 + }, + { + "epoch": 1.721812662619254, + "grad_norm": 2.9296252727508545, + "learning_rate": 8.55215435949514e-06, + "loss": 0.0277, + "step": 15882 + }, + { + "epoch": 1.721921075455334, + "grad_norm": 5.635551929473877, + "learning_rate": 8.551428985927754e-06, + "loss": 0.0354, + "step": 15883 + }, + { + "epoch": 1.7220294882914136, + "grad_norm": 3.7125582695007324, + "learning_rate": 8.550703612360366e-06, + "loss": 0.0204, + "step": 15884 + }, + { + "epoch": 1.7221379011274935, + "grad_norm": 0.3218226730823517, + "learning_rate": 8.549978238792978e-06, + "loss": 0.0028, + "step": 15885 + }, + { + "epoch": 1.7222463139635733, + "grad_norm": 0.5821440815925598, + "learning_rate": 8.549252865225592e-06, + "loss": 0.0052, + "step": 15886 + }, + { + "epoch": 1.722354726799653, + "grad_norm": 0.26623767614364624, + "learning_rate": 8.548527491658206e-06, + "loss": 0.0036, + "step": 15887 + }, + { + "epoch": 1.7224631396357328, + "grad_norm": 3.5460057258605957, + "learning_rate": 8.547802118090818e-06, + "loss": 0.0169, + "step": 15888 + }, + { + "epoch": 1.7225715524718126, + "grad_norm": 8.719625473022461, + "learning_rate": 8.54707674452343e-06, + "loss": 0.0548, + "step": 15889 + }, + { + "epoch": 1.7226799653078926, + "grad_norm": 2.712092399597168, + "learning_rate": 8.546351370956042e-06, + "loss": 0.0154, + "step": 15890 + }, + { + "epoch": 1.7227883781439721, + "grad_norm": 4.441909313201904, + "learning_rate": 8.545625997388656e-06, + "loss": 0.0831, + "step": 15891 + }, + { + "epoch": 1.722896790980052, + "grad_norm": 6.821010112762451, + "learning_rate": 8.544900623821268e-06, + "loss": 0.0232, + "step": 15892 + }, + { + "epoch": 1.7230052038161319, + "grad_norm": 3.0148212909698486, + "learning_rate": 8.54417525025388e-06, + "loss": 0.0109, + "step": 15893 + }, + { + "epoch": 1.7231136166522116, + "grad_norm": 2.3121116161346436, + "learning_rate": 8.543449876686494e-06, + "loss": 0.0334, + "step": 15894 + }, + { + "epoch": 1.7232220294882914, + "grad_norm": 0.697995662689209, + "learning_rate": 8.542724503119108e-06, + "loss": 0.0037, + "step": 15895 + }, + { + "epoch": 1.7233304423243712, + "grad_norm": 0.2564499080181122, + "learning_rate": 8.54199912955172e-06, + "loss": 0.0019, + "step": 15896 + }, + { + "epoch": 1.7234388551604511, + "grad_norm": 0.5705287456512451, + "learning_rate": 8.541273755984332e-06, + "loss": 0.0049, + "step": 15897 + }, + { + "epoch": 1.7235472679965307, + "grad_norm": 1.2794439792633057, + "learning_rate": 8.540548382416946e-06, + "loss": 0.0062, + "step": 15898 + }, + { + "epoch": 1.7236556808326107, + "grad_norm": 1.695432186126709, + "learning_rate": 8.539823008849558e-06, + "loss": 0.0179, + "step": 15899 + }, + { + "epoch": 1.7237640936686902, + "grad_norm": 0.7626966238021851, + "learning_rate": 8.53909763528217e-06, + "loss": 0.0064, + "step": 15900 + }, + { + "epoch": 1.7238725065047702, + "grad_norm": 2.1697704792022705, + "learning_rate": 8.538372261714784e-06, + "loss": 0.0207, + "step": 15901 + }, + { + "epoch": 1.72398091934085, + "grad_norm": 6.670401573181152, + "learning_rate": 8.537646888147396e-06, + "loss": 0.0341, + "step": 15902 + }, + { + "epoch": 1.7240893321769297, + "grad_norm": 2.624791383743286, + "learning_rate": 8.53692151458001e-06, + "loss": 0.0234, + "step": 15903 + }, + { + "epoch": 1.7241977450130095, + "grad_norm": 2.6270670890808105, + "learning_rate": 8.536196141012622e-06, + "loss": 0.0555, + "step": 15904 + }, + { + "epoch": 1.7243061578490892, + "grad_norm": 2.2880351543426514, + "learning_rate": 8.535470767445235e-06, + "loss": 0.0247, + "step": 15905 + }, + { + "epoch": 1.7244145706851692, + "grad_norm": 2.327049970626831, + "learning_rate": 8.534745393877848e-06, + "loss": 0.0102, + "step": 15906 + }, + { + "epoch": 1.7245229835212488, + "grad_norm": 1.425176978111267, + "learning_rate": 8.53402002031046e-06, + "loss": 0.0185, + "step": 15907 + }, + { + "epoch": 1.7246313963573288, + "grad_norm": 5.717260837554932, + "learning_rate": 8.533294646743073e-06, + "loss": 0.0228, + "step": 15908 + }, + { + "epoch": 1.7247398091934085, + "grad_norm": 3.3631668090820312, + "learning_rate": 8.532569273175687e-06, + "loss": 0.0194, + "step": 15909 + }, + { + "epoch": 1.7248482220294883, + "grad_norm": 1.7914252281188965, + "learning_rate": 8.531843899608299e-06, + "loss": 0.0278, + "step": 15910 + }, + { + "epoch": 1.724956634865568, + "grad_norm": 2.0107221603393555, + "learning_rate": 8.531118526040911e-06, + "loss": 0.0531, + "step": 15911 + }, + { + "epoch": 1.7250650477016478, + "grad_norm": 3.526685953140259, + "learning_rate": 8.530393152473525e-06, + "loss": 0.0403, + "step": 15912 + }, + { + "epoch": 1.7251734605377278, + "grad_norm": 5.435093879699707, + "learning_rate": 8.529667778906139e-06, + "loss": 0.0296, + "step": 15913 + }, + { + "epoch": 1.7252818733738073, + "grad_norm": 9.93691635131836, + "learning_rate": 8.52894240533875e-06, + "loss": 0.0384, + "step": 15914 + }, + { + "epoch": 1.7253902862098873, + "grad_norm": 0.41719141602516174, + "learning_rate": 8.528217031771363e-06, + "loss": 0.0019, + "step": 15915 + }, + { + "epoch": 1.725498699045967, + "grad_norm": 1.4318678379058838, + "learning_rate": 8.527491658203975e-06, + "loss": 0.0207, + "step": 15916 + }, + { + "epoch": 1.7256071118820469, + "grad_norm": 0.3692716062068939, + "learning_rate": 8.526766284636589e-06, + "loss": 0.0051, + "step": 15917 + }, + { + "epoch": 1.7257155247181266, + "grad_norm": 1.7008286714553833, + "learning_rate": 8.526040911069201e-06, + "loss": 0.0032, + "step": 15918 + }, + { + "epoch": 1.7258239375542064, + "grad_norm": 3.584996223449707, + "learning_rate": 8.525315537501813e-06, + "loss": 0.0579, + "step": 15919 + }, + { + "epoch": 1.7259323503902864, + "grad_norm": 1.3243951797485352, + "learning_rate": 8.524590163934427e-06, + "loss": 0.0264, + "step": 15920 + }, + { + "epoch": 1.726040763226366, + "grad_norm": 0.28816378116607666, + "learning_rate": 8.52386479036704e-06, + "loss": 0.0018, + "step": 15921 + }, + { + "epoch": 1.726149176062446, + "grad_norm": 1.3604557514190674, + "learning_rate": 8.523139416799653e-06, + "loss": 0.0095, + "step": 15922 + }, + { + "epoch": 1.7262575888985254, + "grad_norm": 3.4263672828674316, + "learning_rate": 8.522414043232265e-06, + "loss": 0.0328, + "step": 15923 + }, + { + "epoch": 1.7263660017346054, + "grad_norm": 0.21578750014305115, + "learning_rate": 8.521688669664879e-06, + "loss": 0.0025, + "step": 15924 + }, + { + "epoch": 1.7264744145706852, + "grad_norm": 5.37947940826416, + "learning_rate": 8.520963296097491e-06, + "loss": 0.0574, + "step": 15925 + }, + { + "epoch": 1.726582827406765, + "grad_norm": 1.402109146118164, + "learning_rate": 8.520237922530103e-06, + "loss": 0.0107, + "step": 15926 + }, + { + "epoch": 1.7266912402428447, + "grad_norm": 3.3280715942382812, + "learning_rate": 8.519512548962717e-06, + "loss": 0.0225, + "step": 15927 + }, + { + "epoch": 1.7267996530789245, + "grad_norm": 1.5200265645980835, + "learning_rate": 8.51878717539533e-06, + "loss": 0.0173, + "step": 15928 + }, + { + "epoch": 1.7269080659150045, + "grad_norm": 2.244985342025757, + "learning_rate": 8.518061801827943e-06, + "loss": 0.0176, + "step": 15929 + }, + { + "epoch": 1.727016478751084, + "grad_norm": 2.13071870803833, + "learning_rate": 8.517336428260555e-06, + "loss": 0.0412, + "step": 15930 + }, + { + "epoch": 1.727124891587164, + "grad_norm": 0.4892231523990631, + "learning_rate": 8.516611054693167e-06, + "loss": 0.0025, + "step": 15931 + }, + { + "epoch": 1.7272333044232437, + "grad_norm": 1.3663396835327148, + "learning_rate": 8.515885681125781e-06, + "loss": 0.0113, + "step": 15932 + }, + { + "epoch": 1.7273417172593235, + "grad_norm": 2.0852746963500977, + "learning_rate": 8.515160307558393e-06, + "loss": 0.0091, + "step": 15933 + }, + { + "epoch": 1.7274501300954033, + "grad_norm": 2.3188321590423584, + "learning_rate": 8.514434933991005e-06, + "loss": 0.009, + "step": 15934 + }, + { + "epoch": 1.727558542931483, + "grad_norm": 0.15118204057216644, + "learning_rate": 8.51370956042362e-06, + "loss": 0.0012, + "step": 15935 + }, + { + "epoch": 1.727666955767563, + "grad_norm": 23.683650970458984, + "learning_rate": 8.512984186856231e-06, + "loss": 0.0568, + "step": 15936 + }, + { + "epoch": 1.7277753686036426, + "grad_norm": 2.1018154621124268, + "learning_rate": 8.512258813288844e-06, + "loss": 0.0076, + "step": 15937 + }, + { + "epoch": 1.7278837814397225, + "grad_norm": 1.2896441221237183, + "learning_rate": 8.511533439721457e-06, + "loss": 0.0033, + "step": 15938 + }, + { + "epoch": 1.7279921942758023, + "grad_norm": 2.2477807998657227, + "learning_rate": 8.510808066154071e-06, + "loss": 0.0396, + "step": 15939 + }, + { + "epoch": 1.728100607111882, + "grad_norm": 3.0515835285186768, + "learning_rate": 8.510082692586683e-06, + "loss": 0.0535, + "step": 15940 + }, + { + "epoch": 1.7282090199479618, + "grad_norm": 0.3846580386161804, + "learning_rate": 8.509357319019295e-06, + "loss": 0.0031, + "step": 15941 + }, + { + "epoch": 1.7283174327840416, + "grad_norm": 4.147937297821045, + "learning_rate": 8.508631945451908e-06, + "loss": 0.0264, + "step": 15942 + }, + { + "epoch": 1.7284258456201216, + "grad_norm": 5.23117208480835, + "learning_rate": 8.507906571884521e-06, + "loss": 0.0222, + "step": 15943 + }, + { + "epoch": 1.7285342584562011, + "grad_norm": 0.24067935347557068, + "learning_rate": 8.507181198317134e-06, + "loss": 0.0021, + "step": 15944 + }, + { + "epoch": 1.728642671292281, + "grad_norm": 1.048008680343628, + "learning_rate": 8.506455824749746e-06, + "loss": 0.006, + "step": 15945 + }, + { + "epoch": 1.7287510841283606, + "grad_norm": 4.006453514099121, + "learning_rate": 8.50573045118236e-06, + "loss": 0.0239, + "step": 15946 + }, + { + "epoch": 1.7288594969644406, + "grad_norm": 1.4054628610610962, + "learning_rate": 8.505005077614973e-06, + "loss": 0.0199, + "step": 15947 + }, + { + "epoch": 1.7289679098005204, + "grad_norm": 0.9388858675956726, + "learning_rate": 8.504279704047586e-06, + "loss": 0.0154, + "step": 15948 + }, + { + "epoch": 1.7290763226366002, + "grad_norm": 0.42575159668922424, + "learning_rate": 8.503554330480198e-06, + "loss": 0.0039, + "step": 15949 + }, + { + "epoch": 1.72918473547268, + "grad_norm": 5.396135330200195, + "learning_rate": 8.502828956912812e-06, + "loss": 0.0531, + "step": 15950 + }, + { + "epoch": 1.7292931483087597, + "grad_norm": 1.6244549751281738, + "learning_rate": 8.502103583345424e-06, + "loss": 0.0065, + "step": 15951 + }, + { + "epoch": 1.7294015611448397, + "grad_norm": 1.5145790576934814, + "learning_rate": 8.501378209778036e-06, + "loss": 0.0094, + "step": 15952 + }, + { + "epoch": 1.7295099739809192, + "grad_norm": 3.108430862426758, + "learning_rate": 8.50065283621065e-06, + "loss": 0.0325, + "step": 15953 + }, + { + "epoch": 1.7296183868169992, + "grad_norm": 1.0960711240768433, + "learning_rate": 8.499927462643262e-06, + "loss": 0.0151, + "step": 15954 + }, + { + "epoch": 1.729726799653079, + "grad_norm": 17.48111343383789, + "learning_rate": 8.499202089075876e-06, + "loss": 0.0159, + "step": 15955 + }, + { + "epoch": 1.7298352124891587, + "grad_norm": 13.059762001037598, + "learning_rate": 8.498476715508488e-06, + "loss": 0.0419, + "step": 15956 + }, + { + "epoch": 1.7299436253252385, + "grad_norm": 10.718212127685547, + "learning_rate": 8.4977513419411e-06, + "loss": 0.0446, + "step": 15957 + }, + { + "epoch": 1.7300520381613183, + "grad_norm": 1.7527117729187012, + "learning_rate": 8.497025968373714e-06, + "loss": 0.0099, + "step": 15958 + }, + { + "epoch": 1.7301604509973982, + "grad_norm": 4.262899398803711, + "learning_rate": 8.496300594806326e-06, + "loss": 0.0378, + "step": 15959 + }, + { + "epoch": 1.7302688638334778, + "grad_norm": 3.408518075942993, + "learning_rate": 8.495575221238938e-06, + "loss": 0.0057, + "step": 15960 + }, + { + "epoch": 1.7303772766695578, + "grad_norm": 2.6513023376464844, + "learning_rate": 8.494849847671552e-06, + "loss": 0.0262, + "step": 15961 + }, + { + "epoch": 1.7304856895056373, + "grad_norm": 6.0464701652526855, + "learning_rate": 8.494124474104164e-06, + "loss": 0.0424, + "step": 15962 + }, + { + "epoch": 1.7305941023417173, + "grad_norm": 6.375553131103516, + "learning_rate": 8.493399100536776e-06, + "loss": 0.0838, + "step": 15963 + }, + { + "epoch": 1.730702515177797, + "grad_norm": 0.8924633860588074, + "learning_rate": 8.49267372696939e-06, + "loss": 0.0043, + "step": 15964 + }, + { + "epoch": 1.7308109280138768, + "grad_norm": 1.6296018362045288, + "learning_rate": 8.491948353402004e-06, + "loss": 0.0049, + "step": 15965 + }, + { + "epoch": 1.7309193408499568, + "grad_norm": 1.9438488483428955, + "learning_rate": 8.491222979834616e-06, + "loss": 0.0129, + "step": 15966 + }, + { + "epoch": 1.7310277536860363, + "grad_norm": 5.638125896453857, + "learning_rate": 8.490497606267228e-06, + "loss": 0.0545, + "step": 15967 + }, + { + "epoch": 1.7311361665221163, + "grad_norm": 5.834023952484131, + "learning_rate": 8.48977223269984e-06, + "loss": 0.0229, + "step": 15968 + }, + { + "epoch": 1.7312445793581959, + "grad_norm": 7.201264381408691, + "learning_rate": 8.489046859132454e-06, + "loss": 0.0477, + "step": 15969 + }, + { + "epoch": 1.7313529921942759, + "grad_norm": 4.634411334991455, + "learning_rate": 8.488321485565066e-06, + "loss": 0.0345, + "step": 15970 + }, + { + "epoch": 1.7314614050303556, + "grad_norm": 5.260780334472656, + "learning_rate": 8.487596111997678e-06, + "loss": 0.0337, + "step": 15971 + }, + { + "epoch": 1.7315698178664354, + "grad_norm": 1.0989910364151, + "learning_rate": 8.486870738430292e-06, + "loss": 0.0033, + "step": 15972 + }, + { + "epoch": 1.7316782307025151, + "grad_norm": 1.5028842687606812, + "learning_rate": 8.486145364862906e-06, + "loss": 0.0369, + "step": 15973 + }, + { + "epoch": 1.731786643538595, + "grad_norm": 6.76132345199585, + "learning_rate": 8.485419991295518e-06, + "loss": 0.0207, + "step": 15974 + }, + { + "epoch": 1.731895056374675, + "grad_norm": 1.9335540533065796, + "learning_rate": 8.48469461772813e-06, + "loss": 0.0121, + "step": 15975 + }, + { + "epoch": 1.7320034692107544, + "grad_norm": 0.7984476685523987, + "learning_rate": 8.483969244160744e-06, + "loss": 0.0054, + "step": 15976 + }, + { + "epoch": 1.7321118820468344, + "grad_norm": 1.8259941339492798, + "learning_rate": 8.483243870593356e-06, + "loss": 0.0202, + "step": 15977 + }, + { + "epoch": 1.7322202948829142, + "grad_norm": 0.23303769528865814, + "learning_rate": 8.482518497025969e-06, + "loss": 0.003, + "step": 15978 + }, + { + "epoch": 1.732328707718994, + "grad_norm": 1.1044617891311646, + "learning_rate": 8.481793123458582e-06, + "loss": 0.0167, + "step": 15979 + }, + { + "epoch": 1.7324371205550737, + "grad_norm": 2.2584080696105957, + "learning_rate": 8.481067749891194e-06, + "loss": 0.015, + "step": 15980 + }, + { + "epoch": 1.7325455333911535, + "grad_norm": 2.271329879760742, + "learning_rate": 8.480342376323808e-06, + "loss": 0.0263, + "step": 15981 + }, + { + "epoch": 1.7326539462272335, + "grad_norm": 1.2400858402252197, + "learning_rate": 8.47961700275642e-06, + "loss": 0.0086, + "step": 15982 + }, + { + "epoch": 1.732762359063313, + "grad_norm": 3.096947431564331, + "learning_rate": 8.478891629189033e-06, + "loss": 0.0657, + "step": 15983 + }, + { + "epoch": 1.732870771899393, + "grad_norm": 9.144967079162598, + "learning_rate": 8.478166255621646e-06, + "loss": 0.0413, + "step": 15984 + }, + { + "epoch": 1.7329791847354725, + "grad_norm": 1.8866615295410156, + "learning_rate": 8.477440882054259e-06, + "loss": 0.0078, + "step": 15985 + }, + { + "epoch": 1.7330875975715525, + "grad_norm": 5.847437858581543, + "learning_rate": 8.47671550848687e-06, + "loss": 0.0117, + "step": 15986 + }, + { + "epoch": 1.7331960104076323, + "grad_norm": 1.9369609355926514, + "learning_rate": 8.475990134919485e-06, + "loss": 0.0343, + "step": 15987 + }, + { + "epoch": 1.733304423243712, + "grad_norm": 1.1946827173233032, + "learning_rate": 8.475264761352097e-06, + "loss": 0.0037, + "step": 15988 + }, + { + "epoch": 1.7334128360797918, + "grad_norm": 3.799032211303711, + "learning_rate": 8.474539387784709e-06, + "loss": 0.0766, + "step": 15989 + }, + { + "epoch": 1.7335212489158716, + "grad_norm": 6.561068534851074, + "learning_rate": 8.473814014217323e-06, + "loss": 0.0279, + "step": 15990 + }, + { + "epoch": 1.7336296617519515, + "grad_norm": 0.6395896077156067, + "learning_rate": 8.473088640649937e-06, + "loss": 0.006, + "step": 15991 + }, + { + "epoch": 1.733738074588031, + "grad_norm": 1.6814181804656982, + "learning_rate": 8.472363267082549e-06, + "loss": 0.0069, + "step": 15992 + }, + { + "epoch": 1.733846487424111, + "grad_norm": 6.066575527191162, + "learning_rate": 8.47163789351516e-06, + "loss": 0.0282, + "step": 15993 + }, + { + "epoch": 1.7339549002601908, + "grad_norm": 16.219940185546875, + "learning_rate": 8.470912519947773e-06, + "loss": 0.0384, + "step": 15994 + }, + { + "epoch": 1.7340633130962706, + "grad_norm": 2.271876573562622, + "learning_rate": 8.470187146380387e-06, + "loss": 0.0252, + "step": 15995 + }, + { + "epoch": 1.7341717259323504, + "grad_norm": 2.38166880607605, + "learning_rate": 8.469461772812999e-06, + "loss": 0.0172, + "step": 15996 + }, + { + "epoch": 1.7342801387684301, + "grad_norm": 3.5298588275909424, + "learning_rate": 8.468736399245611e-06, + "loss": 0.0573, + "step": 15997 + }, + { + "epoch": 1.73438855160451, + "grad_norm": 9.99276065826416, + "learning_rate": 8.468011025678225e-06, + "loss": 0.0136, + "step": 15998 + }, + { + "epoch": 1.7344969644405897, + "grad_norm": 6.042013168334961, + "learning_rate": 8.467285652110839e-06, + "loss": 0.0188, + "step": 15999 + }, + { + "epoch": 1.7346053772766696, + "grad_norm": 2.238722085952759, + "learning_rate": 8.466560278543451e-06, + "loss": 0.0109, + "step": 16000 + }, + { + "epoch": 1.7347137901127494, + "grad_norm": 4.598476409912109, + "learning_rate": 8.465834904976063e-06, + "loss": 0.0277, + "step": 16001 + }, + { + "epoch": 1.7348222029488292, + "grad_norm": 0.9848078489303589, + "learning_rate": 8.465109531408677e-06, + "loss": 0.0052, + "step": 16002 + }, + { + "epoch": 1.734930615784909, + "grad_norm": 3.4099133014678955, + "learning_rate": 8.464384157841289e-06, + "loss": 0.0561, + "step": 16003 + }, + { + "epoch": 1.7350390286209887, + "grad_norm": 7.0311665534973145, + "learning_rate": 8.463658784273901e-06, + "loss": 0.0177, + "step": 16004 + }, + { + "epoch": 1.7351474414570687, + "grad_norm": 2.2224929332733154, + "learning_rate": 8.462933410706515e-06, + "loss": 0.012, + "step": 16005 + }, + { + "epoch": 1.7352558542931482, + "grad_norm": 1.863769292831421, + "learning_rate": 8.462208037139127e-06, + "loss": 0.0143, + "step": 16006 + }, + { + "epoch": 1.7353642671292282, + "grad_norm": 0.24821136891841888, + "learning_rate": 8.461482663571741e-06, + "loss": 0.0022, + "step": 16007 + }, + { + "epoch": 1.7354726799653077, + "grad_norm": 3.5627002716064453, + "learning_rate": 8.460757290004353e-06, + "loss": 0.0116, + "step": 16008 + }, + { + "epoch": 1.7355810928013877, + "grad_norm": 1.9036060571670532, + "learning_rate": 8.460031916436965e-06, + "loss": 0.0289, + "step": 16009 + }, + { + "epoch": 1.7356895056374675, + "grad_norm": 3.8423125743865967, + "learning_rate": 8.459306542869579e-06, + "loss": 0.0296, + "step": 16010 + }, + { + "epoch": 1.7357979184735473, + "grad_norm": 0.08527287095785141, + "learning_rate": 8.458581169302191e-06, + "loss": 0.0008, + "step": 16011 + }, + { + "epoch": 1.735906331309627, + "grad_norm": 2.1630616188049316, + "learning_rate": 8.457855795734803e-06, + "loss": 0.0244, + "step": 16012 + }, + { + "epoch": 1.7360147441457068, + "grad_norm": 1.877642035484314, + "learning_rate": 8.457130422167417e-06, + "loss": 0.0114, + "step": 16013 + }, + { + "epoch": 1.7361231569817868, + "grad_norm": 3.660677909851074, + "learning_rate": 8.45640504860003e-06, + "loss": 0.0278, + "step": 16014 + }, + { + "epoch": 1.7362315698178663, + "grad_norm": 6.0026350021362305, + "learning_rate": 8.455679675032642e-06, + "loss": 0.0299, + "step": 16015 + }, + { + "epoch": 1.7363399826539463, + "grad_norm": 1.184464931488037, + "learning_rate": 8.454954301465255e-06, + "loss": 0.0143, + "step": 16016 + }, + { + "epoch": 1.736448395490026, + "grad_norm": 0.3841865360736847, + "learning_rate": 8.45422892789787e-06, + "loss": 0.0051, + "step": 16017 + }, + { + "epoch": 1.7365568083261058, + "grad_norm": 3.0583791732788086, + "learning_rate": 8.453503554330481e-06, + "loss": 0.0231, + "step": 16018 + }, + { + "epoch": 1.7366652211621856, + "grad_norm": 2.6726362705230713, + "learning_rate": 8.452778180763093e-06, + "loss": 0.0177, + "step": 16019 + }, + { + "epoch": 1.7367736339982653, + "grad_norm": 4.683962821960449, + "learning_rate": 8.452052807195706e-06, + "loss": 0.0446, + "step": 16020 + }, + { + "epoch": 1.7368820468343453, + "grad_norm": 0.25114157795906067, + "learning_rate": 8.45132743362832e-06, + "loss": 0.0026, + "step": 16021 + }, + { + "epoch": 1.7369904596704249, + "grad_norm": 2.417672634124756, + "learning_rate": 8.450602060060932e-06, + "loss": 0.0191, + "step": 16022 + }, + { + "epoch": 1.7370988725065049, + "grad_norm": 1.8921592235565186, + "learning_rate": 8.449876686493544e-06, + "loss": 0.0154, + "step": 16023 + }, + { + "epoch": 1.7372072853425846, + "grad_norm": 2.1629061698913574, + "learning_rate": 8.449151312926158e-06, + "loss": 0.011, + "step": 16024 + }, + { + "epoch": 1.7373156981786644, + "grad_norm": 2.7286927700042725, + "learning_rate": 8.448425939358771e-06, + "loss": 0.0548, + "step": 16025 + }, + { + "epoch": 1.7374241110147441, + "grad_norm": 2.5565266609191895, + "learning_rate": 8.447700565791384e-06, + "loss": 0.0052, + "step": 16026 + }, + { + "epoch": 1.737532523850824, + "grad_norm": 4.38112735748291, + "learning_rate": 8.446975192223996e-06, + "loss": 0.016, + "step": 16027 + }, + { + "epoch": 1.737640936686904, + "grad_norm": 0.6048451662063599, + "learning_rate": 8.44624981865661e-06, + "loss": 0.0027, + "step": 16028 + }, + { + "epoch": 1.7377493495229834, + "grad_norm": 1.771344542503357, + "learning_rate": 8.445524445089222e-06, + "loss": 0.0239, + "step": 16029 + }, + { + "epoch": 1.7378577623590634, + "grad_norm": 0.45615825057029724, + "learning_rate": 8.444799071521834e-06, + "loss": 0.0052, + "step": 16030 + }, + { + "epoch": 1.737966175195143, + "grad_norm": 1.558951497077942, + "learning_rate": 8.444073697954448e-06, + "loss": 0.0214, + "step": 16031 + }, + { + "epoch": 1.738074588031223, + "grad_norm": 5.862049102783203, + "learning_rate": 8.44334832438706e-06, + "loss": 0.055, + "step": 16032 + }, + { + "epoch": 1.7381830008673027, + "grad_norm": 4.942276954650879, + "learning_rate": 8.442622950819674e-06, + "loss": 0.0431, + "step": 16033 + }, + { + "epoch": 1.7382914137033825, + "grad_norm": 2.9692840576171875, + "learning_rate": 8.441897577252286e-06, + "loss": 0.0157, + "step": 16034 + }, + { + "epoch": 1.7383998265394622, + "grad_norm": 1.957351565361023, + "learning_rate": 8.441172203684898e-06, + "loss": 0.0213, + "step": 16035 + }, + { + "epoch": 1.738508239375542, + "grad_norm": 3.052236795425415, + "learning_rate": 8.440446830117512e-06, + "loss": 0.0244, + "step": 16036 + }, + { + "epoch": 1.738616652211622, + "grad_norm": 2.9105095863342285, + "learning_rate": 8.439721456550124e-06, + "loss": 0.0477, + "step": 16037 + }, + { + "epoch": 1.7387250650477015, + "grad_norm": 4.637715816497803, + "learning_rate": 8.438996082982736e-06, + "loss": 0.0558, + "step": 16038 + }, + { + "epoch": 1.7388334778837815, + "grad_norm": 0.15584374964237213, + "learning_rate": 8.43827070941535e-06, + "loss": 0.0014, + "step": 16039 + }, + { + "epoch": 1.7389418907198613, + "grad_norm": 0.1657373607158661, + "learning_rate": 8.437545335847962e-06, + "loss": 0.0012, + "step": 16040 + }, + { + "epoch": 1.739050303555941, + "grad_norm": 0.8142979145050049, + "learning_rate": 8.436819962280574e-06, + "loss": 0.0038, + "step": 16041 + }, + { + "epoch": 1.7391587163920208, + "grad_norm": 2.1066300868988037, + "learning_rate": 8.436094588713188e-06, + "loss": 0.0201, + "step": 16042 + }, + { + "epoch": 1.7392671292281006, + "grad_norm": 3.3370165824890137, + "learning_rate": 8.435369215145802e-06, + "loss": 0.0382, + "step": 16043 + }, + { + "epoch": 1.7393755420641805, + "grad_norm": 4.16104793548584, + "learning_rate": 8.434643841578414e-06, + "loss": 0.0366, + "step": 16044 + }, + { + "epoch": 1.73948395490026, + "grad_norm": 0.1457209438085556, + "learning_rate": 8.433918468011026e-06, + "loss": 0.0011, + "step": 16045 + }, + { + "epoch": 1.73959236773634, + "grad_norm": 2.6039602756500244, + "learning_rate": 8.433193094443638e-06, + "loss": 0.0294, + "step": 16046 + }, + { + "epoch": 1.7397007805724196, + "grad_norm": 8.481547355651855, + "learning_rate": 8.432467720876252e-06, + "loss": 0.0564, + "step": 16047 + }, + { + "epoch": 1.7398091934084996, + "grad_norm": 2.8030076026916504, + "learning_rate": 8.431742347308864e-06, + "loss": 0.0236, + "step": 16048 + }, + { + "epoch": 1.7399176062445794, + "grad_norm": 1.9134355783462524, + "learning_rate": 8.431016973741476e-06, + "loss": 0.0441, + "step": 16049 + }, + { + "epoch": 1.7400260190806591, + "grad_norm": 0.692675769329071, + "learning_rate": 8.43029160017409e-06, + "loss": 0.004, + "step": 16050 + }, + { + "epoch": 1.7401344319167389, + "grad_norm": 2.40633487701416, + "learning_rate": 8.429566226606704e-06, + "loss": 0.0458, + "step": 16051 + }, + { + "epoch": 1.7402428447528187, + "grad_norm": 0.18956424295902252, + "learning_rate": 8.428840853039316e-06, + "loss": 0.0017, + "step": 16052 + }, + { + "epoch": 1.7403512575888986, + "grad_norm": 9.424399375915527, + "learning_rate": 8.428115479471928e-06, + "loss": 0.0382, + "step": 16053 + }, + { + "epoch": 1.7404596704249782, + "grad_norm": 2.1767218112945557, + "learning_rate": 8.427390105904542e-06, + "loss": 0.0046, + "step": 16054 + }, + { + "epoch": 1.7405680832610582, + "grad_norm": 3.807128667831421, + "learning_rate": 8.426664732337154e-06, + "loss": 0.0401, + "step": 16055 + }, + { + "epoch": 1.740676496097138, + "grad_norm": 1.7036343812942505, + "learning_rate": 8.425939358769767e-06, + "loss": 0.0283, + "step": 16056 + }, + { + "epoch": 1.7407849089332177, + "grad_norm": 2.598900079727173, + "learning_rate": 8.42521398520238e-06, + "loss": 0.0292, + "step": 16057 + }, + { + "epoch": 1.7408933217692975, + "grad_norm": 0.9559772610664368, + "learning_rate": 8.424488611634993e-06, + "loss": 0.0027, + "step": 16058 + }, + { + "epoch": 1.7410017346053772, + "grad_norm": 4.46519660949707, + "learning_rate": 8.423763238067606e-06, + "loss": 0.0257, + "step": 16059 + }, + { + "epoch": 1.7411101474414572, + "grad_norm": 1.3502533435821533, + "learning_rate": 8.423037864500218e-06, + "loss": 0.0146, + "step": 16060 + }, + { + "epoch": 1.7412185602775367, + "grad_norm": 4.070329189300537, + "learning_rate": 8.42231249093283e-06, + "loss": 0.0514, + "step": 16061 + }, + { + "epoch": 1.7413269731136167, + "grad_norm": 3.2468769550323486, + "learning_rate": 8.421587117365444e-06, + "loss": 0.008, + "step": 16062 + }, + { + "epoch": 1.7414353859496965, + "grad_norm": 5.180361270904541, + "learning_rate": 8.420861743798057e-06, + "loss": 0.0188, + "step": 16063 + }, + { + "epoch": 1.7415437987857763, + "grad_norm": 3.4304378032684326, + "learning_rate": 8.420136370230669e-06, + "loss": 0.0247, + "step": 16064 + }, + { + "epoch": 1.741652211621856, + "grad_norm": 4.2019524574279785, + "learning_rate": 8.419410996663283e-06, + "loss": 0.0217, + "step": 16065 + }, + { + "epoch": 1.7417606244579358, + "grad_norm": 1.5236316919326782, + "learning_rate": 8.418685623095895e-06, + "loss": 0.0237, + "step": 16066 + }, + { + "epoch": 1.7418690372940158, + "grad_norm": 0.7809454798698425, + "learning_rate": 8.417960249528507e-06, + "loss": 0.0085, + "step": 16067 + }, + { + "epoch": 1.7419774501300953, + "grad_norm": 8.75151252746582, + "learning_rate": 8.41723487596112e-06, + "loss": 0.0356, + "step": 16068 + }, + { + "epoch": 1.7420858629661753, + "grad_norm": 3.6803836822509766, + "learning_rate": 8.416509502393735e-06, + "loss": 0.0173, + "step": 16069 + }, + { + "epoch": 1.7421942758022548, + "grad_norm": 1.241593599319458, + "learning_rate": 8.415784128826347e-06, + "loss": 0.0185, + "step": 16070 + }, + { + "epoch": 1.7423026886383348, + "grad_norm": 1.9761356115341187, + "learning_rate": 8.415058755258959e-06, + "loss": 0.008, + "step": 16071 + }, + { + "epoch": 1.7424111014744146, + "grad_norm": 9.753046989440918, + "learning_rate": 8.414333381691571e-06, + "loss": 0.0371, + "step": 16072 + }, + { + "epoch": 1.7425195143104943, + "grad_norm": 1.6670880317687988, + "learning_rate": 8.413608008124185e-06, + "loss": 0.0295, + "step": 16073 + }, + { + "epoch": 1.742627927146574, + "grad_norm": 6.725226879119873, + "learning_rate": 8.412882634556797e-06, + "loss": 0.0343, + "step": 16074 + }, + { + "epoch": 1.7427363399826539, + "grad_norm": 0.04729966074228287, + "learning_rate": 8.412157260989409e-06, + "loss": 0.0006, + "step": 16075 + }, + { + "epoch": 1.7428447528187339, + "grad_norm": 1.8409193754196167, + "learning_rate": 8.411431887422023e-06, + "loss": 0.026, + "step": 16076 + }, + { + "epoch": 1.7429531656548134, + "grad_norm": 0.4571477770805359, + "learning_rate": 8.410706513854637e-06, + "loss": 0.0038, + "step": 16077 + }, + { + "epoch": 1.7430615784908934, + "grad_norm": 1.729422688484192, + "learning_rate": 8.409981140287249e-06, + "loss": 0.0261, + "step": 16078 + }, + { + "epoch": 1.7431699913269731, + "grad_norm": 0.13751736283302307, + "learning_rate": 8.409255766719861e-06, + "loss": 0.0008, + "step": 16079 + }, + { + "epoch": 1.743278404163053, + "grad_norm": 1.675392985343933, + "learning_rate": 8.408530393152475e-06, + "loss": 0.0103, + "step": 16080 + }, + { + "epoch": 1.7433868169991327, + "grad_norm": 2.2328853607177734, + "learning_rate": 8.407805019585087e-06, + "loss": 0.0141, + "step": 16081 + }, + { + "epoch": 1.7434952298352124, + "grad_norm": 4.4592719078063965, + "learning_rate": 8.4070796460177e-06, + "loss": 0.0589, + "step": 16082 + }, + { + "epoch": 1.7436036426712924, + "grad_norm": 0.5536140203475952, + "learning_rate": 8.406354272450313e-06, + "loss": 0.0019, + "step": 16083 + }, + { + "epoch": 1.743712055507372, + "grad_norm": 2.2792000770568848, + "learning_rate": 8.405628898882925e-06, + "loss": 0.0195, + "step": 16084 + }, + { + "epoch": 1.743820468343452, + "grad_norm": 3.0109198093414307, + "learning_rate": 8.404903525315539e-06, + "loss": 0.0268, + "step": 16085 + }, + { + "epoch": 1.7439288811795317, + "grad_norm": 1.2843379974365234, + "learning_rate": 8.404178151748151e-06, + "loss": 0.0057, + "step": 16086 + }, + { + "epoch": 1.7440372940156115, + "grad_norm": 3.668902635574341, + "learning_rate": 8.403452778180763e-06, + "loss": 0.0134, + "step": 16087 + }, + { + "epoch": 1.7441457068516912, + "grad_norm": 1.7373582124710083, + "learning_rate": 8.402727404613377e-06, + "loss": 0.0315, + "step": 16088 + }, + { + "epoch": 1.744254119687771, + "grad_norm": 0.9347332119941711, + "learning_rate": 8.40200203104599e-06, + "loss": 0.0101, + "step": 16089 + }, + { + "epoch": 1.744362532523851, + "grad_norm": 0.7670233845710754, + "learning_rate": 8.401276657478601e-06, + "loss": 0.011, + "step": 16090 + }, + { + "epoch": 1.7444709453599305, + "grad_norm": 15.049763679504395, + "learning_rate": 8.400551283911215e-06, + "loss": 0.0533, + "step": 16091 + }, + { + "epoch": 1.7445793581960105, + "grad_norm": 4.408330917358398, + "learning_rate": 8.399825910343827e-06, + "loss": 0.0479, + "step": 16092 + }, + { + "epoch": 1.74468777103209, + "grad_norm": 1.703208565711975, + "learning_rate": 8.399100536776441e-06, + "loss": 0.0241, + "step": 16093 + }, + { + "epoch": 1.74479618386817, + "grad_norm": 0.33225932717323303, + "learning_rate": 8.398375163209053e-06, + "loss": 0.0013, + "step": 16094 + }, + { + "epoch": 1.7449045967042498, + "grad_norm": 2.948955774307251, + "learning_rate": 8.397649789641667e-06, + "loss": 0.0176, + "step": 16095 + }, + { + "epoch": 1.7450130095403296, + "grad_norm": 3.37312388420105, + "learning_rate": 8.39692441607428e-06, + "loss": 0.0132, + "step": 16096 + }, + { + "epoch": 1.7451214223764093, + "grad_norm": 2.1678617000579834, + "learning_rate": 8.396199042506892e-06, + "loss": 0.0172, + "step": 16097 + }, + { + "epoch": 1.745229835212489, + "grad_norm": 2.942782163619995, + "learning_rate": 8.395473668939504e-06, + "loss": 0.0376, + "step": 16098 + }, + { + "epoch": 1.745338248048569, + "grad_norm": 1.6106277704238892, + "learning_rate": 8.394748295372117e-06, + "loss": 0.0224, + "step": 16099 + }, + { + "epoch": 1.7454466608846486, + "grad_norm": 2.022420883178711, + "learning_rate": 8.39402292180473e-06, + "loss": 0.0095, + "step": 16100 + }, + { + "epoch": 1.7455550737207286, + "grad_norm": 3.816358804702759, + "learning_rate": 8.393297548237342e-06, + "loss": 0.011, + "step": 16101 + }, + { + "epoch": 1.7456634865568084, + "grad_norm": 4.26327657699585, + "learning_rate": 8.392572174669956e-06, + "loss": 0.0668, + "step": 16102 + }, + { + "epoch": 1.7457718993928881, + "grad_norm": 2.5149850845336914, + "learning_rate": 8.39184680110257e-06, + "loss": 0.0657, + "step": 16103 + }, + { + "epoch": 1.745880312228968, + "grad_norm": 1.2840174436569214, + "learning_rate": 8.391121427535182e-06, + "loss": 0.0034, + "step": 16104 + }, + { + "epoch": 1.7459887250650477, + "grad_norm": 1.4200230836868286, + "learning_rate": 8.390396053967794e-06, + "loss": 0.0078, + "step": 16105 + }, + { + "epoch": 1.7460971379011276, + "grad_norm": 1.1636334657669067, + "learning_rate": 8.389670680400408e-06, + "loss": 0.0066, + "step": 16106 + }, + { + "epoch": 1.7462055507372072, + "grad_norm": 0.5160850882530212, + "learning_rate": 8.38894530683302e-06, + "loss": 0.0065, + "step": 16107 + }, + { + "epoch": 1.7463139635732872, + "grad_norm": 2.9804277420043945, + "learning_rate": 8.388219933265632e-06, + "loss": 0.0218, + "step": 16108 + }, + { + "epoch": 1.7464223764093667, + "grad_norm": 0.3724757432937622, + "learning_rate": 8.387494559698246e-06, + "loss": 0.0038, + "step": 16109 + }, + { + "epoch": 1.7465307892454467, + "grad_norm": 0.815132737159729, + "learning_rate": 8.386769186130858e-06, + "loss": 0.0056, + "step": 16110 + }, + { + "epoch": 1.7466392020815265, + "grad_norm": 0.4080808460712433, + "learning_rate": 8.386043812563472e-06, + "loss": 0.0063, + "step": 16111 + }, + { + "epoch": 1.7467476149176062, + "grad_norm": 4.022413730621338, + "learning_rate": 8.385318438996084e-06, + "loss": 0.0674, + "step": 16112 + }, + { + "epoch": 1.7468560277536862, + "grad_norm": 2.3237321376800537, + "learning_rate": 8.384593065428696e-06, + "loss": 0.0423, + "step": 16113 + }, + { + "epoch": 1.7469644405897657, + "grad_norm": 1.4436379671096802, + "learning_rate": 8.38386769186131e-06, + "loss": 0.0116, + "step": 16114 + }, + { + "epoch": 1.7470728534258457, + "grad_norm": 3.9274771213531494, + "learning_rate": 8.383142318293922e-06, + "loss": 0.0258, + "step": 16115 + }, + { + "epoch": 1.7471812662619253, + "grad_norm": 1.5951932668685913, + "learning_rate": 8.382416944726534e-06, + "loss": 0.0036, + "step": 16116 + }, + { + "epoch": 1.7472896790980053, + "grad_norm": 3.1582000255584717, + "learning_rate": 8.381691571159148e-06, + "loss": 0.0204, + "step": 16117 + }, + { + "epoch": 1.747398091934085, + "grad_norm": 8.522387504577637, + "learning_rate": 8.38096619759176e-06, + "loss": 0.0728, + "step": 16118 + }, + { + "epoch": 1.7475065047701648, + "grad_norm": 0.2574092745780945, + "learning_rate": 8.380240824024374e-06, + "loss": 0.0018, + "step": 16119 + }, + { + "epoch": 1.7476149176062445, + "grad_norm": 0.5491318106651306, + "learning_rate": 8.379515450456986e-06, + "loss": 0.0025, + "step": 16120 + }, + { + "epoch": 1.7477233304423243, + "grad_norm": 0.33726179599761963, + "learning_rate": 8.3787900768896e-06, + "loss": 0.0049, + "step": 16121 + }, + { + "epoch": 1.7478317432784043, + "grad_norm": 7.154021263122559, + "learning_rate": 8.378064703322212e-06, + "loss": 0.0149, + "step": 16122 + }, + { + "epoch": 1.7479401561144838, + "grad_norm": 0.8957324624061584, + "learning_rate": 8.377339329754824e-06, + "loss": 0.0062, + "step": 16123 + }, + { + "epoch": 1.7480485689505638, + "grad_norm": 1.917360544204712, + "learning_rate": 8.376613956187436e-06, + "loss": 0.0203, + "step": 16124 + }, + { + "epoch": 1.7481569817866436, + "grad_norm": 4.125696659088135, + "learning_rate": 8.37588858262005e-06, + "loss": 0.0267, + "step": 16125 + }, + { + "epoch": 1.7482653946227233, + "grad_norm": 2.662194013595581, + "learning_rate": 8.375163209052662e-06, + "loss": 0.0346, + "step": 16126 + }, + { + "epoch": 1.748373807458803, + "grad_norm": 8.093972206115723, + "learning_rate": 8.374437835485274e-06, + "loss": 0.0176, + "step": 16127 + }, + { + "epoch": 1.7484822202948829, + "grad_norm": 3.8238580226898193, + "learning_rate": 8.373712461917888e-06, + "loss": 0.0514, + "step": 16128 + }, + { + "epoch": 1.7485906331309629, + "grad_norm": 4.7925028800964355, + "learning_rate": 8.372987088350502e-06, + "loss": 0.0467, + "step": 16129 + }, + { + "epoch": 1.7486990459670424, + "grad_norm": 0.33747750520706177, + "learning_rate": 8.372261714783114e-06, + "loss": 0.0035, + "step": 16130 + }, + { + "epoch": 1.7488074588031224, + "grad_norm": 1.8178608417510986, + "learning_rate": 8.371536341215726e-06, + "loss": 0.0095, + "step": 16131 + }, + { + "epoch": 1.748915871639202, + "grad_norm": 1.8373874425888062, + "learning_rate": 8.37081096764834e-06, + "loss": 0.0282, + "step": 16132 + }, + { + "epoch": 1.749024284475282, + "grad_norm": 0.28315848112106323, + "learning_rate": 8.370085594080952e-06, + "loss": 0.0038, + "step": 16133 + }, + { + "epoch": 1.7491326973113617, + "grad_norm": 0.2488701492547989, + "learning_rate": 8.369360220513565e-06, + "loss": 0.0029, + "step": 16134 + }, + { + "epoch": 1.7492411101474414, + "grad_norm": 5.90143346786499, + "learning_rate": 8.368634846946178e-06, + "loss": 0.0744, + "step": 16135 + }, + { + "epoch": 1.7493495229835212, + "grad_norm": 0.29641610383987427, + "learning_rate": 8.36790947337879e-06, + "loss": 0.0041, + "step": 16136 + }, + { + "epoch": 1.749457935819601, + "grad_norm": 1.939063549041748, + "learning_rate": 8.367184099811404e-06, + "loss": 0.0484, + "step": 16137 + }, + { + "epoch": 1.749566348655681, + "grad_norm": 0.36329856514930725, + "learning_rate": 8.366458726244016e-06, + "loss": 0.0055, + "step": 16138 + }, + { + "epoch": 1.7496747614917605, + "grad_norm": 0.4608021378517151, + "learning_rate": 8.365733352676629e-06, + "loss": 0.0041, + "step": 16139 + }, + { + "epoch": 1.7497831743278405, + "grad_norm": 0.3601575791835785, + "learning_rate": 8.365007979109242e-06, + "loss": 0.0042, + "step": 16140 + }, + { + "epoch": 1.7498915871639202, + "grad_norm": 2.140667200088501, + "learning_rate": 8.364282605541855e-06, + "loss": 0.0142, + "step": 16141 + }, + { + "epoch": 1.75, + "grad_norm": 2.906708002090454, + "learning_rate": 8.363557231974467e-06, + "loss": 0.0255, + "step": 16142 + }, + { + "epoch": 1.7501084128360798, + "grad_norm": 1.884022831916809, + "learning_rate": 8.36283185840708e-06, + "loss": 0.0169, + "step": 16143 + }, + { + "epoch": 1.7502168256721595, + "grad_norm": 2.510925769805908, + "learning_rate": 8.362106484839693e-06, + "loss": 0.0423, + "step": 16144 + }, + { + "epoch": 1.7503252385082395, + "grad_norm": 3.770388126373291, + "learning_rate": 8.361381111272307e-06, + "loss": 0.0572, + "step": 16145 + }, + { + "epoch": 1.750433651344319, + "grad_norm": 11.913908958435059, + "learning_rate": 8.360655737704919e-06, + "loss": 0.0528, + "step": 16146 + }, + { + "epoch": 1.750542064180399, + "grad_norm": 3.4505085945129395, + "learning_rate": 8.359930364137533e-06, + "loss": 0.0409, + "step": 16147 + }, + { + "epoch": 1.7506504770164788, + "grad_norm": 3.351067304611206, + "learning_rate": 8.359204990570145e-06, + "loss": 0.0209, + "step": 16148 + }, + { + "epoch": 1.7507588898525586, + "grad_norm": 2.5890424251556396, + "learning_rate": 8.358479617002757e-06, + "loss": 0.0379, + "step": 16149 + }, + { + "epoch": 1.7508673026886383, + "grad_norm": 2.0671074390411377, + "learning_rate": 8.357754243435369e-06, + "loss": 0.0318, + "step": 16150 + }, + { + "epoch": 1.750975715524718, + "grad_norm": 2.049940586090088, + "learning_rate": 8.357028869867983e-06, + "loss": 0.0073, + "step": 16151 + }, + { + "epoch": 1.751084128360798, + "grad_norm": 0.979001522064209, + "learning_rate": 8.356303496300595e-06, + "loss": 0.0441, + "step": 16152 + }, + { + "epoch": 1.7511925411968776, + "grad_norm": 0.15953023731708527, + "learning_rate": 8.355578122733207e-06, + "loss": 0.0015, + "step": 16153 + }, + { + "epoch": 1.7513009540329576, + "grad_norm": 0.6959146857261658, + "learning_rate": 8.354852749165821e-06, + "loss": 0.0049, + "step": 16154 + }, + { + "epoch": 1.7514093668690371, + "grad_norm": 2.36146879196167, + "learning_rate": 8.354127375598435e-06, + "loss": 0.0304, + "step": 16155 + }, + { + "epoch": 1.7515177797051171, + "grad_norm": 2.2746329307556152, + "learning_rate": 8.353402002031047e-06, + "loss": 0.055, + "step": 16156 + }, + { + "epoch": 1.751626192541197, + "grad_norm": 0.816102921962738, + "learning_rate": 8.352676628463659e-06, + "loss": 0.0247, + "step": 16157 + }, + { + "epoch": 1.7517346053772767, + "grad_norm": 0.9201303124427795, + "learning_rate": 8.351951254896273e-06, + "loss": 0.0026, + "step": 16158 + }, + { + "epoch": 1.7518430182133564, + "grad_norm": 1.4845672845840454, + "learning_rate": 8.351225881328885e-06, + "loss": 0.0333, + "step": 16159 + }, + { + "epoch": 1.7519514310494362, + "grad_norm": 0.616334855556488, + "learning_rate": 8.350500507761497e-06, + "loss": 0.0034, + "step": 16160 + }, + { + "epoch": 1.7520598438855162, + "grad_norm": 3.2095606327056885, + "learning_rate": 8.349775134194111e-06, + "loss": 0.0164, + "step": 16161 + }, + { + "epoch": 1.7521682567215957, + "grad_norm": 1.688553810119629, + "learning_rate": 8.349049760626723e-06, + "loss": 0.0098, + "step": 16162 + }, + { + "epoch": 1.7522766695576757, + "grad_norm": 1.1858688592910767, + "learning_rate": 8.348324387059337e-06, + "loss": 0.0133, + "step": 16163 + }, + { + "epoch": 1.7523850823937555, + "grad_norm": 2.324489116668701, + "learning_rate": 8.34759901349195e-06, + "loss": 0.0338, + "step": 16164 + }, + { + "epoch": 1.7524934952298352, + "grad_norm": 1.6901997327804565, + "learning_rate": 8.346873639924561e-06, + "loss": 0.0292, + "step": 16165 + }, + { + "epoch": 1.752601908065915, + "grad_norm": 0.37171265482902527, + "learning_rate": 8.346148266357175e-06, + "loss": 0.0037, + "step": 16166 + }, + { + "epoch": 1.7527103209019947, + "grad_norm": 2.9006552696228027, + "learning_rate": 8.345422892789787e-06, + "loss": 0.0141, + "step": 16167 + }, + { + "epoch": 1.7528187337380747, + "grad_norm": 3.3415181636810303, + "learning_rate": 8.3446975192224e-06, + "loss": 0.0334, + "step": 16168 + }, + { + "epoch": 1.7529271465741543, + "grad_norm": 2.6140246391296387, + "learning_rate": 8.343972145655013e-06, + "loss": 0.0484, + "step": 16169 + }, + { + "epoch": 1.7530355594102343, + "grad_norm": 2.92282772064209, + "learning_rate": 8.343246772087625e-06, + "loss": 0.0493, + "step": 16170 + }, + { + "epoch": 1.7531439722463138, + "grad_norm": 7.109988212585449, + "learning_rate": 8.34252139852024e-06, + "loss": 0.0288, + "step": 16171 + }, + { + "epoch": 1.7532523850823938, + "grad_norm": 0.5568246841430664, + "learning_rate": 8.341796024952851e-06, + "loss": 0.0032, + "step": 16172 + }, + { + "epoch": 1.7533607979184735, + "grad_norm": 1.6367096900939941, + "learning_rate": 8.341070651385465e-06, + "loss": 0.0265, + "step": 16173 + }, + { + "epoch": 1.7534692107545533, + "grad_norm": 2.156898260116577, + "learning_rate": 8.340345277818077e-06, + "loss": 0.0198, + "step": 16174 + }, + { + "epoch": 1.7535776235906333, + "grad_norm": 0.8797104954719543, + "learning_rate": 8.33961990425069e-06, + "loss": 0.0156, + "step": 16175 + }, + { + "epoch": 1.7536860364267128, + "grad_norm": 3.8718717098236084, + "learning_rate": 8.338894530683302e-06, + "loss": 0.0341, + "step": 16176 + }, + { + "epoch": 1.7537944492627928, + "grad_norm": 3.1263740062713623, + "learning_rate": 8.338169157115915e-06, + "loss": 0.0259, + "step": 16177 + }, + { + "epoch": 1.7539028620988724, + "grad_norm": 0.1687437742948532, + "learning_rate": 8.337443783548528e-06, + "loss": 0.0012, + "step": 16178 + }, + { + "epoch": 1.7540112749349523, + "grad_norm": 0.4564216434955597, + "learning_rate": 8.33671840998114e-06, + "loss": 0.0029, + "step": 16179 + }, + { + "epoch": 1.754119687771032, + "grad_norm": 3.358074903488159, + "learning_rate": 8.335993036413754e-06, + "loss": 0.0321, + "step": 16180 + }, + { + "epoch": 1.7542281006071119, + "grad_norm": 0.3463810682296753, + "learning_rate": 8.335267662846367e-06, + "loss": 0.0032, + "step": 16181 + }, + { + "epoch": 1.7543365134431916, + "grad_norm": 0.9631817936897278, + "learning_rate": 8.33454228927898e-06, + "loss": 0.0118, + "step": 16182 + }, + { + "epoch": 1.7544449262792714, + "grad_norm": 0.8480017185211182, + "learning_rate": 8.333816915711592e-06, + "loss": 0.0065, + "step": 16183 + }, + { + "epoch": 1.7545533391153514, + "grad_norm": 0.6428835988044739, + "learning_rate": 8.333091542144206e-06, + "loss": 0.0085, + "step": 16184 + }, + { + "epoch": 1.754661751951431, + "grad_norm": 0.9466139674186707, + "learning_rate": 8.332366168576818e-06, + "loss": 0.0061, + "step": 16185 + }, + { + "epoch": 1.754770164787511, + "grad_norm": 9.881875038146973, + "learning_rate": 8.33164079500943e-06, + "loss": 0.0376, + "step": 16186 + }, + { + "epoch": 1.7548785776235907, + "grad_norm": 1.0344595909118652, + "learning_rate": 8.330915421442044e-06, + "loss": 0.007, + "step": 16187 + }, + { + "epoch": 1.7549869904596704, + "grad_norm": 0.20987163484096527, + "learning_rate": 8.330190047874656e-06, + "loss": 0.0017, + "step": 16188 + }, + { + "epoch": 1.7550954032957502, + "grad_norm": 3.2709548473358154, + "learning_rate": 8.32946467430727e-06, + "loss": 0.0493, + "step": 16189 + }, + { + "epoch": 1.75520381613183, + "grad_norm": 7.0550761222839355, + "learning_rate": 8.328739300739882e-06, + "loss": 0.0305, + "step": 16190 + }, + { + "epoch": 1.75531222896791, + "grad_norm": 1.0863468647003174, + "learning_rate": 8.328013927172494e-06, + "loss": 0.0285, + "step": 16191 + }, + { + "epoch": 1.7554206418039895, + "grad_norm": 4.492012977600098, + "learning_rate": 8.327288553605108e-06, + "loss": 0.0513, + "step": 16192 + }, + { + "epoch": 1.7555290546400695, + "grad_norm": 1.0431870222091675, + "learning_rate": 8.32656318003772e-06, + "loss": 0.0111, + "step": 16193 + }, + { + "epoch": 1.755637467476149, + "grad_norm": 3.7664778232574463, + "learning_rate": 8.325837806470332e-06, + "loss": 0.0464, + "step": 16194 + }, + { + "epoch": 1.755745880312229, + "grad_norm": 5.805624485015869, + "learning_rate": 8.325112432902946e-06, + "loss": 0.0234, + "step": 16195 + }, + { + "epoch": 1.7558542931483088, + "grad_norm": 2.0435245037078857, + "learning_rate": 8.324387059335558e-06, + "loss": 0.0262, + "step": 16196 + }, + { + "epoch": 1.7559627059843885, + "grad_norm": 3.8028712272644043, + "learning_rate": 8.323661685768172e-06, + "loss": 0.0197, + "step": 16197 + }, + { + "epoch": 1.7560711188204683, + "grad_norm": 6.307690620422363, + "learning_rate": 8.322936312200784e-06, + "loss": 0.0359, + "step": 16198 + }, + { + "epoch": 1.756179531656548, + "grad_norm": 2.221205711364746, + "learning_rate": 8.322210938633398e-06, + "loss": 0.0635, + "step": 16199 + }, + { + "epoch": 1.756287944492628, + "grad_norm": 1.540716528892517, + "learning_rate": 8.32148556506601e-06, + "loss": 0.014, + "step": 16200 + }, + { + "epoch": 1.7563963573287076, + "grad_norm": 2.6175272464752197, + "learning_rate": 8.320760191498622e-06, + "loss": 0.034, + "step": 16201 + }, + { + "epoch": 1.7565047701647876, + "grad_norm": 0.564348578453064, + "learning_rate": 8.320034817931234e-06, + "loss": 0.0036, + "step": 16202 + }, + { + "epoch": 1.7566131830008673, + "grad_norm": 3.0871901512145996, + "learning_rate": 8.319309444363848e-06, + "loss": 0.0386, + "step": 16203 + }, + { + "epoch": 1.756721595836947, + "grad_norm": 1.4536854028701782, + "learning_rate": 8.31858407079646e-06, + "loss": 0.0147, + "step": 16204 + }, + { + "epoch": 1.7568300086730269, + "grad_norm": 3.3419723510742188, + "learning_rate": 8.317858697229072e-06, + "loss": 0.0875, + "step": 16205 + }, + { + "epoch": 1.7569384215091066, + "grad_norm": 1.1709489822387695, + "learning_rate": 8.317133323661686e-06, + "loss": 0.0171, + "step": 16206 + }, + { + "epoch": 1.7570468343451866, + "grad_norm": 2.922110080718994, + "learning_rate": 8.3164079500943e-06, + "loss": 0.0726, + "step": 16207 + }, + { + "epoch": 1.7571552471812661, + "grad_norm": 5.972740650177002, + "learning_rate": 8.315682576526912e-06, + "loss": 0.0413, + "step": 16208 + }, + { + "epoch": 1.7572636600173461, + "grad_norm": 0.7148534655570984, + "learning_rate": 8.314957202959524e-06, + "loss": 0.0053, + "step": 16209 + }, + { + "epoch": 1.757372072853426, + "grad_norm": 2.819122076034546, + "learning_rate": 8.314231829392138e-06, + "loss": 0.0162, + "step": 16210 + }, + { + "epoch": 1.7574804856895057, + "grad_norm": 3.677607297897339, + "learning_rate": 8.31350645582475e-06, + "loss": 0.0183, + "step": 16211 + }, + { + "epoch": 1.7575888985255854, + "grad_norm": 6.762777805328369, + "learning_rate": 8.312781082257363e-06, + "loss": 0.0526, + "step": 16212 + }, + { + "epoch": 1.7576973113616652, + "grad_norm": 5.249630451202393, + "learning_rate": 8.312055708689976e-06, + "loss": 0.0257, + "step": 16213 + }, + { + "epoch": 1.7578057241977452, + "grad_norm": 4.1381659507751465, + "learning_rate": 8.311330335122589e-06, + "loss": 0.0107, + "step": 16214 + }, + { + "epoch": 1.7579141370338247, + "grad_norm": 1.990561604499817, + "learning_rate": 8.310604961555202e-06, + "loss": 0.0234, + "step": 16215 + }, + { + "epoch": 1.7580225498699047, + "grad_norm": 1.2425320148468018, + "learning_rate": 8.309879587987814e-06, + "loss": 0.0483, + "step": 16216 + }, + { + "epoch": 1.7581309627059842, + "grad_norm": 3.5929880142211914, + "learning_rate": 8.309154214420427e-06, + "loss": 0.0698, + "step": 16217 + }, + { + "epoch": 1.7582393755420642, + "grad_norm": 5.243453025817871, + "learning_rate": 8.30842884085304e-06, + "loss": 0.0141, + "step": 16218 + }, + { + "epoch": 1.758347788378144, + "grad_norm": 2.04848575592041, + "learning_rate": 8.307703467285653e-06, + "loss": 0.0572, + "step": 16219 + }, + { + "epoch": 1.7584562012142237, + "grad_norm": 0.8378921151161194, + "learning_rate": 8.306978093718265e-06, + "loss": 0.0057, + "step": 16220 + }, + { + "epoch": 1.7585646140503035, + "grad_norm": 4.131412982940674, + "learning_rate": 8.306252720150879e-06, + "loss": 0.0245, + "step": 16221 + }, + { + "epoch": 1.7586730268863833, + "grad_norm": 3.4201865196228027, + "learning_rate": 8.30552734658349e-06, + "loss": 0.0078, + "step": 16222 + }, + { + "epoch": 1.7587814397224633, + "grad_norm": 8.520940780639648, + "learning_rate": 8.304801973016105e-06, + "loss": 0.0174, + "step": 16223 + }, + { + "epoch": 1.7588898525585428, + "grad_norm": 1.9624274969100952, + "learning_rate": 8.304076599448717e-06, + "loss": 0.0205, + "step": 16224 + }, + { + "epoch": 1.7589982653946228, + "grad_norm": 0.7274329662322998, + "learning_rate": 8.30335122588133e-06, + "loss": 0.0065, + "step": 16225 + }, + { + "epoch": 1.7591066782307025, + "grad_norm": 1.7048546075820923, + "learning_rate": 8.302625852313943e-06, + "loss": 0.0347, + "step": 16226 + }, + { + "epoch": 1.7592150910667823, + "grad_norm": 1.5593289136886597, + "learning_rate": 8.301900478746555e-06, + "loss": 0.0111, + "step": 16227 + }, + { + "epoch": 1.759323503902862, + "grad_norm": 5.544887542724609, + "learning_rate": 8.301175105179167e-06, + "loss": 0.0354, + "step": 16228 + }, + { + "epoch": 1.7594319167389418, + "grad_norm": 5.056313991546631, + "learning_rate": 8.30044973161178e-06, + "loss": 0.0385, + "step": 16229 + }, + { + "epoch": 1.7595403295750218, + "grad_norm": 1.1559289693832397, + "learning_rate": 8.299724358044393e-06, + "loss": 0.016, + "step": 16230 + }, + { + "epoch": 1.7596487424111014, + "grad_norm": 0.8249550461769104, + "learning_rate": 8.298998984477005e-06, + "loss": 0.0074, + "step": 16231 + }, + { + "epoch": 1.7597571552471813, + "grad_norm": 0.7087788581848145, + "learning_rate": 8.298273610909619e-06, + "loss": 0.0042, + "step": 16232 + }, + { + "epoch": 1.7598655680832611, + "grad_norm": 1.3401097059249878, + "learning_rate": 8.297548237342233e-06, + "loss": 0.013, + "step": 16233 + }, + { + "epoch": 1.7599739809193409, + "grad_norm": 1.5592759847640991, + "learning_rate": 8.296822863774845e-06, + "loss": 0.0393, + "step": 16234 + }, + { + "epoch": 1.7600823937554206, + "grad_norm": 2.5983731746673584, + "learning_rate": 8.296097490207457e-06, + "loss": 0.0534, + "step": 16235 + }, + { + "epoch": 1.7601908065915004, + "grad_norm": 0.6027705073356628, + "learning_rate": 8.295372116640071e-06, + "loss": 0.0052, + "step": 16236 + }, + { + "epoch": 1.7602992194275804, + "grad_norm": 6.158027648925781, + "learning_rate": 8.294646743072683e-06, + "loss": 0.0466, + "step": 16237 + }, + { + "epoch": 1.76040763226366, + "grad_norm": 1.4308844804763794, + "learning_rate": 8.293921369505295e-06, + "loss": 0.0035, + "step": 16238 + }, + { + "epoch": 1.76051604509974, + "grad_norm": 1.1409904956817627, + "learning_rate": 8.293195995937909e-06, + "loss": 0.0078, + "step": 16239 + }, + { + "epoch": 1.7606244579358195, + "grad_norm": 0.38175904750823975, + "learning_rate": 8.292470622370521e-06, + "loss": 0.0044, + "step": 16240 + }, + { + "epoch": 1.7607328707718994, + "grad_norm": 1.8244476318359375, + "learning_rate": 8.291745248803135e-06, + "loss": 0.007, + "step": 16241 + }, + { + "epoch": 1.7608412836079792, + "grad_norm": 0.9802746176719666, + "learning_rate": 8.291019875235747e-06, + "loss": 0.0053, + "step": 16242 + }, + { + "epoch": 1.760949696444059, + "grad_norm": 2.7485713958740234, + "learning_rate": 8.29029450166836e-06, + "loss": 0.0272, + "step": 16243 + }, + { + "epoch": 1.7610581092801387, + "grad_norm": 0.6649872064590454, + "learning_rate": 8.289569128100973e-06, + "loss": 0.0043, + "step": 16244 + }, + { + "epoch": 1.7611665221162185, + "grad_norm": 0.3332729637622833, + "learning_rate": 8.288843754533585e-06, + "loss": 0.0023, + "step": 16245 + }, + { + "epoch": 1.7612749349522985, + "grad_norm": 0.38684800267219543, + "learning_rate": 8.288118380966197e-06, + "loss": 0.0023, + "step": 16246 + }, + { + "epoch": 1.761383347788378, + "grad_norm": 4.337197303771973, + "learning_rate": 8.287393007398811e-06, + "loss": 0.0257, + "step": 16247 + }, + { + "epoch": 1.761491760624458, + "grad_norm": 0.5602723956108093, + "learning_rate": 8.286667633831423e-06, + "loss": 0.0025, + "step": 16248 + }, + { + "epoch": 1.7616001734605378, + "grad_norm": 4.049293518066406, + "learning_rate": 8.285942260264037e-06, + "loss": 0.0079, + "step": 16249 + }, + { + "epoch": 1.7617085862966175, + "grad_norm": 1.6069616079330444, + "learning_rate": 8.28521688669665e-06, + "loss": 0.0316, + "step": 16250 + }, + { + "epoch": 1.7618169991326973, + "grad_norm": 1.3923043012619019, + "learning_rate": 8.284491513129263e-06, + "loss": 0.019, + "step": 16251 + }, + { + "epoch": 1.761925411968777, + "grad_norm": 1.1490485668182373, + "learning_rate": 8.283766139561875e-06, + "loss": 0.0417, + "step": 16252 + }, + { + "epoch": 1.762033824804857, + "grad_norm": 1.110803484916687, + "learning_rate": 8.283040765994488e-06, + "loss": 0.0055, + "step": 16253 + }, + { + "epoch": 1.7621422376409366, + "grad_norm": 1.7360731363296509, + "learning_rate": 8.2823153924271e-06, + "loss": 0.0281, + "step": 16254 + }, + { + "epoch": 1.7622506504770166, + "grad_norm": 1.4626282453536987, + "learning_rate": 8.281590018859713e-06, + "loss": 0.0578, + "step": 16255 + }, + { + "epoch": 1.762359063313096, + "grad_norm": 1.405661702156067, + "learning_rate": 8.280864645292326e-06, + "loss": 0.0615, + "step": 16256 + }, + { + "epoch": 1.762467476149176, + "grad_norm": 0.8931065797805786, + "learning_rate": 8.280139271724938e-06, + "loss": 0.0082, + "step": 16257 + }, + { + "epoch": 1.7625758889852559, + "grad_norm": 0.3299510180950165, + "learning_rate": 8.279413898157552e-06, + "loss": 0.0016, + "step": 16258 + }, + { + "epoch": 1.7626843018213356, + "grad_norm": 6.957776069641113, + "learning_rate": 8.278688524590165e-06, + "loss": 0.0433, + "step": 16259 + }, + { + "epoch": 1.7627927146574154, + "grad_norm": 1.3261775970458984, + "learning_rate": 8.277963151022778e-06, + "loss": 0.0301, + "step": 16260 + }, + { + "epoch": 1.7629011274934951, + "grad_norm": 1.5971221923828125, + "learning_rate": 8.27723777745539e-06, + "loss": 0.0374, + "step": 16261 + }, + { + "epoch": 1.7630095403295751, + "grad_norm": 3.9153892993927, + "learning_rate": 8.276512403888004e-06, + "loss": 0.042, + "step": 16262 + }, + { + "epoch": 1.7631179531656547, + "grad_norm": 4.321839332580566, + "learning_rate": 8.275787030320616e-06, + "loss": 0.0568, + "step": 16263 + }, + { + "epoch": 1.7632263660017347, + "grad_norm": 15.037018775939941, + "learning_rate": 8.275061656753228e-06, + "loss": 0.0449, + "step": 16264 + }, + { + "epoch": 1.7633347788378144, + "grad_norm": 2.1869659423828125, + "learning_rate": 8.274336283185842e-06, + "loss": 0.016, + "step": 16265 + }, + { + "epoch": 1.7634431916738942, + "grad_norm": 0.24572254717350006, + "learning_rate": 8.273610909618456e-06, + "loss": 0.0028, + "step": 16266 + }, + { + "epoch": 1.763551604509974, + "grad_norm": 3.8881192207336426, + "learning_rate": 8.272885536051068e-06, + "loss": 0.0312, + "step": 16267 + }, + { + "epoch": 1.7636600173460537, + "grad_norm": 2.439685106277466, + "learning_rate": 8.27216016248368e-06, + "loss": 0.0088, + "step": 16268 + }, + { + "epoch": 1.7637684301821337, + "grad_norm": 0.362112432718277, + "learning_rate": 8.271434788916292e-06, + "loss": 0.0036, + "step": 16269 + }, + { + "epoch": 1.7638768430182132, + "grad_norm": 5.179840087890625, + "learning_rate": 8.270709415348906e-06, + "loss": 0.0809, + "step": 16270 + }, + { + "epoch": 1.7639852558542932, + "grad_norm": 1.371048927307129, + "learning_rate": 8.269984041781518e-06, + "loss": 0.0218, + "step": 16271 + }, + { + "epoch": 1.764093668690373, + "grad_norm": 1.0915591716766357, + "learning_rate": 8.26925866821413e-06, + "loss": 0.0171, + "step": 16272 + }, + { + "epoch": 1.7642020815264527, + "grad_norm": 1.596052885055542, + "learning_rate": 8.268533294646744e-06, + "loss": 0.0218, + "step": 16273 + }, + { + "epoch": 1.7643104943625325, + "grad_norm": 2.8010194301605225, + "learning_rate": 8.267807921079356e-06, + "loss": 0.0262, + "step": 16274 + }, + { + "epoch": 1.7644189071986123, + "grad_norm": 2.7395870685577393, + "learning_rate": 8.26708254751197e-06, + "loss": 0.0494, + "step": 16275 + }, + { + "epoch": 1.7645273200346923, + "grad_norm": 4.756886959075928, + "learning_rate": 8.266357173944582e-06, + "loss": 0.0336, + "step": 16276 + }, + { + "epoch": 1.7646357328707718, + "grad_norm": 0.476998895406723, + "learning_rate": 8.265631800377196e-06, + "loss": 0.0051, + "step": 16277 + }, + { + "epoch": 1.7647441457068518, + "grad_norm": 12.69876480102539, + "learning_rate": 8.264906426809808e-06, + "loss": 0.0419, + "step": 16278 + }, + { + "epoch": 1.7648525585429313, + "grad_norm": 1.273852825164795, + "learning_rate": 8.26418105324242e-06, + "loss": 0.0222, + "step": 16279 + }, + { + "epoch": 1.7649609713790113, + "grad_norm": 3.9739038944244385, + "learning_rate": 8.263455679675032e-06, + "loss": 0.0204, + "step": 16280 + }, + { + "epoch": 1.765069384215091, + "grad_norm": 0.29295095801353455, + "learning_rate": 8.262730306107646e-06, + "loss": 0.0021, + "step": 16281 + }, + { + "epoch": 1.7651777970511708, + "grad_norm": 4.321127414703369, + "learning_rate": 8.262004932540258e-06, + "loss": 0.0454, + "step": 16282 + }, + { + "epoch": 1.7652862098872506, + "grad_norm": 3.7434301376342773, + "learning_rate": 8.26127955897287e-06, + "loss": 0.0178, + "step": 16283 + }, + { + "epoch": 1.7653946227233304, + "grad_norm": 5.901309490203857, + "learning_rate": 8.260554185405484e-06, + "loss": 0.0143, + "step": 16284 + }, + { + "epoch": 1.7655030355594103, + "grad_norm": 6.852996349334717, + "learning_rate": 8.259828811838098e-06, + "loss": 0.0236, + "step": 16285 + }, + { + "epoch": 1.76561144839549, + "grad_norm": 0.7441725730895996, + "learning_rate": 8.25910343827071e-06, + "loss": 0.0049, + "step": 16286 + }, + { + "epoch": 1.7657198612315699, + "grad_norm": 4.724677562713623, + "learning_rate": 8.258378064703322e-06, + "loss": 0.0507, + "step": 16287 + }, + { + "epoch": 1.7658282740676496, + "grad_norm": 1.2107023000717163, + "learning_rate": 8.257652691135936e-06, + "loss": 0.0084, + "step": 16288 + }, + { + "epoch": 1.7659366869037294, + "grad_norm": 5.2019147872924805, + "learning_rate": 8.256927317568548e-06, + "loss": 0.0148, + "step": 16289 + }, + { + "epoch": 1.7660450997398092, + "grad_norm": 3.538059949874878, + "learning_rate": 8.25620194400116e-06, + "loss": 0.0351, + "step": 16290 + }, + { + "epoch": 1.766153512575889, + "grad_norm": 2.9817771911621094, + "learning_rate": 8.255476570433774e-06, + "loss": 0.0418, + "step": 16291 + }, + { + "epoch": 1.766261925411969, + "grad_norm": 3.1075894832611084, + "learning_rate": 8.254751196866388e-06, + "loss": 0.0279, + "step": 16292 + }, + { + "epoch": 1.7663703382480485, + "grad_norm": 9.126615524291992, + "learning_rate": 8.254025823299e-06, + "loss": 0.028, + "step": 16293 + }, + { + "epoch": 1.7664787510841284, + "grad_norm": 1.0118650197982788, + "learning_rate": 8.253300449731613e-06, + "loss": 0.0066, + "step": 16294 + }, + { + "epoch": 1.7665871639202082, + "grad_norm": 1.319054126739502, + "learning_rate": 8.252575076164225e-06, + "loss": 0.005, + "step": 16295 + }, + { + "epoch": 1.766695576756288, + "grad_norm": 1.4663214683532715, + "learning_rate": 8.251849702596838e-06, + "loss": 0.0163, + "step": 16296 + }, + { + "epoch": 1.7668039895923677, + "grad_norm": 2.1341915130615234, + "learning_rate": 8.25112432902945e-06, + "loss": 0.0588, + "step": 16297 + }, + { + "epoch": 1.7669124024284475, + "grad_norm": 2.639600992202759, + "learning_rate": 8.250398955462063e-06, + "loss": 0.0269, + "step": 16298 + }, + { + "epoch": 1.7670208152645275, + "grad_norm": 0.32470497488975525, + "learning_rate": 8.249673581894677e-06, + "loss": 0.0022, + "step": 16299 + }, + { + "epoch": 1.767129228100607, + "grad_norm": 4.166304588317871, + "learning_rate": 8.248948208327289e-06, + "loss": 0.0224, + "step": 16300 + }, + { + "epoch": 1.767237640936687, + "grad_norm": 4.441570281982422, + "learning_rate": 8.248222834759903e-06, + "loss": 0.0336, + "step": 16301 + }, + { + "epoch": 1.7673460537727665, + "grad_norm": 1.3097015619277954, + "learning_rate": 8.247497461192515e-06, + "loss": 0.0097, + "step": 16302 + }, + { + "epoch": 1.7674544666088465, + "grad_norm": 3.3397276401519775, + "learning_rate": 8.246772087625129e-06, + "loss": 0.0342, + "step": 16303 + }, + { + "epoch": 1.7675628794449263, + "grad_norm": 4.508078098297119, + "learning_rate": 8.24604671405774e-06, + "loss": 0.025, + "step": 16304 + }, + { + "epoch": 1.767671292281006, + "grad_norm": 3.2839324474334717, + "learning_rate": 8.245321340490353e-06, + "loss": 0.065, + "step": 16305 + }, + { + "epoch": 1.7677797051170858, + "grad_norm": 3.5590906143188477, + "learning_rate": 8.244595966922965e-06, + "loss": 0.039, + "step": 16306 + }, + { + "epoch": 1.7678881179531656, + "grad_norm": 0.287065327167511, + "learning_rate": 8.243870593355579e-06, + "loss": 0.0021, + "step": 16307 + }, + { + "epoch": 1.7679965307892456, + "grad_norm": 0.32116568088531494, + "learning_rate": 8.243145219788191e-06, + "loss": 0.0039, + "step": 16308 + }, + { + "epoch": 1.768104943625325, + "grad_norm": 1.6121981143951416, + "learning_rate": 8.242419846220805e-06, + "loss": 0.0078, + "step": 16309 + }, + { + "epoch": 1.768213356461405, + "grad_norm": 1.1384838819503784, + "learning_rate": 8.241694472653417e-06, + "loss": 0.0062, + "step": 16310 + }, + { + "epoch": 1.7683217692974849, + "grad_norm": 0.6973093748092651, + "learning_rate": 8.24096909908603e-06, + "loss": 0.0214, + "step": 16311 + }, + { + "epoch": 1.7684301821335646, + "grad_norm": 4.839329242706299, + "learning_rate": 8.240243725518643e-06, + "loss": 0.0412, + "step": 16312 + }, + { + "epoch": 1.7685385949696444, + "grad_norm": 4.938719272613525, + "learning_rate": 8.239518351951255e-06, + "loss": 0.0251, + "step": 16313 + }, + { + "epoch": 1.7686470078057241, + "grad_norm": 1.129237413406372, + "learning_rate": 8.238792978383869e-06, + "loss": 0.0116, + "step": 16314 + }, + { + "epoch": 1.7687554206418041, + "grad_norm": 0.9677740335464478, + "learning_rate": 8.238067604816481e-06, + "loss": 0.0083, + "step": 16315 + }, + { + "epoch": 1.7688638334778837, + "grad_norm": 5.160699367523193, + "learning_rate": 8.237342231249093e-06, + "loss": 0.0203, + "step": 16316 + }, + { + "epoch": 1.7689722463139637, + "grad_norm": 0.5992897748947144, + "learning_rate": 8.236616857681707e-06, + "loss": 0.0028, + "step": 16317 + }, + { + "epoch": 1.7690806591500432, + "grad_norm": 0.22294802963733673, + "learning_rate": 8.235891484114321e-06, + "loss": 0.0024, + "step": 16318 + }, + { + "epoch": 1.7691890719861232, + "grad_norm": 2.008115768432617, + "learning_rate": 8.235166110546933e-06, + "loss": 0.0057, + "step": 16319 + }, + { + "epoch": 1.769297484822203, + "grad_norm": 0.2732962369918823, + "learning_rate": 8.234440736979545e-06, + "loss": 0.0039, + "step": 16320 + }, + { + "epoch": 1.7694058976582827, + "grad_norm": 2.47880220413208, + "learning_rate": 8.233715363412157e-06, + "loss": 0.0206, + "step": 16321 + }, + { + "epoch": 1.7695143104943627, + "grad_norm": 1.589944839477539, + "learning_rate": 8.232989989844771e-06, + "loss": 0.0212, + "step": 16322 + }, + { + "epoch": 1.7696227233304422, + "grad_norm": 2.386180877685547, + "learning_rate": 8.232264616277383e-06, + "loss": 0.0243, + "step": 16323 + }, + { + "epoch": 1.7697311361665222, + "grad_norm": 5.1591339111328125, + "learning_rate": 8.231539242709995e-06, + "loss": 0.0142, + "step": 16324 + }, + { + "epoch": 1.7698395490026018, + "grad_norm": 3.025263547897339, + "learning_rate": 8.23081386914261e-06, + "loss": 0.0643, + "step": 16325 + }, + { + "epoch": 1.7699479618386817, + "grad_norm": 1.8051432371139526, + "learning_rate": 8.230088495575221e-06, + "loss": 0.0194, + "step": 16326 + }, + { + "epoch": 1.7700563746747615, + "grad_norm": 2.03570556640625, + "learning_rate": 8.229363122007835e-06, + "loss": 0.0265, + "step": 16327 + }, + { + "epoch": 1.7701647875108413, + "grad_norm": 2.7299091815948486, + "learning_rate": 8.228637748440447e-06, + "loss": 0.0223, + "step": 16328 + }, + { + "epoch": 1.770273200346921, + "grad_norm": 0.8681688904762268, + "learning_rate": 8.227912374873061e-06, + "loss": 0.0187, + "step": 16329 + }, + { + "epoch": 1.7703816131830008, + "grad_norm": 1.8225148916244507, + "learning_rate": 8.227187001305673e-06, + "loss": 0.015, + "step": 16330 + }, + { + "epoch": 1.7704900260190808, + "grad_norm": 1.042686939239502, + "learning_rate": 8.226461627738286e-06, + "loss": 0.0087, + "step": 16331 + }, + { + "epoch": 1.7705984388551603, + "grad_norm": 2.2075815200805664, + "learning_rate": 8.225736254170898e-06, + "loss": 0.0336, + "step": 16332 + }, + { + "epoch": 1.7707068516912403, + "grad_norm": 5.1042680740356445, + "learning_rate": 8.225010880603512e-06, + "loss": 0.0367, + "step": 16333 + }, + { + "epoch": 1.77081526452732, + "grad_norm": 5.243442058563232, + "learning_rate": 8.224285507036124e-06, + "loss": 0.0656, + "step": 16334 + }, + { + "epoch": 1.7709236773633998, + "grad_norm": 1.5972756147384644, + "learning_rate": 8.223560133468737e-06, + "loss": 0.0372, + "step": 16335 + }, + { + "epoch": 1.7710320901994796, + "grad_norm": 2.9146718978881836, + "learning_rate": 8.22283475990135e-06, + "loss": 0.0097, + "step": 16336 + }, + { + "epoch": 1.7711405030355594, + "grad_norm": 1.5600725412368774, + "learning_rate": 8.222109386333963e-06, + "loss": 0.0071, + "step": 16337 + }, + { + "epoch": 1.7712489158716394, + "grad_norm": 0.17704182863235474, + "learning_rate": 8.221384012766576e-06, + "loss": 0.001, + "step": 16338 + }, + { + "epoch": 1.771357328707719, + "grad_norm": 2.4491450786590576, + "learning_rate": 8.220658639199188e-06, + "loss": 0.0524, + "step": 16339 + }, + { + "epoch": 1.7714657415437989, + "grad_norm": 0.5792939066886902, + "learning_rate": 8.219933265631802e-06, + "loss": 0.0046, + "step": 16340 + }, + { + "epoch": 1.7715741543798784, + "grad_norm": 0.5331572890281677, + "learning_rate": 8.219207892064414e-06, + "loss": 0.0021, + "step": 16341 + }, + { + "epoch": 1.7716825672159584, + "grad_norm": 0.7658209800720215, + "learning_rate": 8.218482518497026e-06, + "loss": 0.0059, + "step": 16342 + }, + { + "epoch": 1.7717909800520382, + "grad_norm": 3.3050289154052734, + "learning_rate": 8.21775714492964e-06, + "loss": 0.0172, + "step": 16343 + }, + { + "epoch": 1.771899392888118, + "grad_norm": 1.056286096572876, + "learning_rate": 8.217031771362254e-06, + "loss": 0.0223, + "step": 16344 + }, + { + "epoch": 1.7720078057241977, + "grad_norm": 1.9535672664642334, + "learning_rate": 8.216306397794866e-06, + "loss": 0.0276, + "step": 16345 + }, + { + "epoch": 1.7721162185602775, + "grad_norm": 6.437510013580322, + "learning_rate": 8.215581024227478e-06, + "loss": 0.0527, + "step": 16346 + }, + { + "epoch": 1.7722246313963574, + "grad_norm": 2.110567808151245, + "learning_rate": 8.21485565066009e-06, + "loss": 0.0122, + "step": 16347 + }, + { + "epoch": 1.772333044232437, + "grad_norm": 2.4240832328796387, + "learning_rate": 8.214130277092704e-06, + "loss": 0.0454, + "step": 16348 + }, + { + "epoch": 1.772441457068517, + "grad_norm": 6.742363452911377, + "learning_rate": 8.213404903525316e-06, + "loss": 0.0155, + "step": 16349 + }, + { + "epoch": 1.7725498699045967, + "grad_norm": 1.7596938610076904, + "learning_rate": 8.212679529957928e-06, + "loss": 0.0222, + "step": 16350 + }, + { + "epoch": 1.7726582827406765, + "grad_norm": 1.059178352355957, + "learning_rate": 8.211954156390542e-06, + "loss": 0.0054, + "step": 16351 + }, + { + "epoch": 1.7727666955767563, + "grad_norm": 3.9675323963165283, + "learning_rate": 8.211228782823154e-06, + "loss": 0.0263, + "step": 16352 + }, + { + "epoch": 1.772875108412836, + "grad_norm": 1.0430970191955566, + "learning_rate": 8.210503409255768e-06, + "loss": 0.0059, + "step": 16353 + }, + { + "epoch": 1.772983521248916, + "grad_norm": 0.5376417636871338, + "learning_rate": 8.20977803568838e-06, + "loss": 0.0033, + "step": 16354 + }, + { + "epoch": 1.7730919340849955, + "grad_norm": 1.2168364524841309, + "learning_rate": 8.209052662120994e-06, + "loss": 0.0098, + "step": 16355 + }, + { + "epoch": 1.7732003469210755, + "grad_norm": 0.5946487188339233, + "learning_rate": 8.208327288553606e-06, + "loss": 0.009, + "step": 16356 + }, + { + "epoch": 1.7733087597571553, + "grad_norm": 4.831393718719482, + "learning_rate": 8.207601914986218e-06, + "loss": 0.0155, + "step": 16357 + }, + { + "epoch": 1.773417172593235, + "grad_norm": 2.233153820037842, + "learning_rate": 8.20687654141883e-06, + "loss": 0.0255, + "step": 16358 + }, + { + "epoch": 1.7735255854293148, + "grad_norm": 0.8831499814987183, + "learning_rate": 8.206151167851444e-06, + "loss": 0.0074, + "step": 16359 + }, + { + "epoch": 1.7736339982653946, + "grad_norm": 5.529512405395508, + "learning_rate": 8.205425794284056e-06, + "loss": 0.0124, + "step": 16360 + }, + { + "epoch": 1.7737424111014746, + "grad_norm": 3.5413620471954346, + "learning_rate": 8.20470042071667e-06, + "loss": 0.0165, + "step": 16361 + }, + { + "epoch": 1.773850823937554, + "grad_norm": 11.458171844482422, + "learning_rate": 8.203975047149282e-06, + "loss": 0.0449, + "step": 16362 + }, + { + "epoch": 1.773959236773634, + "grad_norm": 1.980036973953247, + "learning_rate": 8.203249673581896e-06, + "loss": 0.027, + "step": 16363 + }, + { + "epoch": 1.7740676496097136, + "grad_norm": 0.9762257933616638, + "learning_rate": 8.202524300014508e-06, + "loss": 0.0068, + "step": 16364 + }, + { + "epoch": 1.7741760624457936, + "grad_norm": 1.1186734437942505, + "learning_rate": 8.20179892644712e-06, + "loss": 0.0145, + "step": 16365 + }, + { + "epoch": 1.7742844752818734, + "grad_norm": 0.14115536212921143, + "learning_rate": 8.201073552879734e-06, + "loss": 0.0018, + "step": 16366 + }, + { + "epoch": 1.7743928881179531, + "grad_norm": 1.110258936882019, + "learning_rate": 8.200348179312346e-06, + "loss": 0.019, + "step": 16367 + }, + { + "epoch": 1.774501300954033, + "grad_norm": 0.246926411986351, + "learning_rate": 8.199622805744959e-06, + "loss": 0.0014, + "step": 16368 + }, + { + "epoch": 1.7746097137901127, + "grad_norm": 2.6149699687957764, + "learning_rate": 8.198897432177572e-06, + "loss": 0.0138, + "step": 16369 + }, + { + "epoch": 1.7747181266261927, + "grad_norm": 0.32613322138786316, + "learning_rate": 8.198172058610186e-06, + "loss": 0.0027, + "step": 16370 + }, + { + "epoch": 1.7748265394622722, + "grad_norm": 1.784284234046936, + "learning_rate": 8.197446685042798e-06, + "loss": 0.0041, + "step": 16371 + }, + { + "epoch": 1.7749349522983522, + "grad_norm": 2.382788896560669, + "learning_rate": 8.19672131147541e-06, + "loss": 0.0455, + "step": 16372 + }, + { + "epoch": 1.775043365134432, + "grad_norm": 3.82094669342041, + "learning_rate": 8.195995937908023e-06, + "loss": 0.0248, + "step": 16373 + }, + { + "epoch": 1.7751517779705117, + "grad_norm": 2.2257044315338135, + "learning_rate": 8.195270564340636e-06, + "loss": 0.0219, + "step": 16374 + }, + { + "epoch": 1.7752601908065915, + "grad_norm": 0.9202601909637451, + "learning_rate": 8.194545190773249e-06, + "loss": 0.0062, + "step": 16375 + }, + { + "epoch": 1.7753686036426712, + "grad_norm": 6.269824028015137, + "learning_rate": 8.19381981720586e-06, + "loss": 0.0241, + "step": 16376 + }, + { + "epoch": 1.7754770164787512, + "grad_norm": 0.054377853870391846, + "learning_rate": 8.193094443638475e-06, + "loss": 0.0006, + "step": 16377 + }, + { + "epoch": 1.7755854293148308, + "grad_norm": 2.017529249191284, + "learning_rate": 8.192369070071087e-06, + "loss": 0.0074, + "step": 16378 + }, + { + "epoch": 1.7756938421509108, + "grad_norm": 0.0835251584649086, + "learning_rate": 8.1916436965037e-06, + "loss": 0.0009, + "step": 16379 + }, + { + "epoch": 1.7758022549869905, + "grad_norm": 3.5450825691223145, + "learning_rate": 8.190918322936313e-06, + "loss": 0.022, + "step": 16380 + }, + { + "epoch": 1.7759106678230703, + "grad_norm": 2.1130259037017822, + "learning_rate": 8.190192949368927e-06, + "loss": 0.0211, + "step": 16381 + }, + { + "epoch": 1.77601908065915, + "grad_norm": 0.9969841837882996, + "learning_rate": 8.189467575801539e-06, + "loss": 0.0161, + "step": 16382 + }, + { + "epoch": 1.7761274934952298, + "grad_norm": 1.1695095300674438, + "learning_rate": 8.188742202234151e-06, + "loss": 0.0129, + "step": 16383 + }, + { + "epoch": 1.7762359063313098, + "grad_norm": 5.230945110321045, + "learning_rate": 8.188016828666763e-06, + "loss": 0.0964, + "step": 16384 + }, + { + "epoch": 1.7763443191673893, + "grad_norm": 3.093369960784912, + "learning_rate": 8.187291455099377e-06, + "loss": 0.0295, + "step": 16385 + }, + { + "epoch": 1.7764527320034693, + "grad_norm": 3.735004425048828, + "learning_rate": 8.186566081531989e-06, + "loss": 0.0194, + "step": 16386 + }, + { + "epoch": 1.7765611448395489, + "grad_norm": 2.0406196117401123, + "learning_rate": 8.185840707964603e-06, + "loss": 0.0244, + "step": 16387 + }, + { + "epoch": 1.7766695576756288, + "grad_norm": 0.8930370807647705, + "learning_rate": 8.185115334397215e-06, + "loss": 0.0097, + "step": 16388 + }, + { + "epoch": 1.7767779705117086, + "grad_norm": 2.47750186920166, + "learning_rate": 8.184389960829829e-06, + "loss": 0.0266, + "step": 16389 + }, + { + "epoch": 1.7768863833477884, + "grad_norm": 0.31108739972114563, + "learning_rate": 8.183664587262441e-06, + "loss": 0.003, + "step": 16390 + }, + { + "epoch": 1.7769947961838681, + "grad_norm": 1.6050124168395996, + "learning_rate": 8.182939213695053e-06, + "loss": 0.0128, + "step": 16391 + }, + { + "epoch": 1.777103209019948, + "grad_norm": 3.1719725131988525, + "learning_rate": 8.182213840127667e-06, + "loss": 0.0078, + "step": 16392 + }, + { + "epoch": 1.7772116218560279, + "grad_norm": 0.22541479766368866, + "learning_rate": 8.181488466560279e-06, + "loss": 0.002, + "step": 16393 + }, + { + "epoch": 1.7773200346921074, + "grad_norm": 1.208242416381836, + "learning_rate": 8.180763092992891e-06, + "loss": 0.0059, + "step": 16394 + }, + { + "epoch": 1.7774284475281874, + "grad_norm": 3.5724666118621826, + "learning_rate": 8.180037719425505e-06, + "loss": 0.0184, + "step": 16395 + }, + { + "epoch": 1.7775368603642672, + "grad_norm": 3.225670337677002, + "learning_rate": 8.179312345858119e-06, + "loss": 0.0319, + "step": 16396 + }, + { + "epoch": 1.777645273200347, + "grad_norm": 1.1287939548492432, + "learning_rate": 8.178586972290731e-06, + "loss": 0.0356, + "step": 16397 + }, + { + "epoch": 1.7777536860364267, + "grad_norm": 1.8219867944717407, + "learning_rate": 8.177861598723343e-06, + "loss": 0.0529, + "step": 16398 + }, + { + "epoch": 1.7778620988725065, + "grad_norm": 1.4264506101608276, + "learning_rate": 8.177136225155955e-06, + "loss": 0.0298, + "step": 16399 + }, + { + "epoch": 1.7779705117085864, + "grad_norm": 10.580500602722168, + "learning_rate": 8.17641085158857e-06, + "loss": 0.0397, + "step": 16400 + }, + { + "epoch": 1.778078924544666, + "grad_norm": 0.6041575074195862, + "learning_rate": 8.175685478021181e-06, + "loss": 0.0072, + "step": 16401 + }, + { + "epoch": 1.778187337380746, + "grad_norm": 12.395112037658691, + "learning_rate": 8.174960104453793e-06, + "loss": 0.0172, + "step": 16402 + }, + { + "epoch": 1.7782957502168255, + "grad_norm": 0.44158294796943665, + "learning_rate": 8.174234730886407e-06, + "loss": 0.0028, + "step": 16403 + }, + { + "epoch": 1.7784041630529055, + "grad_norm": 4.748767852783203, + "learning_rate": 8.173509357319021e-06, + "loss": 0.0186, + "step": 16404 + }, + { + "epoch": 1.7785125758889853, + "grad_norm": 3.582118272781372, + "learning_rate": 8.172783983751633e-06, + "loss": 0.0161, + "step": 16405 + }, + { + "epoch": 1.778620988725065, + "grad_norm": 0.5867692828178406, + "learning_rate": 8.172058610184245e-06, + "loss": 0.0056, + "step": 16406 + }, + { + "epoch": 1.7787294015611448, + "grad_norm": 0.8457837700843811, + "learning_rate": 8.17133323661686e-06, + "loss": 0.0057, + "step": 16407 + }, + { + "epoch": 1.7788378143972245, + "grad_norm": 0.34197378158569336, + "learning_rate": 8.170607863049471e-06, + "loss": 0.0037, + "step": 16408 + }, + { + "epoch": 1.7789462272333045, + "grad_norm": 2.451810598373413, + "learning_rate": 8.169882489482084e-06, + "loss": 0.0323, + "step": 16409 + }, + { + "epoch": 1.779054640069384, + "grad_norm": 1.3605990409851074, + "learning_rate": 8.169157115914696e-06, + "loss": 0.0147, + "step": 16410 + }, + { + "epoch": 1.779163052905464, + "grad_norm": 1.242821455001831, + "learning_rate": 8.16843174234731e-06, + "loss": 0.0062, + "step": 16411 + }, + { + "epoch": 1.7792714657415438, + "grad_norm": 3.1903815269470215, + "learning_rate": 8.167706368779922e-06, + "loss": 0.0133, + "step": 16412 + }, + { + "epoch": 1.7793798785776236, + "grad_norm": 2.7414183616638184, + "learning_rate": 8.166980995212535e-06, + "loss": 0.026, + "step": 16413 + }, + { + "epoch": 1.7794882914137033, + "grad_norm": 3.2936627864837646, + "learning_rate": 8.166255621645148e-06, + "loss": 0.0263, + "step": 16414 + }, + { + "epoch": 1.7795967042497831, + "grad_norm": 5.473005294799805, + "learning_rate": 8.165530248077761e-06, + "loss": 0.064, + "step": 16415 + }, + { + "epoch": 1.779705117085863, + "grad_norm": 1.702149510383606, + "learning_rate": 8.164804874510374e-06, + "loss": 0.0082, + "step": 16416 + }, + { + "epoch": 1.7798135299219426, + "grad_norm": 1.0831447839736938, + "learning_rate": 8.164079500942986e-06, + "loss": 0.0229, + "step": 16417 + }, + { + "epoch": 1.7799219427580226, + "grad_norm": 2.1778178215026855, + "learning_rate": 8.1633541273756e-06, + "loss": 0.0145, + "step": 16418 + }, + { + "epoch": 1.7800303555941024, + "grad_norm": 0.8654398918151855, + "learning_rate": 8.162628753808212e-06, + "loss": 0.0056, + "step": 16419 + }, + { + "epoch": 1.7801387684301822, + "grad_norm": 3.8004093170166016, + "learning_rate": 8.161903380240824e-06, + "loss": 0.0468, + "step": 16420 + }, + { + "epoch": 1.780247181266262, + "grad_norm": 5.3922576904296875, + "learning_rate": 8.161178006673438e-06, + "loss": 0.029, + "step": 16421 + }, + { + "epoch": 1.7803555941023417, + "grad_norm": 0.04661485552787781, + "learning_rate": 8.160452633106052e-06, + "loss": 0.0005, + "step": 16422 + }, + { + "epoch": 1.7804640069384217, + "grad_norm": 0.1884855180978775, + "learning_rate": 8.159727259538664e-06, + "loss": 0.0023, + "step": 16423 + }, + { + "epoch": 1.7805724197745012, + "grad_norm": 1.37618887424469, + "learning_rate": 8.159001885971276e-06, + "loss": 0.0161, + "step": 16424 + }, + { + "epoch": 1.7806808326105812, + "grad_norm": 2.079591989517212, + "learning_rate": 8.158276512403888e-06, + "loss": 0.0104, + "step": 16425 + }, + { + "epoch": 1.7807892454466607, + "grad_norm": 1.242645025253296, + "learning_rate": 8.157551138836502e-06, + "loss": 0.0117, + "step": 16426 + }, + { + "epoch": 1.7808976582827407, + "grad_norm": 3.3976237773895264, + "learning_rate": 8.156825765269114e-06, + "loss": 0.047, + "step": 16427 + }, + { + "epoch": 1.7810060711188205, + "grad_norm": 1.7214099168777466, + "learning_rate": 8.156100391701726e-06, + "loss": 0.0608, + "step": 16428 + }, + { + "epoch": 1.7811144839549002, + "grad_norm": 1.9834104776382446, + "learning_rate": 8.15537501813434e-06, + "loss": 0.0219, + "step": 16429 + }, + { + "epoch": 1.78122289679098, + "grad_norm": 5.845999240875244, + "learning_rate": 8.154649644566954e-06, + "loss": 0.0389, + "step": 16430 + }, + { + "epoch": 1.7813313096270598, + "grad_norm": 0.8480894565582275, + "learning_rate": 8.153924270999566e-06, + "loss": 0.0078, + "step": 16431 + }, + { + "epoch": 1.7814397224631398, + "grad_norm": 12.077768325805664, + "learning_rate": 8.153198897432178e-06, + "loss": 0.0809, + "step": 16432 + }, + { + "epoch": 1.7815481352992193, + "grad_norm": 1.0583150386810303, + "learning_rate": 8.152473523864792e-06, + "loss": 0.0024, + "step": 16433 + }, + { + "epoch": 1.7816565481352993, + "grad_norm": 3.799659490585327, + "learning_rate": 8.151748150297404e-06, + "loss": 0.023, + "step": 16434 + }, + { + "epoch": 1.781764960971379, + "grad_norm": 1.071000099182129, + "learning_rate": 8.151022776730016e-06, + "loss": 0.0101, + "step": 16435 + }, + { + "epoch": 1.7818733738074588, + "grad_norm": 1.907531976699829, + "learning_rate": 8.150297403162628e-06, + "loss": 0.0131, + "step": 16436 + }, + { + "epoch": 1.7819817866435386, + "grad_norm": 1.082221508026123, + "learning_rate": 8.149572029595242e-06, + "loss": 0.0077, + "step": 16437 + }, + { + "epoch": 1.7820901994796183, + "grad_norm": 0.11291930824518204, + "learning_rate": 8.148846656027854e-06, + "loss": 0.0011, + "step": 16438 + }, + { + "epoch": 1.7821986123156983, + "grad_norm": 1.6518642902374268, + "learning_rate": 8.148121282460468e-06, + "loss": 0.0095, + "step": 16439 + }, + { + "epoch": 1.7823070251517779, + "grad_norm": 0.05865398794412613, + "learning_rate": 8.14739590889308e-06, + "loss": 0.0008, + "step": 16440 + }, + { + "epoch": 1.7824154379878578, + "grad_norm": 1.386029839515686, + "learning_rate": 8.146670535325694e-06, + "loss": 0.0123, + "step": 16441 + }, + { + "epoch": 1.7825238508239376, + "grad_norm": 3.9601709842681885, + "learning_rate": 8.145945161758306e-06, + "loss": 0.0624, + "step": 16442 + }, + { + "epoch": 1.7826322636600174, + "grad_norm": 4.393413543701172, + "learning_rate": 8.145219788190918e-06, + "loss": 0.0233, + "step": 16443 + }, + { + "epoch": 1.7827406764960971, + "grad_norm": 1.57350754737854, + "learning_rate": 8.144494414623532e-06, + "loss": 0.0077, + "step": 16444 + }, + { + "epoch": 1.782849089332177, + "grad_norm": 0.8212244510650635, + "learning_rate": 8.143769041056144e-06, + "loss": 0.0059, + "step": 16445 + }, + { + "epoch": 1.7829575021682569, + "grad_norm": 0.59381103515625, + "learning_rate": 8.143043667488757e-06, + "loss": 0.0038, + "step": 16446 + }, + { + "epoch": 1.7830659150043364, + "grad_norm": 3.036330461502075, + "learning_rate": 8.14231829392137e-06, + "loss": 0.0123, + "step": 16447 + }, + { + "epoch": 1.7831743278404164, + "grad_norm": 3.331768274307251, + "learning_rate": 8.141592920353984e-06, + "loss": 0.0333, + "step": 16448 + }, + { + "epoch": 1.783282740676496, + "grad_norm": 0.2725259065628052, + "learning_rate": 8.140867546786596e-06, + "loss": 0.002, + "step": 16449 + }, + { + "epoch": 1.783391153512576, + "grad_norm": 0.8063828349113464, + "learning_rate": 8.140142173219209e-06, + "loss": 0.0091, + "step": 16450 + }, + { + "epoch": 1.7834995663486557, + "grad_norm": 5.715100288391113, + "learning_rate": 8.13941679965182e-06, + "loss": 0.0673, + "step": 16451 + }, + { + "epoch": 1.7836079791847355, + "grad_norm": 2.7216849327087402, + "learning_rate": 8.138691426084434e-06, + "loss": 0.0265, + "step": 16452 + }, + { + "epoch": 1.7837163920208152, + "grad_norm": 1.8402361869812012, + "learning_rate": 8.137966052517047e-06, + "loss": 0.0457, + "step": 16453 + }, + { + "epoch": 1.783824804856895, + "grad_norm": 2.3041622638702393, + "learning_rate": 8.137240678949659e-06, + "loss": 0.0209, + "step": 16454 + }, + { + "epoch": 1.783933217692975, + "grad_norm": 6.583467960357666, + "learning_rate": 8.136515305382273e-06, + "loss": 0.0138, + "step": 16455 + }, + { + "epoch": 1.7840416305290545, + "grad_norm": 1.3047889471054077, + "learning_rate": 8.135789931814886e-06, + "loss": 0.0038, + "step": 16456 + }, + { + "epoch": 1.7841500433651345, + "grad_norm": 2.998115062713623, + "learning_rate": 8.135064558247499e-06, + "loss": 0.0297, + "step": 16457 + }, + { + "epoch": 1.7842584562012143, + "grad_norm": 1.5454810857772827, + "learning_rate": 8.13433918468011e-06, + "loss": 0.0435, + "step": 16458 + }, + { + "epoch": 1.784366869037294, + "grad_norm": 4.380617618560791, + "learning_rate": 8.133613811112725e-06, + "loss": 0.0074, + "step": 16459 + }, + { + "epoch": 1.7844752818733738, + "grad_norm": 2.547482490539551, + "learning_rate": 8.132888437545337e-06, + "loss": 0.0162, + "step": 16460 + }, + { + "epoch": 1.7845836947094535, + "grad_norm": 0.5670425891876221, + "learning_rate": 8.132163063977949e-06, + "loss": 0.0034, + "step": 16461 + }, + { + "epoch": 1.7846921075455335, + "grad_norm": 1.7442164421081543, + "learning_rate": 8.131437690410563e-06, + "loss": 0.0046, + "step": 16462 + }, + { + "epoch": 1.784800520381613, + "grad_norm": 3.1304385662078857, + "learning_rate": 8.130712316843175e-06, + "loss": 0.0456, + "step": 16463 + }, + { + "epoch": 1.784908933217693, + "grad_norm": 0.3788662850856781, + "learning_rate": 8.129986943275787e-06, + "loss": 0.0025, + "step": 16464 + }, + { + "epoch": 1.7850173460537726, + "grad_norm": 1.2997761964797974, + "learning_rate": 8.1292615697084e-06, + "loss": 0.0095, + "step": 16465 + }, + { + "epoch": 1.7851257588898526, + "grad_norm": 1.2948968410491943, + "learning_rate": 8.128536196141013e-06, + "loss": 0.0118, + "step": 16466 + }, + { + "epoch": 1.7852341717259324, + "grad_norm": 5.718914031982422, + "learning_rate": 8.127810822573627e-06, + "loss": 0.0117, + "step": 16467 + }, + { + "epoch": 1.7853425845620121, + "grad_norm": 3.674577236175537, + "learning_rate": 8.127085449006239e-06, + "loss": 0.0312, + "step": 16468 + }, + { + "epoch": 1.785450997398092, + "grad_norm": 3.6640939712524414, + "learning_rate": 8.126360075438851e-06, + "loss": 0.0409, + "step": 16469 + }, + { + "epoch": 1.7855594102341716, + "grad_norm": 1.648939609527588, + "learning_rate": 8.125634701871465e-06, + "loss": 0.0221, + "step": 16470 + }, + { + "epoch": 1.7856678230702516, + "grad_norm": 0.6681633591651917, + "learning_rate": 8.124909328304077e-06, + "loss": 0.0063, + "step": 16471 + }, + { + "epoch": 1.7857762359063312, + "grad_norm": 1.295788288116455, + "learning_rate": 8.12418395473669e-06, + "loss": 0.0402, + "step": 16472 + }, + { + "epoch": 1.7858846487424112, + "grad_norm": 1.383202314376831, + "learning_rate": 8.123458581169303e-06, + "loss": 0.0047, + "step": 16473 + }, + { + "epoch": 1.785993061578491, + "grad_norm": 3.1079671382904053, + "learning_rate": 8.122733207601917e-06, + "loss": 0.022, + "step": 16474 + }, + { + "epoch": 1.7861014744145707, + "grad_norm": 1.9886213541030884, + "learning_rate": 8.122007834034529e-06, + "loss": 0.0302, + "step": 16475 + }, + { + "epoch": 1.7862098872506504, + "grad_norm": 5.856794357299805, + "learning_rate": 8.121282460467141e-06, + "loss": 0.0319, + "step": 16476 + }, + { + "epoch": 1.7863183000867302, + "grad_norm": 1.404396414756775, + "learning_rate": 8.120557086899753e-06, + "loss": 0.0049, + "step": 16477 + }, + { + "epoch": 1.7864267129228102, + "grad_norm": 0.7751547694206238, + "learning_rate": 8.119831713332367e-06, + "loss": 0.0075, + "step": 16478 + }, + { + "epoch": 1.7865351257588897, + "grad_norm": 1.3439909219741821, + "learning_rate": 8.11910633976498e-06, + "loss": 0.0123, + "step": 16479 + }, + { + "epoch": 1.7866435385949697, + "grad_norm": 2.7429816722869873, + "learning_rate": 8.118380966197591e-06, + "loss": 0.0203, + "step": 16480 + }, + { + "epoch": 1.7867519514310495, + "grad_norm": 3.3879995346069336, + "learning_rate": 8.117655592630205e-06, + "loss": 0.0137, + "step": 16481 + }, + { + "epoch": 1.7868603642671292, + "grad_norm": 4.651520252227783, + "learning_rate": 8.116930219062819e-06, + "loss": 0.0536, + "step": 16482 + }, + { + "epoch": 1.786968777103209, + "grad_norm": 1.1789119243621826, + "learning_rate": 8.116204845495431e-06, + "loss": 0.0053, + "step": 16483 + }, + { + "epoch": 1.7870771899392888, + "grad_norm": 2.8045804500579834, + "learning_rate": 8.115479471928043e-06, + "loss": 0.0143, + "step": 16484 + }, + { + "epoch": 1.7871856027753688, + "grad_norm": 0.3025260269641876, + "learning_rate": 8.114754098360657e-06, + "loss": 0.0033, + "step": 16485 + }, + { + "epoch": 1.7872940156114483, + "grad_norm": 0.17683443427085876, + "learning_rate": 8.11402872479327e-06, + "loss": 0.0025, + "step": 16486 + }, + { + "epoch": 1.7874024284475283, + "grad_norm": 5.349608898162842, + "learning_rate": 8.113303351225882e-06, + "loss": 0.0648, + "step": 16487 + }, + { + "epoch": 1.7875108412836078, + "grad_norm": 3.375013828277588, + "learning_rate": 8.112577977658495e-06, + "loss": 0.0212, + "step": 16488 + }, + { + "epoch": 1.7876192541196878, + "grad_norm": 0.1674058437347412, + "learning_rate": 8.111852604091108e-06, + "loss": 0.0024, + "step": 16489 + }, + { + "epoch": 1.7877276669557676, + "grad_norm": 4.342864036560059, + "learning_rate": 8.11112723052372e-06, + "loss": 0.0255, + "step": 16490 + }, + { + "epoch": 1.7878360797918473, + "grad_norm": 8.785323143005371, + "learning_rate": 8.110401856956333e-06, + "loss": 0.0533, + "step": 16491 + }, + { + "epoch": 1.787944492627927, + "grad_norm": 2.813246726989746, + "learning_rate": 8.109676483388946e-06, + "loss": 0.0129, + "step": 16492 + }, + { + "epoch": 1.7880529054640069, + "grad_norm": 1.19400155544281, + "learning_rate": 8.10895110982156e-06, + "loss": 0.0066, + "step": 16493 + }, + { + "epoch": 1.7881613183000868, + "grad_norm": 3.5666286945343018, + "learning_rate": 8.108225736254172e-06, + "loss": 0.0603, + "step": 16494 + }, + { + "epoch": 1.7882697311361664, + "grad_norm": 10.671345710754395, + "learning_rate": 8.107500362686784e-06, + "loss": 0.0305, + "step": 16495 + }, + { + "epoch": 1.7883781439722464, + "grad_norm": 7.499608993530273, + "learning_rate": 8.106774989119398e-06, + "loss": 0.0505, + "step": 16496 + }, + { + "epoch": 1.7884865568083261, + "grad_norm": 0.5048074126243591, + "learning_rate": 8.10604961555201e-06, + "loss": 0.0017, + "step": 16497 + }, + { + "epoch": 1.788594969644406, + "grad_norm": 2.2620437145233154, + "learning_rate": 8.105324241984622e-06, + "loss": 0.009, + "step": 16498 + }, + { + "epoch": 1.7887033824804857, + "grad_norm": 7.296440601348877, + "learning_rate": 8.104598868417236e-06, + "loss": 0.0285, + "step": 16499 + }, + { + "epoch": 1.7888117953165654, + "grad_norm": 1.6741514205932617, + "learning_rate": 8.10387349484985e-06, + "loss": 0.0057, + "step": 16500 + }, + { + "epoch": 1.7889202081526454, + "grad_norm": 2.718654155731201, + "learning_rate": 8.103148121282462e-06, + "loss": 0.0225, + "step": 16501 + }, + { + "epoch": 1.789028620988725, + "grad_norm": 0.3468419015407562, + "learning_rate": 8.102422747715074e-06, + "loss": 0.0035, + "step": 16502 + }, + { + "epoch": 1.789137033824805, + "grad_norm": 2.858196496963501, + "learning_rate": 8.101697374147686e-06, + "loss": 0.0269, + "step": 16503 + }, + { + "epoch": 1.7892454466608847, + "grad_norm": 2.074310779571533, + "learning_rate": 8.1009720005803e-06, + "loss": 0.0096, + "step": 16504 + }, + { + "epoch": 1.7893538594969645, + "grad_norm": 2.016007661819458, + "learning_rate": 8.100246627012912e-06, + "loss": 0.0192, + "step": 16505 + }, + { + "epoch": 1.7894622723330442, + "grad_norm": 4.440334796905518, + "learning_rate": 8.099521253445524e-06, + "loss": 0.0194, + "step": 16506 + }, + { + "epoch": 1.789570685169124, + "grad_norm": 2.7075304985046387, + "learning_rate": 8.098795879878138e-06, + "loss": 0.027, + "step": 16507 + }, + { + "epoch": 1.789679098005204, + "grad_norm": 1.6427432298660278, + "learning_rate": 8.098070506310752e-06, + "loss": 0.0165, + "step": 16508 + }, + { + "epoch": 1.7897875108412835, + "grad_norm": 8.271612167358398, + "learning_rate": 8.097345132743364e-06, + "loss": 0.0408, + "step": 16509 + }, + { + "epoch": 1.7898959236773635, + "grad_norm": 0.3849305212497711, + "learning_rate": 8.096619759175976e-06, + "loss": 0.0022, + "step": 16510 + }, + { + "epoch": 1.790004336513443, + "grad_norm": 1.2956308126449585, + "learning_rate": 8.09589438560859e-06, + "loss": 0.0117, + "step": 16511 + }, + { + "epoch": 1.790112749349523, + "grad_norm": 1.9866316318511963, + "learning_rate": 8.095169012041202e-06, + "loss": 0.046, + "step": 16512 + }, + { + "epoch": 1.7902211621856028, + "grad_norm": 1.8271467685699463, + "learning_rate": 8.094443638473814e-06, + "loss": 0.003, + "step": 16513 + }, + { + "epoch": 1.7903295750216826, + "grad_norm": 1.7161571979522705, + "learning_rate": 8.093718264906428e-06, + "loss": 0.0245, + "step": 16514 + }, + { + "epoch": 1.7904379878577623, + "grad_norm": 1.721628189086914, + "learning_rate": 8.09299289133904e-06, + "loss": 0.0106, + "step": 16515 + }, + { + "epoch": 1.790546400693842, + "grad_norm": 3.359502077102661, + "learning_rate": 8.092267517771652e-06, + "loss": 0.0284, + "step": 16516 + }, + { + "epoch": 1.790654813529922, + "grad_norm": 0.908906877040863, + "learning_rate": 8.091542144204266e-06, + "loss": 0.008, + "step": 16517 + }, + { + "epoch": 1.7907632263660016, + "grad_norm": 1.366455078125, + "learning_rate": 8.090816770636878e-06, + "loss": 0.0111, + "step": 16518 + }, + { + "epoch": 1.7908716392020816, + "grad_norm": 3.097079277038574, + "learning_rate": 8.090091397069492e-06, + "loss": 0.0293, + "step": 16519 + }, + { + "epoch": 1.7909800520381614, + "grad_norm": 1.4869155883789062, + "learning_rate": 8.089366023502104e-06, + "loss": 0.0189, + "step": 16520 + }, + { + "epoch": 1.7910884648742411, + "grad_norm": 9.999404907226562, + "learning_rate": 8.088640649934716e-06, + "loss": 0.0656, + "step": 16521 + }, + { + "epoch": 1.7911968777103209, + "grad_norm": 6.433647155761719, + "learning_rate": 8.08791527636733e-06, + "loss": 0.0219, + "step": 16522 + }, + { + "epoch": 1.7913052905464006, + "grad_norm": 1.2508785724639893, + "learning_rate": 8.087189902799942e-06, + "loss": 0.0084, + "step": 16523 + }, + { + "epoch": 1.7914137033824806, + "grad_norm": 3.3796911239624023, + "learning_rate": 8.086464529232555e-06, + "loss": 0.008, + "step": 16524 + }, + { + "epoch": 1.7915221162185602, + "grad_norm": 3.0147440433502197, + "learning_rate": 8.085739155665168e-06, + "loss": 0.0087, + "step": 16525 + }, + { + "epoch": 1.7916305290546402, + "grad_norm": 2.267052173614502, + "learning_rate": 8.085013782097782e-06, + "loss": 0.0148, + "step": 16526 + }, + { + "epoch": 1.79173894189072, + "grad_norm": 0.24619612097740173, + "learning_rate": 8.084288408530394e-06, + "loss": 0.0019, + "step": 16527 + }, + { + "epoch": 1.7918473547267997, + "grad_norm": 0.21020165085792542, + "learning_rate": 8.083563034963007e-06, + "loss": 0.0006, + "step": 16528 + }, + { + "epoch": 1.7919557675628794, + "grad_norm": 2.212773561477661, + "learning_rate": 8.082837661395619e-06, + "loss": 0.0325, + "step": 16529 + }, + { + "epoch": 1.7920641803989592, + "grad_norm": 0.5876474380493164, + "learning_rate": 8.082112287828233e-06, + "loss": 0.0047, + "step": 16530 + }, + { + "epoch": 1.7921725932350392, + "grad_norm": 2.883462429046631, + "learning_rate": 8.081386914260845e-06, + "loss": 0.0587, + "step": 16531 + }, + { + "epoch": 1.7922810060711187, + "grad_norm": 0.3137849271297455, + "learning_rate": 8.080661540693457e-06, + "loss": 0.0018, + "step": 16532 + }, + { + "epoch": 1.7923894189071987, + "grad_norm": 1.206671118736267, + "learning_rate": 8.07993616712607e-06, + "loss": 0.0064, + "step": 16533 + }, + { + "epoch": 1.7924978317432783, + "grad_norm": 0.9577186107635498, + "learning_rate": 8.079210793558684e-06, + "loss": 0.0101, + "step": 16534 + }, + { + "epoch": 1.7926062445793582, + "grad_norm": 1.4282000064849854, + "learning_rate": 8.078485419991297e-06, + "loss": 0.0132, + "step": 16535 + }, + { + "epoch": 1.792714657415438, + "grad_norm": 4.053256034851074, + "learning_rate": 8.077760046423909e-06, + "loss": 0.0088, + "step": 16536 + }, + { + "epoch": 1.7928230702515178, + "grad_norm": 4.963192939758301, + "learning_rate": 8.077034672856523e-06, + "loss": 0.0228, + "step": 16537 + }, + { + "epoch": 1.7929314830875975, + "grad_norm": 2.2828001976013184, + "learning_rate": 8.076309299289135e-06, + "loss": 0.0513, + "step": 16538 + }, + { + "epoch": 1.7930398959236773, + "grad_norm": 5.61899471282959, + "learning_rate": 8.075583925721747e-06, + "loss": 0.0212, + "step": 16539 + }, + { + "epoch": 1.7931483087597573, + "grad_norm": 0.4846660792827606, + "learning_rate": 8.07485855215436e-06, + "loss": 0.0036, + "step": 16540 + }, + { + "epoch": 1.7932567215958368, + "grad_norm": 1.82940673828125, + "learning_rate": 8.074133178586973e-06, + "loss": 0.0052, + "step": 16541 + }, + { + "epoch": 1.7933651344319168, + "grad_norm": 3.7673394680023193, + "learning_rate": 8.073407805019585e-06, + "loss": 0.0527, + "step": 16542 + }, + { + "epoch": 1.7934735472679966, + "grad_norm": 3.2444286346435547, + "learning_rate": 8.072682431452199e-06, + "loss": 0.0219, + "step": 16543 + }, + { + "epoch": 1.7935819601040763, + "grad_norm": 0.747721254825592, + "learning_rate": 8.071957057884811e-06, + "loss": 0.0033, + "step": 16544 + }, + { + "epoch": 1.793690372940156, + "grad_norm": 1.2027689218521118, + "learning_rate": 8.071231684317425e-06, + "loss": 0.0048, + "step": 16545 + }, + { + "epoch": 1.7937987857762359, + "grad_norm": 0.7601005434989929, + "learning_rate": 8.070506310750037e-06, + "loss": 0.0026, + "step": 16546 + }, + { + "epoch": 1.7939071986123158, + "grad_norm": 1.12105131149292, + "learning_rate": 8.069780937182649e-06, + "loss": 0.0306, + "step": 16547 + }, + { + "epoch": 1.7940156114483954, + "grad_norm": 1.0178344249725342, + "learning_rate": 8.069055563615263e-06, + "loss": 0.0114, + "step": 16548 + }, + { + "epoch": 1.7941240242844754, + "grad_norm": 4.253794193267822, + "learning_rate": 8.068330190047875e-06, + "loss": 0.013, + "step": 16549 + }, + { + "epoch": 1.794232437120555, + "grad_norm": 2.2431509494781494, + "learning_rate": 8.067604816480487e-06, + "loss": 0.0115, + "step": 16550 + }, + { + "epoch": 1.794340849956635, + "grad_norm": 0.769599437713623, + "learning_rate": 8.066879442913101e-06, + "loss": 0.0103, + "step": 16551 + }, + { + "epoch": 1.7944492627927147, + "grad_norm": 5.12437105178833, + "learning_rate": 8.066154069345715e-06, + "loss": 0.0217, + "step": 16552 + }, + { + "epoch": 1.7945576756287944, + "grad_norm": 3.2222700119018555, + "learning_rate": 8.065428695778327e-06, + "loss": 0.0231, + "step": 16553 + }, + { + "epoch": 1.7946660884648742, + "grad_norm": 4.465208530426025, + "learning_rate": 8.06470332221094e-06, + "loss": 0.0227, + "step": 16554 + }, + { + "epoch": 1.794774501300954, + "grad_norm": 3.9796812534332275, + "learning_rate": 8.063977948643551e-06, + "loss": 0.0277, + "step": 16555 + }, + { + "epoch": 1.794882914137034, + "grad_norm": 3.54732084274292, + "learning_rate": 8.063252575076165e-06, + "loss": 0.0583, + "step": 16556 + }, + { + "epoch": 1.7949913269731135, + "grad_norm": 4.256447792053223, + "learning_rate": 8.062527201508777e-06, + "loss": 0.042, + "step": 16557 + }, + { + "epoch": 1.7950997398091935, + "grad_norm": 4.181018829345703, + "learning_rate": 8.06180182794139e-06, + "loss": 0.0248, + "step": 16558 + }, + { + "epoch": 1.7952081526452732, + "grad_norm": 0.3470861315727234, + "learning_rate": 8.061076454374003e-06, + "loss": 0.0028, + "step": 16559 + }, + { + "epoch": 1.795316565481353, + "grad_norm": 2.719325542449951, + "learning_rate": 8.060351080806617e-06, + "loss": 0.0437, + "step": 16560 + }, + { + "epoch": 1.7954249783174328, + "grad_norm": 0.17048227787017822, + "learning_rate": 8.05962570723923e-06, + "loss": 0.0012, + "step": 16561 + }, + { + "epoch": 1.7955333911535125, + "grad_norm": 1.465954065322876, + "learning_rate": 8.058900333671841e-06, + "loss": 0.0139, + "step": 16562 + }, + { + "epoch": 1.7956418039895925, + "grad_norm": 4.811099529266357, + "learning_rate": 8.058174960104455e-06, + "loss": 0.0247, + "step": 16563 + }, + { + "epoch": 1.795750216825672, + "grad_norm": 6.572256088256836, + "learning_rate": 8.057449586537067e-06, + "loss": 0.0198, + "step": 16564 + }, + { + "epoch": 1.795858629661752, + "grad_norm": 1.9451788663864136, + "learning_rate": 8.05672421296968e-06, + "loss": 0.0391, + "step": 16565 + }, + { + "epoch": 1.7959670424978318, + "grad_norm": 3.1051013469696045, + "learning_rate": 8.055998839402293e-06, + "loss": 0.0136, + "step": 16566 + }, + { + "epoch": 1.7960754553339116, + "grad_norm": 3.203094482421875, + "learning_rate": 8.055273465834906e-06, + "loss": 0.0627, + "step": 16567 + }, + { + "epoch": 1.7961838681699913, + "grad_norm": 1.432830572128296, + "learning_rate": 8.054548092267518e-06, + "loss": 0.0053, + "step": 16568 + }, + { + "epoch": 1.796292281006071, + "grad_norm": 7.92611837387085, + "learning_rate": 8.053822718700132e-06, + "loss": 0.0384, + "step": 16569 + }, + { + "epoch": 1.796400693842151, + "grad_norm": 0.49541935324668884, + "learning_rate": 8.053097345132744e-06, + "loss": 0.0021, + "step": 16570 + }, + { + "epoch": 1.7965091066782306, + "grad_norm": 4.189594268798828, + "learning_rate": 8.052371971565357e-06, + "loss": 0.0061, + "step": 16571 + }, + { + "epoch": 1.7966175195143106, + "grad_norm": 1.6813466548919678, + "learning_rate": 8.05164659799797e-06, + "loss": 0.0123, + "step": 16572 + }, + { + "epoch": 1.7967259323503901, + "grad_norm": 6.191125392913818, + "learning_rate": 8.050921224430582e-06, + "loss": 0.0251, + "step": 16573 + }, + { + "epoch": 1.7968343451864701, + "grad_norm": 1.4981576204299927, + "learning_rate": 8.050195850863196e-06, + "loss": 0.0158, + "step": 16574 + }, + { + "epoch": 1.7969427580225499, + "grad_norm": 3.266674518585205, + "learning_rate": 8.049470477295808e-06, + "loss": 0.0259, + "step": 16575 + }, + { + "epoch": 1.7970511708586296, + "grad_norm": 12.401097297668457, + "learning_rate": 8.04874510372842e-06, + "loss": 0.03, + "step": 16576 + }, + { + "epoch": 1.7971595836947094, + "grad_norm": 0.8975309729576111, + "learning_rate": 8.048019730161034e-06, + "loss": 0.0032, + "step": 16577 + }, + { + "epoch": 1.7972679965307892, + "grad_norm": 0.2155502736568451, + "learning_rate": 8.047294356593648e-06, + "loss": 0.0016, + "step": 16578 + }, + { + "epoch": 1.7973764093668692, + "grad_norm": 1.1537443399429321, + "learning_rate": 8.04656898302626e-06, + "loss": 0.0109, + "step": 16579 + }, + { + "epoch": 1.7974848222029487, + "grad_norm": 2.2897422313690186, + "learning_rate": 8.045843609458872e-06, + "loss": 0.0037, + "step": 16580 + }, + { + "epoch": 1.7975932350390287, + "grad_norm": 3.50350022315979, + "learning_rate": 8.045118235891484e-06, + "loss": 0.0635, + "step": 16581 + }, + { + "epoch": 1.7977016478751084, + "grad_norm": 0.7693248987197876, + "learning_rate": 8.044392862324098e-06, + "loss": 0.0037, + "step": 16582 + }, + { + "epoch": 1.7978100607111882, + "grad_norm": 3.170057535171509, + "learning_rate": 8.04366748875671e-06, + "loss": 0.015, + "step": 16583 + }, + { + "epoch": 1.797918473547268, + "grad_norm": 4.917531490325928, + "learning_rate": 8.042942115189322e-06, + "loss": 0.0141, + "step": 16584 + }, + { + "epoch": 1.7980268863833477, + "grad_norm": 0.6784783005714417, + "learning_rate": 8.042216741621936e-06, + "loss": 0.002, + "step": 16585 + }, + { + "epoch": 1.7981352992194277, + "grad_norm": 0.32785889506340027, + "learning_rate": 8.04149136805455e-06, + "loss": 0.0031, + "step": 16586 + }, + { + "epoch": 1.7982437120555073, + "grad_norm": 2.207041025161743, + "learning_rate": 8.040765994487162e-06, + "loss": 0.0393, + "step": 16587 + }, + { + "epoch": 1.7983521248915872, + "grad_norm": 2.5975863933563232, + "learning_rate": 8.040040620919774e-06, + "loss": 0.0342, + "step": 16588 + }, + { + "epoch": 1.798460537727667, + "grad_norm": 0.25265491008758545, + "learning_rate": 8.039315247352388e-06, + "loss": 0.0034, + "step": 16589 + }, + { + "epoch": 1.7985689505637468, + "grad_norm": 2.866347074508667, + "learning_rate": 8.038589873785e-06, + "loss": 0.0258, + "step": 16590 + }, + { + "epoch": 1.7986773633998265, + "grad_norm": 1.6070222854614258, + "learning_rate": 8.037864500217612e-06, + "loss": 0.0118, + "step": 16591 + }, + { + "epoch": 1.7987857762359063, + "grad_norm": 0.5904483795166016, + "learning_rate": 8.037139126650226e-06, + "loss": 0.0028, + "step": 16592 + }, + { + "epoch": 1.7988941890719863, + "grad_norm": 0.043921563774347305, + "learning_rate": 8.036413753082838e-06, + "loss": 0.0005, + "step": 16593 + }, + { + "epoch": 1.7990026019080658, + "grad_norm": 3.3023996353149414, + "learning_rate": 8.035688379515452e-06, + "loss": 0.0445, + "step": 16594 + }, + { + "epoch": 1.7991110147441458, + "grad_norm": 0.32519251108169556, + "learning_rate": 8.034963005948064e-06, + "loss": 0.0042, + "step": 16595 + }, + { + "epoch": 1.7992194275802254, + "grad_norm": 5.141491889953613, + "learning_rate": 8.034237632380676e-06, + "loss": 0.0199, + "step": 16596 + }, + { + "epoch": 1.7993278404163053, + "grad_norm": 4.640902042388916, + "learning_rate": 8.03351225881329e-06, + "loss": 0.0583, + "step": 16597 + }, + { + "epoch": 1.799436253252385, + "grad_norm": 1.1925336122512817, + "learning_rate": 8.032786885245902e-06, + "loss": 0.0369, + "step": 16598 + }, + { + "epoch": 1.7995446660884649, + "grad_norm": 3.606818437576294, + "learning_rate": 8.032061511678514e-06, + "loss": 0.0209, + "step": 16599 + }, + { + "epoch": 1.7996530789245446, + "grad_norm": 7.346982479095459, + "learning_rate": 8.031336138111128e-06, + "loss": 0.0599, + "step": 16600 + }, + { + "epoch": 1.7997614917606244, + "grad_norm": 1.7943114042282104, + "learning_rate": 8.03061076454374e-06, + "loss": 0.0132, + "step": 16601 + }, + { + "epoch": 1.7998699045967044, + "grad_norm": 2.241642951965332, + "learning_rate": 8.029885390976353e-06, + "loss": 0.0393, + "step": 16602 + }, + { + "epoch": 1.799978317432784, + "grad_norm": 1.84901762008667, + "learning_rate": 8.029160017408966e-06, + "loss": 0.0168, + "step": 16603 + }, + { + "epoch": 1.800086730268864, + "grad_norm": 3.9769067764282227, + "learning_rate": 8.02843464384158e-06, + "loss": 0.0459, + "step": 16604 + }, + { + "epoch": 1.8001951431049437, + "grad_norm": 0.21897883713245392, + "learning_rate": 8.027709270274192e-06, + "loss": 0.002, + "step": 16605 + }, + { + "epoch": 1.8003035559410234, + "grad_norm": 2.0051472187042236, + "learning_rate": 8.026983896706805e-06, + "loss": 0.0361, + "step": 16606 + }, + { + "epoch": 1.8004119687771032, + "grad_norm": 0.9284017086029053, + "learning_rate": 8.026258523139417e-06, + "loss": 0.0049, + "step": 16607 + }, + { + "epoch": 1.800520381613183, + "grad_norm": 4.329313278198242, + "learning_rate": 8.02553314957203e-06, + "loss": 0.0106, + "step": 16608 + }, + { + "epoch": 1.800628794449263, + "grad_norm": 4.053309440612793, + "learning_rate": 8.024807776004643e-06, + "loss": 0.0459, + "step": 16609 + }, + { + "epoch": 1.8007372072853425, + "grad_norm": 6.120275020599365, + "learning_rate": 8.024082402437255e-06, + "loss": 0.0237, + "step": 16610 + }, + { + "epoch": 1.8008456201214225, + "grad_norm": 0.6734498143196106, + "learning_rate": 8.023357028869869e-06, + "loss": 0.0014, + "step": 16611 + }, + { + "epoch": 1.800954032957502, + "grad_norm": 3.9455225467681885, + "learning_rate": 8.022631655302482e-06, + "loss": 0.0314, + "step": 16612 + }, + { + "epoch": 1.801062445793582, + "grad_norm": 2.9907565116882324, + "learning_rate": 8.021906281735095e-06, + "loss": 0.0444, + "step": 16613 + }, + { + "epoch": 1.8011708586296618, + "grad_norm": 11.439029693603516, + "learning_rate": 8.021180908167707e-06, + "loss": 0.0361, + "step": 16614 + }, + { + "epoch": 1.8012792714657415, + "grad_norm": 3.409012794494629, + "learning_rate": 8.02045553460032e-06, + "loss": 0.0106, + "step": 16615 + }, + { + "epoch": 1.8013876843018215, + "grad_norm": 1.7736473083496094, + "learning_rate": 8.019730161032933e-06, + "loss": 0.0171, + "step": 16616 + }, + { + "epoch": 1.801496097137901, + "grad_norm": 0.7426741123199463, + "learning_rate": 8.019004787465545e-06, + "loss": 0.0021, + "step": 16617 + }, + { + "epoch": 1.801604509973981, + "grad_norm": 4.95595645904541, + "learning_rate": 8.018279413898159e-06, + "loss": 0.042, + "step": 16618 + }, + { + "epoch": 1.8017129228100606, + "grad_norm": 4.416869640350342, + "learning_rate": 8.017554040330771e-06, + "loss": 0.0203, + "step": 16619 + }, + { + "epoch": 1.8018213356461406, + "grad_norm": 4.110374927520752, + "learning_rate": 8.016828666763385e-06, + "loss": 0.0101, + "step": 16620 + }, + { + "epoch": 1.8019297484822203, + "grad_norm": 0.5137751698493958, + "learning_rate": 8.016103293195997e-06, + "loss": 0.0026, + "step": 16621 + }, + { + "epoch": 1.8020381613183, + "grad_norm": 0.5938875675201416, + "learning_rate": 8.015377919628609e-06, + "loss": 0.0034, + "step": 16622 + }, + { + "epoch": 1.8021465741543798, + "grad_norm": 4.579678535461426, + "learning_rate": 8.014652546061223e-06, + "loss": 0.0353, + "step": 16623 + }, + { + "epoch": 1.8022549869904596, + "grad_norm": 2.285993814468384, + "learning_rate": 8.013927172493835e-06, + "loss": 0.0202, + "step": 16624 + }, + { + "epoch": 1.8023633998265396, + "grad_norm": 0.7188752293586731, + "learning_rate": 8.013201798926447e-06, + "loss": 0.0033, + "step": 16625 + }, + { + "epoch": 1.8024718126626191, + "grad_norm": 0.1921108514070511, + "learning_rate": 8.012476425359061e-06, + "loss": 0.0021, + "step": 16626 + }, + { + "epoch": 1.8025802254986991, + "grad_norm": 3.2535016536712646, + "learning_rate": 8.011751051791673e-06, + "loss": 0.0284, + "step": 16627 + }, + { + "epoch": 1.8026886383347789, + "grad_norm": 0.4178833067417145, + "learning_rate": 8.011025678224285e-06, + "loss": 0.0045, + "step": 16628 + }, + { + "epoch": 1.8027970511708586, + "grad_norm": 3.446779727935791, + "learning_rate": 8.010300304656899e-06, + "loss": 0.0177, + "step": 16629 + }, + { + "epoch": 1.8029054640069384, + "grad_norm": 2.8793745040893555, + "learning_rate": 8.009574931089513e-06, + "loss": 0.0356, + "step": 16630 + }, + { + "epoch": 1.8030138768430182, + "grad_norm": 5.3428144454956055, + "learning_rate": 8.008849557522125e-06, + "loss": 0.0066, + "step": 16631 + }, + { + "epoch": 1.8031222896790982, + "grad_norm": 2.4334516525268555, + "learning_rate": 8.008124183954737e-06, + "loss": 0.026, + "step": 16632 + }, + { + "epoch": 1.8032307025151777, + "grad_norm": 8.045775413513184, + "learning_rate": 8.00739881038735e-06, + "loss": 0.0521, + "step": 16633 + }, + { + "epoch": 1.8033391153512577, + "grad_norm": 3.7833471298217773, + "learning_rate": 8.006673436819963e-06, + "loss": 0.0216, + "step": 16634 + }, + { + "epoch": 1.8034475281873372, + "grad_norm": 1.6184909343719482, + "learning_rate": 8.005948063252575e-06, + "loss": 0.0171, + "step": 16635 + }, + { + "epoch": 1.8035559410234172, + "grad_norm": 3.087205648422241, + "learning_rate": 8.005222689685187e-06, + "loss": 0.0494, + "step": 16636 + }, + { + "epoch": 1.803664353859497, + "grad_norm": 3.2019600868225098, + "learning_rate": 8.004497316117801e-06, + "loss": 0.0338, + "step": 16637 + }, + { + "epoch": 1.8037727666955767, + "grad_norm": 1.5475190877914429, + "learning_rate": 8.003771942550415e-06, + "loss": 0.0072, + "step": 16638 + }, + { + "epoch": 1.8038811795316565, + "grad_norm": 3.706138849258423, + "learning_rate": 8.003046568983027e-06, + "loss": 0.0187, + "step": 16639 + }, + { + "epoch": 1.8039895923677363, + "grad_norm": 0.45301100611686707, + "learning_rate": 8.00232119541564e-06, + "loss": 0.0036, + "step": 16640 + }, + { + "epoch": 1.8040980052038162, + "grad_norm": 10.020462036132812, + "learning_rate": 8.001595821848253e-06, + "loss": 0.0595, + "step": 16641 + }, + { + "epoch": 1.8042064180398958, + "grad_norm": 2.536989450454712, + "learning_rate": 8.000870448280865e-06, + "loss": 0.0193, + "step": 16642 + }, + { + "epoch": 1.8043148308759758, + "grad_norm": 5.4952006340026855, + "learning_rate": 8.000145074713478e-06, + "loss": 0.0202, + "step": 16643 + }, + { + "epoch": 1.8044232437120555, + "grad_norm": 1.8616223335266113, + "learning_rate": 7.999419701146091e-06, + "loss": 0.0104, + "step": 16644 + }, + { + "epoch": 1.8045316565481353, + "grad_norm": 3.877652883529663, + "learning_rate": 7.998694327578704e-06, + "loss": 0.0316, + "step": 16645 + }, + { + "epoch": 1.804640069384215, + "grad_norm": 2.068338632583618, + "learning_rate": 7.997968954011317e-06, + "loss": 0.0172, + "step": 16646 + }, + { + "epoch": 1.8047484822202948, + "grad_norm": 5.0702948570251465, + "learning_rate": 7.99724358044393e-06, + "loss": 0.0417, + "step": 16647 + }, + { + "epoch": 1.8048568950563748, + "grad_norm": 6.638656139373779, + "learning_rate": 7.996518206876542e-06, + "loss": 0.0367, + "step": 16648 + }, + { + "epoch": 1.8049653078924544, + "grad_norm": 0.8151875734329224, + "learning_rate": 7.995792833309155e-06, + "loss": 0.01, + "step": 16649 + }, + { + "epoch": 1.8050737207285343, + "grad_norm": 2.1947882175445557, + "learning_rate": 7.995067459741768e-06, + "loss": 0.017, + "step": 16650 + }, + { + "epoch": 1.805182133564614, + "grad_norm": 0.9074804782867432, + "learning_rate": 7.99434208617438e-06, + "loss": 0.0071, + "step": 16651 + }, + { + "epoch": 1.8052905464006939, + "grad_norm": 1.2488882541656494, + "learning_rate": 7.993616712606994e-06, + "loss": 0.0089, + "step": 16652 + }, + { + "epoch": 1.8053989592367736, + "grad_norm": 2.904578924179077, + "learning_rate": 7.992891339039606e-06, + "loss": 0.0238, + "step": 16653 + }, + { + "epoch": 1.8055073720728534, + "grad_norm": 3.1511073112487793, + "learning_rate": 7.992165965472218e-06, + "loss": 0.0068, + "step": 16654 + }, + { + "epoch": 1.8056157849089334, + "grad_norm": 3.317011594772339, + "learning_rate": 7.991440591904832e-06, + "loss": 0.0184, + "step": 16655 + }, + { + "epoch": 1.805724197745013, + "grad_norm": 1.7239893674850464, + "learning_rate": 7.990715218337446e-06, + "loss": 0.0061, + "step": 16656 + }, + { + "epoch": 1.805832610581093, + "grad_norm": 4.101641654968262, + "learning_rate": 7.989989844770058e-06, + "loss": 0.0198, + "step": 16657 + }, + { + "epoch": 1.8059410234171724, + "grad_norm": 7.973923683166504, + "learning_rate": 7.98926447120267e-06, + "loss": 0.0313, + "step": 16658 + }, + { + "epoch": 1.8060494362532524, + "grad_norm": 4.644124507904053, + "learning_rate": 7.988539097635282e-06, + "loss": 0.0253, + "step": 16659 + }, + { + "epoch": 1.8061578490893322, + "grad_norm": 2.9801149368286133, + "learning_rate": 7.987813724067896e-06, + "loss": 0.0079, + "step": 16660 + }, + { + "epoch": 1.806266261925412, + "grad_norm": 4.596467971801758, + "learning_rate": 7.987088350500508e-06, + "loss": 0.0181, + "step": 16661 + }, + { + "epoch": 1.8063746747614917, + "grad_norm": 13.879189491271973, + "learning_rate": 7.98636297693312e-06, + "loss": 0.0561, + "step": 16662 + }, + { + "epoch": 1.8064830875975715, + "grad_norm": 1.5565265417099, + "learning_rate": 7.985637603365734e-06, + "loss": 0.029, + "step": 16663 + }, + { + "epoch": 1.8065915004336515, + "grad_norm": 1.0940382480621338, + "learning_rate": 7.984912229798348e-06, + "loss": 0.0076, + "step": 16664 + }, + { + "epoch": 1.806699913269731, + "grad_norm": 4.082326412200928, + "learning_rate": 7.98418685623096e-06, + "loss": 0.0474, + "step": 16665 + }, + { + "epoch": 1.806808326105811, + "grad_norm": 0.06991656869649887, + "learning_rate": 7.983461482663572e-06, + "loss": 0.0006, + "step": 16666 + }, + { + "epoch": 1.8069167389418908, + "grad_norm": 2.120616912841797, + "learning_rate": 7.982736109096186e-06, + "loss": 0.0131, + "step": 16667 + }, + { + "epoch": 1.8070251517779705, + "grad_norm": 1.049953818321228, + "learning_rate": 7.982010735528798e-06, + "loss": 0.0052, + "step": 16668 + }, + { + "epoch": 1.8071335646140503, + "grad_norm": 1.0341280698776245, + "learning_rate": 7.98128536196141e-06, + "loss": 0.0054, + "step": 16669 + }, + { + "epoch": 1.80724197745013, + "grad_norm": 0.4636963903903961, + "learning_rate": 7.980559988394024e-06, + "loss": 0.0027, + "step": 16670 + }, + { + "epoch": 1.80735039028621, + "grad_norm": 1.2382317781448364, + "learning_rate": 7.979834614826636e-06, + "loss": 0.0067, + "step": 16671 + }, + { + "epoch": 1.8074588031222896, + "grad_norm": 5.146089553833008, + "learning_rate": 7.97910924125925e-06, + "loss": 0.0291, + "step": 16672 + }, + { + "epoch": 1.8075672159583696, + "grad_norm": 3.5520615577697754, + "learning_rate": 7.978383867691862e-06, + "loss": 0.0219, + "step": 16673 + }, + { + "epoch": 1.807675628794449, + "grad_norm": 2.028567314147949, + "learning_rate": 7.977658494124474e-06, + "loss": 0.0172, + "step": 16674 + }, + { + "epoch": 1.807784041630529, + "grad_norm": 1.6579136848449707, + "learning_rate": 7.976933120557088e-06, + "loss": 0.011, + "step": 16675 + }, + { + "epoch": 1.8078924544666088, + "grad_norm": 5.4987077713012695, + "learning_rate": 7.9762077469897e-06, + "loss": 0.0092, + "step": 16676 + }, + { + "epoch": 1.8080008673026886, + "grad_norm": 5.435853004455566, + "learning_rate": 7.975482373422312e-06, + "loss": 0.0176, + "step": 16677 + }, + { + "epoch": 1.8081092801387686, + "grad_norm": 3.95680832862854, + "learning_rate": 7.974756999854926e-06, + "loss": 0.0255, + "step": 16678 + }, + { + "epoch": 1.8082176929748481, + "grad_norm": 1.352746605873108, + "learning_rate": 7.974031626287538e-06, + "loss": 0.0326, + "step": 16679 + }, + { + "epoch": 1.8083261058109281, + "grad_norm": 8.933673858642578, + "learning_rate": 7.97330625272015e-06, + "loss": 0.0804, + "step": 16680 + }, + { + "epoch": 1.8084345186470077, + "grad_norm": 5.438947677612305, + "learning_rate": 7.972580879152764e-06, + "loss": 0.011, + "step": 16681 + }, + { + "epoch": 1.8085429314830876, + "grad_norm": 5.207435607910156, + "learning_rate": 7.971855505585378e-06, + "loss": 0.0713, + "step": 16682 + }, + { + "epoch": 1.8086513443191674, + "grad_norm": 5.619300842285156, + "learning_rate": 7.97113013201799e-06, + "loss": 0.0458, + "step": 16683 + }, + { + "epoch": 1.8087597571552472, + "grad_norm": 1.4649312496185303, + "learning_rate": 7.970404758450603e-06, + "loss": 0.0082, + "step": 16684 + }, + { + "epoch": 1.808868169991327, + "grad_norm": 1.3834803104400635, + "learning_rate": 7.969679384883215e-06, + "loss": 0.0133, + "step": 16685 + }, + { + "epoch": 1.8089765828274067, + "grad_norm": 4.4434943199157715, + "learning_rate": 7.968954011315829e-06, + "loss": 0.0362, + "step": 16686 + }, + { + "epoch": 1.8090849956634867, + "grad_norm": 2.251622200012207, + "learning_rate": 7.96822863774844e-06, + "loss": 0.0231, + "step": 16687 + }, + { + "epoch": 1.8091934084995662, + "grad_norm": 0.9647461771965027, + "learning_rate": 7.967503264181053e-06, + "loss": 0.0053, + "step": 16688 + }, + { + "epoch": 1.8093018213356462, + "grad_norm": 1.406980276107788, + "learning_rate": 7.966777890613667e-06, + "loss": 0.0046, + "step": 16689 + }, + { + "epoch": 1.809410234171726, + "grad_norm": 0.7685739398002625, + "learning_rate": 7.96605251704628e-06, + "loss": 0.0083, + "step": 16690 + }, + { + "epoch": 1.8095186470078057, + "grad_norm": 6.760863780975342, + "learning_rate": 7.965327143478893e-06, + "loss": 0.0318, + "step": 16691 + }, + { + "epoch": 1.8096270598438855, + "grad_norm": 1.1230660676956177, + "learning_rate": 7.964601769911505e-06, + "loss": 0.0049, + "step": 16692 + }, + { + "epoch": 1.8097354726799653, + "grad_norm": 1.916284441947937, + "learning_rate": 7.963876396344119e-06, + "loss": 0.0661, + "step": 16693 + }, + { + "epoch": 1.8098438855160452, + "grad_norm": 1.1197506189346313, + "learning_rate": 7.96315102277673e-06, + "loss": 0.0043, + "step": 16694 + }, + { + "epoch": 1.8099522983521248, + "grad_norm": 0.19063952565193176, + "learning_rate": 7.962425649209343e-06, + "loss": 0.0012, + "step": 16695 + }, + { + "epoch": 1.8100607111882048, + "grad_norm": 1.467376708984375, + "learning_rate": 7.961700275641957e-06, + "loss": 0.0137, + "step": 16696 + }, + { + "epoch": 1.8101691240242843, + "grad_norm": 3.3131470680236816, + "learning_rate": 7.960974902074569e-06, + "loss": 0.0419, + "step": 16697 + }, + { + "epoch": 1.8102775368603643, + "grad_norm": 2.8492865562438965, + "learning_rate": 7.960249528507183e-06, + "loss": 0.0499, + "step": 16698 + }, + { + "epoch": 1.810385949696444, + "grad_norm": 1.654439926147461, + "learning_rate": 7.959524154939795e-06, + "loss": 0.0071, + "step": 16699 + }, + { + "epoch": 1.8104943625325238, + "grad_norm": 0.3241313397884369, + "learning_rate": 7.958798781372407e-06, + "loss": 0.0034, + "step": 16700 + }, + { + "epoch": 1.8106027753686036, + "grad_norm": 3.5160417556762695, + "learning_rate": 7.95807340780502e-06, + "loss": 0.0187, + "step": 16701 + }, + { + "epoch": 1.8107111882046834, + "grad_norm": 0.049978021532297134, + "learning_rate": 7.957348034237633e-06, + "loss": 0.0006, + "step": 16702 + }, + { + "epoch": 1.8108196010407633, + "grad_norm": 6.4192728996276855, + "learning_rate": 7.956622660670245e-06, + "loss": 0.0147, + "step": 16703 + }, + { + "epoch": 1.8109280138768429, + "grad_norm": 0.1867048293352127, + "learning_rate": 7.955897287102859e-06, + "loss": 0.0011, + "step": 16704 + }, + { + "epoch": 1.8110364267129229, + "grad_norm": 1.4336268901824951, + "learning_rate": 7.955171913535471e-06, + "loss": 0.007, + "step": 16705 + }, + { + "epoch": 1.8111448395490026, + "grad_norm": 2.814788341522217, + "learning_rate": 7.954446539968083e-06, + "loss": 0.0101, + "step": 16706 + }, + { + "epoch": 1.8112532523850824, + "grad_norm": 1.216360330581665, + "learning_rate": 7.953721166400697e-06, + "loss": 0.0142, + "step": 16707 + }, + { + "epoch": 1.8113616652211622, + "grad_norm": 1.4186736345291138, + "learning_rate": 7.952995792833311e-06, + "loss": 0.0093, + "step": 16708 + }, + { + "epoch": 1.811470078057242, + "grad_norm": 1.527870535850525, + "learning_rate": 7.952270419265923e-06, + "loss": 0.0078, + "step": 16709 + }, + { + "epoch": 1.811578490893322, + "grad_norm": 6.716109275817871, + "learning_rate": 7.951545045698535e-06, + "loss": 0.0485, + "step": 16710 + }, + { + "epoch": 1.8116869037294014, + "grad_norm": 2.7530510425567627, + "learning_rate": 7.950819672131147e-06, + "loss": 0.0312, + "step": 16711 + }, + { + "epoch": 1.8117953165654814, + "grad_norm": 3.890123128890991, + "learning_rate": 7.950094298563761e-06, + "loss": 0.0583, + "step": 16712 + }, + { + "epoch": 1.8119037294015612, + "grad_norm": 4.059305667877197, + "learning_rate": 7.949368924996373e-06, + "loss": 0.0116, + "step": 16713 + }, + { + "epoch": 1.812012142237641, + "grad_norm": 8.140238761901855, + "learning_rate": 7.948643551428985e-06, + "loss": 0.0637, + "step": 16714 + }, + { + "epoch": 1.8121205550737207, + "grad_norm": 5.755715370178223, + "learning_rate": 7.9479181778616e-06, + "loss": 0.029, + "step": 16715 + }, + { + "epoch": 1.8122289679098005, + "grad_norm": 6.193628311157227, + "learning_rate": 7.947192804294213e-06, + "loss": 0.0745, + "step": 16716 + }, + { + "epoch": 1.8123373807458805, + "grad_norm": 3.6182327270507812, + "learning_rate": 7.946467430726825e-06, + "loss": 0.0431, + "step": 16717 + }, + { + "epoch": 1.81244579358196, + "grad_norm": 5.230718612670898, + "learning_rate": 7.945742057159437e-06, + "loss": 0.0184, + "step": 16718 + }, + { + "epoch": 1.81255420641804, + "grad_norm": 7.131375789642334, + "learning_rate": 7.945016683592051e-06, + "loss": 0.0646, + "step": 16719 + }, + { + "epoch": 1.8126626192541195, + "grad_norm": 1.1339552402496338, + "learning_rate": 7.944291310024663e-06, + "loss": 0.0103, + "step": 16720 + }, + { + "epoch": 1.8127710320901995, + "grad_norm": 3.2250592708587646, + "learning_rate": 7.943565936457276e-06, + "loss": 0.0112, + "step": 16721 + }, + { + "epoch": 1.8128794449262793, + "grad_norm": 0.35960653424263, + "learning_rate": 7.94284056288989e-06, + "loss": 0.0012, + "step": 16722 + }, + { + "epoch": 1.812987857762359, + "grad_norm": 2.6757407188415527, + "learning_rate": 7.942115189322502e-06, + "loss": 0.0252, + "step": 16723 + }, + { + "epoch": 1.8130962705984388, + "grad_norm": 0.41398683190345764, + "learning_rate": 7.941389815755115e-06, + "loss": 0.0028, + "step": 16724 + }, + { + "epoch": 1.8132046834345186, + "grad_norm": 1.4540307521820068, + "learning_rate": 7.940664442187728e-06, + "loss": 0.0352, + "step": 16725 + }, + { + "epoch": 1.8133130962705986, + "grad_norm": 7.305990695953369, + "learning_rate": 7.93993906862034e-06, + "loss": 0.0159, + "step": 16726 + }, + { + "epoch": 1.813421509106678, + "grad_norm": 3.2440619468688965, + "learning_rate": 7.939213695052953e-06, + "loss": 0.0223, + "step": 16727 + }, + { + "epoch": 1.813529921942758, + "grad_norm": 4.139712810516357, + "learning_rate": 7.938488321485566e-06, + "loss": 0.0232, + "step": 16728 + }, + { + "epoch": 1.8136383347788378, + "grad_norm": 3.265730619430542, + "learning_rate": 7.937762947918178e-06, + "loss": 0.0357, + "step": 16729 + }, + { + "epoch": 1.8137467476149176, + "grad_norm": 5.1838483810424805, + "learning_rate": 7.937037574350792e-06, + "loss": 0.0534, + "step": 16730 + }, + { + "epoch": 1.8138551604509974, + "grad_norm": 1.968052625656128, + "learning_rate": 7.936312200783404e-06, + "loss": 0.0166, + "step": 16731 + }, + { + "epoch": 1.8139635732870771, + "grad_norm": 6.012704849243164, + "learning_rate": 7.935586827216016e-06, + "loss": 0.0048, + "step": 16732 + }, + { + "epoch": 1.8140719861231571, + "grad_norm": 1.8593837022781372, + "learning_rate": 7.93486145364863e-06, + "loss": 0.0415, + "step": 16733 + }, + { + "epoch": 1.8141803989592367, + "grad_norm": 0.22393952310085297, + "learning_rate": 7.934136080081244e-06, + "loss": 0.0025, + "step": 16734 + }, + { + "epoch": 1.8142888117953166, + "grad_norm": 0.1929185837507248, + "learning_rate": 7.933410706513856e-06, + "loss": 0.002, + "step": 16735 + }, + { + "epoch": 1.8143972246313964, + "grad_norm": 2.442511796951294, + "learning_rate": 7.932685332946468e-06, + "loss": 0.0405, + "step": 16736 + }, + { + "epoch": 1.8145056374674762, + "grad_norm": 1.140223503112793, + "learning_rate": 7.93195995937908e-06, + "loss": 0.0051, + "step": 16737 + }, + { + "epoch": 1.814614050303556, + "grad_norm": 2.1155457496643066, + "learning_rate": 7.931234585811694e-06, + "loss": 0.0443, + "step": 16738 + }, + { + "epoch": 1.8147224631396357, + "grad_norm": 0.27464747428894043, + "learning_rate": 7.930509212244306e-06, + "loss": 0.0024, + "step": 16739 + }, + { + "epoch": 1.8148308759757157, + "grad_norm": 4.552560329437256, + "learning_rate": 7.929783838676918e-06, + "loss": 0.048, + "step": 16740 + }, + { + "epoch": 1.8149392888117952, + "grad_norm": 6.593067646026611, + "learning_rate": 7.929058465109532e-06, + "loss": 0.0227, + "step": 16741 + }, + { + "epoch": 1.8150477016478752, + "grad_norm": 3.029500961303711, + "learning_rate": 7.928333091542146e-06, + "loss": 0.0183, + "step": 16742 + }, + { + "epoch": 1.8151561144839548, + "grad_norm": 0.3149774968624115, + "learning_rate": 7.927607717974758e-06, + "loss": 0.0028, + "step": 16743 + }, + { + "epoch": 1.8152645273200347, + "grad_norm": 0.309287965297699, + "learning_rate": 7.92688234440737e-06, + "loss": 0.0018, + "step": 16744 + }, + { + "epoch": 1.8153729401561145, + "grad_norm": 8.604201316833496, + "learning_rate": 7.926156970839984e-06, + "loss": 0.0427, + "step": 16745 + }, + { + "epoch": 1.8154813529921943, + "grad_norm": 3.4580910205841064, + "learning_rate": 7.925431597272596e-06, + "loss": 0.019, + "step": 16746 + }, + { + "epoch": 1.815589765828274, + "grad_norm": 3.8130507469177246, + "learning_rate": 7.924706223705208e-06, + "loss": 0.0336, + "step": 16747 + }, + { + "epoch": 1.8156981786643538, + "grad_norm": 2.423081159591675, + "learning_rate": 7.923980850137822e-06, + "loss": 0.013, + "step": 16748 + }, + { + "epoch": 1.8158065915004338, + "grad_norm": 2.035799026489258, + "learning_rate": 7.923255476570434e-06, + "loss": 0.0343, + "step": 16749 + }, + { + "epoch": 1.8159150043365133, + "grad_norm": 3.226818323135376, + "learning_rate": 7.922530103003048e-06, + "loss": 0.0453, + "step": 16750 + }, + { + "epoch": 1.8160234171725933, + "grad_norm": 1.202027440071106, + "learning_rate": 7.92180472943566e-06, + "loss": 0.0172, + "step": 16751 + }, + { + "epoch": 1.816131830008673, + "grad_norm": 4.013176441192627, + "learning_rate": 7.921079355868272e-06, + "loss": 0.0215, + "step": 16752 + }, + { + "epoch": 1.8162402428447528, + "grad_norm": 0.7996282577514648, + "learning_rate": 7.920353982300886e-06, + "loss": 0.0038, + "step": 16753 + }, + { + "epoch": 1.8163486556808326, + "grad_norm": 3.835674285888672, + "learning_rate": 7.919628608733498e-06, + "loss": 0.0109, + "step": 16754 + }, + { + "epoch": 1.8164570685169124, + "grad_norm": 1.415818214416504, + "learning_rate": 7.91890323516611e-06, + "loss": 0.0054, + "step": 16755 + }, + { + "epoch": 1.8165654813529923, + "grad_norm": 1.541670322418213, + "learning_rate": 7.918177861598724e-06, + "loss": 0.0429, + "step": 16756 + }, + { + "epoch": 1.8166738941890719, + "grad_norm": 4.24556303024292, + "learning_rate": 7.917452488031336e-06, + "loss": 0.0467, + "step": 16757 + }, + { + "epoch": 1.8167823070251519, + "grad_norm": 4.049663543701172, + "learning_rate": 7.916727114463949e-06, + "loss": 0.047, + "step": 16758 + }, + { + "epoch": 1.8168907198612314, + "grad_norm": 1.240966558456421, + "learning_rate": 7.916001740896562e-06, + "loss": 0.0091, + "step": 16759 + }, + { + "epoch": 1.8169991326973114, + "grad_norm": 2.831259250640869, + "learning_rate": 7.915276367329176e-06, + "loss": 0.0108, + "step": 16760 + }, + { + "epoch": 1.8171075455333912, + "grad_norm": 1.2151451110839844, + "learning_rate": 7.914550993761788e-06, + "loss": 0.0065, + "step": 16761 + }, + { + "epoch": 1.817215958369471, + "grad_norm": 0.27166157960891724, + "learning_rate": 7.9138256201944e-06, + "loss": 0.0029, + "step": 16762 + }, + { + "epoch": 1.817324371205551, + "grad_norm": 0.09778092056512833, + "learning_rate": 7.913100246627013e-06, + "loss": 0.0016, + "step": 16763 + }, + { + "epoch": 1.8174327840416304, + "grad_norm": 1.498037576675415, + "learning_rate": 7.912374873059627e-06, + "loss": 0.007, + "step": 16764 + }, + { + "epoch": 1.8175411968777104, + "grad_norm": 6.701542377471924, + "learning_rate": 7.911649499492239e-06, + "loss": 0.0533, + "step": 16765 + }, + { + "epoch": 1.81764960971379, + "grad_norm": 1.646487832069397, + "learning_rate": 7.91092412592485e-06, + "loss": 0.0106, + "step": 16766 + }, + { + "epoch": 1.81775802254987, + "grad_norm": 0.08623085916042328, + "learning_rate": 7.910198752357465e-06, + "loss": 0.0013, + "step": 16767 + }, + { + "epoch": 1.8178664353859497, + "grad_norm": 0.28099045157432556, + "learning_rate": 7.909473378790078e-06, + "loss": 0.0021, + "step": 16768 + }, + { + "epoch": 1.8179748482220295, + "grad_norm": 3.4071333408355713, + "learning_rate": 7.90874800522269e-06, + "loss": 0.0214, + "step": 16769 + }, + { + "epoch": 1.8180832610581092, + "grad_norm": 2.773576259613037, + "learning_rate": 7.908022631655303e-06, + "loss": 0.0232, + "step": 16770 + }, + { + "epoch": 1.818191673894189, + "grad_norm": 4.6704325675964355, + "learning_rate": 7.907297258087917e-06, + "loss": 0.0152, + "step": 16771 + }, + { + "epoch": 1.818300086730269, + "grad_norm": 1.1458066701889038, + "learning_rate": 7.906571884520529e-06, + "loss": 0.0603, + "step": 16772 + }, + { + "epoch": 1.8184084995663485, + "grad_norm": 5.282043933868408, + "learning_rate": 7.905846510953141e-06, + "loss": 0.0191, + "step": 16773 + }, + { + "epoch": 1.8185169124024285, + "grad_norm": 0.26037248969078064, + "learning_rate": 7.905121137385755e-06, + "loss": 0.0025, + "step": 16774 + }, + { + "epoch": 1.8186253252385083, + "grad_norm": 0.9299569725990295, + "learning_rate": 7.904395763818367e-06, + "loss": 0.0247, + "step": 16775 + }, + { + "epoch": 1.818733738074588, + "grad_norm": 2.8093128204345703, + "learning_rate": 7.90367039025098e-06, + "loss": 0.0252, + "step": 16776 + }, + { + "epoch": 1.8188421509106678, + "grad_norm": 2.7386314868927, + "learning_rate": 7.902945016683593e-06, + "loss": 0.0302, + "step": 16777 + }, + { + "epoch": 1.8189505637467476, + "grad_norm": 3.9675981998443604, + "learning_rate": 7.902219643116205e-06, + "loss": 0.0216, + "step": 16778 + }, + { + "epoch": 1.8190589765828276, + "grad_norm": 0.9117150902748108, + "learning_rate": 7.901494269548819e-06, + "loss": 0.0052, + "step": 16779 + }, + { + "epoch": 1.819167389418907, + "grad_norm": 3.258571147918701, + "learning_rate": 7.900768895981431e-06, + "loss": 0.014, + "step": 16780 + }, + { + "epoch": 1.819275802254987, + "grad_norm": 1.1497143507003784, + "learning_rate": 7.900043522414043e-06, + "loss": 0.0049, + "step": 16781 + }, + { + "epoch": 1.8193842150910666, + "grad_norm": 2.2095587253570557, + "learning_rate": 7.899318148846657e-06, + "loss": 0.0232, + "step": 16782 + }, + { + "epoch": 1.8194926279271466, + "grad_norm": 2.471635341644287, + "learning_rate": 7.898592775279269e-06, + "loss": 0.0386, + "step": 16783 + }, + { + "epoch": 1.8196010407632264, + "grad_norm": 2.5582382678985596, + "learning_rate": 7.897867401711883e-06, + "loss": 0.0096, + "step": 16784 + }, + { + "epoch": 1.8197094535993061, + "grad_norm": 1.8183143138885498, + "learning_rate": 7.897142028144495e-06, + "loss": 0.0142, + "step": 16785 + }, + { + "epoch": 1.819817866435386, + "grad_norm": 3.075176239013672, + "learning_rate": 7.896416654577109e-06, + "loss": 0.0375, + "step": 16786 + }, + { + "epoch": 1.8199262792714657, + "grad_norm": 0.7208966016769409, + "learning_rate": 7.895691281009721e-06, + "loss": 0.0069, + "step": 16787 + }, + { + "epoch": 1.8200346921075456, + "grad_norm": 3.9399471282958984, + "learning_rate": 7.894965907442333e-06, + "loss": 0.0426, + "step": 16788 + }, + { + "epoch": 1.8201431049436252, + "grad_norm": 5.400341510772705, + "learning_rate": 7.894240533874945e-06, + "loss": 0.0369, + "step": 16789 + }, + { + "epoch": 1.8202515177797052, + "grad_norm": 3.0667121410369873, + "learning_rate": 7.89351516030756e-06, + "loss": 0.0242, + "step": 16790 + }, + { + "epoch": 1.820359930615785, + "grad_norm": 2.4307048320770264, + "learning_rate": 7.892789786740171e-06, + "loss": 0.0506, + "step": 16791 + }, + { + "epoch": 1.8204683434518647, + "grad_norm": 3.9447247982025146, + "learning_rate": 7.892064413172783e-06, + "loss": 0.0423, + "step": 16792 + }, + { + "epoch": 1.8205767562879445, + "grad_norm": 1.799468994140625, + "learning_rate": 7.891339039605397e-06, + "loss": 0.0062, + "step": 16793 + }, + { + "epoch": 1.8206851691240242, + "grad_norm": 3.0484459400177, + "learning_rate": 7.890613666038011e-06, + "loss": 0.056, + "step": 16794 + }, + { + "epoch": 1.8207935819601042, + "grad_norm": 5.716277599334717, + "learning_rate": 7.889888292470623e-06, + "loss": 0.0601, + "step": 16795 + }, + { + "epoch": 1.8209019947961838, + "grad_norm": 3.012824058532715, + "learning_rate": 7.889162918903235e-06, + "loss": 0.0242, + "step": 16796 + }, + { + "epoch": 1.8210104076322637, + "grad_norm": 1.8453787565231323, + "learning_rate": 7.88843754533585e-06, + "loss": 0.0182, + "step": 16797 + }, + { + "epoch": 1.8211188204683435, + "grad_norm": 0.24206513166427612, + "learning_rate": 7.887712171768461e-06, + "loss": 0.0016, + "step": 16798 + }, + { + "epoch": 1.8212272333044233, + "grad_norm": 1.4819483757019043, + "learning_rate": 7.886986798201074e-06, + "loss": 0.0198, + "step": 16799 + }, + { + "epoch": 1.821335646140503, + "grad_norm": 0.42455360293388367, + "learning_rate": 7.886261424633687e-06, + "loss": 0.0053, + "step": 16800 + }, + { + "epoch": 1.8214440589765828, + "grad_norm": 0.8413497805595398, + "learning_rate": 7.8855360510663e-06, + "loss": 0.0608, + "step": 16801 + }, + { + "epoch": 1.8215524718126628, + "grad_norm": 2.2313506603240967, + "learning_rate": 7.884810677498913e-06, + "loss": 0.0056, + "step": 16802 + }, + { + "epoch": 1.8216608846487423, + "grad_norm": 1.9889330863952637, + "learning_rate": 7.884085303931526e-06, + "loss": 0.0274, + "step": 16803 + }, + { + "epoch": 1.8217692974848223, + "grad_norm": 0.5444937944412231, + "learning_rate": 7.883359930364138e-06, + "loss": 0.0045, + "step": 16804 + }, + { + "epoch": 1.8218777103209018, + "grad_norm": 0.5054301023483276, + "learning_rate": 7.882634556796752e-06, + "loss": 0.0025, + "step": 16805 + }, + { + "epoch": 1.8219861231569818, + "grad_norm": 3.680710792541504, + "learning_rate": 7.881909183229364e-06, + "loss": 0.031, + "step": 16806 + }, + { + "epoch": 1.8220945359930616, + "grad_norm": 3.3517842292785645, + "learning_rate": 7.881183809661976e-06, + "loss": 0.0439, + "step": 16807 + }, + { + "epoch": 1.8222029488291414, + "grad_norm": 0.7070590853691101, + "learning_rate": 7.88045843609459e-06, + "loss": 0.0032, + "step": 16808 + }, + { + "epoch": 1.8223113616652211, + "grad_norm": 3.738896131515503, + "learning_rate": 7.879733062527202e-06, + "loss": 0.0473, + "step": 16809 + }, + { + "epoch": 1.8224197745013009, + "grad_norm": 2.3437306880950928, + "learning_rate": 7.879007688959816e-06, + "loss": 0.0415, + "step": 16810 + }, + { + "epoch": 1.8225281873373809, + "grad_norm": 1.4407638311386108, + "learning_rate": 7.878282315392428e-06, + "loss": 0.0083, + "step": 16811 + }, + { + "epoch": 1.8226366001734604, + "grad_norm": 3.7476754188537598, + "learning_rate": 7.877556941825042e-06, + "loss": 0.0117, + "step": 16812 + }, + { + "epoch": 1.8227450130095404, + "grad_norm": 5.196433067321777, + "learning_rate": 7.876831568257654e-06, + "loss": 0.0341, + "step": 16813 + }, + { + "epoch": 1.8228534258456202, + "grad_norm": 1.1441632509231567, + "learning_rate": 7.876106194690266e-06, + "loss": 0.0109, + "step": 16814 + }, + { + "epoch": 1.8229618386817, + "grad_norm": 5.135570049285889, + "learning_rate": 7.875380821122878e-06, + "loss": 0.0227, + "step": 16815 + }, + { + "epoch": 1.8230702515177797, + "grad_norm": 0.5016562342643738, + "learning_rate": 7.874655447555492e-06, + "loss": 0.0033, + "step": 16816 + }, + { + "epoch": 1.8231786643538594, + "grad_norm": 1.6605417728424072, + "learning_rate": 7.873930073988104e-06, + "loss": 0.0365, + "step": 16817 + }, + { + "epoch": 1.8232870771899394, + "grad_norm": 0.14223583042621613, + "learning_rate": 7.873204700420716e-06, + "loss": 0.0015, + "step": 16818 + }, + { + "epoch": 1.823395490026019, + "grad_norm": 2.3095803260803223, + "learning_rate": 7.87247932685333e-06, + "loss": 0.0281, + "step": 16819 + }, + { + "epoch": 1.823503902862099, + "grad_norm": 5.976850986480713, + "learning_rate": 7.871753953285944e-06, + "loss": 0.0401, + "step": 16820 + }, + { + "epoch": 1.8236123156981785, + "grad_norm": 1.6334353685379028, + "learning_rate": 7.871028579718556e-06, + "loss": 0.0229, + "step": 16821 + }, + { + "epoch": 1.8237207285342585, + "grad_norm": 3.210698127746582, + "learning_rate": 7.870303206151168e-06, + "loss": 0.0489, + "step": 16822 + }, + { + "epoch": 1.8238291413703382, + "grad_norm": 4.088848114013672, + "learning_rate": 7.869577832583782e-06, + "loss": 0.0184, + "step": 16823 + }, + { + "epoch": 1.823937554206418, + "grad_norm": 3.0322346687316895, + "learning_rate": 7.868852459016394e-06, + "loss": 0.0426, + "step": 16824 + }, + { + "epoch": 1.824045967042498, + "grad_norm": 7.078643321990967, + "learning_rate": 7.868127085449006e-06, + "loss": 0.0404, + "step": 16825 + }, + { + "epoch": 1.8241543798785775, + "grad_norm": 3.0068228244781494, + "learning_rate": 7.86740171188162e-06, + "loss": 0.0328, + "step": 16826 + }, + { + "epoch": 1.8242627927146575, + "grad_norm": 5.591822624206543, + "learning_rate": 7.866676338314232e-06, + "loss": 0.0387, + "step": 16827 + }, + { + "epoch": 1.824371205550737, + "grad_norm": 0.6194716691970825, + "learning_rate": 7.865950964746846e-06, + "loss": 0.0146, + "step": 16828 + }, + { + "epoch": 1.824479618386817, + "grad_norm": 0.33540108799934387, + "learning_rate": 7.865225591179458e-06, + "loss": 0.0036, + "step": 16829 + }, + { + "epoch": 1.8245880312228968, + "grad_norm": 8.2205228805542, + "learning_rate": 7.86450021761207e-06, + "loss": 0.0094, + "step": 16830 + }, + { + "epoch": 1.8246964440589766, + "grad_norm": 2.087576389312744, + "learning_rate": 7.863774844044684e-06, + "loss": 0.0085, + "step": 16831 + }, + { + "epoch": 1.8248048568950563, + "grad_norm": 1.7426990270614624, + "learning_rate": 7.863049470477296e-06, + "loss": 0.0068, + "step": 16832 + }, + { + "epoch": 1.824913269731136, + "grad_norm": 4.248772621154785, + "learning_rate": 7.862324096909908e-06, + "loss": 0.0304, + "step": 16833 + }, + { + "epoch": 1.825021682567216, + "grad_norm": 7.263134956359863, + "learning_rate": 7.861598723342522e-06, + "loss": 0.0369, + "step": 16834 + }, + { + "epoch": 1.8251300954032956, + "grad_norm": 0.9484822750091553, + "learning_rate": 7.860873349775134e-06, + "loss": 0.0051, + "step": 16835 + }, + { + "epoch": 1.8252385082393756, + "grad_norm": 3.2886009216308594, + "learning_rate": 7.860147976207748e-06, + "loss": 0.0516, + "step": 16836 + }, + { + "epoch": 1.8253469210754554, + "grad_norm": 1.4189766645431519, + "learning_rate": 7.85942260264036e-06, + "loss": 0.013, + "step": 16837 + }, + { + "epoch": 1.8254553339115351, + "grad_norm": 5.443880558013916, + "learning_rate": 7.858697229072974e-06, + "loss": 0.0227, + "step": 16838 + }, + { + "epoch": 1.825563746747615, + "grad_norm": 0.5090773701667786, + "learning_rate": 7.857971855505586e-06, + "loss": 0.0023, + "step": 16839 + }, + { + "epoch": 1.8256721595836947, + "grad_norm": 0.45321422815322876, + "learning_rate": 7.857246481938199e-06, + "loss": 0.004, + "step": 16840 + }, + { + "epoch": 1.8257805724197746, + "grad_norm": 5.799793243408203, + "learning_rate": 7.85652110837081e-06, + "loss": 0.0217, + "step": 16841 + }, + { + "epoch": 1.8258889852558542, + "grad_norm": 1.3167732954025269, + "learning_rate": 7.855795734803425e-06, + "loss": 0.0096, + "step": 16842 + }, + { + "epoch": 1.8259973980919342, + "grad_norm": 0.2791701853275299, + "learning_rate": 7.855070361236037e-06, + "loss": 0.003, + "step": 16843 + }, + { + "epoch": 1.8261058109280137, + "grad_norm": 3.265789031982422, + "learning_rate": 7.854344987668649e-06, + "loss": 0.0287, + "step": 16844 + }, + { + "epoch": 1.8262142237640937, + "grad_norm": 2.349285364151001, + "learning_rate": 7.853619614101263e-06, + "loss": 0.0258, + "step": 16845 + }, + { + "epoch": 1.8263226366001735, + "grad_norm": 1.5114967823028564, + "learning_rate": 7.852894240533876e-06, + "loss": 0.0103, + "step": 16846 + }, + { + "epoch": 1.8264310494362532, + "grad_norm": 0.3163653612136841, + "learning_rate": 7.852168866966489e-06, + "loss": 0.0031, + "step": 16847 + }, + { + "epoch": 1.826539462272333, + "grad_norm": 0.247807577252388, + "learning_rate": 7.8514434933991e-06, + "loss": 0.0023, + "step": 16848 + }, + { + "epoch": 1.8266478751084128, + "grad_norm": 3.109656810760498, + "learning_rate": 7.850718119831715e-06, + "loss": 0.0619, + "step": 16849 + }, + { + "epoch": 1.8267562879444927, + "grad_norm": 2.3557240962982178, + "learning_rate": 7.849992746264327e-06, + "loss": 0.0313, + "step": 16850 + }, + { + "epoch": 1.8268647007805723, + "grad_norm": 5.1390509605407715, + "learning_rate": 7.849267372696939e-06, + "loss": 0.0417, + "step": 16851 + }, + { + "epoch": 1.8269731136166523, + "grad_norm": 2.1239242553710938, + "learning_rate": 7.848541999129553e-06, + "loss": 0.0282, + "step": 16852 + }, + { + "epoch": 1.827081526452732, + "grad_norm": 3.252464771270752, + "learning_rate": 7.847816625562165e-06, + "loss": 0.0423, + "step": 16853 + }, + { + "epoch": 1.8271899392888118, + "grad_norm": 1.7350847721099854, + "learning_rate": 7.847091251994779e-06, + "loss": 0.03, + "step": 16854 + }, + { + "epoch": 1.8272983521248916, + "grad_norm": 3.6335747241973877, + "learning_rate": 7.846365878427391e-06, + "loss": 0.046, + "step": 16855 + }, + { + "epoch": 1.8274067649609713, + "grad_norm": 2.292689085006714, + "learning_rate": 7.845640504860003e-06, + "loss": 0.0161, + "step": 16856 + }, + { + "epoch": 1.8275151777970513, + "grad_norm": 3.245241403579712, + "learning_rate": 7.844915131292617e-06, + "loss": 0.0265, + "step": 16857 + }, + { + "epoch": 1.8276235906331308, + "grad_norm": 5.420315265655518, + "learning_rate": 7.844189757725229e-06, + "loss": 0.0256, + "step": 16858 + }, + { + "epoch": 1.8277320034692108, + "grad_norm": 4.5335564613342285, + "learning_rate": 7.843464384157841e-06, + "loss": 0.0256, + "step": 16859 + }, + { + "epoch": 1.8278404163052906, + "grad_norm": 2.0931196212768555, + "learning_rate": 7.842739010590455e-06, + "loss": 0.0123, + "step": 16860 + }, + { + "epoch": 1.8279488291413704, + "grad_norm": 4.513577938079834, + "learning_rate": 7.842013637023067e-06, + "loss": 0.0375, + "step": 16861 + }, + { + "epoch": 1.8280572419774501, + "grad_norm": 1.7006466388702393, + "learning_rate": 7.841288263455681e-06, + "loss": 0.0246, + "step": 16862 + }, + { + "epoch": 1.8281656548135299, + "grad_norm": 1.4060866832733154, + "learning_rate": 7.840562889888293e-06, + "loss": 0.0106, + "step": 16863 + }, + { + "epoch": 1.8282740676496099, + "grad_norm": 3.40781307220459, + "learning_rate": 7.839837516320907e-06, + "loss": 0.019, + "step": 16864 + }, + { + "epoch": 1.8283824804856894, + "grad_norm": 2.92976450920105, + "learning_rate": 7.839112142753519e-06, + "loss": 0.0613, + "step": 16865 + }, + { + "epoch": 1.8284908933217694, + "grad_norm": 1.3481382131576538, + "learning_rate": 7.838386769186131e-06, + "loss": 0.0085, + "step": 16866 + }, + { + "epoch": 1.828599306157849, + "grad_norm": 3.792670965194702, + "learning_rate": 7.837661395618743e-06, + "loss": 0.0305, + "step": 16867 + }, + { + "epoch": 1.828707718993929, + "grad_norm": 2.6044156551361084, + "learning_rate": 7.836936022051357e-06, + "loss": 0.0311, + "step": 16868 + }, + { + "epoch": 1.8288161318300087, + "grad_norm": 6.450573921203613, + "learning_rate": 7.83621064848397e-06, + "loss": 0.0141, + "step": 16869 + }, + { + "epoch": 1.8289245446660884, + "grad_norm": 1.6135151386260986, + "learning_rate": 7.835485274916581e-06, + "loss": 0.01, + "step": 16870 + }, + { + "epoch": 1.8290329575021682, + "grad_norm": 5.279016017913818, + "learning_rate": 7.834759901349195e-06, + "loss": 0.0412, + "step": 16871 + }, + { + "epoch": 1.829141370338248, + "grad_norm": 4.683804512023926, + "learning_rate": 7.83403452778181e-06, + "loss": 0.0222, + "step": 16872 + }, + { + "epoch": 1.829249783174328, + "grad_norm": 4.037842273712158, + "learning_rate": 7.833309154214421e-06, + "loss": 0.0081, + "step": 16873 + }, + { + "epoch": 1.8293581960104075, + "grad_norm": 2.7916433811187744, + "learning_rate": 7.832583780647033e-06, + "loss": 0.0276, + "step": 16874 + }, + { + "epoch": 1.8294666088464875, + "grad_norm": 1.0852034091949463, + "learning_rate": 7.831858407079647e-06, + "loss": 0.0058, + "step": 16875 + }, + { + "epoch": 1.8295750216825672, + "grad_norm": 3.624608278274536, + "learning_rate": 7.83113303351226e-06, + "loss": 0.0194, + "step": 16876 + }, + { + "epoch": 1.829683434518647, + "grad_norm": 0.15283821523189545, + "learning_rate": 7.830407659944872e-06, + "loss": 0.0014, + "step": 16877 + }, + { + "epoch": 1.8297918473547268, + "grad_norm": 5.990644931793213, + "learning_rate": 7.829682286377485e-06, + "loss": 0.0215, + "step": 16878 + }, + { + "epoch": 1.8299002601908065, + "grad_norm": 3.28371000289917, + "learning_rate": 7.828956912810098e-06, + "loss": 0.0289, + "step": 16879 + }, + { + "epoch": 1.8300086730268865, + "grad_norm": 3.101146697998047, + "learning_rate": 7.828231539242711e-06, + "loss": 0.0834, + "step": 16880 + }, + { + "epoch": 1.830117085862966, + "grad_norm": 2.0069823265075684, + "learning_rate": 7.827506165675324e-06, + "loss": 0.0109, + "step": 16881 + }, + { + "epoch": 1.830225498699046, + "grad_norm": 0.24347232282161713, + "learning_rate": 7.826780792107936e-06, + "loss": 0.0026, + "step": 16882 + }, + { + "epoch": 1.8303339115351258, + "grad_norm": 4.539474964141846, + "learning_rate": 7.82605541854055e-06, + "loss": 0.0172, + "step": 16883 + }, + { + "epoch": 1.8304423243712056, + "grad_norm": 0.06742426753044128, + "learning_rate": 7.825330044973162e-06, + "loss": 0.0008, + "step": 16884 + }, + { + "epoch": 1.8305507372072853, + "grad_norm": 2.873426914215088, + "learning_rate": 7.824604671405774e-06, + "loss": 0.0135, + "step": 16885 + }, + { + "epoch": 1.830659150043365, + "grad_norm": 2.3858540058135986, + "learning_rate": 7.823879297838388e-06, + "loss": 0.0154, + "step": 16886 + }, + { + "epoch": 1.830767562879445, + "grad_norm": 2.0615437030792236, + "learning_rate": 7.823153924271e-06, + "loss": 0.022, + "step": 16887 + }, + { + "epoch": 1.8308759757155246, + "grad_norm": 1.2577694654464722, + "learning_rate": 7.822428550703614e-06, + "loss": 0.0076, + "step": 16888 + }, + { + "epoch": 1.8309843885516046, + "grad_norm": 3.6421351432800293, + "learning_rate": 7.821703177136226e-06, + "loss": 0.0154, + "step": 16889 + }, + { + "epoch": 1.8310928013876842, + "grad_norm": 2.5502748489379883, + "learning_rate": 7.82097780356884e-06, + "loss": 0.0064, + "step": 16890 + }, + { + "epoch": 1.8312012142237641, + "grad_norm": 0.6208111047744751, + "learning_rate": 7.820252430001452e-06, + "loss": 0.0043, + "step": 16891 + }, + { + "epoch": 1.831309627059844, + "grad_norm": 2.317780017852783, + "learning_rate": 7.819527056434064e-06, + "loss": 0.0291, + "step": 16892 + }, + { + "epoch": 1.8314180398959237, + "grad_norm": 2.1061251163482666, + "learning_rate": 7.818801682866676e-06, + "loss": 0.0084, + "step": 16893 + }, + { + "epoch": 1.8315264527320034, + "grad_norm": 1.2695902585983276, + "learning_rate": 7.81807630929929e-06, + "loss": 0.0195, + "step": 16894 + }, + { + "epoch": 1.8316348655680832, + "grad_norm": 2.330512523651123, + "learning_rate": 7.817350935731902e-06, + "loss": 0.073, + "step": 16895 + }, + { + "epoch": 1.8317432784041632, + "grad_norm": 0.940690815448761, + "learning_rate": 7.816625562164514e-06, + "loss": 0.0046, + "step": 16896 + }, + { + "epoch": 1.8318516912402427, + "grad_norm": 0.268164724111557, + "learning_rate": 7.815900188597128e-06, + "loss": 0.0031, + "step": 16897 + }, + { + "epoch": 1.8319601040763227, + "grad_norm": 3.0309298038482666, + "learning_rate": 7.815174815029742e-06, + "loss": 0.0222, + "step": 16898 + }, + { + "epoch": 1.8320685169124025, + "grad_norm": 5.194852828979492, + "learning_rate": 7.814449441462354e-06, + "loss": 0.0152, + "step": 16899 + }, + { + "epoch": 1.8321769297484822, + "grad_norm": 4.075046062469482, + "learning_rate": 7.813724067894966e-06, + "loss": 0.0258, + "step": 16900 + }, + { + "epoch": 1.832285342584562, + "grad_norm": 10.540878295898438, + "learning_rate": 7.81299869432758e-06, + "loss": 0.0344, + "step": 16901 + }, + { + "epoch": 1.8323937554206418, + "grad_norm": 3.0596835613250732, + "learning_rate": 7.812273320760192e-06, + "loss": 0.0264, + "step": 16902 + }, + { + "epoch": 1.8325021682567217, + "grad_norm": 1.8785648345947266, + "learning_rate": 7.811547947192804e-06, + "loss": 0.0282, + "step": 16903 + }, + { + "epoch": 1.8326105810928013, + "grad_norm": 1.65098237991333, + "learning_rate": 7.810822573625418e-06, + "loss": 0.0168, + "step": 16904 + }, + { + "epoch": 1.8327189939288813, + "grad_norm": 2.189908266067505, + "learning_rate": 7.810097200058032e-06, + "loss": 0.0245, + "step": 16905 + }, + { + "epoch": 1.8328274067649608, + "grad_norm": 4.859023571014404, + "learning_rate": 7.809371826490644e-06, + "loss": 0.0487, + "step": 16906 + }, + { + "epoch": 1.8329358196010408, + "grad_norm": 0.8359833359718323, + "learning_rate": 7.808646452923256e-06, + "loss": 0.0115, + "step": 16907 + }, + { + "epoch": 1.8330442324371206, + "grad_norm": 3.6654820442199707, + "learning_rate": 7.807921079355868e-06, + "loss": 0.0315, + "step": 16908 + }, + { + "epoch": 1.8331526452732003, + "grad_norm": 0.815884530544281, + "learning_rate": 7.807195705788482e-06, + "loss": 0.0024, + "step": 16909 + }, + { + "epoch": 1.83326105810928, + "grad_norm": 1.1830029487609863, + "learning_rate": 7.806470332221094e-06, + "loss": 0.003, + "step": 16910 + }, + { + "epoch": 1.8333694709453598, + "grad_norm": 1.8849260807037354, + "learning_rate": 7.805744958653706e-06, + "loss": 0.0154, + "step": 16911 + }, + { + "epoch": 1.8334778837814398, + "grad_norm": 1.396005630493164, + "learning_rate": 7.80501958508632e-06, + "loss": 0.0037, + "step": 16912 + }, + { + "epoch": 1.8335862966175194, + "grad_norm": 2.9665050506591797, + "learning_rate": 7.804294211518932e-06, + "loss": 0.0645, + "step": 16913 + }, + { + "epoch": 1.8336947094535994, + "grad_norm": 2.998861789703369, + "learning_rate": 7.803568837951546e-06, + "loss": 0.0582, + "step": 16914 + }, + { + "epoch": 1.8338031222896791, + "grad_norm": 1.2020257711410522, + "learning_rate": 7.802843464384158e-06, + "loss": 0.0053, + "step": 16915 + }, + { + "epoch": 1.8339115351257589, + "grad_norm": 2.965211868286133, + "learning_rate": 7.802118090816772e-06, + "loss": 0.0105, + "step": 16916 + }, + { + "epoch": 1.8340199479618386, + "grad_norm": 2.7615556716918945, + "learning_rate": 7.801392717249384e-06, + "loss": 0.0534, + "step": 16917 + }, + { + "epoch": 1.8341283607979184, + "grad_norm": 1.05570650100708, + "learning_rate": 7.800667343681997e-06, + "loss": 0.01, + "step": 16918 + }, + { + "epoch": 1.8342367736339984, + "grad_norm": 0.4318801760673523, + "learning_rate": 7.799941970114609e-06, + "loss": 0.0069, + "step": 16919 + }, + { + "epoch": 1.834345186470078, + "grad_norm": 3.1553385257720947, + "learning_rate": 7.799216596547223e-06, + "loss": 0.0147, + "step": 16920 + }, + { + "epoch": 1.834453599306158, + "grad_norm": 0.4238124489784241, + "learning_rate": 7.798491222979835e-06, + "loss": 0.0034, + "step": 16921 + }, + { + "epoch": 1.8345620121422377, + "grad_norm": 1.5160900354385376, + "learning_rate": 7.797765849412447e-06, + "loss": 0.0278, + "step": 16922 + }, + { + "epoch": 1.8346704249783174, + "grad_norm": 1.5214643478393555, + "learning_rate": 7.79704047584506e-06, + "loss": 0.0289, + "step": 16923 + }, + { + "epoch": 1.8347788378143972, + "grad_norm": 1.674204707145691, + "learning_rate": 7.796315102277674e-06, + "loss": 0.0176, + "step": 16924 + }, + { + "epoch": 1.834887250650477, + "grad_norm": 1.3374613523483276, + "learning_rate": 7.795589728710287e-06, + "loss": 0.0053, + "step": 16925 + }, + { + "epoch": 1.834995663486557, + "grad_norm": 1.8968772888183594, + "learning_rate": 7.794864355142899e-06, + "loss": 0.0146, + "step": 16926 + }, + { + "epoch": 1.8351040763226365, + "grad_norm": 4.635206699371338, + "learning_rate": 7.794138981575513e-06, + "loss": 0.0165, + "step": 16927 + }, + { + "epoch": 1.8352124891587165, + "grad_norm": 10.107677459716797, + "learning_rate": 7.793413608008125e-06, + "loss": 0.0474, + "step": 16928 + }, + { + "epoch": 1.835320901994796, + "grad_norm": 0.23496006429195404, + "learning_rate": 7.792688234440737e-06, + "loss": 0.002, + "step": 16929 + }, + { + "epoch": 1.835429314830876, + "grad_norm": 7.749720573425293, + "learning_rate": 7.79196286087335e-06, + "loss": 0.0161, + "step": 16930 + }, + { + "epoch": 1.8355377276669558, + "grad_norm": 0.38652104139328003, + "learning_rate": 7.791237487305965e-06, + "loss": 0.0036, + "step": 16931 + }, + { + "epoch": 1.8356461405030355, + "grad_norm": 3.355914831161499, + "learning_rate": 7.790512113738577e-06, + "loss": 0.0196, + "step": 16932 + }, + { + "epoch": 1.8357545533391153, + "grad_norm": 1.756614089012146, + "learning_rate": 7.789786740171189e-06, + "loss": 0.0054, + "step": 16933 + }, + { + "epoch": 1.835862966175195, + "grad_norm": 1.3630690574645996, + "learning_rate": 7.789061366603801e-06, + "loss": 0.0234, + "step": 16934 + }, + { + "epoch": 1.835971379011275, + "grad_norm": 2.3831064701080322, + "learning_rate": 7.788335993036415e-06, + "loss": 0.0319, + "step": 16935 + }, + { + "epoch": 1.8360797918473546, + "grad_norm": 0.8492333292961121, + "learning_rate": 7.787610619469027e-06, + "loss": 0.0041, + "step": 16936 + }, + { + "epoch": 1.8361882046834346, + "grad_norm": 4.055136203765869, + "learning_rate": 7.786885245901639e-06, + "loss": 0.0169, + "step": 16937 + }, + { + "epoch": 1.8362966175195143, + "grad_norm": 0.47108563780784607, + "learning_rate": 7.786159872334253e-06, + "loss": 0.0034, + "step": 16938 + }, + { + "epoch": 1.836405030355594, + "grad_norm": 0.5909457206726074, + "learning_rate": 7.785434498766865e-06, + "loss": 0.0071, + "step": 16939 + }, + { + "epoch": 1.8365134431916739, + "grad_norm": 1.7574169635772705, + "learning_rate": 7.784709125199479e-06, + "loss": 0.0072, + "step": 16940 + }, + { + "epoch": 1.8366218560277536, + "grad_norm": 3.1246695518493652, + "learning_rate": 7.783983751632091e-06, + "loss": 0.0563, + "step": 16941 + }, + { + "epoch": 1.8367302688638336, + "grad_norm": 3.2345097064971924, + "learning_rate": 7.783258378064705e-06, + "loss": 0.0125, + "step": 16942 + }, + { + "epoch": 1.8368386816999132, + "grad_norm": 1.1021645069122314, + "learning_rate": 7.782533004497317e-06, + "loss": 0.0354, + "step": 16943 + }, + { + "epoch": 1.8369470945359931, + "grad_norm": 8.821076393127441, + "learning_rate": 7.78180763092993e-06, + "loss": 0.0359, + "step": 16944 + }, + { + "epoch": 1.837055507372073, + "grad_norm": 1.443765640258789, + "learning_rate": 7.781082257362541e-06, + "loss": 0.0275, + "step": 16945 + }, + { + "epoch": 1.8371639202081527, + "grad_norm": 0.23964112997055054, + "learning_rate": 7.780356883795155e-06, + "loss": 0.0025, + "step": 16946 + }, + { + "epoch": 1.8372723330442324, + "grad_norm": 1.660560965538025, + "learning_rate": 7.779631510227767e-06, + "loss": 0.0244, + "step": 16947 + }, + { + "epoch": 1.8373807458803122, + "grad_norm": 2.9719130992889404, + "learning_rate": 7.77890613666038e-06, + "loss": 0.0418, + "step": 16948 + }, + { + "epoch": 1.8374891587163922, + "grad_norm": 3.636019468307495, + "learning_rate": 7.778180763092993e-06, + "loss": 0.0323, + "step": 16949 + }, + { + "epoch": 1.8375975715524717, + "grad_norm": 0.2611631155014038, + "learning_rate": 7.777455389525607e-06, + "loss": 0.0023, + "step": 16950 + }, + { + "epoch": 1.8377059843885517, + "grad_norm": 5.287612438201904, + "learning_rate": 7.77673001595822e-06, + "loss": 0.0568, + "step": 16951 + }, + { + "epoch": 1.8378143972246312, + "grad_norm": 3.70810866355896, + "learning_rate": 7.776004642390831e-06, + "loss": 0.0161, + "step": 16952 + }, + { + "epoch": 1.8379228100607112, + "grad_norm": 1.5182887315750122, + "learning_rate": 7.775279268823445e-06, + "loss": 0.0586, + "step": 16953 + }, + { + "epoch": 1.838031222896791, + "grad_norm": 1.6229517459869385, + "learning_rate": 7.774553895256057e-06, + "loss": 0.0308, + "step": 16954 + }, + { + "epoch": 1.8381396357328708, + "grad_norm": 7.100761413574219, + "learning_rate": 7.77382852168867e-06, + "loss": 0.0382, + "step": 16955 + }, + { + "epoch": 1.8382480485689505, + "grad_norm": 2.0028469562530518, + "learning_rate": 7.773103148121283e-06, + "loss": 0.0254, + "step": 16956 + }, + { + "epoch": 1.8383564614050303, + "grad_norm": 4.7670087814331055, + "learning_rate": 7.772377774553897e-06, + "loss": 0.052, + "step": 16957 + }, + { + "epoch": 1.8384648742411103, + "grad_norm": 2.527862071990967, + "learning_rate": 7.77165240098651e-06, + "loss": 0.036, + "step": 16958 + }, + { + "epoch": 1.8385732870771898, + "grad_norm": 3.5722548961639404, + "learning_rate": 7.770927027419122e-06, + "loss": 0.0446, + "step": 16959 + }, + { + "epoch": 1.8386816999132698, + "grad_norm": 2.606579065322876, + "learning_rate": 7.770201653851734e-06, + "loss": 0.0289, + "step": 16960 + }, + { + "epoch": 1.8387901127493496, + "grad_norm": 2.0364952087402344, + "learning_rate": 7.769476280284348e-06, + "loss": 0.0277, + "step": 16961 + }, + { + "epoch": 1.8388985255854293, + "grad_norm": 3.191138744354248, + "learning_rate": 7.76875090671696e-06, + "loss": 0.0258, + "step": 16962 + }, + { + "epoch": 1.839006938421509, + "grad_norm": 4.074512004852295, + "learning_rate": 7.768025533149572e-06, + "loss": 0.0393, + "step": 16963 + }, + { + "epoch": 1.8391153512575888, + "grad_norm": 1.6126245260238647, + "learning_rate": 7.767300159582186e-06, + "loss": 0.0058, + "step": 16964 + }, + { + "epoch": 1.8392237640936688, + "grad_norm": 1.3360291719436646, + "learning_rate": 7.766574786014798e-06, + "loss": 0.0329, + "step": 16965 + }, + { + "epoch": 1.8393321769297484, + "grad_norm": 0.1953297257423401, + "learning_rate": 7.765849412447412e-06, + "loss": 0.0026, + "step": 16966 + }, + { + "epoch": 1.8394405897658284, + "grad_norm": 1.9230724573135376, + "learning_rate": 7.765124038880024e-06, + "loss": 0.0319, + "step": 16967 + }, + { + "epoch": 1.839549002601908, + "grad_norm": 0.6936145424842834, + "learning_rate": 7.764398665312638e-06, + "loss": 0.0088, + "step": 16968 + }, + { + "epoch": 1.8396574154379879, + "grad_norm": 0.4440220296382904, + "learning_rate": 7.76367329174525e-06, + "loss": 0.0024, + "step": 16969 + }, + { + "epoch": 1.8397658282740676, + "grad_norm": 3.0370213985443115, + "learning_rate": 7.762947918177862e-06, + "loss": 0.0181, + "step": 16970 + }, + { + "epoch": 1.8398742411101474, + "grad_norm": 2.606921911239624, + "learning_rate": 7.762222544610474e-06, + "loss": 0.0582, + "step": 16971 + }, + { + "epoch": 1.8399826539462274, + "grad_norm": 0.2498951256275177, + "learning_rate": 7.761497171043088e-06, + "loss": 0.0024, + "step": 16972 + }, + { + "epoch": 1.840091066782307, + "grad_norm": 1.7013393640518188, + "learning_rate": 7.7607717974757e-06, + "loss": 0.0155, + "step": 16973 + }, + { + "epoch": 1.840199479618387, + "grad_norm": 2.928690195083618, + "learning_rate": 7.760046423908312e-06, + "loss": 0.0096, + "step": 16974 + }, + { + "epoch": 1.8403078924544665, + "grad_norm": 0.4625003933906555, + "learning_rate": 7.759321050340926e-06, + "loss": 0.0031, + "step": 16975 + }, + { + "epoch": 1.8404163052905465, + "grad_norm": 0.20989185571670532, + "learning_rate": 7.75859567677354e-06, + "loss": 0.0025, + "step": 16976 + }, + { + "epoch": 1.8405247181266262, + "grad_norm": 2.5344526767730713, + "learning_rate": 7.757870303206152e-06, + "loss": 0.0135, + "step": 16977 + }, + { + "epoch": 1.840633130962706, + "grad_norm": 0.7153432369232178, + "learning_rate": 7.757144929638764e-06, + "loss": 0.0051, + "step": 16978 + }, + { + "epoch": 1.8407415437987857, + "grad_norm": 3.923452854156494, + "learning_rate": 7.756419556071378e-06, + "loss": 0.0357, + "step": 16979 + }, + { + "epoch": 1.8408499566348655, + "grad_norm": 0.13989272713661194, + "learning_rate": 7.75569418250399e-06, + "loss": 0.0017, + "step": 16980 + }, + { + "epoch": 1.8409583694709455, + "grad_norm": 4.543262958526611, + "learning_rate": 7.754968808936602e-06, + "loss": 0.0657, + "step": 16981 + }, + { + "epoch": 1.841066782307025, + "grad_norm": 0.18476372957229614, + "learning_rate": 7.754243435369216e-06, + "loss": 0.002, + "step": 16982 + }, + { + "epoch": 1.841175195143105, + "grad_norm": 8.478641510009766, + "learning_rate": 7.75351806180183e-06, + "loss": 0.0355, + "step": 16983 + }, + { + "epoch": 1.8412836079791848, + "grad_norm": 1.0004158020019531, + "learning_rate": 7.752792688234442e-06, + "loss": 0.0063, + "step": 16984 + }, + { + "epoch": 1.8413920208152645, + "grad_norm": 3.6481752395629883, + "learning_rate": 7.752067314667054e-06, + "loss": 0.0332, + "step": 16985 + }, + { + "epoch": 1.8415004336513443, + "grad_norm": 0.15517723560333252, + "learning_rate": 7.751341941099666e-06, + "loss": 0.001, + "step": 16986 + }, + { + "epoch": 1.841608846487424, + "grad_norm": 2.0495033264160156, + "learning_rate": 7.75061656753228e-06, + "loss": 0.0786, + "step": 16987 + }, + { + "epoch": 1.841717259323504, + "grad_norm": 1.9920357465744019, + "learning_rate": 7.749891193964892e-06, + "loss": 0.0639, + "step": 16988 + }, + { + "epoch": 1.8418256721595836, + "grad_norm": 4.878334999084473, + "learning_rate": 7.749165820397504e-06, + "loss": 0.0568, + "step": 16989 + }, + { + "epoch": 1.8419340849956636, + "grad_norm": 0.5048913955688477, + "learning_rate": 7.748440446830118e-06, + "loss": 0.0045, + "step": 16990 + }, + { + "epoch": 1.8420424978317431, + "grad_norm": 1.7068718671798706, + "learning_rate": 7.74771507326273e-06, + "loss": 0.0287, + "step": 16991 + }, + { + "epoch": 1.842150910667823, + "grad_norm": 0.5868927836418152, + "learning_rate": 7.746989699695344e-06, + "loss": 0.0043, + "step": 16992 + }, + { + "epoch": 1.8422593235039029, + "grad_norm": 4.823350429534912, + "learning_rate": 7.746264326127956e-06, + "loss": 0.0297, + "step": 16993 + }, + { + "epoch": 1.8423677363399826, + "grad_norm": 2.1165592670440674, + "learning_rate": 7.74553895256057e-06, + "loss": 0.0158, + "step": 16994 + }, + { + "epoch": 1.8424761491760624, + "grad_norm": 1.0401197671890259, + "learning_rate": 7.744813578993182e-06, + "loss": 0.0099, + "step": 16995 + }, + { + "epoch": 1.8425845620121422, + "grad_norm": 3.810914993286133, + "learning_rate": 7.744088205425795e-06, + "loss": 0.085, + "step": 16996 + }, + { + "epoch": 1.8426929748482221, + "grad_norm": 3.124762773513794, + "learning_rate": 7.743362831858407e-06, + "loss": 0.0344, + "step": 16997 + }, + { + "epoch": 1.8428013876843017, + "grad_norm": 1.7116926908493042, + "learning_rate": 7.74263745829102e-06, + "loss": 0.0056, + "step": 16998 + }, + { + "epoch": 1.8429098005203817, + "grad_norm": 5.612717628479004, + "learning_rate": 7.741912084723633e-06, + "loss": 0.0524, + "step": 16999 + }, + { + "epoch": 1.8430182133564614, + "grad_norm": 0.3934377431869507, + "learning_rate": 7.741186711156247e-06, + "loss": 0.0029, + "step": 17000 + }, + { + "epoch": 1.8431266261925412, + "grad_norm": 2.09877347946167, + "learning_rate": 7.740461337588859e-06, + "loss": 0.0317, + "step": 17001 + }, + { + "epoch": 1.843235039028621, + "grad_norm": 4.988891124725342, + "learning_rate": 7.739735964021473e-06, + "loss": 0.0408, + "step": 17002 + }, + { + "epoch": 1.8433434518647007, + "grad_norm": 0.8424440622329712, + "learning_rate": 7.739010590454085e-06, + "loss": 0.0042, + "step": 17003 + }, + { + "epoch": 1.8434518647007807, + "grad_norm": 0.2961026430130005, + "learning_rate": 7.738285216886697e-06, + "loss": 0.0033, + "step": 17004 + }, + { + "epoch": 1.8435602775368602, + "grad_norm": 1.4051121473312378, + "learning_rate": 7.73755984331931e-06, + "loss": 0.0286, + "step": 17005 + }, + { + "epoch": 1.8436686903729402, + "grad_norm": 0.6027923822402954, + "learning_rate": 7.736834469751923e-06, + "loss": 0.0058, + "step": 17006 + }, + { + "epoch": 1.84377710320902, + "grad_norm": 3.5739669799804688, + "learning_rate": 7.736109096184535e-06, + "loss": 0.016, + "step": 17007 + }, + { + "epoch": 1.8438855160450998, + "grad_norm": 1.0692172050476074, + "learning_rate": 7.735383722617149e-06, + "loss": 0.022, + "step": 17008 + }, + { + "epoch": 1.8439939288811795, + "grad_norm": 0.7872333526611328, + "learning_rate": 7.734658349049763e-06, + "loss": 0.0036, + "step": 17009 + }, + { + "epoch": 1.8441023417172593, + "grad_norm": 2.1814286708831787, + "learning_rate": 7.733932975482375e-06, + "loss": 0.004, + "step": 17010 + }, + { + "epoch": 1.8442107545533393, + "grad_norm": 1.7166807651519775, + "learning_rate": 7.733207601914987e-06, + "loss": 0.0154, + "step": 17011 + }, + { + "epoch": 1.8443191673894188, + "grad_norm": 0.4862686097621918, + "learning_rate": 7.732482228347599e-06, + "loss": 0.0065, + "step": 17012 + }, + { + "epoch": 1.8444275802254988, + "grad_norm": 1.749224066734314, + "learning_rate": 7.731756854780213e-06, + "loss": 0.0311, + "step": 17013 + }, + { + "epoch": 1.8445359930615783, + "grad_norm": 7.875115394592285, + "learning_rate": 7.731031481212825e-06, + "loss": 0.0211, + "step": 17014 + }, + { + "epoch": 1.8446444058976583, + "grad_norm": 5.0372748374938965, + "learning_rate": 7.730306107645437e-06, + "loss": 0.0216, + "step": 17015 + }, + { + "epoch": 1.844752818733738, + "grad_norm": 2.1493215560913086, + "learning_rate": 7.729580734078051e-06, + "loss": 0.0215, + "step": 17016 + }, + { + "epoch": 1.8448612315698178, + "grad_norm": 1.5576053857803345, + "learning_rate": 7.728855360510663e-06, + "loss": 0.0074, + "step": 17017 + }, + { + "epoch": 1.8449696444058976, + "grad_norm": 0.3416016101837158, + "learning_rate": 7.728129986943277e-06, + "loss": 0.002, + "step": 17018 + }, + { + "epoch": 1.8450780572419774, + "grad_norm": 0.16152285039424896, + "learning_rate": 7.727404613375889e-06, + "loss": 0.0016, + "step": 17019 + }, + { + "epoch": 1.8451864700780574, + "grad_norm": 5.2266035079956055, + "learning_rate": 7.726679239808503e-06, + "loss": 0.0169, + "step": 17020 + }, + { + "epoch": 1.845294882914137, + "grad_norm": 1.0206882953643799, + "learning_rate": 7.725953866241115e-06, + "loss": 0.0174, + "step": 17021 + }, + { + "epoch": 1.8454032957502169, + "grad_norm": 1.2400610446929932, + "learning_rate": 7.725228492673727e-06, + "loss": 0.0102, + "step": 17022 + }, + { + "epoch": 1.8455117085862967, + "grad_norm": 4.893003463745117, + "learning_rate": 7.72450311910634e-06, + "loss": 0.0166, + "step": 17023 + }, + { + "epoch": 1.8456201214223764, + "grad_norm": 0.5093989372253418, + "learning_rate": 7.723777745538953e-06, + "loss": 0.0026, + "step": 17024 + }, + { + "epoch": 1.8457285342584562, + "grad_norm": 2.271153211593628, + "learning_rate": 7.723052371971565e-06, + "loss": 0.0198, + "step": 17025 + }, + { + "epoch": 1.845836947094536, + "grad_norm": 0.9540947079658508, + "learning_rate": 7.72232699840418e-06, + "loss": 0.0074, + "step": 17026 + }, + { + "epoch": 1.845945359930616, + "grad_norm": 1.7505873441696167, + "learning_rate": 7.721601624836791e-06, + "loss": 0.0335, + "step": 17027 + }, + { + "epoch": 1.8460537727666955, + "grad_norm": 1.6480602025985718, + "learning_rate": 7.720876251269405e-06, + "loss": 0.0234, + "step": 17028 + }, + { + "epoch": 1.8461621856027755, + "grad_norm": 6.895153045654297, + "learning_rate": 7.720150877702017e-06, + "loss": 0.0281, + "step": 17029 + }, + { + "epoch": 1.8462705984388552, + "grad_norm": 4.094840049743652, + "learning_rate": 7.71942550413463e-06, + "loss": 0.0279, + "step": 17030 + }, + { + "epoch": 1.846379011274935, + "grad_norm": 0.8260528445243835, + "learning_rate": 7.718700130567243e-06, + "loss": 0.0064, + "step": 17031 + }, + { + "epoch": 1.8464874241110147, + "grad_norm": 4.8429036140441895, + "learning_rate": 7.717974756999855e-06, + "loss": 0.0191, + "step": 17032 + }, + { + "epoch": 1.8465958369470945, + "grad_norm": 1.7888997793197632, + "learning_rate": 7.717249383432468e-06, + "loss": 0.0067, + "step": 17033 + }, + { + "epoch": 1.8467042497831745, + "grad_norm": 0.5075759887695312, + "learning_rate": 7.716524009865081e-06, + "loss": 0.0031, + "step": 17034 + }, + { + "epoch": 1.846812662619254, + "grad_norm": 0.6440699100494385, + "learning_rate": 7.715798636297695e-06, + "loss": 0.0049, + "step": 17035 + }, + { + "epoch": 1.846921075455334, + "grad_norm": 1.8256666660308838, + "learning_rate": 7.715073262730307e-06, + "loss": 0.0218, + "step": 17036 + }, + { + "epoch": 1.8470294882914136, + "grad_norm": 3.4900171756744385, + "learning_rate": 7.71434788916292e-06, + "loss": 0.0584, + "step": 17037 + }, + { + "epoch": 1.8471379011274935, + "grad_norm": 0.5392968654632568, + "learning_rate": 7.713622515595532e-06, + "loss": 0.0053, + "step": 17038 + }, + { + "epoch": 1.8472463139635733, + "grad_norm": 1.1281819343566895, + "learning_rate": 7.712897142028146e-06, + "loss": 0.0135, + "step": 17039 + }, + { + "epoch": 1.847354726799653, + "grad_norm": 1.2568068504333496, + "learning_rate": 7.712171768460758e-06, + "loss": 0.019, + "step": 17040 + }, + { + "epoch": 1.8474631396357328, + "grad_norm": 4.182677268981934, + "learning_rate": 7.71144639489337e-06, + "loss": 0.013, + "step": 17041 + }, + { + "epoch": 1.8475715524718126, + "grad_norm": 3.7116053104400635, + "learning_rate": 7.710721021325984e-06, + "loss": 0.0325, + "step": 17042 + }, + { + "epoch": 1.8476799653078926, + "grad_norm": 2.773458242416382, + "learning_rate": 7.709995647758596e-06, + "loss": 0.0126, + "step": 17043 + }, + { + "epoch": 1.8477883781439721, + "grad_norm": 2.922165870666504, + "learning_rate": 7.70927027419121e-06, + "loss": 0.0189, + "step": 17044 + }, + { + "epoch": 1.847896790980052, + "grad_norm": 3.0458590984344482, + "learning_rate": 7.708544900623822e-06, + "loss": 0.0178, + "step": 17045 + }, + { + "epoch": 1.8480052038161319, + "grad_norm": 2.969921827316284, + "learning_rate": 7.707819527056436e-06, + "loss": 0.036, + "step": 17046 + }, + { + "epoch": 1.8481136166522116, + "grad_norm": 1.5494881868362427, + "learning_rate": 7.707094153489048e-06, + "loss": 0.0218, + "step": 17047 + }, + { + "epoch": 1.8482220294882914, + "grad_norm": 3.0473437309265137, + "learning_rate": 7.70636877992166e-06, + "loss": 0.0187, + "step": 17048 + }, + { + "epoch": 1.8483304423243712, + "grad_norm": 0.934472918510437, + "learning_rate": 7.705643406354272e-06, + "loss": 0.0098, + "step": 17049 + }, + { + "epoch": 1.8484388551604511, + "grad_norm": 2.737578868865967, + "learning_rate": 7.704918032786886e-06, + "loss": 0.0094, + "step": 17050 + }, + { + "epoch": 1.8485472679965307, + "grad_norm": 2.9393057823181152, + "learning_rate": 7.704192659219498e-06, + "loss": 0.0148, + "step": 17051 + }, + { + "epoch": 1.8486556808326107, + "grad_norm": 2.5274853706359863, + "learning_rate": 7.703467285652112e-06, + "loss": 0.0438, + "step": 17052 + }, + { + "epoch": 1.8487640936686902, + "grad_norm": 0.3845408260822296, + "learning_rate": 7.702741912084724e-06, + "loss": 0.0027, + "step": 17053 + }, + { + "epoch": 1.8488725065047702, + "grad_norm": 0.45636579394340515, + "learning_rate": 7.702016538517338e-06, + "loss": 0.0032, + "step": 17054 + }, + { + "epoch": 1.84898091934085, + "grad_norm": 2.82029128074646, + "learning_rate": 7.70129116494995e-06, + "loss": 0.0257, + "step": 17055 + }, + { + "epoch": 1.8490893321769297, + "grad_norm": 1.9988954067230225, + "learning_rate": 7.700565791382562e-06, + "loss": 0.0337, + "step": 17056 + }, + { + "epoch": 1.8491977450130095, + "grad_norm": 4.905550956726074, + "learning_rate": 7.699840417815176e-06, + "loss": 0.0093, + "step": 17057 + }, + { + "epoch": 1.8493061578490892, + "grad_norm": 0.6225854754447937, + "learning_rate": 7.699115044247788e-06, + "loss": 0.0046, + "step": 17058 + }, + { + "epoch": 1.8494145706851692, + "grad_norm": 0.6450419425964355, + "learning_rate": 7.6983896706804e-06, + "loss": 0.0037, + "step": 17059 + }, + { + "epoch": 1.8495229835212488, + "grad_norm": 0.04330435022711754, + "learning_rate": 7.697664297113014e-06, + "loss": 0.0004, + "step": 17060 + }, + { + "epoch": 1.8496313963573288, + "grad_norm": 2.858809232711792, + "learning_rate": 7.696938923545628e-06, + "loss": 0.0213, + "step": 17061 + }, + { + "epoch": 1.8497398091934085, + "grad_norm": 3.314537763595581, + "learning_rate": 7.69621354997824e-06, + "loss": 0.0239, + "step": 17062 + }, + { + "epoch": 1.8498482220294883, + "grad_norm": 0.20744770765304565, + "learning_rate": 7.695488176410852e-06, + "loss": 0.0009, + "step": 17063 + }, + { + "epoch": 1.849956634865568, + "grad_norm": 0.29395437240600586, + "learning_rate": 7.694762802843464e-06, + "loss": 0.0029, + "step": 17064 + }, + { + "epoch": 1.8500650477016478, + "grad_norm": 2.74692440032959, + "learning_rate": 7.694037429276078e-06, + "loss": 0.0123, + "step": 17065 + }, + { + "epoch": 1.8501734605377278, + "grad_norm": 4.107530117034912, + "learning_rate": 7.69331205570869e-06, + "loss": 0.069, + "step": 17066 + }, + { + "epoch": 1.8502818733738073, + "grad_norm": 0.07702047377824783, + "learning_rate": 7.692586682141302e-06, + "loss": 0.0006, + "step": 17067 + }, + { + "epoch": 1.8503902862098873, + "grad_norm": 0.2451315075159073, + "learning_rate": 7.691861308573916e-06, + "loss": 0.0019, + "step": 17068 + }, + { + "epoch": 1.850498699045967, + "grad_norm": 2.853271007537842, + "learning_rate": 7.691135935006528e-06, + "loss": 0.0198, + "step": 17069 + }, + { + "epoch": 1.8506071118820469, + "grad_norm": 1.298538327217102, + "learning_rate": 7.690410561439142e-06, + "loss": 0.029, + "step": 17070 + }, + { + "epoch": 1.8507155247181266, + "grad_norm": 0.20465801656246185, + "learning_rate": 7.689685187871754e-06, + "loss": 0.0014, + "step": 17071 + }, + { + "epoch": 1.8508239375542064, + "grad_norm": 2.7523465156555176, + "learning_rate": 7.688959814304368e-06, + "loss": 0.0103, + "step": 17072 + }, + { + "epoch": 1.8509323503902864, + "grad_norm": 3.3440966606140137, + "learning_rate": 7.68823444073698e-06, + "loss": 0.0504, + "step": 17073 + }, + { + "epoch": 1.851040763226366, + "grad_norm": 0.8709885478019714, + "learning_rate": 7.687509067169593e-06, + "loss": 0.0049, + "step": 17074 + }, + { + "epoch": 1.851149176062446, + "grad_norm": 0.8814502358436584, + "learning_rate": 7.686783693602205e-06, + "loss": 0.0071, + "step": 17075 + }, + { + "epoch": 1.8512575888985254, + "grad_norm": 3.6308233737945557, + "learning_rate": 7.686058320034819e-06, + "loss": 0.033, + "step": 17076 + }, + { + "epoch": 1.8513660017346054, + "grad_norm": 2.166522264480591, + "learning_rate": 7.68533294646743e-06, + "loss": 0.0163, + "step": 17077 + }, + { + "epoch": 1.8514744145706852, + "grad_norm": 0.6237631440162659, + "learning_rate": 7.684607572900045e-06, + "loss": 0.0026, + "step": 17078 + }, + { + "epoch": 1.851582827406765, + "grad_norm": 2.706387996673584, + "learning_rate": 7.683882199332657e-06, + "loss": 0.0046, + "step": 17079 + }, + { + "epoch": 1.8516912402428447, + "grad_norm": 1.778225302696228, + "learning_rate": 7.68315682576527e-06, + "loss": 0.0109, + "step": 17080 + }, + { + "epoch": 1.8517996530789245, + "grad_norm": 2.8571605682373047, + "learning_rate": 7.682431452197883e-06, + "loss": 0.0515, + "step": 17081 + }, + { + "epoch": 1.8519080659150045, + "grad_norm": 1.8773269653320312, + "learning_rate": 7.681706078630495e-06, + "loss": 0.0098, + "step": 17082 + }, + { + "epoch": 1.852016478751084, + "grad_norm": 2.6090924739837646, + "learning_rate": 7.680980705063109e-06, + "loss": 0.0156, + "step": 17083 + }, + { + "epoch": 1.852124891587164, + "grad_norm": 2.590773582458496, + "learning_rate": 7.68025533149572e-06, + "loss": 0.0114, + "step": 17084 + }, + { + "epoch": 1.8522333044232437, + "grad_norm": 2.1567554473876953, + "learning_rate": 7.679529957928333e-06, + "loss": 0.0428, + "step": 17085 + }, + { + "epoch": 1.8523417172593235, + "grad_norm": 0.12957128882408142, + "learning_rate": 7.678804584360947e-06, + "loss": 0.0007, + "step": 17086 + }, + { + "epoch": 1.8524501300954033, + "grad_norm": 2.1809983253479004, + "learning_rate": 7.67807921079356e-06, + "loss": 0.0367, + "step": 17087 + }, + { + "epoch": 1.852558542931483, + "grad_norm": 0.41210025548934937, + "learning_rate": 7.677353837226173e-06, + "loss": 0.0032, + "step": 17088 + }, + { + "epoch": 1.852666955767563, + "grad_norm": 6.955594539642334, + "learning_rate": 7.676628463658785e-06, + "loss": 0.0385, + "step": 17089 + }, + { + "epoch": 1.8527753686036426, + "grad_norm": 5.3980393409729, + "learning_rate": 7.675903090091397e-06, + "loss": 0.0608, + "step": 17090 + }, + { + "epoch": 1.8528837814397225, + "grad_norm": 0.5649488568305969, + "learning_rate": 7.675177716524011e-06, + "loss": 0.0032, + "step": 17091 + }, + { + "epoch": 1.8529921942758023, + "grad_norm": 5.355048656463623, + "learning_rate": 7.674452342956623e-06, + "loss": 0.0628, + "step": 17092 + }, + { + "epoch": 1.853100607111882, + "grad_norm": 1.8688204288482666, + "learning_rate": 7.673726969389235e-06, + "loss": 0.0496, + "step": 17093 + }, + { + "epoch": 1.8532090199479618, + "grad_norm": 2.68436336517334, + "learning_rate": 7.673001595821849e-06, + "loss": 0.0174, + "step": 17094 + }, + { + "epoch": 1.8533174327840416, + "grad_norm": 3.0699219703674316, + "learning_rate": 7.672276222254463e-06, + "loss": 0.0192, + "step": 17095 + }, + { + "epoch": 1.8534258456201216, + "grad_norm": 1.5993921756744385, + "learning_rate": 7.671550848687075e-06, + "loss": 0.017, + "step": 17096 + }, + { + "epoch": 1.8535342584562011, + "grad_norm": 4.3814473152160645, + "learning_rate": 7.670825475119687e-06, + "loss": 0.0387, + "step": 17097 + }, + { + "epoch": 1.853642671292281, + "grad_norm": 0.8722249269485474, + "learning_rate": 7.670100101552301e-06, + "loss": 0.0088, + "step": 17098 + }, + { + "epoch": 1.8537510841283606, + "grad_norm": 6.968565940856934, + "learning_rate": 7.669374727984913e-06, + "loss": 0.0574, + "step": 17099 + }, + { + "epoch": 1.8538594969644406, + "grad_norm": 2.450767755508423, + "learning_rate": 7.668649354417525e-06, + "loss": 0.0444, + "step": 17100 + }, + { + "epoch": 1.8539679098005204, + "grad_norm": 2.6832001209259033, + "learning_rate": 7.667923980850139e-06, + "loss": 0.0281, + "step": 17101 + }, + { + "epoch": 1.8540763226366002, + "grad_norm": 0.5051764845848083, + "learning_rate": 7.667198607282751e-06, + "loss": 0.0029, + "step": 17102 + }, + { + "epoch": 1.85418473547268, + "grad_norm": 11.918529510498047, + "learning_rate": 7.666473233715363e-06, + "loss": 0.0452, + "step": 17103 + }, + { + "epoch": 1.8542931483087597, + "grad_norm": 0.8441395163536072, + "learning_rate": 7.665747860147977e-06, + "loss": 0.0019, + "step": 17104 + }, + { + "epoch": 1.8544015611448397, + "grad_norm": 0.8241109251976013, + "learning_rate": 7.66502248658059e-06, + "loss": 0.0187, + "step": 17105 + }, + { + "epoch": 1.8545099739809192, + "grad_norm": 3.5880744457244873, + "learning_rate": 7.664297113013203e-06, + "loss": 0.0216, + "step": 17106 + }, + { + "epoch": 1.8546183868169992, + "grad_norm": 2.9501585960388184, + "learning_rate": 7.663571739445815e-06, + "loss": 0.0283, + "step": 17107 + }, + { + "epoch": 1.854726799653079, + "grad_norm": 0.2842288017272949, + "learning_rate": 7.662846365878427e-06, + "loss": 0.0014, + "step": 17108 + }, + { + "epoch": 1.8548352124891587, + "grad_norm": 1.738430142402649, + "learning_rate": 7.662120992311041e-06, + "loss": 0.0172, + "step": 17109 + }, + { + "epoch": 1.8549436253252385, + "grad_norm": 0.8731341361999512, + "learning_rate": 7.661395618743653e-06, + "loss": 0.0046, + "step": 17110 + }, + { + "epoch": 1.8550520381613183, + "grad_norm": 0.4700377881526947, + "learning_rate": 7.660670245176266e-06, + "loss": 0.0063, + "step": 17111 + }, + { + "epoch": 1.8551604509973982, + "grad_norm": 3.138859987258911, + "learning_rate": 7.65994487160888e-06, + "loss": 0.0293, + "step": 17112 + }, + { + "epoch": 1.8552688638334778, + "grad_norm": 5.545475482940674, + "learning_rate": 7.659219498041493e-06, + "loss": 0.0374, + "step": 17113 + }, + { + "epoch": 1.8553772766695578, + "grad_norm": 1.0227470397949219, + "learning_rate": 7.658494124474105e-06, + "loss": 0.0084, + "step": 17114 + }, + { + "epoch": 1.8554856895056373, + "grad_norm": 3.2533886432647705, + "learning_rate": 7.657768750906718e-06, + "loss": 0.0576, + "step": 17115 + }, + { + "epoch": 1.8555941023417173, + "grad_norm": 5.617982864379883, + "learning_rate": 7.65704337733933e-06, + "loss": 0.0854, + "step": 17116 + }, + { + "epoch": 1.855702515177797, + "grad_norm": 1.0097534656524658, + "learning_rate": 7.656318003771944e-06, + "loss": 0.0075, + "step": 17117 + }, + { + "epoch": 1.8558109280138768, + "grad_norm": 0.4537723660469055, + "learning_rate": 7.655592630204556e-06, + "loss": 0.0045, + "step": 17118 + }, + { + "epoch": 1.8559193408499568, + "grad_norm": 2.592259645462036, + "learning_rate": 7.654867256637168e-06, + "loss": 0.0211, + "step": 17119 + }, + { + "epoch": 1.8560277536860363, + "grad_norm": 3.2667019367218018, + "learning_rate": 7.654141883069782e-06, + "loss": 0.0151, + "step": 17120 + }, + { + "epoch": 1.8561361665221163, + "grad_norm": 1.6974509954452515, + "learning_rate": 7.653416509502395e-06, + "loss": 0.0189, + "step": 17121 + }, + { + "epoch": 1.8562445793581959, + "grad_norm": 0.3444674015045166, + "learning_rate": 7.652691135935008e-06, + "loss": 0.0021, + "step": 17122 + }, + { + "epoch": 1.8563529921942759, + "grad_norm": 0.1851755976676941, + "learning_rate": 7.65196576236762e-06, + "loss": 0.0016, + "step": 17123 + }, + { + "epoch": 1.8564614050303556, + "grad_norm": 0.22917339205741882, + "learning_rate": 7.651240388800234e-06, + "loss": 0.0019, + "step": 17124 + }, + { + "epoch": 1.8565698178664354, + "grad_norm": 2.5402889251708984, + "learning_rate": 7.650515015232846e-06, + "loss": 0.0221, + "step": 17125 + }, + { + "epoch": 1.8566782307025151, + "grad_norm": 2.564893960952759, + "learning_rate": 7.649789641665458e-06, + "loss": 0.0502, + "step": 17126 + }, + { + "epoch": 1.856786643538595, + "grad_norm": 1.0001144409179688, + "learning_rate": 7.649064268098072e-06, + "loss": 0.0311, + "step": 17127 + }, + { + "epoch": 1.856895056374675, + "grad_norm": 0.8183296322822571, + "learning_rate": 7.648338894530684e-06, + "loss": 0.0041, + "step": 17128 + }, + { + "epoch": 1.8570034692107544, + "grad_norm": 1.3785284757614136, + "learning_rate": 7.647613520963296e-06, + "loss": 0.0112, + "step": 17129 + }, + { + "epoch": 1.8571118820468344, + "grad_norm": 1.0019334554672241, + "learning_rate": 7.64688814739591e-06, + "loss": 0.0049, + "step": 17130 + }, + { + "epoch": 1.8572202948829142, + "grad_norm": 0.08594702929258347, + "learning_rate": 7.646162773828522e-06, + "loss": 0.0009, + "step": 17131 + }, + { + "epoch": 1.857328707718994, + "grad_norm": 0.4424300491809845, + "learning_rate": 7.645437400261136e-06, + "loss": 0.0034, + "step": 17132 + }, + { + "epoch": 1.8574371205550737, + "grad_norm": 3.483607292175293, + "learning_rate": 7.644712026693748e-06, + "loss": 0.0154, + "step": 17133 + }, + { + "epoch": 1.8575455333911535, + "grad_norm": 3.098414182662964, + "learning_rate": 7.64398665312636e-06, + "loss": 0.0181, + "step": 17134 + }, + { + "epoch": 1.8576539462272335, + "grad_norm": 2.128274917602539, + "learning_rate": 7.643261279558974e-06, + "loss": 0.0181, + "step": 17135 + }, + { + "epoch": 1.857762359063313, + "grad_norm": 3.4645609855651855, + "learning_rate": 7.642535905991586e-06, + "loss": 0.022, + "step": 17136 + }, + { + "epoch": 1.857870771899393, + "grad_norm": 2.973165273666382, + "learning_rate": 7.641810532424198e-06, + "loss": 0.0185, + "step": 17137 + }, + { + "epoch": 1.8579791847354725, + "grad_norm": 3.109414577484131, + "learning_rate": 7.641085158856812e-06, + "loss": 0.0108, + "step": 17138 + }, + { + "epoch": 1.8580875975715525, + "grad_norm": 1.773262619972229, + "learning_rate": 7.640359785289426e-06, + "loss": 0.0346, + "step": 17139 + }, + { + "epoch": 1.8581960104076323, + "grad_norm": 2.9124953746795654, + "learning_rate": 7.639634411722038e-06, + "loss": 0.0204, + "step": 17140 + }, + { + "epoch": 1.858304423243712, + "grad_norm": 1.4948540925979614, + "learning_rate": 7.63890903815465e-06, + "loss": 0.0223, + "step": 17141 + }, + { + "epoch": 1.8584128360797918, + "grad_norm": 0.5317524671554565, + "learning_rate": 7.638183664587262e-06, + "loss": 0.0021, + "step": 17142 + }, + { + "epoch": 1.8585212489158716, + "grad_norm": 2.846189260482788, + "learning_rate": 7.637458291019876e-06, + "loss": 0.0598, + "step": 17143 + }, + { + "epoch": 1.8586296617519515, + "grad_norm": 0.3779308795928955, + "learning_rate": 7.636732917452488e-06, + "loss": 0.0036, + "step": 17144 + }, + { + "epoch": 1.858738074588031, + "grad_norm": 4.229214668273926, + "learning_rate": 7.6360075438851e-06, + "loss": 0.0117, + "step": 17145 + }, + { + "epoch": 1.858846487424111, + "grad_norm": 2.4137163162231445, + "learning_rate": 7.635282170317714e-06, + "loss": 0.0685, + "step": 17146 + }, + { + "epoch": 1.8589549002601908, + "grad_norm": 5.189908504486084, + "learning_rate": 7.634556796750328e-06, + "loss": 0.0317, + "step": 17147 + }, + { + "epoch": 1.8590633130962706, + "grad_norm": 8.77538776397705, + "learning_rate": 7.63383142318294e-06, + "loss": 0.0208, + "step": 17148 + }, + { + "epoch": 1.8591717259323504, + "grad_norm": 2.249643087387085, + "learning_rate": 7.633106049615552e-06, + "loss": 0.0173, + "step": 17149 + }, + { + "epoch": 1.8592801387684301, + "grad_norm": 5.729757308959961, + "learning_rate": 7.632380676048166e-06, + "loss": 0.0272, + "step": 17150 + }, + { + "epoch": 1.85938855160451, + "grad_norm": 0.40154793858528137, + "learning_rate": 7.631655302480778e-06, + "loss": 0.0038, + "step": 17151 + }, + { + "epoch": 1.8594969644405897, + "grad_norm": 1.8129267692565918, + "learning_rate": 7.63092992891339e-06, + "loss": 0.0159, + "step": 17152 + }, + { + "epoch": 1.8596053772766696, + "grad_norm": 2.5254125595092773, + "learning_rate": 7.630204555346004e-06, + "loss": 0.0408, + "step": 17153 + }, + { + "epoch": 1.8597137901127494, + "grad_norm": 0.502342939376831, + "learning_rate": 7.629479181778617e-06, + "loss": 0.0043, + "step": 17154 + }, + { + "epoch": 1.8598222029488292, + "grad_norm": 1.7977581024169922, + "learning_rate": 7.628753808211229e-06, + "loss": 0.0151, + "step": 17155 + }, + { + "epoch": 1.859930615784909, + "grad_norm": 2.954815626144409, + "learning_rate": 7.6280284346438425e-06, + "loss": 0.0119, + "step": 17156 + }, + { + "epoch": 1.8600390286209887, + "grad_norm": 1.765496850013733, + "learning_rate": 7.6273030610764555e-06, + "loss": 0.006, + "step": 17157 + }, + { + "epoch": 1.8601474414570687, + "grad_norm": 0.774015486240387, + "learning_rate": 7.626577687509068e-06, + "loss": 0.0054, + "step": 17158 + }, + { + "epoch": 1.8602558542931482, + "grad_norm": 3.1516783237457275, + "learning_rate": 7.625852313941681e-06, + "loss": 0.0117, + "step": 17159 + }, + { + "epoch": 1.8603642671292282, + "grad_norm": 1.020020842552185, + "learning_rate": 7.625126940374294e-06, + "loss": 0.0105, + "step": 17160 + }, + { + "epoch": 1.8604726799653077, + "grad_norm": 4.504724502563477, + "learning_rate": 7.624401566806906e-06, + "loss": 0.0634, + "step": 17161 + }, + { + "epoch": 1.8605810928013877, + "grad_norm": 3.873692750930786, + "learning_rate": 7.623676193239519e-06, + "loss": 0.0342, + "step": 17162 + }, + { + "epoch": 1.8606895056374675, + "grad_norm": 8.182829856872559, + "learning_rate": 7.622950819672132e-06, + "loss": 0.0103, + "step": 17163 + }, + { + "epoch": 1.8607979184735473, + "grad_norm": 0.18278248608112335, + "learning_rate": 7.622225446104744e-06, + "loss": 0.0014, + "step": 17164 + }, + { + "epoch": 1.860906331309627, + "grad_norm": 3.921388626098633, + "learning_rate": 7.621500072537358e-06, + "loss": 0.0129, + "step": 17165 + }, + { + "epoch": 1.8610147441457068, + "grad_norm": 6.894761085510254, + "learning_rate": 7.620774698969971e-06, + "loss": 0.0302, + "step": 17166 + }, + { + "epoch": 1.8611231569817868, + "grad_norm": 0.19878621399402618, + "learning_rate": 7.620049325402583e-06, + "loss": 0.0025, + "step": 17167 + }, + { + "epoch": 1.8612315698178663, + "grad_norm": 7.963827133178711, + "learning_rate": 7.619323951835196e-06, + "loss": 0.0439, + "step": 17168 + }, + { + "epoch": 1.8613399826539463, + "grad_norm": 0.6563023924827576, + "learning_rate": 7.618598578267809e-06, + "loss": 0.0026, + "step": 17169 + }, + { + "epoch": 1.861448395490026, + "grad_norm": 11.870783805847168, + "learning_rate": 7.617873204700421e-06, + "loss": 0.0472, + "step": 17170 + }, + { + "epoch": 1.8615568083261058, + "grad_norm": 0.3690750300884247, + "learning_rate": 7.617147831133034e-06, + "loss": 0.0029, + "step": 17171 + }, + { + "epoch": 1.8616652211621856, + "grad_norm": 3.558685541152954, + "learning_rate": 7.616422457565646e-06, + "loss": 0.0103, + "step": 17172 + }, + { + "epoch": 1.8617736339982653, + "grad_norm": 2.583116054534912, + "learning_rate": 7.61569708399826e-06, + "loss": 0.0352, + "step": 17173 + }, + { + "epoch": 1.8618820468343453, + "grad_norm": 4.194285869598389, + "learning_rate": 7.614971710430873e-06, + "loss": 0.0198, + "step": 17174 + }, + { + "epoch": 1.8619904596704249, + "grad_norm": 0.9668517112731934, + "learning_rate": 7.614246336863486e-06, + "loss": 0.0414, + "step": 17175 + }, + { + "epoch": 1.8620988725065049, + "grad_norm": 1.5656546354293823, + "learning_rate": 7.613520963296098e-06, + "loss": 0.0108, + "step": 17176 + }, + { + "epoch": 1.8622072853425846, + "grad_norm": 6.332273483276367, + "learning_rate": 7.612795589728711e-06, + "loss": 0.022, + "step": 17177 + }, + { + "epoch": 1.8623156981786644, + "grad_norm": 0.09334912896156311, + "learning_rate": 7.612070216161323e-06, + "loss": 0.0008, + "step": 17178 + }, + { + "epoch": 1.8624241110147441, + "grad_norm": 4.7749128341674805, + "learning_rate": 7.611344842593936e-06, + "loss": 0.0369, + "step": 17179 + }, + { + "epoch": 1.862532523850824, + "grad_norm": 1.2678240537643433, + "learning_rate": 7.610619469026549e-06, + "loss": 0.0051, + "step": 17180 + }, + { + "epoch": 1.862640936686904, + "grad_norm": 3.0647969245910645, + "learning_rate": 7.609894095459161e-06, + "loss": 0.0436, + "step": 17181 + }, + { + "epoch": 1.8627493495229834, + "grad_norm": 0.17114272713661194, + "learning_rate": 7.609168721891775e-06, + "loss": 0.0019, + "step": 17182 + }, + { + "epoch": 1.8628577623590634, + "grad_norm": 0.6394708752632141, + "learning_rate": 7.608443348324388e-06, + "loss": 0.0092, + "step": 17183 + }, + { + "epoch": 1.862966175195143, + "grad_norm": 0.7461534142494202, + "learning_rate": 7.607717974757e-06, + "loss": 0.004, + "step": 17184 + }, + { + "epoch": 1.863074588031223, + "grad_norm": 4.4461212158203125, + "learning_rate": 7.606992601189613e-06, + "loss": 0.0128, + "step": 17185 + }, + { + "epoch": 1.8631830008673027, + "grad_norm": 0.8440335392951965, + "learning_rate": 7.606267227622226e-06, + "loss": 0.0053, + "step": 17186 + }, + { + "epoch": 1.8632914137033825, + "grad_norm": 1.9861390590667725, + "learning_rate": 7.6055418540548385e-06, + "loss": 0.0051, + "step": 17187 + }, + { + "epoch": 1.8633998265394622, + "grad_norm": 4.6782073974609375, + "learning_rate": 7.6048164804874514e-06, + "loss": 0.0182, + "step": 17188 + }, + { + "epoch": 1.863508239375542, + "grad_norm": 0.5189340710639954, + "learning_rate": 7.6040911069200644e-06, + "loss": 0.0042, + "step": 17189 + }, + { + "epoch": 1.863616652211622, + "grad_norm": 2.991816759109497, + "learning_rate": 7.6033657333526774e-06, + "loss": 0.047, + "step": 17190 + }, + { + "epoch": 1.8637250650477015, + "grad_norm": 1.6683536767959595, + "learning_rate": 7.60264035978529e-06, + "loss": 0.0055, + "step": 17191 + }, + { + "epoch": 1.8638334778837815, + "grad_norm": 0.7058501243591309, + "learning_rate": 7.601914986217903e-06, + "loss": 0.0065, + "step": 17192 + }, + { + "epoch": 1.8639418907198613, + "grad_norm": 5.1909637451171875, + "learning_rate": 7.6011896126505156e-06, + "loss": 0.0116, + "step": 17193 + }, + { + "epoch": 1.864050303555941, + "grad_norm": 1.611042857170105, + "learning_rate": 7.6004642390831285e-06, + "loss": 0.0248, + "step": 17194 + }, + { + "epoch": 1.8641587163920208, + "grad_norm": 4.31751823425293, + "learning_rate": 7.5997388655157415e-06, + "loss": 0.1084, + "step": 17195 + }, + { + "epoch": 1.8642671292281006, + "grad_norm": 3.7704174518585205, + "learning_rate": 7.599013491948354e-06, + "loss": 0.009, + "step": 17196 + }, + { + "epoch": 1.8643755420641805, + "grad_norm": 0.3081228733062744, + "learning_rate": 7.598288118380967e-06, + "loss": 0.0023, + "step": 17197 + }, + { + "epoch": 1.86448395490026, + "grad_norm": 1.9857182502746582, + "learning_rate": 7.597562744813579e-06, + "loss": 0.0098, + "step": 17198 + }, + { + "epoch": 1.86459236773634, + "grad_norm": 4.380631446838379, + "learning_rate": 7.596837371246193e-06, + "loss": 0.0204, + "step": 17199 + }, + { + "epoch": 1.8647007805724196, + "grad_norm": 2.9285073280334473, + "learning_rate": 7.596111997678806e-06, + "loss": 0.024, + "step": 17200 + }, + { + "epoch": 1.8648091934084996, + "grad_norm": 2.578831672668457, + "learning_rate": 7.595386624111419e-06, + "loss": 0.0644, + "step": 17201 + }, + { + "epoch": 1.8649176062445794, + "grad_norm": 2.01912522315979, + "learning_rate": 7.594661250544031e-06, + "loss": 0.0592, + "step": 17202 + }, + { + "epoch": 1.8650260190806591, + "grad_norm": 5.163767337799072, + "learning_rate": 7.593935876976644e-06, + "loss": 0.0509, + "step": 17203 + }, + { + "epoch": 1.8651344319167389, + "grad_norm": 3.6563079357147217, + "learning_rate": 7.593210503409256e-06, + "loss": 0.013, + "step": 17204 + }, + { + "epoch": 1.8652428447528187, + "grad_norm": 3.574693202972412, + "learning_rate": 7.592485129841869e-06, + "loss": 0.0094, + "step": 17205 + }, + { + "epoch": 1.8653512575888986, + "grad_norm": 0.07215280830860138, + "learning_rate": 7.591759756274482e-06, + "loss": 0.001, + "step": 17206 + }, + { + "epoch": 1.8654596704249782, + "grad_norm": 8.175588607788086, + "learning_rate": 7.591034382707094e-06, + "loss": 0.0945, + "step": 17207 + }, + { + "epoch": 1.8655680832610582, + "grad_norm": 1.5010508298873901, + "learning_rate": 7.590309009139708e-06, + "loss": 0.006, + "step": 17208 + }, + { + "epoch": 1.865676496097138, + "grad_norm": 1.5943832397460938, + "learning_rate": 7.589583635572321e-06, + "loss": 0.0064, + "step": 17209 + }, + { + "epoch": 1.8657849089332177, + "grad_norm": 3.1404616832733154, + "learning_rate": 7.588858262004933e-06, + "loss": 0.0337, + "step": 17210 + }, + { + "epoch": 1.8658933217692975, + "grad_norm": 0.7775578498840332, + "learning_rate": 7.588132888437546e-06, + "loss": 0.005, + "step": 17211 + }, + { + "epoch": 1.8660017346053772, + "grad_norm": 1.4149235486984253, + "learning_rate": 7.587407514870159e-06, + "loss": 0.0076, + "step": 17212 + }, + { + "epoch": 1.8661101474414572, + "grad_norm": 3.4849472045898438, + "learning_rate": 7.586682141302771e-06, + "loss": 0.0078, + "step": 17213 + }, + { + "epoch": 1.8662185602775367, + "grad_norm": 1.7454825639724731, + "learning_rate": 7.585956767735384e-06, + "loss": 0.0511, + "step": 17214 + }, + { + "epoch": 1.8663269731136167, + "grad_norm": 0.8066596984863281, + "learning_rate": 7.585231394167997e-06, + "loss": 0.0075, + "step": 17215 + }, + { + "epoch": 1.8664353859496965, + "grad_norm": 3.613558292388916, + "learning_rate": 7.58450602060061e-06, + "loss": 0.0228, + "step": 17216 + }, + { + "epoch": 1.8665437987857763, + "grad_norm": 0.5540371537208557, + "learning_rate": 7.583780647033223e-06, + "loss": 0.0096, + "step": 17217 + }, + { + "epoch": 1.866652211621856, + "grad_norm": 0.6277386546134949, + "learning_rate": 7.583055273465836e-06, + "loss": 0.0064, + "step": 17218 + }, + { + "epoch": 1.8667606244579358, + "grad_norm": 1.1077150106430054, + "learning_rate": 7.582329899898448e-06, + "loss": 0.012, + "step": 17219 + }, + { + "epoch": 1.8668690372940158, + "grad_norm": 0.1011885553598404, + "learning_rate": 7.581604526331061e-06, + "loss": 0.001, + "step": 17220 + }, + { + "epoch": 1.8669774501300953, + "grad_norm": 1.1531084775924683, + "learning_rate": 7.580879152763674e-06, + "loss": 0.003, + "step": 17221 + }, + { + "epoch": 1.8670858629661753, + "grad_norm": 0.5111913084983826, + "learning_rate": 7.580153779196286e-06, + "loss": 0.0068, + "step": 17222 + }, + { + "epoch": 1.8671942758022548, + "grad_norm": 2.701083183288574, + "learning_rate": 7.579428405628899e-06, + "loss": 0.0479, + "step": 17223 + }, + { + "epoch": 1.8673026886383348, + "grad_norm": 3.2430293560028076, + "learning_rate": 7.5787030320615115e-06, + "loss": 0.0437, + "step": 17224 + }, + { + "epoch": 1.8674111014744146, + "grad_norm": 0.3842262923717499, + "learning_rate": 7.577977658494125e-06, + "loss": 0.0021, + "step": 17225 + }, + { + "epoch": 1.8675195143104943, + "grad_norm": 0.6925241947174072, + "learning_rate": 7.577252284926738e-06, + "loss": 0.0031, + "step": 17226 + }, + { + "epoch": 1.867627927146574, + "grad_norm": 4.590150356292725, + "learning_rate": 7.576526911359351e-06, + "loss": 0.0199, + "step": 17227 + }, + { + "epoch": 1.8677363399826539, + "grad_norm": 0.15367759764194489, + "learning_rate": 7.5758015377919634e-06, + "loss": 0.0021, + "step": 17228 + }, + { + "epoch": 1.8678447528187339, + "grad_norm": 4.897714138031006, + "learning_rate": 7.5750761642245764e-06, + "loss": 0.0138, + "step": 17229 + }, + { + "epoch": 1.8679531656548134, + "grad_norm": 1.326300859451294, + "learning_rate": 7.574350790657189e-06, + "loss": 0.0144, + "step": 17230 + }, + { + "epoch": 1.8680615784908934, + "grad_norm": 0.6266734600067139, + "learning_rate": 7.5736254170898016e-06, + "loss": 0.0052, + "step": 17231 + }, + { + "epoch": 1.8681699913269731, + "grad_norm": 0.8776223063468933, + "learning_rate": 7.5729000435224146e-06, + "loss": 0.0027, + "step": 17232 + }, + { + "epoch": 1.868278404163053, + "grad_norm": 6.982259750366211, + "learning_rate": 7.572174669955027e-06, + "loss": 0.0283, + "step": 17233 + }, + { + "epoch": 1.8683868169991327, + "grad_norm": 2.2362217903137207, + "learning_rate": 7.5714492963876405e-06, + "loss": 0.0167, + "step": 17234 + }, + { + "epoch": 1.8684952298352124, + "grad_norm": 3.6004459857940674, + "learning_rate": 7.5707239228202535e-06, + "loss": 0.0301, + "step": 17235 + }, + { + "epoch": 1.8686036426712924, + "grad_norm": 0.10372282564640045, + "learning_rate": 7.569998549252866e-06, + "loss": 0.0009, + "step": 17236 + }, + { + "epoch": 1.868712055507372, + "grad_norm": 0.3998453617095947, + "learning_rate": 7.569273175685479e-06, + "loss": 0.0039, + "step": 17237 + }, + { + "epoch": 1.868820468343452, + "grad_norm": 3.3037681579589844, + "learning_rate": 7.568547802118092e-06, + "loss": 0.0348, + "step": 17238 + }, + { + "epoch": 1.8689288811795317, + "grad_norm": 1.6411728858947754, + "learning_rate": 7.567822428550704e-06, + "loss": 0.0157, + "step": 17239 + }, + { + "epoch": 1.8690372940156115, + "grad_norm": 1.5478606224060059, + "learning_rate": 7.567097054983317e-06, + "loss": 0.0141, + "step": 17240 + }, + { + "epoch": 1.8691457068516912, + "grad_norm": 1.3330820798873901, + "learning_rate": 7.56637168141593e-06, + "loss": 0.0038, + "step": 17241 + }, + { + "epoch": 1.869254119687771, + "grad_norm": 0.46707528829574585, + "learning_rate": 7.565646307848543e-06, + "loss": 0.0024, + "step": 17242 + }, + { + "epoch": 1.869362532523851, + "grad_norm": 4.0599045753479, + "learning_rate": 7.564920934281156e-06, + "loss": 0.0194, + "step": 17243 + }, + { + "epoch": 1.8694709453599305, + "grad_norm": 3.1089694499969482, + "learning_rate": 7.564195560713769e-06, + "loss": 0.042, + "step": 17244 + }, + { + "epoch": 1.8695793581960105, + "grad_norm": 1.3496530055999756, + "learning_rate": 7.563470187146381e-06, + "loss": 0.0229, + "step": 17245 + }, + { + "epoch": 1.86968777103209, + "grad_norm": 3.035900592803955, + "learning_rate": 7.562744813578994e-06, + "loss": 0.0296, + "step": 17246 + }, + { + "epoch": 1.86979618386817, + "grad_norm": 4.970493316650391, + "learning_rate": 7.562019440011607e-06, + "loss": 0.0215, + "step": 17247 + }, + { + "epoch": 1.8699045967042498, + "grad_norm": 4.283820629119873, + "learning_rate": 7.561294066444219e-06, + "loss": 0.0073, + "step": 17248 + }, + { + "epoch": 1.8700130095403296, + "grad_norm": 6.270391941070557, + "learning_rate": 7.560568692876832e-06, + "loss": 0.0794, + "step": 17249 + }, + { + "epoch": 1.8701214223764093, + "grad_norm": 2.066253662109375, + "learning_rate": 7.559843319309444e-06, + "loss": 0.0288, + "step": 17250 + }, + { + "epoch": 1.870229835212489, + "grad_norm": 3.193162679672241, + "learning_rate": 7.559117945742058e-06, + "loss": 0.0186, + "step": 17251 + }, + { + "epoch": 1.870338248048569, + "grad_norm": 6.718310832977295, + "learning_rate": 7.558392572174671e-06, + "loss": 0.0593, + "step": 17252 + }, + { + "epoch": 1.8704466608846486, + "grad_norm": 1.5621917247772217, + "learning_rate": 7.557667198607284e-06, + "loss": 0.0237, + "step": 17253 + }, + { + "epoch": 1.8705550737207286, + "grad_norm": 4.866276741027832, + "learning_rate": 7.556941825039896e-06, + "loss": 0.067, + "step": 17254 + }, + { + "epoch": 1.8706634865568084, + "grad_norm": 3.195232629776001, + "learning_rate": 7.556216451472509e-06, + "loss": 0.0259, + "step": 17255 + }, + { + "epoch": 1.8707718993928881, + "grad_norm": 4.391593933105469, + "learning_rate": 7.555491077905121e-06, + "loss": 0.0488, + "step": 17256 + }, + { + "epoch": 1.870880312228968, + "grad_norm": 0.8941395878791809, + "learning_rate": 7.554765704337734e-06, + "loss": 0.0051, + "step": 17257 + }, + { + "epoch": 1.8709887250650477, + "grad_norm": 0.6475184559822083, + "learning_rate": 7.554040330770347e-06, + "loss": 0.0023, + "step": 17258 + }, + { + "epoch": 1.8710971379011276, + "grad_norm": 2.039268970489502, + "learning_rate": 7.553314957202959e-06, + "loss": 0.0333, + "step": 17259 + }, + { + "epoch": 1.8712055507372072, + "grad_norm": 1.6747503280639648, + "learning_rate": 7.552589583635573e-06, + "loss": 0.0086, + "step": 17260 + }, + { + "epoch": 1.8713139635732872, + "grad_norm": 9.82577133178711, + "learning_rate": 7.551864210068186e-06, + "loss": 0.0294, + "step": 17261 + }, + { + "epoch": 1.8714223764093667, + "grad_norm": 4.107144832611084, + "learning_rate": 7.551138836500798e-06, + "loss": 0.0321, + "step": 17262 + }, + { + "epoch": 1.8715307892454467, + "grad_norm": 4.410767555236816, + "learning_rate": 7.550413462933411e-06, + "loss": 0.0317, + "step": 17263 + }, + { + "epoch": 1.8716392020815265, + "grad_norm": 0.9294456243515015, + "learning_rate": 7.549688089366024e-06, + "loss": 0.0089, + "step": 17264 + }, + { + "epoch": 1.8717476149176062, + "grad_norm": 3.6823294162750244, + "learning_rate": 7.5489627157986365e-06, + "loss": 0.065, + "step": 17265 + }, + { + "epoch": 1.8718560277536862, + "grad_norm": 0.13397729396820068, + "learning_rate": 7.5482373422312495e-06, + "loss": 0.0013, + "step": 17266 + }, + { + "epoch": 1.8719644405897657, + "grad_norm": 4.618931293487549, + "learning_rate": 7.5475119686638624e-06, + "loss": 0.0688, + "step": 17267 + }, + { + "epoch": 1.8720728534258457, + "grad_norm": 2.5773122310638428, + "learning_rate": 7.5467865950964754e-06, + "loss": 0.0192, + "step": 17268 + }, + { + "epoch": 1.8721812662619253, + "grad_norm": 0.30036333203315735, + "learning_rate": 7.5460612215290884e-06, + "loss": 0.0038, + "step": 17269 + }, + { + "epoch": 1.8722896790980053, + "grad_norm": 0.7611974477767944, + "learning_rate": 7.545335847961701e-06, + "loss": 0.0034, + "step": 17270 + }, + { + "epoch": 1.872398091934085, + "grad_norm": 2.5700440406799316, + "learning_rate": 7.5446104743943136e-06, + "loss": 0.0153, + "step": 17271 + }, + { + "epoch": 1.8725065047701648, + "grad_norm": 0.13276644051074982, + "learning_rate": 7.5438851008269266e-06, + "loss": 0.0016, + "step": 17272 + }, + { + "epoch": 1.8726149176062445, + "grad_norm": 4.549318313598633, + "learning_rate": 7.5431597272595395e-06, + "loss": 0.0473, + "step": 17273 + }, + { + "epoch": 1.8727233304423243, + "grad_norm": 2.224360942840576, + "learning_rate": 7.542434353692152e-06, + "loss": 0.0074, + "step": 17274 + }, + { + "epoch": 1.8728317432784043, + "grad_norm": 2.4197630882263184, + "learning_rate": 7.541708980124765e-06, + "loss": 0.0138, + "step": 17275 + }, + { + "epoch": 1.8729401561144838, + "grad_norm": 1.071427822113037, + "learning_rate": 7.540983606557377e-06, + "loss": 0.0102, + "step": 17276 + }, + { + "epoch": 1.8730485689505638, + "grad_norm": 1.0246864557266235, + "learning_rate": 7.540258232989991e-06, + "loss": 0.0086, + "step": 17277 + }, + { + "epoch": 1.8731569817866436, + "grad_norm": 1.4551492929458618, + "learning_rate": 7.539532859422604e-06, + "loss": 0.007, + "step": 17278 + }, + { + "epoch": 1.8732653946227233, + "grad_norm": 1.7351402044296265, + "learning_rate": 7.538807485855217e-06, + "loss": 0.0162, + "step": 17279 + }, + { + "epoch": 1.873373807458803, + "grad_norm": 0.05514990910887718, + "learning_rate": 7.538082112287829e-06, + "loss": 0.0005, + "step": 17280 + }, + { + "epoch": 1.8734822202948829, + "grad_norm": 4.695617198944092, + "learning_rate": 7.537356738720442e-06, + "loss": 0.0495, + "step": 17281 + }, + { + "epoch": 1.8735906331309629, + "grad_norm": 1.2546192407608032, + "learning_rate": 7.536631365153054e-06, + "loss": 0.0072, + "step": 17282 + }, + { + "epoch": 1.8736990459670424, + "grad_norm": 0.11432206630706787, + "learning_rate": 7.535905991585667e-06, + "loss": 0.0007, + "step": 17283 + }, + { + "epoch": 1.8738074588031224, + "grad_norm": 3.7071354389190674, + "learning_rate": 7.53518061801828e-06, + "loss": 0.0234, + "step": 17284 + }, + { + "epoch": 1.873915871639202, + "grad_norm": 0.1973024606704712, + "learning_rate": 7.534455244450894e-06, + "loss": 0.0012, + "step": 17285 + }, + { + "epoch": 1.874024284475282, + "grad_norm": 7.16916561126709, + "learning_rate": 7.533729870883506e-06, + "loss": 0.03, + "step": 17286 + }, + { + "epoch": 1.8741326973113617, + "grad_norm": 1.6401010751724243, + "learning_rate": 7.533004497316119e-06, + "loss": 0.0245, + "step": 17287 + }, + { + "epoch": 1.8742411101474414, + "grad_norm": 3.3125288486480713, + "learning_rate": 7.532279123748731e-06, + "loss": 0.0361, + "step": 17288 + }, + { + "epoch": 1.8743495229835212, + "grad_norm": 0.05330287665128708, + "learning_rate": 7.531553750181344e-06, + "loss": 0.0008, + "step": 17289 + }, + { + "epoch": 1.874457935819601, + "grad_norm": 2.7562880516052246, + "learning_rate": 7.530828376613957e-06, + "loss": 0.0532, + "step": 17290 + }, + { + "epoch": 1.874566348655681, + "grad_norm": 0.35510122776031494, + "learning_rate": 7.530103003046569e-06, + "loss": 0.0027, + "step": 17291 + }, + { + "epoch": 1.8746747614917605, + "grad_norm": 4.473546504974365, + "learning_rate": 7.529377629479182e-06, + "loss": 0.0184, + "step": 17292 + }, + { + "epoch": 1.8747831743278405, + "grad_norm": 8.553990364074707, + "learning_rate": 7.528652255911795e-06, + "loss": 0.0662, + "step": 17293 + }, + { + "epoch": 1.8748915871639202, + "grad_norm": 2.024212121963501, + "learning_rate": 7.527926882344408e-06, + "loss": 0.0275, + "step": 17294 + }, + { + "epoch": 1.875, + "grad_norm": 1.2941265106201172, + "learning_rate": 7.527201508777021e-06, + "loss": 0.0126, + "step": 17295 + }, + { + "epoch": 1.8751084128360798, + "grad_norm": 1.352506399154663, + "learning_rate": 7.526476135209634e-06, + "loss": 0.0078, + "step": 17296 + }, + { + "epoch": 1.8752168256721595, + "grad_norm": 5.214744567871094, + "learning_rate": 7.525750761642246e-06, + "loss": 0.0976, + "step": 17297 + }, + { + "epoch": 1.8753252385082395, + "grad_norm": 1.0564732551574707, + "learning_rate": 7.525025388074859e-06, + "loss": 0.0032, + "step": 17298 + }, + { + "epoch": 1.875433651344319, + "grad_norm": 2.8392539024353027, + "learning_rate": 7.524300014507472e-06, + "loss": 0.0329, + "step": 17299 + }, + { + "epoch": 1.875542064180399, + "grad_norm": 2.4988787174224854, + "learning_rate": 7.523574640940084e-06, + "loss": 0.0145, + "step": 17300 + }, + { + "epoch": 1.8756504770164788, + "grad_norm": 2.1062467098236084, + "learning_rate": 7.522849267372697e-06, + "loss": 0.0052, + "step": 17301 + }, + { + "epoch": 1.8757588898525586, + "grad_norm": 2.277268171310425, + "learning_rate": 7.5221238938053095e-06, + "loss": 0.0202, + "step": 17302 + }, + { + "epoch": 1.8758673026886383, + "grad_norm": 1.5539021492004395, + "learning_rate": 7.521398520237923e-06, + "loss": 0.0164, + "step": 17303 + }, + { + "epoch": 1.875975715524718, + "grad_norm": 4.996764183044434, + "learning_rate": 7.520673146670536e-06, + "loss": 0.0737, + "step": 17304 + }, + { + "epoch": 1.876084128360798, + "grad_norm": 2.2481703758239746, + "learning_rate": 7.519947773103149e-06, + "loss": 0.0164, + "step": 17305 + }, + { + "epoch": 1.8761925411968776, + "grad_norm": 0.8444671630859375, + "learning_rate": 7.5192223995357614e-06, + "loss": 0.0057, + "step": 17306 + }, + { + "epoch": 1.8763009540329576, + "grad_norm": 2.0228238105773926, + "learning_rate": 7.5184970259683744e-06, + "loss": 0.007, + "step": 17307 + }, + { + "epoch": 1.8764093668690371, + "grad_norm": 3.325303316116333, + "learning_rate": 7.517771652400987e-06, + "loss": 0.03, + "step": 17308 + }, + { + "epoch": 1.8765177797051171, + "grad_norm": 2.8865764141082764, + "learning_rate": 7.5170462788335996e-06, + "loss": 0.0413, + "step": 17309 + }, + { + "epoch": 1.876626192541197, + "grad_norm": 0.5795623064041138, + "learning_rate": 7.5163209052662126e-06, + "loss": 0.0079, + "step": 17310 + }, + { + "epoch": 1.8767346053772767, + "grad_norm": 2.8512721061706543, + "learning_rate": 7.515595531698826e-06, + "loss": 0.0405, + "step": 17311 + }, + { + "epoch": 1.8768430182133564, + "grad_norm": 4.1286797523498535, + "learning_rate": 7.5148701581314385e-06, + "loss": 0.0397, + "step": 17312 + }, + { + "epoch": 1.8769514310494362, + "grad_norm": 1.537238597869873, + "learning_rate": 7.5141447845640515e-06, + "loss": 0.0173, + "step": 17313 + }, + { + "epoch": 1.8770598438855162, + "grad_norm": 1.4888556003570557, + "learning_rate": 7.513419410996664e-06, + "loss": 0.0162, + "step": 17314 + }, + { + "epoch": 1.8771682567215957, + "grad_norm": 4.363932132720947, + "learning_rate": 7.512694037429277e-06, + "loss": 0.0092, + "step": 17315 + }, + { + "epoch": 1.8772766695576757, + "grad_norm": 4.267935276031494, + "learning_rate": 7.51196866386189e-06, + "loss": 0.0362, + "step": 17316 + }, + { + "epoch": 1.8773850823937555, + "grad_norm": 2.965588092803955, + "learning_rate": 7.511243290294502e-06, + "loss": 0.0138, + "step": 17317 + }, + { + "epoch": 1.8774934952298352, + "grad_norm": 18.59369659423828, + "learning_rate": 7.510517916727115e-06, + "loss": 0.0373, + "step": 17318 + }, + { + "epoch": 1.877601908065915, + "grad_norm": 2.1867213249206543, + "learning_rate": 7.509792543159728e-06, + "loss": 0.0055, + "step": 17319 + }, + { + "epoch": 1.8777103209019947, + "grad_norm": 0.10777326673269272, + "learning_rate": 7.509067169592341e-06, + "loss": 0.0009, + "step": 17320 + }, + { + "epoch": 1.8778187337380747, + "grad_norm": 3.5245745182037354, + "learning_rate": 7.508341796024954e-06, + "loss": 0.0427, + "step": 17321 + }, + { + "epoch": 1.8779271465741543, + "grad_norm": 1.9683669805526733, + "learning_rate": 7.507616422457567e-06, + "loss": 0.0431, + "step": 17322 + }, + { + "epoch": 1.8780355594102343, + "grad_norm": 1.6527140140533447, + "learning_rate": 7.506891048890179e-06, + "loss": 0.0142, + "step": 17323 + }, + { + "epoch": 1.8781439722463138, + "grad_norm": 1.6315016746520996, + "learning_rate": 7.506165675322792e-06, + "loss": 0.0066, + "step": 17324 + }, + { + "epoch": 1.8782523850823938, + "grad_norm": 3.0920515060424805, + "learning_rate": 7.505440301755405e-06, + "loss": 0.0048, + "step": 17325 + }, + { + "epoch": 1.8783607979184735, + "grad_norm": 1.6991171836853027, + "learning_rate": 7.504714928188017e-06, + "loss": 0.009, + "step": 17326 + }, + { + "epoch": 1.8784692107545533, + "grad_norm": 1.5603715181350708, + "learning_rate": 7.50398955462063e-06, + "loss": 0.0256, + "step": 17327 + }, + { + "epoch": 1.8785776235906333, + "grad_norm": 1.4882196187973022, + "learning_rate": 7.503264181053242e-06, + "loss": 0.0096, + "step": 17328 + }, + { + "epoch": 1.8786860364267128, + "grad_norm": 1.091961145401001, + "learning_rate": 7.502538807485856e-06, + "loss": 0.008, + "step": 17329 + }, + { + "epoch": 1.8787944492627928, + "grad_norm": 3.2331762313842773, + "learning_rate": 7.501813433918469e-06, + "loss": 0.0312, + "step": 17330 + }, + { + "epoch": 1.8789028620988724, + "grad_norm": 0.055648889392614365, + "learning_rate": 7.501088060351082e-06, + "loss": 0.0006, + "step": 17331 + }, + { + "epoch": 1.8790112749349523, + "grad_norm": 0.10603823512792587, + "learning_rate": 7.500362686783694e-06, + "loss": 0.0008, + "step": 17332 + }, + { + "epoch": 1.879119687771032, + "grad_norm": 1.3315225839614868, + "learning_rate": 7.499637313216307e-06, + "loss": 0.004, + "step": 17333 + }, + { + "epoch": 1.8792281006071119, + "grad_norm": 2.6935620307922363, + "learning_rate": 7.498911939648919e-06, + "loss": 0.024, + "step": 17334 + }, + { + "epoch": 1.8793365134431916, + "grad_norm": 0.16891679167747498, + "learning_rate": 7.498186566081532e-06, + "loss": 0.0013, + "step": 17335 + }, + { + "epoch": 1.8794449262792714, + "grad_norm": 0.9021861553192139, + "learning_rate": 7.497461192514145e-06, + "loss": 0.0043, + "step": 17336 + }, + { + "epoch": 1.8795533391153514, + "grad_norm": 1.8818501234054565, + "learning_rate": 7.496735818946759e-06, + "loss": 0.0443, + "step": 17337 + }, + { + "epoch": 1.879661751951431, + "grad_norm": 2.1572136878967285, + "learning_rate": 7.496010445379371e-06, + "loss": 0.024, + "step": 17338 + }, + { + "epoch": 1.879770164787511, + "grad_norm": 1.9036645889282227, + "learning_rate": 7.495285071811984e-06, + "loss": 0.0219, + "step": 17339 + }, + { + "epoch": 1.8798785776235907, + "grad_norm": 0.04466194286942482, + "learning_rate": 7.494559698244596e-06, + "loss": 0.0006, + "step": 17340 + }, + { + "epoch": 1.8799869904596704, + "grad_norm": 5.936683177947998, + "learning_rate": 7.493834324677209e-06, + "loss": 0.0237, + "step": 17341 + }, + { + "epoch": 1.8800954032957502, + "grad_norm": 3.482233762741089, + "learning_rate": 7.493108951109822e-06, + "loss": 0.0425, + "step": 17342 + }, + { + "epoch": 1.88020381613183, + "grad_norm": 0.7636136412620544, + "learning_rate": 7.4923835775424345e-06, + "loss": 0.0033, + "step": 17343 + }, + { + "epoch": 1.88031222896791, + "grad_norm": 1.9289474487304688, + "learning_rate": 7.4916582039750475e-06, + "loss": 0.0514, + "step": 17344 + }, + { + "epoch": 1.8804206418039895, + "grad_norm": 2.595578193664551, + "learning_rate": 7.4909328304076605e-06, + "loss": 0.0221, + "step": 17345 + }, + { + "epoch": 1.8805290546400695, + "grad_norm": 0.8645201325416565, + "learning_rate": 7.4902074568402734e-06, + "loss": 0.0055, + "step": 17346 + }, + { + "epoch": 1.880637467476149, + "grad_norm": 7.155272483825684, + "learning_rate": 7.4894820832728864e-06, + "loss": 0.0171, + "step": 17347 + }, + { + "epoch": 1.880745880312229, + "grad_norm": 5.833549499511719, + "learning_rate": 7.4887567097054994e-06, + "loss": 0.0639, + "step": 17348 + }, + { + "epoch": 1.8808542931483088, + "grad_norm": 2.311291456222534, + "learning_rate": 7.4880313361381116e-06, + "loss": 0.037, + "step": 17349 + }, + { + "epoch": 1.8809627059843885, + "grad_norm": 2.013927698135376, + "learning_rate": 7.4873059625707246e-06, + "loss": 0.0251, + "step": 17350 + }, + { + "epoch": 1.8810711188204683, + "grad_norm": 4.3479413986206055, + "learning_rate": 7.4865805890033375e-06, + "loss": 0.0528, + "step": 17351 + }, + { + "epoch": 1.881179531656548, + "grad_norm": 0.42808860540390015, + "learning_rate": 7.48585521543595e-06, + "loss": 0.0036, + "step": 17352 + }, + { + "epoch": 1.881287944492628, + "grad_norm": 0.4192502796649933, + "learning_rate": 7.485129841868563e-06, + "loss": 0.003, + "step": 17353 + }, + { + "epoch": 1.8813963573287076, + "grad_norm": 2.0486128330230713, + "learning_rate": 7.484404468301175e-06, + "loss": 0.0193, + "step": 17354 + }, + { + "epoch": 1.8815047701647876, + "grad_norm": 0.7880284190177917, + "learning_rate": 7.483679094733789e-06, + "loss": 0.0067, + "step": 17355 + }, + { + "epoch": 1.8816131830008673, + "grad_norm": 4.960996627807617, + "learning_rate": 7.482953721166402e-06, + "loss": 0.0561, + "step": 17356 + }, + { + "epoch": 1.881721595836947, + "grad_norm": 2.38726544380188, + "learning_rate": 7.482228347599015e-06, + "loss": 0.0147, + "step": 17357 + }, + { + "epoch": 1.8818300086730269, + "grad_norm": 3.3213918209075928, + "learning_rate": 7.481502974031627e-06, + "loss": 0.0109, + "step": 17358 + }, + { + "epoch": 1.8819384215091066, + "grad_norm": 0.22575394809246063, + "learning_rate": 7.48077760046424e-06, + "loss": 0.0027, + "step": 17359 + }, + { + "epoch": 1.8820468343451866, + "grad_norm": 0.9715144038200378, + "learning_rate": 7.480052226896852e-06, + "loss": 0.0173, + "step": 17360 + }, + { + "epoch": 1.8821552471812661, + "grad_norm": 1.736095666885376, + "learning_rate": 7.479326853329465e-06, + "loss": 0.0151, + "step": 17361 + }, + { + "epoch": 1.8822636600173461, + "grad_norm": 1.2937730550765991, + "learning_rate": 7.478601479762078e-06, + "loss": 0.0528, + "step": 17362 + }, + { + "epoch": 1.882372072853426, + "grad_norm": 1.4507845640182495, + "learning_rate": 7.477876106194692e-06, + "loss": 0.0115, + "step": 17363 + }, + { + "epoch": 1.8824804856895057, + "grad_norm": 1.1545215845108032, + "learning_rate": 7.477150732627304e-06, + "loss": 0.032, + "step": 17364 + }, + { + "epoch": 1.8825888985255854, + "grad_norm": 0.6761497259140015, + "learning_rate": 7.476425359059917e-06, + "loss": 0.0071, + "step": 17365 + }, + { + "epoch": 1.8826973113616652, + "grad_norm": 2.859927177429199, + "learning_rate": 7.475699985492529e-06, + "loss": 0.0409, + "step": 17366 + }, + { + "epoch": 1.8828057241977452, + "grad_norm": 0.4606533348560333, + "learning_rate": 7.474974611925142e-06, + "loss": 0.0039, + "step": 17367 + }, + { + "epoch": 1.8829141370338247, + "grad_norm": 3.3956642150878906, + "learning_rate": 7.474249238357755e-06, + "loss": 0.0266, + "step": 17368 + }, + { + "epoch": 1.8830225498699047, + "grad_norm": 2.6618456840515137, + "learning_rate": 7.473523864790367e-06, + "loss": 0.0244, + "step": 17369 + }, + { + "epoch": 1.8831309627059842, + "grad_norm": 2.7171199321746826, + "learning_rate": 7.47279849122298e-06, + "loss": 0.0252, + "step": 17370 + }, + { + "epoch": 1.8832393755420642, + "grad_norm": 0.17447051405906677, + "learning_rate": 7.472073117655593e-06, + "loss": 0.0014, + "step": 17371 + }, + { + "epoch": 1.883347788378144, + "grad_norm": 2.519047975540161, + "learning_rate": 7.471347744088206e-06, + "loss": 0.0225, + "step": 17372 + }, + { + "epoch": 1.8834562012142237, + "grad_norm": 1.9099239110946655, + "learning_rate": 7.470622370520819e-06, + "loss": 0.0162, + "step": 17373 + }, + { + "epoch": 1.8835646140503035, + "grad_norm": 0.09847882390022278, + "learning_rate": 7.469896996953432e-06, + "loss": 0.0008, + "step": 17374 + }, + { + "epoch": 1.8836730268863833, + "grad_norm": 6.934532642364502, + "learning_rate": 7.469171623386044e-06, + "loss": 0.0992, + "step": 17375 + }, + { + "epoch": 1.8837814397224633, + "grad_norm": 1.2279016971588135, + "learning_rate": 7.468446249818657e-06, + "loss": 0.0217, + "step": 17376 + }, + { + "epoch": 1.8838898525585428, + "grad_norm": 1.5360244512557983, + "learning_rate": 7.46772087625127e-06, + "loss": 0.0194, + "step": 17377 + }, + { + "epoch": 1.8839982653946228, + "grad_norm": 0.5730947256088257, + "learning_rate": 7.466995502683882e-06, + "loss": 0.003, + "step": 17378 + }, + { + "epoch": 1.8841066782307025, + "grad_norm": 4.710333347320557, + "learning_rate": 7.466270129116495e-06, + "loss": 0.0291, + "step": 17379 + }, + { + "epoch": 1.8842150910667823, + "grad_norm": 1.1371976137161255, + "learning_rate": 7.4655447555491075e-06, + "loss": 0.0059, + "step": 17380 + }, + { + "epoch": 1.884323503902862, + "grad_norm": 0.2525654435157776, + "learning_rate": 7.464819381981721e-06, + "loss": 0.0027, + "step": 17381 + }, + { + "epoch": 1.8844319167389418, + "grad_norm": 2.7018136978149414, + "learning_rate": 7.464094008414334e-06, + "loss": 0.0293, + "step": 17382 + }, + { + "epoch": 1.8845403295750218, + "grad_norm": 5.710321426391602, + "learning_rate": 7.463368634846947e-06, + "loss": 0.0269, + "step": 17383 + }, + { + "epoch": 1.8846487424111014, + "grad_norm": 1.435370683670044, + "learning_rate": 7.4626432612795595e-06, + "loss": 0.0291, + "step": 17384 + }, + { + "epoch": 1.8847571552471813, + "grad_norm": 14.429083824157715, + "learning_rate": 7.4619178877121724e-06, + "loss": 0.0145, + "step": 17385 + }, + { + "epoch": 1.8848655680832611, + "grad_norm": 1.7255991697311401, + "learning_rate": 7.461192514144785e-06, + "loss": 0.0089, + "step": 17386 + }, + { + "epoch": 1.8849739809193409, + "grad_norm": 2.4091315269470215, + "learning_rate": 7.460467140577398e-06, + "loss": 0.0118, + "step": 17387 + }, + { + "epoch": 1.8850823937554206, + "grad_norm": 2.30946683883667, + "learning_rate": 7.4597417670100106e-06, + "loss": 0.0284, + "step": 17388 + }, + { + "epoch": 1.8851908065915004, + "grad_norm": 1.4221382141113281, + "learning_rate": 7.459016393442624e-06, + "loss": 0.0122, + "step": 17389 + }, + { + "epoch": 1.8852992194275804, + "grad_norm": 6.247662544250488, + "learning_rate": 7.4582910198752366e-06, + "loss": 0.0216, + "step": 17390 + }, + { + "epoch": 1.88540763226366, + "grad_norm": 1.404632806777954, + "learning_rate": 7.4575656463078495e-06, + "loss": 0.0112, + "step": 17391 + }, + { + "epoch": 1.88551604509974, + "grad_norm": 1.147642731666565, + "learning_rate": 7.456840272740462e-06, + "loss": 0.0071, + "step": 17392 + }, + { + "epoch": 1.8856244579358195, + "grad_norm": 0.4834684133529663, + "learning_rate": 7.456114899173075e-06, + "loss": 0.0021, + "step": 17393 + }, + { + "epoch": 1.8857328707718994, + "grad_norm": 2.161438226699829, + "learning_rate": 7.455389525605688e-06, + "loss": 0.0712, + "step": 17394 + }, + { + "epoch": 1.8858412836079792, + "grad_norm": 3.8350696563720703, + "learning_rate": 7.4546641520383e-06, + "loss": 0.0148, + "step": 17395 + }, + { + "epoch": 1.885949696444059, + "grad_norm": 1.5314545631408691, + "learning_rate": 7.453938778470913e-06, + "loss": 0.0363, + "step": 17396 + }, + { + "epoch": 1.8860581092801387, + "grad_norm": 0.6576012372970581, + "learning_rate": 7.453213404903526e-06, + "loss": 0.0077, + "step": 17397 + }, + { + "epoch": 1.8861665221162185, + "grad_norm": 0.29112508893013, + "learning_rate": 7.452488031336139e-06, + "loss": 0.0039, + "step": 17398 + }, + { + "epoch": 1.8862749349522985, + "grad_norm": 1.5359702110290527, + "learning_rate": 7.451762657768752e-06, + "loss": 0.0062, + "step": 17399 + }, + { + "epoch": 1.886383347788378, + "grad_norm": 5.49754524230957, + "learning_rate": 7.451037284201365e-06, + "loss": 0.0456, + "step": 17400 + }, + { + "epoch": 1.886491760624458, + "grad_norm": 0.6934069395065308, + "learning_rate": 7.450311910633977e-06, + "loss": 0.0114, + "step": 17401 + }, + { + "epoch": 1.8866001734605378, + "grad_norm": 3.2438957691192627, + "learning_rate": 7.44958653706659e-06, + "loss": 0.0236, + "step": 17402 + }, + { + "epoch": 1.8867085862966175, + "grad_norm": 1.9719233512878418, + "learning_rate": 7.448861163499203e-06, + "loss": 0.0389, + "step": 17403 + }, + { + "epoch": 1.8868169991326973, + "grad_norm": 4.622819900512695, + "learning_rate": 7.448135789931815e-06, + "loss": 0.009, + "step": 17404 + }, + { + "epoch": 1.886925411968777, + "grad_norm": 3.693091869354248, + "learning_rate": 7.447410416364428e-06, + "loss": 0.0119, + "step": 17405 + }, + { + "epoch": 1.887033824804857, + "grad_norm": 1.6652040481567383, + "learning_rate": 7.446685042797042e-06, + "loss": 0.0042, + "step": 17406 + }, + { + "epoch": 1.8871422376409366, + "grad_norm": 1.6897125244140625, + "learning_rate": 7.445959669229654e-06, + "loss": 0.0241, + "step": 17407 + }, + { + "epoch": 1.8872506504770166, + "grad_norm": 1.2606251239776611, + "learning_rate": 7.445234295662267e-06, + "loss": 0.0156, + "step": 17408 + }, + { + "epoch": 1.887359063313096, + "grad_norm": 1.9722497463226318, + "learning_rate": 7.44450892209488e-06, + "loss": 0.0174, + "step": 17409 + }, + { + "epoch": 1.887467476149176, + "grad_norm": 2.796170473098755, + "learning_rate": 7.443783548527492e-06, + "loss": 0.0307, + "step": 17410 + }, + { + "epoch": 1.8875758889852559, + "grad_norm": 0.2526628375053406, + "learning_rate": 7.443058174960105e-06, + "loss": 0.0017, + "step": 17411 + }, + { + "epoch": 1.8876843018213356, + "grad_norm": 1.2743130922317505, + "learning_rate": 7.442332801392717e-06, + "loss": 0.0211, + "step": 17412 + }, + { + "epoch": 1.8877927146574154, + "grad_norm": 7.716072082519531, + "learning_rate": 7.44160742782533e-06, + "loss": 0.0447, + "step": 17413 + }, + { + "epoch": 1.8879011274934951, + "grad_norm": 2.462378740310669, + "learning_rate": 7.440882054257943e-06, + "loss": 0.0278, + "step": 17414 + }, + { + "epoch": 1.8880095403295751, + "grad_norm": 1.8310927152633667, + "learning_rate": 7.440156680690557e-06, + "loss": 0.0111, + "step": 17415 + }, + { + "epoch": 1.8881179531656547, + "grad_norm": 1.4308953285217285, + "learning_rate": 7.439431307123169e-06, + "loss": 0.0055, + "step": 17416 + }, + { + "epoch": 1.8882263660017347, + "grad_norm": 1.648371696472168, + "learning_rate": 7.438705933555782e-06, + "loss": 0.0514, + "step": 17417 + }, + { + "epoch": 1.8883347788378144, + "grad_norm": 4.368140697479248, + "learning_rate": 7.437980559988394e-06, + "loss": 0.0312, + "step": 17418 + }, + { + "epoch": 1.8884431916738942, + "grad_norm": 0.09296803176403046, + "learning_rate": 7.437255186421007e-06, + "loss": 0.0006, + "step": 17419 + }, + { + "epoch": 1.888551604509974, + "grad_norm": 2.063948631286621, + "learning_rate": 7.43652981285362e-06, + "loss": 0.0319, + "step": 17420 + }, + { + "epoch": 1.8886600173460537, + "grad_norm": 2.878528594970703, + "learning_rate": 7.4358044392862325e-06, + "loss": 0.0194, + "step": 17421 + }, + { + "epoch": 1.8887684301821337, + "grad_norm": 2.1240899562835693, + "learning_rate": 7.4350790657188455e-06, + "loss": 0.0222, + "step": 17422 + }, + { + "epoch": 1.8888768430182132, + "grad_norm": 0.34869495034217834, + "learning_rate": 7.4343536921514585e-06, + "loss": 0.0032, + "step": 17423 + }, + { + "epoch": 1.8889852558542932, + "grad_norm": 1.279698133468628, + "learning_rate": 7.4336283185840714e-06, + "loss": 0.011, + "step": 17424 + }, + { + "epoch": 1.889093668690373, + "grad_norm": 2.736962080001831, + "learning_rate": 7.4329029450166844e-06, + "loss": 0.0397, + "step": 17425 + }, + { + "epoch": 1.8892020815264527, + "grad_norm": 5.85358190536499, + "learning_rate": 7.4321775714492974e-06, + "loss": 0.0681, + "step": 17426 + }, + { + "epoch": 1.8893104943625325, + "grad_norm": 2.7397642135620117, + "learning_rate": 7.4314521978819096e-06, + "loss": 0.0119, + "step": 17427 + }, + { + "epoch": 1.8894189071986123, + "grad_norm": 2.0946731567382812, + "learning_rate": 7.4307268243145226e-06, + "loss": 0.0616, + "step": 17428 + }, + { + "epoch": 1.8895273200346923, + "grad_norm": 0.8564862608909607, + "learning_rate": 7.4300014507471356e-06, + "loss": 0.0046, + "step": 17429 + }, + { + "epoch": 1.8896357328707718, + "grad_norm": 1.1197240352630615, + "learning_rate": 7.429276077179748e-06, + "loss": 0.0092, + "step": 17430 + }, + { + "epoch": 1.8897441457068518, + "grad_norm": 1.07188880443573, + "learning_rate": 7.428550703612361e-06, + "loss": 0.0276, + "step": 17431 + }, + { + "epoch": 1.8898525585429313, + "grad_norm": 6.429769992828369, + "learning_rate": 7.4278253300449745e-06, + "loss": 0.0307, + "step": 17432 + }, + { + "epoch": 1.8899609713790113, + "grad_norm": 0.15812969207763672, + "learning_rate": 7.427099956477587e-06, + "loss": 0.0011, + "step": 17433 + }, + { + "epoch": 1.890069384215091, + "grad_norm": 1.654502272605896, + "learning_rate": 7.4263745829102e-06, + "loss": 0.0198, + "step": 17434 + }, + { + "epoch": 1.8901777970511708, + "grad_norm": 5.360489368438721, + "learning_rate": 7.425649209342813e-06, + "loss": 0.0411, + "step": 17435 + }, + { + "epoch": 1.8902862098872506, + "grad_norm": 0.8052114844322205, + "learning_rate": 7.424923835775425e-06, + "loss": 0.0063, + "step": 17436 + }, + { + "epoch": 1.8903946227233304, + "grad_norm": 1.5710906982421875, + "learning_rate": 7.424198462208038e-06, + "loss": 0.0278, + "step": 17437 + }, + { + "epoch": 1.8905030355594103, + "grad_norm": 4.091601848602295, + "learning_rate": 7.42347308864065e-06, + "loss": 0.0263, + "step": 17438 + }, + { + "epoch": 1.89061144839549, + "grad_norm": 3.252335548400879, + "learning_rate": 7.422747715073263e-06, + "loss": 0.0128, + "step": 17439 + }, + { + "epoch": 1.8907198612315699, + "grad_norm": 0.23648996651172638, + "learning_rate": 7.422022341505876e-06, + "loss": 0.0027, + "step": 17440 + }, + { + "epoch": 1.8908282740676496, + "grad_norm": 2.9777774810791016, + "learning_rate": 7.42129696793849e-06, + "loss": 0.0317, + "step": 17441 + }, + { + "epoch": 1.8909366869037294, + "grad_norm": 1.7606195211410522, + "learning_rate": 7.420571594371102e-06, + "loss": 0.0241, + "step": 17442 + }, + { + "epoch": 1.8910450997398092, + "grad_norm": 4.986486911773682, + "learning_rate": 7.419846220803715e-06, + "loss": 0.0793, + "step": 17443 + }, + { + "epoch": 1.891153512575889, + "grad_norm": 4.352983474731445, + "learning_rate": 7.419120847236327e-06, + "loss": 0.0602, + "step": 17444 + }, + { + "epoch": 1.891261925411969, + "grad_norm": 0.22270509600639343, + "learning_rate": 7.41839547366894e-06, + "loss": 0.0015, + "step": 17445 + }, + { + "epoch": 1.8913703382480485, + "grad_norm": 0.810692310333252, + "learning_rate": 7.417670100101553e-06, + "loss": 0.0017, + "step": 17446 + }, + { + "epoch": 1.8914787510841284, + "grad_norm": 0.3764718770980835, + "learning_rate": 7.416944726534165e-06, + "loss": 0.0034, + "step": 17447 + }, + { + "epoch": 1.8915871639202082, + "grad_norm": 0.7203431725502014, + "learning_rate": 7.416219352966778e-06, + "loss": 0.0058, + "step": 17448 + }, + { + "epoch": 1.891695576756288, + "grad_norm": 2.422447681427002, + "learning_rate": 7.415493979399391e-06, + "loss": 0.0309, + "step": 17449 + }, + { + "epoch": 1.8918039895923677, + "grad_norm": 1.3449013233184814, + "learning_rate": 7.414768605832004e-06, + "loss": 0.007, + "step": 17450 + }, + { + "epoch": 1.8919124024284475, + "grad_norm": 0.41114503145217896, + "learning_rate": 7.414043232264617e-06, + "loss": 0.0024, + "step": 17451 + }, + { + "epoch": 1.8920208152645275, + "grad_norm": 3.87420916557312, + "learning_rate": 7.41331785869723e-06, + "loss": 0.0254, + "step": 17452 + }, + { + "epoch": 1.892129228100607, + "grad_norm": 2.171104669570923, + "learning_rate": 7.412592485129842e-06, + "loss": 0.0319, + "step": 17453 + }, + { + "epoch": 1.892237640936687, + "grad_norm": 2.5321121215820312, + "learning_rate": 7.411867111562455e-06, + "loss": 0.0086, + "step": 17454 + }, + { + "epoch": 1.8923460537727665, + "grad_norm": 0.2788935601711273, + "learning_rate": 7.411141737995068e-06, + "loss": 0.0017, + "step": 17455 + }, + { + "epoch": 1.8924544666088465, + "grad_norm": 5.479169845581055, + "learning_rate": 7.41041636442768e-06, + "loss": 0.0269, + "step": 17456 + }, + { + "epoch": 1.8925628794449263, + "grad_norm": 0.2527967691421509, + "learning_rate": 7.409690990860293e-06, + "loss": 0.0014, + "step": 17457 + }, + { + "epoch": 1.892671292281006, + "grad_norm": 5.844311714172363, + "learning_rate": 7.408965617292907e-06, + "loss": 0.0361, + "step": 17458 + }, + { + "epoch": 1.8927797051170858, + "grad_norm": 0.15758632123470306, + "learning_rate": 7.408240243725519e-06, + "loss": 0.0013, + "step": 17459 + }, + { + "epoch": 1.8928881179531656, + "grad_norm": 2.0839109420776367, + "learning_rate": 7.407514870158132e-06, + "loss": 0.0109, + "step": 17460 + }, + { + "epoch": 1.8929965307892456, + "grad_norm": 1.3872370719909668, + "learning_rate": 7.406789496590745e-06, + "loss": 0.0148, + "step": 17461 + }, + { + "epoch": 1.893104943625325, + "grad_norm": 10.123394966125488, + "learning_rate": 7.4060641230233575e-06, + "loss": 0.0476, + "step": 17462 + }, + { + "epoch": 1.893213356461405, + "grad_norm": 10.335095405578613, + "learning_rate": 7.4053387494559705e-06, + "loss": 0.0439, + "step": 17463 + }, + { + "epoch": 1.8933217692974849, + "grad_norm": 0.632166862487793, + "learning_rate": 7.404613375888583e-06, + "loss": 0.003, + "step": 17464 + }, + { + "epoch": 1.8934301821335646, + "grad_norm": 2.685464382171631, + "learning_rate": 7.403888002321196e-06, + "loss": 0.0135, + "step": 17465 + }, + { + "epoch": 1.8935385949696444, + "grad_norm": 1.3851423263549805, + "learning_rate": 7.403162628753809e-06, + "loss": 0.0118, + "step": 17466 + }, + { + "epoch": 1.8936470078057241, + "grad_norm": 1.5518893003463745, + "learning_rate": 7.402437255186422e-06, + "loss": 0.0221, + "step": 17467 + }, + { + "epoch": 1.8937554206418041, + "grad_norm": 0.7361441254615784, + "learning_rate": 7.4017118816190346e-06, + "loss": 0.0047, + "step": 17468 + }, + { + "epoch": 1.8938638334778837, + "grad_norm": 1.334823727607727, + "learning_rate": 7.4009865080516475e-06, + "loss": 0.0185, + "step": 17469 + }, + { + "epoch": 1.8939722463139637, + "grad_norm": 0.21788519620895386, + "learning_rate": 7.40026113448426e-06, + "loss": 0.0014, + "step": 17470 + }, + { + "epoch": 1.8940806591500432, + "grad_norm": 5.13580846786499, + "learning_rate": 7.399535760916873e-06, + "loss": 0.031, + "step": 17471 + }, + { + "epoch": 1.8941890719861232, + "grad_norm": 11.798277854919434, + "learning_rate": 7.398810387349486e-06, + "loss": 0.0335, + "step": 17472 + }, + { + "epoch": 1.894297484822203, + "grad_norm": 0.69049471616745, + "learning_rate": 7.398085013782098e-06, + "loss": 0.0051, + "step": 17473 + }, + { + "epoch": 1.8944058976582827, + "grad_norm": 2.4877710342407227, + "learning_rate": 7.397359640214711e-06, + "loss": 0.0074, + "step": 17474 + }, + { + "epoch": 1.8945143104943627, + "grad_norm": 0.09950143098831177, + "learning_rate": 7.396634266647324e-06, + "loss": 0.0014, + "step": 17475 + }, + { + "epoch": 1.8946227233304422, + "grad_norm": 0.7492039799690247, + "learning_rate": 7.395908893079937e-06, + "loss": 0.0058, + "step": 17476 + }, + { + "epoch": 1.8947311361665222, + "grad_norm": 8.109402656555176, + "learning_rate": 7.39518351951255e-06, + "loss": 0.0187, + "step": 17477 + }, + { + "epoch": 1.8948395490026018, + "grad_norm": 1.2185269594192505, + "learning_rate": 7.394458145945163e-06, + "loss": 0.0344, + "step": 17478 + }, + { + "epoch": 1.8949479618386817, + "grad_norm": 4.781225681304932, + "learning_rate": 7.393732772377775e-06, + "loss": 0.0222, + "step": 17479 + }, + { + "epoch": 1.8950563746747615, + "grad_norm": 4.24320125579834, + "learning_rate": 7.393007398810388e-06, + "loss": 0.0279, + "step": 17480 + }, + { + "epoch": 1.8951647875108413, + "grad_norm": 10.070176124572754, + "learning_rate": 7.392282025243001e-06, + "loss": 0.0263, + "step": 17481 + }, + { + "epoch": 1.895273200346921, + "grad_norm": 3.8749566078186035, + "learning_rate": 7.391556651675613e-06, + "loss": 0.0266, + "step": 17482 + }, + { + "epoch": 1.8953816131830008, + "grad_norm": 0.4587323069572449, + "learning_rate": 7.390831278108226e-06, + "loss": 0.0032, + "step": 17483 + }, + { + "epoch": 1.8954900260190808, + "grad_norm": 0.9812219738960266, + "learning_rate": 7.39010590454084e-06, + "loss": 0.0068, + "step": 17484 + }, + { + "epoch": 1.8955984388551603, + "grad_norm": 0.17159704864025116, + "learning_rate": 7.389380530973452e-06, + "loss": 0.0026, + "step": 17485 + }, + { + "epoch": 1.8957068516912403, + "grad_norm": 0.7975817322731018, + "learning_rate": 7.388655157406065e-06, + "loss": 0.0052, + "step": 17486 + }, + { + "epoch": 1.89581526452732, + "grad_norm": 1.7152094841003418, + "learning_rate": 7.387929783838678e-06, + "loss": 0.0321, + "step": 17487 + }, + { + "epoch": 1.8959236773633998, + "grad_norm": 1.0001168251037598, + "learning_rate": 7.38720441027129e-06, + "loss": 0.0098, + "step": 17488 + }, + { + "epoch": 1.8960320901994796, + "grad_norm": 7.694911003112793, + "learning_rate": 7.386479036703903e-06, + "loss": 0.0468, + "step": 17489 + }, + { + "epoch": 1.8961405030355594, + "grad_norm": 1.5088293552398682, + "learning_rate": 7.385753663136515e-06, + "loss": 0.0434, + "step": 17490 + }, + { + "epoch": 1.8962489158716394, + "grad_norm": 0.3801543414592743, + "learning_rate": 7.385028289569128e-06, + "loss": 0.0056, + "step": 17491 + }, + { + "epoch": 1.896357328707719, + "grad_norm": 0.13889333605766296, + "learning_rate": 7.384302916001741e-06, + "loss": 0.0018, + "step": 17492 + }, + { + "epoch": 1.8964657415437989, + "grad_norm": 10.393970489501953, + "learning_rate": 7.383577542434355e-06, + "loss": 0.0254, + "step": 17493 + }, + { + "epoch": 1.8965741543798784, + "grad_norm": 1.467965006828308, + "learning_rate": 7.382852168866967e-06, + "loss": 0.003, + "step": 17494 + }, + { + "epoch": 1.8966825672159584, + "grad_norm": 0.34045520424842834, + "learning_rate": 7.38212679529958e-06, + "loss": 0.0018, + "step": 17495 + }, + { + "epoch": 1.8967909800520382, + "grad_norm": 0.28792768716812134, + "learning_rate": 7.381401421732192e-06, + "loss": 0.0016, + "step": 17496 + }, + { + "epoch": 1.896899392888118, + "grad_norm": 7.7563042640686035, + "learning_rate": 7.380676048164805e-06, + "loss": 0.0127, + "step": 17497 + }, + { + "epoch": 1.8970078057241977, + "grad_norm": 0.6800099015235901, + "learning_rate": 7.379950674597418e-06, + "loss": 0.0026, + "step": 17498 + }, + { + "epoch": 1.8971162185602775, + "grad_norm": 5.657742977142334, + "learning_rate": 7.3792253010300305e-06, + "loss": 0.0324, + "step": 17499 + }, + { + "epoch": 1.8972246313963574, + "grad_norm": 4.05962610244751, + "learning_rate": 7.3784999274626435e-06, + "loss": 0.046, + "step": 17500 + }, + { + "epoch": 1.897333044232437, + "grad_norm": 7.324914932250977, + "learning_rate": 7.377774553895257e-06, + "loss": 0.0172, + "step": 17501 + }, + { + "epoch": 1.897441457068517, + "grad_norm": 0.4565218687057495, + "learning_rate": 7.3770491803278695e-06, + "loss": 0.0038, + "step": 17502 + }, + { + "epoch": 1.8975498699045967, + "grad_norm": 4.742059707641602, + "learning_rate": 7.3763238067604824e-06, + "loss": 0.0132, + "step": 17503 + }, + { + "epoch": 1.8976582827406765, + "grad_norm": 1.8583139181137085, + "learning_rate": 7.3755984331930954e-06, + "loss": 0.0426, + "step": 17504 + }, + { + "epoch": 1.8977666955767563, + "grad_norm": 2.5399231910705566, + "learning_rate": 7.374873059625708e-06, + "loss": 0.0334, + "step": 17505 + }, + { + "epoch": 1.897875108412836, + "grad_norm": 4.259757995605469, + "learning_rate": 7.3741476860583206e-06, + "loss": 0.01, + "step": 17506 + }, + { + "epoch": 1.897983521248916, + "grad_norm": 10.33245849609375, + "learning_rate": 7.3734223124909336e-06, + "loss": 0.0281, + "step": 17507 + }, + { + "epoch": 1.8980919340849955, + "grad_norm": 0.6957699060440063, + "learning_rate": 7.372696938923546e-06, + "loss": 0.015, + "step": 17508 + }, + { + "epoch": 1.8982003469210755, + "grad_norm": 1.8124042749404907, + "learning_rate": 7.371971565356159e-06, + "loss": 0.0105, + "step": 17509 + }, + { + "epoch": 1.8983087597571553, + "grad_norm": 0.06226573511958122, + "learning_rate": 7.3712461917887725e-06, + "loss": 0.0005, + "step": 17510 + }, + { + "epoch": 1.898417172593235, + "grad_norm": 2.9911468029022217, + "learning_rate": 7.370520818221385e-06, + "loss": 0.0151, + "step": 17511 + }, + { + "epoch": 1.8985255854293148, + "grad_norm": 3.2425143718719482, + "learning_rate": 7.369795444653998e-06, + "loss": 0.0063, + "step": 17512 + }, + { + "epoch": 1.8986339982653946, + "grad_norm": 2.862318515777588, + "learning_rate": 7.369070071086611e-06, + "loss": 0.0124, + "step": 17513 + }, + { + "epoch": 1.8987424111014746, + "grad_norm": 3.410914659500122, + "learning_rate": 7.368344697519223e-06, + "loss": 0.0316, + "step": 17514 + }, + { + "epoch": 1.898850823937554, + "grad_norm": 5.536984920501709, + "learning_rate": 7.367619323951836e-06, + "loss": 0.0072, + "step": 17515 + }, + { + "epoch": 1.898959236773634, + "grad_norm": 4.606231212615967, + "learning_rate": 7.366893950384448e-06, + "loss": 0.0259, + "step": 17516 + }, + { + "epoch": 1.8990676496097136, + "grad_norm": 5.310461044311523, + "learning_rate": 7.366168576817061e-06, + "loss": 0.0277, + "step": 17517 + }, + { + "epoch": 1.8991760624457936, + "grad_norm": 1.5646096467971802, + "learning_rate": 7.365443203249674e-06, + "loss": 0.0088, + "step": 17518 + }, + { + "epoch": 1.8992844752818734, + "grad_norm": 3.3263778686523438, + "learning_rate": 7.364717829682288e-06, + "loss": 0.0194, + "step": 17519 + }, + { + "epoch": 1.8993928881179531, + "grad_norm": 0.5566691160202026, + "learning_rate": 7.3639924561149e-06, + "loss": 0.0026, + "step": 17520 + }, + { + "epoch": 1.899501300954033, + "grad_norm": 0.27627038955688477, + "learning_rate": 7.363267082547513e-06, + "loss": 0.0017, + "step": 17521 + }, + { + "epoch": 1.8996097137901127, + "grad_norm": 0.4080143868923187, + "learning_rate": 7.362541708980125e-06, + "loss": 0.0032, + "step": 17522 + }, + { + "epoch": 1.8997181266261927, + "grad_norm": 3.008302927017212, + "learning_rate": 7.361816335412738e-06, + "loss": 0.0237, + "step": 17523 + }, + { + "epoch": 1.8998265394622722, + "grad_norm": 4.70236873626709, + "learning_rate": 7.361090961845351e-06, + "loss": 0.105, + "step": 17524 + }, + { + "epoch": 1.8999349522983522, + "grad_norm": 0.05502861365675926, + "learning_rate": 7.360365588277963e-06, + "loss": 0.0006, + "step": 17525 + }, + { + "epoch": 1.900043365134432, + "grad_norm": 0.18635234236717224, + "learning_rate": 7.359640214710576e-06, + "loss": 0.0025, + "step": 17526 + }, + { + "epoch": 1.9001517779705117, + "grad_norm": 2.3514177799224854, + "learning_rate": 7.35891484114319e-06, + "loss": 0.0043, + "step": 17527 + }, + { + "epoch": 1.9002601908065915, + "grad_norm": 1.4507683515548706, + "learning_rate": 7.358189467575802e-06, + "loss": 0.0111, + "step": 17528 + }, + { + "epoch": 1.9003686036426712, + "grad_norm": 2.193624258041382, + "learning_rate": 7.357464094008415e-06, + "loss": 0.0479, + "step": 17529 + }, + { + "epoch": 1.9004770164787512, + "grad_norm": 2.9661037921905518, + "learning_rate": 7.356738720441028e-06, + "loss": 0.0236, + "step": 17530 + }, + { + "epoch": 1.9005854293148308, + "grad_norm": 1.848466157913208, + "learning_rate": 7.35601334687364e-06, + "loss": 0.0145, + "step": 17531 + }, + { + "epoch": 1.9006938421509108, + "grad_norm": 2.8707141876220703, + "learning_rate": 7.355287973306253e-06, + "loss": 0.0357, + "step": 17532 + }, + { + "epoch": 1.9008022549869905, + "grad_norm": 5.16810941696167, + "learning_rate": 7.354562599738866e-06, + "loss": 0.0147, + "step": 17533 + }, + { + "epoch": 1.9009106678230703, + "grad_norm": 15.641731262207031, + "learning_rate": 7.353837226171478e-06, + "loss": 0.0504, + "step": 17534 + }, + { + "epoch": 1.90101908065915, + "grad_norm": 2.6474740505218506, + "learning_rate": 7.353111852604091e-06, + "loss": 0.0514, + "step": 17535 + }, + { + "epoch": 1.9011274934952298, + "grad_norm": 2.9874985218048096, + "learning_rate": 7.352386479036705e-06, + "loss": 0.0501, + "step": 17536 + }, + { + "epoch": 1.9012359063313098, + "grad_norm": 3.0567069053649902, + "learning_rate": 7.351661105469317e-06, + "loss": 0.0301, + "step": 17537 + }, + { + "epoch": 1.9013443191673893, + "grad_norm": 0.3048593997955322, + "learning_rate": 7.35093573190193e-06, + "loss": 0.0023, + "step": 17538 + }, + { + "epoch": 1.9014527320034693, + "grad_norm": 1.6217148303985596, + "learning_rate": 7.350210358334543e-06, + "loss": 0.0111, + "step": 17539 + }, + { + "epoch": 1.9015611448395489, + "grad_norm": 0.7280465960502625, + "learning_rate": 7.3494849847671555e-06, + "loss": 0.0031, + "step": 17540 + }, + { + "epoch": 1.9016695576756288, + "grad_norm": 1.887380838394165, + "learning_rate": 7.3487596111997685e-06, + "loss": 0.021, + "step": 17541 + }, + { + "epoch": 1.9017779705117086, + "grad_norm": 9.22594928741455, + "learning_rate": 7.348034237632381e-06, + "loss": 0.0537, + "step": 17542 + }, + { + "epoch": 1.9018863833477884, + "grad_norm": 8.194561004638672, + "learning_rate": 7.347308864064994e-06, + "loss": 0.0481, + "step": 17543 + }, + { + "epoch": 1.9019947961838681, + "grad_norm": 2.4320175647735596, + "learning_rate": 7.346583490497607e-06, + "loss": 0.0178, + "step": 17544 + }, + { + "epoch": 1.902103209019948, + "grad_norm": 3.109595775604248, + "learning_rate": 7.34585811693022e-06, + "loss": 0.0212, + "step": 17545 + }, + { + "epoch": 1.9022116218560279, + "grad_norm": 0.7429375648498535, + "learning_rate": 7.3451327433628326e-06, + "loss": 0.0087, + "step": 17546 + }, + { + "epoch": 1.9023200346921074, + "grad_norm": 3.9666337966918945, + "learning_rate": 7.3444073697954456e-06, + "loss": 0.0251, + "step": 17547 + }, + { + "epoch": 1.9024284475281874, + "grad_norm": 0.4743509590625763, + "learning_rate": 7.343681996228058e-06, + "loss": 0.0021, + "step": 17548 + }, + { + "epoch": 1.9025368603642672, + "grad_norm": 1.3134931325912476, + "learning_rate": 7.342956622660671e-06, + "loss": 0.0296, + "step": 17549 + }, + { + "epoch": 1.902645273200347, + "grad_norm": 1.5297369956970215, + "learning_rate": 7.342231249093284e-06, + "loss": 0.0233, + "step": 17550 + }, + { + "epoch": 1.9027536860364267, + "grad_norm": 0.9614200592041016, + "learning_rate": 7.341505875525896e-06, + "loss": 0.0029, + "step": 17551 + }, + { + "epoch": 1.9028620988725065, + "grad_norm": 2.6818227767944336, + "learning_rate": 7.340780501958509e-06, + "loss": 0.0531, + "step": 17552 + }, + { + "epoch": 1.9029705117085864, + "grad_norm": 1.6482627391815186, + "learning_rate": 7.340055128391123e-06, + "loss": 0.0094, + "step": 17553 + }, + { + "epoch": 1.903078924544666, + "grad_norm": 0.2591401934623718, + "learning_rate": 7.339329754823736e-06, + "loss": 0.002, + "step": 17554 + }, + { + "epoch": 1.903187337380746, + "grad_norm": 3.2794976234436035, + "learning_rate": 7.338604381256348e-06, + "loss": 0.0149, + "step": 17555 + }, + { + "epoch": 1.9032957502168255, + "grad_norm": 2.5034472942352295, + "learning_rate": 7.337879007688961e-06, + "loss": 0.0235, + "step": 17556 + }, + { + "epoch": 1.9034041630529055, + "grad_norm": 3.1863889694213867, + "learning_rate": 7.337153634121573e-06, + "loss": 0.0402, + "step": 17557 + }, + { + "epoch": 1.9035125758889853, + "grad_norm": 2.542295455932617, + "learning_rate": 7.336428260554186e-06, + "loss": 0.0085, + "step": 17558 + }, + { + "epoch": 1.903620988725065, + "grad_norm": 2.012583017349243, + "learning_rate": 7.335702886986799e-06, + "loss": 0.0142, + "step": 17559 + }, + { + "epoch": 1.9037294015611448, + "grad_norm": 2.4263343811035156, + "learning_rate": 7.334977513419411e-06, + "loss": 0.0084, + "step": 17560 + }, + { + "epoch": 1.9038378143972245, + "grad_norm": 0.12059416621923447, + "learning_rate": 7.334252139852024e-06, + "loss": 0.001, + "step": 17561 + }, + { + "epoch": 1.9039462272333045, + "grad_norm": 1.1151505708694458, + "learning_rate": 7.333526766284638e-06, + "loss": 0.0047, + "step": 17562 + }, + { + "epoch": 1.904054640069384, + "grad_norm": 0.9125990271568298, + "learning_rate": 7.33280139271725e-06, + "loss": 0.0096, + "step": 17563 + }, + { + "epoch": 1.904163052905464, + "grad_norm": 3.3207085132598877, + "learning_rate": 7.332076019149863e-06, + "loss": 0.0672, + "step": 17564 + }, + { + "epoch": 1.9042714657415438, + "grad_norm": 0.2965448200702667, + "learning_rate": 7.331350645582476e-06, + "loss": 0.0022, + "step": 17565 + }, + { + "epoch": 1.9043798785776236, + "grad_norm": 11.943753242492676, + "learning_rate": 7.330625272015088e-06, + "loss": 0.0192, + "step": 17566 + }, + { + "epoch": 1.9044882914137033, + "grad_norm": 1.3521921634674072, + "learning_rate": 7.329899898447701e-06, + "loss": 0.017, + "step": 17567 + }, + { + "epoch": 1.9045967042497831, + "grad_norm": 3.595081090927124, + "learning_rate": 7.329174524880313e-06, + "loss": 0.0162, + "step": 17568 + }, + { + "epoch": 1.904705117085863, + "grad_norm": 1.2720329761505127, + "learning_rate": 7.328449151312926e-06, + "loss": 0.0051, + "step": 17569 + }, + { + "epoch": 1.9048135299219426, + "grad_norm": 2.7604944705963135, + "learning_rate": 7.327723777745539e-06, + "loss": 0.0151, + "step": 17570 + }, + { + "epoch": 1.9049219427580226, + "grad_norm": 2.227463722229004, + "learning_rate": 7.326998404178153e-06, + "loss": 0.0467, + "step": 17571 + }, + { + "epoch": 1.9050303555941024, + "grad_norm": 0.1616968959569931, + "learning_rate": 7.326273030610765e-06, + "loss": 0.0014, + "step": 17572 + }, + { + "epoch": 1.9051387684301822, + "grad_norm": 4.1878886222839355, + "learning_rate": 7.325547657043378e-06, + "loss": 0.0101, + "step": 17573 + }, + { + "epoch": 1.905247181266262, + "grad_norm": 5.522558212280273, + "learning_rate": 7.32482228347599e-06, + "loss": 0.062, + "step": 17574 + }, + { + "epoch": 1.9053555941023417, + "grad_norm": 1.033692479133606, + "learning_rate": 7.324096909908603e-06, + "loss": 0.0033, + "step": 17575 + }, + { + "epoch": 1.9054640069384217, + "grad_norm": 0.44079825282096863, + "learning_rate": 7.323371536341216e-06, + "loss": 0.0062, + "step": 17576 + }, + { + "epoch": 1.9055724197745012, + "grad_norm": 2.0052716732025146, + "learning_rate": 7.3226461627738285e-06, + "loss": 0.0178, + "step": 17577 + }, + { + "epoch": 1.9056808326105812, + "grad_norm": 5.190842628479004, + "learning_rate": 7.3219207892064415e-06, + "loss": 0.062, + "step": 17578 + }, + { + "epoch": 1.9057892454466607, + "grad_norm": 1.971475601196289, + "learning_rate": 7.321195415639055e-06, + "loss": 0.0305, + "step": 17579 + }, + { + "epoch": 1.9058976582827407, + "grad_norm": 8.49673080444336, + "learning_rate": 7.320470042071668e-06, + "loss": 0.1681, + "step": 17580 + }, + { + "epoch": 1.9060060711188205, + "grad_norm": 2.071378469467163, + "learning_rate": 7.3197446685042805e-06, + "loss": 0.0466, + "step": 17581 + }, + { + "epoch": 1.9061144839549002, + "grad_norm": 2.3880562782287598, + "learning_rate": 7.3190192949368934e-06, + "loss": 0.0275, + "step": 17582 + }, + { + "epoch": 1.90622289679098, + "grad_norm": 2.1534135341644287, + "learning_rate": 7.318293921369506e-06, + "loss": 0.012, + "step": 17583 + }, + { + "epoch": 1.9063313096270598, + "grad_norm": 0.8595160245895386, + "learning_rate": 7.317568547802119e-06, + "loss": 0.0129, + "step": 17584 + }, + { + "epoch": 1.9064397224631398, + "grad_norm": 1.1045082807540894, + "learning_rate": 7.3168431742347316e-06, + "loss": 0.0124, + "step": 17585 + }, + { + "epoch": 1.9065481352992193, + "grad_norm": 4.454914569854736, + "learning_rate": 7.316117800667344e-06, + "loss": 0.0788, + "step": 17586 + }, + { + "epoch": 1.9066565481352993, + "grad_norm": 1.0160235166549683, + "learning_rate": 7.315392427099957e-06, + "loss": 0.0146, + "step": 17587 + }, + { + "epoch": 1.906764960971379, + "grad_norm": 5.352335453033447, + "learning_rate": 7.3146670535325705e-06, + "loss": 0.0375, + "step": 17588 + }, + { + "epoch": 1.9068733738074588, + "grad_norm": 2.173600912094116, + "learning_rate": 7.313941679965183e-06, + "loss": 0.0282, + "step": 17589 + }, + { + "epoch": 1.9069817866435386, + "grad_norm": 0.9725740551948547, + "learning_rate": 7.313216306397796e-06, + "loss": 0.0286, + "step": 17590 + }, + { + "epoch": 1.9070901994796183, + "grad_norm": 3.041795492172241, + "learning_rate": 7.312490932830409e-06, + "loss": 0.0449, + "step": 17591 + }, + { + "epoch": 1.9071986123156983, + "grad_norm": 3.002014398574829, + "learning_rate": 7.311765559263021e-06, + "loss": 0.0385, + "step": 17592 + }, + { + "epoch": 1.9073070251517779, + "grad_norm": 1.913503646850586, + "learning_rate": 7.311040185695634e-06, + "loss": 0.0286, + "step": 17593 + }, + { + "epoch": 1.9074154379878578, + "grad_norm": 2.5078957080841064, + "learning_rate": 7.310314812128246e-06, + "loss": 0.0187, + "step": 17594 + }, + { + "epoch": 1.9075238508239376, + "grad_norm": 1.7437571287155151, + "learning_rate": 7.309589438560859e-06, + "loss": 0.0199, + "step": 17595 + }, + { + "epoch": 1.9076322636600174, + "grad_norm": 3.8677632808685303, + "learning_rate": 7.308864064993473e-06, + "loss": 0.0378, + "step": 17596 + }, + { + "epoch": 1.9077406764960971, + "grad_norm": 4.2900896072387695, + "learning_rate": 7.308138691426086e-06, + "loss": 0.0371, + "step": 17597 + }, + { + "epoch": 1.907849089332177, + "grad_norm": 0.06791095435619354, + "learning_rate": 7.307413317858698e-06, + "loss": 0.0012, + "step": 17598 + }, + { + "epoch": 1.9079575021682569, + "grad_norm": 4.0522871017456055, + "learning_rate": 7.306687944291311e-06, + "loss": 0.0138, + "step": 17599 + }, + { + "epoch": 1.9080659150043364, + "grad_norm": 3.926652431488037, + "learning_rate": 7.305962570723923e-06, + "loss": 0.0071, + "step": 17600 + }, + { + "epoch": 1.9081743278404164, + "grad_norm": 0.13614331185817719, + "learning_rate": 7.305237197156536e-06, + "loss": 0.0014, + "step": 17601 + }, + { + "epoch": 1.908282740676496, + "grad_norm": 6.513924598693848, + "learning_rate": 7.304511823589149e-06, + "loss": 0.0528, + "step": 17602 + }, + { + "epoch": 1.908391153512576, + "grad_norm": 0.8294581174850464, + "learning_rate": 7.303786450021761e-06, + "loss": 0.0115, + "step": 17603 + }, + { + "epoch": 1.9084995663486557, + "grad_norm": 4.252645969390869, + "learning_rate": 7.303061076454374e-06, + "loss": 0.0244, + "step": 17604 + }, + { + "epoch": 1.9086079791847355, + "grad_norm": 0.8266773819923401, + "learning_rate": 7.302335702886988e-06, + "loss": 0.006, + "step": 17605 + }, + { + "epoch": 1.9087163920208152, + "grad_norm": 1.8473769426345825, + "learning_rate": 7.301610329319601e-06, + "loss": 0.0425, + "step": 17606 + }, + { + "epoch": 1.908824804856895, + "grad_norm": 0.20661334693431854, + "learning_rate": 7.300884955752213e-06, + "loss": 0.0023, + "step": 17607 + }, + { + "epoch": 1.908933217692975, + "grad_norm": 1.697445034980774, + "learning_rate": 7.300159582184826e-06, + "loss": 0.0313, + "step": 17608 + }, + { + "epoch": 1.9090416305290545, + "grad_norm": 0.47486841678619385, + "learning_rate": 7.299434208617438e-06, + "loss": 0.0063, + "step": 17609 + }, + { + "epoch": 1.9091500433651345, + "grad_norm": 0.5315905213356018, + "learning_rate": 7.298708835050051e-06, + "loss": 0.0064, + "step": 17610 + }, + { + "epoch": 1.9092584562012143, + "grad_norm": 2.7729074954986572, + "learning_rate": 7.297983461482664e-06, + "loss": 0.0139, + "step": 17611 + }, + { + "epoch": 1.909366869037294, + "grad_norm": 2.7173943519592285, + "learning_rate": 7.297258087915276e-06, + "loss": 0.0432, + "step": 17612 + }, + { + "epoch": 1.9094752818733738, + "grad_norm": 6.313873291015625, + "learning_rate": 7.296532714347889e-06, + "loss": 0.039, + "step": 17613 + }, + { + "epoch": 1.9095836947094535, + "grad_norm": 8.725112915039062, + "learning_rate": 7.295807340780503e-06, + "loss": 0.0664, + "step": 17614 + }, + { + "epoch": 1.9096921075455335, + "grad_norm": 4.320244312286377, + "learning_rate": 7.295081967213115e-06, + "loss": 0.0144, + "step": 17615 + }, + { + "epoch": 1.909800520381613, + "grad_norm": 0.878203809261322, + "learning_rate": 7.294356593645728e-06, + "loss": 0.0163, + "step": 17616 + }, + { + "epoch": 1.909908933217693, + "grad_norm": 0.35818037390708923, + "learning_rate": 7.293631220078341e-06, + "loss": 0.004, + "step": 17617 + }, + { + "epoch": 1.9100173460537726, + "grad_norm": 3.0741684436798096, + "learning_rate": 7.2929058465109535e-06, + "loss": 0.0078, + "step": 17618 + }, + { + "epoch": 1.9101257588898526, + "grad_norm": 1.857864260673523, + "learning_rate": 7.2921804729435665e-06, + "loss": 0.0343, + "step": 17619 + }, + { + "epoch": 1.9102341717259324, + "grad_norm": 2.71174693107605, + "learning_rate": 7.291455099376179e-06, + "loss": 0.0126, + "step": 17620 + }, + { + "epoch": 1.9103425845620121, + "grad_norm": 1.5774987936019897, + "learning_rate": 7.290729725808792e-06, + "loss": 0.0142, + "step": 17621 + }, + { + "epoch": 1.910450997398092, + "grad_norm": 1.0865199565887451, + "learning_rate": 7.2900043522414054e-06, + "loss": 0.0474, + "step": 17622 + }, + { + "epoch": 1.9105594102341716, + "grad_norm": 1.9301451444625854, + "learning_rate": 7.2892789786740184e-06, + "loss": 0.0408, + "step": 17623 + }, + { + "epoch": 1.9106678230702516, + "grad_norm": 0.47360843420028687, + "learning_rate": 7.2885536051066306e-06, + "loss": 0.0044, + "step": 17624 + }, + { + "epoch": 1.9107762359063312, + "grad_norm": 1.4533090591430664, + "learning_rate": 7.2878282315392436e-06, + "loss": 0.0043, + "step": 17625 + }, + { + "epoch": 1.9108846487424112, + "grad_norm": 1.3742783069610596, + "learning_rate": 7.287102857971856e-06, + "loss": 0.0076, + "step": 17626 + }, + { + "epoch": 1.910993061578491, + "grad_norm": 0.042914170771837234, + "learning_rate": 7.286377484404469e-06, + "loss": 0.0005, + "step": 17627 + }, + { + "epoch": 1.9111014744145707, + "grad_norm": 0.23131847381591797, + "learning_rate": 7.285652110837082e-06, + "loss": 0.0027, + "step": 17628 + }, + { + "epoch": 1.9112098872506504, + "grad_norm": 1.849045991897583, + "learning_rate": 7.284926737269694e-06, + "loss": 0.0223, + "step": 17629 + }, + { + "epoch": 1.9113183000867302, + "grad_norm": 13.613546371459961, + "learning_rate": 7.284201363702307e-06, + "loss": 0.0274, + "step": 17630 + }, + { + "epoch": 1.9114267129228102, + "grad_norm": 0.5751510858535767, + "learning_rate": 7.283475990134921e-06, + "loss": 0.001, + "step": 17631 + }, + { + "epoch": 1.9115351257588897, + "grad_norm": 0.8559345602989197, + "learning_rate": 7.282750616567534e-06, + "loss": 0.0063, + "step": 17632 + }, + { + "epoch": 1.9116435385949697, + "grad_norm": 0.7232636213302612, + "learning_rate": 7.282025243000146e-06, + "loss": 0.012, + "step": 17633 + }, + { + "epoch": 1.9117519514310495, + "grad_norm": 3.4318883419036865, + "learning_rate": 7.281299869432759e-06, + "loss": 0.0115, + "step": 17634 + }, + { + "epoch": 1.9118603642671292, + "grad_norm": 2.7101690769195557, + "learning_rate": 7.280574495865371e-06, + "loss": 0.0472, + "step": 17635 + }, + { + "epoch": 1.911968777103209, + "grad_norm": 1.548892855644226, + "learning_rate": 7.279849122297984e-06, + "loss": 0.0149, + "step": 17636 + }, + { + "epoch": 1.9120771899392888, + "grad_norm": 1.3058998584747314, + "learning_rate": 7.279123748730597e-06, + "loss": 0.0177, + "step": 17637 + }, + { + "epoch": 1.9121856027753688, + "grad_norm": 0.8966976404190063, + "learning_rate": 7.278398375163209e-06, + "loss": 0.0097, + "step": 17638 + }, + { + "epoch": 1.9122940156114483, + "grad_norm": 1.723224401473999, + "learning_rate": 7.277673001595822e-06, + "loss": 0.0168, + "step": 17639 + }, + { + "epoch": 1.9124024284475283, + "grad_norm": 3.3808515071868896, + "learning_rate": 7.276947628028436e-06, + "loss": 0.0057, + "step": 17640 + }, + { + "epoch": 1.9125108412836078, + "grad_norm": 3.4720871448516846, + "learning_rate": 7.276222254461048e-06, + "loss": 0.0207, + "step": 17641 + }, + { + "epoch": 1.9126192541196878, + "grad_norm": 0.5091440677642822, + "learning_rate": 7.275496880893661e-06, + "loss": 0.0061, + "step": 17642 + }, + { + "epoch": 1.9127276669557676, + "grad_norm": 1.4925696849822998, + "learning_rate": 7.274771507326274e-06, + "loss": 0.0165, + "step": 17643 + }, + { + "epoch": 1.9128360797918473, + "grad_norm": 0.9750671982765198, + "learning_rate": 7.274046133758886e-06, + "loss": 0.0101, + "step": 17644 + }, + { + "epoch": 1.912944492627927, + "grad_norm": 1.5446659326553345, + "learning_rate": 7.273320760191499e-06, + "loss": 0.019, + "step": 17645 + }, + { + "epoch": 1.9130529054640069, + "grad_norm": 3.1763689517974854, + "learning_rate": 7.272595386624111e-06, + "loss": 0.0531, + "step": 17646 + }, + { + "epoch": 1.9131613183000868, + "grad_norm": 0.4772098660469055, + "learning_rate": 7.271870013056724e-06, + "loss": 0.0048, + "step": 17647 + }, + { + "epoch": 1.9132697311361664, + "grad_norm": 3.266179323196411, + "learning_rate": 7.271144639489338e-06, + "loss": 0.0122, + "step": 17648 + }, + { + "epoch": 1.9133781439722464, + "grad_norm": 1.621878743171692, + "learning_rate": 7.270419265921951e-06, + "loss": 0.0111, + "step": 17649 + }, + { + "epoch": 1.9134865568083261, + "grad_norm": 0.4708080589771271, + "learning_rate": 7.269693892354563e-06, + "loss": 0.0022, + "step": 17650 + }, + { + "epoch": 1.913594969644406, + "grad_norm": 1.1792548894882202, + "learning_rate": 7.268968518787176e-06, + "loss": 0.009, + "step": 17651 + }, + { + "epoch": 1.9137033824804857, + "grad_norm": 1.736434817314148, + "learning_rate": 7.268243145219789e-06, + "loss": 0.0092, + "step": 17652 + }, + { + "epoch": 1.9138117953165654, + "grad_norm": 4.511679649353027, + "learning_rate": 7.267517771652401e-06, + "loss": 0.0481, + "step": 17653 + }, + { + "epoch": 1.9139202081526454, + "grad_norm": 6.434362888336182, + "learning_rate": 7.266792398085014e-06, + "loss": 0.0555, + "step": 17654 + }, + { + "epoch": 1.914028620988725, + "grad_norm": 1.7096104621887207, + "learning_rate": 7.2660670245176265e-06, + "loss": 0.0049, + "step": 17655 + }, + { + "epoch": 1.914137033824805, + "grad_norm": 5.373225212097168, + "learning_rate": 7.2653416509502395e-06, + "loss": 0.018, + "step": 17656 + }, + { + "epoch": 1.9142454466608847, + "grad_norm": 5.569773197174072, + "learning_rate": 7.264616277382853e-06, + "loss": 0.012, + "step": 17657 + }, + { + "epoch": 1.9143538594969645, + "grad_norm": 1.2967921495437622, + "learning_rate": 7.263890903815466e-06, + "loss": 0.017, + "step": 17658 + }, + { + "epoch": 1.9144622723330442, + "grad_norm": 3.1297638416290283, + "learning_rate": 7.2631655302480785e-06, + "loss": 0.0296, + "step": 17659 + }, + { + "epoch": 1.914570685169124, + "grad_norm": 3.382472515106201, + "learning_rate": 7.2624401566806914e-06, + "loss": 0.0224, + "step": 17660 + }, + { + "epoch": 1.914679098005204, + "grad_norm": 0.700944185256958, + "learning_rate": 7.261714783113304e-06, + "loss": 0.0036, + "step": 17661 + }, + { + "epoch": 1.9147875108412835, + "grad_norm": 0.33326709270477295, + "learning_rate": 7.260989409545917e-06, + "loss": 0.0037, + "step": 17662 + }, + { + "epoch": 1.9148959236773635, + "grad_norm": 14.166193008422852, + "learning_rate": 7.2602640359785296e-06, + "loss": 0.0331, + "step": 17663 + }, + { + "epoch": 1.915004336513443, + "grad_norm": 2.0090866088867188, + "learning_rate": 7.259538662411142e-06, + "loss": 0.0093, + "step": 17664 + }, + { + "epoch": 1.915112749349523, + "grad_norm": 3.017490863800049, + "learning_rate": 7.258813288843755e-06, + "loss": 0.0263, + "step": 17665 + }, + { + "epoch": 1.9152211621856028, + "grad_norm": 2.646695852279663, + "learning_rate": 7.2580879152763685e-06, + "loss": 0.0396, + "step": 17666 + }, + { + "epoch": 1.9153295750216826, + "grad_norm": 1.2609057426452637, + "learning_rate": 7.257362541708981e-06, + "loss": 0.0129, + "step": 17667 + }, + { + "epoch": 1.9154379878577623, + "grad_norm": 1.8653489351272583, + "learning_rate": 7.256637168141594e-06, + "loss": 0.0303, + "step": 17668 + }, + { + "epoch": 1.915546400693842, + "grad_norm": 4.814354419708252, + "learning_rate": 7.255911794574207e-06, + "loss": 0.0143, + "step": 17669 + }, + { + "epoch": 1.915654813529922, + "grad_norm": 0.8719190359115601, + "learning_rate": 7.255186421006819e-06, + "loss": 0.0078, + "step": 17670 + }, + { + "epoch": 1.9157632263660016, + "grad_norm": 0.1017179936170578, + "learning_rate": 7.254461047439432e-06, + "loss": 0.001, + "step": 17671 + }, + { + "epoch": 1.9158716392020816, + "grad_norm": 3.8602874279022217, + "learning_rate": 7.253735673872044e-06, + "loss": 0.0249, + "step": 17672 + }, + { + "epoch": 1.9159800520381614, + "grad_norm": 1.3783738613128662, + "learning_rate": 7.253010300304657e-06, + "loss": 0.0177, + "step": 17673 + }, + { + "epoch": 1.9160884648742411, + "grad_norm": 2.804351806640625, + "learning_rate": 7.252284926737271e-06, + "loss": 0.0316, + "step": 17674 + }, + { + "epoch": 1.9161968777103209, + "grad_norm": 0.8306018710136414, + "learning_rate": 7.251559553169884e-06, + "loss": 0.0041, + "step": 17675 + }, + { + "epoch": 1.9163052905464006, + "grad_norm": 0.9342455267906189, + "learning_rate": 7.250834179602496e-06, + "loss": 0.0044, + "step": 17676 + }, + { + "epoch": 1.9164137033824806, + "grad_norm": 3.966465950012207, + "learning_rate": 7.250108806035109e-06, + "loss": 0.0177, + "step": 17677 + }, + { + "epoch": 1.9165221162185602, + "grad_norm": 2.4364752769470215, + "learning_rate": 7.249383432467722e-06, + "loss": 0.0101, + "step": 17678 + }, + { + "epoch": 1.9166305290546402, + "grad_norm": 3.224694013595581, + "learning_rate": 7.248658058900334e-06, + "loss": 0.0593, + "step": 17679 + }, + { + "epoch": 1.91673894189072, + "grad_norm": 2.5665383338928223, + "learning_rate": 7.247932685332947e-06, + "loss": 0.0719, + "step": 17680 + }, + { + "epoch": 1.9168473547267997, + "grad_norm": 1.689003825187683, + "learning_rate": 7.247207311765559e-06, + "loss": 0.0122, + "step": 17681 + }, + { + "epoch": 1.9169557675628794, + "grad_norm": 3.4394001960754395, + "learning_rate": 7.246481938198172e-06, + "loss": 0.0299, + "step": 17682 + }, + { + "epoch": 1.9170641803989592, + "grad_norm": 3.71834397315979, + "learning_rate": 7.245756564630786e-06, + "loss": 0.0089, + "step": 17683 + }, + { + "epoch": 1.9171725932350392, + "grad_norm": 0.9000577926635742, + "learning_rate": 7.245031191063399e-06, + "loss": 0.0059, + "step": 17684 + }, + { + "epoch": 1.9172810060711187, + "grad_norm": 0.9576524496078491, + "learning_rate": 7.244305817496011e-06, + "loss": 0.0068, + "step": 17685 + }, + { + "epoch": 1.9173894189071987, + "grad_norm": 2.8944199085235596, + "learning_rate": 7.243580443928624e-06, + "loss": 0.0477, + "step": 17686 + }, + { + "epoch": 1.9174978317432783, + "grad_norm": 4.589733600616455, + "learning_rate": 7.242855070361236e-06, + "loss": 0.0079, + "step": 17687 + }, + { + "epoch": 1.9176062445793582, + "grad_norm": 0.09971656650304794, + "learning_rate": 7.242129696793849e-06, + "loss": 0.0006, + "step": 17688 + }, + { + "epoch": 1.917714657415438, + "grad_norm": 2.1641969680786133, + "learning_rate": 7.241404323226462e-06, + "loss": 0.0188, + "step": 17689 + }, + { + "epoch": 1.9178230702515178, + "grad_norm": 9.688586235046387, + "learning_rate": 7.240678949659074e-06, + "loss": 0.055, + "step": 17690 + }, + { + "epoch": 1.9179314830875975, + "grad_norm": 8.624701499938965, + "learning_rate": 7.239953576091688e-06, + "loss": 0.0593, + "step": 17691 + }, + { + "epoch": 1.9180398959236773, + "grad_norm": 1.0292290449142456, + "learning_rate": 7.239228202524301e-06, + "loss": 0.0184, + "step": 17692 + }, + { + "epoch": 1.9181483087597573, + "grad_norm": 2.6090221405029297, + "learning_rate": 7.238502828956913e-06, + "loss": 0.014, + "step": 17693 + }, + { + "epoch": 1.9182567215958368, + "grad_norm": 0.7186252474784851, + "learning_rate": 7.237777455389526e-06, + "loss": 0.0073, + "step": 17694 + }, + { + "epoch": 1.9183651344319168, + "grad_norm": 8.430694580078125, + "learning_rate": 7.237052081822139e-06, + "loss": 0.0669, + "step": 17695 + }, + { + "epoch": 1.9184735472679966, + "grad_norm": 2.2882816791534424, + "learning_rate": 7.2363267082547515e-06, + "loss": 0.0203, + "step": 17696 + }, + { + "epoch": 1.9185819601040763, + "grad_norm": 2.2503600120544434, + "learning_rate": 7.2356013346873645e-06, + "loss": 0.0648, + "step": 17697 + }, + { + "epoch": 1.918690372940156, + "grad_norm": 6.145445823669434, + "learning_rate": 7.234875961119977e-06, + "loss": 0.0591, + "step": 17698 + }, + { + "epoch": 1.9187987857762359, + "grad_norm": 4.131433010101318, + "learning_rate": 7.23415058755259e-06, + "loss": 0.0761, + "step": 17699 + }, + { + "epoch": 1.9189071986123158, + "grad_norm": 0.7497051954269409, + "learning_rate": 7.2334252139852034e-06, + "loss": 0.0128, + "step": 17700 + }, + { + "epoch": 1.9190156114483954, + "grad_norm": 4.916934490203857, + "learning_rate": 7.2326998404178164e-06, + "loss": 0.0095, + "step": 17701 + }, + { + "epoch": 1.9191240242844754, + "grad_norm": 0.6993110179901123, + "learning_rate": 7.231974466850429e-06, + "loss": 0.0118, + "step": 17702 + }, + { + "epoch": 1.919232437120555, + "grad_norm": 6.8662800788879395, + "learning_rate": 7.2312490932830416e-06, + "loss": 0.0765, + "step": 17703 + }, + { + "epoch": 1.919340849956635, + "grad_norm": 3.468895435333252, + "learning_rate": 7.2305237197156546e-06, + "loss": 0.0562, + "step": 17704 + }, + { + "epoch": 1.9194492627927147, + "grad_norm": 0.7413585782051086, + "learning_rate": 7.229798346148267e-06, + "loss": 0.0014, + "step": 17705 + }, + { + "epoch": 1.9195576756287944, + "grad_norm": 1.1869516372680664, + "learning_rate": 7.22907297258088e-06, + "loss": 0.0061, + "step": 17706 + }, + { + "epoch": 1.9196660884648742, + "grad_norm": 3.0103201866149902, + "learning_rate": 7.228347599013492e-06, + "loss": 0.0178, + "step": 17707 + }, + { + "epoch": 1.919774501300954, + "grad_norm": 7.101152420043945, + "learning_rate": 7.227622225446105e-06, + "loss": 0.0301, + "step": 17708 + }, + { + "epoch": 1.919882914137034, + "grad_norm": 2.619123935699463, + "learning_rate": 7.226896851878719e-06, + "loss": 0.0729, + "step": 17709 + }, + { + "epoch": 1.9199913269731135, + "grad_norm": 1.9426698684692383, + "learning_rate": 7.226171478311332e-06, + "loss": 0.0024, + "step": 17710 + }, + { + "epoch": 1.9200997398091935, + "grad_norm": 3.644775152206421, + "learning_rate": 7.225446104743944e-06, + "loss": 0.0251, + "step": 17711 + }, + { + "epoch": 1.9202081526452732, + "grad_norm": 2.9878437519073486, + "learning_rate": 7.224720731176557e-06, + "loss": 0.0138, + "step": 17712 + }, + { + "epoch": 1.920316565481353, + "grad_norm": 0.9629692435264587, + "learning_rate": 7.223995357609169e-06, + "loss": 0.0043, + "step": 17713 + }, + { + "epoch": 1.9204249783174328, + "grad_norm": 2.9062936305999756, + "learning_rate": 7.223269984041782e-06, + "loss": 0.0256, + "step": 17714 + }, + { + "epoch": 1.9205333911535125, + "grad_norm": 2.0128960609436035, + "learning_rate": 7.222544610474395e-06, + "loss": 0.0149, + "step": 17715 + }, + { + "epoch": 1.9206418039895925, + "grad_norm": 5.842562675476074, + "learning_rate": 7.221819236907007e-06, + "loss": 0.031, + "step": 17716 + }, + { + "epoch": 1.920750216825672, + "grad_norm": 0.726429283618927, + "learning_rate": 7.221093863339621e-06, + "loss": 0.0071, + "step": 17717 + }, + { + "epoch": 1.920858629661752, + "grad_norm": 0.34484395384788513, + "learning_rate": 7.220368489772234e-06, + "loss": 0.0033, + "step": 17718 + }, + { + "epoch": 1.9209670424978318, + "grad_norm": 5.693498611450195, + "learning_rate": 7.219643116204846e-06, + "loss": 0.0827, + "step": 17719 + }, + { + "epoch": 1.9210754553339116, + "grad_norm": 1.8382140398025513, + "learning_rate": 7.218917742637459e-06, + "loss": 0.0072, + "step": 17720 + }, + { + "epoch": 1.9211838681699913, + "grad_norm": 4.1626763343811035, + "learning_rate": 7.218192369070072e-06, + "loss": 0.0413, + "step": 17721 + }, + { + "epoch": 1.921292281006071, + "grad_norm": 2.4034512042999268, + "learning_rate": 7.217466995502684e-06, + "loss": 0.0262, + "step": 17722 + }, + { + "epoch": 1.921400693842151, + "grad_norm": 0.990497887134552, + "learning_rate": 7.216741621935297e-06, + "loss": 0.0225, + "step": 17723 + }, + { + "epoch": 1.9215091066782306, + "grad_norm": 1.283307671546936, + "learning_rate": 7.216016248367909e-06, + "loss": 0.0239, + "step": 17724 + }, + { + "epoch": 1.9216175195143106, + "grad_norm": 1.0558297634124756, + "learning_rate": 7.215290874800522e-06, + "loss": 0.0117, + "step": 17725 + }, + { + "epoch": 1.9217259323503901, + "grad_norm": 0.07921455800533295, + "learning_rate": 7.214565501233136e-06, + "loss": 0.0011, + "step": 17726 + }, + { + "epoch": 1.9218343451864701, + "grad_norm": 3.1200525760650635, + "learning_rate": 7.213840127665749e-06, + "loss": 0.0293, + "step": 17727 + }, + { + "epoch": 1.9219427580225499, + "grad_norm": 0.5389177203178406, + "learning_rate": 7.213114754098361e-06, + "loss": 0.0063, + "step": 17728 + }, + { + "epoch": 1.9220511708586296, + "grad_norm": 0.6616811752319336, + "learning_rate": 7.212389380530974e-06, + "loss": 0.0056, + "step": 17729 + }, + { + "epoch": 1.9221595836947094, + "grad_norm": 0.5366150736808777, + "learning_rate": 7.211664006963587e-06, + "loss": 0.0072, + "step": 17730 + }, + { + "epoch": 1.9222679965307892, + "grad_norm": 4.420082092285156, + "learning_rate": 7.210938633396199e-06, + "loss": 0.0226, + "step": 17731 + }, + { + "epoch": 1.9223764093668692, + "grad_norm": 2.0497138500213623, + "learning_rate": 7.210213259828812e-06, + "loss": 0.0237, + "step": 17732 + }, + { + "epoch": 1.9224848222029487, + "grad_norm": 3.999870538711548, + "learning_rate": 7.2094878862614245e-06, + "loss": 0.0287, + "step": 17733 + }, + { + "epoch": 1.9225932350390287, + "grad_norm": 0.1961522102355957, + "learning_rate": 7.2087625126940375e-06, + "loss": 0.0025, + "step": 17734 + }, + { + "epoch": 1.9227016478751084, + "grad_norm": 0.5060821175575256, + "learning_rate": 7.208037139126651e-06, + "loss": 0.0048, + "step": 17735 + }, + { + "epoch": 1.9228100607111882, + "grad_norm": 8.409379959106445, + "learning_rate": 7.207311765559264e-06, + "loss": 0.0532, + "step": 17736 + }, + { + "epoch": 1.922918473547268, + "grad_norm": 3.1012792587280273, + "learning_rate": 7.2065863919918765e-06, + "loss": 0.0263, + "step": 17737 + }, + { + "epoch": 1.9230268863833477, + "grad_norm": 1.361943006515503, + "learning_rate": 7.2058610184244895e-06, + "loss": 0.0107, + "step": 17738 + }, + { + "epoch": 1.9231352992194277, + "grad_norm": 0.7966051697731018, + "learning_rate": 7.205135644857102e-06, + "loss": 0.0056, + "step": 17739 + }, + { + "epoch": 1.9232437120555073, + "grad_norm": 3.1304142475128174, + "learning_rate": 7.204410271289715e-06, + "loss": 0.0114, + "step": 17740 + }, + { + "epoch": 1.9233521248915872, + "grad_norm": 0.837926983833313, + "learning_rate": 7.203684897722328e-06, + "loss": 0.0125, + "step": 17741 + }, + { + "epoch": 1.923460537727667, + "grad_norm": 4.4936394691467285, + "learning_rate": 7.20295952415494e-06, + "loss": 0.0179, + "step": 17742 + }, + { + "epoch": 1.9235689505637468, + "grad_norm": 0.5207909941673279, + "learning_rate": 7.2022341505875536e-06, + "loss": 0.008, + "step": 17743 + }, + { + "epoch": 1.9236773633998265, + "grad_norm": 4.113725662231445, + "learning_rate": 7.2015087770201666e-06, + "loss": 0.0157, + "step": 17744 + }, + { + "epoch": 1.9237857762359063, + "grad_norm": 1.0272070169448853, + "learning_rate": 7.200783403452779e-06, + "loss": 0.0116, + "step": 17745 + }, + { + "epoch": 1.9238941890719863, + "grad_norm": 1.6008085012435913, + "learning_rate": 7.200058029885392e-06, + "loss": 0.0061, + "step": 17746 + }, + { + "epoch": 1.9240026019080658, + "grad_norm": 2.118412971496582, + "learning_rate": 7.199332656318005e-06, + "loss": 0.0078, + "step": 17747 + }, + { + "epoch": 1.9241110147441458, + "grad_norm": 4.464573860168457, + "learning_rate": 7.198607282750617e-06, + "loss": 0.0291, + "step": 17748 + }, + { + "epoch": 1.9242194275802254, + "grad_norm": 0.03689097985625267, + "learning_rate": 7.19788190918323e-06, + "loss": 0.0004, + "step": 17749 + }, + { + "epoch": 1.9243278404163053, + "grad_norm": 1.4100878238677979, + "learning_rate": 7.197156535615843e-06, + "loss": 0.0432, + "step": 17750 + }, + { + "epoch": 1.924436253252385, + "grad_norm": 4.416857719421387, + "learning_rate": 7.196431162048455e-06, + "loss": 0.0065, + "step": 17751 + }, + { + "epoch": 1.9245446660884649, + "grad_norm": 3.760972738265991, + "learning_rate": 7.195705788481069e-06, + "loss": 0.0566, + "step": 17752 + }, + { + "epoch": 1.9246530789245446, + "grad_norm": 1.748471975326538, + "learning_rate": 7.194980414913682e-06, + "loss": 0.0095, + "step": 17753 + }, + { + "epoch": 1.9247614917606244, + "grad_norm": 1.1900475025177002, + "learning_rate": 7.194255041346294e-06, + "loss": 0.012, + "step": 17754 + }, + { + "epoch": 1.9248699045967044, + "grad_norm": 0.02980171889066696, + "learning_rate": 7.193529667778907e-06, + "loss": 0.0004, + "step": 17755 + }, + { + "epoch": 1.924978317432784, + "grad_norm": 0.12036708742380142, + "learning_rate": 7.19280429421152e-06, + "loss": 0.0006, + "step": 17756 + }, + { + "epoch": 1.925086730268864, + "grad_norm": 2.227057695388794, + "learning_rate": 7.192078920644132e-06, + "loss": 0.0161, + "step": 17757 + }, + { + "epoch": 1.9251951431049437, + "grad_norm": 2.9454379081726074, + "learning_rate": 7.191353547076745e-06, + "loss": 0.0095, + "step": 17758 + }, + { + "epoch": 1.9253035559410234, + "grad_norm": 3.0305681228637695, + "learning_rate": 7.190628173509357e-06, + "loss": 0.0548, + "step": 17759 + }, + { + "epoch": 1.9254119687771032, + "grad_norm": 3.9469170570373535, + "learning_rate": 7.18990279994197e-06, + "loss": 0.0148, + "step": 17760 + }, + { + "epoch": 1.925520381613183, + "grad_norm": 7.56870174407959, + "learning_rate": 7.189177426374584e-06, + "loss": 0.0124, + "step": 17761 + }, + { + "epoch": 1.925628794449263, + "grad_norm": 3.0699374675750732, + "learning_rate": 7.188452052807197e-06, + "loss": 0.0347, + "step": 17762 + }, + { + "epoch": 1.9257372072853425, + "grad_norm": 3.3416593074798584, + "learning_rate": 7.187726679239809e-06, + "loss": 0.0686, + "step": 17763 + }, + { + "epoch": 1.9258456201214225, + "grad_norm": 1.5406593084335327, + "learning_rate": 7.187001305672422e-06, + "loss": 0.0255, + "step": 17764 + }, + { + "epoch": 1.925954032957502, + "grad_norm": 0.13578832149505615, + "learning_rate": 7.186275932105034e-06, + "loss": 0.0007, + "step": 17765 + }, + { + "epoch": 1.926062445793582, + "grad_norm": 0.5233749747276306, + "learning_rate": 7.185550558537647e-06, + "loss": 0.0034, + "step": 17766 + }, + { + "epoch": 1.9261708586296618, + "grad_norm": 1.433671474456787, + "learning_rate": 7.18482518497026e-06, + "loss": 0.0101, + "step": 17767 + }, + { + "epoch": 1.9262792714657415, + "grad_norm": 1.816810965538025, + "learning_rate": 7.184099811402872e-06, + "loss": 0.0278, + "step": 17768 + }, + { + "epoch": 1.9263876843018215, + "grad_norm": 6.480623722076416, + "learning_rate": 7.183374437835486e-06, + "loss": 0.0094, + "step": 17769 + }, + { + "epoch": 1.926496097137901, + "grad_norm": 3.5737054347991943, + "learning_rate": 7.182649064268099e-06, + "loss": 0.0151, + "step": 17770 + }, + { + "epoch": 1.926604509973981, + "grad_norm": 0.47590672969818115, + "learning_rate": 7.181923690700711e-06, + "loss": 0.003, + "step": 17771 + }, + { + "epoch": 1.9267129228100606, + "grad_norm": 3.8738951683044434, + "learning_rate": 7.181198317133324e-06, + "loss": 0.0427, + "step": 17772 + }, + { + "epoch": 1.9268213356461406, + "grad_norm": 1.5010297298431396, + "learning_rate": 7.180472943565937e-06, + "loss": 0.0182, + "step": 17773 + }, + { + "epoch": 1.9269297484822203, + "grad_norm": 2.2758264541625977, + "learning_rate": 7.1797475699985495e-06, + "loss": 0.0651, + "step": 17774 + }, + { + "epoch": 1.9270381613183, + "grad_norm": 0.27565860748291016, + "learning_rate": 7.1790221964311625e-06, + "loss": 0.0024, + "step": 17775 + }, + { + "epoch": 1.9271465741543798, + "grad_norm": 1.5785781145095825, + "learning_rate": 7.1782968228637755e-06, + "loss": 0.0244, + "step": 17776 + }, + { + "epoch": 1.9272549869904596, + "grad_norm": 0.23529307544231415, + "learning_rate": 7.177571449296388e-06, + "loss": 0.004, + "step": 17777 + }, + { + "epoch": 1.9273633998265396, + "grad_norm": 0.11435429751873016, + "learning_rate": 7.1768460757290014e-06, + "loss": 0.0006, + "step": 17778 + }, + { + "epoch": 1.9274718126626191, + "grad_norm": 1.1005375385284424, + "learning_rate": 7.1761207021616144e-06, + "loss": 0.0382, + "step": 17779 + }, + { + "epoch": 1.9275802254986991, + "grad_norm": 1.6209169626235962, + "learning_rate": 7.175395328594227e-06, + "loss": 0.0123, + "step": 17780 + }, + { + "epoch": 1.9276886383347789, + "grad_norm": 5.099432945251465, + "learning_rate": 7.1746699550268396e-06, + "loss": 0.0283, + "step": 17781 + }, + { + "epoch": 1.9277970511708586, + "grad_norm": 5.322148323059082, + "learning_rate": 7.1739445814594526e-06, + "loss": 0.0308, + "step": 17782 + }, + { + "epoch": 1.9279054640069384, + "grad_norm": 3.0546953678131104, + "learning_rate": 7.173219207892065e-06, + "loss": 0.0347, + "step": 17783 + }, + { + "epoch": 1.9280138768430182, + "grad_norm": 8.661346435546875, + "learning_rate": 7.172493834324678e-06, + "loss": 0.0522, + "step": 17784 + }, + { + "epoch": 1.9281222896790982, + "grad_norm": 1.3162715435028076, + "learning_rate": 7.17176846075729e-06, + "loss": 0.0114, + "step": 17785 + }, + { + "epoch": 1.9282307025151777, + "grad_norm": 1.9759479761123657, + "learning_rate": 7.171043087189904e-06, + "loss": 0.0175, + "step": 17786 + }, + { + "epoch": 1.9283391153512577, + "grad_norm": 0.4362335205078125, + "learning_rate": 7.170317713622517e-06, + "loss": 0.0046, + "step": 17787 + }, + { + "epoch": 1.9284475281873372, + "grad_norm": 1.9399478435516357, + "learning_rate": 7.16959234005513e-06, + "loss": 0.021, + "step": 17788 + }, + { + "epoch": 1.9285559410234172, + "grad_norm": 0.8141745328903198, + "learning_rate": 7.168866966487742e-06, + "loss": 0.0036, + "step": 17789 + }, + { + "epoch": 1.928664353859497, + "grad_norm": 3.293255567550659, + "learning_rate": 7.168141592920355e-06, + "loss": 0.0337, + "step": 17790 + }, + { + "epoch": 1.9287727666955767, + "grad_norm": 0.9396902322769165, + "learning_rate": 7.167416219352967e-06, + "loss": 0.0105, + "step": 17791 + }, + { + "epoch": 1.9288811795316565, + "grad_norm": 0.3970957100391388, + "learning_rate": 7.16669084578558e-06, + "loss": 0.0047, + "step": 17792 + }, + { + "epoch": 1.9289895923677363, + "grad_norm": 3.4789021015167236, + "learning_rate": 7.165965472218193e-06, + "loss": 0.0565, + "step": 17793 + }, + { + "epoch": 1.9290980052038162, + "grad_norm": 0.556842565536499, + "learning_rate": 7.165240098650805e-06, + "loss": 0.0047, + "step": 17794 + }, + { + "epoch": 1.9292064180398958, + "grad_norm": 4.844944953918457, + "learning_rate": 7.164514725083419e-06, + "loss": 0.0079, + "step": 17795 + }, + { + "epoch": 1.9293148308759758, + "grad_norm": 5.407127380371094, + "learning_rate": 7.163789351516032e-06, + "loss": 0.057, + "step": 17796 + }, + { + "epoch": 1.9294232437120555, + "grad_norm": 1.3739293813705444, + "learning_rate": 7.163063977948644e-06, + "loss": 0.0068, + "step": 17797 + }, + { + "epoch": 1.9295316565481353, + "grad_norm": 2.1363723278045654, + "learning_rate": 7.162338604381257e-06, + "loss": 0.0388, + "step": 17798 + }, + { + "epoch": 1.929640069384215, + "grad_norm": 2.9807982444763184, + "learning_rate": 7.16161323081387e-06, + "loss": 0.0359, + "step": 17799 + }, + { + "epoch": 1.9297484822202948, + "grad_norm": 2.2795276641845703, + "learning_rate": 7.160887857246482e-06, + "loss": 0.0509, + "step": 17800 + }, + { + "epoch": 1.9298568950563748, + "grad_norm": 0.041844069957733154, + "learning_rate": 7.160162483679095e-06, + "loss": 0.0005, + "step": 17801 + }, + { + "epoch": 1.9299653078924544, + "grad_norm": 1.6718147993087769, + "learning_rate": 7.159437110111708e-06, + "loss": 0.0259, + "step": 17802 + }, + { + "epoch": 1.9300737207285343, + "grad_norm": 0.4861878454685211, + "learning_rate": 7.15871173654432e-06, + "loss": 0.0044, + "step": 17803 + }, + { + "epoch": 1.930182133564614, + "grad_norm": 2.7899558544158936, + "learning_rate": 7.157986362976934e-06, + "loss": 0.0146, + "step": 17804 + }, + { + "epoch": 1.9302905464006939, + "grad_norm": 5.844955921173096, + "learning_rate": 7.157260989409547e-06, + "loss": 0.042, + "step": 17805 + }, + { + "epoch": 1.9303989592367736, + "grad_norm": 2.111783266067505, + "learning_rate": 7.156535615842159e-06, + "loss": 0.0477, + "step": 17806 + }, + { + "epoch": 1.9305073720728534, + "grad_norm": 0.43776530027389526, + "learning_rate": 7.155810242274772e-06, + "loss": 0.0038, + "step": 17807 + }, + { + "epoch": 1.9306157849089334, + "grad_norm": 5.280577182769775, + "learning_rate": 7.155084868707385e-06, + "loss": 0.0301, + "step": 17808 + }, + { + "epoch": 1.930724197745013, + "grad_norm": 1.920559048652649, + "learning_rate": 7.154359495139997e-06, + "loss": 0.0067, + "step": 17809 + }, + { + "epoch": 1.930832610581093, + "grad_norm": 1.4987348318099976, + "learning_rate": 7.15363412157261e-06, + "loss": 0.0146, + "step": 17810 + }, + { + "epoch": 1.9309410234171724, + "grad_norm": 3.2978084087371826, + "learning_rate": 7.1529087480052225e-06, + "loss": 0.0105, + "step": 17811 + }, + { + "epoch": 1.9310494362532524, + "grad_norm": 0.39696601033210754, + "learning_rate": 7.152183374437836e-06, + "loss": 0.0018, + "step": 17812 + }, + { + "epoch": 1.9311578490893322, + "grad_norm": 3.5348384380340576, + "learning_rate": 7.151458000870449e-06, + "loss": 0.0561, + "step": 17813 + }, + { + "epoch": 1.931266261925412, + "grad_norm": 4.501952171325684, + "learning_rate": 7.150732627303062e-06, + "loss": 0.0436, + "step": 17814 + }, + { + "epoch": 1.9313746747614917, + "grad_norm": 0.21443969011306763, + "learning_rate": 7.1500072537356745e-06, + "loss": 0.0026, + "step": 17815 + }, + { + "epoch": 1.9314830875975715, + "grad_norm": 0.5208040475845337, + "learning_rate": 7.1492818801682875e-06, + "loss": 0.0033, + "step": 17816 + }, + { + "epoch": 1.9315915004336515, + "grad_norm": 0.06730398535728455, + "learning_rate": 7.1485565066009e-06, + "loss": 0.0012, + "step": 17817 + }, + { + "epoch": 1.931699913269731, + "grad_norm": 3.1661806106567383, + "learning_rate": 7.147831133033513e-06, + "loss": 0.0428, + "step": 17818 + }, + { + "epoch": 1.931808326105811, + "grad_norm": 1.5808079242706299, + "learning_rate": 7.147105759466126e-06, + "loss": 0.0056, + "step": 17819 + }, + { + "epoch": 1.9319167389418908, + "grad_norm": 0.4011942446231842, + "learning_rate": 7.146380385898738e-06, + "loss": 0.0057, + "step": 17820 + }, + { + "epoch": 1.9320251517779705, + "grad_norm": 2.1103110313415527, + "learning_rate": 7.1456550123313516e-06, + "loss": 0.0065, + "step": 17821 + }, + { + "epoch": 1.9321335646140503, + "grad_norm": 5.153995037078857, + "learning_rate": 7.1449296387639646e-06, + "loss": 0.0528, + "step": 17822 + }, + { + "epoch": 1.93224197745013, + "grad_norm": 1.1995837688446045, + "learning_rate": 7.144204265196577e-06, + "loss": 0.0229, + "step": 17823 + }, + { + "epoch": 1.93235039028621, + "grad_norm": 0.6656681299209595, + "learning_rate": 7.14347889162919e-06, + "loss": 0.0038, + "step": 17824 + }, + { + "epoch": 1.9324588031222896, + "grad_norm": 6.165970802307129, + "learning_rate": 7.142753518061803e-06, + "loss": 0.0586, + "step": 17825 + }, + { + "epoch": 1.9325672159583696, + "grad_norm": 1.6878609657287598, + "learning_rate": 7.142028144494415e-06, + "loss": 0.0195, + "step": 17826 + }, + { + "epoch": 1.932675628794449, + "grad_norm": 2.1828951835632324, + "learning_rate": 7.141302770927028e-06, + "loss": 0.0376, + "step": 17827 + }, + { + "epoch": 1.932784041630529, + "grad_norm": 5.456162452697754, + "learning_rate": 7.140577397359641e-06, + "loss": 0.0755, + "step": 17828 + }, + { + "epoch": 1.9328924544666088, + "grad_norm": 2.1699752807617188, + "learning_rate": 7.139852023792253e-06, + "loss": 0.0117, + "step": 17829 + }, + { + "epoch": 1.9330008673026886, + "grad_norm": 1.9259119033813477, + "learning_rate": 7.139126650224867e-06, + "loss": 0.0042, + "step": 17830 + }, + { + "epoch": 1.9331092801387686, + "grad_norm": 7.360161781311035, + "learning_rate": 7.13840127665748e-06, + "loss": 0.0221, + "step": 17831 + }, + { + "epoch": 1.9332176929748481, + "grad_norm": 1.8502365350723267, + "learning_rate": 7.137675903090092e-06, + "loss": 0.0182, + "step": 17832 + }, + { + "epoch": 1.9333261058109281, + "grad_norm": 0.292630672454834, + "learning_rate": 7.136950529522705e-06, + "loss": 0.0034, + "step": 17833 + }, + { + "epoch": 1.9334345186470077, + "grad_norm": 1.4310081005096436, + "learning_rate": 7.136225155955318e-06, + "loss": 0.0239, + "step": 17834 + }, + { + "epoch": 1.9335429314830876, + "grad_norm": 1.9456027746200562, + "learning_rate": 7.13549978238793e-06, + "loss": 0.035, + "step": 17835 + }, + { + "epoch": 1.9336513443191674, + "grad_norm": 3.710242986679077, + "learning_rate": 7.134774408820543e-06, + "loss": 0.0459, + "step": 17836 + }, + { + "epoch": 1.9337597571552472, + "grad_norm": 5.640117168426514, + "learning_rate": 7.134049035253155e-06, + "loss": 0.0459, + "step": 17837 + }, + { + "epoch": 1.933868169991327, + "grad_norm": 0.5473605990409851, + "learning_rate": 7.133323661685769e-06, + "loss": 0.0033, + "step": 17838 + }, + { + "epoch": 1.9339765828274067, + "grad_norm": 2.716561794281006, + "learning_rate": 7.132598288118382e-06, + "loss": 0.0232, + "step": 17839 + }, + { + "epoch": 1.9340849956634867, + "grad_norm": 3.10180401802063, + "learning_rate": 7.131872914550995e-06, + "loss": 0.0473, + "step": 17840 + }, + { + "epoch": 1.9341934084995662, + "grad_norm": 2.103353500366211, + "learning_rate": 7.131147540983607e-06, + "loss": 0.0093, + "step": 17841 + }, + { + "epoch": 1.9343018213356462, + "grad_norm": 4.281601905822754, + "learning_rate": 7.13042216741622e-06, + "loss": 0.0665, + "step": 17842 + }, + { + "epoch": 1.934410234171726, + "grad_norm": 1.5817317962646484, + "learning_rate": 7.129696793848832e-06, + "loss": 0.0275, + "step": 17843 + }, + { + "epoch": 1.9345186470078057, + "grad_norm": 14.171974182128906, + "learning_rate": 7.128971420281445e-06, + "loss": 0.0323, + "step": 17844 + }, + { + "epoch": 1.9346270598438855, + "grad_norm": 0.3173104226589203, + "learning_rate": 7.128246046714058e-06, + "loss": 0.0021, + "step": 17845 + }, + { + "epoch": 1.9347354726799653, + "grad_norm": 3.4060096740722656, + "learning_rate": 7.12752067314667e-06, + "loss": 0.0138, + "step": 17846 + }, + { + "epoch": 1.9348438855160452, + "grad_norm": 0.9548139572143555, + "learning_rate": 7.126795299579284e-06, + "loss": 0.0029, + "step": 17847 + }, + { + "epoch": 1.9349522983521248, + "grad_norm": 0.9036409854888916, + "learning_rate": 7.126069926011897e-06, + "loss": 0.0052, + "step": 17848 + }, + { + "epoch": 1.9350607111882048, + "grad_norm": 2.116328477859497, + "learning_rate": 7.125344552444509e-06, + "loss": 0.0181, + "step": 17849 + }, + { + "epoch": 1.9351691240242843, + "grad_norm": 2.4912445545196533, + "learning_rate": 7.124619178877122e-06, + "loss": 0.0579, + "step": 17850 + }, + { + "epoch": 1.9352775368603643, + "grad_norm": 1.8364886045455933, + "learning_rate": 7.123893805309735e-06, + "loss": 0.0401, + "step": 17851 + }, + { + "epoch": 1.935385949696444, + "grad_norm": 2.3338189125061035, + "learning_rate": 7.1231684317423475e-06, + "loss": 0.0541, + "step": 17852 + }, + { + "epoch": 1.9354943625325238, + "grad_norm": 2.9464592933654785, + "learning_rate": 7.1224430581749605e-06, + "loss": 0.0335, + "step": 17853 + }, + { + "epoch": 1.9356027753686036, + "grad_norm": 1.8363378047943115, + "learning_rate": 7.1217176846075735e-06, + "loss": 0.0693, + "step": 17854 + }, + { + "epoch": 1.9357111882046834, + "grad_norm": 1.6077451705932617, + "learning_rate": 7.120992311040186e-06, + "loss": 0.0051, + "step": 17855 + }, + { + "epoch": 1.9358196010407633, + "grad_norm": 3.8721954822540283, + "learning_rate": 7.1202669374727995e-06, + "loss": 0.0154, + "step": 17856 + }, + { + "epoch": 1.9359280138768429, + "grad_norm": 2.1329421997070312, + "learning_rate": 7.1195415639054124e-06, + "loss": 0.0249, + "step": 17857 + }, + { + "epoch": 1.9360364267129229, + "grad_norm": 2.736107587814331, + "learning_rate": 7.118816190338025e-06, + "loss": 0.0604, + "step": 17858 + }, + { + "epoch": 1.9361448395490026, + "grad_norm": 4.525880813598633, + "learning_rate": 7.118090816770638e-06, + "loss": 0.0285, + "step": 17859 + }, + { + "epoch": 1.9362532523850824, + "grad_norm": 2.56652569770813, + "learning_rate": 7.1173654432032506e-06, + "loss": 0.0153, + "step": 17860 + }, + { + "epoch": 1.9363616652211622, + "grad_norm": 4.184438705444336, + "learning_rate": 7.116640069635863e-06, + "loss": 0.0256, + "step": 17861 + }, + { + "epoch": 1.936470078057242, + "grad_norm": 0.6008393168449402, + "learning_rate": 7.115914696068476e-06, + "loss": 0.0049, + "step": 17862 + }, + { + "epoch": 1.936578490893322, + "grad_norm": 3.164583683013916, + "learning_rate": 7.115189322501088e-06, + "loss": 0.0209, + "step": 17863 + }, + { + "epoch": 1.9366869037294014, + "grad_norm": 1.6409564018249512, + "learning_rate": 7.114463948933702e-06, + "loss": 0.0115, + "step": 17864 + }, + { + "epoch": 1.9367953165654814, + "grad_norm": 1.839571237564087, + "learning_rate": 7.113738575366315e-06, + "loss": 0.0207, + "step": 17865 + }, + { + "epoch": 1.9369037294015612, + "grad_norm": 0.8897638916969299, + "learning_rate": 7.113013201798928e-06, + "loss": 0.0099, + "step": 17866 + }, + { + "epoch": 1.937012142237641, + "grad_norm": 1.0703274011611938, + "learning_rate": 7.11228782823154e-06, + "loss": 0.0083, + "step": 17867 + }, + { + "epoch": 1.9371205550737207, + "grad_norm": 3.4932258129119873, + "learning_rate": 7.111562454664153e-06, + "loss": 0.0141, + "step": 17868 + }, + { + "epoch": 1.9372289679098005, + "grad_norm": 1.598246455192566, + "learning_rate": 7.110837081096765e-06, + "loss": 0.029, + "step": 17869 + }, + { + "epoch": 1.9373373807458805, + "grad_norm": 2.867039918899536, + "learning_rate": 7.110111707529378e-06, + "loss": 0.023, + "step": 17870 + }, + { + "epoch": 1.93744579358196, + "grad_norm": 0.45683205127716064, + "learning_rate": 7.109386333961991e-06, + "loss": 0.0036, + "step": 17871 + }, + { + "epoch": 1.93755420641804, + "grad_norm": 2.7852091789245605, + "learning_rate": 7.108660960394603e-06, + "loss": 0.0073, + "step": 17872 + }, + { + "epoch": 1.9376626192541195, + "grad_norm": 2.2845170497894287, + "learning_rate": 7.107935586827217e-06, + "loss": 0.0196, + "step": 17873 + }, + { + "epoch": 1.9377710320901995, + "grad_norm": 1.7853636741638184, + "learning_rate": 7.10721021325983e-06, + "loss": 0.0352, + "step": 17874 + }, + { + "epoch": 1.9378794449262793, + "grad_norm": 4.556045055389404, + "learning_rate": 7.106484839692442e-06, + "loss": 0.015, + "step": 17875 + }, + { + "epoch": 1.937987857762359, + "grad_norm": 0.451761931180954, + "learning_rate": 7.105759466125055e-06, + "loss": 0.003, + "step": 17876 + }, + { + "epoch": 1.9380962705984388, + "grad_norm": 0.4672797620296478, + "learning_rate": 7.105034092557668e-06, + "loss": 0.0036, + "step": 17877 + }, + { + "epoch": 1.9382046834345186, + "grad_norm": 1.6008493900299072, + "learning_rate": 7.10430871899028e-06, + "loss": 0.0323, + "step": 17878 + }, + { + "epoch": 1.9383130962705986, + "grad_norm": 1.655887484550476, + "learning_rate": 7.103583345422893e-06, + "loss": 0.0194, + "step": 17879 + }, + { + "epoch": 1.938421509106678, + "grad_norm": 3.3316946029663086, + "learning_rate": 7.102857971855506e-06, + "loss": 0.0735, + "step": 17880 + }, + { + "epoch": 1.938529921942758, + "grad_norm": 1.032768726348877, + "learning_rate": 7.102132598288118e-06, + "loss": 0.0073, + "step": 17881 + }, + { + "epoch": 1.9386383347788378, + "grad_norm": 9.592865943908691, + "learning_rate": 7.101407224720732e-06, + "loss": 0.0233, + "step": 17882 + }, + { + "epoch": 1.9387467476149176, + "grad_norm": 0.8976743221282959, + "learning_rate": 7.100681851153345e-06, + "loss": 0.0191, + "step": 17883 + }, + { + "epoch": 1.9388551604509974, + "grad_norm": 0.5942256450653076, + "learning_rate": 7.099956477585957e-06, + "loss": 0.0041, + "step": 17884 + }, + { + "epoch": 1.9389635732870771, + "grad_norm": 5.133370876312256, + "learning_rate": 7.09923110401857e-06, + "loss": 0.0308, + "step": 17885 + }, + { + "epoch": 1.9390719861231571, + "grad_norm": 0.27432140707969666, + "learning_rate": 7.098505730451183e-06, + "loss": 0.0028, + "step": 17886 + }, + { + "epoch": 1.9391803989592367, + "grad_norm": 0.2164924293756485, + "learning_rate": 7.097780356883795e-06, + "loss": 0.002, + "step": 17887 + }, + { + "epoch": 1.9392888117953166, + "grad_norm": 1.365417242050171, + "learning_rate": 7.097054983316408e-06, + "loss": 0.0032, + "step": 17888 + }, + { + "epoch": 1.9393972246313964, + "grad_norm": 3.547004222869873, + "learning_rate": 7.0963296097490205e-06, + "loss": 0.0447, + "step": 17889 + }, + { + "epoch": 1.9395056374674762, + "grad_norm": 3.278665065765381, + "learning_rate": 7.095604236181634e-06, + "loss": 0.0699, + "step": 17890 + }, + { + "epoch": 1.939614050303556, + "grad_norm": 0.2464112937450409, + "learning_rate": 7.094878862614247e-06, + "loss": 0.0027, + "step": 17891 + }, + { + "epoch": 1.9397224631396357, + "grad_norm": 3.0806498527526855, + "learning_rate": 7.09415348904686e-06, + "loss": 0.0554, + "step": 17892 + }, + { + "epoch": 1.9398308759757157, + "grad_norm": 5.997413158416748, + "learning_rate": 7.0934281154794725e-06, + "loss": 0.0361, + "step": 17893 + }, + { + "epoch": 1.9399392888117952, + "grad_norm": 0.61603182554245, + "learning_rate": 7.0927027419120855e-06, + "loss": 0.0135, + "step": 17894 + }, + { + "epoch": 1.9400477016478752, + "grad_norm": 2.2982587814331055, + "learning_rate": 7.091977368344698e-06, + "loss": 0.0192, + "step": 17895 + }, + { + "epoch": 1.9401561144839548, + "grad_norm": 3.7241458892822266, + "learning_rate": 7.091251994777311e-06, + "loss": 0.0464, + "step": 17896 + }, + { + "epoch": 1.9402645273200347, + "grad_norm": 2.3412530422210693, + "learning_rate": 7.090526621209924e-06, + "loss": 0.0111, + "step": 17897 + }, + { + "epoch": 1.9403729401561145, + "grad_norm": 2.2214467525482178, + "learning_rate": 7.089801247642536e-06, + "loss": 0.0177, + "step": 17898 + }, + { + "epoch": 1.9404813529921943, + "grad_norm": 2.522385358810425, + "learning_rate": 7.0890758740751496e-06, + "loss": 0.0128, + "step": 17899 + }, + { + "epoch": 1.940589765828274, + "grad_norm": 0.7788289189338684, + "learning_rate": 7.0883505005077626e-06, + "loss": 0.0109, + "step": 17900 + }, + { + "epoch": 1.9406981786643538, + "grad_norm": 2.2245795726776123, + "learning_rate": 7.087625126940375e-06, + "loss": 0.0218, + "step": 17901 + }, + { + "epoch": 1.9408065915004338, + "grad_norm": 2.6522467136383057, + "learning_rate": 7.086899753372988e-06, + "loss": 0.0349, + "step": 17902 + }, + { + "epoch": 1.9409150043365133, + "grad_norm": 3.272397518157959, + "learning_rate": 7.086174379805601e-06, + "loss": 0.0295, + "step": 17903 + }, + { + "epoch": 1.9410234171725933, + "grad_norm": 2.8810815811157227, + "learning_rate": 7.085449006238213e-06, + "loss": 0.0331, + "step": 17904 + }, + { + "epoch": 1.941131830008673, + "grad_norm": 3.4392759799957275, + "learning_rate": 7.084723632670826e-06, + "loss": 0.0765, + "step": 17905 + }, + { + "epoch": 1.9412402428447528, + "grad_norm": 4.151577949523926, + "learning_rate": 7.083998259103439e-06, + "loss": 0.0231, + "step": 17906 + }, + { + "epoch": 1.9413486556808326, + "grad_norm": 0.7085950374603271, + "learning_rate": 7.083272885536052e-06, + "loss": 0.0099, + "step": 17907 + }, + { + "epoch": 1.9414570685169124, + "grad_norm": 3.852869749069214, + "learning_rate": 7.082547511968665e-06, + "loss": 0.0459, + "step": 17908 + }, + { + "epoch": 1.9415654813529923, + "grad_norm": 2.309512138366699, + "learning_rate": 7.081822138401278e-06, + "loss": 0.015, + "step": 17909 + }, + { + "epoch": 1.9416738941890719, + "grad_norm": 2.9206900596618652, + "learning_rate": 7.08109676483389e-06, + "loss": 0.0085, + "step": 17910 + }, + { + "epoch": 1.9417823070251519, + "grad_norm": 0.40998655557632446, + "learning_rate": 7.080371391266503e-06, + "loss": 0.0025, + "step": 17911 + }, + { + "epoch": 1.9418907198612314, + "grad_norm": 4.610427379608154, + "learning_rate": 7.079646017699116e-06, + "loss": 0.049, + "step": 17912 + }, + { + "epoch": 1.9419991326973114, + "grad_norm": 4.091304302215576, + "learning_rate": 7.078920644131728e-06, + "loss": 0.0746, + "step": 17913 + }, + { + "epoch": 1.9421075455333912, + "grad_norm": 5.307658672332764, + "learning_rate": 7.078195270564341e-06, + "loss": 0.0576, + "step": 17914 + }, + { + "epoch": 1.942215958369471, + "grad_norm": 0.3107485771179199, + "learning_rate": 7.077469896996953e-06, + "loss": 0.0017, + "step": 17915 + }, + { + "epoch": 1.942324371205551, + "grad_norm": 1.5138742923736572, + "learning_rate": 7.076744523429567e-06, + "loss": 0.0346, + "step": 17916 + }, + { + "epoch": 1.9424327840416304, + "grad_norm": 6.712928771972656, + "learning_rate": 7.07601914986218e-06, + "loss": 0.0375, + "step": 17917 + }, + { + "epoch": 1.9425411968777104, + "grad_norm": 3.419071912765503, + "learning_rate": 7.075293776294793e-06, + "loss": 0.0222, + "step": 17918 + }, + { + "epoch": 1.94264960971379, + "grad_norm": 3.5883162021636963, + "learning_rate": 7.074568402727405e-06, + "loss": 0.0218, + "step": 17919 + }, + { + "epoch": 1.94275802254987, + "grad_norm": 0.7614724040031433, + "learning_rate": 7.073843029160018e-06, + "loss": 0.0145, + "step": 17920 + }, + { + "epoch": 1.9428664353859497, + "grad_norm": 2.190537214279175, + "learning_rate": 7.07311765559263e-06, + "loss": 0.0051, + "step": 17921 + }, + { + "epoch": 1.9429748482220295, + "grad_norm": 0.7923017144203186, + "learning_rate": 7.072392282025243e-06, + "loss": 0.0065, + "step": 17922 + }, + { + "epoch": 1.9430832610581092, + "grad_norm": 0.8216182589530945, + "learning_rate": 7.071666908457856e-06, + "loss": 0.0068, + "step": 17923 + }, + { + "epoch": 1.943191673894189, + "grad_norm": 0.18384702503681183, + "learning_rate": 7.070941534890468e-06, + "loss": 0.0014, + "step": 17924 + }, + { + "epoch": 1.943300086730269, + "grad_norm": 0.6939510703086853, + "learning_rate": 7.070216161323082e-06, + "loss": 0.0103, + "step": 17925 + }, + { + "epoch": 1.9434084995663485, + "grad_norm": 3.999584197998047, + "learning_rate": 7.069490787755695e-06, + "loss": 0.0309, + "step": 17926 + }, + { + "epoch": 1.9435169124024285, + "grad_norm": 3.6129214763641357, + "learning_rate": 7.068765414188307e-06, + "loss": 0.0381, + "step": 17927 + }, + { + "epoch": 1.9436253252385083, + "grad_norm": 1.2599356174468994, + "learning_rate": 7.06804004062092e-06, + "loss": 0.0065, + "step": 17928 + }, + { + "epoch": 1.943733738074588, + "grad_norm": 1.1877398490905762, + "learning_rate": 7.067314667053533e-06, + "loss": 0.0142, + "step": 17929 + }, + { + "epoch": 1.9438421509106678, + "grad_norm": 0.3551640212535858, + "learning_rate": 7.0665892934861455e-06, + "loss": 0.0028, + "step": 17930 + }, + { + "epoch": 1.9439505637467476, + "grad_norm": 0.2846727669239044, + "learning_rate": 7.0658639199187585e-06, + "loss": 0.0016, + "step": 17931 + }, + { + "epoch": 1.9440589765828276, + "grad_norm": 2.5478131771087646, + "learning_rate": 7.0651385463513715e-06, + "loss": 0.0068, + "step": 17932 + }, + { + "epoch": 1.944167389418907, + "grad_norm": 1.811079978942871, + "learning_rate": 7.0644131727839845e-06, + "loss": 0.0123, + "step": 17933 + }, + { + "epoch": 1.944275802254987, + "grad_norm": 0.7366249561309814, + "learning_rate": 7.0636877992165975e-06, + "loss": 0.0146, + "step": 17934 + }, + { + "epoch": 1.9443842150910666, + "grad_norm": 7.525284767150879, + "learning_rate": 7.0629624256492105e-06, + "loss": 0.0397, + "step": 17935 + }, + { + "epoch": 1.9444926279271466, + "grad_norm": 2.120971441268921, + "learning_rate": 7.062237052081823e-06, + "loss": 0.0075, + "step": 17936 + }, + { + "epoch": 1.9446010407632264, + "grad_norm": 0.2364509105682373, + "learning_rate": 7.061511678514436e-06, + "loss": 0.0026, + "step": 17937 + }, + { + "epoch": 1.9447094535993061, + "grad_norm": 3.316072463989258, + "learning_rate": 7.060786304947049e-06, + "loss": 0.0237, + "step": 17938 + }, + { + "epoch": 1.944817866435386, + "grad_norm": 5.153099536895752, + "learning_rate": 7.060060931379661e-06, + "loss": 0.0236, + "step": 17939 + }, + { + "epoch": 1.9449262792714657, + "grad_norm": 4.984487056732178, + "learning_rate": 7.059335557812274e-06, + "loss": 0.0192, + "step": 17940 + }, + { + "epoch": 1.9450346921075456, + "grad_norm": 5.077080726623535, + "learning_rate": 7.058610184244886e-06, + "loss": 0.042, + "step": 17941 + }, + { + "epoch": 1.9451431049436252, + "grad_norm": 2.6743111610412598, + "learning_rate": 7.0578848106775e-06, + "loss": 0.0394, + "step": 17942 + }, + { + "epoch": 1.9452515177797052, + "grad_norm": 5.442409515380859, + "learning_rate": 7.057159437110113e-06, + "loss": 0.0413, + "step": 17943 + }, + { + "epoch": 1.945359930615785, + "grad_norm": 0.8329946994781494, + "learning_rate": 7.056434063542726e-06, + "loss": 0.0057, + "step": 17944 + }, + { + "epoch": 1.9454683434518647, + "grad_norm": 4.209498405456543, + "learning_rate": 7.055708689975338e-06, + "loss": 0.0386, + "step": 17945 + }, + { + "epoch": 1.9455767562879445, + "grad_norm": 5.393014430999756, + "learning_rate": 7.054983316407951e-06, + "loss": 0.0515, + "step": 17946 + }, + { + "epoch": 1.9456851691240242, + "grad_norm": 3.0671212673187256, + "learning_rate": 7.054257942840563e-06, + "loss": 0.0097, + "step": 17947 + }, + { + "epoch": 1.9457935819601042, + "grad_norm": 3.1297810077667236, + "learning_rate": 7.053532569273176e-06, + "loss": 0.0166, + "step": 17948 + }, + { + "epoch": 1.9459019947961838, + "grad_norm": 0.6400583386421204, + "learning_rate": 7.052807195705789e-06, + "loss": 0.003, + "step": 17949 + }, + { + "epoch": 1.9460104076322637, + "grad_norm": 0.2877322733402252, + "learning_rate": 7.052081822138401e-06, + "loss": 0.0031, + "step": 17950 + }, + { + "epoch": 1.9461188204683435, + "grad_norm": 0.8814184069633484, + "learning_rate": 7.051356448571015e-06, + "loss": 0.0042, + "step": 17951 + }, + { + "epoch": 1.9462272333044233, + "grad_norm": 0.2421300858259201, + "learning_rate": 7.050631075003628e-06, + "loss": 0.0032, + "step": 17952 + }, + { + "epoch": 1.946335646140503, + "grad_norm": 2.703881025314331, + "learning_rate": 7.04990570143624e-06, + "loss": 0.0174, + "step": 17953 + }, + { + "epoch": 1.9464440589765828, + "grad_norm": 2.702272415161133, + "learning_rate": 7.049180327868853e-06, + "loss": 0.0115, + "step": 17954 + }, + { + "epoch": 1.9465524718126628, + "grad_norm": 1.2564514875411987, + "learning_rate": 7.048454954301466e-06, + "loss": 0.0066, + "step": 17955 + }, + { + "epoch": 1.9466608846487423, + "grad_norm": 2.51248836517334, + "learning_rate": 7.047729580734078e-06, + "loss": 0.0259, + "step": 17956 + }, + { + "epoch": 1.9467692974848223, + "grad_norm": 0.6255232691764832, + "learning_rate": 7.047004207166691e-06, + "loss": 0.0037, + "step": 17957 + }, + { + "epoch": 1.9468777103209018, + "grad_norm": 4.504007339477539, + "learning_rate": 7.046278833599304e-06, + "loss": 0.0395, + "step": 17958 + }, + { + "epoch": 1.9469861231569818, + "grad_norm": 3.601879119873047, + "learning_rate": 7.045553460031917e-06, + "loss": 0.0415, + "step": 17959 + }, + { + "epoch": 1.9470945359930616, + "grad_norm": 3.128690481185913, + "learning_rate": 7.04482808646453e-06, + "loss": 0.0403, + "step": 17960 + }, + { + "epoch": 1.9472029488291414, + "grad_norm": 2.305919885635376, + "learning_rate": 7.044102712897143e-06, + "loss": 0.011, + "step": 17961 + }, + { + "epoch": 1.9473113616652211, + "grad_norm": 1.5975371599197388, + "learning_rate": 7.043377339329755e-06, + "loss": 0.0258, + "step": 17962 + }, + { + "epoch": 1.9474197745013009, + "grad_norm": 0.22332113981246948, + "learning_rate": 7.042651965762368e-06, + "loss": 0.001, + "step": 17963 + }, + { + "epoch": 1.9475281873373809, + "grad_norm": 2.498232126235962, + "learning_rate": 7.041926592194981e-06, + "loss": 0.0055, + "step": 17964 + }, + { + "epoch": 1.9476366001734604, + "grad_norm": 1.9903645515441895, + "learning_rate": 7.041201218627593e-06, + "loss": 0.0267, + "step": 17965 + }, + { + "epoch": 1.9477450130095404, + "grad_norm": 5.4696044921875, + "learning_rate": 7.040475845060206e-06, + "loss": 0.051, + "step": 17966 + }, + { + "epoch": 1.9478534258456202, + "grad_norm": 3.695209503173828, + "learning_rate": 7.0397504714928185e-06, + "loss": 0.0132, + "step": 17967 + }, + { + "epoch": 1.9479618386817, + "grad_norm": 6.778685092926025, + "learning_rate": 7.039025097925432e-06, + "loss": 0.0544, + "step": 17968 + }, + { + "epoch": 1.9480702515177797, + "grad_norm": 0.29973304271698, + "learning_rate": 7.038299724358045e-06, + "loss": 0.0033, + "step": 17969 + }, + { + "epoch": 1.9481786643538594, + "grad_norm": 2.9321846961975098, + "learning_rate": 7.037574350790658e-06, + "loss": 0.0127, + "step": 17970 + }, + { + "epoch": 1.9482870771899394, + "grad_norm": 0.22963719069957733, + "learning_rate": 7.0368489772232705e-06, + "loss": 0.0017, + "step": 17971 + }, + { + "epoch": 1.948395490026019, + "grad_norm": 1.5808957815170288, + "learning_rate": 7.0361236036558835e-06, + "loss": 0.008, + "step": 17972 + }, + { + "epoch": 1.948503902862099, + "grad_norm": 3.8887860774993896, + "learning_rate": 7.035398230088496e-06, + "loss": 0.0816, + "step": 17973 + }, + { + "epoch": 1.9486123156981785, + "grad_norm": 2.042099952697754, + "learning_rate": 7.034672856521109e-06, + "loss": 0.0062, + "step": 17974 + }, + { + "epoch": 1.9487207285342585, + "grad_norm": 0.15767449140548706, + "learning_rate": 7.033947482953722e-06, + "loss": 0.0013, + "step": 17975 + }, + { + "epoch": 1.9488291413703382, + "grad_norm": 3.8195717334747314, + "learning_rate": 7.033222109386334e-06, + "loss": 0.0272, + "step": 17976 + }, + { + "epoch": 1.948937554206418, + "grad_norm": 1.2166852951049805, + "learning_rate": 7.032496735818948e-06, + "loss": 0.0444, + "step": 17977 + }, + { + "epoch": 1.949045967042498, + "grad_norm": 3.5684194564819336, + "learning_rate": 7.0317713622515606e-06, + "loss": 0.0252, + "step": 17978 + }, + { + "epoch": 1.9491543798785775, + "grad_norm": 5.393867015838623, + "learning_rate": 7.031045988684173e-06, + "loss": 0.0679, + "step": 17979 + }, + { + "epoch": 1.9492627927146575, + "grad_norm": 0.2715085446834564, + "learning_rate": 7.030320615116786e-06, + "loss": 0.0024, + "step": 17980 + }, + { + "epoch": 1.949371205550737, + "grad_norm": 3.0745606422424316, + "learning_rate": 7.029595241549399e-06, + "loss": 0.022, + "step": 17981 + }, + { + "epoch": 1.949479618386817, + "grad_norm": 1.526695966720581, + "learning_rate": 7.028869867982011e-06, + "loss": 0.0368, + "step": 17982 + }, + { + "epoch": 1.9495880312228968, + "grad_norm": 2.1502156257629395, + "learning_rate": 7.028144494414624e-06, + "loss": 0.0177, + "step": 17983 + }, + { + "epoch": 1.9496964440589766, + "grad_norm": 2.176203489303589, + "learning_rate": 7.027419120847237e-06, + "loss": 0.0203, + "step": 17984 + }, + { + "epoch": 1.9498048568950563, + "grad_norm": 1.9100863933563232, + "learning_rate": 7.02669374727985e-06, + "loss": 0.0179, + "step": 17985 + }, + { + "epoch": 1.949913269731136, + "grad_norm": 1.3424097299575806, + "learning_rate": 7.025968373712463e-06, + "loss": 0.0067, + "step": 17986 + }, + { + "epoch": 1.950021682567216, + "grad_norm": 0.314275324344635, + "learning_rate": 7.025243000145076e-06, + "loss": 0.0017, + "step": 17987 + }, + { + "epoch": 1.9501300954032956, + "grad_norm": 3.424131155014038, + "learning_rate": 7.024517626577688e-06, + "loss": 0.0329, + "step": 17988 + }, + { + "epoch": 1.9502385082393756, + "grad_norm": 5.680782318115234, + "learning_rate": 7.023792253010301e-06, + "loss": 0.0116, + "step": 17989 + }, + { + "epoch": 1.9503469210754554, + "grad_norm": 7.286358833312988, + "learning_rate": 7.023066879442914e-06, + "loss": 0.0165, + "step": 17990 + }, + { + "epoch": 1.9504553339115351, + "grad_norm": 0.46069273352622986, + "learning_rate": 7.022341505875526e-06, + "loss": 0.0025, + "step": 17991 + }, + { + "epoch": 1.950563746747615, + "grad_norm": 0.29210081696510315, + "learning_rate": 7.021616132308139e-06, + "loss": 0.0029, + "step": 17992 + }, + { + "epoch": 1.9506721595836947, + "grad_norm": 2.3617608547210693, + "learning_rate": 7.020890758740751e-06, + "loss": 0.028, + "step": 17993 + }, + { + "epoch": 1.9507805724197746, + "grad_norm": 2.3512957096099854, + "learning_rate": 7.020165385173365e-06, + "loss": 0.0252, + "step": 17994 + }, + { + "epoch": 1.9508889852558542, + "grad_norm": 5.730383396148682, + "learning_rate": 7.019440011605978e-06, + "loss": 0.0781, + "step": 17995 + }, + { + "epoch": 1.9509973980919342, + "grad_norm": 1.334599494934082, + "learning_rate": 7.018714638038591e-06, + "loss": 0.0033, + "step": 17996 + }, + { + "epoch": 1.9511058109280137, + "grad_norm": 1.7369194030761719, + "learning_rate": 7.017989264471203e-06, + "loss": 0.0129, + "step": 17997 + }, + { + "epoch": 1.9512142237640937, + "grad_norm": 11.695250511169434, + "learning_rate": 7.017263890903816e-06, + "loss": 0.0435, + "step": 17998 + }, + { + "epoch": 1.9513226366001735, + "grad_norm": 5.854657173156738, + "learning_rate": 7.016538517336428e-06, + "loss": 0.0155, + "step": 17999 + }, + { + "epoch": 1.9514310494362532, + "grad_norm": 3.1488254070281982, + "learning_rate": 7.015813143769041e-06, + "loss": 0.0136, + "step": 18000 + }, + { + "epoch": 1.951539462272333, + "grad_norm": 2.77109694480896, + "learning_rate": 7.015087770201654e-06, + "loss": 0.006, + "step": 18001 + }, + { + "epoch": 1.9516478751084128, + "grad_norm": 0.466431200504303, + "learning_rate": 7.014362396634268e-06, + "loss": 0.0028, + "step": 18002 + }, + { + "epoch": 1.9517562879444927, + "grad_norm": 3.0265026092529297, + "learning_rate": 7.01363702306688e-06, + "loss": 0.0164, + "step": 18003 + }, + { + "epoch": 1.9518647007805723, + "grad_norm": 7.218968391418457, + "learning_rate": 7.012911649499493e-06, + "loss": 0.0183, + "step": 18004 + }, + { + "epoch": 1.9519731136166523, + "grad_norm": 11.069844245910645, + "learning_rate": 7.012186275932105e-06, + "loss": 0.0474, + "step": 18005 + }, + { + "epoch": 1.952081526452732, + "grad_norm": 4.669299125671387, + "learning_rate": 7.011460902364718e-06, + "loss": 0.0444, + "step": 18006 + }, + { + "epoch": 1.9521899392888118, + "grad_norm": 1.9427632093429565, + "learning_rate": 7.010735528797331e-06, + "loss": 0.0239, + "step": 18007 + }, + { + "epoch": 1.9522983521248916, + "grad_norm": 3.1955552101135254, + "learning_rate": 7.0100101552299435e-06, + "loss": 0.0153, + "step": 18008 + }, + { + "epoch": 1.9524067649609713, + "grad_norm": 2.8423633575439453, + "learning_rate": 7.0092847816625565e-06, + "loss": 0.0129, + "step": 18009 + }, + { + "epoch": 1.9525151777970513, + "grad_norm": 7.317274570465088, + "learning_rate": 7.0085594080951695e-06, + "loss": 0.009, + "step": 18010 + }, + { + "epoch": 1.9526235906331308, + "grad_norm": 1.6267818212509155, + "learning_rate": 7.0078340345277825e-06, + "loss": 0.0214, + "step": 18011 + }, + { + "epoch": 1.9527320034692108, + "grad_norm": 0.3325774073600769, + "learning_rate": 7.0071086609603955e-06, + "loss": 0.0024, + "step": 18012 + }, + { + "epoch": 1.9528404163052906, + "grad_norm": 4.6141276359558105, + "learning_rate": 7.0063832873930085e-06, + "loss": 0.0372, + "step": 18013 + }, + { + "epoch": 1.9529488291413704, + "grad_norm": 3.089883327484131, + "learning_rate": 7.005657913825621e-06, + "loss": 0.0458, + "step": 18014 + }, + { + "epoch": 1.9530572419774501, + "grad_norm": 0.43684282898902893, + "learning_rate": 7.004932540258234e-06, + "loss": 0.0049, + "step": 18015 + }, + { + "epoch": 1.9531656548135299, + "grad_norm": 4.667379856109619, + "learning_rate": 7.004207166690847e-06, + "loss": 0.0484, + "step": 18016 + }, + { + "epoch": 1.9532740676496099, + "grad_norm": 5.784405708312988, + "learning_rate": 7.003481793123459e-06, + "loss": 0.0768, + "step": 18017 + }, + { + "epoch": 1.9533824804856894, + "grad_norm": 4.670140266418457, + "learning_rate": 7.002756419556072e-06, + "loss": 0.0239, + "step": 18018 + }, + { + "epoch": 1.9534908933217694, + "grad_norm": 0.7994441390037537, + "learning_rate": 7.002031045988684e-06, + "loss": 0.005, + "step": 18019 + }, + { + "epoch": 1.953599306157849, + "grad_norm": 3.7910146713256836, + "learning_rate": 7.001305672421298e-06, + "loss": 0.0456, + "step": 18020 + }, + { + "epoch": 1.953707718993929, + "grad_norm": 0.5172565579414368, + "learning_rate": 7.000580298853911e-06, + "loss": 0.0087, + "step": 18021 + }, + { + "epoch": 1.9538161318300087, + "grad_norm": 1.9690558910369873, + "learning_rate": 6.999854925286524e-06, + "loss": 0.0331, + "step": 18022 + }, + { + "epoch": 1.9539245446660884, + "grad_norm": 2.7263710498809814, + "learning_rate": 6.999129551719136e-06, + "loss": 0.0413, + "step": 18023 + }, + { + "epoch": 1.9540329575021682, + "grad_norm": 1.2378730773925781, + "learning_rate": 6.998404178151749e-06, + "loss": 0.0089, + "step": 18024 + }, + { + "epoch": 1.954141370338248, + "grad_norm": 4.330865859985352, + "learning_rate": 6.997678804584361e-06, + "loss": 0.0306, + "step": 18025 + }, + { + "epoch": 1.954249783174328, + "grad_norm": 3.6690070629119873, + "learning_rate": 6.996953431016974e-06, + "loss": 0.0396, + "step": 18026 + }, + { + "epoch": 1.9543581960104075, + "grad_norm": 2.163217067718506, + "learning_rate": 6.996228057449587e-06, + "loss": 0.0115, + "step": 18027 + }, + { + "epoch": 1.9544666088464875, + "grad_norm": 1.8995044231414795, + "learning_rate": 6.995502683882201e-06, + "loss": 0.0198, + "step": 18028 + }, + { + "epoch": 1.9545750216825672, + "grad_norm": 2.6886472702026367, + "learning_rate": 6.994777310314813e-06, + "loss": 0.0386, + "step": 18029 + }, + { + "epoch": 1.954683434518647, + "grad_norm": 3.490140676498413, + "learning_rate": 6.994051936747426e-06, + "loss": 0.0429, + "step": 18030 + }, + { + "epoch": 1.9547918473547268, + "grad_norm": 5.908344745635986, + "learning_rate": 6.993326563180038e-06, + "loss": 0.0158, + "step": 18031 + }, + { + "epoch": 1.9549002601908065, + "grad_norm": 0.8011672496795654, + "learning_rate": 6.992601189612651e-06, + "loss": 0.0432, + "step": 18032 + }, + { + "epoch": 1.9550086730268865, + "grad_norm": 0.30601343512535095, + "learning_rate": 6.991875816045264e-06, + "loss": 0.0017, + "step": 18033 + }, + { + "epoch": 1.955117085862966, + "grad_norm": 2.1149215698242188, + "learning_rate": 6.991150442477876e-06, + "loss": 0.0099, + "step": 18034 + }, + { + "epoch": 1.955225498699046, + "grad_norm": 2.612107753753662, + "learning_rate": 6.990425068910489e-06, + "loss": 0.0169, + "step": 18035 + }, + { + "epoch": 1.9553339115351258, + "grad_norm": 2.2135939598083496, + "learning_rate": 6.989699695343102e-06, + "loss": 0.005, + "step": 18036 + }, + { + "epoch": 1.9554423243712056, + "grad_norm": 14.047680854797363, + "learning_rate": 6.988974321775715e-06, + "loss": 0.0328, + "step": 18037 + }, + { + "epoch": 1.9555507372072853, + "grad_norm": 1.6258882284164429, + "learning_rate": 6.988248948208328e-06, + "loss": 0.0071, + "step": 18038 + }, + { + "epoch": 1.955659150043365, + "grad_norm": 3.543567419052124, + "learning_rate": 6.987523574640941e-06, + "loss": 0.0435, + "step": 18039 + }, + { + "epoch": 1.955767562879445, + "grad_norm": 0.5213366746902466, + "learning_rate": 6.986798201073553e-06, + "loss": 0.0061, + "step": 18040 + }, + { + "epoch": 1.9558759757155246, + "grad_norm": 0.9194290637969971, + "learning_rate": 6.986072827506166e-06, + "loss": 0.0364, + "step": 18041 + }, + { + "epoch": 1.9559843885516046, + "grad_norm": 1.3302556276321411, + "learning_rate": 6.985347453938779e-06, + "loss": 0.0109, + "step": 18042 + }, + { + "epoch": 1.9560928013876842, + "grad_norm": 0.8235804438591003, + "learning_rate": 6.984622080371391e-06, + "loss": 0.0054, + "step": 18043 + }, + { + "epoch": 1.9562012142237641, + "grad_norm": 0.4055383503437042, + "learning_rate": 6.983896706804004e-06, + "loss": 0.0032, + "step": 18044 + }, + { + "epoch": 1.956309627059844, + "grad_norm": 6.109950065612793, + "learning_rate": 6.9831713332366165e-06, + "loss": 0.0238, + "step": 18045 + }, + { + "epoch": 1.9564180398959237, + "grad_norm": 1.8139020204544067, + "learning_rate": 6.98244595966923e-06, + "loss": 0.0216, + "step": 18046 + }, + { + "epoch": 1.9565264527320034, + "grad_norm": 1.3892537355422974, + "learning_rate": 6.981720586101843e-06, + "loss": 0.0086, + "step": 18047 + }, + { + "epoch": 1.9566348655680832, + "grad_norm": 2.0387721061706543, + "learning_rate": 6.980995212534456e-06, + "loss": 0.0172, + "step": 18048 + }, + { + "epoch": 1.9567432784041632, + "grad_norm": 0.5508498549461365, + "learning_rate": 6.9802698389670685e-06, + "loss": 0.0114, + "step": 18049 + }, + { + "epoch": 1.9568516912402427, + "grad_norm": 0.6930046081542969, + "learning_rate": 6.9795444653996815e-06, + "loss": 0.0025, + "step": 18050 + }, + { + "epoch": 1.9569601040763227, + "grad_norm": 2.328777551651001, + "learning_rate": 6.978819091832294e-06, + "loss": 0.0476, + "step": 18051 + }, + { + "epoch": 1.9570685169124025, + "grad_norm": 7.914749622344971, + "learning_rate": 6.978093718264907e-06, + "loss": 0.015, + "step": 18052 + }, + { + "epoch": 1.9571769297484822, + "grad_norm": 0.7742642164230347, + "learning_rate": 6.97736834469752e-06, + "loss": 0.0032, + "step": 18053 + }, + { + "epoch": 1.957285342584562, + "grad_norm": 0.7382991313934326, + "learning_rate": 6.9766429711301334e-06, + "loss": 0.0075, + "step": 18054 + }, + { + "epoch": 1.9573937554206418, + "grad_norm": 0.4008611738681793, + "learning_rate": 6.975917597562746e-06, + "loss": 0.0042, + "step": 18055 + }, + { + "epoch": 1.9575021682567217, + "grad_norm": 2.054840564727783, + "learning_rate": 6.975192223995359e-06, + "loss": 0.0147, + "step": 18056 + }, + { + "epoch": 1.9576105810928013, + "grad_norm": 3.548201560974121, + "learning_rate": 6.974466850427971e-06, + "loss": 0.0229, + "step": 18057 + }, + { + "epoch": 1.9577189939288813, + "grad_norm": 8.5314302444458, + "learning_rate": 6.973741476860584e-06, + "loss": 0.0288, + "step": 18058 + }, + { + "epoch": 1.9578274067649608, + "grad_norm": 1.159806251525879, + "learning_rate": 6.973016103293197e-06, + "loss": 0.0215, + "step": 18059 + }, + { + "epoch": 1.9579358196010408, + "grad_norm": 9.977088928222656, + "learning_rate": 6.972290729725809e-06, + "loss": 0.0216, + "step": 18060 + }, + { + "epoch": 1.9580442324371206, + "grad_norm": 1.158306360244751, + "learning_rate": 6.971565356158422e-06, + "loss": 0.0044, + "step": 18061 + }, + { + "epoch": 1.9581526452732003, + "grad_norm": 3.4070510864257812, + "learning_rate": 6.970839982591035e-06, + "loss": 0.0267, + "step": 18062 + }, + { + "epoch": 1.95826105810928, + "grad_norm": 0.7784650325775146, + "learning_rate": 6.970114609023648e-06, + "loss": 0.0058, + "step": 18063 + }, + { + "epoch": 1.9583694709453598, + "grad_norm": 1.8081300258636475, + "learning_rate": 6.969389235456261e-06, + "loss": 0.014, + "step": 18064 + }, + { + "epoch": 1.9584778837814398, + "grad_norm": 5.734401226043701, + "learning_rate": 6.968663861888874e-06, + "loss": 0.0185, + "step": 18065 + }, + { + "epoch": 1.9585862966175194, + "grad_norm": 0.4213066101074219, + "learning_rate": 6.967938488321486e-06, + "loss": 0.0021, + "step": 18066 + }, + { + "epoch": 1.9586947094535994, + "grad_norm": 0.26416969299316406, + "learning_rate": 6.967213114754099e-06, + "loss": 0.0014, + "step": 18067 + }, + { + "epoch": 1.9588031222896791, + "grad_norm": 5.6432204246521, + "learning_rate": 6.966487741186712e-06, + "loss": 0.0254, + "step": 18068 + }, + { + "epoch": 1.9589115351257589, + "grad_norm": 3.539344310760498, + "learning_rate": 6.965762367619324e-06, + "loss": 0.0205, + "step": 18069 + }, + { + "epoch": 1.9590199479618386, + "grad_norm": 2.9353036880493164, + "learning_rate": 6.965036994051937e-06, + "loss": 0.0057, + "step": 18070 + }, + { + "epoch": 1.9591283607979184, + "grad_norm": 0.13975006341934204, + "learning_rate": 6.964311620484549e-06, + "loss": 0.0015, + "step": 18071 + }, + { + "epoch": 1.9592367736339984, + "grad_norm": 1.796552300453186, + "learning_rate": 6.963586246917163e-06, + "loss": 0.0301, + "step": 18072 + }, + { + "epoch": 1.959345186470078, + "grad_norm": 0.20322339236736298, + "learning_rate": 6.962860873349776e-06, + "loss": 0.0033, + "step": 18073 + }, + { + "epoch": 1.959453599306158, + "grad_norm": 1.4858442544937134, + "learning_rate": 6.962135499782389e-06, + "loss": 0.0162, + "step": 18074 + }, + { + "epoch": 1.9595620121422377, + "grad_norm": 3.829460382461548, + "learning_rate": 6.961410126215001e-06, + "loss": 0.0277, + "step": 18075 + }, + { + "epoch": 1.9596704249783174, + "grad_norm": 1.2371139526367188, + "learning_rate": 6.960684752647614e-06, + "loss": 0.0094, + "step": 18076 + }, + { + "epoch": 1.9597788378143972, + "grad_norm": 0.36601367592811584, + "learning_rate": 6.959959379080226e-06, + "loss": 0.0036, + "step": 18077 + }, + { + "epoch": 1.959887250650477, + "grad_norm": 1.2058249711990356, + "learning_rate": 6.959234005512839e-06, + "loss": 0.0101, + "step": 18078 + }, + { + "epoch": 1.959995663486557, + "grad_norm": 0.9611486792564392, + "learning_rate": 6.958508631945452e-06, + "loss": 0.0074, + "step": 18079 + }, + { + "epoch": 1.9601040763226365, + "grad_norm": 3.031730890274048, + "learning_rate": 6.957783258378066e-06, + "loss": 0.0511, + "step": 18080 + }, + { + "epoch": 1.9602124891587165, + "grad_norm": 1.7247095108032227, + "learning_rate": 6.957057884810678e-06, + "loss": 0.0216, + "step": 18081 + }, + { + "epoch": 1.960320901994796, + "grad_norm": 3.4595158100128174, + "learning_rate": 6.956332511243291e-06, + "loss": 0.0056, + "step": 18082 + }, + { + "epoch": 1.960429314830876, + "grad_norm": 0.5211765170097351, + "learning_rate": 6.955607137675903e-06, + "loss": 0.0033, + "step": 18083 + }, + { + "epoch": 1.9605377276669558, + "grad_norm": 0.6636249423027039, + "learning_rate": 6.954881764108516e-06, + "loss": 0.0065, + "step": 18084 + }, + { + "epoch": 1.9606461405030355, + "grad_norm": 4.052069664001465, + "learning_rate": 6.954156390541129e-06, + "loss": 0.0299, + "step": 18085 + }, + { + "epoch": 1.9607545533391153, + "grad_norm": 2.1571006774902344, + "learning_rate": 6.9534310169737415e-06, + "loss": 0.0472, + "step": 18086 + }, + { + "epoch": 1.960862966175195, + "grad_norm": 2.7163891792297363, + "learning_rate": 6.9527056434063545e-06, + "loss": 0.0664, + "step": 18087 + }, + { + "epoch": 1.960971379011275, + "grad_norm": 3.1930811405181885, + "learning_rate": 6.9519802698389675e-06, + "loss": 0.0396, + "step": 18088 + }, + { + "epoch": 1.9610797918473546, + "grad_norm": 2.1636781692504883, + "learning_rate": 6.9512548962715805e-06, + "loss": 0.0495, + "step": 18089 + }, + { + "epoch": 1.9611882046834346, + "grad_norm": 2.3498027324676514, + "learning_rate": 6.9505295227041935e-06, + "loss": 0.0352, + "step": 18090 + }, + { + "epoch": 1.9612966175195143, + "grad_norm": 2.8468029499053955, + "learning_rate": 6.9498041491368065e-06, + "loss": 0.0512, + "step": 18091 + }, + { + "epoch": 1.961405030355594, + "grad_norm": 2.6603150367736816, + "learning_rate": 6.949078775569419e-06, + "loss": 0.0042, + "step": 18092 + }, + { + "epoch": 1.9615134431916739, + "grad_norm": 2.0581324100494385, + "learning_rate": 6.948353402002032e-06, + "loss": 0.0054, + "step": 18093 + }, + { + "epoch": 1.9616218560277536, + "grad_norm": 8.968249320983887, + "learning_rate": 6.947628028434645e-06, + "loss": 0.0158, + "step": 18094 + }, + { + "epoch": 1.9617302688638336, + "grad_norm": 1.986219048500061, + "learning_rate": 6.946902654867257e-06, + "loss": 0.0037, + "step": 18095 + }, + { + "epoch": 1.9618386816999132, + "grad_norm": 2.1602976322174072, + "learning_rate": 6.94617728129987e-06, + "loss": 0.0253, + "step": 18096 + }, + { + "epoch": 1.9619470945359931, + "grad_norm": 7.571694850921631, + "learning_rate": 6.9454519077324836e-06, + "loss": 0.0242, + "step": 18097 + }, + { + "epoch": 1.962055507372073, + "grad_norm": 1.8197344541549683, + "learning_rate": 6.944726534165096e-06, + "loss": 0.0063, + "step": 18098 + }, + { + "epoch": 1.9621639202081527, + "grad_norm": 0.4753418564796448, + "learning_rate": 6.944001160597709e-06, + "loss": 0.0053, + "step": 18099 + }, + { + "epoch": 1.9622723330442324, + "grad_norm": 5.590159893035889, + "learning_rate": 6.943275787030322e-06, + "loss": 0.0619, + "step": 18100 + }, + { + "epoch": 1.9623807458803122, + "grad_norm": 1.199746012687683, + "learning_rate": 6.942550413462934e-06, + "loss": 0.0093, + "step": 18101 + }, + { + "epoch": 1.9624891587163922, + "grad_norm": 7.253795623779297, + "learning_rate": 6.941825039895547e-06, + "loss": 0.0429, + "step": 18102 + }, + { + "epoch": 1.9625975715524717, + "grad_norm": 1.2388869524002075, + "learning_rate": 6.941099666328159e-06, + "loss": 0.0068, + "step": 18103 + }, + { + "epoch": 1.9627059843885517, + "grad_norm": 4.470452308654785, + "learning_rate": 6.940374292760772e-06, + "loss": 0.0171, + "step": 18104 + }, + { + "epoch": 1.9628143972246312, + "grad_norm": 3.9604244232177734, + "learning_rate": 6.939648919193385e-06, + "loss": 0.0213, + "step": 18105 + }, + { + "epoch": 1.9629228100607112, + "grad_norm": 2.7260336875915527, + "learning_rate": 6.938923545625999e-06, + "loss": 0.0248, + "step": 18106 + }, + { + "epoch": 1.963031222896791, + "grad_norm": 2.766223907470703, + "learning_rate": 6.938198172058611e-06, + "loss": 0.0286, + "step": 18107 + }, + { + "epoch": 1.9631396357328708, + "grad_norm": 0.15221340954303741, + "learning_rate": 6.937472798491224e-06, + "loss": 0.0016, + "step": 18108 + }, + { + "epoch": 1.9632480485689505, + "grad_norm": 2.6092166900634766, + "learning_rate": 6.936747424923836e-06, + "loss": 0.0384, + "step": 18109 + }, + { + "epoch": 1.9633564614050303, + "grad_norm": 1.0283039808273315, + "learning_rate": 6.936022051356449e-06, + "loss": 0.008, + "step": 18110 + }, + { + "epoch": 1.9634648742411103, + "grad_norm": 1.93833589553833, + "learning_rate": 6.935296677789062e-06, + "loss": 0.0308, + "step": 18111 + }, + { + "epoch": 1.9635732870771898, + "grad_norm": 1.2037087678909302, + "learning_rate": 6.934571304221674e-06, + "loss": 0.0336, + "step": 18112 + }, + { + "epoch": 1.9636816999132698, + "grad_norm": 5.467023849487305, + "learning_rate": 6.933845930654287e-06, + "loss": 0.0612, + "step": 18113 + }, + { + "epoch": 1.9637901127493496, + "grad_norm": 6.5742878913879395, + "learning_rate": 6.9331205570869e-06, + "loss": 0.0267, + "step": 18114 + }, + { + "epoch": 1.9638985255854293, + "grad_norm": 1.0209726095199585, + "learning_rate": 6.932395183519513e-06, + "loss": 0.0039, + "step": 18115 + }, + { + "epoch": 1.964006938421509, + "grad_norm": 1.5348386764526367, + "learning_rate": 6.931669809952126e-06, + "loss": 0.0032, + "step": 18116 + }, + { + "epoch": 1.9641153512575888, + "grad_norm": 3.205228805541992, + "learning_rate": 6.930944436384739e-06, + "loss": 0.0123, + "step": 18117 + }, + { + "epoch": 1.9642237640936688, + "grad_norm": 0.6248228549957275, + "learning_rate": 6.930219062817351e-06, + "loss": 0.009, + "step": 18118 + }, + { + "epoch": 1.9643321769297484, + "grad_norm": 0.039940472692251205, + "learning_rate": 6.929493689249964e-06, + "loss": 0.0004, + "step": 18119 + }, + { + "epoch": 1.9644405897658284, + "grad_norm": 0.20021435618400574, + "learning_rate": 6.928768315682577e-06, + "loss": 0.0017, + "step": 18120 + }, + { + "epoch": 1.964549002601908, + "grad_norm": 4.5185723304748535, + "learning_rate": 6.928042942115189e-06, + "loss": 0.0393, + "step": 18121 + }, + { + "epoch": 1.9646574154379879, + "grad_norm": 1.0328295230865479, + "learning_rate": 6.927317568547802e-06, + "loss": 0.0067, + "step": 18122 + }, + { + "epoch": 1.9647658282740676, + "grad_norm": 3.2604572772979736, + "learning_rate": 6.926592194980416e-06, + "loss": 0.011, + "step": 18123 + }, + { + "epoch": 1.9648742411101474, + "grad_norm": 4.7924041748046875, + "learning_rate": 6.925866821413028e-06, + "loss": 0.0154, + "step": 18124 + }, + { + "epoch": 1.9649826539462274, + "grad_norm": 3.1931722164154053, + "learning_rate": 6.925141447845641e-06, + "loss": 0.0215, + "step": 18125 + }, + { + "epoch": 1.965091066782307, + "grad_norm": 3.4325926303863525, + "learning_rate": 6.924416074278254e-06, + "loss": 0.016, + "step": 18126 + }, + { + "epoch": 1.965199479618387, + "grad_norm": 3.77339768409729, + "learning_rate": 6.9236907007108665e-06, + "loss": 0.0581, + "step": 18127 + }, + { + "epoch": 1.9653078924544665, + "grad_norm": 8.159463882446289, + "learning_rate": 6.9229653271434795e-06, + "loss": 0.0157, + "step": 18128 + }, + { + "epoch": 1.9654163052905465, + "grad_norm": 0.15717381238937378, + "learning_rate": 6.922239953576092e-06, + "loss": 0.0026, + "step": 18129 + }, + { + "epoch": 1.9655247181266262, + "grad_norm": 3.2496445178985596, + "learning_rate": 6.921514580008705e-06, + "loss": 0.0153, + "step": 18130 + }, + { + "epoch": 1.965633130962706, + "grad_norm": 2.635040044784546, + "learning_rate": 6.920789206441318e-06, + "loss": 0.046, + "step": 18131 + }, + { + "epoch": 1.9657415437987857, + "grad_norm": 4.934934616088867, + "learning_rate": 6.9200638328739314e-06, + "loss": 0.0121, + "step": 18132 + }, + { + "epoch": 1.9658499566348655, + "grad_norm": 0.09405948966741562, + "learning_rate": 6.919338459306544e-06, + "loss": 0.001, + "step": 18133 + }, + { + "epoch": 1.9659583694709455, + "grad_norm": 0.298768550157547, + "learning_rate": 6.918613085739157e-06, + "loss": 0.0032, + "step": 18134 + }, + { + "epoch": 1.966066782307025, + "grad_norm": 4.35528039932251, + "learning_rate": 6.917887712171769e-06, + "loss": 0.0403, + "step": 18135 + }, + { + "epoch": 1.966175195143105, + "grad_norm": 4.81379508972168, + "learning_rate": 6.917162338604382e-06, + "loss": 0.0588, + "step": 18136 + }, + { + "epoch": 1.9662836079791848, + "grad_norm": 0.21374641358852386, + "learning_rate": 6.916436965036995e-06, + "loss": 0.0026, + "step": 18137 + }, + { + "epoch": 1.9663920208152645, + "grad_norm": 1.3782192468643188, + "learning_rate": 6.915711591469607e-06, + "loss": 0.0114, + "step": 18138 + }, + { + "epoch": 1.9665004336513443, + "grad_norm": 1.333816647529602, + "learning_rate": 6.91498621790222e-06, + "loss": 0.0061, + "step": 18139 + }, + { + "epoch": 1.966608846487424, + "grad_norm": 1.7088955640792847, + "learning_rate": 6.914260844334833e-06, + "loss": 0.0126, + "step": 18140 + }, + { + "epoch": 1.966717259323504, + "grad_norm": 2.706822395324707, + "learning_rate": 6.913535470767446e-06, + "loss": 0.0156, + "step": 18141 + }, + { + "epoch": 1.9668256721595836, + "grad_norm": 5.07773494720459, + "learning_rate": 6.912810097200059e-06, + "loss": 0.0576, + "step": 18142 + }, + { + "epoch": 1.9669340849956636, + "grad_norm": 0.8486219048500061, + "learning_rate": 6.912084723632672e-06, + "loss": 0.0029, + "step": 18143 + }, + { + "epoch": 1.9670424978317431, + "grad_norm": 1.617170810699463, + "learning_rate": 6.911359350065284e-06, + "loss": 0.0663, + "step": 18144 + }, + { + "epoch": 1.967150910667823, + "grad_norm": 3.1751010417938232, + "learning_rate": 6.910633976497897e-06, + "loss": 0.0172, + "step": 18145 + }, + { + "epoch": 1.9672593235039029, + "grad_norm": 5.765750885009766, + "learning_rate": 6.90990860293051e-06, + "loss": 0.0776, + "step": 18146 + }, + { + "epoch": 1.9673677363399826, + "grad_norm": 1.0850611925125122, + "learning_rate": 6.909183229363122e-06, + "loss": 0.0081, + "step": 18147 + }, + { + "epoch": 1.9674761491760624, + "grad_norm": 2.132922649383545, + "learning_rate": 6.908457855795735e-06, + "loss": 0.0231, + "step": 18148 + }, + { + "epoch": 1.9675845620121422, + "grad_norm": 10.184698104858398, + "learning_rate": 6.907732482228349e-06, + "loss": 0.0198, + "step": 18149 + }, + { + "epoch": 1.9676929748482221, + "grad_norm": 2.0415186882019043, + "learning_rate": 6.907007108660961e-06, + "loss": 0.0134, + "step": 18150 + }, + { + "epoch": 1.9678013876843017, + "grad_norm": 4.6721954345703125, + "learning_rate": 6.906281735093574e-06, + "loss": 0.0479, + "step": 18151 + }, + { + "epoch": 1.9679098005203817, + "grad_norm": 2.348785400390625, + "learning_rate": 6.905556361526187e-06, + "loss": 0.033, + "step": 18152 + }, + { + "epoch": 1.9680182133564614, + "grad_norm": 0.2522225081920624, + "learning_rate": 6.904830987958799e-06, + "loss": 0.0018, + "step": 18153 + }, + { + "epoch": 1.9681266261925412, + "grad_norm": 2.6739280223846436, + "learning_rate": 6.904105614391412e-06, + "loss": 0.0392, + "step": 18154 + }, + { + "epoch": 1.968235039028621, + "grad_norm": 3.916896343231201, + "learning_rate": 6.903380240824024e-06, + "loss": 0.0737, + "step": 18155 + }, + { + "epoch": 1.9683434518647007, + "grad_norm": 2.090334177017212, + "learning_rate": 6.902654867256637e-06, + "loss": 0.0128, + "step": 18156 + }, + { + "epoch": 1.9684518647007807, + "grad_norm": 2.1582441329956055, + "learning_rate": 6.90192949368925e-06, + "loss": 0.0133, + "step": 18157 + }, + { + "epoch": 1.9685602775368602, + "grad_norm": 1.7267030477523804, + "learning_rate": 6.901204120121864e-06, + "loss": 0.023, + "step": 18158 + }, + { + "epoch": 1.9686686903729402, + "grad_norm": 0.13589432835578918, + "learning_rate": 6.900478746554476e-06, + "loss": 0.0022, + "step": 18159 + }, + { + "epoch": 1.96877710320902, + "grad_norm": 2.559746742248535, + "learning_rate": 6.899753372987089e-06, + "loss": 0.0169, + "step": 18160 + }, + { + "epoch": 1.9688855160450998, + "grad_norm": 2.030557870864868, + "learning_rate": 6.899027999419701e-06, + "loss": 0.0064, + "step": 18161 + }, + { + "epoch": 1.9689939288811795, + "grad_norm": 0.38147902488708496, + "learning_rate": 6.898302625852314e-06, + "loss": 0.004, + "step": 18162 + }, + { + "epoch": 1.9691023417172593, + "grad_norm": 1.5727322101593018, + "learning_rate": 6.897577252284927e-06, + "loss": 0.0148, + "step": 18163 + }, + { + "epoch": 1.9692107545533393, + "grad_norm": 2.0003254413604736, + "learning_rate": 6.8968518787175395e-06, + "loss": 0.0157, + "step": 18164 + }, + { + "epoch": 1.9693191673894188, + "grad_norm": 0.5459766983985901, + "learning_rate": 6.8961265051501525e-06, + "loss": 0.0049, + "step": 18165 + }, + { + "epoch": 1.9694275802254988, + "grad_norm": 1.6356441974639893, + "learning_rate": 6.8954011315827655e-06, + "loss": 0.0214, + "step": 18166 + }, + { + "epoch": 1.9695359930615783, + "grad_norm": 6.4368085861206055, + "learning_rate": 6.894675758015379e-06, + "loss": 0.0495, + "step": 18167 + }, + { + "epoch": 1.9696444058976583, + "grad_norm": 2.2771522998809814, + "learning_rate": 6.8939503844479915e-06, + "loss": 0.0172, + "step": 18168 + }, + { + "epoch": 1.969752818733738, + "grad_norm": 9.386906623840332, + "learning_rate": 6.8932250108806045e-06, + "loss": 0.0342, + "step": 18169 + }, + { + "epoch": 1.9698612315698178, + "grad_norm": 2.22684645652771, + "learning_rate": 6.892499637313217e-06, + "loss": 0.0093, + "step": 18170 + }, + { + "epoch": 1.9699696444058976, + "grad_norm": 6.318423748016357, + "learning_rate": 6.89177426374583e-06, + "loss": 0.0517, + "step": 18171 + }, + { + "epoch": 1.9700780572419774, + "grad_norm": 1.8482770919799805, + "learning_rate": 6.891048890178443e-06, + "loss": 0.0091, + "step": 18172 + }, + { + "epoch": 1.9701864700780574, + "grad_norm": 1.5798143148422241, + "learning_rate": 6.890323516611055e-06, + "loss": 0.0221, + "step": 18173 + }, + { + "epoch": 1.970294882914137, + "grad_norm": 4.238429546356201, + "learning_rate": 6.889598143043668e-06, + "loss": 0.0287, + "step": 18174 + }, + { + "epoch": 1.9704032957502169, + "grad_norm": 2.8962225914001465, + "learning_rate": 6.8888727694762816e-06, + "loss": 0.004, + "step": 18175 + }, + { + "epoch": 1.9705117085862967, + "grad_norm": 8.513588905334473, + "learning_rate": 6.888147395908894e-06, + "loss": 0.035, + "step": 18176 + }, + { + "epoch": 1.9706201214223764, + "grad_norm": 1.3772287368774414, + "learning_rate": 6.887422022341507e-06, + "loss": 0.0365, + "step": 18177 + }, + { + "epoch": 1.9707285342584562, + "grad_norm": 2.6776235103607178, + "learning_rate": 6.88669664877412e-06, + "loss": 0.0278, + "step": 18178 + }, + { + "epoch": 1.970836947094536, + "grad_norm": 0.5900930166244507, + "learning_rate": 6.885971275206732e-06, + "loss": 0.0033, + "step": 18179 + }, + { + "epoch": 1.970945359930616, + "grad_norm": 4.751371383666992, + "learning_rate": 6.885245901639345e-06, + "loss": 0.0774, + "step": 18180 + }, + { + "epoch": 1.9710537727666955, + "grad_norm": 5.36806583404541, + "learning_rate": 6.884520528071957e-06, + "loss": 0.0641, + "step": 18181 + }, + { + "epoch": 1.9711621856027755, + "grad_norm": 5.2987847328186035, + "learning_rate": 6.88379515450457e-06, + "loss": 0.0348, + "step": 18182 + }, + { + "epoch": 1.9712705984388552, + "grad_norm": 5.602938652038574, + "learning_rate": 6.883069780937183e-06, + "loss": 0.0244, + "step": 18183 + }, + { + "epoch": 1.971379011274935, + "grad_norm": 0.8492171168327332, + "learning_rate": 6.882344407369797e-06, + "loss": 0.0074, + "step": 18184 + }, + { + "epoch": 1.9714874241110147, + "grad_norm": 6.906880855560303, + "learning_rate": 6.881619033802409e-06, + "loss": 0.0369, + "step": 18185 + }, + { + "epoch": 1.9715958369470945, + "grad_norm": 0.7888902425765991, + "learning_rate": 6.880893660235022e-06, + "loss": 0.0057, + "step": 18186 + }, + { + "epoch": 1.9717042497831745, + "grad_norm": 0.07789464294910431, + "learning_rate": 6.880168286667634e-06, + "loss": 0.001, + "step": 18187 + }, + { + "epoch": 1.971812662619254, + "grad_norm": 3.4373698234558105, + "learning_rate": 6.879442913100247e-06, + "loss": 0.023, + "step": 18188 + }, + { + "epoch": 1.971921075455334, + "grad_norm": 2.983248710632324, + "learning_rate": 6.87871753953286e-06, + "loss": 0.0255, + "step": 18189 + }, + { + "epoch": 1.9720294882914136, + "grad_norm": 0.8779605627059937, + "learning_rate": 6.877992165965472e-06, + "loss": 0.026, + "step": 18190 + }, + { + "epoch": 1.9721379011274935, + "grad_norm": 10.82855224609375, + "learning_rate": 6.877266792398085e-06, + "loss": 0.0338, + "step": 18191 + }, + { + "epoch": 1.9722463139635733, + "grad_norm": 2.061117649078369, + "learning_rate": 6.876541418830699e-06, + "loss": 0.0133, + "step": 18192 + }, + { + "epoch": 1.972354726799653, + "grad_norm": 4.96034574508667, + "learning_rate": 6.875816045263312e-06, + "loss": 0.1166, + "step": 18193 + }, + { + "epoch": 1.9724631396357328, + "grad_norm": 0.15065662562847137, + "learning_rate": 6.875090671695924e-06, + "loss": 0.0012, + "step": 18194 + }, + { + "epoch": 1.9725715524718126, + "grad_norm": 0.9548807740211487, + "learning_rate": 6.874365298128537e-06, + "loss": 0.0074, + "step": 18195 + }, + { + "epoch": 1.9726799653078926, + "grad_norm": 1.247611165046692, + "learning_rate": 6.873639924561149e-06, + "loss": 0.0118, + "step": 18196 + }, + { + "epoch": 1.9727883781439721, + "grad_norm": 1.189578652381897, + "learning_rate": 6.872914550993762e-06, + "loss": 0.008, + "step": 18197 + }, + { + "epoch": 1.972896790980052, + "grad_norm": 0.5059071183204651, + "learning_rate": 6.872189177426375e-06, + "loss": 0.0053, + "step": 18198 + }, + { + "epoch": 1.9730052038161319, + "grad_norm": 4.48055362701416, + "learning_rate": 6.871463803858987e-06, + "loss": 0.0116, + "step": 18199 + }, + { + "epoch": 1.9731136166522116, + "grad_norm": 1.9486905336380005, + "learning_rate": 6.8707384302916e-06, + "loss": 0.0142, + "step": 18200 + }, + { + "epoch": 1.9732220294882914, + "grad_norm": 2.253478527069092, + "learning_rate": 6.870013056724214e-06, + "loss": 0.019, + "step": 18201 + }, + { + "epoch": 1.9733304423243712, + "grad_norm": 0.20932653546333313, + "learning_rate": 6.869287683156826e-06, + "loss": 0.0025, + "step": 18202 + }, + { + "epoch": 1.9734388551604511, + "grad_norm": 0.16233490407466888, + "learning_rate": 6.868562309589439e-06, + "loss": 0.0013, + "step": 18203 + }, + { + "epoch": 1.9735472679965307, + "grad_norm": 2.6034128665924072, + "learning_rate": 6.867836936022052e-06, + "loss": 0.0053, + "step": 18204 + }, + { + "epoch": 1.9736556808326107, + "grad_norm": 3.116729974746704, + "learning_rate": 6.8671115624546645e-06, + "loss": 0.0188, + "step": 18205 + }, + { + "epoch": 1.9737640936686902, + "grad_norm": 0.12601837515830994, + "learning_rate": 6.8663861888872775e-06, + "loss": 0.0016, + "step": 18206 + }, + { + "epoch": 1.9738725065047702, + "grad_norm": 3.4856035709381104, + "learning_rate": 6.86566081531989e-06, + "loss": 0.0421, + "step": 18207 + }, + { + "epoch": 1.97398091934085, + "grad_norm": 0.3552723526954651, + "learning_rate": 6.864935441752503e-06, + "loss": 0.003, + "step": 18208 + }, + { + "epoch": 1.9740893321769297, + "grad_norm": 0.37327679991722107, + "learning_rate": 6.864210068185116e-06, + "loss": 0.0036, + "step": 18209 + }, + { + "epoch": 1.9741977450130095, + "grad_norm": 1.634547233581543, + "learning_rate": 6.8634846946177295e-06, + "loss": 0.0253, + "step": 18210 + }, + { + "epoch": 1.9743061578490892, + "grad_norm": 2.3039298057556152, + "learning_rate": 6.862759321050342e-06, + "loss": 0.0163, + "step": 18211 + }, + { + "epoch": 1.9744145706851692, + "grad_norm": 3.7631828784942627, + "learning_rate": 6.862033947482955e-06, + "loss": 0.0725, + "step": 18212 + }, + { + "epoch": 1.9745229835212488, + "grad_norm": 1.6084954738616943, + "learning_rate": 6.861308573915567e-06, + "loss": 0.0654, + "step": 18213 + }, + { + "epoch": 1.9746313963573288, + "grad_norm": 6.682602405548096, + "learning_rate": 6.86058320034818e-06, + "loss": 0.0111, + "step": 18214 + }, + { + "epoch": 1.9747398091934085, + "grad_norm": 1.2076237201690674, + "learning_rate": 6.859857826780793e-06, + "loss": 0.0245, + "step": 18215 + }, + { + "epoch": 1.9748482220294883, + "grad_norm": 3.845414400100708, + "learning_rate": 6.859132453213405e-06, + "loss": 0.0102, + "step": 18216 + }, + { + "epoch": 1.974956634865568, + "grad_norm": 0.11717523634433746, + "learning_rate": 6.858407079646018e-06, + "loss": 0.0013, + "step": 18217 + }, + { + "epoch": 1.9750650477016478, + "grad_norm": 3.8134396076202393, + "learning_rate": 6.857681706078632e-06, + "loss": 0.0095, + "step": 18218 + }, + { + "epoch": 1.9751734605377278, + "grad_norm": 1.7340079545974731, + "learning_rate": 6.856956332511245e-06, + "loss": 0.0171, + "step": 18219 + }, + { + "epoch": 1.9752818733738073, + "grad_norm": 4.004318714141846, + "learning_rate": 6.856230958943857e-06, + "loss": 0.0054, + "step": 18220 + }, + { + "epoch": 1.9753902862098873, + "grad_norm": 6.842373847961426, + "learning_rate": 6.85550558537647e-06, + "loss": 0.0293, + "step": 18221 + }, + { + "epoch": 1.975498699045967, + "grad_norm": 0.6996886134147644, + "learning_rate": 6.854780211809082e-06, + "loss": 0.0038, + "step": 18222 + }, + { + "epoch": 1.9756071118820469, + "grad_norm": 2.0870423316955566, + "learning_rate": 6.854054838241695e-06, + "loss": 0.0124, + "step": 18223 + }, + { + "epoch": 1.9757155247181266, + "grad_norm": 4.22847318649292, + "learning_rate": 6.853329464674308e-06, + "loss": 0.0302, + "step": 18224 + }, + { + "epoch": 1.9758239375542064, + "grad_norm": 3.7154693603515625, + "learning_rate": 6.85260409110692e-06, + "loss": 0.0638, + "step": 18225 + }, + { + "epoch": 1.9759323503902864, + "grad_norm": 6.037874698638916, + "learning_rate": 6.851878717539533e-06, + "loss": 0.0485, + "step": 18226 + }, + { + "epoch": 1.976040763226366, + "grad_norm": 0.5169271230697632, + "learning_rate": 6.851153343972147e-06, + "loss": 0.0033, + "step": 18227 + }, + { + "epoch": 1.976149176062446, + "grad_norm": 1.7216277122497559, + "learning_rate": 6.850427970404759e-06, + "loss": 0.0131, + "step": 18228 + }, + { + "epoch": 1.9762575888985254, + "grad_norm": 4.51123046875, + "learning_rate": 6.849702596837372e-06, + "loss": 0.0298, + "step": 18229 + }, + { + "epoch": 1.9763660017346054, + "grad_norm": 0.08267836272716522, + "learning_rate": 6.848977223269985e-06, + "loss": 0.0009, + "step": 18230 + }, + { + "epoch": 1.9764744145706852, + "grad_norm": 2.3178670406341553, + "learning_rate": 6.848251849702597e-06, + "loss": 0.0543, + "step": 18231 + }, + { + "epoch": 1.976582827406765, + "grad_norm": 2.721768856048584, + "learning_rate": 6.84752647613521e-06, + "loss": 0.0264, + "step": 18232 + }, + { + "epoch": 1.9766912402428447, + "grad_norm": 1.8028295040130615, + "learning_rate": 6.846801102567822e-06, + "loss": 0.0231, + "step": 18233 + }, + { + "epoch": 1.9767996530789245, + "grad_norm": 3.2159082889556885, + "learning_rate": 6.846075729000435e-06, + "loss": 0.0451, + "step": 18234 + }, + { + "epoch": 1.9769080659150045, + "grad_norm": 0.7171160578727722, + "learning_rate": 6.845350355433048e-06, + "loss": 0.0073, + "step": 18235 + }, + { + "epoch": 1.977016478751084, + "grad_norm": 1.879333734512329, + "learning_rate": 6.844624981865662e-06, + "loss": 0.0266, + "step": 18236 + }, + { + "epoch": 1.977124891587164, + "grad_norm": 4.2617268562316895, + "learning_rate": 6.843899608298274e-06, + "loss": 0.035, + "step": 18237 + }, + { + "epoch": 1.9772333044232437, + "grad_norm": 0.41264140605926514, + "learning_rate": 6.843174234730887e-06, + "loss": 0.0038, + "step": 18238 + }, + { + "epoch": 1.9773417172593235, + "grad_norm": 2.1078004837036133, + "learning_rate": 6.842448861163499e-06, + "loss": 0.0093, + "step": 18239 + }, + { + "epoch": 1.9774501300954033, + "grad_norm": 3.8388149738311768, + "learning_rate": 6.841723487596112e-06, + "loss": 0.0475, + "step": 18240 + }, + { + "epoch": 1.977558542931483, + "grad_norm": 1.580928921699524, + "learning_rate": 6.840998114028725e-06, + "loss": 0.0512, + "step": 18241 + }, + { + "epoch": 1.977666955767563, + "grad_norm": 7.365316390991211, + "learning_rate": 6.8402727404613375e-06, + "loss": 0.0363, + "step": 18242 + }, + { + "epoch": 1.9777753686036426, + "grad_norm": 0.22416448593139648, + "learning_rate": 6.8395473668939505e-06, + "loss": 0.0025, + "step": 18243 + }, + { + "epoch": 1.9778837814397225, + "grad_norm": 3.2048726081848145, + "learning_rate": 6.838821993326564e-06, + "loss": 0.013, + "step": 18244 + }, + { + "epoch": 1.9779921942758023, + "grad_norm": 0.3061470091342926, + "learning_rate": 6.838096619759177e-06, + "loss": 0.0019, + "step": 18245 + }, + { + "epoch": 1.978100607111882, + "grad_norm": 0.7042983174324036, + "learning_rate": 6.8373712461917895e-06, + "loss": 0.0094, + "step": 18246 + }, + { + "epoch": 1.9782090199479618, + "grad_norm": 0.3844785690307617, + "learning_rate": 6.8366458726244025e-06, + "loss": 0.0037, + "step": 18247 + }, + { + "epoch": 1.9783174327840416, + "grad_norm": 1.7172006368637085, + "learning_rate": 6.835920499057015e-06, + "loss": 0.0187, + "step": 18248 + }, + { + "epoch": 1.9784258456201216, + "grad_norm": 4.494373321533203, + "learning_rate": 6.835195125489628e-06, + "loss": 0.0299, + "step": 18249 + }, + { + "epoch": 1.9785342584562011, + "grad_norm": 0.9409076571464539, + "learning_rate": 6.834469751922241e-06, + "loss": 0.0131, + "step": 18250 + }, + { + "epoch": 1.978642671292281, + "grad_norm": 1.1138637065887451, + "learning_rate": 6.833744378354853e-06, + "loss": 0.0192, + "step": 18251 + }, + { + "epoch": 1.9787510841283606, + "grad_norm": 4.5414910316467285, + "learning_rate": 6.833019004787466e-06, + "loss": 0.0114, + "step": 18252 + }, + { + "epoch": 1.9788594969644406, + "grad_norm": 3.2402360439300537, + "learning_rate": 6.8322936312200796e-06, + "loss": 0.0091, + "step": 18253 + }, + { + "epoch": 1.9789679098005204, + "grad_norm": 5.578888893127441, + "learning_rate": 6.831568257652692e-06, + "loss": 0.0691, + "step": 18254 + }, + { + "epoch": 1.9790763226366002, + "grad_norm": 2.081972599029541, + "learning_rate": 6.830842884085305e-06, + "loss": 0.0069, + "step": 18255 + }, + { + "epoch": 1.97918473547268, + "grad_norm": 6.463046550750732, + "learning_rate": 6.830117510517918e-06, + "loss": 0.0445, + "step": 18256 + }, + { + "epoch": 1.9792931483087597, + "grad_norm": 7.940140724182129, + "learning_rate": 6.82939213695053e-06, + "loss": 0.0603, + "step": 18257 + }, + { + "epoch": 1.9794015611448397, + "grad_norm": 0.5365954637527466, + "learning_rate": 6.828666763383143e-06, + "loss": 0.0058, + "step": 18258 + }, + { + "epoch": 1.9795099739809192, + "grad_norm": 1.4699550867080688, + "learning_rate": 6.827941389815755e-06, + "loss": 0.0046, + "step": 18259 + }, + { + "epoch": 1.9796183868169992, + "grad_norm": 2.8502988815307617, + "learning_rate": 6.827216016248368e-06, + "loss": 0.0252, + "step": 18260 + }, + { + "epoch": 1.979726799653079, + "grad_norm": 2.222459316253662, + "learning_rate": 6.826490642680981e-06, + "loss": 0.0208, + "step": 18261 + }, + { + "epoch": 1.9798352124891587, + "grad_norm": 0.33876657485961914, + "learning_rate": 6.825765269113595e-06, + "loss": 0.005, + "step": 18262 + }, + { + "epoch": 1.9799436253252385, + "grad_norm": 1.233499526977539, + "learning_rate": 6.825039895546207e-06, + "loss": 0.0064, + "step": 18263 + }, + { + "epoch": 1.9800520381613183, + "grad_norm": 0.3466504216194153, + "learning_rate": 6.82431452197882e-06, + "loss": 0.0036, + "step": 18264 + }, + { + "epoch": 1.9801604509973982, + "grad_norm": 1.915102481842041, + "learning_rate": 6.823589148411433e-06, + "loss": 0.0068, + "step": 18265 + }, + { + "epoch": 1.9802688638334778, + "grad_norm": 0.12522560358047485, + "learning_rate": 6.822863774844045e-06, + "loss": 0.0016, + "step": 18266 + }, + { + "epoch": 1.9803772766695578, + "grad_norm": 0.749691903591156, + "learning_rate": 6.822138401276658e-06, + "loss": 0.0296, + "step": 18267 + }, + { + "epoch": 1.9804856895056373, + "grad_norm": 0.475411057472229, + "learning_rate": 6.82141302770927e-06, + "loss": 0.006, + "step": 18268 + }, + { + "epoch": 1.9805941023417173, + "grad_norm": 1.9634389877319336, + "learning_rate": 6.820687654141883e-06, + "loss": 0.0226, + "step": 18269 + }, + { + "epoch": 1.980702515177797, + "grad_norm": 0.4668957591056824, + "learning_rate": 6.819962280574497e-06, + "loss": 0.0051, + "step": 18270 + }, + { + "epoch": 1.9808109280138768, + "grad_norm": 1.238060712814331, + "learning_rate": 6.81923690700711e-06, + "loss": 0.0086, + "step": 18271 + }, + { + "epoch": 1.9809193408499568, + "grad_norm": 1.8624546527862549, + "learning_rate": 6.818511533439722e-06, + "loss": 0.0114, + "step": 18272 + }, + { + "epoch": 1.9810277536860363, + "grad_norm": 0.542323648929596, + "learning_rate": 6.817786159872335e-06, + "loss": 0.0054, + "step": 18273 + }, + { + "epoch": 1.9811361665221163, + "grad_norm": 3.148494243621826, + "learning_rate": 6.817060786304947e-06, + "loss": 0.014, + "step": 18274 + }, + { + "epoch": 1.9812445793581959, + "grad_norm": 4.597592353820801, + "learning_rate": 6.81633541273756e-06, + "loss": 0.0111, + "step": 18275 + }, + { + "epoch": 1.9813529921942759, + "grad_norm": 0.5547083616256714, + "learning_rate": 6.815610039170173e-06, + "loss": 0.053, + "step": 18276 + }, + { + "epoch": 1.9814614050303556, + "grad_norm": 2.486635446548462, + "learning_rate": 6.814884665602785e-06, + "loss": 0.0124, + "step": 18277 + }, + { + "epoch": 1.9815698178664354, + "grad_norm": 4.274168491363525, + "learning_rate": 6.814159292035398e-06, + "loss": 0.0283, + "step": 18278 + }, + { + "epoch": 1.9816782307025151, + "grad_norm": 0.2401869148015976, + "learning_rate": 6.813433918468012e-06, + "loss": 0.0015, + "step": 18279 + }, + { + "epoch": 1.981786643538595, + "grad_norm": 2.031097173690796, + "learning_rate": 6.812708544900624e-06, + "loss": 0.0258, + "step": 18280 + }, + { + "epoch": 1.981895056374675, + "grad_norm": 0.10355248302221298, + "learning_rate": 6.811983171333237e-06, + "loss": 0.0012, + "step": 18281 + }, + { + "epoch": 1.9820034692107544, + "grad_norm": 2.853149890899658, + "learning_rate": 6.81125779776585e-06, + "loss": 0.0338, + "step": 18282 + }, + { + "epoch": 1.9821118820468344, + "grad_norm": 7.087488651275635, + "learning_rate": 6.8105324241984625e-06, + "loss": 0.0243, + "step": 18283 + }, + { + "epoch": 1.9822202948829142, + "grad_norm": 7.717792510986328, + "learning_rate": 6.8098070506310755e-06, + "loss": 0.0439, + "step": 18284 + }, + { + "epoch": 1.982328707718994, + "grad_norm": 3.7167654037475586, + "learning_rate": 6.809081677063688e-06, + "loss": 0.0546, + "step": 18285 + }, + { + "epoch": 1.9824371205550737, + "grad_norm": 3.4458985328674316, + "learning_rate": 6.808356303496301e-06, + "loss": 0.0651, + "step": 18286 + }, + { + "epoch": 1.9825455333911535, + "grad_norm": 0.8481817841529846, + "learning_rate": 6.8076309299289145e-06, + "loss": 0.0057, + "step": 18287 + }, + { + "epoch": 1.9826539462272335, + "grad_norm": 5.432412624359131, + "learning_rate": 6.8069055563615275e-06, + "loss": 0.035, + "step": 18288 + }, + { + "epoch": 1.982762359063313, + "grad_norm": 3.812976598739624, + "learning_rate": 6.80618018279414e-06, + "loss": 0.0359, + "step": 18289 + }, + { + "epoch": 1.982870771899393, + "grad_norm": 2.547898530960083, + "learning_rate": 6.805454809226753e-06, + "loss": 0.0345, + "step": 18290 + }, + { + "epoch": 1.9829791847354725, + "grad_norm": 0.11551161855459213, + "learning_rate": 6.804729435659366e-06, + "loss": 0.0015, + "step": 18291 + }, + { + "epoch": 1.9830875975715525, + "grad_norm": 1.9502047300338745, + "learning_rate": 6.804004062091978e-06, + "loss": 0.0232, + "step": 18292 + }, + { + "epoch": 1.9831960104076323, + "grad_norm": 3.9314892292022705, + "learning_rate": 6.803278688524591e-06, + "loss": 0.0352, + "step": 18293 + }, + { + "epoch": 1.983304423243712, + "grad_norm": 2.02229380607605, + "learning_rate": 6.802553314957203e-06, + "loss": 0.0111, + "step": 18294 + }, + { + "epoch": 1.9834128360797918, + "grad_norm": 1.2534425258636475, + "learning_rate": 6.801827941389816e-06, + "loss": 0.0074, + "step": 18295 + }, + { + "epoch": 1.9835212489158716, + "grad_norm": 0.2660433351993561, + "learning_rate": 6.80110256782243e-06, + "loss": 0.0021, + "step": 18296 + }, + { + "epoch": 1.9836296617519515, + "grad_norm": 3.841221809387207, + "learning_rate": 6.800377194255043e-06, + "loss": 0.0218, + "step": 18297 + }, + { + "epoch": 1.983738074588031, + "grad_norm": 8.101666450500488, + "learning_rate": 6.799651820687655e-06, + "loss": 0.0114, + "step": 18298 + }, + { + "epoch": 1.983846487424111, + "grad_norm": 0.6409260034561157, + "learning_rate": 6.798926447120268e-06, + "loss": 0.0042, + "step": 18299 + }, + { + "epoch": 1.9839549002601908, + "grad_norm": 2.5382707118988037, + "learning_rate": 6.79820107355288e-06, + "loss": 0.0107, + "step": 18300 + }, + { + "epoch": 1.9840633130962706, + "grad_norm": 1.0703351497650146, + "learning_rate": 6.797475699985493e-06, + "loss": 0.0084, + "step": 18301 + }, + { + "epoch": 1.9841717259323504, + "grad_norm": 0.9720844626426697, + "learning_rate": 6.796750326418106e-06, + "loss": 0.017, + "step": 18302 + }, + { + "epoch": 1.9842801387684301, + "grad_norm": 3.0050957202911377, + "learning_rate": 6.796024952850718e-06, + "loss": 0.0151, + "step": 18303 + }, + { + "epoch": 1.98438855160451, + "grad_norm": 3.1100852489471436, + "learning_rate": 6.795299579283331e-06, + "loss": 0.0466, + "step": 18304 + }, + { + "epoch": 1.9844969644405897, + "grad_norm": 3.0466246604919434, + "learning_rate": 6.794574205715945e-06, + "loss": 0.0098, + "step": 18305 + }, + { + "epoch": 1.9846053772766696, + "grad_norm": 3.0572991371154785, + "learning_rate": 6.793848832148557e-06, + "loss": 0.0396, + "step": 18306 + }, + { + "epoch": 1.9847137901127494, + "grad_norm": 0.40548497438430786, + "learning_rate": 6.79312345858117e-06, + "loss": 0.0021, + "step": 18307 + }, + { + "epoch": 1.9848222029488292, + "grad_norm": 1.8278870582580566, + "learning_rate": 6.792398085013783e-06, + "loss": 0.0304, + "step": 18308 + }, + { + "epoch": 1.984930615784909, + "grad_norm": 2.6325314044952393, + "learning_rate": 6.791672711446395e-06, + "loss": 0.0096, + "step": 18309 + }, + { + "epoch": 1.9850390286209887, + "grad_norm": 2.89683198928833, + "learning_rate": 6.790947337879008e-06, + "loss": 0.0122, + "step": 18310 + }, + { + "epoch": 1.9851474414570687, + "grad_norm": 1.891961932182312, + "learning_rate": 6.79022196431162e-06, + "loss": 0.0231, + "step": 18311 + }, + { + "epoch": 1.9852558542931482, + "grad_norm": 1.9132896661758423, + "learning_rate": 6.789496590744233e-06, + "loss": 0.0038, + "step": 18312 + }, + { + "epoch": 1.9853642671292282, + "grad_norm": 0.14449627697467804, + "learning_rate": 6.788771217176847e-06, + "loss": 0.0015, + "step": 18313 + }, + { + "epoch": 1.9854726799653077, + "grad_norm": 5.277917861938477, + "learning_rate": 6.78804584360946e-06, + "loss": 0.0246, + "step": 18314 + }, + { + "epoch": 1.9855810928013877, + "grad_norm": 1.0848866701126099, + "learning_rate": 6.787320470042072e-06, + "loss": 0.002, + "step": 18315 + }, + { + "epoch": 1.9856895056374675, + "grad_norm": 0.6254496574401855, + "learning_rate": 6.786595096474685e-06, + "loss": 0.0052, + "step": 18316 + }, + { + "epoch": 1.9857979184735473, + "grad_norm": 1.4818778038024902, + "learning_rate": 6.785869722907298e-06, + "loss": 0.016, + "step": 18317 + }, + { + "epoch": 1.985906331309627, + "grad_norm": 1.760986328125, + "learning_rate": 6.78514434933991e-06, + "loss": 0.0343, + "step": 18318 + }, + { + "epoch": 1.9860147441457068, + "grad_norm": 5.393112659454346, + "learning_rate": 6.784418975772523e-06, + "loss": 0.0334, + "step": 18319 + }, + { + "epoch": 1.9861231569817868, + "grad_norm": 7.6300578117370605, + "learning_rate": 6.7836936022051355e-06, + "loss": 0.0386, + "step": 18320 + }, + { + "epoch": 1.9862315698178663, + "grad_norm": 2.370274782180786, + "learning_rate": 6.7829682286377485e-06, + "loss": 0.0138, + "step": 18321 + }, + { + "epoch": 1.9863399826539463, + "grad_norm": 5.170624256134033, + "learning_rate": 6.782242855070362e-06, + "loss": 0.027, + "step": 18322 + }, + { + "epoch": 1.986448395490026, + "grad_norm": 4.7317094802856445, + "learning_rate": 6.781517481502975e-06, + "loss": 0.0404, + "step": 18323 + }, + { + "epoch": 1.9865568083261058, + "grad_norm": 3.948882818222046, + "learning_rate": 6.7807921079355875e-06, + "loss": 0.0102, + "step": 18324 + }, + { + "epoch": 1.9866652211621856, + "grad_norm": 1.046417474746704, + "learning_rate": 6.7800667343682005e-06, + "loss": 0.0258, + "step": 18325 + }, + { + "epoch": 1.9867736339982653, + "grad_norm": 2.156822919845581, + "learning_rate": 6.779341360800813e-06, + "loss": 0.0152, + "step": 18326 + }, + { + "epoch": 1.9868820468343453, + "grad_norm": 0.6602420210838318, + "learning_rate": 6.778615987233426e-06, + "loss": 0.0053, + "step": 18327 + }, + { + "epoch": 1.9869904596704249, + "grad_norm": 0.1374022364616394, + "learning_rate": 6.777890613666039e-06, + "loss": 0.0016, + "step": 18328 + }, + { + "epoch": 1.9870988725065049, + "grad_norm": 3.7065720558166504, + "learning_rate": 6.777165240098651e-06, + "loss": 0.0362, + "step": 18329 + }, + { + "epoch": 1.9872072853425846, + "grad_norm": 2.201439142227173, + "learning_rate": 6.776439866531264e-06, + "loss": 0.0464, + "step": 18330 + }, + { + "epoch": 1.9873156981786644, + "grad_norm": 1.808409333229065, + "learning_rate": 6.775714492963878e-06, + "loss": 0.0255, + "step": 18331 + }, + { + "epoch": 1.9874241110147441, + "grad_norm": 4.000988483428955, + "learning_rate": 6.77498911939649e-06, + "loss": 0.0057, + "step": 18332 + }, + { + "epoch": 1.987532523850824, + "grad_norm": 1.7983585596084595, + "learning_rate": 6.774263745829103e-06, + "loss": 0.025, + "step": 18333 + }, + { + "epoch": 1.987640936686904, + "grad_norm": 2.163036346435547, + "learning_rate": 6.773538372261716e-06, + "loss": 0.0328, + "step": 18334 + }, + { + "epoch": 1.9877493495229834, + "grad_norm": 0.38621607422828674, + "learning_rate": 6.772812998694328e-06, + "loss": 0.0017, + "step": 18335 + }, + { + "epoch": 1.9878577623590634, + "grad_norm": 2.726454019546509, + "learning_rate": 6.772087625126941e-06, + "loss": 0.0411, + "step": 18336 + }, + { + "epoch": 1.987966175195143, + "grad_norm": 3.241682291030884, + "learning_rate": 6.771362251559553e-06, + "loss": 0.023, + "step": 18337 + }, + { + "epoch": 1.988074588031223, + "grad_norm": 7.360343933105469, + "learning_rate": 6.770636877992166e-06, + "loss": 0.02, + "step": 18338 + }, + { + "epoch": 1.9881830008673027, + "grad_norm": 0.9278129935264587, + "learning_rate": 6.76991150442478e-06, + "loss": 0.0025, + "step": 18339 + }, + { + "epoch": 1.9882914137033825, + "grad_norm": 5.795496940612793, + "learning_rate": 6.769186130857393e-06, + "loss": 0.0182, + "step": 18340 + }, + { + "epoch": 1.9883998265394622, + "grad_norm": 1.5483973026275635, + "learning_rate": 6.768460757290005e-06, + "loss": 0.0348, + "step": 18341 + }, + { + "epoch": 1.988508239375542, + "grad_norm": 1.5945111513137817, + "learning_rate": 6.767735383722618e-06, + "loss": 0.0287, + "step": 18342 + }, + { + "epoch": 1.988616652211622, + "grad_norm": 3.44338059425354, + "learning_rate": 6.767010010155231e-06, + "loss": 0.0278, + "step": 18343 + }, + { + "epoch": 1.9887250650477015, + "grad_norm": 0.07284332811832428, + "learning_rate": 6.766284636587843e-06, + "loss": 0.001, + "step": 18344 + }, + { + "epoch": 1.9888334778837815, + "grad_norm": 4.230914115905762, + "learning_rate": 6.765559263020456e-06, + "loss": 0.0556, + "step": 18345 + }, + { + "epoch": 1.9889418907198613, + "grad_norm": 0.4354470670223236, + "learning_rate": 6.764833889453068e-06, + "loss": 0.0042, + "step": 18346 + }, + { + "epoch": 1.989050303555941, + "grad_norm": 4.238332748413086, + "learning_rate": 6.764108515885681e-06, + "loss": 0.0776, + "step": 18347 + }, + { + "epoch": 1.9891587163920208, + "grad_norm": 1.155198574066162, + "learning_rate": 6.763383142318295e-06, + "loss": 0.0127, + "step": 18348 + }, + { + "epoch": 1.9892671292281006, + "grad_norm": 5.619193077087402, + "learning_rate": 6.762657768750908e-06, + "loss": 0.0209, + "step": 18349 + }, + { + "epoch": 1.9893755420641805, + "grad_norm": 5.0104451179504395, + "learning_rate": 6.76193239518352e-06, + "loss": 0.0472, + "step": 18350 + }, + { + "epoch": 1.98948395490026, + "grad_norm": 2.379591464996338, + "learning_rate": 6.761207021616133e-06, + "loss": 0.0216, + "step": 18351 + }, + { + "epoch": 1.98959236773634, + "grad_norm": 1.3068268299102783, + "learning_rate": 6.760481648048745e-06, + "loss": 0.0227, + "step": 18352 + }, + { + "epoch": 1.9897007805724196, + "grad_norm": 1.5566902160644531, + "learning_rate": 6.759756274481358e-06, + "loss": 0.0318, + "step": 18353 + }, + { + "epoch": 1.9898091934084996, + "grad_norm": 0.922030508518219, + "learning_rate": 6.759030900913971e-06, + "loss": 0.0101, + "step": 18354 + }, + { + "epoch": 1.9899176062445794, + "grad_norm": 1.8738360404968262, + "learning_rate": 6.758305527346583e-06, + "loss": 0.0361, + "step": 18355 + }, + { + "epoch": 1.9900260190806591, + "grad_norm": 0.7404544949531555, + "learning_rate": 6.757580153779196e-06, + "loss": 0.0175, + "step": 18356 + }, + { + "epoch": 1.9901344319167389, + "grad_norm": 6.3782877922058105, + "learning_rate": 6.75685478021181e-06, + "loss": 0.0445, + "step": 18357 + }, + { + "epoch": 1.9902428447528187, + "grad_norm": 5.147428035736084, + "learning_rate": 6.756129406644422e-06, + "loss": 0.0147, + "step": 18358 + }, + { + "epoch": 1.9903512575888986, + "grad_norm": 0.7117996215820312, + "learning_rate": 6.755404033077035e-06, + "loss": 0.0085, + "step": 18359 + }, + { + "epoch": 1.9904596704249782, + "grad_norm": 0.25867632031440735, + "learning_rate": 6.754678659509648e-06, + "loss": 0.0024, + "step": 18360 + }, + { + "epoch": 1.9905680832610582, + "grad_norm": 6.723546504974365, + "learning_rate": 6.7539532859422605e-06, + "loss": 0.0143, + "step": 18361 + }, + { + "epoch": 1.990676496097138, + "grad_norm": 1.3487112522125244, + "learning_rate": 6.7532279123748735e-06, + "loss": 0.0229, + "step": 18362 + }, + { + "epoch": 1.9907849089332177, + "grad_norm": 4.459421157836914, + "learning_rate": 6.752502538807486e-06, + "loss": 0.0533, + "step": 18363 + }, + { + "epoch": 1.9908933217692975, + "grad_norm": 0.16204535961151123, + "learning_rate": 6.751777165240099e-06, + "loss": 0.0012, + "step": 18364 + }, + { + "epoch": 1.9910017346053772, + "grad_norm": 1.8898965120315552, + "learning_rate": 6.7510517916727125e-06, + "loss": 0.0198, + "step": 18365 + }, + { + "epoch": 1.9911101474414572, + "grad_norm": 0.34257951378822327, + "learning_rate": 6.7503264181053255e-06, + "loss": 0.0053, + "step": 18366 + }, + { + "epoch": 1.9912185602775367, + "grad_norm": 5.624545097351074, + "learning_rate": 6.749601044537938e-06, + "loss": 0.0373, + "step": 18367 + }, + { + "epoch": 1.9913269731136167, + "grad_norm": 0.3363441824913025, + "learning_rate": 6.748875670970551e-06, + "loss": 0.0038, + "step": 18368 + }, + { + "epoch": 1.9914353859496965, + "grad_norm": 2.888662815093994, + "learning_rate": 6.748150297403164e-06, + "loss": 0.0052, + "step": 18369 + }, + { + "epoch": 1.9915437987857763, + "grad_norm": 1.8283997774124146, + "learning_rate": 6.747424923835776e-06, + "loss": 0.0274, + "step": 18370 + }, + { + "epoch": 1.991652211621856, + "grad_norm": 3.1849253177642822, + "learning_rate": 6.746699550268389e-06, + "loss": 0.0097, + "step": 18371 + }, + { + "epoch": 1.9917606244579358, + "grad_norm": 4.271515846252441, + "learning_rate": 6.745974176701001e-06, + "loss": 0.0127, + "step": 18372 + }, + { + "epoch": 1.9918690372940158, + "grad_norm": 4.704086780548096, + "learning_rate": 6.745248803133614e-06, + "loss": 0.0258, + "step": 18373 + }, + { + "epoch": 1.9919774501300953, + "grad_norm": 1.2411640882492065, + "learning_rate": 6.744523429566228e-06, + "loss": 0.0157, + "step": 18374 + }, + { + "epoch": 1.9920858629661753, + "grad_norm": 2.1267213821411133, + "learning_rate": 6.743798055998841e-06, + "loss": 0.0121, + "step": 18375 + }, + { + "epoch": 1.9921942758022548, + "grad_norm": 1.2208819389343262, + "learning_rate": 6.743072682431453e-06, + "loss": 0.0113, + "step": 18376 + }, + { + "epoch": 1.9923026886383348, + "grad_norm": 2.5694739818573, + "learning_rate": 6.742347308864066e-06, + "loss": 0.012, + "step": 18377 + }, + { + "epoch": 1.9924111014744146, + "grad_norm": 0.8827324509620667, + "learning_rate": 6.741621935296678e-06, + "loss": 0.0046, + "step": 18378 + }, + { + "epoch": 1.9925195143104943, + "grad_norm": 1.8182326555252075, + "learning_rate": 6.740896561729291e-06, + "loss": 0.0154, + "step": 18379 + }, + { + "epoch": 1.992627927146574, + "grad_norm": 3.0440468788146973, + "learning_rate": 6.740171188161904e-06, + "loss": 0.0172, + "step": 18380 + }, + { + "epoch": 1.9927363399826539, + "grad_norm": 1.1161972284317017, + "learning_rate": 6.739445814594516e-06, + "loss": 0.0149, + "step": 18381 + }, + { + "epoch": 1.9928447528187339, + "grad_norm": 3.935944080352783, + "learning_rate": 6.73872044102713e-06, + "loss": 0.031, + "step": 18382 + }, + { + "epoch": 1.9929531656548134, + "grad_norm": 0.701568067073822, + "learning_rate": 6.737995067459743e-06, + "loss": 0.0048, + "step": 18383 + }, + { + "epoch": 1.9930615784908934, + "grad_norm": 5.1116814613342285, + "learning_rate": 6.737269693892355e-06, + "loss": 0.0121, + "step": 18384 + }, + { + "epoch": 1.9931699913269731, + "grad_norm": 2.1677870750427246, + "learning_rate": 6.736544320324968e-06, + "loss": 0.0406, + "step": 18385 + }, + { + "epoch": 1.993278404163053, + "grad_norm": 0.505874752998352, + "learning_rate": 6.735818946757581e-06, + "loss": 0.0055, + "step": 18386 + }, + { + "epoch": 1.9933868169991327, + "grad_norm": 2.3533775806427, + "learning_rate": 6.735093573190193e-06, + "loss": 0.0191, + "step": 18387 + }, + { + "epoch": 1.9934952298352124, + "grad_norm": 0.14374029636383057, + "learning_rate": 6.734368199622806e-06, + "loss": 0.0012, + "step": 18388 + }, + { + "epoch": 1.9936036426712924, + "grad_norm": 0.7120038866996765, + "learning_rate": 6.733642826055419e-06, + "loss": 0.0034, + "step": 18389 + }, + { + "epoch": 1.993712055507372, + "grad_norm": 3.634190559387207, + "learning_rate": 6.732917452488031e-06, + "loss": 0.0162, + "step": 18390 + }, + { + "epoch": 1.993820468343452, + "grad_norm": 2.354193925857544, + "learning_rate": 6.732192078920645e-06, + "loss": 0.0467, + "step": 18391 + }, + { + "epoch": 1.9939288811795317, + "grad_norm": 1.6400816440582275, + "learning_rate": 6.731466705353258e-06, + "loss": 0.007, + "step": 18392 + }, + { + "epoch": 1.9940372940156115, + "grad_norm": 0.39880257844924927, + "learning_rate": 6.73074133178587e-06, + "loss": 0.0038, + "step": 18393 + }, + { + "epoch": 1.9941457068516912, + "grad_norm": 0.3133695721626282, + "learning_rate": 6.730015958218483e-06, + "loss": 0.0031, + "step": 18394 + }, + { + "epoch": 1.994254119687771, + "grad_norm": 3.4526922702789307, + "learning_rate": 6.729290584651096e-06, + "loss": 0.0113, + "step": 18395 + }, + { + "epoch": 1.994362532523851, + "grad_norm": 2.289522171020508, + "learning_rate": 6.728565211083708e-06, + "loss": 0.0092, + "step": 18396 + }, + { + "epoch": 1.9944709453599305, + "grad_norm": 2.5627784729003906, + "learning_rate": 6.727839837516321e-06, + "loss": 0.0178, + "step": 18397 + }, + { + "epoch": 1.9945793581960105, + "grad_norm": 3.7465202808380127, + "learning_rate": 6.7271144639489335e-06, + "loss": 0.051, + "step": 18398 + }, + { + "epoch": 1.99468777103209, + "grad_norm": 1.4376968145370483, + "learning_rate": 6.7263890903815465e-06, + "loss": 0.0251, + "step": 18399 + }, + { + "epoch": 1.99479618386817, + "grad_norm": 0.864587664604187, + "learning_rate": 6.72566371681416e-06, + "loss": 0.0206, + "step": 18400 + }, + { + "epoch": 1.9949045967042498, + "grad_norm": 0.9285323023796082, + "learning_rate": 6.724938343246773e-06, + "loss": 0.0038, + "step": 18401 + }, + { + "epoch": 1.9950130095403296, + "grad_norm": 7.782749652862549, + "learning_rate": 6.7242129696793855e-06, + "loss": 0.0665, + "step": 18402 + }, + { + "epoch": 1.9951214223764093, + "grad_norm": 1.6024417877197266, + "learning_rate": 6.7234875961119985e-06, + "loss": 0.0116, + "step": 18403 + }, + { + "epoch": 1.995229835212489, + "grad_norm": 3.314088821411133, + "learning_rate": 6.722762222544611e-06, + "loss": 0.0256, + "step": 18404 + }, + { + "epoch": 1.995338248048569, + "grad_norm": 2.644761085510254, + "learning_rate": 6.722036848977224e-06, + "loss": 0.0169, + "step": 18405 + }, + { + "epoch": 1.9954466608846486, + "grad_norm": 4.58701753616333, + "learning_rate": 6.721311475409837e-06, + "loss": 0.0171, + "step": 18406 + }, + { + "epoch": 1.9955550737207286, + "grad_norm": 5.337417125701904, + "learning_rate": 6.720586101842449e-06, + "loss": 0.0512, + "step": 18407 + }, + { + "epoch": 1.9956634865568084, + "grad_norm": 1.491666555404663, + "learning_rate": 6.719860728275063e-06, + "loss": 0.0078, + "step": 18408 + }, + { + "epoch": 1.9957718993928881, + "grad_norm": 3.4180824756622314, + "learning_rate": 6.719135354707676e-06, + "loss": 0.0262, + "step": 18409 + }, + { + "epoch": 1.995880312228968, + "grad_norm": 3.9994258880615234, + "learning_rate": 6.718409981140288e-06, + "loss": 0.0337, + "step": 18410 + }, + { + "epoch": 1.9959887250650477, + "grad_norm": 1.0023213624954224, + "learning_rate": 6.717684607572901e-06, + "loss": 0.0266, + "step": 18411 + }, + { + "epoch": 1.9960971379011276, + "grad_norm": 7.469665050506592, + "learning_rate": 6.716959234005514e-06, + "loss": 0.0915, + "step": 18412 + }, + { + "epoch": 1.9962055507372072, + "grad_norm": 0.8656811118125916, + "learning_rate": 6.716233860438126e-06, + "loss": 0.0063, + "step": 18413 + }, + { + "epoch": 1.9963139635732872, + "grad_norm": 1.3319110870361328, + "learning_rate": 6.715508486870739e-06, + "loss": 0.0116, + "step": 18414 + }, + { + "epoch": 1.9964223764093667, + "grad_norm": 6.059831142425537, + "learning_rate": 6.714783113303352e-06, + "loss": 0.0322, + "step": 18415 + }, + { + "epoch": 1.9965307892454467, + "grad_norm": 0.3630380928516388, + "learning_rate": 6.714057739735964e-06, + "loss": 0.0009, + "step": 18416 + }, + { + "epoch": 1.9966392020815265, + "grad_norm": 3.852360248565674, + "learning_rate": 6.713332366168578e-06, + "loss": 0.0485, + "step": 18417 + }, + { + "epoch": 1.9967476149176062, + "grad_norm": 1.8950847387313843, + "learning_rate": 6.712606992601191e-06, + "loss": 0.04, + "step": 18418 + }, + { + "epoch": 1.9968560277536862, + "grad_norm": 2.9020543098449707, + "learning_rate": 6.711881619033803e-06, + "loss": 0.0247, + "step": 18419 + }, + { + "epoch": 1.9969644405897657, + "grad_norm": 1.4694385528564453, + "learning_rate": 6.711156245466416e-06, + "loss": 0.0068, + "step": 18420 + }, + { + "epoch": 1.9970728534258457, + "grad_norm": 4.494928359985352, + "learning_rate": 6.710430871899029e-06, + "loss": 0.0235, + "step": 18421 + }, + { + "epoch": 1.9971812662619253, + "grad_norm": 0.9014991521835327, + "learning_rate": 6.709705498331641e-06, + "loss": 0.0061, + "step": 18422 + }, + { + "epoch": 1.9972896790980053, + "grad_norm": 4.51359224319458, + "learning_rate": 6.708980124764254e-06, + "loss": 0.034, + "step": 18423 + }, + { + "epoch": 1.997398091934085, + "grad_norm": 1.616974949836731, + "learning_rate": 6.708254751196866e-06, + "loss": 0.0234, + "step": 18424 + }, + { + "epoch": 1.9975065047701648, + "grad_norm": 0.9505412578582764, + "learning_rate": 6.707529377629479e-06, + "loss": 0.0059, + "step": 18425 + }, + { + "epoch": 1.9976149176062445, + "grad_norm": 2.1638200283050537, + "learning_rate": 6.706804004062093e-06, + "loss": 0.0112, + "step": 18426 + }, + { + "epoch": 1.9977233304423243, + "grad_norm": 1.2955764532089233, + "learning_rate": 6.706078630494706e-06, + "loss": 0.0135, + "step": 18427 + }, + { + "epoch": 1.9978317432784043, + "grad_norm": 0.9034083485603333, + "learning_rate": 6.705353256927318e-06, + "loss": 0.0077, + "step": 18428 + }, + { + "epoch": 1.9979401561144838, + "grad_norm": 1.0243275165557861, + "learning_rate": 6.704627883359931e-06, + "loss": 0.0083, + "step": 18429 + }, + { + "epoch": 1.9980485689505638, + "grad_norm": 2.901423215866089, + "learning_rate": 6.703902509792543e-06, + "loss": 0.0219, + "step": 18430 + }, + { + "epoch": 1.9981569817866436, + "grad_norm": 3.1981453895568848, + "learning_rate": 6.703177136225156e-06, + "loss": 0.0162, + "step": 18431 + }, + { + "epoch": 1.9982653946227233, + "grad_norm": 5.934446334838867, + "learning_rate": 6.702451762657769e-06, + "loss": 0.0268, + "step": 18432 + }, + { + "epoch": 1.998373807458803, + "grad_norm": 0.9904366731643677, + "learning_rate": 6.7017263890903814e-06, + "loss": 0.0132, + "step": 18433 + }, + { + "epoch": 1.9984822202948829, + "grad_norm": 1.8133232593536377, + "learning_rate": 6.701001015522995e-06, + "loss": 0.0153, + "step": 18434 + }, + { + "epoch": 1.9985906331309629, + "grad_norm": 2.8980178833007812, + "learning_rate": 6.700275641955608e-06, + "loss": 0.0375, + "step": 18435 + }, + { + "epoch": 1.9986990459670424, + "grad_norm": 0.863410472869873, + "learning_rate": 6.69955026838822e-06, + "loss": 0.0037, + "step": 18436 + }, + { + "epoch": 1.9988074588031224, + "grad_norm": 0.30839377641677856, + "learning_rate": 6.698824894820833e-06, + "loss": 0.0025, + "step": 18437 + }, + { + "epoch": 1.998915871639202, + "grad_norm": 1.5581347942352295, + "learning_rate": 6.698099521253446e-06, + "loss": 0.0299, + "step": 18438 + }, + { + "epoch": 1.999024284475282, + "grad_norm": 1.6731927394866943, + "learning_rate": 6.6973741476860585e-06, + "loss": 0.0081, + "step": 18439 + }, + { + "epoch": 1.9991326973113617, + "grad_norm": 4.763548374176025, + "learning_rate": 6.6966487741186715e-06, + "loss": 0.0478, + "step": 18440 + }, + { + "epoch": 1.9992411101474414, + "grad_norm": 9.164690971374512, + "learning_rate": 6.6959234005512845e-06, + "loss": 0.033, + "step": 18441 + }, + { + "epoch": 1.9993495229835212, + "grad_norm": 0.2954227030277252, + "learning_rate": 6.695198026983897e-06, + "loss": 0.0025, + "step": 18442 + }, + { + "epoch": 1.999457935819601, + "grad_norm": 7.383391857147217, + "learning_rate": 6.6944726534165105e-06, + "loss": 0.0114, + "step": 18443 + }, + { + "epoch": 1.999566348655681, + "grad_norm": 1.8250772953033447, + "learning_rate": 6.6937472798491235e-06, + "loss": 0.0265, + "step": 18444 + }, + { + "epoch": 1.9996747614917605, + "grad_norm": 4.971262454986572, + "learning_rate": 6.693021906281736e-06, + "loss": 0.0182, + "step": 18445 + }, + { + "epoch": 1.9997831743278405, + "grad_norm": 1.684003233909607, + "learning_rate": 6.692296532714349e-06, + "loss": 0.0266, + "step": 18446 + }, + { + "epoch": 1.9998915871639202, + "grad_norm": 3.978257179260254, + "learning_rate": 6.691571159146962e-06, + "loss": 0.028, + "step": 18447 + }, + { + "epoch": 2.0, + "grad_norm": 0.3508358895778656, + "learning_rate": 6.690845785579574e-06, + "loss": 0.0031, + "step": 18448 + }, + { + "epoch": 2.00010841283608, + "grad_norm": 0.4964468479156494, + "learning_rate": 6.690120412012187e-06, + "loss": 0.0021, + "step": 18449 + }, + { + "epoch": 2.0002168256721595, + "grad_norm": 0.03044886142015457, + "learning_rate": 6.689395038444799e-06, + "loss": 0.0004, + "step": 18450 + }, + { + "epoch": 2.0003252385082395, + "grad_norm": 2.5706889629364014, + "learning_rate": 6.688669664877412e-06, + "loss": 0.0326, + "step": 18451 + }, + { + "epoch": 2.000433651344319, + "grad_norm": 5.0189642906188965, + "learning_rate": 6.687944291310026e-06, + "loss": 0.0062, + "step": 18452 + }, + { + "epoch": 2.000542064180399, + "grad_norm": 1.7198379039764404, + "learning_rate": 6.687218917742639e-06, + "loss": 0.0111, + "step": 18453 + }, + { + "epoch": 2.0006504770164786, + "grad_norm": 0.32070571184158325, + "learning_rate": 6.686493544175251e-06, + "loss": 0.0026, + "step": 18454 + }, + { + "epoch": 2.0007588898525586, + "grad_norm": 2.156562089920044, + "learning_rate": 6.685768170607864e-06, + "loss": 0.0071, + "step": 18455 + }, + { + "epoch": 2.0008673026886385, + "grad_norm": 2.0333807468414307, + "learning_rate": 6.685042797040476e-06, + "loss": 0.0628, + "step": 18456 + }, + { + "epoch": 2.000975715524718, + "grad_norm": 4.2782883644104, + "learning_rate": 6.684317423473089e-06, + "loss": 0.0381, + "step": 18457 + }, + { + "epoch": 2.001084128360798, + "grad_norm": 0.1406174898147583, + "learning_rate": 6.683592049905702e-06, + "loss": 0.0009, + "step": 18458 + }, + { + "epoch": 2.0011925411968776, + "grad_norm": 2.3186819553375244, + "learning_rate": 6.682866676338314e-06, + "loss": 0.0216, + "step": 18459 + }, + { + "epoch": 2.0013009540329576, + "grad_norm": 1.9477479457855225, + "learning_rate": 6.682141302770928e-06, + "loss": 0.0308, + "step": 18460 + }, + { + "epoch": 2.001409366869037, + "grad_norm": 0.45714882016181946, + "learning_rate": 6.681415929203541e-06, + "loss": 0.0039, + "step": 18461 + }, + { + "epoch": 2.001517779705117, + "grad_norm": 1.647862195968628, + "learning_rate": 6.680690555636153e-06, + "loss": 0.0205, + "step": 18462 + }, + { + "epoch": 2.0016261925411967, + "grad_norm": 0.44668084383010864, + "learning_rate": 6.679965182068766e-06, + "loss": 0.0049, + "step": 18463 + }, + { + "epoch": 2.0017346053772767, + "grad_norm": 1.8720343112945557, + "learning_rate": 6.679239808501379e-06, + "loss": 0.0233, + "step": 18464 + }, + { + "epoch": 2.0018430182133566, + "grad_norm": 0.6600099802017212, + "learning_rate": 6.678514434933991e-06, + "loss": 0.0057, + "step": 18465 + }, + { + "epoch": 2.001951431049436, + "grad_norm": 0.5002751350402832, + "learning_rate": 6.677789061366604e-06, + "loss": 0.0046, + "step": 18466 + }, + { + "epoch": 2.002059843885516, + "grad_norm": 2.323695659637451, + "learning_rate": 6.677063687799217e-06, + "loss": 0.0455, + "step": 18467 + }, + { + "epoch": 2.0021682567215957, + "grad_norm": 0.2332804650068283, + "learning_rate": 6.676338314231829e-06, + "loss": 0.0029, + "step": 18468 + }, + { + "epoch": 2.0022766695576757, + "grad_norm": 1.490610957145691, + "learning_rate": 6.675612940664443e-06, + "loss": 0.0119, + "step": 18469 + }, + { + "epoch": 2.0023850823937552, + "grad_norm": 2.651865005493164, + "learning_rate": 6.674887567097056e-06, + "loss": 0.0113, + "step": 18470 + }, + { + "epoch": 2.002493495229835, + "grad_norm": 1.1212273836135864, + "learning_rate": 6.674162193529668e-06, + "loss": 0.0056, + "step": 18471 + }, + { + "epoch": 2.002601908065915, + "grad_norm": 0.7122533321380615, + "learning_rate": 6.673436819962281e-06, + "loss": 0.0064, + "step": 18472 + }, + { + "epoch": 2.0027103209019947, + "grad_norm": 1.1471941471099854, + "learning_rate": 6.672711446394894e-06, + "loss": 0.0136, + "step": 18473 + }, + { + "epoch": 2.0028187337380747, + "grad_norm": 0.7262167930603027, + "learning_rate": 6.671986072827506e-06, + "loss": 0.0072, + "step": 18474 + }, + { + "epoch": 2.0029271465741543, + "grad_norm": 3.63851261138916, + "learning_rate": 6.671260699260119e-06, + "loss": 0.0112, + "step": 18475 + }, + { + "epoch": 2.0030355594102343, + "grad_norm": 1.5445904731750488, + "learning_rate": 6.6705353256927315e-06, + "loss": 0.0213, + "step": 18476 + }, + { + "epoch": 2.003143972246314, + "grad_norm": 0.22184322774410248, + "learning_rate": 6.6698099521253445e-06, + "loss": 0.0023, + "step": 18477 + }, + { + "epoch": 2.003252385082394, + "grad_norm": 1.0321300029754639, + "learning_rate": 6.669084578557958e-06, + "loss": 0.0106, + "step": 18478 + }, + { + "epoch": 2.0033607979184738, + "grad_norm": 0.9807632565498352, + "learning_rate": 6.668359204990571e-06, + "loss": 0.0021, + "step": 18479 + }, + { + "epoch": 2.0034692107545533, + "grad_norm": 0.03605649620294571, + "learning_rate": 6.6676338314231835e-06, + "loss": 0.0004, + "step": 18480 + }, + { + "epoch": 2.0035776235906333, + "grad_norm": 0.3361533284187317, + "learning_rate": 6.6669084578557965e-06, + "loss": 0.0029, + "step": 18481 + }, + { + "epoch": 2.003686036426713, + "grad_norm": 0.14926370978355408, + "learning_rate": 6.666183084288409e-06, + "loss": 0.0015, + "step": 18482 + }, + { + "epoch": 2.003794449262793, + "grad_norm": 8.423782348632812, + "learning_rate": 6.665457710721022e-06, + "loss": 0.0178, + "step": 18483 + }, + { + "epoch": 2.0039028620988724, + "grad_norm": 1.4658628702163696, + "learning_rate": 6.664732337153635e-06, + "loss": 0.007, + "step": 18484 + }, + { + "epoch": 2.0040112749349523, + "grad_norm": 0.22157125174999237, + "learning_rate": 6.664006963586247e-06, + "loss": 0.0021, + "step": 18485 + }, + { + "epoch": 2.004119687771032, + "grad_norm": 0.2848951518535614, + "learning_rate": 6.663281590018861e-06, + "loss": 0.0027, + "step": 18486 + }, + { + "epoch": 2.004228100607112, + "grad_norm": 0.07108508050441742, + "learning_rate": 6.662556216451474e-06, + "loss": 0.0004, + "step": 18487 + }, + { + "epoch": 2.004336513443192, + "grad_norm": 1.0444304943084717, + "learning_rate": 6.661830842884086e-06, + "loss": 0.0161, + "step": 18488 + }, + { + "epoch": 2.0044449262792714, + "grad_norm": 0.1358906626701355, + "learning_rate": 6.661105469316699e-06, + "loss": 0.0008, + "step": 18489 + }, + { + "epoch": 2.0045533391153514, + "grad_norm": 2.7008399963378906, + "learning_rate": 6.660380095749312e-06, + "loss": 0.0437, + "step": 18490 + }, + { + "epoch": 2.004661751951431, + "grad_norm": 0.08824130892753601, + "learning_rate": 6.659654722181924e-06, + "loss": 0.0012, + "step": 18491 + }, + { + "epoch": 2.004770164787511, + "grad_norm": 0.6138978600502014, + "learning_rate": 6.658929348614537e-06, + "loss": 0.0049, + "step": 18492 + }, + { + "epoch": 2.0048785776235905, + "grad_norm": 0.1445169597864151, + "learning_rate": 6.65820397504715e-06, + "loss": 0.0019, + "step": 18493 + }, + { + "epoch": 2.0049869904596704, + "grad_norm": 1.2192506790161133, + "learning_rate": 6.657478601479762e-06, + "loss": 0.0063, + "step": 18494 + }, + { + "epoch": 2.0050954032957504, + "grad_norm": 1.3391400575637817, + "learning_rate": 6.656753227912376e-06, + "loss": 0.0264, + "step": 18495 + }, + { + "epoch": 2.00520381613183, + "grad_norm": 2.7429027557373047, + "learning_rate": 6.656027854344989e-06, + "loss": 0.0142, + "step": 18496 + }, + { + "epoch": 2.00531222896791, + "grad_norm": 0.5519901514053345, + "learning_rate": 6.655302480777601e-06, + "loss": 0.0018, + "step": 18497 + }, + { + "epoch": 2.0054206418039895, + "grad_norm": 0.22047942876815796, + "learning_rate": 6.654577107210214e-06, + "loss": 0.0025, + "step": 18498 + }, + { + "epoch": 2.0055290546400695, + "grad_norm": 0.576540470123291, + "learning_rate": 6.653851733642827e-06, + "loss": 0.0061, + "step": 18499 + }, + { + "epoch": 2.005637467476149, + "grad_norm": 1.3041043281555176, + "learning_rate": 6.653126360075439e-06, + "loss": 0.0084, + "step": 18500 + }, + { + "epoch": 2.005745880312229, + "grad_norm": 1.0522547960281372, + "learning_rate": 6.652400986508052e-06, + "loss": 0.0099, + "step": 18501 + }, + { + "epoch": 2.0058542931483085, + "grad_norm": 2.776217460632324, + "learning_rate": 6.651675612940664e-06, + "loss": 0.0086, + "step": 18502 + }, + { + "epoch": 2.0059627059843885, + "grad_norm": 0.27754634618759155, + "learning_rate": 6.650950239373278e-06, + "loss": 0.0016, + "step": 18503 + }, + { + "epoch": 2.0060711188204685, + "grad_norm": 0.189733624458313, + "learning_rate": 6.650224865805891e-06, + "loss": 0.0015, + "step": 18504 + }, + { + "epoch": 2.006179531656548, + "grad_norm": 0.42447423934936523, + "learning_rate": 6.649499492238504e-06, + "loss": 0.0031, + "step": 18505 + }, + { + "epoch": 2.006287944492628, + "grad_norm": 0.6270830035209656, + "learning_rate": 6.648774118671116e-06, + "loss": 0.0079, + "step": 18506 + }, + { + "epoch": 2.0063963573287076, + "grad_norm": 0.012530108913779259, + "learning_rate": 6.648048745103729e-06, + "loss": 0.0002, + "step": 18507 + }, + { + "epoch": 2.0065047701647876, + "grad_norm": 0.3430273234844208, + "learning_rate": 6.647323371536341e-06, + "loss": 0.0029, + "step": 18508 + }, + { + "epoch": 2.006613183000867, + "grad_norm": 0.7716532945632935, + "learning_rate": 6.646597997968954e-06, + "loss": 0.0067, + "step": 18509 + }, + { + "epoch": 2.006721595836947, + "grad_norm": 1.34398353099823, + "learning_rate": 6.645872624401567e-06, + "loss": 0.0367, + "step": 18510 + }, + { + "epoch": 2.006830008673027, + "grad_norm": 0.09824229031801224, + "learning_rate": 6.6451472508341794e-06, + "loss": 0.001, + "step": 18511 + }, + { + "epoch": 2.0069384215091066, + "grad_norm": 2.2529830932617188, + "learning_rate": 6.644421877266793e-06, + "loss": 0.0054, + "step": 18512 + }, + { + "epoch": 2.0070468343451866, + "grad_norm": 0.28189900517463684, + "learning_rate": 6.643696503699406e-06, + "loss": 0.0039, + "step": 18513 + }, + { + "epoch": 2.007155247181266, + "grad_norm": 5.4319281578063965, + "learning_rate": 6.642971130132018e-06, + "loss": 0.0082, + "step": 18514 + }, + { + "epoch": 2.007263660017346, + "grad_norm": 0.7626683115959167, + "learning_rate": 6.642245756564631e-06, + "loss": 0.0038, + "step": 18515 + }, + { + "epoch": 2.0073720728534257, + "grad_norm": 1.9000705480575562, + "learning_rate": 6.641520382997244e-06, + "loss": 0.0223, + "step": 18516 + }, + { + "epoch": 2.0074804856895057, + "grad_norm": 2.3863890171051025, + "learning_rate": 6.6407950094298565e-06, + "loss": 0.012, + "step": 18517 + }, + { + "epoch": 2.0075888985255856, + "grad_norm": 0.049320049583911896, + "learning_rate": 6.6400696358624695e-06, + "loss": 0.0004, + "step": 18518 + }, + { + "epoch": 2.007697311361665, + "grad_norm": 5.304440975189209, + "learning_rate": 6.6393442622950825e-06, + "loss": 0.0102, + "step": 18519 + }, + { + "epoch": 2.007805724197745, + "grad_norm": 0.34471818804740906, + "learning_rate": 6.638618888727695e-06, + "loss": 0.0018, + "step": 18520 + }, + { + "epoch": 2.0079141370338247, + "grad_norm": 4.275972366333008, + "learning_rate": 6.6378935151603085e-06, + "loss": 0.0049, + "step": 18521 + }, + { + "epoch": 2.0080225498699047, + "grad_norm": 0.06396615505218506, + "learning_rate": 6.6371681415929215e-06, + "loss": 0.0006, + "step": 18522 + }, + { + "epoch": 2.0081309627059842, + "grad_norm": 0.03186366707086563, + "learning_rate": 6.636442768025534e-06, + "loss": 0.0002, + "step": 18523 + }, + { + "epoch": 2.008239375542064, + "grad_norm": 0.4651433527469635, + "learning_rate": 6.635717394458147e-06, + "loss": 0.0025, + "step": 18524 + }, + { + "epoch": 2.0083477883781438, + "grad_norm": 1.7938899993896484, + "learning_rate": 6.63499202089076e-06, + "loss": 0.0099, + "step": 18525 + }, + { + "epoch": 2.0084562012142237, + "grad_norm": 0.15333862602710724, + "learning_rate": 6.634266647323372e-06, + "loss": 0.0014, + "step": 18526 + }, + { + "epoch": 2.0085646140503037, + "grad_norm": 2.388469696044922, + "learning_rate": 6.633541273755985e-06, + "loss": 0.0122, + "step": 18527 + }, + { + "epoch": 2.0086730268863833, + "grad_norm": 1.1378278732299805, + "learning_rate": 6.632815900188597e-06, + "loss": 0.0061, + "step": 18528 + }, + { + "epoch": 2.0087814397224633, + "grad_norm": 0.1855357438325882, + "learning_rate": 6.632090526621211e-06, + "loss": 0.0015, + "step": 18529 + }, + { + "epoch": 2.008889852558543, + "grad_norm": 0.9320123195648193, + "learning_rate": 6.631365153053824e-06, + "loss": 0.0023, + "step": 18530 + }, + { + "epoch": 2.008998265394623, + "grad_norm": 0.5097618699073792, + "learning_rate": 6.630639779486437e-06, + "loss": 0.0029, + "step": 18531 + }, + { + "epoch": 2.0091066782307023, + "grad_norm": 0.716097891330719, + "learning_rate": 6.629914405919049e-06, + "loss": 0.0035, + "step": 18532 + }, + { + "epoch": 2.0092150910667823, + "grad_norm": 2.1237943172454834, + "learning_rate": 6.629189032351662e-06, + "loss": 0.0152, + "step": 18533 + }, + { + "epoch": 2.0093235039028623, + "grad_norm": 0.3992577791213989, + "learning_rate": 6.628463658784274e-06, + "loss": 0.0025, + "step": 18534 + }, + { + "epoch": 2.009431916738942, + "grad_norm": 0.2725825905799866, + "learning_rate": 6.627738285216887e-06, + "loss": 0.0014, + "step": 18535 + }, + { + "epoch": 2.009540329575022, + "grad_norm": 0.230485200881958, + "learning_rate": 6.6270129116495e-06, + "loss": 0.0014, + "step": 18536 + }, + { + "epoch": 2.0096487424111014, + "grad_norm": 4.323365211486816, + "learning_rate": 6.626287538082112e-06, + "loss": 0.0285, + "step": 18537 + }, + { + "epoch": 2.0097571552471813, + "grad_norm": 0.6451985836029053, + "learning_rate": 6.625562164514726e-06, + "loss": 0.0037, + "step": 18538 + }, + { + "epoch": 2.009865568083261, + "grad_norm": 0.4646909236907959, + "learning_rate": 6.624836790947339e-06, + "loss": 0.0024, + "step": 18539 + }, + { + "epoch": 2.009973980919341, + "grad_norm": 0.14167048037052155, + "learning_rate": 6.624111417379951e-06, + "loss": 0.0009, + "step": 18540 + }, + { + "epoch": 2.010082393755421, + "grad_norm": 1.4869590997695923, + "learning_rate": 6.623386043812564e-06, + "loss": 0.0083, + "step": 18541 + }, + { + "epoch": 2.0101908065915004, + "grad_norm": 0.07291337847709656, + "learning_rate": 6.622660670245177e-06, + "loss": 0.0004, + "step": 18542 + }, + { + "epoch": 2.0102992194275804, + "grad_norm": 0.7143075466156006, + "learning_rate": 6.621935296677789e-06, + "loss": 0.0042, + "step": 18543 + }, + { + "epoch": 2.01040763226366, + "grad_norm": 0.10967700928449631, + "learning_rate": 6.621209923110402e-06, + "loss": 0.0007, + "step": 18544 + }, + { + "epoch": 2.01051604509974, + "grad_norm": 1.4673599004745483, + "learning_rate": 6.620484549543015e-06, + "loss": 0.0048, + "step": 18545 + }, + { + "epoch": 2.0106244579358195, + "grad_norm": 2.5878586769104004, + "learning_rate": 6.619759175975627e-06, + "loss": 0.0157, + "step": 18546 + }, + { + "epoch": 2.0107328707718994, + "grad_norm": 1.6239805221557617, + "learning_rate": 6.619033802408241e-06, + "loss": 0.0099, + "step": 18547 + }, + { + "epoch": 2.010841283607979, + "grad_norm": 0.8063681125640869, + "learning_rate": 6.618308428840854e-06, + "loss": 0.0068, + "step": 18548 + }, + { + "epoch": 2.010949696444059, + "grad_norm": 2.9885592460632324, + "learning_rate": 6.617583055273466e-06, + "loss": 0.0143, + "step": 18549 + }, + { + "epoch": 2.011058109280139, + "grad_norm": 1.646081805229187, + "learning_rate": 6.616857681706079e-06, + "loss": 0.0054, + "step": 18550 + }, + { + "epoch": 2.0111665221162185, + "grad_norm": 0.3517983853816986, + "learning_rate": 6.616132308138692e-06, + "loss": 0.0014, + "step": 18551 + }, + { + "epoch": 2.0112749349522985, + "grad_norm": 2.948936700820923, + "learning_rate": 6.615406934571304e-06, + "loss": 0.0183, + "step": 18552 + }, + { + "epoch": 2.011383347788378, + "grad_norm": 2.2620108127593994, + "learning_rate": 6.614681561003917e-06, + "loss": 0.0337, + "step": 18553 + }, + { + "epoch": 2.011491760624458, + "grad_norm": 1.6092047691345215, + "learning_rate": 6.6139561874365296e-06, + "loss": 0.009, + "step": 18554 + }, + { + "epoch": 2.0116001734605375, + "grad_norm": 3.300182580947876, + "learning_rate": 6.613230813869143e-06, + "loss": 0.061, + "step": 18555 + }, + { + "epoch": 2.0117085862966175, + "grad_norm": 0.7343549132347107, + "learning_rate": 6.612505440301756e-06, + "loss": 0.0017, + "step": 18556 + }, + { + "epoch": 2.0118169991326975, + "grad_norm": 0.8593565821647644, + "learning_rate": 6.611780066734369e-06, + "loss": 0.0032, + "step": 18557 + }, + { + "epoch": 2.011925411968777, + "grad_norm": 0.10898040235042572, + "learning_rate": 6.6110546931669815e-06, + "loss": 0.0008, + "step": 18558 + }, + { + "epoch": 2.012033824804857, + "grad_norm": 0.06468439102172852, + "learning_rate": 6.6103293195995945e-06, + "loss": 0.0003, + "step": 18559 + }, + { + "epoch": 2.0121422376409366, + "grad_norm": 5.299201965332031, + "learning_rate": 6.609603946032207e-06, + "loss": 0.0165, + "step": 18560 + }, + { + "epoch": 2.0122506504770166, + "grad_norm": 0.15259811282157898, + "learning_rate": 6.60887857246482e-06, + "loss": 0.0019, + "step": 18561 + }, + { + "epoch": 2.012359063313096, + "grad_norm": 0.06718608736991882, + "learning_rate": 6.608153198897433e-06, + "loss": 0.0002, + "step": 18562 + }, + { + "epoch": 2.012467476149176, + "grad_norm": 2.013509511947632, + "learning_rate": 6.607427825330045e-06, + "loss": 0.0176, + "step": 18563 + }, + { + "epoch": 2.0125758889852556, + "grad_norm": 0.31167709827423096, + "learning_rate": 6.606702451762659e-06, + "loss": 0.0016, + "step": 18564 + }, + { + "epoch": 2.0126843018213356, + "grad_norm": 0.11626353114843369, + "learning_rate": 6.605977078195272e-06, + "loss": 0.0005, + "step": 18565 + }, + { + "epoch": 2.0127927146574156, + "grad_norm": 0.02247200347483158, + "learning_rate": 6.605251704627884e-06, + "loss": 0.0002, + "step": 18566 + }, + { + "epoch": 2.012901127493495, + "grad_norm": 0.5721685886383057, + "learning_rate": 6.604526331060497e-06, + "loss": 0.0014, + "step": 18567 + }, + { + "epoch": 2.013009540329575, + "grad_norm": 3.334084987640381, + "learning_rate": 6.60380095749311e-06, + "loss": 0.0461, + "step": 18568 + }, + { + "epoch": 2.0131179531656547, + "grad_norm": 4.3890485763549805, + "learning_rate": 6.603075583925722e-06, + "loss": 0.0468, + "step": 18569 + }, + { + "epoch": 2.0132263660017347, + "grad_norm": 0.09314166009426117, + "learning_rate": 6.602350210358335e-06, + "loss": 0.0008, + "step": 18570 + }, + { + "epoch": 2.013334778837814, + "grad_norm": 0.17891807854175568, + "learning_rate": 6.601624836790948e-06, + "loss": 0.0003, + "step": 18571 + }, + { + "epoch": 2.013443191673894, + "grad_norm": 0.05134731903672218, + "learning_rate": 6.60089946322356e-06, + "loss": 0.0006, + "step": 18572 + }, + { + "epoch": 2.013551604509974, + "grad_norm": 0.053546320647001266, + "learning_rate": 6.600174089656174e-06, + "loss": 0.0003, + "step": 18573 + }, + { + "epoch": 2.0136600173460537, + "grad_norm": 0.34953656792640686, + "learning_rate": 6.599448716088787e-06, + "loss": 0.0014, + "step": 18574 + }, + { + "epoch": 2.0137684301821337, + "grad_norm": 0.5562894940376282, + "learning_rate": 6.598723342521399e-06, + "loss": 0.0029, + "step": 18575 + }, + { + "epoch": 2.0138768430182132, + "grad_norm": 0.47374582290649414, + "learning_rate": 6.597997968954012e-06, + "loss": 0.0021, + "step": 18576 + }, + { + "epoch": 2.013985255854293, + "grad_norm": 0.2276098132133484, + "learning_rate": 6.597272595386625e-06, + "loss": 0.0015, + "step": 18577 + }, + { + "epoch": 2.0140936686903728, + "grad_norm": 3.8372116088867188, + "learning_rate": 6.596547221819237e-06, + "loss": 0.0192, + "step": 18578 + }, + { + "epoch": 2.0142020815264527, + "grad_norm": 1.7504231929779053, + "learning_rate": 6.59582184825185e-06, + "loss": 0.0051, + "step": 18579 + }, + { + "epoch": 2.0143104943625327, + "grad_norm": 1.3691591024398804, + "learning_rate": 6.595096474684462e-06, + "loss": 0.0109, + "step": 18580 + }, + { + "epoch": 2.0144189071986123, + "grad_norm": 0.32927030324935913, + "learning_rate": 6.594371101117076e-06, + "loss": 0.0012, + "step": 18581 + }, + { + "epoch": 2.0145273200346923, + "grad_norm": 0.8484163284301758, + "learning_rate": 6.593645727549689e-06, + "loss": 0.0013, + "step": 18582 + }, + { + "epoch": 2.014635732870772, + "grad_norm": 0.38158613443374634, + "learning_rate": 6.592920353982302e-06, + "loss": 0.0016, + "step": 18583 + }, + { + "epoch": 2.014744145706852, + "grad_norm": 4.96052360534668, + "learning_rate": 6.592194980414914e-06, + "loss": 0.0018, + "step": 18584 + }, + { + "epoch": 2.0148525585429313, + "grad_norm": 0.28227499127388, + "learning_rate": 6.591469606847527e-06, + "loss": 0.0019, + "step": 18585 + }, + { + "epoch": 2.0149609713790113, + "grad_norm": 0.19707751274108887, + "learning_rate": 6.590744233280139e-06, + "loss": 0.0007, + "step": 18586 + }, + { + "epoch": 2.015069384215091, + "grad_norm": 0.15414045751094818, + "learning_rate": 6.590018859712752e-06, + "loss": 0.0013, + "step": 18587 + }, + { + "epoch": 2.015177797051171, + "grad_norm": 7.551421642303467, + "learning_rate": 6.589293486145365e-06, + "loss": 0.0137, + "step": 18588 + }, + { + "epoch": 2.015286209887251, + "grad_norm": 0.07417432218790054, + "learning_rate": 6.5885681125779774e-06, + "loss": 0.0004, + "step": 18589 + }, + { + "epoch": 2.0153946227233304, + "grad_norm": 0.23855935037136078, + "learning_rate": 6.587842739010591e-06, + "loss": 0.001, + "step": 18590 + }, + { + "epoch": 2.0155030355594103, + "grad_norm": 8.834012031555176, + "learning_rate": 6.587117365443204e-06, + "loss": 0.0173, + "step": 18591 + }, + { + "epoch": 2.01561144839549, + "grad_norm": 0.3338504433631897, + "learning_rate": 6.586391991875816e-06, + "loss": 0.0016, + "step": 18592 + }, + { + "epoch": 2.01571986123157, + "grad_norm": 1.3768389225006104, + "learning_rate": 6.585666618308429e-06, + "loss": 0.021, + "step": 18593 + }, + { + "epoch": 2.0158282740676494, + "grad_norm": 2.7026946544647217, + "learning_rate": 6.584941244741042e-06, + "loss": 0.0282, + "step": 18594 + }, + { + "epoch": 2.0159366869037294, + "grad_norm": 0.019482716917991638, + "learning_rate": 6.5842158711736545e-06, + "loss": 0.0001, + "step": 18595 + }, + { + "epoch": 2.0160450997398094, + "grad_norm": 0.1805102378129959, + "learning_rate": 6.5834904976062675e-06, + "loss": 0.0011, + "step": 18596 + }, + { + "epoch": 2.016153512575889, + "grad_norm": 0.1543784886598587, + "learning_rate": 6.5827651240388805e-06, + "loss": 0.0009, + "step": 18597 + }, + { + "epoch": 2.016261925411969, + "grad_norm": 0.3443467319011688, + "learning_rate": 6.5820397504714935e-06, + "loss": 0.001, + "step": 18598 + }, + { + "epoch": 2.0163703382480485, + "grad_norm": 0.015645582228899002, + "learning_rate": 6.5813143769041065e-06, + "loss": 0.0001, + "step": 18599 + }, + { + "epoch": 2.0164787510841284, + "grad_norm": 19.343303680419922, + "learning_rate": 6.5805890033367195e-06, + "loss": 0.0272, + "step": 18600 + }, + { + "epoch": 2.016587163920208, + "grad_norm": 0.052216701209545135, + "learning_rate": 6.579863629769332e-06, + "loss": 0.0004, + "step": 18601 + }, + { + "epoch": 2.016695576756288, + "grad_norm": 2.5459952354431152, + "learning_rate": 6.579138256201945e-06, + "loss": 0.0133, + "step": 18602 + }, + { + "epoch": 2.016803989592368, + "grad_norm": 1.02968430519104, + "learning_rate": 6.578412882634558e-06, + "loss": 0.0104, + "step": 18603 + }, + { + "epoch": 2.0169124024284475, + "grad_norm": 0.6363314390182495, + "learning_rate": 6.57768750906717e-06, + "loss": 0.0027, + "step": 18604 + }, + { + "epoch": 2.0170208152645275, + "grad_norm": 1.1022241115570068, + "learning_rate": 6.576962135499783e-06, + "loss": 0.0061, + "step": 18605 + }, + { + "epoch": 2.017129228100607, + "grad_norm": 0.6632934212684631, + "learning_rate": 6.576236761932395e-06, + "loss": 0.0028, + "step": 18606 + }, + { + "epoch": 2.017237640936687, + "grad_norm": 9.51482105255127, + "learning_rate": 6.575511388365009e-06, + "loss": 0.0208, + "step": 18607 + }, + { + "epoch": 2.0173460537727665, + "grad_norm": 0.05343969166278839, + "learning_rate": 6.574786014797622e-06, + "loss": 0.0003, + "step": 18608 + }, + { + "epoch": 2.0174544666088465, + "grad_norm": 0.7265685796737671, + "learning_rate": 6.574060641230235e-06, + "loss": 0.0008, + "step": 18609 + }, + { + "epoch": 2.017562879444926, + "grad_norm": 0.06576091051101685, + "learning_rate": 6.573335267662847e-06, + "loss": 0.0005, + "step": 18610 + }, + { + "epoch": 2.017671292281006, + "grad_norm": 1.3779760599136353, + "learning_rate": 6.57260989409546e-06, + "loss": 0.0024, + "step": 18611 + }, + { + "epoch": 2.017779705117086, + "grad_norm": 0.3408734202384949, + "learning_rate": 6.571884520528072e-06, + "loss": 0.0022, + "step": 18612 + }, + { + "epoch": 2.0178881179531656, + "grad_norm": 1.231330156326294, + "learning_rate": 6.571159146960685e-06, + "loss": 0.0038, + "step": 18613 + }, + { + "epoch": 2.0179965307892456, + "grad_norm": 0.8243698477745056, + "learning_rate": 6.570433773393298e-06, + "loss": 0.0014, + "step": 18614 + }, + { + "epoch": 2.018104943625325, + "grad_norm": 0.5691836476325989, + "learning_rate": 6.56970839982591e-06, + "loss": 0.0029, + "step": 18615 + }, + { + "epoch": 2.018213356461405, + "grad_norm": 0.8725003004074097, + "learning_rate": 6.568983026258524e-06, + "loss": 0.0052, + "step": 18616 + }, + { + "epoch": 2.0183217692974846, + "grad_norm": 2.1459972858428955, + "learning_rate": 6.568257652691137e-06, + "loss": 0.0362, + "step": 18617 + }, + { + "epoch": 2.0184301821335646, + "grad_norm": 1.6166421175003052, + "learning_rate": 6.567532279123749e-06, + "loss": 0.0192, + "step": 18618 + }, + { + "epoch": 2.0185385949696446, + "grad_norm": 0.014138883911073208, + "learning_rate": 6.566806905556362e-06, + "loss": 0.0001, + "step": 18619 + }, + { + "epoch": 2.018647007805724, + "grad_norm": 1.7648435831069946, + "learning_rate": 6.566081531988975e-06, + "loss": 0.0085, + "step": 18620 + }, + { + "epoch": 2.018755420641804, + "grad_norm": 0.8531116843223572, + "learning_rate": 6.565356158421587e-06, + "loss": 0.0028, + "step": 18621 + }, + { + "epoch": 2.0188638334778837, + "grad_norm": 0.29473182559013367, + "learning_rate": 6.5646307848542e-06, + "loss": 0.0012, + "step": 18622 + }, + { + "epoch": 2.0189722463139637, + "grad_norm": 10.827079772949219, + "learning_rate": 6.563905411286813e-06, + "loss": 0.0103, + "step": 18623 + }, + { + "epoch": 2.019080659150043, + "grad_norm": 0.0320918969810009, + "learning_rate": 6.563180037719426e-06, + "loss": 0.0002, + "step": 18624 + }, + { + "epoch": 2.019189071986123, + "grad_norm": 0.11286934465169907, + "learning_rate": 6.562454664152039e-06, + "loss": 0.0007, + "step": 18625 + }, + { + "epoch": 2.0192974848222027, + "grad_norm": 0.06383765488862991, + "learning_rate": 6.561729290584652e-06, + "loss": 0.0005, + "step": 18626 + }, + { + "epoch": 2.0194058976582827, + "grad_norm": 10.271451950073242, + "learning_rate": 6.561003917017264e-06, + "loss": 0.0392, + "step": 18627 + }, + { + "epoch": 2.0195143104943627, + "grad_norm": 4.089141845703125, + "learning_rate": 6.560278543449877e-06, + "loss": 0.0335, + "step": 18628 + }, + { + "epoch": 2.0196227233304422, + "grad_norm": 0.01772397570312023, + "learning_rate": 6.55955316988249e-06, + "loss": 0.0002, + "step": 18629 + }, + { + "epoch": 2.0197311361665222, + "grad_norm": 1.5620975494384766, + "learning_rate": 6.5588277963151024e-06, + "loss": 0.007, + "step": 18630 + }, + { + "epoch": 2.0198395490026018, + "grad_norm": 0.8303648829460144, + "learning_rate": 6.558102422747715e-06, + "loss": 0.0017, + "step": 18631 + }, + { + "epoch": 2.0199479618386817, + "grad_norm": 0.057734549045562744, + "learning_rate": 6.5573770491803276e-06, + "loss": 0.0005, + "step": 18632 + }, + { + "epoch": 2.0200563746747613, + "grad_norm": 2.440584659576416, + "learning_rate": 6.556651675612941e-06, + "loss": 0.0081, + "step": 18633 + }, + { + "epoch": 2.0201647875108413, + "grad_norm": 0.2443157136440277, + "learning_rate": 6.555926302045554e-06, + "loss": 0.0005, + "step": 18634 + }, + { + "epoch": 2.0202732003469213, + "grad_norm": 0.20115087926387787, + "learning_rate": 6.555200928478167e-06, + "loss": 0.0016, + "step": 18635 + }, + { + "epoch": 2.020381613183001, + "grad_norm": 0.7796831130981445, + "learning_rate": 6.5544755549107795e-06, + "loss": 0.0006, + "step": 18636 + }, + { + "epoch": 2.020490026019081, + "grad_norm": 2.3285861015319824, + "learning_rate": 6.5537501813433925e-06, + "loss": 0.0486, + "step": 18637 + }, + { + "epoch": 2.0205984388551603, + "grad_norm": 0.07955816388130188, + "learning_rate": 6.553024807776005e-06, + "loss": 0.0005, + "step": 18638 + }, + { + "epoch": 2.0207068516912403, + "grad_norm": 9.124492645263672, + "learning_rate": 6.552299434208618e-06, + "loss": 0.0234, + "step": 18639 + }, + { + "epoch": 2.02081526452732, + "grad_norm": 3.0609207153320312, + "learning_rate": 6.551574060641231e-06, + "loss": 0.0139, + "step": 18640 + }, + { + "epoch": 2.0209236773634, + "grad_norm": 0.022300204262137413, + "learning_rate": 6.550848687073843e-06, + "loss": 0.0002, + "step": 18641 + }, + { + "epoch": 2.02103209019948, + "grad_norm": 0.11332827061414719, + "learning_rate": 6.550123313506457e-06, + "loss": 0.0007, + "step": 18642 + }, + { + "epoch": 2.0211405030355594, + "grad_norm": 0.2887389659881592, + "learning_rate": 6.54939793993907e-06, + "loss": 0.0019, + "step": 18643 + }, + { + "epoch": 2.0212489158716394, + "grad_norm": 0.043468717485666275, + "learning_rate": 6.548672566371682e-06, + "loss": 0.0003, + "step": 18644 + }, + { + "epoch": 2.021357328707719, + "grad_norm": 0.2935291826725006, + "learning_rate": 6.547947192804295e-06, + "loss": 0.0016, + "step": 18645 + }, + { + "epoch": 2.021465741543799, + "grad_norm": 2.9551563262939453, + "learning_rate": 6.547221819236908e-06, + "loss": 0.0289, + "step": 18646 + }, + { + "epoch": 2.0215741543798784, + "grad_norm": 5.876340866088867, + "learning_rate": 6.54649644566952e-06, + "loss": 0.0142, + "step": 18647 + }, + { + "epoch": 2.0216825672159584, + "grad_norm": 0.06886307150125504, + "learning_rate": 6.545771072102133e-06, + "loss": 0.0003, + "step": 18648 + }, + { + "epoch": 2.021790980052038, + "grad_norm": 0.5638512969017029, + "learning_rate": 6.545045698534746e-06, + "loss": 0.0022, + "step": 18649 + }, + { + "epoch": 2.021899392888118, + "grad_norm": 0.4590442180633545, + "learning_rate": 6.544320324967359e-06, + "loss": 0.0031, + "step": 18650 + }, + { + "epoch": 2.022007805724198, + "grad_norm": 4.017760753631592, + "learning_rate": 6.543594951399972e-06, + "loss": 0.0447, + "step": 18651 + }, + { + "epoch": 2.0221162185602775, + "grad_norm": 0.21996378898620605, + "learning_rate": 6.542869577832585e-06, + "loss": 0.0003, + "step": 18652 + }, + { + "epoch": 2.0222246313963574, + "grad_norm": 6.374166011810303, + "learning_rate": 6.542144204265197e-06, + "loss": 0.0228, + "step": 18653 + }, + { + "epoch": 2.022333044232437, + "grad_norm": 0.03293846920132637, + "learning_rate": 6.54141883069781e-06, + "loss": 0.0003, + "step": 18654 + }, + { + "epoch": 2.022441457068517, + "grad_norm": 1.6746481657028198, + "learning_rate": 6.540693457130423e-06, + "loss": 0.0079, + "step": 18655 + }, + { + "epoch": 2.0225498699045965, + "grad_norm": 11.565004348754883, + "learning_rate": 6.539968083563035e-06, + "loss": 0.0261, + "step": 18656 + }, + { + "epoch": 2.0226582827406765, + "grad_norm": 12.489969253540039, + "learning_rate": 6.539242709995648e-06, + "loss": 0.0244, + "step": 18657 + }, + { + "epoch": 2.0227666955767565, + "grad_norm": 3.0146377086639404, + "learning_rate": 6.53851733642826e-06, + "loss": 0.0112, + "step": 18658 + }, + { + "epoch": 2.022875108412836, + "grad_norm": 0.07089720666408539, + "learning_rate": 6.537791962860874e-06, + "loss": 0.0003, + "step": 18659 + }, + { + "epoch": 2.022983521248916, + "grad_norm": 0.11066249758005142, + "learning_rate": 6.537066589293487e-06, + "loss": 0.0005, + "step": 18660 + }, + { + "epoch": 2.0230919340849955, + "grad_norm": 1.702069878578186, + "learning_rate": 6.5363412157261e-06, + "loss": 0.0083, + "step": 18661 + }, + { + "epoch": 2.0232003469210755, + "grad_norm": 0.013187970034778118, + "learning_rate": 6.535615842158712e-06, + "loss": 0.0001, + "step": 18662 + }, + { + "epoch": 2.023308759757155, + "grad_norm": 0.520649790763855, + "learning_rate": 6.534890468591325e-06, + "loss": 0.0033, + "step": 18663 + }, + { + "epoch": 2.023417172593235, + "grad_norm": 2.207847833633423, + "learning_rate": 6.534165095023937e-06, + "loss": 0.0028, + "step": 18664 + }, + { + "epoch": 2.023525585429315, + "grad_norm": 1.4343968629837036, + "learning_rate": 6.53343972145655e-06, + "loss": 0.0026, + "step": 18665 + }, + { + "epoch": 2.0236339982653946, + "grad_norm": 1.136368989944458, + "learning_rate": 6.532714347889163e-06, + "loss": 0.0059, + "step": 18666 + }, + { + "epoch": 2.0237424111014746, + "grad_norm": 10.801992416381836, + "learning_rate": 6.5319889743217754e-06, + "loss": 0.0456, + "step": 18667 + }, + { + "epoch": 2.023850823937554, + "grad_norm": 0.5666526556015015, + "learning_rate": 6.531263600754389e-06, + "loss": 0.003, + "step": 18668 + }, + { + "epoch": 2.023959236773634, + "grad_norm": 0.6116659045219421, + "learning_rate": 6.530538227187002e-06, + "loss": 0.0024, + "step": 18669 + }, + { + "epoch": 2.0240676496097136, + "grad_norm": 1.9202709197998047, + "learning_rate": 6.529812853619614e-06, + "loss": 0.0033, + "step": 18670 + }, + { + "epoch": 2.0241760624457936, + "grad_norm": 1.1997109651565552, + "learning_rate": 6.529087480052227e-06, + "loss": 0.0044, + "step": 18671 + }, + { + "epoch": 2.024284475281873, + "grad_norm": 0.02007790096104145, + "learning_rate": 6.52836210648484e-06, + "loss": 0.0002, + "step": 18672 + }, + { + "epoch": 2.024392888117953, + "grad_norm": 6.198502540588379, + "learning_rate": 6.5276367329174525e-06, + "loss": 0.0249, + "step": 18673 + }, + { + "epoch": 2.024501300954033, + "grad_norm": 3.36203932762146, + "learning_rate": 6.5269113593500655e-06, + "loss": 0.0109, + "step": 18674 + }, + { + "epoch": 2.0246097137901127, + "grad_norm": 2.2310526371002197, + "learning_rate": 6.5261859857826785e-06, + "loss": 0.0214, + "step": 18675 + }, + { + "epoch": 2.0247181266261927, + "grad_norm": 0.15276820957660675, + "learning_rate": 6.5254606122152915e-06, + "loss": 0.0004, + "step": 18676 + }, + { + "epoch": 2.024826539462272, + "grad_norm": 1.8818448781967163, + "learning_rate": 6.5247352386479045e-06, + "loss": 0.0257, + "step": 18677 + }, + { + "epoch": 2.024934952298352, + "grad_norm": 6.965218544006348, + "learning_rate": 6.5240098650805175e-06, + "loss": 0.0199, + "step": 18678 + }, + { + "epoch": 2.0250433651344317, + "grad_norm": 0.02351764775812626, + "learning_rate": 6.52328449151313e-06, + "loss": 0.0001, + "step": 18679 + }, + { + "epoch": 2.0251517779705117, + "grad_norm": 0.08679568022489548, + "learning_rate": 6.522559117945743e-06, + "loss": 0.0004, + "step": 18680 + }, + { + "epoch": 2.0252601908065917, + "grad_norm": 0.022488027811050415, + "learning_rate": 6.521833744378356e-06, + "loss": 0.0002, + "step": 18681 + }, + { + "epoch": 2.0253686036426712, + "grad_norm": 1.0966233015060425, + "learning_rate": 6.521108370810968e-06, + "loss": 0.0014, + "step": 18682 + }, + { + "epoch": 2.0254770164787512, + "grad_norm": 1.324588656425476, + "learning_rate": 6.520382997243581e-06, + "loss": 0.0041, + "step": 18683 + }, + { + "epoch": 2.0255854293148308, + "grad_norm": 0.3334045708179474, + "learning_rate": 6.519657623676193e-06, + "loss": 0.0005, + "step": 18684 + }, + { + "epoch": 2.0256938421509108, + "grad_norm": 15.879961967468262, + "learning_rate": 6.518932250108807e-06, + "loss": 0.0708, + "step": 18685 + }, + { + "epoch": 2.0258022549869903, + "grad_norm": 3.40887451171875, + "learning_rate": 6.51820687654142e-06, + "loss": 0.0158, + "step": 18686 + }, + { + "epoch": 2.0259106678230703, + "grad_norm": 1.8491315841674805, + "learning_rate": 6.517481502974033e-06, + "loss": 0.004, + "step": 18687 + }, + { + "epoch": 2.0260190806591503, + "grad_norm": 4.732418060302734, + "learning_rate": 6.516756129406645e-06, + "loss": 0.0324, + "step": 18688 + }, + { + "epoch": 2.02612749349523, + "grad_norm": 0.9091562032699585, + "learning_rate": 6.516030755839258e-06, + "loss": 0.0033, + "step": 18689 + }, + { + "epoch": 2.02623590633131, + "grad_norm": 0.06739146262407303, + "learning_rate": 6.51530538227187e-06, + "loss": 0.0004, + "step": 18690 + }, + { + "epoch": 2.0263443191673893, + "grad_norm": 1.2700550556182861, + "learning_rate": 6.514580008704483e-06, + "loss": 0.0021, + "step": 18691 + }, + { + "epoch": 2.0264527320034693, + "grad_norm": 1.012264609336853, + "learning_rate": 6.513854635137096e-06, + "loss": 0.0027, + "step": 18692 + }, + { + "epoch": 2.026561144839549, + "grad_norm": 3.5203371047973633, + "learning_rate": 6.51312926156971e-06, + "loss": 0.021, + "step": 18693 + }, + { + "epoch": 2.026669557675629, + "grad_norm": 1.2604912519454956, + "learning_rate": 6.512403888002322e-06, + "loss": 0.0061, + "step": 18694 + }, + { + "epoch": 2.0267779705117084, + "grad_norm": 0.2313988208770752, + "learning_rate": 6.511678514434935e-06, + "loss": 0.0004, + "step": 18695 + }, + { + "epoch": 2.0268863833477884, + "grad_norm": 0.27465102076530457, + "learning_rate": 6.510953140867547e-06, + "loss": 0.0007, + "step": 18696 + }, + { + "epoch": 2.0269947961838684, + "grad_norm": 0.09387357532978058, + "learning_rate": 6.51022776730016e-06, + "loss": 0.0003, + "step": 18697 + }, + { + "epoch": 2.027103209019948, + "grad_norm": 0.08043474704027176, + "learning_rate": 6.509502393732773e-06, + "loss": 0.0004, + "step": 18698 + }, + { + "epoch": 2.027211621856028, + "grad_norm": 0.609093189239502, + "learning_rate": 6.508777020165385e-06, + "loss": 0.0012, + "step": 18699 + }, + { + "epoch": 2.0273200346921074, + "grad_norm": 0.13295882940292358, + "learning_rate": 6.508051646597998e-06, + "loss": 0.0008, + "step": 18700 + }, + { + "epoch": 2.0274284475281874, + "grad_norm": 0.5846439003944397, + "learning_rate": 6.507326273030611e-06, + "loss": 0.0009, + "step": 18701 + }, + { + "epoch": 2.027536860364267, + "grad_norm": 0.13627927005290985, + "learning_rate": 6.506600899463224e-06, + "loss": 0.0007, + "step": 18702 + }, + { + "epoch": 2.027645273200347, + "grad_norm": 0.39176145195961, + "learning_rate": 6.505875525895837e-06, + "loss": 0.0012, + "step": 18703 + }, + { + "epoch": 2.027753686036427, + "grad_norm": 5.401750564575195, + "learning_rate": 6.50515015232845e-06, + "loss": 0.0066, + "step": 18704 + }, + { + "epoch": 2.0278620988725065, + "grad_norm": 1.2856343984603882, + "learning_rate": 6.504424778761062e-06, + "loss": 0.0029, + "step": 18705 + }, + { + "epoch": 2.0279705117085864, + "grad_norm": 0.01179332286119461, + "learning_rate": 6.503699405193675e-06, + "loss": 0.0001, + "step": 18706 + }, + { + "epoch": 2.028078924544666, + "grad_norm": 2.40677809715271, + "learning_rate": 6.502974031626288e-06, + "loss": 0.0192, + "step": 18707 + }, + { + "epoch": 2.028187337380746, + "grad_norm": 7.637319564819336, + "learning_rate": 6.5022486580589004e-06, + "loss": 0.0061, + "step": 18708 + }, + { + "epoch": 2.0282957502168255, + "grad_norm": 12.229296684265137, + "learning_rate": 6.501523284491513e-06, + "loss": 0.043, + "step": 18709 + }, + { + "epoch": 2.0284041630529055, + "grad_norm": 0.05225649103522301, + "learning_rate": 6.5007979109241256e-06, + "loss": 0.0003, + "step": 18710 + }, + { + "epoch": 2.028512575888985, + "grad_norm": 0.08187175542116165, + "learning_rate": 6.500072537356739e-06, + "loss": 0.0004, + "step": 18711 + }, + { + "epoch": 2.028620988725065, + "grad_norm": 2.4133150577545166, + "learning_rate": 6.499347163789352e-06, + "loss": 0.0445, + "step": 18712 + }, + { + "epoch": 2.028729401561145, + "grad_norm": 0.0031242554541677237, + "learning_rate": 6.498621790221965e-06, + "loss": 0.0, + "step": 18713 + }, + { + "epoch": 2.0288378143972245, + "grad_norm": 0.331929087638855, + "learning_rate": 6.4978964166545775e-06, + "loss": 0.0009, + "step": 18714 + }, + { + "epoch": 2.0289462272333045, + "grad_norm": 3.7811784744262695, + "learning_rate": 6.4971710430871905e-06, + "loss": 0.0231, + "step": 18715 + }, + { + "epoch": 2.029054640069384, + "grad_norm": 2.989675283432007, + "learning_rate": 6.496445669519803e-06, + "loss": 0.0118, + "step": 18716 + }, + { + "epoch": 2.029163052905464, + "grad_norm": 11.16855525970459, + "learning_rate": 6.495720295952416e-06, + "loss": 0.0251, + "step": 18717 + }, + { + "epoch": 2.0292714657415436, + "grad_norm": 0.20982353389263153, + "learning_rate": 6.494994922385029e-06, + "loss": 0.0004, + "step": 18718 + }, + { + "epoch": 2.0293798785776236, + "grad_norm": 0.03403811901807785, + "learning_rate": 6.4942695488176425e-06, + "loss": 0.0003, + "step": 18719 + }, + { + "epoch": 2.0294882914137036, + "grad_norm": 1.2338175773620605, + "learning_rate": 6.493544175250255e-06, + "loss": 0.0021, + "step": 18720 + }, + { + "epoch": 2.029596704249783, + "grad_norm": 0.12290249764919281, + "learning_rate": 6.492818801682868e-06, + "loss": 0.0002, + "step": 18721 + }, + { + "epoch": 2.029705117085863, + "grad_norm": 2.3440916538238525, + "learning_rate": 6.49209342811548e-06, + "loss": 0.0363, + "step": 18722 + }, + { + "epoch": 2.0298135299219426, + "grad_norm": 27.133934020996094, + "learning_rate": 6.491368054548093e-06, + "loss": 0.0259, + "step": 18723 + }, + { + "epoch": 2.0299219427580226, + "grad_norm": 5.041303634643555, + "learning_rate": 6.490642680980706e-06, + "loss": 0.0091, + "step": 18724 + }, + { + "epoch": 2.030030355594102, + "grad_norm": 1.368054747581482, + "learning_rate": 6.489917307413318e-06, + "loss": 0.0087, + "step": 18725 + }, + { + "epoch": 2.030138768430182, + "grad_norm": 3.6198418140411377, + "learning_rate": 6.489191933845931e-06, + "loss": 0.0126, + "step": 18726 + }, + { + "epoch": 2.030247181266262, + "grad_norm": 0.14133672416210175, + "learning_rate": 6.488466560278544e-06, + "loss": 0.0005, + "step": 18727 + }, + { + "epoch": 2.0303555941023417, + "grad_norm": 0.016111021861433983, + "learning_rate": 6.487741186711157e-06, + "loss": 0.0001, + "step": 18728 + }, + { + "epoch": 2.0304640069384217, + "grad_norm": 1.8058829307556152, + "learning_rate": 6.48701581314377e-06, + "loss": 0.0023, + "step": 18729 + }, + { + "epoch": 2.030572419774501, + "grad_norm": 2.8087856769561768, + "learning_rate": 6.486290439576383e-06, + "loss": 0.0107, + "step": 18730 + }, + { + "epoch": 2.030680832610581, + "grad_norm": 2.151254177093506, + "learning_rate": 6.485565066008995e-06, + "loss": 0.003, + "step": 18731 + }, + { + "epoch": 2.0307892454466607, + "grad_norm": 4.613700866699219, + "learning_rate": 6.484839692441608e-06, + "loss": 0.0088, + "step": 18732 + }, + { + "epoch": 2.0308976582827407, + "grad_norm": 1.821536660194397, + "learning_rate": 6.484114318874221e-06, + "loss": 0.0171, + "step": 18733 + }, + { + "epoch": 2.0310060711188203, + "grad_norm": 3.9527487754821777, + "learning_rate": 6.483388945306833e-06, + "loss": 0.0687, + "step": 18734 + }, + { + "epoch": 2.0311144839549002, + "grad_norm": 4.058011531829834, + "learning_rate": 6.482663571739446e-06, + "loss": 0.0038, + "step": 18735 + }, + { + "epoch": 2.0312228967909802, + "grad_norm": 1.4950122833251953, + "learning_rate": 6.481938198172058e-06, + "loss": 0.0049, + "step": 18736 + }, + { + "epoch": 2.0313313096270598, + "grad_norm": 0.3638964593410492, + "learning_rate": 6.481212824604672e-06, + "loss": 0.0011, + "step": 18737 + }, + { + "epoch": 2.0314397224631398, + "grad_norm": 0.05492753908038139, + "learning_rate": 6.480487451037285e-06, + "loss": 0.0004, + "step": 18738 + }, + { + "epoch": 2.0315481352992193, + "grad_norm": 0.1727476567029953, + "learning_rate": 6.479762077469898e-06, + "loss": 0.0004, + "step": 18739 + }, + { + "epoch": 2.0316565481352993, + "grad_norm": 3.8485140800476074, + "learning_rate": 6.47903670390251e-06, + "loss": 0.0135, + "step": 18740 + }, + { + "epoch": 2.031764960971379, + "grad_norm": 0.22330600023269653, + "learning_rate": 6.478311330335123e-06, + "loss": 0.0009, + "step": 18741 + }, + { + "epoch": 2.031873373807459, + "grad_norm": 2.547804594039917, + "learning_rate": 6.477585956767735e-06, + "loss": 0.019, + "step": 18742 + }, + { + "epoch": 2.031981786643539, + "grad_norm": 2.389547348022461, + "learning_rate": 6.476860583200348e-06, + "loss": 0.0145, + "step": 18743 + }, + { + "epoch": 2.0320901994796183, + "grad_norm": 0.01379484124481678, + "learning_rate": 6.476135209632961e-06, + "loss": 0.0001, + "step": 18744 + }, + { + "epoch": 2.0321986123156983, + "grad_norm": 1.554388403892517, + "learning_rate": 6.475409836065575e-06, + "loss": 0.0498, + "step": 18745 + }, + { + "epoch": 2.032307025151778, + "grad_norm": 6.405264377593994, + "learning_rate": 6.474684462498187e-06, + "loss": 0.0161, + "step": 18746 + }, + { + "epoch": 2.032415437987858, + "grad_norm": 0.016162950545549393, + "learning_rate": 6.4739590889308e-06, + "loss": 0.0002, + "step": 18747 + }, + { + "epoch": 2.0325238508239374, + "grad_norm": 3.549459934234619, + "learning_rate": 6.4732337153634124e-06, + "loss": 0.0062, + "step": 18748 + }, + { + "epoch": 2.0326322636600174, + "grad_norm": 0.9980521202087402, + "learning_rate": 6.472508341796025e-06, + "loss": 0.0065, + "step": 18749 + }, + { + "epoch": 2.0327406764960974, + "grad_norm": 0.6452032923698425, + "learning_rate": 6.471782968228638e-06, + "loss": 0.0011, + "step": 18750 + }, + { + "epoch": 2.032849089332177, + "grad_norm": 0.03409669175744057, + "learning_rate": 6.4710575946612505e-06, + "loss": 0.0001, + "step": 18751 + }, + { + "epoch": 2.032957502168257, + "grad_norm": 0.08803828805685043, + "learning_rate": 6.4703322210938635e-06, + "loss": 0.0006, + "step": 18752 + }, + { + "epoch": 2.0330659150043364, + "grad_norm": 3.6170380115509033, + "learning_rate": 6.4696068475264765e-06, + "loss": 0.0043, + "step": 18753 + }, + { + "epoch": 2.0331743278404164, + "grad_norm": 0.2692127823829651, + "learning_rate": 6.4688814739590895e-06, + "loss": 0.0012, + "step": 18754 + }, + { + "epoch": 2.033282740676496, + "grad_norm": 2.6923129558563232, + "learning_rate": 6.4681561003917025e-06, + "loss": 0.0077, + "step": 18755 + }, + { + "epoch": 2.033391153512576, + "grad_norm": 2.7551820278167725, + "learning_rate": 6.4674307268243155e-06, + "loss": 0.018, + "step": 18756 + }, + { + "epoch": 2.0334995663486555, + "grad_norm": 0.13297751545906067, + "learning_rate": 6.466705353256928e-06, + "loss": 0.0002, + "step": 18757 + }, + { + "epoch": 2.0336079791847355, + "grad_norm": 0.2015664428472519, + "learning_rate": 6.465979979689541e-06, + "loss": 0.0005, + "step": 18758 + }, + { + "epoch": 2.0337163920208154, + "grad_norm": 4.082460403442383, + "learning_rate": 6.465254606122154e-06, + "loss": 0.0324, + "step": 18759 + }, + { + "epoch": 2.033824804856895, + "grad_norm": 0.10436727851629257, + "learning_rate": 6.464529232554766e-06, + "loss": 0.0006, + "step": 18760 + }, + { + "epoch": 2.033933217692975, + "grad_norm": 3.2576980590820312, + "learning_rate": 6.463803858987379e-06, + "loss": 0.0028, + "step": 18761 + }, + { + "epoch": 2.0340416305290545, + "grad_norm": 5.173908233642578, + "learning_rate": 6.463078485419991e-06, + "loss": 0.0018, + "step": 18762 + }, + { + "epoch": 2.0341500433651345, + "grad_norm": 6.986522197723389, + "learning_rate": 6.462353111852605e-06, + "loss": 0.045, + "step": 18763 + }, + { + "epoch": 2.034258456201214, + "grad_norm": 0.16821689903736115, + "learning_rate": 6.461627738285218e-06, + "loss": 0.0008, + "step": 18764 + }, + { + "epoch": 2.034366869037294, + "grad_norm": 0.12191004306077957, + "learning_rate": 6.460902364717831e-06, + "loss": 0.0006, + "step": 18765 + }, + { + "epoch": 2.034475281873374, + "grad_norm": 0.21150444447994232, + "learning_rate": 6.460176991150443e-06, + "loss": 0.0009, + "step": 18766 + }, + { + "epoch": 2.0345836947094535, + "grad_norm": 0.5716129541397095, + "learning_rate": 6.459451617583056e-06, + "loss": 0.0025, + "step": 18767 + }, + { + "epoch": 2.0346921075455335, + "grad_norm": 6.102184295654297, + "learning_rate": 6.458726244015668e-06, + "loss": 0.0247, + "step": 18768 + }, + { + "epoch": 2.034800520381613, + "grad_norm": 13.637646675109863, + "learning_rate": 6.458000870448281e-06, + "loss": 0.016, + "step": 18769 + }, + { + "epoch": 2.034908933217693, + "grad_norm": 0.06823049485683441, + "learning_rate": 6.457275496880894e-06, + "loss": 0.0002, + "step": 18770 + }, + { + "epoch": 2.0350173460537726, + "grad_norm": 0.09927927702665329, + "learning_rate": 6.456550123313508e-06, + "loss": 0.0006, + "step": 18771 + }, + { + "epoch": 2.0351257588898526, + "grad_norm": 1.5925853252410889, + "learning_rate": 6.45582474974612e-06, + "loss": 0.0402, + "step": 18772 + }, + { + "epoch": 2.035234171725932, + "grad_norm": 0.6196544766426086, + "learning_rate": 6.455099376178733e-06, + "loss": 0.0014, + "step": 18773 + }, + { + "epoch": 2.035342584562012, + "grad_norm": 1.4917303323745728, + "learning_rate": 6.454374002611345e-06, + "loss": 0.0078, + "step": 18774 + }, + { + "epoch": 2.035450997398092, + "grad_norm": 0.042686279863119125, + "learning_rate": 6.453648629043958e-06, + "loss": 0.0002, + "step": 18775 + }, + { + "epoch": 2.0355594102341716, + "grad_norm": 0.3968972861766815, + "learning_rate": 6.452923255476571e-06, + "loss": 0.0018, + "step": 18776 + }, + { + "epoch": 2.0356678230702516, + "grad_norm": 2.905211925506592, + "learning_rate": 6.452197881909183e-06, + "loss": 0.0216, + "step": 18777 + }, + { + "epoch": 2.035776235906331, + "grad_norm": 1.7876393795013428, + "learning_rate": 6.451472508341796e-06, + "loss": 0.0034, + "step": 18778 + }, + { + "epoch": 2.035884648742411, + "grad_norm": 3.1471805572509766, + "learning_rate": 6.450747134774409e-06, + "loss": 0.0209, + "step": 18779 + }, + { + "epoch": 2.0359930615784907, + "grad_norm": 0.408040851354599, + "learning_rate": 6.450021761207023e-06, + "loss": 0.0012, + "step": 18780 + }, + { + "epoch": 2.0361014744145707, + "grad_norm": 4.961700439453125, + "learning_rate": 6.449296387639635e-06, + "loss": 0.0309, + "step": 18781 + }, + { + "epoch": 2.0362098872506507, + "grad_norm": 2.023571729660034, + "learning_rate": 6.448571014072248e-06, + "loss": 0.0295, + "step": 18782 + }, + { + "epoch": 2.03631830008673, + "grad_norm": 0.22327473759651184, + "learning_rate": 6.44784564050486e-06, + "loss": 0.001, + "step": 18783 + }, + { + "epoch": 2.03642671292281, + "grad_norm": 0.02022206038236618, + "learning_rate": 6.447120266937473e-06, + "loss": 0.0001, + "step": 18784 + }, + { + "epoch": 2.0365351257588897, + "grad_norm": 0.01987610198557377, + "learning_rate": 6.446394893370086e-06, + "loss": 0.0001, + "step": 18785 + }, + { + "epoch": 2.0366435385949697, + "grad_norm": 0.4401594400405884, + "learning_rate": 6.4456695198026984e-06, + "loss": 0.002, + "step": 18786 + }, + { + "epoch": 2.0367519514310493, + "grad_norm": 1.7713006734848022, + "learning_rate": 6.4449441462353114e-06, + "loss": 0.0177, + "step": 18787 + }, + { + "epoch": 2.0368603642671292, + "grad_norm": 2.942910671234131, + "learning_rate": 6.444218772667925e-06, + "loss": 0.009, + "step": 18788 + }, + { + "epoch": 2.0369687771032092, + "grad_norm": 0.2338400036096573, + "learning_rate": 6.443493399100537e-06, + "loss": 0.0012, + "step": 18789 + }, + { + "epoch": 2.0370771899392888, + "grad_norm": 6.001022815704346, + "learning_rate": 6.44276802553315e-06, + "loss": 0.0388, + "step": 18790 + }, + { + "epoch": 2.0371856027753688, + "grad_norm": 0.24918904900550842, + "learning_rate": 6.442042651965763e-06, + "loss": 0.0006, + "step": 18791 + }, + { + "epoch": 2.0372940156114483, + "grad_norm": 4.047100067138672, + "learning_rate": 6.4413172783983755e-06, + "loss": 0.0151, + "step": 18792 + }, + { + "epoch": 2.0374024284475283, + "grad_norm": 0.248369038105011, + "learning_rate": 6.4405919048309885e-06, + "loss": 0.0008, + "step": 18793 + }, + { + "epoch": 2.037510841283608, + "grad_norm": 0.2788814604282379, + "learning_rate": 6.439866531263601e-06, + "loss": 0.0009, + "step": 18794 + }, + { + "epoch": 2.037619254119688, + "grad_norm": 0.03950970619916916, + "learning_rate": 6.439141157696214e-06, + "loss": 0.0002, + "step": 18795 + }, + { + "epoch": 2.0377276669557673, + "grad_norm": 1.0295324325561523, + "learning_rate": 6.438415784128827e-06, + "loss": 0.0013, + "step": 18796 + }, + { + "epoch": 2.0378360797918473, + "grad_norm": 1.3103675842285156, + "learning_rate": 6.4376904105614405e-06, + "loss": 0.0391, + "step": 18797 + }, + { + "epoch": 2.0379444926279273, + "grad_norm": 3.895226001739502, + "learning_rate": 6.436965036994053e-06, + "loss": 0.0124, + "step": 18798 + }, + { + "epoch": 2.038052905464007, + "grad_norm": 0.07686520367860794, + "learning_rate": 6.436239663426666e-06, + "loss": 0.0003, + "step": 18799 + }, + { + "epoch": 2.038161318300087, + "grad_norm": 0.04898472875356674, + "learning_rate": 6.435514289859278e-06, + "loss": 0.0002, + "step": 18800 + }, + { + "epoch": 2.0382697311361664, + "grad_norm": 0.15956075489521027, + "learning_rate": 6.434788916291891e-06, + "loss": 0.0005, + "step": 18801 + }, + { + "epoch": 2.0383781439722464, + "grad_norm": 1.0867488384246826, + "learning_rate": 6.434063542724504e-06, + "loss": 0.0024, + "step": 18802 + }, + { + "epoch": 2.038486556808326, + "grad_norm": 0.6239096522331238, + "learning_rate": 6.433338169157116e-06, + "loss": 0.0023, + "step": 18803 + }, + { + "epoch": 2.038594969644406, + "grad_norm": 0.15529945492744446, + "learning_rate": 6.432612795589729e-06, + "loss": 0.0004, + "step": 18804 + }, + { + "epoch": 2.038703382480486, + "grad_norm": 0.2584588825702667, + "learning_rate": 6.431887422022342e-06, + "loss": 0.0009, + "step": 18805 + }, + { + "epoch": 2.0388117953165654, + "grad_norm": 8.744546890258789, + "learning_rate": 6.431162048454956e-06, + "loss": 0.0168, + "step": 18806 + }, + { + "epoch": 2.0389202081526454, + "grad_norm": 0.7973740696907043, + "learning_rate": 6.430436674887568e-06, + "loss": 0.003, + "step": 18807 + }, + { + "epoch": 2.039028620988725, + "grad_norm": 0.9410597085952759, + "learning_rate": 6.429711301320181e-06, + "loss": 0.0048, + "step": 18808 + }, + { + "epoch": 2.039137033824805, + "grad_norm": 0.26370808482170105, + "learning_rate": 6.428985927752793e-06, + "loss": 0.0012, + "step": 18809 + }, + { + "epoch": 2.0392454466608845, + "grad_norm": 0.34171348810195923, + "learning_rate": 6.428260554185406e-06, + "loss": 0.0007, + "step": 18810 + }, + { + "epoch": 2.0393538594969645, + "grad_norm": 2.542588710784912, + "learning_rate": 6.427535180618019e-06, + "loss": 0.0118, + "step": 18811 + }, + { + "epoch": 2.0394622723330444, + "grad_norm": 1.1680340766906738, + "learning_rate": 6.426809807050631e-06, + "loss": 0.0056, + "step": 18812 + }, + { + "epoch": 2.039570685169124, + "grad_norm": 0.27636778354644775, + "learning_rate": 6.426084433483244e-06, + "loss": 0.0006, + "step": 18813 + }, + { + "epoch": 2.039679098005204, + "grad_norm": 0.3447084426879883, + "learning_rate": 6.425359059915858e-06, + "loss": 0.0018, + "step": 18814 + }, + { + "epoch": 2.0397875108412835, + "grad_norm": 0.020495988428592682, + "learning_rate": 6.42463368634847e-06, + "loss": 0.0001, + "step": 18815 + }, + { + "epoch": 2.0398959236773635, + "grad_norm": 13.41328239440918, + "learning_rate": 6.423908312781083e-06, + "loss": 0.0244, + "step": 18816 + }, + { + "epoch": 2.040004336513443, + "grad_norm": 3.8356077671051025, + "learning_rate": 6.423182939213696e-06, + "loss": 0.0117, + "step": 18817 + }, + { + "epoch": 2.040112749349523, + "grad_norm": 0.01630081795156002, + "learning_rate": 6.422457565646308e-06, + "loss": 0.0001, + "step": 18818 + }, + { + "epoch": 2.0402211621856026, + "grad_norm": 4.961716651916504, + "learning_rate": 6.421732192078921e-06, + "loss": 0.0429, + "step": 18819 + }, + { + "epoch": 2.0403295750216826, + "grad_norm": 3.978060007095337, + "learning_rate": 6.421006818511533e-06, + "loss": 0.0153, + "step": 18820 + }, + { + "epoch": 2.0404379878577625, + "grad_norm": 2.5028798580169678, + "learning_rate": 6.420281444944146e-06, + "loss": 0.0314, + "step": 18821 + }, + { + "epoch": 2.040546400693842, + "grad_norm": 1.9674476385116577, + "learning_rate": 6.419556071376759e-06, + "loss": 0.0042, + "step": 18822 + }, + { + "epoch": 2.040654813529922, + "grad_norm": 0.04487277939915657, + "learning_rate": 6.418830697809373e-06, + "loss": 0.0004, + "step": 18823 + }, + { + "epoch": 2.0407632263660016, + "grad_norm": 0.1738506555557251, + "learning_rate": 6.418105324241985e-06, + "loss": 0.001, + "step": 18824 + }, + { + "epoch": 2.0408716392020816, + "grad_norm": 0.3804285526275635, + "learning_rate": 6.417379950674598e-06, + "loss": 0.001, + "step": 18825 + }, + { + "epoch": 2.040980052038161, + "grad_norm": 3.1239829063415527, + "learning_rate": 6.4166545771072104e-06, + "loss": 0.0028, + "step": 18826 + }, + { + "epoch": 2.041088464874241, + "grad_norm": 2.4029853343963623, + "learning_rate": 6.415929203539823e-06, + "loss": 0.0026, + "step": 18827 + }, + { + "epoch": 2.041196877710321, + "grad_norm": 0.009497162885963917, + "learning_rate": 6.415203829972436e-06, + "loss": 0.0001, + "step": 18828 + }, + { + "epoch": 2.0413052905464006, + "grad_norm": 4.122460842132568, + "learning_rate": 6.4144784564050486e-06, + "loss": 0.0206, + "step": 18829 + }, + { + "epoch": 2.0414137033824806, + "grad_norm": 0.40241357684135437, + "learning_rate": 6.4137530828376615e-06, + "loss": 0.0014, + "step": 18830 + }, + { + "epoch": 2.04152211621856, + "grad_norm": 0.03136124461889267, + "learning_rate": 6.4130277092702745e-06, + "loss": 0.0001, + "step": 18831 + }, + { + "epoch": 2.04163052905464, + "grad_norm": 1.1800568103790283, + "learning_rate": 6.412302335702888e-06, + "loss": 0.003, + "step": 18832 + }, + { + "epoch": 2.0417389418907197, + "grad_norm": 0.05585680902004242, + "learning_rate": 6.4115769621355005e-06, + "loss": 0.0002, + "step": 18833 + }, + { + "epoch": 2.0418473547267997, + "grad_norm": 1.5652605295181274, + "learning_rate": 6.4108515885681135e-06, + "loss": 0.0288, + "step": 18834 + }, + { + "epoch": 2.0419557675628797, + "grad_norm": 1.9270910024642944, + "learning_rate": 6.410126215000726e-06, + "loss": 0.0055, + "step": 18835 + }, + { + "epoch": 2.042064180398959, + "grad_norm": 0.05667564272880554, + "learning_rate": 6.409400841433339e-06, + "loss": 0.0002, + "step": 18836 + }, + { + "epoch": 2.042172593235039, + "grad_norm": 0.04337547346949577, + "learning_rate": 6.408675467865952e-06, + "loss": 0.0003, + "step": 18837 + }, + { + "epoch": 2.0422810060711187, + "grad_norm": 12.918886184692383, + "learning_rate": 6.407950094298564e-06, + "loss": 0.0273, + "step": 18838 + }, + { + "epoch": 2.0423894189071987, + "grad_norm": 1.6630816459655762, + "learning_rate": 6.407224720731177e-06, + "loss": 0.0185, + "step": 18839 + }, + { + "epoch": 2.0424978317432783, + "grad_norm": 1.6834344863891602, + "learning_rate": 6.406499347163791e-06, + "loss": 0.035, + "step": 18840 + }, + { + "epoch": 2.0426062445793582, + "grad_norm": 0.0354786217212677, + "learning_rate": 6.405773973596403e-06, + "loss": 0.0003, + "step": 18841 + }, + { + "epoch": 2.042714657415438, + "grad_norm": 0.541279673576355, + "learning_rate": 6.405048600029016e-06, + "loss": 0.0013, + "step": 18842 + }, + { + "epoch": 2.0428230702515178, + "grad_norm": 5.071564197540283, + "learning_rate": 6.404323226461629e-06, + "loss": 0.0164, + "step": 18843 + }, + { + "epoch": 2.0429314830875978, + "grad_norm": 4.240568161010742, + "learning_rate": 6.403597852894241e-06, + "loss": 0.0227, + "step": 18844 + }, + { + "epoch": 2.0430398959236773, + "grad_norm": 0.028799237683415413, + "learning_rate": 6.402872479326854e-06, + "loss": 0.0001, + "step": 18845 + }, + { + "epoch": 2.0431483087597573, + "grad_norm": 0.5010915994644165, + "learning_rate": 6.402147105759466e-06, + "loss": 0.002, + "step": 18846 + }, + { + "epoch": 2.043256721595837, + "grad_norm": 0.16707395017147064, + "learning_rate": 6.401421732192079e-06, + "loss": 0.0006, + "step": 18847 + }, + { + "epoch": 2.043365134431917, + "grad_norm": 0.6922922730445862, + "learning_rate": 6.400696358624692e-06, + "loss": 0.0012, + "step": 18848 + }, + { + "epoch": 2.0434735472679963, + "grad_norm": 0.4179852306842804, + "learning_rate": 6.399970985057306e-06, + "loss": 0.0019, + "step": 18849 + }, + { + "epoch": 2.0435819601040763, + "grad_norm": 1.5359461307525635, + "learning_rate": 6.399245611489918e-06, + "loss": 0.0281, + "step": 18850 + }, + { + "epoch": 2.0436903729401563, + "grad_norm": 0.01894688419997692, + "learning_rate": 6.398520237922531e-06, + "loss": 0.0001, + "step": 18851 + }, + { + "epoch": 2.043798785776236, + "grad_norm": 7.681092739105225, + "learning_rate": 6.397794864355143e-06, + "loss": 0.027, + "step": 18852 + }, + { + "epoch": 2.043907198612316, + "grad_norm": 0.018769392743706703, + "learning_rate": 6.397069490787756e-06, + "loss": 0.0001, + "step": 18853 + }, + { + "epoch": 2.0440156114483954, + "grad_norm": 0.8901007771492004, + "learning_rate": 6.396344117220369e-06, + "loss": 0.0048, + "step": 18854 + }, + { + "epoch": 2.0441240242844754, + "grad_norm": 6.866899013519287, + "learning_rate": 6.395618743652981e-06, + "loss": 0.0106, + "step": 18855 + }, + { + "epoch": 2.044232437120555, + "grad_norm": 6.560752868652344, + "learning_rate": 6.394893370085594e-06, + "loss": 0.0112, + "step": 18856 + }, + { + "epoch": 2.044340849956635, + "grad_norm": 0.7836070656776428, + "learning_rate": 6.394167996518207e-06, + "loss": 0.0025, + "step": 18857 + }, + { + "epoch": 2.0444492627927144, + "grad_norm": 0.5010836124420166, + "learning_rate": 6.393442622950821e-06, + "loss": 0.0047, + "step": 18858 + }, + { + "epoch": 2.0445576756287944, + "grad_norm": 0.21631452441215515, + "learning_rate": 6.392717249383433e-06, + "loss": 0.0008, + "step": 18859 + }, + { + "epoch": 2.0446660884648744, + "grad_norm": 9.648273468017578, + "learning_rate": 6.391991875816046e-06, + "loss": 0.0348, + "step": 18860 + }, + { + "epoch": 2.044774501300954, + "grad_norm": 4.651060581207275, + "learning_rate": 6.391266502248658e-06, + "loss": 0.0103, + "step": 18861 + }, + { + "epoch": 2.044882914137034, + "grad_norm": 0.12043676525354385, + "learning_rate": 6.390541128681271e-06, + "loss": 0.0006, + "step": 18862 + }, + { + "epoch": 2.0449913269731135, + "grad_norm": 0.03581549972295761, + "learning_rate": 6.389815755113884e-06, + "loss": 0.0003, + "step": 18863 + }, + { + "epoch": 2.0450997398091935, + "grad_norm": 0.5463348031044006, + "learning_rate": 6.3890903815464964e-06, + "loss": 0.0019, + "step": 18864 + }, + { + "epoch": 2.045208152645273, + "grad_norm": 0.38055381178855896, + "learning_rate": 6.3883650079791094e-06, + "loss": 0.0011, + "step": 18865 + }, + { + "epoch": 2.045316565481353, + "grad_norm": 0.16208690404891968, + "learning_rate": 6.387639634411723e-06, + "loss": 0.0009, + "step": 18866 + }, + { + "epoch": 2.045424978317433, + "grad_norm": 0.09931011497974396, + "learning_rate": 6.386914260844335e-06, + "loss": 0.0006, + "step": 18867 + }, + { + "epoch": 2.0455333911535125, + "grad_norm": 0.15313243865966797, + "learning_rate": 6.386188887276948e-06, + "loss": 0.0007, + "step": 18868 + }, + { + "epoch": 2.0456418039895925, + "grad_norm": 2.72422194480896, + "learning_rate": 6.385463513709561e-06, + "loss": 0.0083, + "step": 18869 + }, + { + "epoch": 2.045750216825672, + "grad_norm": 6.670919895172119, + "learning_rate": 6.3847381401421735e-06, + "loss": 0.0311, + "step": 18870 + }, + { + "epoch": 2.045858629661752, + "grad_norm": 15.031359672546387, + "learning_rate": 6.3840127665747865e-06, + "loss": 0.0132, + "step": 18871 + }, + { + "epoch": 2.0459670424978316, + "grad_norm": 3.775804281234741, + "learning_rate": 6.383287393007399e-06, + "loss": 0.0091, + "step": 18872 + }, + { + "epoch": 2.0460754553339116, + "grad_norm": 2.296100616455078, + "learning_rate": 6.382562019440012e-06, + "loss": 0.0396, + "step": 18873 + }, + { + "epoch": 2.0461838681699915, + "grad_norm": 1.512836217880249, + "learning_rate": 6.381836645872625e-06, + "loss": 0.0136, + "step": 18874 + }, + { + "epoch": 2.046292281006071, + "grad_norm": 5.429705619812012, + "learning_rate": 6.3811112723052385e-06, + "loss": 0.0392, + "step": 18875 + }, + { + "epoch": 2.046400693842151, + "grad_norm": 0.8446491956710815, + "learning_rate": 6.380385898737851e-06, + "loss": 0.001, + "step": 18876 + }, + { + "epoch": 2.0465091066782306, + "grad_norm": 0.9730046987533569, + "learning_rate": 6.379660525170464e-06, + "loss": 0.0028, + "step": 18877 + }, + { + "epoch": 2.0466175195143106, + "grad_norm": 0.5062848329544067, + "learning_rate": 6.378935151603077e-06, + "loss": 0.002, + "step": 18878 + }, + { + "epoch": 2.04672593235039, + "grad_norm": 2.5760042667388916, + "learning_rate": 6.378209778035689e-06, + "loss": 0.025, + "step": 18879 + }, + { + "epoch": 2.04683434518647, + "grad_norm": 0.5794180035591125, + "learning_rate": 6.377484404468302e-06, + "loss": 0.0018, + "step": 18880 + }, + { + "epoch": 2.0469427580225497, + "grad_norm": 0.03717683628201485, + "learning_rate": 6.376759030900914e-06, + "loss": 0.0002, + "step": 18881 + }, + { + "epoch": 2.0470511708586296, + "grad_norm": 0.7657269835472107, + "learning_rate": 6.376033657333527e-06, + "loss": 0.0024, + "step": 18882 + }, + { + "epoch": 2.0471595836947096, + "grad_norm": 5.779501914978027, + "learning_rate": 6.375308283766141e-06, + "loss": 0.0711, + "step": 18883 + }, + { + "epoch": 2.047267996530789, + "grad_norm": 2.7096362113952637, + "learning_rate": 6.374582910198754e-06, + "loss": 0.0151, + "step": 18884 + }, + { + "epoch": 2.047376409366869, + "grad_norm": 0.7221409678459167, + "learning_rate": 6.373857536631366e-06, + "loss": 0.0038, + "step": 18885 + }, + { + "epoch": 2.0474848222029487, + "grad_norm": 1.017755389213562, + "learning_rate": 6.373132163063979e-06, + "loss": 0.0041, + "step": 18886 + }, + { + "epoch": 2.0475932350390287, + "grad_norm": 1.1781315803527832, + "learning_rate": 6.372406789496591e-06, + "loss": 0.0069, + "step": 18887 + }, + { + "epoch": 2.047701647875108, + "grad_norm": 5.559932231903076, + "learning_rate": 6.371681415929204e-06, + "loss": 0.0156, + "step": 18888 + }, + { + "epoch": 2.047810060711188, + "grad_norm": 0.10888484865427017, + "learning_rate": 6.370956042361817e-06, + "loss": 0.0003, + "step": 18889 + }, + { + "epoch": 2.047918473547268, + "grad_norm": 0.13727892935276031, + "learning_rate": 6.370230668794429e-06, + "loss": 0.0006, + "step": 18890 + }, + { + "epoch": 2.0480268863833477, + "grad_norm": 5.313038349151611, + "learning_rate": 6.369505295227042e-06, + "loss": 0.0252, + "step": 18891 + }, + { + "epoch": 2.0481352992194277, + "grad_norm": 0.4430289566516876, + "learning_rate": 6.368779921659656e-06, + "loss": 0.0023, + "step": 18892 + }, + { + "epoch": 2.0482437120555073, + "grad_norm": 0.3053323030471802, + "learning_rate": 6.368054548092268e-06, + "loss": 0.0013, + "step": 18893 + }, + { + "epoch": 2.0483521248915872, + "grad_norm": 0.028518743813037872, + "learning_rate": 6.367329174524881e-06, + "loss": 0.0002, + "step": 18894 + }, + { + "epoch": 2.048460537727667, + "grad_norm": 0.2554190456867218, + "learning_rate": 6.366603800957494e-06, + "loss": 0.0012, + "step": 18895 + }, + { + "epoch": 2.0485689505637468, + "grad_norm": 0.3541918098926544, + "learning_rate": 6.365878427390106e-06, + "loss": 0.0009, + "step": 18896 + }, + { + "epoch": 2.0486773633998268, + "grad_norm": 0.3454388976097107, + "learning_rate": 6.365153053822719e-06, + "loss": 0.0007, + "step": 18897 + }, + { + "epoch": 2.0487857762359063, + "grad_norm": 0.2806771695613861, + "learning_rate": 6.364427680255331e-06, + "loss": 0.0014, + "step": 18898 + }, + { + "epoch": 2.0488941890719863, + "grad_norm": 1.4948887825012207, + "learning_rate": 6.363702306687944e-06, + "loss": 0.0112, + "step": 18899 + }, + { + "epoch": 2.049002601908066, + "grad_norm": 0.14377617835998535, + "learning_rate": 6.362976933120557e-06, + "loss": 0.001, + "step": 18900 + }, + { + "epoch": 2.049111014744146, + "grad_norm": 0.1219293475151062, + "learning_rate": 6.362251559553171e-06, + "loss": 0.0004, + "step": 18901 + }, + { + "epoch": 2.0492194275802254, + "grad_norm": 2.2714617252349854, + "learning_rate": 6.361526185985783e-06, + "loss": 0.0045, + "step": 18902 + }, + { + "epoch": 2.0493278404163053, + "grad_norm": 0.3839043974876404, + "learning_rate": 6.360800812418396e-06, + "loss": 0.0015, + "step": 18903 + }, + { + "epoch": 2.049436253252385, + "grad_norm": 0.12627285718917847, + "learning_rate": 6.360075438851009e-06, + "loss": 0.0004, + "step": 18904 + }, + { + "epoch": 2.049544666088465, + "grad_norm": 3.1294946670532227, + "learning_rate": 6.3593500652836214e-06, + "loss": 0.0051, + "step": 18905 + }, + { + "epoch": 2.049653078924545, + "grad_norm": 2.4863691329956055, + "learning_rate": 6.358624691716234e-06, + "loss": 0.0214, + "step": 18906 + }, + { + "epoch": 2.0497614917606244, + "grad_norm": 2.0094027519226074, + "learning_rate": 6.3578993181488466e-06, + "loss": 0.0358, + "step": 18907 + }, + { + "epoch": 2.0498699045967044, + "grad_norm": 0.3312288522720337, + "learning_rate": 6.3571739445814596e-06, + "loss": 0.0006, + "step": 18908 + }, + { + "epoch": 2.049978317432784, + "grad_norm": 3.530421495437622, + "learning_rate": 6.356448571014073e-06, + "loss": 0.0096, + "step": 18909 + }, + { + "epoch": 2.050086730268864, + "grad_norm": 6.507027626037598, + "learning_rate": 6.355723197446686e-06, + "loss": 0.0274, + "step": 18910 + }, + { + "epoch": 2.0501951431049434, + "grad_norm": 0.1202240064740181, + "learning_rate": 6.3549978238792985e-06, + "loss": 0.0007, + "step": 18911 + }, + { + "epoch": 2.0503035559410234, + "grad_norm": 0.2709461748600006, + "learning_rate": 6.3542724503119115e-06, + "loss": 0.001, + "step": 18912 + }, + { + "epoch": 2.0504119687771034, + "grad_norm": 0.0070637548342347145, + "learning_rate": 6.353547076744524e-06, + "loss": 0.0, + "step": 18913 + }, + { + "epoch": 2.050520381613183, + "grad_norm": 1.311302900314331, + "learning_rate": 6.352821703177137e-06, + "loss": 0.0023, + "step": 18914 + }, + { + "epoch": 2.050628794449263, + "grad_norm": 1.1164631843566895, + "learning_rate": 6.35209632960975e-06, + "loss": 0.004, + "step": 18915 + }, + { + "epoch": 2.0507372072853425, + "grad_norm": 6.80495023727417, + "learning_rate": 6.351370956042362e-06, + "loss": 0.0219, + "step": 18916 + }, + { + "epoch": 2.0508456201214225, + "grad_norm": 1.4555063247680664, + "learning_rate": 6.350645582474975e-06, + "loss": 0.0091, + "step": 18917 + }, + { + "epoch": 2.050954032957502, + "grad_norm": 5.611311912536621, + "learning_rate": 6.349920208907589e-06, + "loss": 0.0164, + "step": 18918 + }, + { + "epoch": 2.051062445793582, + "grad_norm": 0.4938640892505646, + "learning_rate": 6.349194835340201e-06, + "loss": 0.002, + "step": 18919 + }, + { + "epoch": 2.0511708586296615, + "grad_norm": 3.441410779953003, + "learning_rate": 6.348469461772814e-06, + "loss": 0.0089, + "step": 18920 + }, + { + "epoch": 2.0512792714657415, + "grad_norm": 0.9432587623596191, + "learning_rate": 6.347744088205427e-06, + "loss": 0.0019, + "step": 18921 + }, + { + "epoch": 2.0513876843018215, + "grad_norm": 3.3767454624176025, + "learning_rate": 6.347018714638039e-06, + "loss": 0.0155, + "step": 18922 + }, + { + "epoch": 2.051496097137901, + "grad_norm": 0.00984665472060442, + "learning_rate": 6.346293341070652e-06, + "loss": 0.0001, + "step": 18923 + }, + { + "epoch": 2.051604509973981, + "grad_norm": 10.680902481079102, + "learning_rate": 6.345567967503264e-06, + "loss": 0.0399, + "step": 18924 + }, + { + "epoch": 2.0517129228100606, + "grad_norm": 0.1908004879951477, + "learning_rate": 6.344842593935877e-06, + "loss": 0.0007, + "step": 18925 + }, + { + "epoch": 2.0518213356461406, + "grad_norm": 0.04947599396109581, + "learning_rate": 6.34411722036849e-06, + "loss": 0.0002, + "step": 18926 + }, + { + "epoch": 2.05192974848222, + "grad_norm": 10.098468780517578, + "learning_rate": 6.343391846801104e-06, + "loss": 0.0743, + "step": 18927 + }, + { + "epoch": 2.0520381613183, + "grad_norm": 1.036812663078308, + "learning_rate": 6.342666473233716e-06, + "loss": 0.0029, + "step": 18928 + }, + { + "epoch": 2.05214657415438, + "grad_norm": 0.4031166434288025, + "learning_rate": 6.341941099666329e-06, + "loss": 0.0012, + "step": 18929 + }, + { + "epoch": 2.0522549869904596, + "grad_norm": 0.19359460473060608, + "learning_rate": 6.341215726098942e-06, + "loss": 0.0006, + "step": 18930 + }, + { + "epoch": 2.0523633998265396, + "grad_norm": 3.4582901000976562, + "learning_rate": 6.340490352531554e-06, + "loss": 0.0265, + "step": 18931 + }, + { + "epoch": 2.052471812662619, + "grad_norm": 0.07610389590263367, + "learning_rate": 6.339764978964167e-06, + "loss": 0.0004, + "step": 18932 + }, + { + "epoch": 2.052580225498699, + "grad_norm": 5.8344902992248535, + "learning_rate": 6.339039605396779e-06, + "loss": 0.0127, + "step": 18933 + }, + { + "epoch": 2.0526886383347787, + "grad_norm": 13.748663902282715, + "learning_rate": 6.338314231829392e-06, + "loss": 0.0227, + "step": 18934 + }, + { + "epoch": 2.0527970511708586, + "grad_norm": 3.309340000152588, + "learning_rate": 6.337588858262006e-06, + "loss": 0.011, + "step": 18935 + }, + { + "epoch": 2.0529054640069386, + "grad_norm": 0.493764728307724, + "learning_rate": 6.336863484694619e-06, + "loss": 0.0007, + "step": 18936 + }, + { + "epoch": 2.053013876843018, + "grad_norm": 0.09566020220518112, + "learning_rate": 6.336138111127231e-06, + "loss": 0.0006, + "step": 18937 + }, + { + "epoch": 2.053122289679098, + "grad_norm": 0.2596309185028076, + "learning_rate": 6.335412737559844e-06, + "loss": 0.0007, + "step": 18938 + }, + { + "epoch": 2.0532307025151777, + "grad_norm": 0.6949326395988464, + "learning_rate": 6.334687363992456e-06, + "loss": 0.002, + "step": 18939 + }, + { + "epoch": 2.0533391153512577, + "grad_norm": 4.8643269538879395, + "learning_rate": 6.333961990425069e-06, + "loss": 0.0264, + "step": 18940 + }, + { + "epoch": 2.0534475281873372, + "grad_norm": 1.2763304710388184, + "learning_rate": 6.333236616857682e-06, + "loss": 0.0113, + "step": 18941 + }, + { + "epoch": 2.053555941023417, + "grad_norm": 0.034826070070266724, + "learning_rate": 6.3325112432902945e-06, + "loss": 0.0003, + "step": 18942 + }, + { + "epoch": 2.0536643538594967, + "grad_norm": 0.08398748934268951, + "learning_rate": 6.3317858697229074e-06, + "loss": 0.0004, + "step": 18943 + }, + { + "epoch": 2.0537727666955767, + "grad_norm": 0.008263272233307362, + "learning_rate": 6.331060496155521e-06, + "loss": 0.0001, + "step": 18944 + }, + { + "epoch": 2.0538811795316567, + "grad_norm": 3.5829646587371826, + "learning_rate": 6.330335122588133e-06, + "loss": 0.0027, + "step": 18945 + }, + { + "epoch": 2.0539895923677363, + "grad_norm": 1.077473759651184, + "learning_rate": 6.329609749020746e-06, + "loss": 0.0031, + "step": 18946 + }, + { + "epoch": 2.0540980052038162, + "grad_norm": 0.09920331090688705, + "learning_rate": 6.328884375453359e-06, + "loss": 0.0003, + "step": 18947 + }, + { + "epoch": 2.054206418039896, + "grad_norm": 1.4562740325927734, + "learning_rate": 6.3281590018859715e-06, + "loss": 0.0053, + "step": 18948 + }, + { + "epoch": 2.0543148308759758, + "grad_norm": 0.3965950310230255, + "learning_rate": 6.3274336283185845e-06, + "loss": 0.0016, + "step": 18949 + }, + { + "epoch": 2.0544232437120553, + "grad_norm": 0.04330059885978699, + "learning_rate": 6.326708254751197e-06, + "loss": 0.0002, + "step": 18950 + }, + { + "epoch": 2.0545316565481353, + "grad_norm": 0.4975176751613617, + "learning_rate": 6.32598288118381e-06, + "loss": 0.0011, + "step": 18951 + }, + { + "epoch": 2.0546400693842153, + "grad_norm": 0.022275133058428764, + "learning_rate": 6.325257507616423e-06, + "loss": 0.0002, + "step": 18952 + }, + { + "epoch": 2.054748482220295, + "grad_norm": 5.979119300842285, + "learning_rate": 6.3245321340490365e-06, + "loss": 0.0193, + "step": 18953 + }, + { + "epoch": 2.054856895056375, + "grad_norm": 0.13791847229003906, + "learning_rate": 6.323806760481649e-06, + "loss": 0.0005, + "step": 18954 + }, + { + "epoch": 2.0549653078924544, + "grad_norm": 0.11586098372936249, + "learning_rate": 6.323081386914262e-06, + "loss": 0.0003, + "step": 18955 + }, + { + "epoch": 2.0550737207285343, + "grad_norm": 1.046440839767456, + "learning_rate": 6.322356013346875e-06, + "loss": 0.0015, + "step": 18956 + }, + { + "epoch": 2.055182133564614, + "grad_norm": 0.10409262031316757, + "learning_rate": 6.321630639779487e-06, + "loss": 0.0007, + "step": 18957 + }, + { + "epoch": 2.055290546400694, + "grad_norm": 1.0942884683609009, + "learning_rate": 6.3209052662121e-06, + "loss": 0.0033, + "step": 18958 + }, + { + "epoch": 2.055398959236774, + "grad_norm": 0.7314119338989258, + "learning_rate": 6.320179892644712e-06, + "loss": 0.0018, + "step": 18959 + }, + { + "epoch": 2.0555073720728534, + "grad_norm": 3.6879148483276367, + "learning_rate": 6.319454519077325e-06, + "loss": 0.0028, + "step": 18960 + }, + { + "epoch": 2.0556157849089334, + "grad_norm": 0.013137657195329666, + "learning_rate": 6.318729145509939e-06, + "loss": 0.0, + "step": 18961 + }, + { + "epoch": 2.055724197745013, + "grad_norm": 6.117608547210693, + "learning_rate": 6.318003771942552e-06, + "loss": 0.01, + "step": 18962 + }, + { + "epoch": 2.055832610581093, + "grad_norm": 0.03691850230097771, + "learning_rate": 6.317278398375164e-06, + "loss": 0.0002, + "step": 18963 + }, + { + "epoch": 2.0559410234171724, + "grad_norm": 0.8247113823890686, + "learning_rate": 6.316553024807777e-06, + "loss": 0.0011, + "step": 18964 + }, + { + "epoch": 2.0560494362532524, + "grad_norm": 2.3403828144073486, + "learning_rate": 6.315827651240389e-06, + "loss": 0.0583, + "step": 18965 + }, + { + "epoch": 2.056157849089332, + "grad_norm": 6.676750183105469, + "learning_rate": 6.315102277673002e-06, + "loss": 0.0085, + "step": 18966 + }, + { + "epoch": 2.056266261925412, + "grad_norm": 3.1205320358276367, + "learning_rate": 6.314376904105615e-06, + "loss": 0.0168, + "step": 18967 + }, + { + "epoch": 2.056374674761492, + "grad_norm": 21.440073013305664, + "learning_rate": 6.313651530538227e-06, + "loss": 0.0137, + "step": 18968 + }, + { + "epoch": 2.0564830875975715, + "grad_norm": 0.022768931463360786, + "learning_rate": 6.31292615697084e-06, + "loss": 0.0001, + "step": 18969 + }, + { + "epoch": 2.0565915004336515, + "grad_norm": 0.09920427948236465, + "learning_rate": 6.312200783403454e-06, + "loss": 0.0004, + "step": 18970 + }, + { + "epoch": 2.056699913269731, + "grad_norm": 1.2014827728271484, + "learning_rate": 6.311475409836066e-06, + "loss": 0.0009, + "step": 18971 + }, + { + "epoch": 2.056808326105811, + "grad_norm": 3.4605891704559326, + "learning_rate": 6.310750036268679e-06, + "loss": 0.0718, + "step": 18972 + }, + { + "epoch": 2.0569167389418905, + "grad_norm": 0.9604441523551941, + "learning_rate": 6.310024662701292e-06, + "loss": 0.0013, + "step": 18973 + }, + { + "epoch": 2.0570251517779705, + "grad_norm": 3.244666814804077, + "learning_rate": 6.309299289133904e-06, + "loss": 0.009, + "step": 18974 + }, + { + "epoch": 2.0571335646140505, + "grad_norm": 2.68182110786438, + "learning_rate": 6.308573915566517e-06, + "loss": 0.0097, + "step": 18975 + }, + { + "epoch": 2.05724197745013, + "grad_norm": 0.6008235812187195, + "learning_rate": 6.307848541999129e-06, + "loss": 0.0005, + "step": 18976 + }, + { + "epoch": 2.05735039028621, + "grad_norm": 0.0802575871348381, + "learning_rate": 6.307123168431742e-06, + "loss": 0.0003, + "step": 18977 + }, + { + "epoch": 2.0574588031222896, + "grad_norm": 8.803848266601562, + "learning_rate": 6.306397794864355e-06, + "loss": 0.0058, + "step": 18978 + }, + { + "epoch": 2.0575672159583696, + "grad_norm": 1.4122554063796997, + "learning_rate": 6.305672421296969e-06, + "loss": 0.0143, + "step": 18979 + }, + { + "epoch": 2.057675628794449, + "grad_norm": 0.06989064812660217, + "learning_rate": 6.304947047729581e-06, + "loss": 0.0003, + "step": 18980 + }, + { + "epoch": 2.057784041630529, + "grad_norm": 1.6755207777023315, + "learning_rate": 6.304221674162194e-06, + "loss": 0.0182, + "step": 18981 + }, + { + "epoch": 2.057892454466609, + "grad_norm": 0.0314132496714592, + "learning_rate": 6.303496300594807e-06, + "loss": 0.0001, + "step": 18982 + }, + { + "epoch": 2.0580008673026886, + "grad_norm": 0.8520947098731995, + "learning_rate": 6.3027709270274194e-06, + "loss": 0.0029, + "step": 18983 + }, + { + "epoch": 2.0581092801387686, + "grad_norm": 0.43148043751716614, + "learning_rate": 6.3020455534600324e-06, + "loss": 0.0014, + "step": 18984 + }, + { + "epoch": 2.058217692974848, + "grad_norm": 1.7710967063903809, + "learning_rate": 6.3013201798926446e-06, + "loss": 0.0103, + "step": 18985 + }, + { + "epoch": 2.058326105810928, + "grad_norm": 3.0263495445251465, + "learning_rate": 6.3005948063252576e-06, + "loss": 0.0141, + "step": 18986 + }, + { + "epoch": 2.0584345186470077, + "grad_norm": 0.01871432177722454, + "learning_rate": 6.299869432757871e-06, + "loss": 0.0001, + "step": 18987 + }, + { + "epoch": 2.0585429314830876, + "grad_norm": 1.4698753356933594, + "learning_rate": 6.299144059190484e-06, + "loss": 0.0018, + "step": 18988 + }, + { + "epoch": 2.058651344319167, + "grad_norm": 0.08365479856729507, + "learning_rate": 6.2984186856230965e-06, + "loss": 0.0002, + "step": 18989 + }, + { + "epoch": 2.058759757155247, + "grad_norm": 0.16533228754997253, + "learning_rate": 6.2976933120557095e-06, + "loss": 0.0005, + "step": 18990 + }, + { + "epoch": 2.058868169991327, + "grad_norm": 4.8042497634887695, + "learning_rate": 6.296967938488322e-06, + "loss": 0.0509, + "step": 18991 + }, + { + "epoch": 2.0589765828274067, + "grad_norm": 0.02156057395040989, + "learning_rate": 6.296242564920935e-06, + "loss": 0.0001, + "step": 18992 + }, + { + "epoch": 2.0590849956634867, + "grad_norm": 0.176982581615448, + "learning_rate": 6.295517191353548e-06, + "loss": 0.0007, + "step": 18993 + }, + { + "epoch": 2.0591934084995662, + "grad_norm": 1.1738901138305664, + "learning_rate": 6.29479181778616e-06, + "loss": 0.0027, + "step": 18994 + }, + { + "epoch": 2.059301821335646, + "grad_norm": 5.83311128616333, + "learning_rate": 6.294066444218773e-06, + "loss": 0.0062, + "step": 18995 + }, + { + "epoch": 2.0594102341717258, + "grad_norm": 0.09089570492506027, + "learning_rate": 6.293341070651387e-06, + "loss": 0.0003, + "step": 18996 + }, + { + "epoch": 2.0595186470078057, + "grad_norm": 4.573829174041748, + "learning_rate": 6.292615697083999e-06, + "loss": 0.0285, + "step": 18997 + }, + { + "epoch": 2.0596270598438857, + "grad_norm": 0.056326724588871, + "learning_rate": 6.291890323516612e-06, + "loss": 0.0003, + "step": 18998 + }, + { + "epoch": 2.0597354726799653, + "grad_norm": 0.8252557516098022, + "learning_rate": 6.291164949949225e-06, + "loss": 0.0051, + "step": 18999 + }, + { + "epoch": 2.0598438855160452, + "grad_norm": 0.8242493271827698, + "learning_rate": 6.290439576381837e-06, + "loss": 0.0029, + "step": 19000 + }, + { + "epoch": 2.059952298352125, + "grad_norm": 0.12671856582164764, + "learning_rate": 6.28971420281445e-06, + "loss": 0.0004, + "step": 19001 + }, + { + "epoch": 2.0600607111882048, + "grad_norm": 6.769077777862549, + "learning_rate": 6.288988829247063e-06, + "loss": 0.0159, + "step": 19002 + }, + { + "epoch": 2.0601691240242843, + "grad_norm": 0.9398286938667297, + "learning_rate": 6.288263455679675e-06, + "loss": 0.0018, + "step": 19003 + }, + { + "epoch": 2.0602775368603643, + "grad_norm": 0.020061520859599113, + "learning_rate": 6.287538082112289e-06, + "loss": 0.0002, + "step": 19004 + }, + { + "epoch": 2.060385949696444, + "grad_norm": 0.15763899683952332, + "learning_rate": 6.286812708544902e-06, + "loss": 0.0006, + "step": 19005 + }, + { + "epoch": 2.060494362532524, + "grad_norm": 0.19491910934448242, + "learning_rate": 6.286087334977514e-06, + "loss": 0.0004, + "step": 19006 + }, + { + "epoch": 2.060602775368604, + "grad_norm": 8.754105567932129, + "learning_rate": 6.285361961410127e-06, + "loss": 0.0197, + "step": 19007 + }, + { + "epoch": 2.0607111882046834, + "grad_norm": 2.794715404510498, + "learning_rate": 6.28463658784274e-06, + "loss": 0.0223, + "step": 19008 + }, + { + "epoch": 2.0608196010407633, + "grad_norm": 12.819300651550293, + "learning_rate": 6.283911214275352e-06, + "loss": 0.0182, + "step": 19009 + }, + { + "epoch": 2.060928013876843, + "grad_norm": 0.18336215615272522, + "learning_rate": 6.283185840707965e-06, + "loss": 0.0005, + "step": 19010 + }, + { + "epoch": 2.061036426712923, + "grad_norm": 2.267434597015381, + "learning_rate": 6.282460467140577e-06, + "loss": 0.0018, + "step": 19011 + }, + { + "epoch": 2.0611448395490024, + "grad_norm": 3.001725673675537, + "learning_rate": 6.28173509357319e-06, + "loss": 0.0091, + "step": 19012 + }, + { + "epoch": 2.0612532523850824, + "grad_norm": 2.934792995452881, + "learning_rate": 6.281009720005804e-06, + "loss": 0.0064, + "step": 19013 + }, + { + "epoch": 2.0613616652211624, + "grad_norm": 0.12533317506313324, + "learning_rate": 6.280284346438417e-06, + "loss": 0.0008, + "step": 19014 + }, + { + "epoch": 2.061470078057242, + "grad_norm": 0.19322866201400757, + "learning_rate": 6.279558972871029e-06, + "loss": 0.0005, + "step": 19015 + }, + { + "epoch": 2.061578490893322, + "grad_norm": 0.7694514393806458, + "learning_rate": 6.278833599303642e-06, + "loss": 0.0013, + "step": 19016 + }, + { + "epoch": 2.0616869037294014, + "grad_norm": 0.30577850341796875, + "learning_rate": 6.278108225736254e-06, + "loss": 0.0013, + "step": 19017 + }, + { + "epoch": 2.0617953165654814, + "grad_norm": 0.19491241872310638, + "learning_rate": 6.277382852168867e-06, + "loss": 0.0005, + "step": 19018 + }, + { + "epoch": 2.061903729401561, + "grad_norm": 0.5932580232620239, + "learning_rate": 6.27665747860148e-06, + "loss": 0.0042, + "step": 19019 + }, + { + "epoch": 2.062012142237641, + "grad_norm": 0.220469668507576, + "learning_rate": 6.2759321050340925e-06, + "loss": 0.0045, + "step": 19020 + }, + { + "epoch": 2.062120555073721, + "grad_norm": 0.07453195750713348, + "learning_rate": 6.2752067314667054e-06, + "loss": 0.0005, + "step": 19021 + }, + { + "epoch": 2.0622289679098005, + "grad_norm": 0.04614759609103203, + "learning_rate": 6.274481357899319e-06, + "loss": 0.0003, + "step": 19022 + }, + { + "epoch": 2.0623373807458805, + "grad_norm": 8.050545692443848, + "learning_rate": 6.2737559843319314e-06, + "loss": 0.0175, + "step": 19023 + }, + { + "epoch": 2.06244579358196, + "grad_norm": 0.10334652662277222, + "learning_rate": 6.273030610764544e-06, + "loss": 0.0003, + "step": 19024 + }, + { + "epoch": 2.06255420641804, + "grad_norm": 2.511455774307251, + "learning_rate": 6.272305237197157e-06, + "loss": 0.0083, + "step": 19025 + }, + { + "epoch": 2.0626626192541195, + "grad_norm": 2.4099836349487305, + "learning_rate": 6.2715798636297696e-06, + "loss": 0.0176, + "step": 19026 + }, + { + "epoch": 2.0627710320901995, + "grad_norm": 0.761248767375946, + "learning_rate": 6.2708544900623825e-06, + "loss": 0.0029, + "step": 19027 + }, + { + "epoch": 2.062879444926279, + "grad_norm": 3.6863770484924316, + "learning_rate": 6.2701291164949955e-06, + "loss": 0.006, + "step": 19028 + }, + { + "epoch": 2.062987857762359, + "grad_norm": 4.514988422393799, + "learning_rate": 6.269403742927608e-06, + "loss": 0.0404, + "step": 19029 + }, + { + "epoch": 2.063096270598439, + "grad_norm": 2.130574941635132, + "learning_rate": 6.2686783693602215e-06, + "loss": 0.0179, + "step": 19030 + }, + { + "epoch": 2.0632046834345186, + "grad_norm": 0.06498458236455917, + "learning_rate": 6.2679529957928345e-06, + "loss": 0.0002, + "step": 19031 + }, + { + "epoch": 2.0633130962705986, + "grad_norm": 0.17808955907821655, + "learning_rate": 6.267227622225447e-06, + "loss": 0.0008, + "step": 19032 + }, + { + "epoch": 2.063421509106678, + "grad_norm": 0.9100040197372437, + "learning_rate": 6.26650224865806e-06, + "loss": 0.0019, + "step": 19033 + }, + { + "epoch": 2.063529921942758, + "grad_norm": 1.2865829467773438, + "learning_rate": 6.265776875090673e-06, + "loss": 0.0011, + "step": 19034 + }, + { + "epoch": 2.0636383347788376, + "grad_norm": 3.577596426010132, + "learning_rate": 6.265051501523285e-06, + "loss": 0.0283, + "step": 19035 + }, + { + "epoch": 2.0637467476149176, + "grad_norm": 0.19311641156673431, + "learning_rate": 6.264326127955898e-06, + "loss": 0.0009, + "step": 19036 + }, + { + "epoch": 2.0638551604509976, + "grad_norm": 0.16446538269519806, + "learning_rate": 6.26360075438851e-06, + "loss": 0.0005, + "step": 19037 + }, + { + "epoch": 2.063963573287077, + "grad_norm": 1.2613228559494019, + "learning_rate": 6.262875380821123e-06, + "loss": 0.0024, + "step": 19038 + }, + { + "epoch": 2.064071986123157, + "grad_norm": 0.308817595243454, + "learning_rate": 6.262150007253737e-06, + "loss": 0.0008, + "step": 19039 + }, + { + "epoch": 2.0641803989592367, + "grad_norm": 1.9096930027008057, + "learning_rate": 6.26142463368635e-06, + "loss": 0.0047, + "step": 19040 + }, + { + "epoch": 2.0642888117953166, + "grad_norm": 0.07198663055896759, + "learning_rate": 6.260699260118962e-06, + "loss": 0.0002, + "step": 19041 + }, + { + "epoch": 2.064397224631396, + "grad_norm": 0.02812138758599758, + "learning_rate": 6.259973886551575e-06, + "loss": 0.0001, + "step": 19042 + }, + { + "epoch": 2.064505637467476, + "grad_norm": 1.5166094303131104, + "learning_rate": 6.259248512984187e-06, + "loss": 0.0374, + "step": 19043 + }, + { + "epoch": 2.0646140503035557, + "grad_norm": 0.6620254516601562, + "learning_rate": 6.2585231394168e-06, + "loss": 0.001, + "step": 19044 + }, + { + "epoch": 2.0647224631396357, + "grad_norm": 3.521578788757324, + "learning_rate": 6.257797765849413e-06, + "loss": 0.0107, + "step": 19045 + }, + { + "epoch": 2.0648308759757157, + "grad_norm": 0.8372009992599487, + "learning_rate": 6.257072392282025e-06, + "loss": 0.0016, + "step": 19046 + }, + { + "epoch": 2.0649392888117952, + "grad_norm": 0.053017567843198776, + "learning_rate": 6.256347018714638e-06, + "loss": 0.0002, + "step": 19047 + }, + { + "epoch": 2.065047701647875, + "grad_norm": 0.18339544534683228, + "learning_rate": 6.255621645147252e-06, + "loss": 0.0003, + "step": 19048 + }, + { + "epoch": 2.0651561144839548, + "grad_norm": 0.2965924143791199, + "learning_rate": 6.254896271579864e-06, + "loss": 0.0006, + "step": 19049 + }, + { + "epoch": 2.0652645273200347, + "grad_norm": 0.22839957475662231, + "learning_rate": 6.254170898012477e-06, + "loss": 0.0006, + "step": 19050 + }, + { + "epoch": 2.0653729401561143, + "grad_norm": 4.7246928215026855, + "learning_rate": 6.25344552444509e-06, + "loss": 0.0129, + "step": 19051 + }, + { + "epoch": 2.0654813529921943, + "grad_norm": 2.4282448291778564, + "learning_rate": 6.252720150877702e-06, + "loss": 0.0227, + "step": 19052 + }, + { + "epoch": 2.0655897658282742, + "grad_norm": 0.2127101868391037, + "learning_rate": 6.251994777310315e-06, + "loss": 0.0008, + "step": 19053 + }, + { + "epoch": 2.065698178664354, + "grad_norm": 3.9577462673187256, + "learning_rate": 6.251269403742928e-06, + "loss": 0.0181, + "step": 19054 + }, + { + "epoch": 2.0658065915004338, + "grad_norm": 5.86686372756958, + "learning_rate": 6.25054403017554e-06, + "loss": 0.058, + "step": 19055 + }, + { + "epoch": 2.0659150043365133, + "grad_norm": 0.07917342334985733, + "learning_rate": 6.249818656608154e-06, + "loss": 0.0003, + "step": 19056 + }, + { + "epoch": 2.0660234171725933, + "grad_norm": 6.134185791015625, + "learning_rate": 6.249093283040767e-06, + "loss": 0.0336, + "step": 19057 + }, + { + "epoch": 2.066131830008673, + "grad_norm": 2.0670015811920166, + "learning_rate": 6.248367909473379e-06, + "loss": 0.0157, + "step": 19058 + }, + { + "epoch": 2.066240242844753, + "grad_norm": 0.25838351249694824, + "learning_rate": 6.247642535905992e-06, + "loss": 0.0005, + "step": 19059 + }, + { + "epoch": 2.066348655680833, + "grad_norm": 1.882718801498413, + "learning_rate": 6.246917162338605e-06, + "loss": 0.0198, + "step": 19060 + }, + { + "epoch": 2.0664570685169124, + "grad_norm": 0.256164014339447, + "learning_rate": 6.2461917887712174e-06, + "loss": 0.0008, + "step": 19061 + }, + { + "epoch": 2.0665654813529923, + "grad_norm": 5.281441688537598, + "learning_rate": 6.2454664152038304e-06, + "loss": 0.0158, + "step": 19062 + }, + { + "epoch": 2.066673894189072, + "grad_norm": 0.06518898159265518, + "learning_rate": 6.2447410416364426e-06, + "loss": 0.0005, + "step": 19063 + }, + { + "epoch": 2.066782307025152, + "grad_norm": 0.4485059678554535, + "learning_rate": 6.2440156680690556e-06, + "loss": 0.0014, + "step": 19064 + }, + { + "epoch": 2.0668907198612314, + "grad_norm": 0.04752745479345322, + "learning_rate": 6.243290294501669e-06, + "loss": 0.0004, + "step": 19065 + }, + { + "epoch": 2.0669991326973114, + "grad_norm": 0.07249169796705246, + "learning_rate": 6.242564920934282e-06, + "loss": 0.0004, + "step": 19066 + }, + { + "epoch": 2.067107545533391, + "grad_norm": 3.214597463607788, + "learning_rate": 6.2418395473668945e-06, + "loss": 0.0316, + "step": 19067 + }, + { + "epoch": 2.067215958369471, + "grad_norm": 0.010564216412603855, + "learning_rate": 6.2411141737995075e-06, + "loss": 0.0001, + "step": 19068 + }, + { + "epoch": 2.067324371205551, + "grad_norm": 4.952144145965576, + "learning_rate": 6.24038880023212e-06, + "loss": 0.0059, + "step": 19069 + }, + { + "epoch": 2.0674327840416304, + "grad_norm": 0.09444385766983032, + "learning_rate": 6.239663426664733e-06, + "loss": 0.0006, + "step": 19070 + }, + { + "epoch": 2.0675411968777104, + "grad_norm": 0.14937320351600647, + "learning_rate": 6.238938053097346e-06, + "loss": 0.0004, + "step": 19071 + }, + { + "epoch": 2.06764960971379, + "grad_norm": 2.691903591156006, + "learning_rate": 6.238212679529958e-06, + "loss": 0.0344, + "step": 19072 + }, + { + "epoch": 2.06775802254987, + "grad_norm": 0.03354082256555557, + "learning_rate": 6.237487305962571e-06, + "loss": 0.0003, + "step": 19073 + }, + { + "epoch": 2.0678664353859495, + "grad_norm": 2.4423606395721436, + "learning_rate": 6.236761932395185e-06, + "loss": 0.0246, + "step": 19074 + }, + { + "epoch": 2.0679748482220295, + "grad_norm": 0.07968629151582718, + "learning_rate": 6.236036558827797e-06, + "loss": 0.0004, + "step": 19075 + }, + { + "epoch": 2.0680832610581095, + "grad_norm": 0.6118687391281128, + "learning_rate": 6.23531118526041e-06, + "loss": 0.0034, + "step": 19076 + }, + { + "epoch": 2.068191673894189, + "grad_norm": 3.4626007080078125, + "learning_rate": 6.234585811693023e-06, + "loss": 0.0031, + "step": 19077 + }, + { + "epoch": 2.068300086730269, + "grad_norm": 1.224106788635254, + "learning_rate": 6.233860438125635e-06, + "loss": 0.0021, + "step": 19078 + }, + { + "epoch": 2.0684084995663485, + "grad_norm": 2.7156484127044678, + "learning_rate": 6.233135064558248e-06, + "loss": 0.0041, + "step": 19079 + }, + { + "epoch": 2.0685169124024285, + "grad_norm": 1.5424985885620117, + "learning_rate": 6.232409690990861e-06, + "loss": 0.0049, + "step": 19080 + }, + { + "epoch": 2.068625325238508, + "grad_norm": 0.17537756264209747, + "learning_rate": 6.231684317423473e-06, + "loss": 0.0009, + "step": 19081 + }, + { + "epoch": 2.068733738074588, + "grad_norm": 2.866488456726074, + "learning_rate": 6.230958943856087e-06, + "loss": 0.0138, + "step": 19082 + }, + { + "epoch": 2.068842150910668, + "grad_norm": 0.33593499660491943, + "learning_rate": 6.2302335702887e-06, + "loss": 0.0014, + "step": 19083 + }, + { + "epoch": 2.0689505637467476, + "grad_norm": 1.5868899822235107, + "learning_rate": 6.229508196721312e-06, + "loss": 0.0094, + "step": 19084 + }, + { + "epoch": 2.0690589765828276, + "grad_norm": 14.894401550292969, + "learning_rate": 6.228782823153925e-06, + "loss": 0.0062, + "step": 19085 + }, + { + "epoch": 2.069167389418907, + "grad_norm": 0.42844444513320923, + "learning_rate": 6.228057449586538e-06, + "loss": 0.0011, + "step": 19086 + }, + { + "epoch": 2.069275802254987, + "grad_norm": 2.0792734622955322, + "learning_rate": 6.22733207601915e-06, + "loss": 0.0025, + "step": 19087 + }, + { + "epoch": 2.0693842150910666, + "grad_norm": 10.468894958496094, + "learning_rate": 6.226606702451763e-06, + "loss": 0.052, + "step": 19088 + }, + { + "epoch": 2.0694926279271466, + "grad_norm": 0.06332951039075851, + "learning_rate": 6.225881328884375e-06, + "loss": 0.0001, + "step": 19089 + }, + { + "epoch": 2.069601040763226, + "grad_norm": 1.8012526035308838, + "learning_rate": 6.225155955316988e-06, + "loss": 0.0014, + "step": 19090 + }, + { + "epoch": 2.069709453599306, + "grad_norm": 0.20099587738513947, + "learning_rate": 6.224430581749602e-06, + "loss": 0.0005, + "step": 19091 + }, + { + "epoch": 2.069817866435386, + "grad_norm": 0.04981957748532295, + "learning_rate": 6.223705208182215e-06, + "loss": 0.0004, + "step": 19092 + }, + { + "epoch": 2.0699262792714657, + "grad_norm": 0.8214426636695862, + "learning_rate": 6.222979834614827e-06, + "loss": 0.0008, + "step": 19093 + }, + { + "epoch": 2.0700346921075456, + "grad_norm": 0.6389716267585754, + "learning_rate": 6.22225446104744e-06, + "loss": 0.0047, + "step": 19094 + }, + { + "epoch": 2.070143104943625, + "grad_norm": 5.176590442657471, + "learning_rate": 6.221529087480052e-06, + "loss": 0.0066, + "step": 19095 + }, + { + "epoch": 2.070251517779705, + "grad_norm": 0.021762987598776817, + "learning_rate": 6.220803713912665e-06, + "loss": 0.0002, + "step": 19096 + }, + { + "epoch": 2.0703599306157847, + "grad_norm": 0.20906540751457214, + "learning_rate": 6.220078340345278e-06, + "loss": 0.0009, + "step": 19097 + }, + { + "epoch": 2.0704683434518647, + "grad_norm": 0.9119434356689453, + "learning_rate": 6.2193529667778905e-06, + "loss": 0.0029, + "step": 19098 + }, + { + "epoch": 2.0705767562879447, + "grad_norm": 1.3446764945983887, + "learning_rate": 6.218627593210504e-06, + "loss": 0.0229, + "step": 19099 + }, + { + "epoch": 2.0706851691240242, + "grad_norm": 0.25191694498062134, + "learning_rate": 6.217902219643117e-06, + "loss": 0.0003, + "step": 19100 + }, + { + "epoch": 2.070793581960104, + "grad_norm": 1.9650359153747559, + "learning_rate": 6.2171768460757294e-06, + "loss": 0.0023, + "step": 19101 + }, + { + "epoch": 2.0709019947961838, + "grad_norm": 3.2835347652435303, + "learning_rate": 6.2164514725083424e-06, + "loss": 0.0226, + "step": 19102 + }, + { + "epoch": 2.0710104076322637, + "grad_norm": 2.364715099334717, + "learning_rate": 6.215726098940955e-06, + "loss": 0.0135, + "step": 19103 + }, + { + "epoch": 2.0711188204683433, + "grad_norm": 0.1982649862766266, + "learning_rate": 6.2150007253735676e-06, + "loss": 0.0009, + "step": 19104 + }, + { + "epoch": 2.0712272333044233, + "grad_norm": 0.0037525349762290716, + "learning_rate": 6.2142753518061805e-06, + "loss": 0.0, + "step": 19105 + }, + { + "epoch": 2.0713356461405033, + "grad_norm": 1.6050996780395508, + "learning_rate": 6.2135499782387935e-06, + "loss": 0.0054, + "step": 19106 + }, + { + "epoch": 2.071444058976583, + "grad_norm": 0.9564023613929749, + "learning_rate": 6.212824604671406e-06, + "loss": 0.0027, + "step": 19107 + }, + { + "epoch": 2.0715524718126628, + "grad_norm": 0.2027723342180252, + "learning_rate": 6.2120992311040195e-06, + "loss": 0.0008, + "step": 19108 + }, + { + "epoch": 2.0716608846487423, + "grad_norm": 0.10327666997909546, + "learning_rate": 6.2113738575366325e-06, + "loss": 0.0006, + "step": 19109 + }, + { + "epoch": 2.0717692974848223, + "grad_norm": 1.305595874786377, + "learning_rate": 6.210648483969245e-06, + "loss": 0.0013, + "step": 19110 + }, + { + "epoch": 2.071877710320902, + "grad_norm": 2.6974239349365234, + "learning_rate": 6.209923110401858e-06, + "loss": 0.0053, + "step": 19111 + }, + { + "epoch": 2.071986123156982, + "grad_norm": 0.6459739804267883, + "learning_rate": 6.209197736834471e-06, + "loss": 0.0017, + "step": 19112 + }, + { + "epoch": 2.0720945359930614, + "grad_norm": 6.985661506652832, + "learning_rate": 6.208472363267083e-06, + "loss": 0.048, + "step": 19113 + }, + { + "epoch": 2.0722029488291414, + "grad_norm": 4.714519500732422, + "learning_rate": 6.207746989699696e-06, + "loss": 0.0828, + "step": 19114 + }, + { + "epoch": 2.0723113616652213, + "grad_norm": 0.12109322845935822, + "learning_rate": 6.207021616132308e-06, + "loss": 0.0005, + "step": 19115 + }, + { + "epoch": 2.072419774501301, + "grad_norm": 2.1161110401153564, + "learning_rate": 6.206296242564921e-06, + "loss": 0.0051, + "step": 19116 + }, + { + "epoch": 2.072528187337381, + "grad_norm": 0.08944063633680344, + "learning_rate": 6.205570868997535e-06, + "loss": 0.0002, + "step": 19117 + }, + { + "epoch": 2.0726366001734604, + "grad_norm": 1.9470975399017334, + "learning_rate": 6.204845495430148e-06, + "loss": 0.0251, + "step": 19118 + }, + { + "epoch": 2.0727450130095404, + "grad_norm": 6.517491340637207, + "learning_rate": 6.20412012186276e-06, + "loss": 0.0274, + "step": 19119 + }, + { + "epoch": 2.07285342584562, + "grad_norm": 1.5561761856079102, + "learning_rate": 6.203394748295373e-06, + "loss": 0.0015, + "step": 19120 + }, + { + "epoch": 2.0729618386817, + "grad_norm": 1.7334063053131104, + "learning_rate": 6.202669374727985e-06, + "loss": 0.0018, + "step": 19121 + }, + { + "epoch": 2.07307025151778, + "grad_norm": 0.2684393525123596, + "learning_rate": 6.201944001160598e-06, + "loss": 0.0007, + "step": 19122 + }, + { + "epoch": 2.0731786643538594, + "grad_norm": 0.005401456728577614, + "learning_rate": 6.201218627593211e-06, + "loss": 0.0, + "step": 19123 + }, + { + "epoch": 2.0732870771899394, + "grad_norm": 5.166639804840088, + "learning_rate": 6.200493254025823e-06, + "loss": 0.0299, + "step": 19124 + }, + { + "epoch": 2.073395490026019, + "grad_norm": 10.042672157287598, + "learning_rate": 6.199767880458437e-06, + "loss": 0.0227, + "step": 19125 + }, + { + "epoch": 2.073503902862099, + "grad_norm": 0.44078201055526733, + "learning_rate": 6.19904250689105e-06, + "loss": 0.0016, + "step": 19126 + }, + { + "epoch": 2.0736123156981785, + "grad_norm": 4.810307025909424, + "learning_rate": 6.198317133323662e-06, + "loss": 0.0078, + "step": 19127 + }, + { + "epoch": 2.0737207285342585, + "grad_norm": 2.2611303329467773, + "learning_rate": 6.197591759756275e-06, + "loss": 0.0266, + "step": 19128 + }, + { + "epoch": 2.0738291413703385, + "grad_norm": 0.2309209555387497, + "learning_rate": 6.196866386188888e-06, + "loss": 0.001, + "step": 19129 + }, + { + "epoch": 2.073937554206418, + "grad_norm": 0.6584689617156982, + "learning_rate": 6.1961410126215e-06, + "loss": 0.0018, + "step": 19130 + }, + { + "epoch": 2.074045967042498, + "grad_norm": 0.04471209645271301, + "learning_rate": 6.195415639054113e-06, + "loss": 0.0002, + "step": 19131 + }, + { + "epoch": 2.0741543798785775, + "grad_norm": 3.0235743522644043, + "learning_rate": 6.194690265486726e-06, + "loss": 0.0167, + "step": 19132 + }, + { + "epoch": 2.0742627927146575, + "grad_norm": 13.345998764038086, + "learning_rate": 6.193964891919338e-06, + "loss": 0.0478, + "step": 19133 + }, + { + "epoch": 2.074371205550737, + "grad_norm": 0.1367868334054947, + "learning_rate": 6.193239518351952e-06, + "loss": 0.0005, + "step": 19134 + }, + { + "epoch": 2.074479618386817, + "grad_norm": 2.5852127075195312, + "learning_rate": 6.192514144784565e-06, + "loss": 0.0205, + "step": 19135 + }, + { + "epoch": 2.0745880312228966, + "grad_norm": 3.659139394760132, + "learning_rate": 6.191788771217177e-06, + "loss": 0.015, + "step": 19136 + }, + { + "epoch": 2.0746964440589766, + "grad_norm": 0.014022755436599255, + "learning_rate": 6.19106339764979e-06, + "loss": 0.0001, + "step": 19137 + }, + { + "epoch": 2.0748048568950566, + "grad_norm": 1.5876268148422241, + "learning_rate": 6.190338024082403e-06, + "loss": 0.0043, + "step": 19138 + }, + { + "epoch": 2.074913269731136, + "grad_norm": 0.7666358351707458, + "learning_rate": 6.1896126505150154e-06, + "loss": 0.0023, + "step": 19139 + }, + { + "epoch": 2.075021682567216, + "grad_norm": 3.4436769485473633, + "learning_rate": 6.1888872769476284e-06, + "loss": 0.0454, + "step": 19140 + }, + { + "epoch": 2.0751300954032956, + "grad_norm": 12.978992462158203, + "learning_rate": 6.188161903380241e-06, + "loss": 0.004, + "step": 19141 + }, + { + "epoch": 2.0752385082393756, + "grad_norm": 0.14148952066898346, + "learning_rate": 6.1874365298128536e-06, + "loss": 0.0004, + "step": 19142 + }, + { + "epoch": 2.075346921075455, + "grad_norm": 4.580589294433594, + "learning_rate": 6.186711156245467e-06, + "loss": 0.0276, + "step": 19143 + }, + { + "epoch": 2.075455333911535, + "grad_norm": 3.60467529296875, + "learning_rate": 6.18598578267808e-06, + "loss": 0.0038, + "step": 19144 + }, + { + "epoch": 2.075563746747615, + "grad_norm": 3.253394365310669, + "learning_rate": 6.1852604091106925e-06, + "loss": 0.0484, + "step": 19145 + }, + { + "epoch": 2.0756721595836947, + "grad_norm": 0.978121817111969, + "learning_rate": 6.1845350355433055e-06, + "loss": 0.003, + "step": 19146 + }, + { + "epoch": 2.0757805724197746, + "grad_norm": 11.157135963439941, + "learning_rate": 6.183809661975918e-06, + "loss": 0.0222, + "step": 19147 + }, + { + "epoch": 2.075888985255854, + "grad_norm": 0.1408448964357376, + "learning_rate": 6.183084288408531e-06, + "loss": 0.0009, + "step": 19148 + }, + { + "epoch": 2.075997398091934, + "grad_norm": 0.6759411692619324, + "learning_rate": 6.182358914841144e-06, + "loss": 0.0014, + "step": 19149 + }, + { + "epoch": 2.0761058109280137, + "grad_norm": 2.136432409286499, + "learning_rate": 6.181633541273756e-06, + "loss": 0.0318, + "step": 19150 + }, + { + "epoch": 2.0762142237640937, + "grad_norm": 0.02470509335398674, + "learning_rate": 6.18090816770637e-06, + "loss": 0.0001, + "step": 19151 + }, + { + "epoch": 2.0763226366001737, + "grad_norm": 0.06506118923425674, + "learning_rate": 6.180182794138983e-06, + "loss": 0.0002, + "step": 19152 + }, + { + "epoch": 2.0764310494362532, + "grad_norm": 0.006175572983920574, + "learning_rate": 6.179457420571595e-06, + "loss": 0.0, + "step": 19153 + }, + { + "epoch": 2.076539462272333, + "grad_norm": 0.7635856866836548, + "learning_rate": 6.178732047004208e-06, + "loss": 0.0041, + "step": 19154 + }, + { + "epoch": 2.0766478751084128, + "grad_norm": 0.5378313660621643, + "learning_rate": 6.178006673436821e-06, + "loss": 0.0007, + "step": 19155 + }, + { + "epoch": 2.0767562879444927, + "grad_norm": 0.010870282538235188, + "learning_rate": 6.177281299869433e-06, + "loss": 0.0001, + "step": 19156 + }, + { + "epoch": 2.0768647007805723, + "grad_norm": 1.4550918340682983, + "learning_rate": 6.176555926302046e-06, + "loss": 0.0055, + "step": 19157 + }, + { + "epoch": 2.0769731136166523, + "grad_norm": 0.06025285646319389, + "learning_rate": 6.175830552734659e-06, + "loss": 0.0005, + "step": 19158 + }, + { + "epoch": 2.077081526452732, + "grad_norm": 0.2620980143547058, + "learning_rate": 6.175105179167271e-06, + "loss": 0.0004, + "step": 19159 + }, + { + "epoch": 2.077189939288812, + "grad_norm": 0.5401647686958313, + "learning_rate": 6.174379805599885e-06, + "loss": 0.0021, + "step": 19160 + }, + { + "epoch": 2.077298352124892, + "grad_norm": 0.6552072763442993, + "learning_rate": 6.173654432032498e-06, + "loss": 0.0018, + "step": 19161 + }, + { + "epoch": 2.0774067649609713, + "grad_norm": 0.44593629240989685, + "learning_rate": 6.17292905846511e-06, + "loss": 0.0005, + "step": 19162 + }, + { + "epoch": 2.0775151777970513, + "grad_norm": 6.669849872589111, + "learning_rate": 6.172203684897723e-06, + "loss": 0.0334, + "step": 19163 + }, + { + "epoch": 2.077623590633131, + "grad_norm": 9.236757278442383, + "learning_rate": 6.171478311330336e-06, + "loss": 0.0069, + "step": 19164 + }, + { + "epoch": 2.077732003469211, + "grad_norm": 0.29841265082359314, + "learning_rate": 6.170752937762948e-06, + "loss": 0.001, + "step": 19165 + }, + { + "epoch": 2.0778404163052904, + "grad_norm": 5.693528175354004, + "learning_rate": 6.170027564195561e-06, + "loss": 0.0225, + "step": 19166 + }, + { + "epoch": 2.0779488291413704, + "grad_norm": 0.04493898153305054, + "learning_rate": 6.169302190628173e-06, + "loss": 0.0003, + "step": 19167 + }, + { + "epoch": 2.0780572419774503, + "grad_norm": 0.12051566690206528, + "learning_rate": 6.168576817060786e-06, + "loss": 0.0006, + "step": 19168 + }, + { + "epoch": 2.07816565481353, + "grad_norm": 0.3756926357746124, + "learning_rate": 6.1678514434934e-06, + "loss": 0.0003, + "step": 19169 + }, + { + "epoch": 2.07827406764961, + "grad_norm": 2.57643461227417, + "learning_rate": 6.167126069926013e-06, + "loss": 0.0056, + "step": 19170 + }, + { + "epoch": 2.0783824804856894, + "grad_norm": 2.2418055534362793, + "learning_rate": 6.166400696358625e-06, + "loss": 0.0163, + "step": 19171 + }, + { + "epoch": 2.0784908933217694, + "grad_norm": 4.897542476654053, + "learning_rate": 6.165675322791238e-06, + "loss": 0.0041, + "step": 19172 + }, + { + "epoch": 2.078599306157849, + "grad_norm": 3.56813907623291, + "learning_rate": 6.16494994922385e-06, + "loss": 0.0142, + "step": 19173 + }, + { + "epoch": 2.078707718993929, + "grad_norm": 0.047466766089200974, + "learning_rate": 6.164224575656463e-06, + "loss": 0.0003, + "step": 19174 + }, + { + "epoch": 2.0788161318300085, + "grad_norm": 5.476528644561768, + "learning_rate": 6.163499202089076e-06, + "loss": 0.0169, + "step": 19175 + }, + { + "epoch": 2.0789245446660884, + "grad_norm": 0.01296007726341486, + "learning_rate": 6.1627738285216885e-06, + "loss": 0.0001, + "step": 19176 + }, + { + "epoch": 2.0790329575021684, + "grad_norm": 2.1924688816070557, + "learning_rate": 6.162048454954302e-06, + "loss": 0.008, + "step": 19177 + }, + { + "epoch": 2.079141370338248, + "grad_norm": 0.04690363630652428, + "learning_rate": 6.161323081386915e-06, + "loss": 0.0003, + "step": 19178 + }, + { + "epoch": 2.079249783174328, + "grad_norm": 0.8855699300765991, + "learning_rate": 6.1605977078195274e-06, + "loss": 0.0018, + "step": 19179 + }, + { + "epoch": 2.0793581960104075, + "grad_norm": 0.17314748466014862, + "learning_rate": 6.1598723342521404e-06, + "loss": 0.0005, + "step": 19180 + }, + { + "epoch": 2.0794666088464875, + "grad_norm": 0.79498291015625, + "learning_rate": 6.159146960684753e-06, + "loss": 0.0019, + "step": 19181 + }, + { + "epoch": 2.079575021682567, + "grad_norm": 3.2158749103546143, + "learning_rate": 6.1584215871173656e-06, + "loss": 0.0366, + "step": 19182 + }, + { + "epoch": 2.079683434518647, + "grad_norm": 0.7219035029411316, + "learning_rate": 6.1576962135499786e-06, + "loss": 0.0017, + "step": 19183 + }, + { + "epoch": 2.079791847354727, + "grad_norm": 3.620344400405884, + "learning_rate": 6.1569708399825915e-06, + "loss": 0.006, + "step": 19184 + }, + { + "epoch": 2.0799002601908065, + "grad_norm": 0.04464882239699364, + "learning_rate": 6.156245466415204e-06, + "loss": 0.0002, + "step": 19185 + }, + { + "epoch": 2.0800086730268865, + "grad_norm": 0.010242905467748642, + "learning_rate": 6.1555200928478175e-06, + "loss": 0.0001, + "step": 19186 + }, + { + "epoch": 2.080117085862966, + "grad_norm": 0.24012553691864014, + "learning_rate": 6.1547947192804305e-06, + "loss": 0.0005, + "step": 19187 + }, + { + "epoch": 2.080225498699046, + "grad_norm": 2.6577088832855225, + "learning_rate": 6.154069345713043e-06, + "loss": 0.0286, + "step": 19188 + }, + { + "epoch": 2.0803339115351256, + "grad_norm": 2.8363990783691406, + "learning_rate": 6.153343972145656e-06, + "loss": 0.0023, + "step": 19189 + }, + { + "epoch": 2.0804423243712056, + "grad_norm": 0.6052420139312744, + "learning_rate": 6.152618598578269e-06, + "loss": 0.0013, + "step": 19190 + }, + { + "epoch": 2.080550737207285, + "grad_norm": 0.17806850373744965, + "learning_rate": 6.151893225010881e-06, + "loss": 0.0007, + "step": 19191 + }, + { + "epoch": 2.080659150043365, + "grad_norm": 1.2225998640060425, + "learning_rate": 6.151167851443494e-06, + "loss": 0.004, + "step": 19192 + }, + { + "epoch": 2.080767562879445, + "grad_norm": 0.02946256846189499, + "learning_rate": 6.150442477876106e-06, + "loss": 0.0001, + "step": 19193 + }, + { + "epoch": 2.0808759757155246, + "grad_norm": 9.74030590057373, + "learning_rate": 6.14971710430872e-06, + "loss": 0.0147, + "step": 19194 + }, + { + "epoch": 2.0809843885516046, + "grad_norm": 0.5366898775100708, + "learning_rate": 6.148991730741333e-06, + "loss": 0.0018, + "step": 19195 + }, + { + "epoch": 2.081092801387684, + "grad_norm": 0.16383324563503265, + "learning_rate": 6.148266357173946e-06, + "loss": 0.0006, + "step": 19196 + }, + { + "epoch": 2.081201214223764, + "grad_norm": 0.02769354358315468, + "learning_rate": 6.147540983606558e-06, + "loss": 0.0001, + "step": 19197 + }, + { + "epoch": 2.0813096270598437, + "grad_norm": 0.023392442613840103, + "learning_rate": 6.146815610039171e-06, + "loss": 0.0001, + "step": 19198 + }, + { + "epoch": 2.0814180398959237, + "grad_norm": 0.05902770906686783, + "learning_rate": 6.146090236471783e-06, + "loss": 0.0005, + "step": 19199 + }, + { + "epoch": 2.0815264527320037, + "grad_norm": 3.369704246520996, + "learning_rate": 6.145364862904396e-06, + "loss": 0.0207, + "step": 19200 + }, + { + "epoch": 2.081634865568083, + "grad_norm": 3.6084561347961426, + "learning_rate": 6.144639489337009e-06, + "loss": 0.0142, + "step": 19201 + }, + { + "epoch": 2.081743278404163, + "grad_norm": 5.0675201416015625, + "learning_rate": 6.143914115769621e-06, + "loss": 0.0178, + "step": 19202 + }, + { + "epoch": 2.0818516912402427, + "grad_norm": 1.1692694425582886, + "learning_rate": 6.143188742202235e-06, + "loss": 0.0033, + "step": 19203 + }, + { + "epoch": 2.0819601040763227, + "grad_norm": 0.9250870943069458, + "learning_rate": 6.142463368634848e-06, + "loss": 0.0006, + "step": 19204 + }, + { + "epoch": 2.0820685169124022, + "grad_norm": 1.121946930885315, + "learning_rate": 6.14173799506746e-06, + "loss": 0.0028, + "step": 19205 + }, + { + "epoch": 2.0821769297484822, + "grad_norm": 7.940593719482422, + "learning_rate": 6.141012621500073e-06, + "loss": 0.0196, + "step": 19206 + }, + { + "epoch": 2.082285342584562, + "grad_norm": 0.11110678315162659, + "learning_rate": 6.140287247932686e-06, + "loss": 0.0006, + "step": 19207 + }, + { + "epoch": 2.0823937554206418, + "grad_norm": 2.8534762859344482, + "learning_rate": 6.139561874365298e-06, + "loss": 0.0078, + "step": 19208 + }, + { + "epoch": 2.0825021682567217, + "grad_norm": 6.327472686767578, + "learning_rate": 6.138836500797911e-06, + "loss": 0.009, + "step": 19209 + }, + { + "epoch": 2.0826105810928013, + "grad_norm": 13.324808120727539, + "learning_rate": 6.138111127230524e-06, + "loss": 0.0099, + "step": 19210 + }, + { + "epoch": 2.0827189939288813, + "grad_norm": 3.584418296813965, + "learning_rate": 6.137385753663136e-06, + "loss": 0.0188, + "step": 19211 + }, + { + "epoch": 2.082827406764961, + "grad_norm": 0.08054833114147186, + "learning_rate": 6.13666038009575e-06, + "loss": 0.0002, + "step": 19212 + }, + { + "epoch": 2.082935819601041, + "grad_norm": 0.43989965319633484, + "learning_rate": 6.135935006528363e-06, + "loss": 0.0029, + "step": 19213 + }, + { + "epoch": 2.0830442324371203, + "grad_norm": 1.2001495361328125, + "learning_rate": 6.135209632960975e-06, + "loss": 0.0128, + "step": 19214 + }, + { + "epoch": 2.0831526452732003, + "grad_norm": 0.927834153175354, + "learning_rate": 6.134484259393588e-06, + "loss": 0.0006, + "step": 19215 + }, + { + "epoch": 2.0832610581092803, + "grad_norm": 2.309269905090332, + "learning_rate": 6.133758885826201e-06, + "loss": 0.0076, + "step": 19216 + }, + { + "epoch": 2.08336947094536, + "grad_norm": 0.014893104322254658, + "learning_rate": 6.1330335122588135e-06, + "loss": 0.0001, + "step": 19217 + }, + { + "epoch": 2.08347788378144, + "grad_norm": 2.262019634246826, + "learning_rate": 6.1323081386914264e-06, + "loss": 0.0037, + "step": 19218 + }, + { + "epoch": 2.0835862966175194, + "grad_norm": 0.1116771474480629, + "learning_rate": 6.131582765124039e-06, + "loss": 0.0003, + "step": 19219 + }, + { + "epoch": 2.0836947094535994, + "grad_norm": 0.25054988265037537, + "learning_rate": 6.1308573915566524e-06, + "loss": 0.0006, + "step": 19220 + }, + { + "epoch": 2.083803122289679, + "grad_norm": 0.9807422757148743, + "learning_rate": 6.130132017989265e-06, + "loss": 0.004, + "step": 19221 + }, + { + "epoch": 2.083911535125759, + "grad_norm": 4.393210411071777, + "learning_rate": 6.129406644421878e-06, + "loss": 0.096, + "step": 19222 + }, + { + "epoch": 2.084019947961839, + "grad_norm": 6.317525386810303, + "learning_rate": 6.1286812708544905e-06, + "loss": 0.0208, + "step": 19223 + }, + { + "epoch": 2.0841283607979184, + "grad_norm": 0.031336709856987, + "learning_rate": 6.1279558972871035e-06, + "loss": 0.0002, + "step": 19224 + }, + { + "epoch": 2.0842367736339984, + "grad_norm": 0.10349750518798828, + "learning_rate": 6.127230523719716e-06, + "loss": 0.0006, + "step": 19225 + }, + { + "epoch": 2.084345186470078, + "grad_norm": 0.9262800812721252, + "learning_rate": 6.126505150152329e-06, + "loss": 0.0036, + "step": 19226 + }, + { + "epoch": 2.084453599306158, + "grad_norm": 6.150193214416504, + "learning_rate": 6.125779776584942e-06, + "loss": 0.0036, + "step": 19227 + }, + { + "epoch": 2.0845620121422375, + "grad_norm": 2.2526309490203857, + "learning_rate": 6.125054403017554e-06, + "loss": 0.0046, + "step": 19228 + }, + { + "epoch": 2.0846704249783174, + "grad_norm": 0.43579256534576416, + "learning_rate": 6.124329029450168e-06, + "loss": 0.0013, + "step": 19229 + }, + { + "epoch": 2.0847788378143974, + "grad_norm": 0.0908232256770134, + "learning_rate": 6.123603655882781e-06, + "loss": 0.0002, + "step": 19230 + }, + { + "epoch": 2.084887250650477, + "grad_norm": 0.13596895337104797, + "learning_rate": 6.122878282315393e-06, + "loss": 0.0006, + "step": 19231 + }, + { + "epoch": 2.084995663486557, + "grad_norm": 0.24466052651405334, + "learning_rate": 6.122152908748006e-06, + "loss": 0.0005, + "step": 19232 + }, + { + "epoch": 2.0851040763226365, + "grad_norm": 4.613708972930908, + "learning_rate": 6.121427535180619e-06, + "loss": 0.0553, + "step": 19233 + }, + { + "epoch": 2.0852124891587165, + "grad_norm": 1.763842225074768, + "learning_rate": 6.120702161613231e-06, + "loss": 0.0031, + "step": 19234 + }, + { + "epoch": 2.085320901994796, + "grad_norm": 2.2642340660095215, + "learning_rate": 6.119976788045844e-06, + "loss": 0.0507, + "step": 19235 + }, + { + "epoch": 2.085429314830876, + "grad_norm": 0.011973864398896694, + "learning_rate": 6.119251414478457e-06, + "loss": 0.0001, + "step": 19236 + }, + { + "epoch": 2.0855377276669556, + "grad_norm": 3.004464626312256, + "learning_rate": 6.118526040911069e-06, + "loss": 0.0326, + "step": 19237 + }, + { + "epoch": 2.0856461405030355, + "grad_norm": 0.009244700893759727, + "learning_rate": 6.117800667343683e-06, + "loss": 0.0, + "step": 19238 + }, + { + "epoch": 2.0857545533391155, + "grad_norm": 0.09964755177497864, + "learning_rate": 6.117075293776296e-06, + "loss": 0.0003, + "step": 19239 + }, + { + "epoch": 2.085862966175195, + "grad_norm": 0.37071067094802856, + "learning_rate": 6.116349920208908e-06, + "loss": 0.0011, + "step": 19240 + }, + { + "epoch": 2.085971379011275, + "grad_norm": 0.014697132632136345, + "learning_rate": 6.115624546641521e-06, + "loss": 0.0001, + "step": 19241 + }, + { + "epoch": 2.0860797918473546, + "grad_norm": 0.802575945854187, + "learning_rate": 6.114899173074134e-06, + "loss": 0.0025, + "step": 19242 + }, + { + "epoch": 2.0861882046834346, + "grad_norm": 0.05781499296426773, + "learning_rate": 6.114173799506746e-06, + "loss": 0.0004, + "step": 19243 + }, + { + "epoch": 2.086296617519514, + "grad_norm": 11.949151039123535, + "learning_rate": 6.113448425939359e-06, + "loss": 0.0475, + "step": 19244 + }, + { + "epoch": 2.086405030355594, + "grad_norm": 0.6857742667198181, + "learning_rate": 6.112723052371971e-06, + "loss": 0.0015, + "step": 19245 + }, + { + "epoch": 2.086513443191674, + "grad_norm": 0.5499278903007507, + "learning_rate": 6.111997678804585e-06, + "loss": 0.0007, + "step": 19246 + }, + { + "epoch": 2.0866218560277536, + "grad_norm": 1.218713641166687, + "learning_rate": 6.111272305237198e-06, + "loss": 0.0508, + "step": 19247 + }, + { + "epoch": 2.0867302688638336, + "grad_norm": 0.34133151173591614, + "learning_rate": 6.110546931669811e-06, + "loss": 0.0007, + "step": 19248 + }, + { + "epoch": 2.086838681699913, + "grad_norm": 0.8091108798980713, + "learning_rate": 6.109821558102423e-06, + "loss": 0.001, + "step": 19249 + }, + { + "epoch": 2.086947094535993, + "grad_norm": 3.4850070476531982, + "learning_rate": 6.109096184535036e-06, + "loss": 0.0157, + "step": 19250 + }, + { + "epoch": 2.0870555073720727, + "grad_norm": 2.861135721206665, + "learning_rate": 6.108370810967648e-06, + "loss": 0.0344, + "step": 19251 + }, + { + "epoch": 2.0871639202081527, + "grad_norm": 0.4674265682697296, + "learning_rate": 6.107645437400261e-06, + "loss": 0.0022, + "step": 19252 + }, + { + "epoch": 2.0872723330442327, + "grad_norm": 0.5559673309326172, + "learning_rate": 6.106920063832874e-06, + "loss": 0.0014, + "step": 19253 + }, + { + "epoch": 2.087380745880312, + "grad_norm": 0.02040271833539009, + "learning_rate": 6.1061946902654865e-06, + "loss": 0.0001, + "step": 19254 + }, + { + "epoch": 2.087489158716392, + "grad_norm": 0.09754348546266556, + "learning_rate": 6.1054693166981e-06, + "loss": 0.0004, + "step": 19255 + }, + { + "epoch": 2.0875975715524717, + "grad_norm": 1.9597253799438477, + "learning_rate": 6.104743943130713e-06, + "loss": 0.015, + "step": 19256 + }, + { + "epoch": 2.0877059843885517, + "grad_norm": 1.742524266242981, + "learning_rate": 6.1040185695633254e-06, + "loss": 0.0035, + "step": 19257 + }, + { + "epoch": 2.0878143972246312, + "grad_norm": 1.5427697896957397, + "learning_rate": 6.1032931959959384e-06, + "loss": 0.0035, + "step": 19258 + }, + { + "epoch": 2.0879228100607112, + "grad_norm": 4.310063362121582, + "learning_rate": 6.1025678224285514e-06, + "loss": 0.0588, + "step": 19259 + }, + { + "epoch": 2.0880312228967908, + "grad_norm": 0.893061637878418, + "learning_rate": 6.1018424488611636e-06, + "loss": 0.0032, + "step": 19260 + }, + { + "epoch": 2.0881396357328708, + "grad_norm": 0.4175538122653961, + "learning_rate": 6.1011170752937766e-06, + "loss": 0.001, + "step": 19261 + }, + { + "epoch": 2.0882480485689507, + "grad_norm": 4.674749374389648, + "learning_rate": 6.1003917017263896e-06, + "loss": 0.0097, + "step": 19262 + }, + { + "epoch": 2.0883564614050303, + "grad_norm": 1.765085220336914, + "learning_rate": 6.099666328159002e-06, + "loss": 0.0021, + "step": 19263 + }, + { + "epoch": 2.0884648742411103, + "grad_norm": 0.5125213861465454, + "learning_rate": 6.0989409545916155e-06, + "loss": 0.0018, + "step": 19264 + }, + { + "epoch": 2.08857328707719, + "grad_norm": 0.823847234249115, + "learning_rate": 6.0982155810242285e-06, + "loss": 0.0015, + "step": 19265 + }, + { + "epoch": 2.08868169991327, + "grad_norm": 1.2091057300567627, + "learning_rate": 6.097490207456841e-06, + "loss": 0.0055, + "step": 19266 + }, + { + "epoch": 2.0887901127493493, + "grad_norm": 0.8047729730606079, + "learning_rate": 6.096764833889454e-06, + "loss": 0.0037, + "step": 19267 + }, + { + "epoch": 2.0888985255854293, + "grad_norm": 0.24400155246257782, + "learning_rate": 6.096039460322067e-06, + "loss": 0.0015, + "step": 19268 + }, + { + "epoch": 2.0890069384215093, + "grad_norm": 0.041903216391801834, + "learning_rate": 6.095314086754679e-06, + "loss": 0.0003, + "step": 19269 + }, + { + "epoch": 2.089115351257589, + "grad_norm": 0.13157528638839722, + "learning_rate": 6.094588713187292e-06, + "loss": 0.0004, + "step": 19270 + }, + { + "epoch": 2.089223764093669, + "grad_norm": 0.08663345873355865, + "learning_rate": 6.093863339619904e-06, + "loss": 0.0003, + "step": 19271 + }, + { + "epoch": 2.0893321769297484, + "grad_norm": 5.696247577667236, + "learning_rate": 6.093137966052518e-06, + "loss": 0.0478, + "step": 19272 + }, + { + "epoch": 2.0894405897658284, + "grad_norm": 0.6885361075401306, + "learning_rate": 6.092412592485131e-06, + "loss": 0.0018, + "step": 19273 + }, + { + "epoch": 2.089549002601908, + "grad_norm": 0.6801154017448425, + "learning_rate": 6.091687218917744e-06, + "loss": 0.0024, + "step": 19274 + }, + { + "epoch": 2.089657415437988, + "grad_norm": 3.053947925567627, + "learning_rate": 6.090961845350356e-06, + "loss": 0.0421, + "step": 19275 + }, + { + "epoch": 2.089765828274068, + "grad_norm": 14.006684303283691, + "learning_rate": 6.090236471782969e-06, + "loss": 0.0113, + "step": 19276 + }, + { + "epoch": 2.0898742411101474, + "grad_norm": 2.141289234161377, + "learning_rate": 6.089511098215581e-06, + "loss": 0.0102, + "step": 19277 + }, + { + "epoch": 2.0899826539462274, + "grad_norm": 0.5064996480941772, + "learning_rate": 6.088785724648194e-06, + "loss": 0.0008, + "step": 19278 + }, + { + "epoch": 2.090091066782307, + "grad_norm": 0.03185025975108147, + "learning_rate": 6.088060351080807e-06, + "loss": 0.0002, + "step": 19279 + }, + { + "epoch": 2.090199479618387, + "grad_norm": 2.258162260055542, + "learning_rate": 6.087334977513419e-06, + "loss": 0.0087, + "step": 19280 + }, + { + "epoch": 2.0903078924544665, + "grad_norm": 0.039085905998945236, + "learning_rate": 6.086609603946033e-06, + "loss": 0.0003, + "step": 19281 + }, + { + "epoch": 2.0904163052905465, + "grad_norm": 0.04870269075036049, + "learning_rate": 6.085884230378646e-06, + "loss": 0.0002, + "step": 19282 + }, + { + "epoch": 2.090524718126626, + "grad_norm": 0.6363827586174011, + "learning_rate": 6.085158856811258e-06, + "loss": 0.001, + "step": 19283 + }, + { + "epoch": 2.090633130962706, + "grad_norm": 0.7181784510612488, + "learning_rate": 6.084433483243871e-06, + "loss": 0.0025, + "step": 19284 + }, + { + "epoch": 2.090741543798786, + "grad_norm": 10.196160316467285, + "learning_rate": 6.083708109676484e-06, + "loss": 0.0112, + "step": 19285 + }, + { + "epoch": 2.0908499566348655, + "grad_norm": 0.27029600739479065, + "learning_rate": 6.082982736109096e-06, + "loss": 0.0006, + "step": 19286 + }, + { + "epoch": 2.0909583694709455, + "grad_norm": 0.3164670169353485, + "learning_rate": 6.082257362541709e-06, + "loss": 0.0004, + "step": 19287 + }, + { + "epoch": 2.091066782307025, + "grad_norm": 0.21154145896434784, + "learning_rate": 6.081531988974322e-06, + "loss": 0.0005, + "step": 19288 + }, + { + "epoch": 2.091175195143105, + "grad_norm": 0.07956738024950027, + "learning_rate": 6.080806615406935e-06, + "loss": 0.0003, + "step": 19289 + }, + { + "epoch": 2.0912836079791846, + "grad_norm": 0.007008002605289221, + "learning_rate": 6.080081241839548e-06, + "loss": 0.0001, + "step": 19290 + }, + { + "epoch": 2.0913920208152645, + "grad_norm": 1.170174479484558, + "learning_rate": 6.079355868272161e-06, + "loss": 0.0009, + "step": 19291 + }, + { + "epoch": 2.0915004336513445, + "grad_norm": 0.20974798500537872, + "learning_rate": 6.078630494704773e-06, + "loss": 0.0004, + "step": 19292 + }, + { + "epoch": 2.091608846487424, + "grad_norm": 2.068467617034912, + "learning_rate": 6.077905121137386e-06, + "loss": 0.003, + "step": 19293 + }, + { + "epoch": 2.091717259323504, + "grad_norm": 5.732614517211914, + "learning_rate": 6.077179747569999e-06, + "loss": 0.0125, + "step": 19294 + }, + { + "epoch": 2.0918256721595836, + "grad_norm": 3.211466073989868, + "learning_rate": 6.0764543740026115e-06, + "loss": 0.0112, + "step": 19295 + }, + { + "epoch": 2.0919340849956636, + "grad_norm": 0.2507769763469696, + "learning_rate": 6.0757290004352245e-06, + "loss": 0.001, + "step": 19296 + }, + { + "epoch": 2.092042497831743, + "grad_norm": 0.18577000498771667, + "learning_rate": 6.075003626867837e-06, + "loss": 0.0009, + "step": 19297 + }, + { + "epoch": 2.092150910667823, + "grad_norm": 3.056701421737671, + "learning_rate": 6.0742782533004504e-06, + "loss": 0.004, + "step": 19298 + }, + { + "epoch": 2.092259323503903, + "grad_norm": 0.127404123544693, + "learning_rate": 6.073552879733063e-06, + "loss": 0.0005, + "step": 19299 + }, + { + "epoch": 2.0923677363399826, + "grad_norm": 0.6700406074523926, + "learning_rate": 6.072827506165676e-06, + "loss": 0.0037, + "step": 19300 + }, + { + "epoch": 2.0924761491760626, + "grad_norm": 0.3988882899284363, + "learning_rate": 6.0721021325982886e-06, + "loss": 0.0012, + "step": 19301 + }, + { + "epoch": 2.092584562012142, + "grad_norm": 0.4665353298187256, + "learning_rate": 6.0713767590309015e-06, + "loss": 0.003, + "step": 19302 + }, + { + "epoch": 2.092692974848222, + "grad_norm": 14.077842712402344, + "learning_rate": 6.070651385463514e-06, + "loss": 0.0262, + "step": 19303 + }, + { + "epoch": 2.0928013876843017, + "grad_norm": 6.295045852661133, + "learning_rate": 6.069926011896127e-06, + "loss": 0.0587, + "step": 19304 + }, + { + "epoch": 2.0929098005203817, + "grad_norm": 1.984465479850769, + "learning_rate": 6.06920063832874e-06, + "loss": 0.0052, + "step": 19305 + }, + { + "epoch": 2.093018213356461, + "grad_norm": 0.022643037140369415, + "learning_rate": 6.068475264761352e-06, + "loss": 0.0002, + "step": 19306 + }, + { + "epoch": 2.093126626192541, + "grad_norm": 0.3241461217403412, + "learning_rate": 6.067749891193966e-06, + "loss": 0.0006, + "step": 19307 + }, + { + "epoch": 2.093235039028621, + "grad_norm": 1.22286856174469, + "learning_rate": 6.067024517626579e-06, + "loss": 0.0036, + "step": 19308 + }, + { + "epoch": 2.0933434518647007, + "grad_norm": 3.964738607406616, + "learning_rate": 6.066299144059191e-06, + "loss": 0.0021, + "step": 19309 + }, + { + "epoch": 2.0934518647007807, + "grad_norm": 0.0513804629445076, + "learning_rate": 6.065573770491804e-06, + "loss": 0.0002, + "step": 19310 + }, + { + "epoch": 2.0935602775368602, + "grad_norm": 0.09195239841938019, + "learning_rate": 6.064848396924417e-06, + "loss": 0.0002, + "step": 19311 + }, + { + "epoch": 2.0936686903729402, + "grad_norm": 0.08050008118152618, + "learning_rate": 6.064123023357029e-06, + "loss": 0.0003, + "step": 19312 + }, + { + "epoch": 2.0937771032090198, + "grad_norm": 8.808756828308105, + "learning_rate": 6.063397649789642e-06, + "loss": 0.0517, + "step": 19313 + }, + { + "epoch": 2.0938855160450998, + "grad_norm": 4.611393451690674, + "learning_rate": 6.062672276222255e-06, + "loss": 0.049, + "step": 19314 + }, + { + "epoch": 2.0939939288811797, + "grad_norm": 0.5188845992088318, + "learning_rate": 6.061946902654868e-06, + "loss": 0.003, + "step": 19315 + }, + { + "epoch": 2.0941023417172593, + "grad_norm": 2.747842311859131, + "learning_rate": 6.061221529087481e-06, + "loss": 0.0256, + "step": 19316 + }, + { + "epoch": 2.0942107545533393, + "grad_norm": 1.8598155975341797, + "learning_rate": 6.060496155520094e-06, + "loss": 0.0598, + "step": 19317 + }, + { + "epoch": 2.094319167389419, + "grad_norm": 1.1646839380264282, + "learning_rate": 6.059770781952706e-06, + "loss": 0.0039, + "step": 19318 + }, + { + "epoch": 2.094427580225499, + "grad_norm": 2.0476200580596924, + "learning_rate": 6.059045408385319e-06, + "loss": 0.0311, + "step": 19319 + }, + { + "epoch": 2.0945359930615783, + "grad_norm": 1.903738260269165, + "learning_rate": 6.058320034817932e-06, + "loss": 0.0412, + "step": 19320 + }, + { + "epoch": 2.0946444058976583, + "grad_norm": 0.9575248956680298, + "learning_rate": 6.057594661250544e-06, + "loss": 0.002, + "step": 19321 + }, + { + "epoch": 2.094752818733738, + "grad_norm": 4.992489814758301, + "learning_rate": 6.056869287683157e-06, + "loss": 0.0227, + "step": 19322 + }, + { + "epoch": 2.094861231569818, + "grad_norm": 0.013106731697916985, + "learning_rate": 6.056143914115769e-06, + "loss": 0.0001, + "step": 19323 + }, + { + "epoch": 2.094969644405898, + "grad_norm": 0.15951202809810638, + "learning_rate": 6.055418540548383e-06, + "loss": 0.0008, + "step": 19324 + }, + { + "epoch": 2.0950780572419774, + "grad_norm": 5.756031513214111, + "learning_rate": 6.054693166980996e-06, + "loss": 0.007, + "step": 19325 + }, + { + "epoch": 2.0951864700780574, + "grad_norm": 0.005080994684249163, + "learning_rate": 6.053967793413609e-06, + "loss": 0.0, + "step": 19326 + }, + { + "epoch": 2.095294882914137, + "grad_norm": 1.9680194854736328, + "learning_rate": 6.053242419846221e-06, + "loss": 0.0113, + "step": 19327 + }, + { + "epoch": 2.095403295750217, + "grad_norm": 2.713099718093872, + "learning_rate": 6.052517046278834e-06, + "loss": 0.0041, + "step": 19328 + }, + { + "epoch": 2.0955117085862964, + "grad_norm": 0.04805809631943703, + "learning_rate": 6.051791672711446e-06, + "loss": 0.0002, + "step": 19329 + }, + { + "epoch": 2.0956201214223764, + "grad_norm": 1.522087812423706, + "learning_rate": 6.051066299144059e-06, + "loss": 0.01, + "step": 19330 + }, + { + "epoch": 2.0957285342584564, + "grad_norm": 0.32750412821769714, + "learning_rate": 6.050340925576672e-06, + "loss": 0.0013, + "step": 19331 + }, + { + "epoch": 2.095836947094536, + "grad_norm": 0.6630679368972778, + "learning_rate": 6.0496155520092845e-06, + "loss": 0.0024, + "step": 19332 + }, + { + "epoch": 2.095945359930616, + "grad_norm": 5.705472469329834, + "learning_rate": 6.048890178441898e-06, + "loss": 0.0054, + "step": 19333 + }, + { + "epoch": 2.0960537727666955, + "grad_norm": 9.136138916015625, + "learning_rate": 6.048164804874511e-06, + "loss": 0.0084, + "step": 19334 + }, + { + "epoch": 2.0961621856027755, + "grad_norm": 1.1546299457550049, + "learning_rate": 6.0474394313071235e-06, + "loss": 0.0015, + "step": 19335 + }, + { + "epoch": 2.096270598438855, + "grad_norm": 0.07897788286209106, + "learning_rate": 6.0467140577397364e-06, + "loss": 0.0005, + "step": 19336 + }, + { + "epoch": 2.096379011274935, + "grad_norm": 0.006263121962547302, + "learning_rate": 6.0459886841723494e-06, + "loss": 0.0, + "step": 19337 + }, + { + "epoch": 2.0964874241110145, + "grad_norm": 1.7321404218673706, + "learning_rate": 6.045263310604962e-06, + "loss": 0.008, + "step": 19338 + }, + { + "epoch": 2.0965958369470945, + "grad_norm": 0.04233039915561676, + "learning_rate": 6.0445379370375746e-06, + "loss": 0.0001, + "step": 19339 + }, + { + "epoch": 2.0967042497831745, + "grad_norm": 0.2796729505062103, + "learning_rate": 6.0438125634701876e-06, + "loss": 0.0015, + "step": 19340 + }, + { + "epoch": 2.096812662619254, + "grad_norm": 2.592383861541748, + "learning_rate": 6.0430871899028005e-06, + "loss": 0.0121, + "step": 19341 + }, + { + "epoch": 2.096921075455334, + "grad_norm": 0.29324114322662354, + "learning_rate": 6.0423618163354135e-06, + "loss": 0.0006, + "step": 19342 + }, + { + "epoch": 2.0970294882914136, + "grad_norm": 3.089217185974121, + "learning_rate": 6.0416364427680265e-06, + "loss": 0.0404, + "step": 19343 + }, + { + "epoch": 2.0971379011274935, + "grad_norm": 0.039816755801439285, + "learning_rate": 6.040911069200639e-06, + "loss": 0.0002, + "step": 19344 + }, + { + "epoch": 2.097246313963573, + "grad_norm": 1.5089136362075806, + "learning_rate": 6.040185695633252e-06, + "loss": 0.042, + "step": 19345 + }, + { + "epoch": 2.097354726799653, + "grad_norm": 0.6980702877044678, + "learning_rate": 6.039460322065865e-06, + "loss": 0.0099, + "step": 19346 + }, + { + "epoch": 2.097463139635733, + "grad_norm": 0.014885188080370426, + "learning_rate": 6.038734948498477e-06, + "loss": 0.0001, + "step": 19347 + }, + { + "epoch": 2.0975715524718126, + "grad_norm": 1.1186933517456055, + "learning_rate": 6.03800957493109e-06, + "loss": 0.0045, + "step": 19348 + }, + { + "epoch": 2.0976799653078926, + "grad_norm": 0.07119004428386688, + "learning_rate": 6.037284201363702e-06, + "loss": 0.0004, + "step": 19349 + }, + { + "epoch": 2.097788378143972, + "grad_norm": 0.15881390869617462, + "learning_rate": 6.036558827796316e-06, + "loss": 0.0001, + "step": 19350 + }, + { + "epoch": 2.097896790980052, + "grad_norm": 0.36556118726730347, + "learning_rate": 6.035833454228929e-06, + "loss": 0.0009, + "step": 19351 + }, + { + "epoch": 2.0980052038161316, + "grad_norm": 8.8663330078125, + "learning_rate": 6.035108080661542e-06, + "loss": 0.0179, + "step": 19352 + }, + { + "epoch": 2.0981136166522116, + "grad_norm": 5.08295202255249, + "learning_rate": 6.034382707094154e-06, + "loss": 0.0082, + "step": 19353 + }, + { + "epoch": 2.0982220294882916, + "grad_norm": 0.006478389725089073, + "learning_rate": 6.033657333526767e-06, + "loss": 0.0001, + "step": 19354 + }, + { + "epoch": 2.098330442324371, + "grad_norm": 0.12072692811489105, + "learning_rate": 6.032931959959379e-06, + "loss": 0.0009, + "step": 19355 + }, + { + "epoch": 2.098438855160451, + "grad_norm": 6.746103763580322, + "learning_rate": 6.032206586391992e-06, + "loss": 0.0072, + "step": 19356 + }, + { + "epoch": 2.0985472679965307, + "grad_norm": 0.033195726573467255, + "learning_rate": 6.031481212824605e-06, + "loss": 0.0004, + "step": 19357 + }, + { + "epoch": 2.0986556808326107, + "grad_norm": 1.047116994857788, + "learning_rate": 6.030755839257217e-06, + "loss": 0.0049, + "step": 19358 + }, + { + "epoch": 2.09876409366869, + "grad_norm": 8.513188362121582, + "learning_rate": 6.030030465689831e-06, + "loss": 0.0904, + "step": 19359 + }, + { + "epoch": 2.09887250650477, + "grad_norm": 0.5373373031616211, + "learning_rate": 6.029305092122444e-06, + "loss": 0.0018, + "step": 19360 + }, + { + "epoch": 2.0989809193408497, + "grad_norm": 0.10652627050876617, + "learning_rate": 6.028579718555056e-06, + "loss": 0.0009, + "step": 19361 + }, + { + "epoch": 2.0990893321769297, + "grad_norm": 0.40197718143463135, + "learning_rate": 6.027854344987669e-06, + "loss": 0.0016, + "step": 19362 + }, + { + "epoch": 2.0991977450130097, + "grad_norm": 2.993007183074951, + "learning_rate": 6.027128971420282e-06, + "loss": 0.0265, + "step": 19363 + }, + { + "epoch": 2.0993061578490892, + "grad_norm": 12.662276268005371, + "learning_rate": 6.026403597852894e-06, + "loss": 0.0424, + "step": 19364 + }, + { + "epoch": 2.0994145706851692, + "grad_norm": 6.216653347015381, + "learning_rate": 6.025678224285507e-06, + "loss": 0.0071, + "step": 19365 + }, + { + "epoch": 2.0995229835212488, + "grad_norm": 3.5258004665374756, + "learning_rate": 6.02495285071812e-06, + "loss": 0.0406, + "step": 19366 + }, + { + "epoch": 2.0996313963573288, + "grad_norm": 3.185147285461426, + "learning_rate": 6.024227477150733e-06, + "loss": 0.0091, + "step": 19367 + }, + { + "epoch": 2.0997398091934083, + "grad_norm": 0.0016381823224946856, + "learning_rate": 6.023502103583346e-06, + "loss": 0.0, + "step": 19368 + }, + { + "epoch": 2.0998482220294883, + "grad_norm": 0.0608690120279789, + "learning_rate": 6.022776730015959e-06, + "loss": 0.0003, + "step": 19369 + }, + { + "epoch": 2.0999566348655683, + "grad_norm": 0.05756315961480141, + "learning_rate": 6.022051356448571e-06, + "loss": 0.0003, + "step": 19370 + }, + { + "epoch": 2.100065047701648, + "grad_norm": 0.09394017606973648, + "learning_rate": 6.021325982881184e-06, + "loss": 0.0002, + "step": 19371 + }, + { + "epoch": 2.100173460537728, + "grad_norm": 0.023054057732224464, + "learning_rate": 6.020600609313797e-06, + "loss": 0.0001, + "step": 19372 + }, + { + "epoch": 2.1002818733738073, + "grad_norm": 0.9477412700653076, + "learning_rate": 6.0198752357464095e-06, + "loss": 0.0026, + "step": 19373 + }, + { + "epoch": 2.1003902862098873, + "grad_norm": 2.340324640274048, + "learning_rate": 6.0191498621790225e-06, + "loss": 0.038, + "step": 19374 + }, + { + "epoch": 2.100498699045967, + "grad_norm": 4.802672386169434, + "learning_rate": 6.018424488611635e-06, + "loss": 0.0256, + "step": 19375 + }, + { + "epoch": 2.100607111882047, + "grad_norm": 0.2990272641181946, + "learning_rate": 6.0176991150442484e-06, + "loss": 0.0007, + "step": 19376 + }, + { + "epoch": 2.100715524718127, + "grad_norm": 0.04455355182290077, + "learning_rate": 6.0169737414768614e-06, + "loss": 0.0004, + "step": 19377 + }, + { + "epoch": 2.1008239375542064, + "grad_norm": 1.6259846687316895, + "learning_rate": 6.016248367909474e-06, + "loss": 0.0005, + "step": 19378 + }, + { + "epoch": 2.1009323503902864, + "grad_norm": 0.9741738438606262, + "learning_rate": 6.0155229943420866e-06, + "loss": 0.0022, + "step": 19379 + }, + { + "epoch": 2.101040763226366, + "grad_norm": 0.12061681598424911, + "learning_rate": 6.0147976207746996e-06, + "loss": 0.0004, + "step": 19380 + }, + { + "epoch": 2.101149176062446, + "grad_norm": 3.003678321838379, + "learning_rate": 6.014072247207312e-06, + "loss": 0.0043, + "step": 19381 + }, + { + "epoch": 2.1012575888985254, + "grad_norm": 0.15561437606811523, + "learning_rate": 6.013346873639925e-06, + "loss": 0.0004, + "step": 19382 + }, + { + "epoch": 2.1013660017346054, + "grad_norm": 0.031844574958086014, + "learning_rate": 6.012621500072538e-06, + "loss": 0.0001, + "step": 19383 + }, + { + "epoch": 2.101474414570685, + "grad_norm": 1.0745612382888794, + "learning_rate": 6.0118961265051515e-06, + "loss": 0.0048, + "step": 19384 + }, + { + "epoch": 2.101582827406765, + "grad_norm": 0.6381373405456543, + "learning_rate": 6.011170752937764e-06, + "loss": 0.003, + "step": 19385 + }, + { + "epoch": 2.101691240242845, + "grad_norm": 0.875654399394989, + "learning_rate": 6.010445379370377e-06, + "loss": 0.0046, + "step": 19386 + }, + { + "epoch": 2.1017996530789245, + "grad_norm": 0.26558345556259155, + "learning_rate": 6.009720005802989e-06, + "loss": 0.0008, + "step": 19387 + }, + { + "epoch": 2.1019080659150045, + "grad_norm": 5.541418552398682, + "learning_rate": 6.008994632235602e-06, + "loss": 0.0178, + "step": 19388 + }, + { + "epoch": 2.102016478751084, + "grad_norm": 4.76491641998291, + "learning_rate": 6.008269258668215e-06, + "loss": 0.0082, + "step": 19389 + }, + { + "epoch": 2.102124891587164, + "grad_norm": 3.6863157749176025, + "learning_rate": 6.007543885100827e-06, + "loss": 0.0363, + "step": 19390 + }, + { + "epoch": 2.1022333044232435, + "grad_norm": 2.4228978157043457, + "learning_rate": 6.00681851153344e-06, + "loss": 0.0095, + "step": 19391 + }, + { + "epoch": 2.1023417172593235, + "grad_norm": 0.22969822585582733, + "learning_rate": 6.006093137966053e-06, + "loss": 0.002, + "step": 19392 + }, + { + "epoch": 2.1024501300954035, + "grad_norm": 0.011365426704287529, + "learning_rate": 6.005367764398667e-06, + "loss": 0.0001, + "step": 19393 + }, + { + "epoch": 2.102558542931483, + "grad_norm": 5.119755268096924, + "learning_rate": 6.004642390831279e-06, + "loss": 0.0459, + "step": 19394 + }, + { + "epoch": 2.102666955767563, + "grad_norm": 0.9961339235305786, + "learning_rate": 6.003917017263892e-06, + "loss": 0.0036, + "step": 19395 + }, + { + "epoch": 2.1027753686036426, + "grad_norm": 0.4663887023925781, + "learning_rate": 6.003191643696504e-06, + "loss": 0.0017, + "step": 19396 + }, + { + "epoch": 2.1028837814397225, + "grad_norm": 0.21366927027702332, + "learning_rate": 6.002466270129117e-06, + "loss": 0.0011, + "step": 19397 + }, + { + "epoch": 2.102992194275802, + "grad_norm": 2.003498077392578, + "learning_rate": 6.00174089656173e-06, + "loss": 0.006, + "step": 19398 + }, + { + "epoch": 2.103100607111882, + "grad_norm": 0.2829146087169647, + "learning_rate": 6.001015522994342e-06, + "loss": 0.0008, + "step": 19399 + }, + { + "epoch": 2.103209019947962, + "grad_norm": 6.669656753540039, + "learning_rate": 6.000290149426955e-06, + "loss": 0.0078, + "step": 19400 + }, + { + "epoch": 2.1033174327840416, + "grad_norm": 0.9664578437805176, + "learning_rate": 5.999564775859567e-06, + "loss": 0.0022, + "step": 19401 + }, + { + "epoch": 2.1034258456201216, + "grad_norm": 1.9351873397827148, + "learning_rate": 5.998839402292181e-06, + "loss": 0.0312, + "step": 19402 + }, + { + "epoch": 2.103534258456201, + "grad_norm": 0.782366931438446, + "learning_rate": 5.998114028724794e-06, + "loss": 0.0033, + "step": 19403 + }, + { + "epoch": 2.103642671292281, + "grad_norm": 4.973428726196289, + "learning_rate": 5.997388655157407e-06, + "loss": 0.0217, + "step": 19404 + }, + { + "epoch": 2.1037510841283606, + "grad_norm": 0.6911669373512268, + "learning_rate": 5.996663281590019e-06, + "loss": 0.0005, + "step": 19405 + }, + { + "epoch": 2.1038594969644406, + "grad_norm": 1.9871448278427124, + "learning_rate": 5.995937908022632e-06, + "loss": 0.0102, + "step": 19406 + }, + { + "epoch": 2.10396790980052, + "grad_norm": 0.1168290227651596, + "learning_rate": 5.995212534455244e-06, + "loss": 0.0003, + "step": 19407 + }, + { + "epoch": 2.1040763226366, + "grad_norm": 0.04064435884356499, + "learning_rate": 5.994487160887857e-06, + "loss": 0.0003, + "step": 19408 + }, + { + "epoch": 2.10418473547268, + "grad_norm": 0.23655471205711365, + "learning_rate": 5.99376178732047e-06, + "loss": 0.0007, + "step": 19409 + }, + { + "epoch": 2.1042931483087597, + "grad_norm": 0.8415039777755737, + "learning_rate": 5.993036413753084e-06, + "loss": 0.0015, + "step": 19410 + }, + { + "epoch": 2.1044015611448397, + "grad_norm": 0.10446134209632874, + "learning_rate": 5.992311040185696e-06, + "loss": 0.0007, + "step": 19411 + }, + { + "epoch": 2.104509973980919, + "grad_norm": 2.6180334091186523, + "learning_rate": 5.991585666618309e-06, + "loss": 0.0336, + "step": 19412 + }, + { + "epoch": 2.104618386816999, + "grad_norm": 0.0887809470295906, + "learning_rate": 5.9908602930509215e-06, + "loss": 0.0008, + "step": 19413 + }, + { + "epoch": 2.1047267996530787, + "grad_norm": 0.1821269392967224, + "learning_rate": 5.9901349194835345e-06, + "loss": 0.0004, + "step": 19414 + }, + { + "epoch": 2.1048352124891587, + "grad_norm": 0.11708133667707443, + "learning_rate": 5.9894095459161474e-06, + "loss": 0.0005, + "step": 19415 + }, + { + "epoch": 2.1049436253252387, + "grad_norm": 0.017536146566271782, + "learning_rate": 5.98868417234876e-06, + "loss": 0.0001, + "step": 19416 + }, + { + "epoch": 2.1050520381613183, + "grad_norm": 1.7239267826080322, + "learning_rate": 5.9879587987813726e-06, + "loss": 0.0105, + "step": 19417 + }, + { + "epoch": 2.1051604509973982, + "grad_norm": 6.036871433258057, + "learning_rate": 5.9872334252139856e-06, + "loss": 0.019, + "step": 19418 + }, + { + "epoch": 2.1052688638334778, + "grad_norm": 2.965053081512451, + "learning_rate": 5.986508051646599e-06, + "loss": 0.0024, + "step": 19419 + }, + { + "epoch": 2.1053772766695578, + "grad_norm": 1.552775502204895, + "learning_rate": 5.9857826780792115e-06, + "loss": 0.0123, + "step": 19420 + }, + { + "epoch": 2.1054856895056373, + "grad_norm": 0.034690290689468384, + "learning_rate": 5.9850573045118245e-06, + "loss": 0.0002, + "step": 19421 + }, + { + "epoch": 2.1055941023417173, + "grad_norm": 4.6046462059021, + "learning_rate": 5.984331930944437e-06, + "loss": 0.0135, + "step": 19422 + }, + { + "epoch": 2.1057025151777973, + "grad_norm": 0.5768752098083496, + "learning_rate": 5.98360655737705e-06, + "loss": 0.0015, + "step": 19423 + }, + { + "epoch": 2.105810928013877, + "grad_norm": 3.815626621246338, + "learning_rate": 5.982881183809663e-06, + "loss": 0.0218, + "step": 19424 + }, + { + "epoch": 2.105919340849957, + "grad_norm": 1.525823950767517, + "learning_rate": 5.982155810242275e-06, + "loss": 0.0076, + "step": 19425 + }, + { + "epoch": 2.1060277536860363, + "grad_norm": 0.263746052980423, + "learning_rate": 5.981430436674888e-06, + "loss": 0.0006, + "step": 19426 + }, + { + "epoch": 2.1061361665221163, + "grad_norm": 0.09691056609153748, + "learning_rate": 5.9807050631075e-06, + "loss": 0.0004, + "step": 19427 + }, + { + "epoch": 2.106244579358196, + "grad_norm": 22.926090240478516, + "learning_rate": 5.979979689540114e-06, + "loss": 0.0121, + "step": 19428 + }, + { + "epoch": 2.106352992194276, + "grad_norm": 10.111135482788086, + "learning_rate": 5.979254315972727e-06, + "loss": 0.0393, + "step": 19429 + }, + { + "epoch": 2.1064614050303554, + "grad_norm": 4.53515625, + "learning_rate": 5.97852894240534e-06, + "loss": 0.0332, + "step": 19430 + }, + { + "epoch": 2.1065698178664354, + "grad_norm": 0.17489999532699585, + "learning_rate": 5.977803568837952e-06, + "loss": 0.0012, + "step": 19431 + }, + { + "epoch": 2.1066782307025154, + "grad_norm": 5.624704360961914, + "learning_rate": 5.977078195270565e-06, + "loss": 0.0142, + "step": 19432 + }, + { + "epoch": 2.106786643538595, + "grad_norm": 2.456501007080078, + "learning_rate": 5.976352821703177e-06, + "loss": 0.0015, + "step": 19433 + }, + { + "epoch": 2.106895056374675, + "grad_norm": 4.295406818389893, + "learning_rate": 5.97562744813579e-06, + "loss": 0.0349, + "step": 19434 + }, + { + "epoch": 2.1070034692107544, + "grad_norm": 0.062028609216213226, + "learning_rate": 5.974902074568403e-06, + "loss": 0.0001, + "step": 19435 + }, + { + "epoch": 2.1071118820468344, + "grad_norm": 4.689085006713867, + "learning_rate": 5.974176701001017e-06, + "loss": 0.0349, + "step": 19436 + }, + { + "epoch": 2.107220294882914, + "grad_norm": 0.7996684312820435, + "learning_rate": 5.973451327433629e-06, + "loss": 0.0031, + "step": 19437 + }, + { + "epoch": 2.107328707718994, + "grad_norm": 0.9444612860679626, + "learning_rate": 5.972725953866242e-06, + "loss": 0.0015, + "step": 19438 + }, + { + "epoch": 2.107437120555074, + "grad_norm": 0.5523749589920044, + "learning_rate": 5.972000580298854e-06, + "loss": 0.0014, + "step": 19439 + }, + { + "epoch": 2.1075455333911535, + "grad_norm": 0.006660949904471636, + "learning_rate": 5.971275206731467e-06, + "loss": 0.0, + "step": 19440 + }, + { + "epoch": 2.1076539462272335, + "grad_norm": 0.2924245595932007, + "learning_rate": 5.97054983316408e-06, + "loss": 0.0013, + "step": 19441 + }, + { + "epoch": 2.107762359063313, + "grad_norm": 0.7706390619277954, + "learning_rate": 5.969824459596692e-06, + "loss": 0.0431, + "step": 19442 + }, + { + "epoch": 2.107870771899393, + "grad_norm": 12.309609413146973, + "learning_rate": 5.969099086029305e-06, + "loss": 0.0219, + "step": 19443 + }, + { + "epoch": 2.1079791847354725, + "grad_norm": 1.053802251815796, + "learning_rate": 5.968373712461918e-06, + "loss": 0.0069, + "step": 19444 + }, + { + "epoch": 2.1080875975715525, + "grad_norm": 0.05384593456983566, + "learning_rate": 5.967648338894532e-06, + "loss": 0.0003, + "step": 19445 + }, + { + "epoch": 2.1081960104076325, + "grad_norm": 0.265300989151001, + "learning_rate": 5.966922965327144e-06, + "loss": 0.0025, + "step": 19446 + }, + { + "epoch": 2.108304423243712, + "grad_norm": 0.22339583933353424, + "learning_rate": 5.966197591759757e-06, + "loss": 0.0007, + "step": 19447 + }, + { + "epoch": 2.108412836079792, + "grad_norm": 0.0029499479569494724, + "learning_rate": 5.965472218192369e-06, + "loss": 0.0, + "step": 19448 + }, + { + "epoch": 2.1085212489158716, + "grad_norm": 0.9263308644294739, + "learning_rate": 5.964746844624982e-06, + "loss": 0.002, + "step": 19449 + }, + { + "epoch": 2.1086296617519515, + "grad_norm": 4.772894382476807, + "learning_rate": 5.964021471057595e-06, + "loss": 0.0192, + "step": 19450 + }, + { + "epoch": 2.108738074588031, + "grad_norm": 0.2033633291721344, + "learning_rate": 5.9632960974902075e-06, + "loss": 0.0008, + "step": 19451 + }, + { + "epoch": 2.108846487424111, + "grad_norm": 3.1401398181915283, + "learning_rate": 5.9625707239228205e-06, + "loss": 0.0187, + "step": 19452 + }, + { + "epoch": 2.1089549002601906, + "grad_norm": 9.380565643310547, + "learning_rate": 5.961845350355433e-06, + "loss": 0.0133, + "step": 19453 + }, + { + "epoch": 2.1090633130962706, + "grad_norm": 1.3655492067337036, + "learning_rate": 5.9611199767880464e-06, + "loss": 0.0024, + "step": 19454 + }, + { + "epoch": 2.1091717259323506, + "grad_norm": 0.6195462942123413, + "learning_rate": 5.9603946032206594e-06, + "loss": 0.0017, + "step": 19455 + }, + { + "epoch": 2.10928013876843, + "grad_norm": 1.155300498008728, + "learning_rate": 5.9596692296532724e-06, + "loss": 0.0016, + "step": 19456 + }, + { + "epoch": 2.10938855160451, + "grad_norm": 0.33595606684684753, + "learning_rate": 5.9589438560858846e-06, + "loss": 0.0014, + "step": 19457 + }, + { + "epoch": 2.1094969644405897, + "grad_norm": 0.9176729321479797, + "learning_rate": 5.9582184825184976e-06, + "loss": 0.0017, + "step": 19458 + }, + { + "epoch": 2.1096053772766696, + "grad_norm": 0.3199469745159149, + "learning_rate": 5.95749310895111e-06, + "loss": 0.0017, + "step": 19459 + }, + { + "epoch": 2.109713790112749, + "grad_norm": 1.7479063272476196, + "learning_rate": 5.956767735383723e-06, + "loss": 0.0072, + "step": 19460 + }, + { + "epoch": 2.109822202948829, + "grad_norm": 0.09386589378118515, + "learning_rate": 5.956042361816336e-06, + "loss": 0.0002, + "step": 19461 + }, + { + "epoch": 2.109930615784909, + "grad_norm": 0.056663453578948975, + "learning_rate": 5.9553169882489495e-06, + "loss": 0.0003, + "step": 19462 + }, + { + "epoch": 2.1100390286209887, + "grad_norm": 0.16395564377307892, + "learning_rate": 5.954591614681562e-06, + "loss": 0.0002, + "step": 19463 + }, + { + "epoch": 2.1101474414570687, + "grad_norm": 8.611160278320312, + "learning_rate": 5.953866241114175e-06, + "loss": 0.0276, + "step": 19464 + }, + { + "epoch": 2.110255854293148, + "grad_norm": 3.0531809329986572, + "learning_rate": 5.953140867546787e-06, + "loss": 0.0497, + "step": 19465 + }, + { + "epoch": 2.110364267129228, + "grad_norm": 0.06271495670080185, + "learning_rate": 5.9524154939794e-06, + "loss": 0.0003, + "step": 19466 + }, + { + "epoch": 2.1104726799653077, + "grad_norm": 7.114628791809082, + "learning_rate": 5.951690120412013e-06, + "loss": 0.0761, + "step": 19467 + }, + { + "epoch": 2.1105810928013877, + "grad_norm": 5.75114631652832, + "learning_rate": 5.950964746844625e-06, + "loss": 0.0078, + "step": 19468 + }, + { + "epoch": 2.1106895056374673, + "grad_norm": 1.2638630867004395, + "learning_rate": 5.950239373277238e-06, + "loss": 0.0048, + "step": 19469 + }, + { + "epoch": 2.1107979184735473, + "grad_norm": 1.9547498226165771, + "learning_rate": 5.949513999709851e-06, + "loss": 0.0032, + "step": 19470 + }, + { + "epoch": 2.1109063313096272, + "grad_norm": 1.240515947341919, + "learning_rate": 5.948788626142465e-06, + "loss": 0.0045, + "step": 19471 + }, + { + "epoch": 2.111014744145707, + "grad_norm": 0.6977297067642212, + "learning_rate": 5.948063252575077e-06, + "loss": 0.0011, + "step": 19472 + }, + { + "epoch": 2.1111231569817868, + "grad_norm": 0.09665647149085999, + "learning_rate": 5.94733787900769e-06, + "loss": 0.0004, + "step": 19473 + }, + { + "epoch": 2.1112315698178663, + "grad_norm": 0.021888354793190956, + "learning_rate": 5.946612505440302e-06, + "loss": 0.0001, + "step": 19474 + }, + { + "epoch": 2.1113399826539463, + "grad_norm": 22.901905059814453, + "learning_rate": 5.945887131872915e-06, + "loss": 0.0136, + "step": 19475 + }, + { + "epoch": 2.111448395490026, + "grad_norm": 0.8254827260971069, + "learning_rate": 5.945161758305528e-06, + "loss": 0.0034, + "step": 19476 + }, + { + "epoch": 2.111556808326106, + "grad_norm": 0.08708079904317856, + "learning_rate": 5.94443638473814e-06, + "loss": 0.0003, + "step": 19477 + }, + { + "epoch": 2.111665221162186, + "grad_norm": 4.552458763122559, + "learning_rate": 5.943711011170753e-06, + "loss": 0.0049, + "step": 19478 + }, + { + "epoch": 2.1117736339982653, + "grad_norm": 3.2928414344787598, + "learning_rate": 5.942985637603365e-06, + "loss": 0.0356, + "step": 19479 + }, + { + "epoch": 2.1118820468343453, + "grad_norm": 0.031915705651044846, + "learning_rate": 5.942260264035979e-06, + "loss": 0.0001, + "step": 19480 + }, + { + "epoch": 2.111990459670425, + "grad_norm": 0.06772616505622864, + "learning_rate": 5.941534890468592e-06, + "loss": 0.0004, + "step": 19481 + }, + { + "epoch": 2.112098872506505, + "grad_norm": 0.30912017822265625, + "learning_rate": 5.940809516901205e-06, + "loss": 0.0009, + "step": 19482 + }, + { + "epoch": 2.1122072853425844, + "grad_norm": 0.03688455745577812, + "learning_rate": 5.940084143333817e-06, + "loss": 0.0003, + "step": 19483 + }, + { + "epoch": 2.1123156981786644, + "grad_norm": 6.684323310852051, + "learning_rate": 5.93935876976643e-06, + "loss": 0.0307, + "step": 19484 + }, + { + "epoch": 2.112424111014744, + "grad_norm": 0.012071030214428902, + "learning_rate": 5.938633396199042e-06, + "loss": 0.0001, + "step": 19485 + }, + { + "epoch": 2.112532523850824, + "grad_norm": 0.054130468517541885, + "learning_rate": 5.937908022631655e-06, + "loss": 0.0002, + "step": 19486 + }, + { + "epoch": 2.112640936686904, + "grad_norm": 0.013396736234426498, + "learning_rate": 5.937182649064268e-06, + "loss": 0.0001, + "step": 19487 + }, + { + "epoch": 2.1127493495229834, + "grad_norm": 3.104262113571167, + "learning_rate": 5.936457275496882e-06, + "loss": 0.0345, + "step": 19488 + }, + { + "epoch": 2.1128577623590634, + "grad_norm": 11.240019798278809, + "learning_rate": 5.935731901929494e-06, + "loss": 0.045, + "step": 19489 + }, + { + "epoch": 2.112966175195143, + "grad_norm": 0.23914851248264313, + "learning_rate": 5.935006528362107e-06, + "loss": 0.0024, + "step": 19490 + }, + { + "epoch": 2.113074588031223, + "grad_norm": 0.07641922682523727, + "learning_rate": 5.9342811547947195e-06, + "loss": 0.0002, + "step": 19491 + }, + { + "epoch": 2.1131830008673025, + "grad_norm": 0.03764388710260391, + "learning_rate": 5.9335557812273325e-06, + "loss": 0.0001, + "step": 19492 + }, + { + "epoch": 2.1132914137033825, + "grad_norm": 3.061718702316284, + "learning_rate": 5.9328304076599454e-06, + "loss": 0.0082, + "step": 19493 + }, + { + "epoch": 2.1133998265394625, + "grad_norm": 0.006364813074469566, + "learning_rate": 5.932105034092558e-06, + "loss": 0.0, + "step": 19494 + }, + { + "epoch": 2.113508239375542, + "grad_norm": 0.11210162192583084, + "learning_rate": 5.931379660525171e-06, + "loss": 0.0006, + "step": 19495 + }, + { + "epoch": 2.113616652211622, + "grad_norm": 1.0790976285934448, + "learning_rate": 5.9306542869577836e-06, + "loss": 0.0025, + "step": 19496 + }, + { + "epoch": 2.1137250650477015, + "grad_norm": 0.32875725626945496, + "learning_rate": 5.929928913390397e-06, + "loss": 0.0019, + "step": 19497 + }, + { + "epoch": 2.1138334778837815, + "grad_norm": 11.71218204498291, + "learning_rate": 5.9292035398230096e-06, + "loss": 0.006, + "step": 19498 + }, + { + "epoch": 2.113941890719861, + "grad_norm": 1.0940184593200684, + "learning_rate": 5.9284781662556225e-06, + "loss": 0.0048, + "step": 19499 + }, + { + "epoch": 2.114050303555941, + "grad_norm": 1.401677131652832, + "learning_rate": 5.927752792688235e-06, + "loss": 0.003, + "step": 19500 + }, + { + "epoch": 2.114158716392021, + "grad_norm": 0.1749657839536667, + "learning_rate": 5.927027419120848e-06, + "loss": 0.0011, + "step": 19501 + }, + { + "epoch": 2.1142671292281006, + "grad_norm": 4.734978199005127, + "learning_rate": 5.926302045553461e-06, + "loss": 0.0192, + "step": 19502 + }, + { + "epoch": 2.1143755420641805, + "grad_norm": 4.302065849304199, + "learning_rate": 5.925576671986073e-06, + "loss": 0.0385, + "step": 19503 + }, + { + "epoch": 2.11448395490026, + "grad_norm": 0.008590414188802242, + "learning_rate": 5.924851298418686e-06, + "loss": 0.0001, + "step": 19504 + }, + { + "epoch": 2.11459236773634, + "grad_norm": 0.2831566631793976, + "learning_rate": 5.9241259248513e-06, + "loss": 0.0004, + "step": 19505 + }, + { + "epoch": 2.1147007805724196, + "grad_norm": 3.4173200130462646, + "learning_rate": 5.923400551283912e-06, + "loss": 0.0183, + "step": 19506 + }, + { + "epoch": 2.1148091934084996, + "grad_norm": 0.2579134702682495, + "learning_rate": 5.922675177716525e-06, + "loss": 0.0011, + "step": 19507 + }, + { + "epoch": 2.114917606244579, + "grad_norm": 0.032173868268728256, + "learning_rate": 5.921949804149138e-06, + "loss": 0.0002, + "step": 19508 + }, + { + "epoch": 2.115026019080659, + "grad_norm": 4.449549198150635, + "learning_rate": 5.92122443058175e-06, + "loss": 0.0289, + "step": 19509 + }, + { + "epoch": 2.115134431916739, + "grad_norm": 0.06026701256632805, + "learning_rate": 5.920499057014363e-06, + "loss": 0.0002, + "step": 19510 + }, + { + "epoch": 2.1152428447528187, + "grad_norm": 0.4951646029949188, + "learning_rate": 5.919773683446975e-06, + "loss": 0.001, + "step": 19511 + }, + { + "epoch": 2.1153512575888986, + "grad_norm": 0.6856756210327148, + "learning_rate": 5.919048309879588e-06, + "loss": 0.0011, + "step": 19512 + }, + { + "epoch": 2.115459670424978, + "grad_norm": 0.12135796993970871, + "learning_rate": 5.918322936312201e-06, + "loss": 0.0004, + "step": 19513 + }, + { + "epoch": 2.115568083261058, + "grad_norm": 0.8578844666481018, + "learning_rate": 5.917597562744815e-06, + "loss": 0.0055, + "step": 19514 + }, + { + "epoch": 2.1156764960971377, + "grad_norm": 0.0196828730404377, + "learning_rate": 5.916872189177427e-06, + "loss": 0.0001, + "step": 19515 + }, + { + "epoch": 2.1157849089332177, + "grad_norm": 0.25672829151153564, + "learning_rate": 5.91614681561004e-06, + "loss": 0.0015, + "step": 19516 + }, + { + "epoch": 2.1158933217692977, + "grad_norm": 0.03603978455066681, + "learning_rate": 5.915421442042653e-06, + "loss": 0.0001, + "step": 19517 + }, + { + "epoch": 2.116001734605377, + "grad_norm": 4.66955041885376, + "learning_rate": 5.914696068475265e-06, + "loss": 0.0074, + "step": 19518 + }, + { + "epoch": 2.116110147441457, + "grad_norm": 0.07216504216194153, + "learning_rate": 5.913970694907878e-06, + "loss": 0.0005, + "step": 19519 + }, + { + "epoch": 2.1162185602775367, + "grad_norm": 0.12499520182609558, + "learning_rate": 5.91324532134049e-06, + "loss": 0.0004, + "step": 19520 + }, + { + "epoch": 2.1163269731136167, + "grad_norm": 0.42281195521354675, + "learning_rate": 5.912519947773103e-06, + "loss": 0.0026, + "step": 19521 + }, + { + "epoch": 2.1164353859496963, + "grad_norm": 0.11346721649169922, + "learning_rate": 5.911794574205716e-06, + "loss": 0.0005, + "step": 19522 + }, + { + "epoch": 2.1165437987857763, + "grad_norm": 2.9439823627471924, + "learning_rate": 5.91106920063833e-06, + "loss": 0.0509, + "step": 19523 + }, + { + "epoch": 2.1166522116218562, + "grad_norm": 0.08602991700172424, + "learning_rate": 5.910343827070942e-06, + "loss": 0.0005, + "step": 19524 + }, + { + "epoch": 2.116760624457936, + "grad_norm": 1.28598952293396, + "learning_rate": 5.909618453503555e-06, + "loss": 0.0032, + "step": 19525 + }, + { + "epoch": 2.1168690372940158, + "grad_norm": 1.8731495141983032, + "learning_rate": 5.908893079936167e-06, + "loss": 0.0156, + "step": 19526 + }, + { + "epoch": 2.1169774501300953, + "grad_norm": 3.7575676441192627, + "learning_rate": 5.90816770636878e-06, + "loss": 0.0146, + "step": 19527 + }, + { + "epoch": 2.1170858629661753, + "grad_norm": 14.326639175415039, + "learning_rate": 5.907442332801393e-06, + "loss": 0.0134, + "step": 19528 + }, + { + "epoch": 2.117194275802255, + "grad_norm": 1.3965343236923218, + "learning_rate": 5.9067169592340055e-06, + "loss": 0.0213, + "step": 19529 + }, + { + "epoch": 2.117302688638335, + "grad_norm": 0.17190559208393097, + "learning_rate": 5.9059915856666185e-06, + "loss": 0.0009, + "step": 19530 + }, + { + "epoch": 2.1174111014744144, + "grad_norm": 0.007905954495072365, + "learning_rate": 5.905266212099232e-06, + "loss": 0.0001, + "step": 19531 + }, + { + "epoch": 2.1175195143104943, + "grad_norm": 2.145578384399414, + "learning_rate": 5.9045408385318445e-06, + "loss": 0.0033, + "step": 19532 + }, + { + "epoch": 2.1176279271465743, + "grad_norm": 0.3446119427680969, + "learning_rate": 5.9038154649644574e-06, + "loss": 0.0008, + "step": 19533 + }, + { + "epoch": 2.117736339982654, + "grad_norm": 0.09506022185087204, + "learning_rate": 5.9030900913970704e-06, + "loss": 0.0006, + "step": 19534 + }, + { + "epoch": 2.117844752818734, + "grad_norm": 0.35751470923423767, + "learning_rate": 5.9023647178296826e-06, + "loss": 0.001, + "step": 19535 + }, + { + "epoch": 2.1179531656548134, + "grad_norm": 0.9040983319282532, + "learning_rate": 5.9016393442622956e-06, + "loss": 0.0044, + "step": 19536 + }, + { + "epoch": 2.1180615784908934, + "grad_norm": 5.54755163192749, + "learning_rate": 5.900913970694908e-06, + "loss": 0.0208, + "step": 19537 + }, + { + "epoch": 2.118169991326973, + "grad_norm": 2.128141164779663, + "learning_rate": 5.900188597127521e-06, + "loss": 0.0065, + "step": 19538 + }, + { + "epoch": 2.118278404163053, + "grad_norm": 13.013775825500488, + "learning_rate": 5.899463223560134e-06, + "loss": 0.0619, + "step": 19539 + }, + { + "epoch": 2.118386816999133, + "grad_norm": 0.03424437716603279, + "learning_rate": 5.8987378499927475e-06, + "loss": 0.0003, + "step": 19540 + }, + { + "epoch": 2.1184952298352124, + "grad_norm": 0.028796972706913948, + "learning_rate": 5.89801247642536e-06, + "loss": 0.0001, + "step": 19541 + }, + { + "epoch": 2.1186036426712924, + "grad_norm": 0.05683949217200279, + "learning_rate": 5.897287102857973e-06, + "loss": 0.0002, + "step": 19542 + }, + { + "epoch": 2.118712055507372, + "grad_norm": 0.06515245884656906, + "learning_rate": 5.896561729290586e-06, + "loss": 0.0002, + "step": 19543 + }, + { + "epoch": 2.118820468343452, + "grad_norm": 0.028956707566976547, + "learning_rate": 5.895836355723198e-06, + "loss": 0.0002, + "step": 19544 + }, + { + "epoch": 2.1189288811795315, + "grad_norm": 0.21526408195495605, + "learning_rate": 5.895110982155811e-06, + "loss": 0.0006, + "step": 19545 + }, + { + "epoch": 2.1190372940156115, + "grad_norm": 2.4519853591918945, + "learning_rate": 5.894385608588423e-06, + "loss": 0.0251, + "step": 19546 + }, + { + "epoch": 2.1191457068516915, + "grad_norm": 0.20207124948501587, + "learning_rate": 5.893660235021036e-06, + "loss": 0.0005, + "step": 19547 + }, + { + "epoch": 2.119254119687771, + "grad_norm": 0.1604815423488617, + "learning_rate": 5.892934861453649e-06, + "loss": 0.0003, + "step": 19548 + }, + { + "epoch": 2.119362532523851, + "grad_norm": 0.03605787083506584, + "learning_rate": 5.892209487886263e-06, + "loss": 0.0001, + "step": 19549 + }, + { + "epoch": 2.1194709453599305, + "grad_norm": 0.2698861360549927, + "learning_rate": 5.891484114318875e-06, + "loss": 0.0008, + "step": 19550 + }, + { + "epoch": 2.1195793581960105, + "grad_norm": 0.3803677558898926, + "learning_rate": 5.890758740751488e-06, + "loss": 0.0013, + "step": 19551 + }, + { + "epoch": 2.11968777103209, + "grad_norm": 0.07747866213321686, + "learning_rate": 5.8900333671841e-06, + "loss": 0.0003, + "step": 19552 + }, + { + "epoch": 2.11979618386817, + "grad_norm": 0.018259340897202492, + "learning_rate": 5.889307993616713e-06, + "loss": 0.0002, + "step": 19553 + }, + { + "epoch": 2.1199045967042496, + "grad_norm": 8.512858390808105, + "learning_rate": 5.888582620049326e-06, + "loss": 0.0296, + "step": 19554 + }, + { + "epoch": 2.1200130095403296, + "grad_norm": 4.7862396240234375, + "learning_rate": 5.887857246481938e-06, + "loss": 0.0348, + "step": 19555 + }, + { + "epoch": 2.1201214223764095, + "grad_norm": 0.09903135150671005, + "learning_rate": 5.887131872914551e-06, + "loss": 0.0003, + "step": 19556 + }, + { + "epoch": 2.120229835212489, + "grad_norm": 5.284798622131348, + "learning_rate": 5.886406499347165e-06, + "loss": 0.0318, + "step": 19557 + }, + { + "epoch": 2.120338248048569, + "grad_norm": 5.344504356384277, + "learning_rate": 5.885681125779777e-06, + "loss": 0.0362, + "step": 19558 + }, + { + "epoch": 2.1204466608846486, + "grad_norm": 0.05061676725745201, + "learning_rate": 5.88495575221239e-06, + "loss": 0.0003, + "step": 19559 + }, + { + "epoch": 2.1205550737207286, + "grad_norm": 0.14562761783599854, + "learning_rate": 5.884230378645003e-06, + "loss": 0.0008, + "step": 19560 + }, + { + "epoch": 2.120663486556808, + "grad_norm": 0.1369703859090805, + "learning_rate": 5.883505005077615e-06, + "loss": 0.0004, + "step": 19561 + }, + { + "epoch": 2.120771899392888, + "grad_norm": 3.045746326446533, + "learning_rate": 5.882779631510228e-06, + "loss": 0.0414, + "step": 19562 + }, + { + "epoch": 2.120880312228968, + "grad_norm": 0.11154891550540924, + "learning_rate": 5.88205425794284e-06, + "loss": 0.0003, + "step": 19563 + }, + { + "epoch": 2.1209887250650477, + "grad_norm": 0.14408130943775177, + "learning_rate": 5.881328884375453e-06, + "loss": 0.0006, + "step": 19564 + }, + { + "epoch": 2.1210971379011276, + "grad_norm": 0.011659645475447178, + "learning_rate": 5.880603510808066e-06, + "loss": 0.0, + "step": 19565 + }, + { + "epoch": 2.121205550737207, + "grad_norm": 0.14958855509757996, + "learning_rate": 5.87987813724068e-06, + "loss": 0.0003, + "step": 19566 + }, + { + "epoch": 2.121313963573287, + "grad_norm": 0.8414465188980103, + "learning_rate": 5.879152763673292e-06, + "loss": 0.001, + "step": 19567 + }, + { + "epoch": 2.1214223764093667, + "grad_norm": 0.20739832520484924, + "learning_rate": 5.878427390105905e-06, + "loss": 0.0004, + "step": 19568 + }, + { + "epoch": 2.1215307892454467, + "grad_norm": 7.82926082611084, + "learning_rate": 5.877702016538518e-06, + "loss": 0.0588, + "step": 19569 + }, + { + "epoch": 2.1216392020815267, + "grad_norm": 0.29830700159072876, + "learning_rate": 5.8769766429711305e-06, + "loss": 0.0003, + "step": 19570 + }, + { + "epoch": 2.121747614917606, + "grad_norm": 0.4616304337978363, + "learning_rate": 5.8762512694037435e-06, + "loss": 0.0202, + "step": 19571 + }, + { + "epoch": 2.121856027753686, + "grad_norm": 0.3945835828781128, + "learning_rate": 5.875525895836356e-06, + "loss": 0.002, + "step": 19572 + }, + { + "epoch": 2.1219644405897657, + "grad_norm": 2.2883479595184326, + "learning_rate": 5.874800522268969e-06, + "loss": 0.0036, + "step": 19573 + }, + { + "epoch": 2.1220728534258457, + "grad_norm": 4.694845199584961, + "learning_rate": 5.874075148701582e-06, + "loss": 0.0482, + "step": 19574 + }, + { + "epoch": 2.1221812662619253, + "grad_norm": 0.05188356339931488, + "learning_rate": 5.873349775134195e-06, + "loss": 0.0003, + "step": 19575 + }, + { + "epoch": 2.1222896790980053, + "grad_norm": 5.094089984893799, + "learning_rate": 5.8726244015668076e-06, + "loss": 0.0783, + "step": 19576 + }, + { + "epoch": 2.122398091934085, + "grad_norm": 0.40151435136795044, + "learning_rate": 5.8718990279994205e-06, + "loss": 0.0009, + "step": 19577 + }, + { + "epoch": 2.122506504770165, + "grad_norm": 0.10361406952142715, + "learning_rate": 5.871173654432033e-06, + "loss": 0.0003, + "step": 19578 + }, + { + "epoch": 2.1226149176062448, + "grad_norm": 0.28572526574134827, + "learning_rate": 5.870448280864646e-06, + "loss": 0.0006, + "step": 19579 + }, + { + "epoch": 2.1227233304423243, + "grad_norm": 0.1961161345243454, + "learning_rate": 5.869722907297259e-06, + "loss": 0.0004, + "step": 19580 + }, + { + "epoch": 2.1228317432784043, + "grad_norm": 5.5882439613342285, + "learning_rate": 5.868997533729871e-06, + "loss": 0.0566, + "step": 19581 + }, + { + "epoch": 2.122940156114484, + "grad_norm": 4.844860553741455, + "learning_rate": 5.868272160162484e-06, + "loss": 0.0344, + "step": 19582 + }, + { + "epoch": 2.123048568950564, + "grad_norm": 0.08277827501296997, + "learning_rate": 5.867546786595098e-06, + "loss": 0.0003, + "step": 19583 + }, + { + "epoch": 2.1231569817866434, + "grad_norm": 4.3880815505981445, + "learning_rate": 5.86682141302771e-06, + "loss": 0.1155, + "step": 19584 + }, + { + "epoch": 2.1232653946227233, + "grad_norm": 0.012680264189839363, + "learning_rate": 5.866096039460323e-06, + "loss": 0.0001, + "step": 19585 + }, + { + "epoch": 2.1233738074588033, + "grad_norm": 0.8251170516014099, + "learning_rate": 5.865370665892936e-06, + "loss": 0.0028, + "step": 19586 + }, + { + "epoch": 2.123482220294883, + "grad_norm": 2.5077009201049805, + "learning_rate": 5.864645292325548e-06, + "loss": 0.0048, + "step": 19587 + }, + { + "epoch": 2.123590633130963, + "grad_norm": 0.08975028991699219, + "learning_rate": 5.863919918758161e-06, + "loss": 0.0005, + "step": 19588 + }, + { + "epoch": 2.1236990459670424, + "grad_norm": 2.079392433166504, + "learning_rate": 5.863194545190773e-06, + "loss": 0.0193, + "step": 19589 + }, + { + "epoch": 2.1238074588031224, + "grad_norm": 3.3987646102905273, + "learning_rate": 5.862469171623386e-06, + "loss": 0.0206, + "step": 19590 + }, + { + "epoch": 2.123915871639202, + "grad_norm": 4.915334224700928, + "learning_rate": 5.861743798055999e-06, + "loss": 0.0347, + "step": 19591 + }, + { + "epoch": 2.124024284475282, + "grad_norm": 3.400820016860962, + "learning_rate": 5.861018424488613e-06, + "loss": 0.0031, + "step": 19592 + }, + { + "epoch": 2.1241326973113615, + "grad_norm": 1.181724190711975, + "learning_rate": 5.860293050921225e-06, + "loss": 0.0036, + "step": 19593 + }, + { + "epoch": 2.1242411101474414, + "grad_norm": 4.617620468139648, + "learning_rate": 5.859567677353838e-06, + "loss": 0.0243, + "step": 19594 + }, + { + "epoch": 2.1243495229835214, + "grad_norm": 0.011673773638904095, + "learning_rate": 5.858842303786451e-06, + "loss": 0.0001, + "step": 19595 + }, + { + "epoch": 2.124457935819601, + "grad_norm": 0.42756780982017517, + "learning_rate": 5.858116930219063e-06, + "loss": 0.0012, + "step": 19596 + }, + { + "epoch": 2.124566348655681, + "grad_norm": 1.0794055461883545, + "learning_rate": 5.857391556651676e-06, + "loss": 0.0377, + "step": 19597 + }, + { + "epoch": 2.1246747614917605, + "grad_norm": 3.2530429363250732, + "learning_rate": 5.856666183084288e-06, + "loss": 0.0119, + "step": 19598 + }, + { + "epoch": 2.1247831743278405, + "grad_norm": 0.03779491037130356, + "learning_rate": 5.855940809516901e-06, + "loss": 0.0003, + "step": 19599 + }, + { + "epoch": 2.12489158716392, + "grad_norm": 0.425022155046463, + "learning_rate": 5.855215435949515e-06, + "loss": 0.0009, + "step": 19600 + }, + { + "epoch": 2.125, + "grad_norm": 0.0033112706150859594, + "learning_rate": 5.854490062382128e-06, + "loss": 0.0, + "step": 19601 + }, + { + "epoch": 2.12510841283608, + "grad_norm": 0.05306894704699516, + "learning_rate": 5.85376468881474e-06, + "loss": 0.0002, + "step": 19602 + }, + { + "epoch": 2.1252168256721595, + "grad_norm": 0.2039579600095749, + "learning_rate": 5.853039315247353e-06, + "loss": 0.0003, + "step": 19603 + }, + { + "epoch": 2.1253252385082395, + "grad_norm": 3.8829286098480225, + "learning_rate": 5.852313941679965e-06, + "loss": 0.0028, + "step": 19604 + }, + { + "epoch": 2.125433651344319, + "grad_norm": 5.509915828704834, + "learning_rate": 5.851588568112578e-06, + "loss": 0.014, + "step": 19605 + }, + { + "epoch": 2.125542064180399, + "grad_norm": 0.2273344248533249, + "learning_rate": 5.850863194545191e-06, + "loss": 0.0013, + "step": 19606 + }, + { + "epoch": 2.1256504770164786, + "grad_norm": 0.6732335686683655, + "learning_rate": 5.8501378209778035e-06, + "loss": 0.0035, + "step": 19607 + }, + { + "epoch": 2.1257588898525586, + "grad_norm": 0.09439724683761597, + "learning_rate": 5.8494124474104165e-06, + "loss": 0.0007, + "step": 19608 + }, + { + "epoch": 2.125867302688638, + "grad_norm": 1.3525370359420776, + "learning_rate": 5.84868707384303e-06, + "loss": 0.0035, + "step": 19609 + }, + { + "epoch": 2.125975715524718, + "grad_norm": 4.733677387237549, + "learning_rate": 5.8479617002756425e-06, + "loss": 0.0437, + "step": 19610 + }, + { + "epoch": 2.126084128360798, + "grad_norm": 0.21350304782390594, + "learning_rate": 5.8472363267082554e-06, + "loss": 0.0014, + "step": 19611 + }, + { + "epoch": 2.1261925411968776, + "grad_norm": 4.202937126159668, + "learning_rate": 5.8465109531408684e-06, + "loss": 0.0211, + "step": 19612 + }, + { + "epoch": 2.1263009540329576, + "grad_norm": 0.4457376003265381, + "learning_rate": 5.845785579573481e-06, + "loss": 0.0019, + "step": 19613 + }, + { + "epoch": 2.126409366869037, + "grad_norm": 0.18099521100521088, + "learning_rate": 5.8450602060060936e-06, + "loss": 0.0007, + "step": 19614 + }, + { + "epoch": 2.126517779705117, + "grad_norm": 1.9559937715530396, + "learning_rate": 5.8443348324387066e-06, + "loss": 0.0022, + "step": 19615 + }, + { + "epoch": 2.126626192541197, + "grad_norm": 0.1579161137342453, + "learning_rate": 5.843609458871319e-06, + "loss": 0.0006, + "step": 19616 + }, + { + "epoch": 2.1267346053772767, + "grad_norm": 6.219748497009277, + "learning_rate": 5.842884085303932e-06, + "loss": 0.0605, + "step": 19617 + }, + { + "epoch": 2.1268430182133566, + "grad_norm": 0.5337584018707275, + "learning_rate": 5.8421587117365455e-06, + "loss": 0.0016, + "step": 19618 + }, + { + "epoch": 2.126951431049436, + "grad_norm": 1.1415497064590454, + "learning_rate": 5.841433338169158e-06, + "loss": 0.0036, + "step": 19619 + }, + { + "epoch": 2.127059843885516, + "grad_norm": 0.11270876228809357, + "learning_rate": 5.840707964601771e-06, + "loss": 0.0002, + "step": 19620 + }, + { + "epoch": 2.1271682567215957, + "grad_norm": 3.825568437576294, + "learning_rate": 5.839982591034384e-06, + "loss": 0.0061, + "step": 19621 + }, + { + "epoch": 2.1272766695576757, + "grad_norm": 1.6047368049621582, + "learning_rate": 5.839257217466996e-06, + "loss": 0.0121, + "step": 19622 + }, + { + "epoch": 2.1273850823937552, + "grad_norm": 0.10302332788705826, + "learning_rate": 5.838531843899609e-06, + "loss": 0.0003, + "step": 19623 + }, + { + "epoch": 2.127493495229835, + "grad_norm": 4.218433856964111, + "learning_rate": 5.837806470332221e-06, + "loss": 0.0085, + "step": 19624 + }, + { + "epoch": 2.127601908065915, + "grad_norm": 0.053287964314222336, + "learning_rate": 5.837081096764834e-06, + "loss": 0.0002, + "step": 19625 + }, + { + "epoch": 2.1277103209019947, + "grad_norm": 0.23735332489013672, + "learning_rate": 5.836355723197448e-06, + "loss": 0.0011, + "step": 19626 + }, + { + "epoch": 2.1278187337380747, + "grad_norm": 0.06698185950517654, + "learning_rate": 5.835630349630061e-06, + "loss": 0.0005, + "step": 19627 + }, + { + "epoch": 2.1279271465741543, + "grad_norm": 0.018410691991448402, + "learning_rate": 5.834904976062673e-06, + "loss": 0.0001, + "step": 19628 + }, + { + "epoch": 2.1280355594102343, + "grad_norm": 9.37020206451416, + "learning_rate": 5.834179602495286e-06, + "loss": 0.0159, + "step": 19629 + }, + { + "epoch": 2.128143972246314, + "grad_norm": 3.4394373893737793, + "learning_rate": 5.833454228927898e-06, + "loss": 0.0047, + "step": 19630 + }, + { + "epoch": 2.128252385082394, + "grad_norm": 1.4851738214492798, + "learning_rate": 5.832728855360511e-06, + "loss": 0.0065, + "step": 19631 + }, + { + "epoch": 2.1283607979184733, + "grad_norm": 4.754059791564941, + "learning_rate": 5.832003481793124e-06, + "loss": 0.0052, + "step": 19632 + }, + { + "epoch": 2.1284692107545533, + "grad_norm": 2.4023752212524414, + "learning_rate": 5.831278108225736e-06, + "loss": 0.0051, + "step": 19633 + }, + { + "epoch": 2.1285776235906333, + "grad_norm": 0.07227041572332382, + "learning_rate": 5.830552734658349e-06, + "loss": 0.0003, + "step": 19634 + }, + { + "epoch": 2.128686036426713, + "grad_norm": 4.420975208282471, + "learning_rate": 5.829827361090963e-06, + "loss": 0.0122, + "step": 19635 + }, + { + "epoch": 2.128794449262793, + "grad_norm": 0.08418333530426025, + "learning_rate": 5.829101987523575e-06, + "loss": 0.0004, + "step": 19636 + }, + { + "epoch": 2.1289028620988724, + "grad_norm": 0.10581695288419724, + "learning_rate": 5.828376613956188e-06, + "loss": 0.0004, + "step": 19637 + }, + { + "epoch": 2.1290112749349523, + "grad_norm": 0.03503155708312988, + "learning_rate": 5.827651240388801e-06, + "loss": 0.0003, + "step": 19638 + }, + { + "epoch": 2.129119687771032, + "grad_norm": 1.150239109992981, + "learning_rate": 5.826925866821413e-06, + "loss": 0.0042, + "step": 19639 + }, + { + "epoch": 2.129228100607112, + "grad_norm": 5.118253231048584, + "learning_rate": 5.826200493254026e-06, + "loss": 0.0297, + "step": 19640 + }, + { + "epoch": 2.129336513443192, + "grad_norm": 2.2745320796966553, + "learning_rate": 5.825475119686639e-06, + "loss": 0.0042, + "step": 19641 + }, + { + "epoch": 2.1294449262792714, + "grad_norm": 4.255001544952393, + "learning_rate": 5.824749746119251e-06, + "loss": 0.005, + "step": 19642 + }, + { + "epoch": 2.1295533391153514, + "grad_norm": 2.8538386821746826, + "learning_rate": 5.824024372551864e-06, + "loss": 0.0202, + "step": 19643 + }, + { + "epoch": 2.129661751951431, + "grad_norm": 0.1928347647190094, + "learning_rate": 5.823298998984478e-06, + "loss": 0.0008, + "step": 19644 + }, + { + "epoch": 2.129770164787511, + "grad_norm": 3.933119535446167, + "learning_rate": 5.82257362541709e-06, + "loss": 0.024, + "step": 19645 + }, + { + "epoch": 2.1298785776235905, + "grad_norm": 3.0342793464660645, + "learning_rate": 5.821848251849703e-06, + "loss": 0.0327, + "step": 19646 + }, + { + "epoch": 2.1299869904596704, + "grad_norm": 0.19461314380168915, + "learning_rate": 5.821122878282316e-06, + "loss": 0.0009, + "step": 19647 + }, + { + "epoch": 2.1300954032957504, + "grad_norm": 0.25233322381973267, + "learning_rate": 5.8203975047149285e-06, + "loss": 0.0006, + "step": 19648 + }, + { + "epoch": 2.13020381613183, + "grad_norm": 1.932983636856079, + "learning_rate": 5.8196721311475415e-06, + "loss": 0.0056, + "step": 19649 + }, + { + "epoch": 2.13031222896791, + "grad_norm": 0.23011702299118042, + "learning_rate": 5.818946757580154e-06, + "loss": 0.0006, + "step": 19650 + }, + { + "epoch": 2.1304206418039895, + "grad_norm": 1.6632630825042725, + "learning_rate": 5.818221384012767e-06, + "loss": 0.0019, + "step": 19651 + }, + { + "epoch": 2.1305290546400695, + "grad_norm": 0.16319769620895386, + "learning_rate": 5.8174960104453804e-06, + "loss": 0.0013, + "step": 19652 + }, + { + "epoch": 2.130637467476149, + "grad_norm": 13.831377983093262, + "learning_rate": 5.816770636877993e-06, + "loss": 0.0304, + "step": 19653 + }, + { + "epoch": 2.130745880312229, + "grad_norm": 1.2558974027633667, + "learning_rate": 5.8160452633106056e-06, + "loss": 0.0024, + "step": 19654 + }, + { + "epoch": 2.1308542931483085, + "grad_norm": 7.896440505981445, + "learning_rate": 5.8153198897432186e-06, + "loss": 0.0287, + "step": 19655 + }, + { + "epoch": 2.1309627059843885, + "grad_norm": 0.04505263268947601, + "learning_rate": 5.814594516175831e-06, + "loss": 0.0001, + "step": 19656 + }, + { + "epoch": 2.1310711188204685, + "grad_norm": 0.8446289300918579, + "learning_rate": 5.813869142608444e-06, + "loss": 0.0039, + "step": 19657 + }, + { + "epoch": 2.131179531656548, + "grad_norm": 0.018705196678638458, + "learning_rate": 5.813143769041057e-06, + "loss": 0.0001, + "step": 19658 + }, + { + "epoch": 2.131287944492628, + "grad_norm": 4.263066291809082, + "learning_rate": 5.812418395473669e-06, + "loss": 0.0079, + "step": 19659 + }, + { + "epoch": 2.1313963573287076, + "grad_norm": 2.70243239402771, + "learning_rate": 5.811693021906282e-06, + "loss": 0.0145, + "step": 19660 + }, + { + "epoch": 2.1315047701647876, + "grad_norm": 1.596971035003662, + "learning_rate": 5.810967648338896e-06, + "loss": 0.0043, + "step": 19661 + }, + { + "epoch": 2.131613183000867, + "grad_norm": 2.597722053527832, + "learning_rate": 5.810242274771508e-06, + "loss": 0.0284, + "step": 19662 + }, + { + "epoch": 2.131721595836947, + "grad_norm": 0.06528617441654205, + "learning_rate": 5.809516901204121e-06, + "loss": 0.0004, + "step": 19663 + }, + { + "epoch": 2.131830008673027, + "grad_norm": 2.067721128463745, + "learning_rate": 5.808791527636734e-06, + "loss": 0.0071, + "step": 19664 + }, + { + "epoch": 2.1319384215091066, + "grad_norm": 2.242870807647705, + "learning_rate": 5.808066154069346e-06, + "loss": 0.0145, + "step": 19665 + }, + { + "epoch": 2.1320468343451866, + "grad_norm": 0.23205214738845825, + "learning_rate": 5.807340780501959e-06, + "loss": 0.0005, + "step": 19666 + }, + { + "epoch": 2.132155247181266, + "grad_norm": 0.1654878407716751, + "learning_rate": 5.806615406934572e-06, + "loss": 0.0011, + "step": 19667 + }, + { + "epoch": 2.132263660017346, + "grad_norm": 2.1289210319519043, + "learning_rate": 5.805890033367184e-06, + "loss": 0.0023, + "step": 19668 + }, + { + "epoch": 2.1323720728534257, + "grad_norm": 5.707645416259766, + "learning_rate": 5.805164659799797e-06, + "loss": 0.0113, + "step": 19669 + }, + { + "epoch": 2.1324804856895057, + "grad_norm": 2.786818265914917, + "learning_rate": 5.804439286232411e-06, + "loss": 0.0129, + "step": 19670 + }, + { + "epoch": 2.1325888985255856, + "grad_norm": 8.91886043548584, + "learning_rate": 5.803713912665023e-06, + "loss": 0.0639, + "step": 19671 + }, + { + "epoch": 2.132697311361665, + "grad_norm": 3.2423346042633057, + "learning_rate": 5.802988539097636e-06, + "loss": 0.0241, + "step": 19672 + }, + { + "epoch": 2.132805724197745, + "grad_norm": 1.2994496822357178, + "learning_rate": 5.802263165530249e-06, + "loss": 0.0063, + "step": 19673 + }, + { + "epoch": 2.1329141370338247, + "grad_norm": 2.827897310256958, + "learning_rate": 5.801537791962861e-06, + "loss": 0.0052, + "step": 19674 + }, + { + "epoch": 2.1330225498699047, + "grad_norm": 1.8089507818222046, + "learning_rate": 5.800812418395474e-06, + "loss": 0.0028, + "step": 19675 + }, + { + "epoch": 2.1331309627059842, + "grad_norm": 0.7201588749885559, + "learning_rate": 5.800087044828086e-06, + "loss": 0.0027, + "step": 19676 + }, + { + "epoch": 2.133239375542064, + "grad_norm": 3.4989261627197266, + "learning_rate": 5.799361671260699e-06, + "loss": 0.0027, + "step": 19677 + }, + { + "epoch": 2.1333477883781438, + "grad_norm": 0.37544944882392883, + "learning_rate": 5.798636297693313e-06, + "loss": 0.0014, + "step": 19678 + }, + { + "epoch": 2.1334562012142237, + "grad_norm": 1.9511234760284424, + "learning_rate": 5.797910924125926e-06, + "loss": 0.0175, + "step": 19679 + }, + { + "epoch": 2.1335646140503037, + "grad_norm": 14.30266284942627, + "learning_rate": 5.797185550558538e-06, + "loss": 0.0115, + "step": 19680 + }, + { + "epoch": 2.1336730268863833, + "grad_norm": 2.1308858394622803, + "learning_rate": 5.796460176991151e-06, + "loss": 0.0054, + "step": 19681 + }, + { + "epoch": 2.1337814397224633, + "grad_norm": 0.04877961426973343, + "learning_rate": 5.795734803423763e-06, + "loss": 0.0002, + "step": 19682 + }, + { + "epoch": 2.133889852558543, + "grad_norm": 0.08198950439691544, + "learning_rate": 5.795009429856376e-06, + "loss": 0.0003, + "step": 19683 + }, + { + "epoch": 2.133998265394623, + "grad_norm": 0.3220725953578949, + "learning_rate": 5.794284056288989e-06, + "loss": 0.0014, + "step": 19684 + }, + { + "epoch": 2.1341066782307023, + "grad_norm": 0.8588458895683289, + "learning_rate": 5.7935586827216015e-06, + "loss": 0.001, + "step": 19685 + }, + { + "epoch": 2.1342150910667823, + "grad_norm": 0.9264843463897705, + "learning_rate": 5.7928333091542145e-06, + "loss": 0.0013, + "step": 19686 + }, + { + "epoch": 2.1343235039028623, + "grad_norm": 4.419605731964111, + "learning_rate": 5.792107935586828e-06, + "loss": 0.0188, + "step": 19687 + }, + { + "epoch": 2.134431916738942, + "grad_norm": 0.06482166051864624, + "learning_rate": 5.7913825620194405e-06, + "loss": 0.0004, + "step": 19688 + }, + { + "epoch": 2.134540329575022, + "grad_norm": 0.44498327374458313, + "learning_rate": 5.7906571884520535e-06, + "loss": 0.0003, + "step": 19689 + }, + { + "epoch": 2.1346487424111014, + "grad_norm": 1.4154695272445679, + "learning_rate": 5.7899318148846664e-06, + "loss": 0.0068, + "step": 19690 + }, + { + "epoch": 2.1347571552471813, + "grad_norm": 0.7359958291053772, + "learning_rate": 5.789206441317279e-06, + "loss": 0.001, + "step": 19691 + }, + { + "epoch": 2.134865568083261, + "grad_norm": 4.5247111320495605, + "learning_rate": 5.788481067749892e-06, + "loss": 0.0252, + "step": 19692 + }, + { + "epoch": 2.134973980919341, + "grad_norm": 6.3400559425354, + "learning_rate": 5.7877556941825046e-06, + "loss": 0.0301, + "step": 19693 + }, + { + "epoch": 2.135082393755421, + "grad_norm": 2.6076385974884033, + "learning_rate": 5.787030320615117e-06, + "loss": 0.0487, + "step": 19694 + }, + { + "epoch": 2.1351908065915004, + "grad_norm": 8.77540111541748, + "learning_rate": 5.7863049470477305e-06, + "loss": 0.0144, + "step": 19695 + }, + { + "epoch": 2.1352992194275804, + "grad_norm": 0.08170247822999954, + "learning_rate": 5.7855795734803435e-06, + "loss": 0.0005, + "step": 19696 + }, + { + "epoch": 2.13540763226366, + "grad_norm": 1.2083550691604614, + "learning_rate": 5.784854199912956e-06, + "loss": 0.0031, + "step": 19697 + }, + { + "epoch": 2.13551604509974, + "grad_norm": 2.757920980453491, + "learning_rate": 5.784128826345569e-06, + "loss": 0.0148, + "step": 19698 + }, + { + "epoch": 2.1356244579358195, + "grad_norm": 2.823730230331421, + "learning_rate": 5.783403452778182e-06, + "loss": 0.0102, + "step": 19699 + }, + { + "epoch": 2.1357328707718994, + "grad_norm": 2.0715110301971436, + "learning_rate": 5.782678079210794e-06, + "loss": 0.0155, + "step": 19700 + }, + { + "epoch": 2.135841283607979, + "grad_norm": 0.5688716769218445, + "learning_rate": 5.781952705643407e-06, + "loss": 0.0013, + "step": 19701 + }, + { + "epoch": 2.135949696444059, + "grad_norm": 0.37104332447052, + "learning_rate": 5.781227332076019e-06, + "loss": 0.0017, + "step": 19702 + }, + { + "epoch": 2.136058109280139, + "grad_norm": 0.03212011605501175, + "learning_rate": 5.780501958508632e-06, + "loss": 0.0002, + "step": 19703 + }, + { + "epoch": 2.1361665221162185, + "grad_norm": 1.10909104347229, + "learning_rate": 5.779776584941246e-06, + "loss": 0.0021, + "step": 19704 + }, + { + "epoch": 2.1362749349522985, + "grad_norm": 0.4603535234928131, + "learning_rate": 5.779051211373859e-06, + "loss": 0.0012, + "step": 19705 + }, + { + "epoch": 2.136383347788378, + "grad_norm": 4.375990390777588, + "learning_rate": 5.778325837806471e-06, + "loss": 0.0031, + "step": 19706 + }, + { + "epoch": 2.136491760624458, + "grad_norm": 1.9836628437042236, + "learning_rate": 5.777600464239084e-06, + "loss": 0.0033, + "step": 19707 + }, + { + "epoch": 2.1366001734605375, + "grad_norm": 5.287342071533203, + "learning_rate": 5.776875090671696e-06, + "loss": 0.0046, + "step": 19708 + }, + { + "epoch": 2.1367085862966175, + "grad_norm": 2.269042730331421, + "learning_rate": 5.776149717104309e-06, + "loss": 0.0245, + "step": 19709 + }, + { + "epoch": 2.1368169991326975, + "grad_norm": 0.9195126295089722, + "learning_rate": 5.775424343536922e-06, + "loss": 0.0057, + "step": 19710 + }, + { + "epoch": 2.136925411968777, + "grad_norm": 11.582008361816406, + "learning_rate": 5.774698969969534e-06, + "loss": 0.0172, + "step": 19711 + }, + { + "epoch": 2.137033824804857, + "grad_norm": 2.40091609954834, + "learning_rate": 5.773973596402147e-06, + "loss": 0.0132, + "step": 19712 + }, + { + "epoch": 2.1371422376409366, + "grad_norm": 0.0690389946103096, + "learning_rate": 5.773248222834761e-06, + "loss": 0.0003, + "step": 19713 + }, + { + "epoch": 2.1372506504770166, + "grad_norm": 0.027851000428199768, + "learning_rate": 5.772522849267373e-06, + "loss": 0.0002, + "step": 19714 + }, + { + "epoch": 2.137359063313096, + "grad_norm": 8.234052658081055, + "learning_rate": 5.771797475699986e-06, + "loss": 0.0217, + "step": 19715 + }, + { + "epoch": 2.137467476149176, + "grad_norm": 2.5956497192382812, + "learning_rate": 5.771072102132599e-06, + "loss": 0.0189, + "step": 19716 + }, + { + "epoch": 2.137575888985256, + "grad_norm": 1.0015811920166016, + "learning_rate": 5.770346728565211e-06, + "loss": 0.0036, + "step": 19717 + }, + { + "epoch": 2.1376843018213356, + "grad_norm": 0.04500436782836914, + "learning_rate": 5.769621354997824e-06, + "loss": 0.0001, + "step": 19718 + }, + { + "epoch": 2.1377927146574156, + "grad_norm": 3.9697015285491943, + "learning_rate": 5.768895981430437e-06, + "loss": 0.0064, + "step": 19719 + }, + { + "epoch": 2.137901127493495, + "grad_norm": 3.089132785797119, + "learning_rate": 5.768170607863049e-06, + "loss": 0.0516, + "step": 19720 + }, + { + "epoch": 2.138009540329575, + "grad_norm": 0.03944950923323631, + "learning_rate": 5.767445234295663e-06, + "loss": 0.0002, + "step": 19721 + }, + { + "epoch": 2.1381179531656547, + "grad_norm": 2.8327741622924805, + "learning_rate": 5.766719860728276e-06, + "loss": 0.011, + "step": 19722 + }, + { + "epoch": 2.1382263660017347, + "grad_norm": 2.644796371459961, + "learning_rate": 5.765994487160888e-06, + "loss": 0.019, + "step": 19723 + }, + { + "epoch": 2.138334778837814, + "grad_norm": 0.281480073928833, + "learning_rate": 5.765269113593501e-06, + "loss": 0.0006, + "step": 19724 + }, + { + "epoch": 2.138443191673894, + "grad_norm": 1.9246200323104858, + "learning_rate": 5.764543740026114e-06, + "loss": 0.0035, + "step": 19725 + }, + { + "epoch": 2.138551604509974, + "grad_norm": 0.4662042558193207, + "learning_rate": 5.7638183664587265e-06, + "loss": 0.0012, + "step": 19726 + }, + { + "epoch": 2.1386600173460537, + "grad_norm": 2.4867091178894043, + "learning_rate": 5.7630929928913395e-06, + "loss": 0.012, + "step": 19727 + }, + { + "epoch": 2.1387684301821337, + "grad_norm": 11.670679092407227, + "learning_rate": 5.762367619323952e-06, + "loss": 0.0195, + "step": 19728 + }, + { + "epoch": 2.1388768430182132, + "grad_norm": 0.3131524920463562, + "learning_rate": 5.761642245756565e-06, + "loss": 0.001, + "step": 19729 + }, + { + "epoch": 2.138985255854293, + "grad_norm": 5.266709804534912, + "learning_rate": 5.7609168721891784e-06, + "loss": 0.0235, + "step": 19730 + }, + { + "epoch": 2.1390936686903728, + "grad_norm": 7.688248157501221, + "learning_rate": 5.7601914986217914e-06, + "loss": 0.0174, + "step": 19731 + }, + { + "epoch": 2.1392020815264527, + "grad_norm": 0.2317696213722229, + "learning_rate": 5.7594661250544036e-06, + "loss": 0.001, + "step": 19732 + }, + { + "epoch": 2.1393104943625323, + "grad_norm": 0.690538227558136, + "learning_rate": 5.7587407514870166e-06, + "loss": 0.001, + "step": 19733 + }, + { + "epoch": 2.1394189071986123, + "grad_norm": 0.02625308744609356, + "learning_rate": 5.758015377919629e-06, + "loss": 0.0002, + "step": 19734 + }, + { + "epoch": 2.1395273200346923, + "grad_norm": 0.6592370271682739, + "learning_rate": 5.757290004352242e-06, + "loss": 0.0007, + "step": 19735 + }, + { + "epoch": 2.139635732870772, + "grad_norm": 0.03988271951675415, + "learning_rate": 5.756564630784855e-06, + "loss": 0.0004, + "step": 19736 + }, + { + "epoch": 2.139744145706852, + "grad_norm": 0.15137237310409546, + "learning_rate": 5.755839257217467e-06, + "loss": 0.0006, + "step": 19737 + }, + { + "epoch": 2.1398525585429313, + "grad_norm": 0.04596845805644989, + "learning_rate": 5.75511388365008e-06, + "loss": 0.0002, + "step": 19738 + }, + { + "epoch": 2.1399609713790113, + "grad_norm": 0.7652718424797058, + "learning_rate": 5.754388510082694e-06, + "loss": 0.0026, + "step": 19739 + }, + { + "epoch": 2.1400693842150913, + "grad_norm": 1.4559264183044434, + "learning_rate": 5.753663136515306e-06, + "loss": 0.0073, + "step": 19740 + }, + { + "epoch": 2.140177797051171, + "grad_norm": 0.036580462008714676, + "learning_rate": 5.752937762947919e-06, + "loss": 0.0003, + "step": 19741 + }, + { + "epoch": 2.140286209887251, + "grad_norm": 0.03278676047921181, + "learning_rate": 5.752212389380532e-06, + "loss": 0.0001, + "step": 19742 + }, + { + "epoch": 2.1403946227233304, + "grad_norm": 0.1271015703678131, + "learning_rate": 5.751487015813144e-06, + "loss": 0.0004, + "step": 19743 + }, + { + "epoch": 2.1405030355594103, + "grad_norm": 2.4133143424987793, + "learning_rate": 5.750761642245757e-06, + "loss": 0.0431, + "step": 19744 + }, + { + "epoch": 2.14061144839549, + "grad_norm": 2.955178737640381, + "learning_rate": 5.75003626867837e-06, + "loss": 0.0054, + "step": 19745 + }, + { + "epoch": 2.14071986123157, + "grad_norm": 0.13497014343738556, + "learning_rate": 5.749310895110982e-06, + "loss": 0.0003, + "step": 19746 + }, + { + "epoch": 2.1408282740676494, + "grad_norm": 2.200404405593872, + "learning_rate": 5.748585521543596e-06, + "loss": 0.0077, + "step": 19747 + }, + { + "epoch": 2.1409366869037294, + "grad_norm": 0.260225385427475, + "learning_rate": 5.747860147976209e-06, + "loss": 0.0012, + "step": 19748 + }, + { + "epoch": 2.1410450997398094, + "grad_norm": 1.0295796394348145, + "learning_rate": 5.747134774408821e-06, + "loss": 0.0042, + "step": 19749 + }, + { + "epoch": 2.141153512575889, + "grad_norm": 0.6595892906188965, + "learning_rate": 5.746409400841434e-06, + "loss": 0.0015, + "step": 19750 + }, + { + "epoch": 2.141261925411969, + "grad_norm": 0.11548660695552826, + "learning_rate": 5.745684027274047e-06, + "loss": 0.0006, + "step": 19751 + }, + { + "epoch": 2.1413703382480485, + "grad_norm": 0.2734123766422272, + "learning_rate": 5.744958653706659e-06, + "loss": 0.0006, + "step": 19752 + }, + { + "epoch": 2.1414787510841284, + "grad_norm": 0.06443721055984497, + "learning_rate": 5.744233280139272e-06, + "loss": 0.0002, + "step": 19753 + }, + { + "epoch": 2.141587163920208, + "grad_norm": 0.023822791874408722, + "learning_rate": 5.743507906571884e-06, + "loss": 0.0002, + "step": 19754 + }, + { + "epoch": 2.141695576756288, + "grad_norm": 4.250552654266357, + "learning_rate": 5.742782533004497e-06, + "loss": 0.0095, + "step": 19755 + }, + { + "epoch": 2.1418039895923675, + "grad_norm": 0.1432420015335083, + "learning_rate": 5.742057159437111e-06, + "loss": 0.0005, + "step": 19756 + }, + { + "epoch": 2.1419124024284475, + "grad_norm": 0.4930631220340729, + "learning_rate": 5.741331785869724e-06, + "loss": 0.002, + "step": 19757 + }, + { + "epoch": 2.1420208152645275, + "grad_norm": 1.132720947265625, + "learning_rate": 5.740606412302336e-06, + "loss": 0.0048, + "step": 19758 + }, + { + "epoch": 2.142129228100607, + "grad_norm": 0.1717052012681961, + "learning_rate": 5.739881038734949e-06, + "loss": 0.0005, + "step": 19759 + }, + { + "epoch": 2.142237640936687, + "grad_norm": 0.030165493488311768, + "learning_rate": 5.739155665167561e-06, + "loss": 0.0001, + "step": 19760 + }, + { + "epoch": 2.1423460537727665, + "grad_norm": 4.572456359863281, + "learning_rate": 5.738430291600174e-06, + "loss": 0.0205, + "step": 19761 + }, + { + "epoch": 2.1424544666088465, + "grad_norm": 0.04425022378563881, + "learning_rate": 5.737704918032787e-06, + "loss": 0.0003, + "step": 19762 + }, + { + "epoch": 2.142562879444926, + "grad_norm": 0.15944816172122955, + "learning_rate": 5.7369795444653995e-06, + "loss": 0.0003, + "step": 19763 + }, + { + "epoch": 2.142671292281006, + "grad_norm": 0.02524840272963047, + "learning_rate": 5.7362541708980125e-06, + "loss": 0.0002, + "step": 19764 + }, + { + "epoch": 2.142779705117086, + "grad_norm": 5.811352252960205, + "learning_rate": 5.735528797330626e-06, + "loss": 0.0078, + "step": 19765 + }, + { + "epoch": 2.1428881179531656, + "grad_norm": 0.4337367117404938, + "learning_rate": 5.7348034237632385e-06, + "loss": 0.001, + "step": 19766 + }, + { + "epoch": 2.1429965307892456, + "grad_norm": 0.12914453446865082, + "learning_rate": 5.7340780501958515e-06, + "loss": 0.0005, + "step": 19767 + }, + { + "epoch": 2.143104943625325, + "grad_norm": 0.043062444776296616, + "learning_rate": 5.7333526766284645e-06, + "loss": 0.0002, + "step": 19768 + }, + { + "epoch": 2.143213356461405, + "grad_norm": 0.03322603181004524, + "learning_rate": 5.732627303061077e-06, + "loss": 0.0001, + "step": 19769 + }, + { + "epoch": 2.1433217692974846, + "grad_norm": 3.7497143745422363, + "learning_rate": 5.73190192949369e-06, + "loss": 0.0339, + "step": 19770 + }, + { + "epoch": 2.1434301821335646, + "grad_norm": 4.516135215759277, + "learning_rate": 5.7311765559263026e-06, + "loss": 0.0041, + "step": 19771 + }, + { + "epoch": 2.1435385949696446, + "grad_norm": 0.08687383681535721, + "learning_rate": 5.730451182358915e-06, + "loss": 0.0004, + "step": 19772 + }, + { + "epoch": 2.143647007805724, + "grad_norm": 2.9335408210754395, + "learning_rate": 5.7297258087915286e-06, + "loss": 0.0111, + "step": 19773 + }, + { + "epoch": 2.143755420641804, + "grad_norm": 4.366841793060303, + "learning_rate": 5.7290004352241415e-06, + "loss": 0.0104, + "step": 19774 + }, + { + "epoch": 2.1438638334778837, + "grad_norm": 1.3127535581588745, + "learning_rate": 5.728275061656754e-06, + "loss": 0.0022, + "step": 19775 + }, + { + "epoch": 2.1439722463139637, + "grad_norm": 1.6478285789489746, + "learning_rate": 5.727549688089367e-06, + "loss": 0.0026, + "step": 19776 + }, + { + "epoch": 2.144080659150043, + "grad_norm": 2.176504373550415, + "learning_rate": 5.72682431452198e-06, + "loss": 0.0179, + "step": 19777 + }, + { + "epoch": 2.144189071986123, + "grad_norm": 0.03111339546740055, + "learning_rate": 5.726098940954592e-06, + "loss": 0.0002, + "step": 19778 + }, + { + "epoch": 2.1442974848222027, + "grad_norm": 0.010559496469795704, + "learning_rate": 5.725373567387205e-06, + "loss": 0.0, + "step": 19779 + }, + { + "epoch": 2.1444058976582827, + "grad_norm": 2.6708009243011475, + "learning_rate": 5.724648193819817e-06, + "loss": 0.0326, + "step": 19780 + }, + { + "epoch": 2.1445143104943627, + "grad_norm": 0.01830609329044819, + "learning_rate": 5.72392282025243e-06, + "loss": 0.0001, + "step": 19781 + }, + { + "epoch": 2.1446227233304422, + "grad_norm": 2.1267032623291016, + "learning_rate": 5.723197446685044e-06, + "loss": 0.0033, + "step": 19782 + }, + { + "epoch": 2.1447311361665222, + "grad_norm": 4.002037048339844, + "learning_rate": 5.722472073117657e-06, + "loss": 0.0188, + "step": 19783 + }, + { + "epoch": 2.1448395490026018, + "grad_norm": 0.025329075753688812, + "learning_rate": 5.721746699550269e-06, + "loss": 0.0001, + "step": 19784 + }, + { + "epoch": 2.1449479618386817, + "grad_norm": 6.345882892608643, + "learning_rate": 5.721021325982882e-06, + "loss": 0.006, + "step": 19785 + }, + { + "epoch": 2.1450563746747613, + "grad_norm": 0.10389330983161926, + "learning_rate": 5.720295952415494e-06, + "loss": 0.0005, + "step": 19786 + }, + { + "epoch": 2.1451647875108413, + "grad_norm": 0.18827685713768005, + "learning_rate": 5.719570578848107e-06, + "loss": 0.0005, + "step": 19787 + }, + { + "epoch": 2.1452732003469213, + "grad_norm": 2.3743419647216797, + "learning_rate": 5.71884520528072e-06, + "loss": 0.0016, + "step": 19788 + }, + { + "epoch": 2.145381613183001, + "grad_norm": 10.375718116760254, + "learning_rate": 5.718119831713332e-06, + "loss": 0.0058, + "step": 19789 + }, + { + "epoch": 2.145490026019081, + "grad_norm": 0.07000277191400528, + "learning_rate": 5.717394458145946e-06, + "loss": 0.0006, + "step": 19790 + }, + { + "epoch": 2.1455984388551603, + "grad_norm": 0.9185609221458435, + "learning_rate": 5.716669084578559e-06, + "loss": 0.0429, + "step": 19791 + }, + { + "epoch": 2.1457068516912403, + "grad_norm": 6.855888366699219, + "learning_rate": 5.715943711011171e-06, + "loss": 0.0131, + "step": 19792 + }, + { + "epoch": 2.14581526452732, + "grad_norm": 2.5098209381103516, + "learning_rate": 5.715218337443784e-06, + "loss": 0.0383, + "step": 19793 + }, + { + "epoch": 2.1459236773634, + "grad_norm": 0.023207196965813637, + "learning_rate": 5.714492963876397e-06, + "loss": 0.0001, + "step": 19794 + }, + { + "epoch": 2.14603209019948, + "grad_norm": 4.360188961029053, + "learning_rate": 5.713767590309009e-06, + "loss": 0.0198, + "step": 19795 + }, + { + "epoch": 2.1461405030355594, + "grad_norm": 0.11779629439115524, + "learning_rate": 5.713042216741622e-06, + "loss": 0.0006, + "step": 19796 + }, + { + "epoch": 2.1462489158716394, + "grad_norm": 3.564255714416504, + "learning_rate": 5.712316843174235e-06, + "loss": 0.008, + "step": 19797 + }, + { + "epoch": 2.146357328707719, + "grad_norm": 2.778287172317505, + "learning_rate": 5.711591469606847e-06, + "loss": 0.0179, + "step": 19798 + }, + { + "epoch": 2.146465741543799, + "grad_norm": 0.4671679735183716, + "learning_rate": 5.710866096039461e-06, + "loss": 0.0002, + "step": 19799 + }, + { + "epoch": 2.1465741543798784, + "grad_norm": 4.909853458404541, + "learning_rate": 5.710140722472074e-06, + "loss": 0.0804, + "step": 19800 + }, + { + "epoch": 2.1466825672159584, + "grad_norm": 0.08127504587173462, + "learning_rate": 5.709415348904686e-06, + "loss": 0.0004, + "step": 19801 + }, + { + "epoch": 2.146790980052038, + "grad_norm": 1.268305778503418, + "learning_rate": 5.708689975337299e-06, + "loss": 0.0044, + "step": 19802 + }, + { + "epoch": 2.146899392888118, + "grad_norm": 0.18900513648986816, + "learning_rate": 5.707964601769912e-06, + "loss": 0.0008, + "step": 19803 + }, + { + "epoch": 2.147007805724198, + "grad_norm": 0.06845526397228241, + "learning_rate": 5.7072392282025245e-06, + "loss": 0.0007, + "step": 19804 + }, + { + "epoch": 2.1471162185602775, + "grad_norm": 0.10449780523777008, + "learning_rate": 5.7065138546351375e-06, + "loss": 0.0005, + "step": 19805 + }, + { + "epoch": 2.1472246313963574, + "grad_norm": 0.6977389454841614, + "learning_rate": 5.70578848106775e-06, + "loss": 0.0016, + "step": 19806 + }, + { + "epoch": 2.147333044232437, + "grad_norm": 0.24108146131038666, + "learning_rate": 5.705063107500363e-06, + "loss": 0.0012, + "step": 19807 + }, + { + "epoch": 2.147441457068517, + "grad_norm": 0.09700272977352142, + "learning_rate": 5.7043377339329764e-06, + "loss": 0.0007, + "step": 19808 + }, + { + "epoch": 2.1475498699045965, + "grad_norm": 0.06929830461740494, + "learning_rate": 5.7036123603655894e-06, + "loss": 0.0002, + "step": 19809 + }, + { + "epoch": 2.1476582827406765, + "grad_norm": 3.6615195274353027, + "learning_rate": 5.702886986798202e-06, + "loss": 0.0054, + "step": 19810 + }, + { + "epoch": 2.1477666955767565, + "grad_norm": 0.010763317346572876, + "learning_rate": 5.7021616132308146e-06, + "loss": 0.0001, + "step": 19811 + }, + { + "epoch": 2.147875108412836, + "grad_norm": 0.2036326378583908, + "learning_rate": 5.701436239663427e-06, + "loss": 0.0008, + "step": 19812 + }, + { + "epoch": 2.147983521248916, + "grad_norm": 4.493849754333496, + "learning_rate": 5.70071086609604e-06, + "loss": 0.0083, + "step": 19813 + }, + { + "epoch": 2.1480919340849955, + "grad_norm": 0.04804937168955803, + "learning_rate": 5.699985492528653e-06, + "loss": 0.0003, + "step": 19814 + }, + { + "epoch": 2.1482003469210755, + "grad_norm": 3.7605035305023193, + "learning_rate": 5.699260118961265e-06, + "loss": 0.0094, + "step": 19815 + }, + { + "epoch": 2.148308759757155, + "grad_norm": 0.04438960924744606, + "learning_rate": 5.698534745393879e-06, + "loss": 0.0001, + "step": 19816 + }, + { + "epoch": 2.148417172593235, + "grad_norm": 1.1606789827346802, + "learning_rate": 5.697809371826492e-06, + "loss": 0.0009, + "step": 19817 + }, + { + "epoch": 2.148525585429315, + "grad_norm": 3.3367583751678467, + "learning_rate": 5.697083998259104e-06, + "loss": 0.0082, + "step": 19818 + }, + { + "epoch": 2.1486339982653946, + "grad_norm": 2.6247167587280273, + "learning_rate": 5.696358624691717e-06, + "loss": 0.0403, + "step": 19819 + }, + { + "epoch": 2.1487424111014746, + "grad_norm": 0.2852841317653656, + "learning_rate": 5.69563325112433e-06, + "loss": 0.0014, + "step": 19820 + }, + { + "epoch": 2.148850823937554, + "grad_norm": 0.4455898702144623, + "learning_rate": 5.694907877556942e-06, + "loss": 0.001, + "step": 19821 + }, + { + "epoch": 2.148959236773634, + "grad_norm": 4.324100971221924, + "learning_rate": 5.694182503989555e-06, + "loss": 0.0255, + "step": 19822 + }, + { + "epoch": 2.1490676496097136, + "grad_norm": 0.07867651432752609, + "learning_rate": 5.693457130422168e-06, + "loss": 0.0004, + "step": 19823 + }, + { + "epoch": 2.1491760624457936, + "grad_norm": 8.847952842712402, + "learning_rate": 5.69273175685478e-06, + "loss": 0.0565, + "step": 19824 + }, + { + "epoch": 2.149284475281873, + "grad_norm": 0.24572306871414185, + "learning_rate": 5.692006383287394e-06, + "loss": 0.0004, + "step": 19825 + }, + { + "epoch": 2.149392888117953, + "grad_norm": 8.337120056152344, + "learning_rate": 5.691281009720007e-06, + "loss": 0.0076, + "step": 19826 + }, + { + "epoch": 2.149501300954033, + "grad_norm": 0.10976522415876389, + "learning_rate": 5.690555636152619e-06, + "loss": 0.0005, + "step": 19827 + }, + { + "epoch": 2.1496097137901127, + "grad_norm": 0.011670464649796486, + "learning_rate": 5.689830262585232e-06, + "loss": 0.0001, + "step": 19828 + }, + { + "epoch": 2.1497181266261927, + "grad_norm": 2.612131118774414, + "learning_rate": 5.689104889017845e-06, + "loss": 0.0047, + "step": 19829 + }, + { + "epoch": 2.149826539462272, + "grad_norm": 0.005283951759338379, + "learning_rate": 5.688379515450457e-06, + "loss": 0.0, + "step": 19830 + }, + { + "epoch": 2.149934952298352, + "grad_norm": 8.133138656616211, + "learning_rate": 5.68765414188307e-06, + "loss": 0.011, + "step": 19831 + }, + { + "epoch": 2.1500433651344317, + "grad_norm": 1.8669984340667725, + "learning_rate": 5.686928768315682e-06, + "loss": 0.0112, + "step": 19832 + }, + { + "epoch": 2.1501517779705117, + "grad_norm": 2.8045449256896973, + "learning_rate": 5.686203394748295e-06, + "loss": 0.0057, + "step": 19833 + }, + { + "epoch": 2.1502601908065917, + "grad_norm": 5.377172946929932, + "learning_rate": 5.685478021180909e-06, + "loss": 0.0527, + "step": 19834 + }, + { + "epoch": 2.1503686036426712, + "grad_norm": 4.01722526550293, + "learning_rate": 5.684752647613522e-06, + "loss": 0.0077, + "step": 19835 + }, + { + "epoch": 2.1504770164787512, + "grad_norm": 0.10160455107688904, + "learning_rate": 5.684027274046134e-06, + "loss": 0.0004, + "step": 19836 + }, + { + "epoch": 2.1505854293148308, + "grad_norm": 0.2592533230781555, + "learning_rate": 5.683301900478747e-06, + "loss": 0.0006, + "step": 19837 + }, + { + "epoch": 2.1506938421509108, + "grad_norm": 0.05729679390788078, + "learning_rate": 5.682576526911359e-06, + "loss": 0.0002, + "step": 19838 + }, + { + "epoch": 2.1508022549869903, + "grad_norm": 0.19894817471504211, + "learning_rate": 5.681851153343972e-06, + "loss": 0.0005, + "step": 19839 + }, + { + "epoch": 2.1509106678230703, + "grad_norm": 5.7025251388549805, + "learning_rate": 5.681125779776585e-06, + "loss": 0.0443, + "step": 19840 + }, + { + "epoch": 2.1510190806591503, + "grad_norm": 1.288241982460022, + "learning_rate": 5.6804004062091975e-06, + "loss": 0.0251, + "step": 19841 + }, + { + "epoch": 2.15112749349523, + "grad_norm": 0.26475951075553894, + "learning_rate": 5.679675032641811e-06, + "loss": 0.0018, + "step": 19842 + }, + { + "epoch": 2.15123590633131, + "grad_norm": 1.2785786390304565, + "learning_rate": 5.678949659074424e-06, + "loss": 0.0036, + "step": 19843 + }, + { + "epoch": 2.1513443191673893, + "grad_norm": 6.109763145446777, + "learning_rate": 5.6782242855070365e-06, + "loss": 0.0078, + "step": 19844 + }, + { + "epoch": 2.1514527320034693, + "grad_norm": 10.167523384094238, + "learning_rate": 5.6774989119396495e-06, + "loss": 0.0049, + "step": 19845 + }, + { + "epoch": 2.151561144839549, + "grad_norm": 0.2945234179496765, + "learning_rate": 5.6767735383722625e-06, + "loss": 0.0009, + "step": 19846 + }, + { + "epoch": 2.151669557675629, + "grad_norm": 8.818950653076172, + "learning_rate": 5.676048164804875e-06, + "loss": 0.0068, + "step": 19847 + }, + { + "epoch": 2.1517779705117084, + "grad_norm": 0.3568745255470276, + "learning_rate": 5.675322791237488e-06, + "loss": 0.0004, + "step": 19848 + }, + { + "epoch": 2.1518863833477884, + "grad_norm": 0.06659932434558868, + "learning_rate": 5.674597417670101e-06, + "loss": 0.0004, + "step": 19849 + }, + { + "epoch": 2.1519947961838684, + "grad_norm": 2.228907585144043, + "learning_rate": 5.673872044102713e-06, + "loss": 0.0037, + "step": 19850 + }, + { + "epoch": 2.152103209019948, + "grad_norm": 0.5369800925254822, + "learning_rate": 5.6731466705353266e-06, + "loss": 0.0014, + "step": 19851 + }, + { + "epoch": 2.152211621856028, + "grad_norm": 0.033966064453125, + "learning_rate": 5.6724212969679396e-06, + "loss": 0.0002, + "step": 19852 + }, + { + "epoch": 2.1523200346921074, + "grad_norm": 0.07771220058202744, + "learning_rate": 5.671695923400552e-06, + "loss": 0.0008, + "step": 19853 + }, + { + "epoch": 2.1524284475281874, + "grad_norm": 3.4429879188537598, + "learning_rate": 5.670970549833165e-06, + "loss": 0.0332, + "step": 19854 + }, + { + "epoch": 2.152536860364267, + "grad_norm": 0.13239075243473053, + "learning_rate": 5.670245176265778e-06, + "loss": 0.0006, + "step": 19855 + }, + { + "epoch": 2.152645273200347, + "grad_norm": 0.030316567048430443, + "learning_rate": 5.66951980269839e-06, + "loss": 0.0002, + "step": 19856 + }, + { + "epoch": 2.152753686036427, + "grad_norm": 4.098126411437988, + "learning_rate": 5.668794429131003e-06, + "loss": 0.0017, + "step": 19857 + }, + { + "epoch": 2.1528620988725065, + "grad_norm": 4.063714981079102, + "learning_rate": 5.668069055563615e-06, + "loss": 0.0016, + "step": 19858 + }, + { + "epoch": 2.1529705117085864, + "grad_norm": 0.020195040851831436, + "learning_rate": 5.667343681996228e-06, + "loss": 0.0002, + "step": 19859 + }, + { + "epoch": 2.153078924544666, + "grad_norm": 4.632701396942139, + "learning_rate": 5.666618308428842e-06, + "loss": 0.0332, + "step": 19860 + }, + { + "epoch": 2.153187337380746, + "grad_norm": 1.9733989238739014, + "learning_rate": 5.665892934861455e-06, + "loss": 0.0083, + "step": 19861 + }, + { + "epoch": 2.1532957502168255, + "grad_norm": 0.9578573107719421, + "learning_rate": 5.665167561294067e-06, + "loss": 0.0026, + "step": 19862 + }, + { + "epoch": 2.1534041630529055, + "grad_norm": 1.1705073118209839, + "learning_rate": 5.66444218772668e-06, + "loss": 0.006, + "step": 19863 + }, + { + "epoch": 2.1535125758889855, + "grad_norm": 0.17355695366859436, + "learning_rate": 5.663716814159292e-06, + "loss": 0.0008, + "step": 19864 + }, + { + "epoch": 2.153620988725065, + "grad_norm": 0.6114488840103149, + "learning_rate": 5.662991440591905e-06, + "loss": 0.0024, + "step": 19865 + }, + { + "epoch": 2.153729401561145, + "grad_norm": 0.8990299105644226, + "learning_rate": 5.662266067024518e-06, + "loss": 0.0033, + "step": 19866 + }, + { + "epoch": 2.1538378143972245, + "grad_norm": 0.6683996915817261, + "learning_rate": 5.66154069345713e-06, + "loss": 0.0016, + "step": 19867 + }, + { + "epoch": 2.1539462272333045, + "grad_norm": 5.326638698577881, + "learning_rate": 5.660815319889744e-06, + "loss": 0.0107, + "step": 19868 + }, + { + "epoch": 2.154054640069384, + "grad_norm": 0.12067290395498276, + "learning_rate": 5.660089946322357e-06, + "loss": 0.0006, + "step": 19869 + }, + { + "epoch": 2.154163052905464, + "grad_norm": 0.033057793974876404, + "learning_rate": 5.659364572754969e-06, + "loss": 0.0003, + "step": 19870 + }, + { + "epoch": 2.1542714657415436, + "grad_norm": 5.062098979949951, + "learning_rate": 5.658639199187582e-06, + "loss": 0.0131, + "step": 19871 + }, + { + "epoch": 2.1543798785776236, + "grad_norm": 0.06465044617652893, + "learning_rate": 5.657913825620195e-06, + "loss": 0.0007, + "step": 19872 + }, + { + "epoch": 2.1544882914137036, + "grad_norm": 0.0188591368496418, + "learning_rate": 5.657188452052807e-06, + "loss": 0.0001, + "step": 19873 + }, + { + "epoch": 2.154596704249783, + "grad_norm": 2.663572072982788, + "learning_rate": 5.65646307848542e-06, + "loss": 0.0029, + "step": 19874 + }, + { + "epoch": 2.154705117085863, + "grad_norm": 0.01973140798509121, + "learning_rate": 5.655737704918033e-06, + "loss": 0.0002, + "step": 19875 + }, + { + "epoch": 2.1548135299219426, + "grad_norm": 13.558332443237305, + "learning_rate": 5.655012331350645e-06, + "loss": 0.0111, + "step": 19876 + }, + { + "epoch": 2.1549219427580226, + "grad_norm": 0.0737745389342308, + "learning_rate": 5.654286957783259e-06, + "loss": 0.0006, + "step": 19877 + }, + { + "epoch": 2.155030355594102, + "grad_norm": 0.022163433954119682, + "learning_rate": 5.653561584215872e-06, + "loss": 0.0001, + "step": 19878 + }, + { + "epoch": 2.155138768430182, + "grad_norm": 1.8059715032577515, + "learning_rate": 5.652836210648484e-06, + "loss": 0.0032, + "step": 19879 + }, + { + "epoch": 2.1552471812662617, + "grad_norm": 1.9241652488708496, + "learning_rate": 5.652110837081097e-06, + "loss": 0.0066, + "step": 19880 + }, + { + "epoch": 2.1553555941023417, + "grad_norm": 24.446884155273438, + "learning_rate": 5.65138546351371e-06, + "loss": 0.0254, + "step": 19881 + }, + { + "epoch": 2.1554640069384217, + "grad_norm": 4.745500564575195, + "learning_rate": 5.6506600899463225e-06, + "loss": 0.0151, + "step": 19882 + }, + { + "epoch": 2.155572419774501, + "grad_norm": 0.014006126672029495, + "learning_rate": 5.6499347163789355e-06, + "loss": 0.0001, + "step": 19883 + }, + { + "epoch": 2.155680832610581, + "grad_norm": 2.3615057468414307, + "learning_rate": 5.649209342811548e-06, + "loss": 0.0031, + "step": 19884 + }, + { + "epoch": 2.1557892454466607, + "grad_norm": 2.4240288734436035, + "learning_rate": 5.6484839692441615e-06, + "loss": 0.0089, + "step": 19885 + }, + { + "epoch": 2.1558976582827407, + "grad_norm": 3.6370487213134766, + "learning_rate": 5.6477585956767745e-06, + "loss": 0.0134, + "step": 19886 + }, + { + "epoch": 2.1560060711188207, + "grad_norm": 0.33183249831199646, + "learning_rate": 5.6470332221093874e-06, + "loss": 0.0012, + "step": 19887 + }, + { + "epoch": 2.1561144839549002, + "grad_norm": 0.07945815473794937, + "learning_rate": 5.646307848542e-06, + "loss": 0.0005, + "step": 19888 + }, + { + "epoch": 2.1562228967909802, + "grad_norm": 0.3558999300003052, + "learning_rate": 5.6455824749746126e-06, + "loss": 0.001, + "step": 19889 + }, + { + "epoch": 2.1563313096270598, + "grad_norm": 1.8637237548828125, + "learning_rate": 5.644857101407225e-06, + "loss": 0.0009, + "step": 19890 + }, + { + "epoch": 2.1564397224631398, + "grad_norm": 0.12396173179149628, + "learning_rate": 5.644131727839838e-06, + "loss": 0.0007, + "step": 19891 + }, + { + "epoch": 2.1565481352992193, + "grad_norm": 0.3853014409542084, + "learning_rate": 5.643406354272451e-06, + "loss": 0.0013, + "step": 19892 + }, + { + "epoch": 2.1566565481352993, + "grad_norm": 0.4285689890384674, + "learning_rate": 5.642680980705063e-06, + "loss": 0.0006, + "step": 19893 + }, + { + "epoch": 2.156764960971379, + "grad_norm": 0.10335876792669296, + "learning_rate": 5.641955607137677e-06, + "loss": 0.0006, + "step": 19894 + }, + { + "epoch": 2.156873373807459, + "grad_norm": 2.9979352951049805, + "learning_rate": 5.64123023357029e-06, + "loss": 0.0025, + "step": 19895 + }, + { + "epoch": 2.156981786643539, + "grad_norm": 2.3691670894622803, + "learning_rate": 5.640504860002902e-06, + "loss": 0.0249, + "step": 19896 + }, + { + "epoch": 2.1570901994796183, + "grad_norm": 0.009424986317753792, + "learning_rate": 5.639779486435515e-06, + "loss": 0.0, + "step": 19897 + }, + { + "epoch": 2.1571986123156983, + "grad_norm": 0.036436133086681366, + "learning_rate": 5.639054112868128e-06, + "loss": 0.0002, + "step": 19898 + }, + { + "epoch": 2.157307025151778, + "grad_norm": 2.6513328552246094, + "learning_rate": 5.63832873930074e-06, + "loss": 0.0153, + "step": 19899 + }, + { + "epoch": 2.157415437987858, + "grad_norm": 0.01793477311730385, + "learning_rate": 5.637603365733353e-06, + "loss": 0.0002, + "step": 19900 + }, + { + "epoch": 2.1575238508239374, + "grad_norm": 2.326345205307007, + "learning_rate": 5.636877992165966e-06, + "loss": 0.0055, + "step": 19901 + }, + { + "epoch": 2.1576322636600174, + "grad_norm": 6.522136688232422, + "learning_rate": 5.636152618598578e-06, + "loss": 0.0236, + "step": 19902 + }, + { + "epoch": 2.157740676496097, + "grad_norm": 23.036741256713867, + "learning_rate": 5.635427245031192e-06, + "loss": 0.0326, + "step": 19903 + }, + { + "epoch": 2.157849089332177, + "grad_norm": 2.254833221435547, + "learning_rate": 5.634701871463805e-06, + "loss": 0.0065, + "step": 19904 + }, + { + "epoch": 2.157957502168257, + "grad_norm": 1.273329257965088, + "learning_rate": 5.633976497896417e-06, + "loss": 0.0042, + "step": 19905 + }, + { + "epoch": 2.1580659150043364, + "grad_norm": 0.3520783483982086, + "learning_rate": 5.63325112432903e-06, + "loss": 0.0011, + "step": 19906 + }, + { + "epoch": 2.1581743278404164, + "grad_norm": 0.3363227844238281, + "learning_rate": 5.632525750761643e-06, + "loss": 0.0012, + "step": 19907 + }, + { + "epoch": 2.158282740676496, + "grad_norm": 11.230568885803223, + "learning_rate": 5.631800377194255e-06, + "loss": 0.0851, + "step": 19908 + }, + { + "epoch": 2.158391153512576, + "grad_norm": 4.11521053314209, + "learning_rate": 5.631075003626868e-06, + "loss": 0.0768, + "step": 19909 + }, + { + "epoch": 2.1584995663486555, + "grad_norm": 9.563329696655273, + "learning_rate": 5.63034963005948e-06, + "loss": 0.0267, + "step": 19910 + }, + { + "epoch": 2.1586079791847355, + "grad_norm": 0.06244460865855217, + "learning_rate": 5.629624256492094e-06, + "loss": 0.0002, + "step": 19911 + }, + { + "epoch": 2.1587163920208154, + "grad_norm": 0.023868342861533165, + "learning_rate": 5.628898882924707e-06, + "loss": 0.0002, + "step": 19912 + }, + { + "epoch": 2.158824804856895, + "grad_norm": 2.703150510787964, + "learning_rate": 5.62817350935732e-06, + "loss": 0.0045, + "step": 19913 + }, + { + "epoch": 2.158933217692975, + "grad_norm": 0.023246297612786293, + "learning_rate": 5.627448135789932e-06, + "loss": 0.0002, + "step": 19914 + }, + { + "epoch": 2.1590416305290545, + "grad_norm": 0.2932427227497101, + "learning_rate": 5.626722762222545e-06, + "loss": 0.0009, + "step": 19915 + }, + { + "epoch": 2.1591500433651345, + "grad_norm": 0.16242505609989166, + "learning_rate": 5.625997388655157e-06, + "loss": 0.0004, + "step": 19916 + }, + { + "epoch": 2.159258456201214, + "grad_norm": 1.7205923795700073, + "learning_rate": 5.62527201508777e-06, + "loss": 0.0062, + "step": 19917 + }, + { + "epoch": 2.159366869037294, + "grad_norm": 2.311243772506714, + "learning_rate": 5.624546641520383e-06, + "loss": 0.0033, + "step": 19918 + }, + { + "epoch": 2.159475281873374, + "grad_norm": 0.6206225156784058, + "learning_rate": 5.6238212679529955e-06, + "loss": 0.0017, + "step": 19919 + }, + { + "epoch": 2.1595836947094535, + "grad_norm": 0.6019399166107178, + "learning_rate": 5.623095894385609e-06, + "loss": 0.0012, + "step": 19920 + }, + { + "epoch": 2.1596921075455335, + "grad_norm": 0.02898857183754444, + "learning_rate": 5.622370520818222e-06, + "loss": 0.0001, + "step": 19921 + }, + { + "epoch": 2.159800520381613, + "grad_norm": 6.0489068031311035, + "learning_rate": 5.6216451472508345e-06, + "loss": 0.0109, + "step": 19922 + }, + { + "epoch": 2.159908933217693, + "grad_norm": 0.1419995129108429, + "learning_rate": 5.6209197736834475e-06, + "loss": 0.0003, + "step": 19923 + }, + { + "epoch": 2.1600173460537726, + "grad_norm": 0.5111286640167236, + "learning_rate": 5.6201944001160605e-06, + "loss": 0.0014, + "step": 19924 + }, + { + "epoch": 2.1601257588898526, + "grad_norm": 0.4661004841327667, + "learning_rate": 5.619469026548673e-06, + "loss": 0.0013, + "step": 19925 + }, + { + "epoch": 2.160234171725932, + "grad_norm": 6.733764171600342, + "learning_rate": 5.618743652981286e-06, + "loss": 0.0032, + "step": 19926 + }, + { + "epoch": 2.160342584562012, + "grad_norm": 0.0747656375169754, + "learning_rate": 5.618018279413899e-06, + "loss": 0.0003, + "step": 19927 + }, + { + "epoch": 2.160450997398092, + "grad_norm": 0.036080095916986465, + "learning_rate": 5.617292905846511e-06, + "loss": 0.0002, + "step": 19928 + }, + { + "epoch": 2.1605594102341716, + "grad_norm": 0.11191173642873764, + "learning_rate": 5.6165675322791246e-06, + "loss": 0.0005, + "step": 19929 + }, + { + "epoch": 2.1606678230702516, + "grad_norm": 0.4832759201526642, + "learning_rate": 5.6158421587117376e-06, + "loss": 0.0018, + "step": 19930 + }, + { + "epoch": 2.160776235906331, + "grad_norm": 2.34214448928833, + "learning_rate": 5.61511678514435e-06, + "loss": 0.0133, + "step": 19931 + }, + { + "epoch": 2.160884648742411, + "grad_norm": 2.524643659591675, + "learning_rate": 5.614391411576963e-06, + "loss": 0.0074, + "step": 19932 + }, + { + "epoch": 2.1609930615784907, + "grad_norm": 0.22885942459106445, + "learning_rate": 5.613666038009576e-06, + "loss": 0.0009, + "step": 19933 + }, + { + "epoch": 2.1611014744145707, + "grad_norm": 8.884482383728027, + "learning_rate": 5.612940664442188e-06, + "loss": 0.0299, + "step": 19934 + }, + { + "epoch": 2.1612098872506507, + "grad_norm": 5.860199451446533, + "learning_rate": 5.612215290874801e-06, + "loss": 0.024, + "step": 19935 + }, + { + "epoch": 2.16131830008673, + "grad_norm": 0.11090889573097229, + "learning_rate": 5.611489917307413e-06, + "loss": 0.0004, + "step": 19936 + }, + { + "epoch": 2.16142671292281, + "grad_norm": 5.359134197235107, + "learning_rate": 5.610764543740027e-06, + "loss": 0.054, + "step": 19937 + }, + { + "epoch": 2.1615351257588897, + "grad_norm": 9.588120460510254, + "learning_rate": 5.61003917017264e-06, + "loss": 0.0174, + "step": 19938 + }, + { + "epoch": 2.1616435385949697, + "grad_norm": 0.7305837869644165, + "learning_rate": 5.609313796605253e-06, + "loss": 0.0346, + "step": 19939 + }, + { + "epoch": 2.1617519514310493, + "grad_norm": 12.051384925842285, + "learning_rate": 5.608588423037865e-06, + "loss": 0.0125, + "step": 19940 + }, + { + "epoch": 2.1618603642671292, + "grad_norm": 0.052073586732149124, + "learning_rate": 5.607863049470478e-06, + "loss": 0.0004, + "step": 19941 + }, + { + "epoch": 2.1619687771032092, + "grad_norm": 0.3279462456703186, + "learning_rate": 5.60713767590309e-06, + "loss": 0.0011, + "step": 19942 + }, + { + "epoch": 2.1620771899392888, + "grad_norm": 0.7476127743721008, + "learning_rate": 5.606412302335703e-06, + "loss": 0.0017, + "step": 19943 + }, + { + "epoch": 2.1621856027753688, + "grad_norm": 1.1240631341934204, + "learning_rate": 5.605686928768316e-06, + "loss": 0.0015, + "step": 19944 + }, + { + "epoch": 2.1622940156114483, + "grad_norm": 2.2242274284362793, + "learning_rate": 5.604961555200928e-06, + "loss": 0.0046, + "step": 19945 + }, + { + "epoch": 2.1624024284475283, + "grad_norm": 0.010035368613898754, + "learning_rate": 5.604236181633542e-06, + "loss": 0.0, + "step": 19946 + }, + { + "epoch": 2.162510841283608, + "grad_norm": 0.6455418467521667, + "learning_rate": 5.603510808066155e-06, + "loss": 0.001, + "step": 19947 + }, + { + "epoch": 2.162619254119688, + "grad_norm": 0.08161503821611404, + "learning_rate": 5.602785434498767e-06, + "loss": 0.0008, + "step": 19948 + }, + { + "epoch": 2.1627276669557673, + "grad_norm": 0.2798725366592407, + "learning_rate": 5.60206006093138e-06, + "loss": 0.0009, + "step": 19949 + }, + { + "epoch": 2.1628360797918473, + "grad_norm": 5.710607051849365, + "learning_rate": 5.601334687363993e-06, + "loss": 0.0526, + "step": 19950 + }, + { + "epoch": 2.1629444926279273, + "grad_norm": 0.18978558480739594, + "learning_rate": 5.600609313796605e-06, + "loss": 0.0004, + "step": 19951 + }, + { + "epoch": 2.163052905464007, + "grad_norm": 0.9837374687194824, + "learning_rate": 5.599883940229218e-06, + "loss": 0.0012, + "step": 19952 + }, + { + "epoch": 2.163161318300087, + "grad_norm": 7.374551296234131, + "learning_rate": 5.599158566661831e-06, + "loss": 0.044, + "step": 19953 + }, + { + "epoch": 2.1632697311361664, + "grad_norm": 4.98203706741333, + "learning_rate": 5.598433193094443e-06, + "loss": 0.0046, + "step": 19954 + }, + { + "epoch": 2.1633781439722464, + "grad_norm": 12.705409049987793, + "learning_rate": 5.597707819527057e-06, + "loss": 0.0339, + "step": 19955 + }, + { + "epoch": 2.163486556808326, + "grad_norm": 0.008155686780810356, + "learning_rate": 5.59698244595967e-06, + "loss": 0.0001, + "step": 19956 + }, + { + "epoch": 2.163594969644406, + "grad_norm": 3.45749831199646, + "learning_rate": 5.596257072392282e-06, + "loss": 0.0182, + "step": 19957 + }, + { + "epoch": 2.163703382480486, + "grad_norm": 3.2398879528045654, + "learning_rate": 5.595531698824895e-06, + "loss": 0.0128, + "step": 19958 + }, + { + "epoch": 2.1638117953165654, + "grad_norm": 0.5616328716278076, + "learning_rate": 5.594806325257508e-06, + "loss": 0.002, + "step": 19959 + }, + { + "epoch": 2.1639202081526454, + "grad_norm": 0.019790569320321083, + "learning_rate": 5.5940809516901205e-06, + "loss": 0.0001, + "step": 19960 + }, + { + "epoch": 2.164028620988725, + "grad_norm": 0.07510396093130112, + "learning_rate": 5.5933555781227335e-06, + "loss": 0.0002, + "step": 19961 + }, + { + "epoch": 2.164137033824805, + "grad_norm": 0.8516612648963928, + "learning_rate": 5.592630204555346e-06, + "loss": 0.0014, + "step": 19962 + }, + { + "epoch": 2.1642454466608845, + "grad_norm": 0.08816173672676086, + "learning_rate": 5.5919048309879595e-06, + "loss": 0.0002, + "step": 19963 + }, + { + "epoch": 2.1643538594969645, + "grad_norm": 14.9603910446167, + "learning_rate": 5.5911794574205725e-06, + "loss": 0.0356, + "step": 19964 + }, + { + "epoch": 2.1644622723330444, + "grad_norm": 1.6487094163894653, + "learning_rate": 5.5904540838531854e-06, + "loss": 0.0016, + "step": 19965 + }, + { + "epoch": 2.164570685169124, + "grad_norm": 4.998103141784668, + "learning_rate": 5.589728710285798e-06, + "loss": 0.0143, + "step": 19966 + }, + { + "epoch": 2.164679098005204, + "grad_norm": 0.6069791316986084, + "learning_rate": 5.589003336718411e-06, + "loss": 0.0013, + "step": 19967 + }, + { + "epoch": 2.1647875108412835, + "grad_norm": 1.253454327583313, + "learning_rate": 5.588277963151023e-06, + "loss": 0.0031, + "step": 19968 + }, + { + "epoch": 2.1648959236773635, + "grad_norm": 2.0781240463256836, + "learning_rate": 5.587552589583636e-06, + "loss": 0.0038, + "step": 19969 + }, + { + "epoch": 2.165004336513443, + "grad_norm": 0.002752120839431882, + "learning_rate": 5.586827216016249e-06, + "loss": 0.0, + "step": 19970 + }, + { + "epoch": 2.165112749349523, + "grad_norm": 0.07438617944717407, + "learning_rate": 5.586101842448861e-06, + "loss": 0.0003, + "step": 19971 + }, + { + "epoch": 2.1652211621856026, + "grad_norm": 1.7063941955566406, + "learning_rate": 5.585376468881475e-06, + "loss": 0.0075, + "step": 19972 + }, + { + "epoch": 2.1653295750216826, + "grad_norm": 2.9128499031066895, + "learning_rate": 5.584651095314088e-06, + "loss": 0.0102, + "step": 19973 + }, + { + "epoch": 2.1654379878577625, + "grad_norm": 3.2600386142730713, + "learning_rate": 5.5839257217467e-06, + "loss": 0.0103, + "step": 19974 + }, + { + "epoch": 2.165546400693842, + "grad_norm": 2.8363945484161377, + "learning_rate": 5.583200348179313e-06, + "loss": 0.0432, + "step": 19975 + }, + { + "epoch": 2.165654813529922, + "grad_norm": 0.016052041202783585, + "learning_rate": 5.582474974611926e-06, + "loss": 0.0001, + "step": 19976 + }, + { + "epoch": 2.1657632263660016, + "grad_norm": 0.7095513939857483, + "learning_rate": 5.581749601044538e-06, + "loss": 0.0038, + "step": 19977 + }, + { + "epoch": 2.1658716392020816, + "grad_norm": 3.167172908782959, + "learning_rate": 5.581024227477151e-06, + "loss": 0.0138, + "step": 19978 + }, + { + "epoch": 2.165980052038161, + "grad_norm": 2.5519275665283203, + "learning_rate": 5.580298853909764e-06, + "loss": 0.0279, + "step": 19979 + }, + { + "epoch": 2.166088464874241, + "grad_norm": 0.035156697034835815, + "learning_rate": 5.579573480342377e-06, + "loss": 0.0001, + "step": 19980 + }, + { + "epoch": 2.166196877710321, + "grad_norm": 0.6716212034225464, + "learning_rate": 5.57884810677499e-06, + "loss": 0.0014, + "step": 19981 + }, + { + "epoch": 2.1663052905464006, + "grad_norm": 1.2671916484832764, + "learning_rate": 5.578122733207603e-06, + "loss": 0.0035, + "step": 19982 + }, + { + "epoch": 2.1664137033824806, + "grad_norm": 1.0496814250946045, + "learning_rate": 5.577397359640215e-06, + "loss": 0.0032, + "step": 19983 + }, + { + "epoch": 2.16652211621856, + "grad_norm": 0.08047924190759659, + "learning_rate": 5.576671986072828e-06, + "loss": 0.0002, + "step": 19984 + }, + { + "epoch": 2.16663052905464, + "grad_norm": 1.3046810626983643, + "learning_rate": 5.575946612505441e-06, + "loss": 0.0443, + "step": 19985 + }, + { + "epoch": 2.1667389418907197, + "grad_norm": 0.8113373517990112, + "learning_rate": 5.575221238938053e-06, + "loss": 0.0014, + "step": 19986 + }, + { + "epoch": 2.1668473547267997, + "grad_norm": 2.9172511100769043, + "learning_rate": 5.574495865370666e-06, + "loss": 0.0212, + "step": 19987 + }, + { + "epoch": 2.1669557675628797, + "grad_norm": 1.6392709016799927, + "learning_rate": 5.573770491803278e-06, + "loss": 0.002, + "step": 19988 + }, + { + "epoch": 2.167064180398959, + "grad_norm": 1.7557908296585083, + "learning_rate": 5.573045118235892e-06, + "loss": 0.0597, + "step": 19989 + }, + { + "epoch": 2.167172593235039, + "grad_norm": 0.007495351135730743, + "learning_rate": 5.572319744668505e-06, + "loss": 0.0, + "step": 19990 + }, + { + "epoch": 2.1672810060711187, + "grad_norm": 0.0488029308617115, + "learning_rate": 5.571594371101118e-06, + "loss": 0.0005, + "step": 19991 + }, + { + "epoch": 2.1673894189071987, + "grad_norm": 8.587331771850586, + "learning_rate": 5.57086899753373e-06, + "loss": 0.0476, + "step": 19992 + }, + { + "epoch": 2.1674978317432783, + "grad_norm": 1.22908353805542, + "learning_rate": 5.570143623966343e-06, + "loss": 0.0036, + "step": 19993 + }, + { + "epoch": 2.1676062445793582, + "grad_norm": 7.0282182693481445, + "learning_rate": 5.569418250398955e-06, + "loss": 0.0049, + "step": 19994 + }, + { + "epoch": 2.167714657415438, + "grad_norm": 0.028950203210115433, + "learning_rate": 5.568692876831568e-06, + "loss": 0.0001, + "step": 19995 + }, + { + "epoch": 2.1678230702515178, + "grad_norm": 3.5472936630249023, + "learning_rate": 5.567967503264181e-06, + "loss": 0.005, + "step": 19996 + }, + { + "epoch": 2.1679314830875978, + "grad_norm": 0.14066511392593384, + "learning_rate": 5.5672421296967935e-06, + "loss": 0.0009, + "step": 19997 + }, + { + "epoch": 2.1680398959236773, + "grad_norm": 0.07056903839111328, + "learning_rate": 5.566516756129407e-06, + "loss": 0.0004, + "step": 19998 + }, + { + "epoch": 2.1681483087597573, + "grad_norm": 0.023632237687706947, + "learning_rate": 5.56579138256202e-06, + "loss": 0.0003, + "step": 19999 + }, + { + "epoch": 2.168256721595837, + "grad_norm": 0.09254578500986099, + "learning_rate": 5.5650660089946325e-06, + "loss": 0.0007, + "step": 20000 + }, + { + "epoch": 2.168365134431917, + "grad_norm": 4.646196365356445, + "learning_rate": 5.5643406354272455e-06, + "loss": 0.0119, + "step": 20001 + }, + { + "epoch": 2.1684735472679963, + "grad_norm": 0.6117221117019653, + "learning_rate": 5.5636152618598585e-06, + "loss": 0.0021, + "step": 20002 + }, + { + "epoch": 2.1685819601040763, + "grad_norm": 7.688277721405029, + "learning_rate": 5.562889888292471e-06, + "loss": 0.0546, + "step": 20003 + }, + { + "epoch": 2.1686903729401563, + "grad_norm": 2.9015965461730957, + "learning_rate": 5.562164514725084e-06, + "loss": 0.0088, + "step": 20004 + }, + { + "epoch": 2.168798785776236, + "grad_norm": 2.037796974182129, + "learning_rate": 5.561439141157697e-06, + "loss": 0.0311, + "step": 20005 + }, + { + "epoch": 2.168907198612316, + "grad_norm": 0.8913277387619019, + "learning_rate": 5.5607137675903104e-06, + "loss": 0.0015, + "step": 20006 + }, + { + "epoch": 2.1690156114483954, + "grad_norm": 0.17161424458026886, + "learning_rate": 5.5599883940229226e-06, + "loss": 0.0009, + "step": 20007 + }, + { + "epoch": 2.1691240242844754, + "grad_norm": 0.5458976030349731, + "learning_rate": 5.5592630204555356e-06, + "loss": 0.0022, + "step": 20008 + }, + { + "epoch": 2.169232437120555, + "grad_norm": 4.0275559425354, + "learning_rate": 5.558537646888148e-06, + "loss": 0.0183, + "step": 20009 + }, + { + "epoch": 2.169340849956635, + "grad_norm": 0.01184480357915163, + "learning_rate": 5.557812273320761e-06, + "loss": 0.0001, + "step": 20010 + }, + { + "epoch": 2.169449262792715, + "grad_norm": 0.011512067168951035, + "learning_rate": 5.557086899753374e-06, + "loss": 0.0002, + "step": 20011 + }, + { + "epoch": 2.1695576756287944, + "grad_norm": 4.259092330932617, + "learning_rate": 5.556361526185986e-06, + "loss": 0.0121, + "step": 20012 + }, + { + "epoch": 2.1696660884648744, + "grad_norm": 0.909593939781189, + "learning_rate": 5.555636152618599e-06, + "loss": 0.0028, + "step": 20013 + }, + { + "epoch": 2.169774501300954, + "grad_norm": 1.5768929719924927, + "learning_rate": 5.554910779051211e-06, + "loss": 0.0015, + "step": 20014 + }, + { + "epoch": 2.169882914137034, + "grad_norm": 0.6893465518951416, + "learning_rate": 5.554185405483825e-06, + "loss": 0.0027, + "step": 20015 + }, + { + "epoch": 2.1699913269731135, + "grad_norm": 3.9608840942382812, + "learning_rate": 5.553460031916438e-06, + "loss": 0.0015, + "step": 20016 + }, + { + "epoch": 2.1700997398091935, + "grad_norm": 1.3779135942459106, + "learning_rate": 5.552734658349051e-06, + "loss": 0.0044, + "step": 20017 + }, + { + "epoch": 2.170208152645273, + "grad_norm": 0.802505373954773, + "learning_rate": 5.552009284781663e-06, + "loss": 0.001, + "step": 20018 + }, + { + "epoch": 2.170316565481353, + "grad_norm": 0.08493396639823914, + "learning_rate": 5.551283911214276e-06, + "loss": 0.0003, + "step": 20019 + }, + { + "epoch": 2.170424978317433, + "grad_norm": 3.3304905891418457, + "learning_rate": 5.550558537646888e-06, + "loss": 0.0453, + "step": 20020 + }, + { + "epoch": 2.1705333911535125, + "grad_norm": 3.0916643142700195, + "learning_rate": 5.549833164079501e-06, + "loss": 0.0052, + "step": 20021 + }, + { + "epoch": 2.1706418039895925, + "grad_norm": 0.914795458316803, + "learning_rate": 5.549107790512114e-06, + "loss": 0.0037, + "step": 20022 + }, + { + "epoch": 2.170750216825672, + "grad_norm": 0.04417124390602112, + "learning_rate": 5.548382416944726e-06, + "loss": 0.0003, + "step": 20023 + }, + { + "epoch": 2.170858629661752, + "grad_norm": 2.909655809402466, + "learning_rate": 5.54765704337734e-06, + "loss": 0.0131, + "step": 20024 + }, + { + "epoch": 2.1709670424978316, + "grad_norm": 3.1875414848327637, + "learning_rate": 5.546931669809953e-06, + "loss": 0.0248, + "step": 20025 + }, + { + "epoch": 2.1710754553339116, + "grad_norm": 5.3998236656188965, + "learning_rate": 5.546206296242565e-06, + "loss": 0.0195, + "step": 20026 + }, + { + "epoch": 2.171183868169991, + "grad_norm": 2.9656927585601807, + "learning_rate": 5.545480922675178e-06, + "loss": 0.0263, + "step": 20027 + }, + { + "epoch": 2.171292281006071, + "grad_norm": 4.240262508392334, + "learning_rate": 5.544755549107791e-06, + "loss": 0.0076, + "step": 20028 + }, + { + "epoch": 2.171400693842151, + "grad_norm": 3.2994329929351807, + "learning_rate": 5.544030175540403e-06, + "loss": 0.0112, + "step": 20029 + }, + { + "epoch": 2.1715091066782306, + "grad_norm": 9.980669021606445, + "learning_rate": 5.543304801973016e-06, + "loss": 0.0246, + "step": 20030 + }, + { + "epoch": 2.1716175195143106, + "grad_norm": 0.11450455337762833, + "learning_rate": 5.542579428405629e-06, + "loss": 0.0007, + "step": 20031 + }, + { + "epoch": 2.17172593235039, + "grad_norm": 6.925998210906982, + "learning_rate": 5.541854054838243e-06, + "loss": 0.0424, + "step": 20032 + }, + { + "epoch": 2.17183434518647, + "grad_norm": 3.681312322616577, + "learning_rate": 5.541128681270855e-06, + "loss": 0.0194, + "step": 20033 + }, + { + "epoch": 2.17194275802255, + "grad_norm": 9.413405418395996, + "learning_rate": 5.540403307703468e-06, + "loss": 0.0525, + "step": 20034 + }, + { + "epoch": 2.1720511708586296, + "grad_norm": 8.662440299987793, + "learning_rate": 5.53967793413608e-06, + "loss": 0.0528, + "step": 20035 + }, + { + "epoch": 2.1721595836947096, + "grad_norm": 0.014104098081588745, + "learning_rate": 5.538952560568693e-06, + "loss": 0.0001, + "step": 20036 + }, + { + "epoch": 2.172267996530789, + "grad_norm": 0.13146573305130005, + "learning_rate": 5.538227187001306e-06, + "loss": 0.0004, + "step": 20037 + }, + { + "epoch": 2.172376409366869, + "grad_norm": 0.13726864755153656, + "learning_rate": 5.5375018134339185e-06, + "loss": 0.0002, + "step": 20038 + }, + { + "epoch": 2.1724848222029487, + "grad_norm": 0.12243802845478058, + "learning_rate": 5.5367764398665315e-06, + "loss": 0.0003, + "step": 20039 + }, + { + "epoch": 2.1725932350390287, + "grad_norm": 2.3506898880004883, + "learning_rate": 5.536051066299144e-06, + "loss": 0.0067, + "step": 20040 + }, + { + "epoch": 2.172701647875108, + "grad_norm": 0.04181715473532677, + "learning_rate": 5.5353256927317575e-06, + "loss": 0.0001, + "step": 20041 + }, + { + "epoch": 2.172810060711188, + "grad_norm": 0.2637527585029602, + "learning_rate": 5.5346003191643705e-06, + "loss": 0.0006, + "step": 20042 + }, + { + "epoch": 2.172918473547268, + "grad_norm": 0.48207783699035645, + "learning_rate": 5.5338749455969835e-06, + "loss": 0.0011, + "step": 20043 + }, + { + "epoch": 2.1730268863833477, + "grad_norm": 5.022196292877197, + "learning_rate": 5.533149572029596e-06, + "loss": 0.0359, + "step": 20044 + }, + { + "epoch": 2.1731352992194277, + "grad_norm": 0.027513882145285606, + "learning_rate": 5.532424198462209e-06, + "loss": 0.0004, + "step": 20045 + }, + { + "epoch": 2.1732437120555073, + "grad_norm": 0.01385269034653902, + "learning_rate": 5.531698824894821e-06, + "loss": 0.0002, + "step": 20046 + }, + { + "epoch": 2.1733521248915872, + "grad_norm": 2.713045597076416, + "learning_rate": 5.530973451327434e-06, + "loss": 0.0344, + "step": 20047 + }, + { + "epoch": 2.173460537727667, + "grad_norm": 0.28894874453544617, + "learning_rate": 5.530248077760047e-06, + "loss": 0.0012, + "step": 20048 + }, + { + "epoch": 2.1735689505637468, + "grad_norm": 0.5126231908798218, + "learning_rate": 5.529522704192659e-06, + "loss": 0.0026, + "step": 20049 + }, + { + "epoch": 2.1736773633998263, + "grad_norm": 2.0116922855377197, + "learning_rate": 5.528797330625273e-06, + "loss": 0.0043, + "step": 20050 + }, + { + "epoch": 2.1737857762359063, + "grad_norm": 0.6081538796424866, + "learning_rate": 5.528071957057886e-06, + "loss": 0.0017, + "step": 20051 + }, + { + "epoch": 2.1738941890719863, + "grad_norm": 0.6138612627983093, + "learning_rate": 5.527346583490498e-06, + "loss": 0.0025, + "step": 20052 + }, + { + "epoch": 2.174002601908066, + "grad_norm": 0.18293382227420807, + "learning_rate": 5.526621209923111e-06, + "loss": 0.0008, + "step": 20053 + }, + { + "epoch": 2.174111014744146, + "grad_norm": 1.1486986875534058, + "learning_rate": 5.525895836355724e-06, + "loss": 0.0301, + "step": 20054 + }, + { + "epoch": 2.1742194275802254, + "grad_norm": 0.08087959885597229, + "learning_rate": 5.525170462788336e-06, + "loss": 0.0005, + "step": 20055 + }, + { + "epoch": 2.1743278404163053, + "grad_norm": 0.17557698488235474, + "learning_rate": 5.524445089220949e-06, + "loss": 0.0012, + "step": 20056 + }, + { + "epoch": 2.174436253252385, + "grad_norm": 2.2793095111846924, + "learning_rate": 5.523719715653562e-06, + "loss": 0.0061, + "step": 20057 + }, + { + "epoch": 2.174544666088465, + "grad_norm": 3.5989186763763428, + "learning_rate": 5.522994342086176e-06, + "loss": 0.0552, + "step": 20058 + }, + { + "epoch": 2.174653078924545, + "grad_norm": 0.08349957317113876, + "learning_rate": 5.522268968518788e-06, + "loss": 0.0005, + "step": 20059 + }, + { + "epoch": 2.1747614917606244, + "grad_norm": 3.1514134407043457, + "learning_rate": 5.521543594951401e-06, + "loss": 0.0374, + "step": 20060 + }, + { + "epoch": 2.1748699045967044, + "grad_norm": 0.3233421742916107, + "learning_rate": 5.520818221384013e-06, + "loss": 0.0006, + "step": 20061 + }, + { + "epoch": 2.174978317432784, + "grad_norm": 2.0574281215667725, + "learning_rate": 5.520092847816626e-06, + "loss": 0.0332, + "step": 20062 + }, + { + "epoch": 2.175086730268864, + "grad_norm": 1.756177306175232, + "learning_rate": 5.519367474249239e-06, + "loss": 0.0062, + "step": 20063 + }, + { + "epoch": 2.1751951431049434, + "grad_norm": 2.3490405082702637, + "learning_rate": 5.518642100681851e-06, + "loss": 0.0046, + "step": 20064 + }, + { + "epoch": 2.1753035559410234, + "grad_norm": 0.5831623077392578, + "learning_rate": 5.517916727114464e-06, + "loss": 0.001, + "step": 20065 + }, + { + "epoch": 2.1754119687771034, + "grad_norm": 7.597879409790039, + "learning_rate": 5.517191353547076e-06, + "loss": 0.0706, + "step": 20066 + }, + { + "epoch": 2.175520381613183, + "grad_norm": 0.1737648844718933, + "learning_rate": 5.51646597997969e-06, + "loss": 0.0008, + "step": 20067 + }, + { + "epoch": 2.175628794449263, + "grad_norm": 3.8349130153656006, + "learning_rate": 5.515740606412303e-06, + "loss": 0.0105, + "step": 20068 + }, + { + "epoch": 2.1757372072853425, + "grad_norm": 12.866076469421387, + "learning_rate": 5.515015232844916e-06, + "loss": 0.009, + "step": 20069 + }, + { + "epoch": 2.1758456201214225, + "grad_norm": 5.015256881713867, + "learning_rate": 5.514289859277528e-06, + "loss": 0.0584, + "step": 20070 + }, + { + "epoch": 2.175954032957502, + "grad_norm": 0.031213993206620216, + "learning_rate": 5.513564485710141e-06, + "loss": 0.0004, + "step": 20071 + }, + { + "epoch": 2.176062445793582, + "grad_norm": 0.1898341178894043, + "learning_rate": 5.512839112142753e-06, + "loss": 0.0014, + "step": 20072 + }, + { + "epoch": 2.1761708586296615, + "grad_norm": 0.9069072008132935, + "learning_rate": 5.512113738575366e-06, + "loss": 0.0063, + "step": 20073 + }, + { + "epoch": 2.1762792714657415, + "grad_norm": 12.090933799743652, + "learning_rate": 5.511388365007979e-06, + "loss": 0.0156, + "step": 20074 + }, + { + "epoch": 2.1763876843018215, + "grad_norm": 5.125833988189697, + "learning_rate": 5.5106629914405915e-06, + "loss": 0.0365, + "step": 20075 + }, + { + "epoch": 2.176496097137901, + "grad_norm": 15.840896606445312, + "learning_rate": 5.509937617873205e-06, + "loss": 0.016, + "step": 20076 + }, + { + "epoch": 2.176604509973981, + "grad_norm": 5.83714485168457, + "learning_rate": 5.509212244305818e-06, + "loss": 0.082, + "step": 20077 + }, + { + "epoch": 2.1767129228100606, + "grad_norm": 5.7520036697387695, + "learning_rate": 5.5084868707384305e-06, + "loss": 0.0118, + "step": 20078 + }, + { + "epoch": 2.1768213356461406, + "grad_norm": 6.371212482452393, + "learning_rate": 5.5077614971710435e-06, + "loss": 0.0571, + "step": 20079 + }, + { + "epoch": 2.17692974848222, + "grad_norm": 3.587423801422119, + "learning_rate": 5.5070361236036565e-06, + "loss": 0.0074, + "step": 20080 + }, + { + "epoch": 2.1770381613183, + "grad_norm": 4.279173851013184, + "learning_rate": 5.506310750036269e-06, + "loss": 0.0044, + "step": 20081 + }, + { + "epoch": 2.17714657415438, + "grad_norm": 0.23520688712596893, + "learning_rate": 5.505585376468882e-06, + "loss": 0.0007, + "step": 20082 + }, + { + "epoch": 2.1772549869904596, + "grad_norm": 0.20164811611175537, + "learning_rate": 5.504860002901495e-06, + "loss": 0.0007, + "step": 20083 + }, + { + "epoch": 2.1773633998265396, + "grad_norm": 0.2571551501750946, + "learning_rate": 5.5041346293341084e-06, + "loss": 0.0007, + "step": 20084 + }, + { + "epoch": 2.177471812662619, + "grad_norm": 0.1752665936946869, + "learning_rate": 5.503409255766721e-06, + "loss": 0.001, + "step": 20085 + }, + { + "epoch": 2.177580225498699, + "grad_norm": 0.07003214955329895, + "learning_rate": 5.5026838821993336e-06, + "loss": 0.0002, + "step": 20086 + }, + { + "epoch": 2.1776886383347787, + "grad_norm": 0.04863588511943817, + "learning_rate": 5.501958508631946e-06, + "loss": 0.0004, + "step": 20087 + }, + { + "epoch": 2.1777970511708586, + "grad_norm": 1.4220497608184814, + "learning_rate": 5.501233135064559e-06, + "loss": 0.0563, + "step": 20088 + }, + { + "epoch": 2.1779054640069386, + "grad_norm": 0.08186084032058716, + "learning_rate": 5.500507761497172e-06, + "loss": 0.0005, + "step": 20089 + }, + { + "epoch": 2.178013876843018, + "grad_norm": 0.0037679974921047688, + "learning_rate": 5.499782387929784e-06, + "loss": 0.0001, + "step": 20090 + }, + { + "epoch": 2.178122289679098, + "grad_norm": 0.44998791813850403, + "learning_rate": 5.499057014362397e-06, + "loss": 0.0004, + "step": 20091 + }, + { + "epoch": 2.1782307025151777, + "grad_norm": 0.3331068456172943, + "learning_rate": 5.498331640795009e-06, + "loss": 0.0006, + "step": 20092 + }, + { + "epoch": 2.1783391153512577, + "grad_norm": 1.4621268510818481, + "learning_rate": 5.497606267227623e-06, + "loss": 0.0068, + "step": 20093 + }, + { + "epoch": 2.1784475281873372, + "grad_norm": 3.9933743476867676, + "learning_rate": 5.496880893660236e-06, + "loss": 0.0194, + "step": 20094 + }, + { + "epoch": 2.178555941023417, + "grad_norm": 0.23760317265987396, + "learning_rate": 5.496155520092849e-06, + "loss": 0.001, + "step": 20095 + }, + { + "epoch": 2.1786643538594967, + "grad_norm": 0.7590295672416687, + "learning_rate": 5.495430146525461e-06, + "loss": 0.0012, + "step": 20096 + }, + { + "epoch": 2.1787727666955767, + "grad_norm": 3.0619924068450928, + "learning_rate": 5.494704772958074e-06, + "loss": 0.0109, + "step": 20097 + }, + { + "epoch": 2.1788811795316567, + "grad_norm": 4.156509876251221, + "learning_rate": 5.493979399390686e-06, + "loss": 0.0094, + "step": 20098 + }, + { + "epoch": 2.1789895923677363, + "grad_norm": 0.0857233926653862, + "learning_rate": 5.493254025823299e-06, + "loss": 0.0005, + "step": 20099 + }, + { + "epoch": 2.1790980052038162, + "grad_norm": 0.03914235904812813, + "learning_rate": 5.492528652255912e-06, + "loss": 0.0002, + "step": 20100 + }, + { + "epoch": 2.179206418039896, + "grad_norm": 0.0478145033121109, + "learning_rate": 5.491803278688526e-06, + "loss": 0.0003, + "step": 20101 + }, + { + "epoch": 2.1793148308759758, + "grad_norm": 2.686694383621216, + "learning_rate": 5.491077905121138e-06, + "loss": 0.0231, + "step": 20102 + }, + { + "epoch": 2.1794232437120553, + "grad_norm": 0.08907538652420044, + "learning_rate": 5.490352531553751e-06, + "loss": 0.0006, + "step": 20103 + }, + { + "epoch": 2.1795316565481353, + "grad_norm": 0.05072629079222679, + "learning_rate": 5.489627157986363e-06, + "loss": 0.0005, + "step": 20104 + }, + { + "epoch": 2.1796400693842153, + "grad_norm": 3.7845892906188965, + "learning_rate": 5.488901784418976e-06, + "loss": 0.0373, + "step": 20105 + }, + { + "epoch": 2.179748482220295, + "grad_norm": 0.03918808326125145, + "learning_rate": 5.488176410851589e-06, + "loss": 0.0003, + "step": 20106 + }, + { + "epoch": 2.179856895056375, + "grad_norm": 0.21860253810882568, + "learning_rate": 5.487451037284201e-06, + "loss": 0.0011, + "step": 20107 + }, + { + "epoch": 2.1799653078924544, + "grad_norm": 0.027097908779978752, + "learning_rate": 5.486725663716814e-06, + "loss": 0.0002, + "step": 20108 + }, + { + "epoch": 2.1800737207285343, + "grad_norm": 6.947897434234619, + "learning_rate": 5.486000290149427e-06, + "loss": 0.0525, + "step": 20109 + }, + { + "epoch": 2.180182133564614, + "grad_norm": 0.4507273733615875, + "learning_rate": 5.485274916582041e-06, + "loss": 0.0009, + "step": 20110 + }, + { + "epoch": 2.180290546400694, + "grad_norm": 0.2681770622730255, + "learning_rate": 5.484549543014653e-06, + "loss": 0.0014, + "step": 20111 + }, + { + "epoch": 2.180398959236774, + "grad_norm": 0.3081367611885071, + "learning_rate": 5.483824169447266e-06, + "loss": 0.0015, + "step": 20112 + }, + { + "epoch": 2.1805073720728534, + "grad_norm": 4.2315192222595215, + "learning_rate": 5.483098795879878e-06, + "loss": 0.0075, + "step": 20113 + }, + { + "epoch": 2.1806157849089334, + "grad_norm": 1.5670034885406494, + "learning_rate": 5.482373422312491e-06, + "loss": 0.0027, + "step": 20114 + }, + { + "epoch": 2.180724197745013, + "grad_norm": 0.05818193405866623, + "learning_rate": 5.481648048745104e-06, + "loss": 0.0002, + "step": 20115 + }, + { + "epoch": 2.180832610581093, + "grad_norm": 0.22127021849155426, + "learning_rate": 5.4809226751777165e-06, + "loss": 0.001, + "step": 20116 + }, + { + "epoch": 2.1809410234171724, + "grad_norm": 1.300008773803711, + "learning_rate": 5.4801973016103295e-06, + "loss": 0.0367, + "step": 20117 + }, + { + "epoch": 2.1810494362532524, + "grad_norm": 1.7781375646591187, + "learning_rate": 5.479471928042942e-06, + "loss": 0.0297, + "step": 20118 + }, + { + "epoch": 2.181157849089332, + "grad_norm": 0.1384054571390152, + "learning_rate": 5.4787465544755555e-06, + "loss": 0.0009, + "step": 20119 + }, + { + "epoch": 2.181266261925412, + "grad_norm": 1.2826451063156128, + "learning_rate": 5.4780211809081685e-06, + "loss": 0.0022, + "step": 20120 + }, + { + "epoch": 2.181374674761492, + "grad_norm": 0.07195335626602173, + "learning_rate": 5.4772958073407815e-06, + "loss": 0.0003, + "step": 20121 + }, + { + "epoch": 2.1814830875975715, + "grad_norm": 2.3053205013275146, + "learning_rate": 5.476570433773394e-06, + "loss": 0.006, + "step": 20122 + }, + { + "epoch": 2.1815915004336515, + "grad_norm": 1.8842216730117798, + "learning_rate": 5.475845060206007e-06, + "loss": 0.0026, + "step": 20123 + }, + { + "epoch": 2.181699913269731, + "grad_norm": 2.432974100112915, + "learning_rate": 5.475119686638619e-06, + "loss": 0.0079, + "step": 20124 + }, + { + "epoch": 2.181808326105811, + "grad_norm": 0.00780695304274559, + "learning_rate": 5.474394313071232e-06, + "loss": 0.0, + "step": 20125 + }, + { + "epoch": 2.1819167389418905, + "grad_norm": 0.5610815286636353, + "learning_rate": 5.473668939503845e-06, + "loss": 0.0034, + "step": 20126 + }, + { + "epoch": 2.1820251517779705, + "grad_norm": 8.543147087097168, + "learning_rate": 5.4729435659364586e-06, + "loss": 0.0111, + "step": 20127 + }, + { + "epoch": 2.1821335646140505, + "grad_norm": 5.350042343139648, + "learning_rate": 5.472218192369071e-06, + "loss": 0.0098, + "step": 20128 + }, + { + "epoch": 2.18224197745013, + "grad_norm": 0.0883241593837738, + "learning_rate": 5.471492818801684e-06, + "loss": 0.0004, + "step": 20129 + }, + { + "epoch": 2.18235039028621, + "grad_norm": 3.4551124572753906, + "learning_rate": 5.470767445234297e-06, + "loss": 0.0278, + "step": 20130 + }, + { + "epoch": 2.1824588031222896, + "grad_norm": 0.24258257448673248, + "learning_rate": 5.470042071666909e-06, + "loss": 0.0014, + "step": 20131 + }, + { + "epoch": 2.1825672159583696, + "grad_norm": 0.23871934413909912, + "learning_rate": 5.469316698099522e-06, + "loss": 0.001, + "step": 20132 + }, + { + "epoch": 2.182675628794449, + "grad_norm": 3.2397634983062744, + "learning_rate": 5.468591324532134e-06, + "loss": 0.0174, + "step": 20133 + }, + { + "epoch": 2.182784041630529, + "grad_norm": 2.5074493885040283, + "learning_rate": 5.467865950964747e-06, + "loss": 0.0053, + "step": 20134 + }, + { + "epoch": 2.182892454466609, + "grad_norm": 5.353522777557373, + "learning_rate": 5.46714057739736e-06, + "loss": 0.025, + "step": 20135 + }, + { + "epoch": 2.1830008673026886, + "grad_norm": 0.1574198454618454, + "learning_rate": 5.466415203829974e-06, + "loss": 0.0006, + "step": 20136 + }, + { + "epoch": 2.1831092801387686, + "grad_norm": 2.594895124435425, + "learning_rate": 5.465689830262586e-06, + "loss": 0.0206, + "step": 20137 + }, + { + "epoch": 2.183217692974848, + "grad_norm": 3.3381524085998535, + "learning_rate": 5.464964456695199e-06, + "loss": 0.0391, + "step": 20138 + }, + { + "epoch": 2.183326105810928, + "grad_norm": 3.3658108711242676, + "learning_rate": 5.464239083127811e-06, + "loss": 0.0169, + "step": 20139 + }, + { + "epoch": 2.1834345186470077, + "grad_norm": 1.5926156044006348, + "learning_rate": 5.463513709560424e-06, + "loss": 0.0198, + "step": 20140 + }, + { + "epoch": 2.1835429314830876, + "grad_norm": 0.19514186680316925, + "learning_rate": 5.462788335993037e-06, + "loss": 0.0015, + "step": 20141 + }, + { + "epoch": 2.183651344319167, + "grad_norm": 0.033865585923194885, + "learning_rate": 5.462062962425649e-06, + "loss": 0.0003, + "step": 20142 + }, + { + "epoch": 2.183759757155247, + "grad_norm": 2.134080648422241, + "learning_rate": 5.461337588858262e-06, + "loss": 0.0101, + "step": 20143 + }, + { + "epoch": 2.183868169991327, + "grad_norm": 0.06876815855503082, + "learning_rate": 5.460612215290874e-06, + "loss": 0.0008, + "step": 20144 + }, + { + "epoch": 2.1839765828274067, + "grad_norm": 1.1643122434616089, + "learning_rate": 5.459886841723488e-06, + "loss": 0.0022, + "step": 20145 + }, + { + "epoch": 2.1840849956634867, + "grad_norm": 0.1729271560907364, + "learning_rate": 5.459161468156101e-06, + "loss": 0.0009, + "step": 20146 + }, + { + "epoch": 2.1841934084995662, + "grad_norm": 1.6004548072814941, + "learning_rate": 5.458436094588714e-06, + "loss": 0.0113, + "step": 20147 + }, + { + "epoch": 2.184301821335646, + "grad_norm": 0.40636059641838074, + "learning_rate": 5.457710721021326e-06, + "loss": 0.0017, + "step": 20148 + }, + { + "epoch": 2.1844102341717258, + "grad_norm": 0.03310847282409668, + "learning_rate": 5.456985347453939e-06, + "loss": 0.0002, + "step": 20149 + }, + { + "epoch": 2.1845186470078057, + "grad_norm": 0.05862048268318176, + "learning_rate": 5.456259973886551e-06, + "loss": 0.0006, + "step": 20150 + }, + { + "epoch": 2.1846270598438857, + "grad_norm": 2.550314426422119, + "learning_rate": 5.455534600319164e-06, + "loss": 0.0021, + "step": 20151 + }, + { + "epoch": 2.1847354726799653, + "grad_norm": 0.0791744738817215, + "learning_rate": 5.454809226751777e-06, + "loss": 0.0003, + "step": 20152 + }, + { + "epoch": 2.1848438855160452, + "grad_norm": 0.7503364682197571, + "learning_rate": 5.454083853184391e-06, + "loss": 0.0012, + "step": 20153 + }, + { + "epoch": 2.184952298352125, + "grad_norm": 2.1885879039764404, + "learning_rate": 5.453358479617003e-06, + "loss": 0.0374, + "step": 20154 + }, + { + "epoch": 2.1850607111882048, + "grad_norm": 2.283381938934326, + "learning_rate": 5.452633106049616e-06, + "loss": 0.0085, + "step": 20155 + }, + { + "epoch": 2.1851691240242843, + "grad_norm": 0.22055721282958984, + "learning_rate": 5.451907732482229e-06, + "loss": 0.0011, + "step": 20156 + }, + { + "epoch": 2.1852775368603643, + "grad_norm": 0.2001473754644394, + "learning_rate": 5.4511823589148415e-06, + "loss": 0.0018, + "step": 20157 + }, + { + "epoch": 2.1853859496964443, + "grad_norm": 3.882800579071045, + "learning_rate": 5.4504569853474545e-06, + "loss": 0.0136, + "step": 20158 + }, + { + "epoch": 2.185494362532524, + "grad_norm": 4.1449503898620605, + "learning_rate": 5.449731611780067e-06, + "loss": 0.0078, + "step": 20159 + }, + { + "epoch": 2.185602775368604, + "grad_norm": 0.43191635608673096, + "learning_rate": 5.44900623821268e-06, + "loss": 0.0015, + "step": 20160 + }, + { + "epoch": 2.1857111882046834, + "grad_norm": 0.9159272313117981, + "learning_rate": 5.448280864645293e-06, + "loss": 0.0019, + "step": 20161 + }, + { + "epoch": 2.1858196010407633, + "grad_norm": 0.016608932986855507, + "learning_rate": 5.4475554910779064e-06, + "loss": 0.0001, + "step": 20162 + }, + { + "epoch": 2.185928013876843, + "grad_norm": 2.7252159118652344, + "learning_rate": 5.446830117510519e-06, + "loss": 0.0238, + "step": 20163 + }, + { + "epoch": 2.186036426712923, + "grad_norm": 0.17520318925380707, + "learning_rate": 5.446104743943132e-06, + "loss": 0.0006, + "step": 20164 + }, + { + "epoch": 2.1861448395490024, + "grad_norm": 6.287834167480469, + "learning_rate": 5.445379370375744e-06, + "loss": 0.025, + "step": 20165 + }, + { + "epoch": 2.1862532523850824, + "grad_norm": 0.12327856570482254, + "learning_rate": 5.444653996808357e-06, + "loss": 0.0002, + "step": 20166 + }, + { + "epoch": 2.1863616652211624, + "grad_norm": 0.21578526496887207, + "learning_rate": 5.44392862324097e-06, + "loss": 0.0004, + "step": 20167 + }, + { + "epoch": 2.186470078057242, + "grad_norm": 1.4469900131225586, + "learning_rate": 5.443203249673582e-06, + "loss": 0.0022, + "step": 20168 + }, + { + "epoch": 2.186578490893322, + "grad_norm": 9.12828254699707, + "learning_rate": 5.442477876106195e-06, + "loss": 0.0311, + "step": 20169 + }, + { + "epoch": 2.1866869037294014, + "grad_norm": 6.232733726501465, + "learning_rate": 5.441752502538807e-06, + "loss": 0.0069, + "step": 20170 + }, + { + "epoch": 2.1867953165654814, + "grad_norm": 7.020461559295654, + "learning_rate": 5.441027128971421e-06, + "loss": 0.0106, + "step": 20171 + }, + { + "epoch": 2.186903729401561, + "grad_norm": 1.0276638269424438, + "learning_rate": 5.440301755404034e-06, + "loss": 0.0334, + "step": 20172 + }, + { + "epoch": 2.187012142237641, + "grad_norm": 1.424970030784607, + "learning_rate": 5.439576381836647e-06, + "loss": 0.0048, + "step": 20173 + }, + { + "epoch": 2.1871205550737205, + "grad_norm": 0.12314943969249725, + "learning_rate": 5.438851008269259e-06, + "loss": 0.0009, + "step": 20174 + }, + { + "epoch": 2.1872289679098005, + "grad_norm": 0.43753352761268616, + "learning_rate": 5.438125634701872e-06, + "loss": 0.0014, + "step": 20175 + }, + { + "epoch": 2.1873373807458805, + "grad_norm": 0.01312904991209507, + "learning_rate": 5.437400261134484e-06, + "loss": 0.0001, + "step": 20176 + }, + { + "epoch": 2.18744579358196, + "grad_norm": 0.05156337842345238, + "learning_rate": 5.436674887567097e-06, + "loss": 0.0003, + "step": 20177 + }, + { + "epoch": 2.18755420641804, + "grad_norm": 0.16630062460899353, + "learning_rate": 5.43594951399971e-06, + "loss": 0.0004, + "step": 20178 + }, + { + "epoch": 2.1876626192541195, + "grad_norm": 0.07780543714761734, + "learning_rate": 5.435224140432324e-06, + "loss": 0.0004, + "step": 20179 + }, + { + "epoch": 2.1877710320901995, + "grad_norm": 0.44064536690711975, + "learning_rate": 5.434498766864936e-06, + "loss": 0.0006, + "step": 20180 + }, + { + "epoch": 2.1878794449262795, + "grad_norm": 0.12666848301887512, + "learning_rate": 5.433773393297549e-06, + "loss": 0.0009, + "step": 20181 + }, + { + "epoch": 2.187987857762359, + "grad_norm": 0.01792093552649021, + "learning_rate": 5.433048019730162e-06, + "loss": 0.0001, + "step": 20182 + }, + { + "epoch": 2.188096270598439, + "grad_norm": 1.0321683883666992, + "learning_rate": 5.432322646162774e-06, + "loss": 0.0038, + "step": 20183 + }, + { + "epoch": 2.1882046834345186, + "grad_norm": 2.9645984172821045, + "learning_rate": 5.431597272595387e-06, + "loss": 0.018, + "step": 20184 + }, + { + "epoch": 2.1883130962705986, + "grad_norm": 4.032095432281494, + "learning_rate": 5.430871899027999e-06, + "loss": 0.006, + "step": 20185 + }, + { + "epoch": 2.188421509106678, + "grad_norm": 0.2509101629257202, + "learning_rate": 5.430146525460612e-06, + "loss": 0.001, + "step": 20186 + }, + { + "epoch": 2.188529921942758, + "grad_norm": 3.373368501663208, + "learning_rate": 5.429421151893225e-06, + "loss": 0.0091, + "step": 20187 + }, + { + "epoch": 2.1886383347788376, + "grad_norm": 4.545988082885742, + "learning_rate": 5.428695778325839e-06, + "loss": 0.0112, + "step": 20188 + }, + { + "epoch": 2.1887467476149176, + "grad_norm": 1.6020435094833374, + "learning_rate": 5.427970404758451e-06, + "loss": 0.0162, + "step": 20189 + }, + { + "epoch": 2.1888551604509976, + "grad_norm": 7.091740608215332, + "learning_rate": 5.427245031191064e-06, + "loss": 0.0136, + "step": 20190 + }, + { + "epoch": 2.188963573287077, + "grad_norm": 1.9329725503921509, + "learning_rate": 5.426519657623676e-06, + "loss": 0.0089, + "step": 20191 + }, + { + "epoch": 2.189071986123157, + "grad_norm": 0.0968501940369606, + "learning_rate": 5.425794284056289e-06, + "loss": 0.0007, + "step": 20192 + }, + { + "epoch": 2.1891803989592367, + "grad_norm": 0.11178457736968994, + "learning_rate": 5.425068910488902e-06, + "loss": 0.0002, + "step": 20193 + }, + { + "epoch": 2.1892888117953166, + "grad_norm": 0.044083792716264725, + "learning_rate": 5.4243435369215145e-06, + "loss": 0.0004, + "step": 20194 + }, + { + "epoch": 2.189397224631396, + "grad_norm": 3.189351797103882, + "learning_rate": 5.4236181633541275e-06, + "loss": 0.0368, + "step": 20195 + }, + { + "epoch": 2.189505637467476, + "grad_norm": 0.08896519243717194, + "learning_rate": 5.422892789786741e-06, + "loss": 0.0005, + "step": 20196 + }, + { + "epoch": 2.1896140503035557, + "grad_norm": 0.4188985526561737, + "learning_rate": 5.4221674162193535e-06, + "loss": 0.0007, + "step": 20197 + }, + { + "epoch": 2.1897224631396357, + "grad_norm": 0.30087903141975403, + "learning_rate": 5.4214420426519665e-06, + "loss": 0.0016, + "step": 20198 + }, + { + "epoch": 2.1898308759757157, + "grad_norm": 6.111227035522461, + "learning_rate": 5.4207166690845795e-06, + "loss": 0.0089, + "step": 20199 + }, + { + "epoch": 2.1899392888117952, + "grad_norm": 0.20047377049922943, + "learning_rate": 5.419991295517192e-06, + "loss": 0.0008, + "step": 20200 + }, + { + "epoch": 2.190047701647875, + "grad_norm": 2.131592273712158, + "learning_rate": 5.419265921949805e-06, + "loss": 0.0041, + "step": 20201 + }, + { + "epoch": 2.1901561144839548, + "grad_norm": 3.0184104442596436, + "learning_rate": 5.418540548382417e-06, + "loss": 0.0015, + "step": 20202 + }, + { + "epoch": 2.1902645273200347, + "grad_norm": 2.6963510513305664, + "learning_rate": 5.41781517481503e-06, + "loss": 0.0226, + "step": 20203 + }, + { + "epoch": 2.1903729401561143, + "grad_norm": 12.280187606811523, + "learning_rate": 5.417089801247643e-06, + "loss": 0.0353, + "step": 20204 + }, + { + "epoch": 2.1904813529921943, + "grad_norm": 0.0993400439620018, + "learning_rate": 5.4163644276802566e-06, + "loss": 0.0004, + "step": 20205 + }, + { + "epoch": 2.1905897658282742, + "grad_norm": 2.5845866203308105, + "learning_rate": 5.415639054112869e-06, + "loss": 0.029, + "step": 20206 + }, + { + "epoch": 2.190698178664354, + "grad_norm": 0.22957362234592438, + "learning_rate": 5.414913680545482e-06, + "loss": 0.0008, + "step": 20207 + }, + { + "epoch": 2.1908065915004338, + "grad_norm": 4.814138889312744, + "learning_rate": 5.414188306978095e-06, + "loss": 0.0054, + "step": 20208 + }, + { + "epoch": 2.1909150043365133, + "grad_norm": 0.1955481469631195, + "learning_rate": 5.413462933410707e-06, + "loss": 0.0008, + "step": 20209 + }, + { + "epoch": 2.1910234171725933, + "grad_norm": 10.374035835266113, + "learning_rate": 5.41273755984332e-06, + "loss": 0.0465, + "step": 20210 + }, + { + "epoch": 2.191131830008673, + "grad_norm": 0.7477644085884094, + "learning_rate": 5.412012186275932e-06, + "loss": 0.0008, + "step": 20211 + }, + { + "epoch": 2.191240242844753, + "grad_norm": 0.11074865609407425, + "learning_rate": 5.411286812708545e-06, + "loss": 0.0008, + "step": 20212 + }, + { + "epoch": 2.191348655680833, + "grad_norm": 0.06353865563869476, + "learning_rate": 5.410561439141158e-06, + "loss": 0.0004, + "step": 20213 + }, + { + "epoch": 2.1914570685169124, + "grad_norm": 0.015143689699470997, + "learning_rate": 5.409836065573772e-06, + "loss": 0.0, + "step": 20214 + }, + { + "epoch": 2.1915654813529923, + "grad_norm": 0.007637216243892908, + "learning_rate": 5.409110692006384e-06, + "loss": 0.0, + "step": 20215 + }, + { + "epoch": 2.191673894189072, + "grad_norm": 2.0315141677856445, + "learning_rate": 5.408385318438997e-06, + "loss": 0.0145, + "step": 20216 + }, + { + "epoch": 2.191782307025152, + "grad_norm": 0.43855899572372437, + "learning_rate": 5.407659944871609e-06, + "loss": 0.0014, + "step": 20217 + }, + { + "epoch": 2.1918907198612314, + "grad_norm": 1.4950817823410034, + "learning_rate": 5.406934571304222e-06, + "loss": 0.0114, + "step": 20218 + }, + { + "epoch": 2.1919991326973114, + "grad_norm": 3.249809503555298, + "learning_rate": 5.406209197736835e-06, + "loss": 0.0675, + "step": 20219 + }, + { + "epoch": 2.192107545533391, + "grad_norm": 0.033341340720653534, + "learning_rate": 5.405483824169447e-06, + "loss": 0.0004, + "step": 20220 + }, + { + "epoch": 2.192215958369471, + "grad_norm": 0.11959325522184372, + "learning_rate": 5.40475845060206e-06, + "loss": 0.0009, + "step": 20221 + }, + { + "epoch": 2.192324371205551, + "grad_norm": 0.03217657655477524, + "learning_rate": 5.404033077034674e-06, + "loss": 0.0003, + "step": 20222 + }, + { + "epoch": 2.1924327840416304, + "grad_norm": 0.22569864988327026, + "learning_rate": 5.403307703467286e-06, + "loss": 0.0008, + "step": 20223 + }, + { + "epoch": 2.1925411968777104, + "grad_norm": 0.08833900839090347, + "learning_rate": 5.402582329899899e-06, + "loss": 0.0006, + "step": 20224 + }, + { + "epoch": 2.19264960971379, + "grad_norm": 2.385723114013672, + "learning_rate": 5.401856956332512e-06, + "loss": 0.0086, + "step": 20225 + }, + { + "epoch": 2.19275802254987, + "grad_norm": 4.100849151611328, + "learning_rate": 5.401131582765124e-06, + "loss": 0.0305, + "step": 20226 + }, + { + "epoch": 2.1928664353859495, + "grad_norm": 0.2240605503320694, + "learning_rate": 5.400406209197737e-06, + "loss": 0.0007, + "step": 20227 + }, + { + "epoch": 2.1929748482220295, + "grad_norm": 3.995842933654785, + "learning_rate": 5.39968083563035e-06, + "loss": 0.0153, + "step": 20228 + }, + { + "epoch": 2.1930832610581095, + "grad_norm": 0.03290051221847534, + "learning_rate": 5.398955462062962e-06, + "loss": 0.0003, + "step": 20229 + }, + { + "epoch": 2.193191673894189, + "grad_norm": 2.090465545654297, + "learning_rate": 5.398230088495575e-06, + "loss": 0.0036, + "step": 20230 + }, + { + "epoch": 2.193300086730269, + "grad_norm": 0.603006899356842, + "learning_rate": 5.397504714928189e-06, + "loss": 0.0015, + "step": 20231 + }, + { + "epoch": 2.1934084995663485, + "grad_norm": 0.3045329451560974, + "learning_rate": 5.396779341360801e-06, + "loss": 0.0012, + "step": 20232 + }, + { + "epoch": 2.1935169124024285, + "grad_norm": 0.022842830047011375, + "learning_rate": 5.396053967793414e-06, + "loss": 0.0001, + "step": 20233 + }, + { + "epoch": 2.193625325238508, + "grad_norm": 1.713684320449829, + "learning_rate": 5.395328594226027e-06, + "loss": 0.0015, + "step": 20234 + }, + { + "epoch": 2.193733738074588, + "grad_norm": 0.015808966010808945, + "learning_rate": 5.3946032206586395e-06, + "loss": 0.0001, + "step": 20235 + }, + { + "epoch": 2.193842150910668, + "grad_norm": 0.5628136396408081, + "learning_rate": 5.3938778470912525e-06, + "loss": 0.001, + "step": 20236 + }, + { + "epoch": 2.1939505637467476, + "grad_norm": 0.07879015803337097, + "learning_rate": 5.393152473523865e-06, + "loss": 0.0005, + "step": 20237 + }, + { + "epoch": 2.1940589765828276, + "grad_norm": 0.2706174850463867, + "learning_rate": 5.392427099956478e-06, + "loss": 0.0009, + "step": 20238 + }, + { + "epoch": 2.194167389418907, + "grad_norm": 11.165109634399414, + "learning_rate": 5.391701726389091e-06, + "loss": 0.0076, + "step": 20239 + }, + { + "epoch": 2.194275802254987, + "grad_norm": 3.887545585632324, + "learning_rate": 5.3909763528217045e-06, + "loss": 0.0188, + "step": 20240 + }, + { + "epoch": 2.1943842150910666, + "grad_norm": 2.551687240600586, + "learning_rate": 5.390250979254317e-06, + "loss": 0.0053, + "step": 20241 + }, + { + "epoch": 2.1944926279271466, + "grad_norm": 0.07978727668523788, + "learning_rate": 5.38952560568693e-06, + "loss": 0.0003, + "step": 20242 + }, + { + "epoch": 2.194601040763226, + "grad_norm": 4.042350769042969, + "learning_rate": 5.388800232119542e-06, + "loss": 0.0275, + "step": 20243 + }, + { + "epoch": 2.194709453599306, + "grad_norm": 4.3185248374938965, + "learning_rate": 5.388074858552155e-06, + "loss": 0.0111, + "step": 20244 + }, + { + "epoch": 2.194817866435386, + "grad_norm": 0.07118580490350723, + "learning_rate": 5.387349484984768e-06, + "loss": 0.0006, + "step": 20245 + }, + { + "epoch": 2.1949262792714657, + "grad_norm": 1.5200122594833374, + "learning_rate": 5.38662411141738e-06, + "loss": 0.0103, + "step": 20246 + }, + { + "epoch": 2.1950346921075456, + "grad_norm": 0.35040637850761414, + "learning_rate": 5.385898737849993e-06, + "loss": 0.0013, + "step": 20247 + }, + { + "epoch": 2.195143104943625, + "grad_norm": 0.010994818061590195, + "learning_rate": 5.385173364282607e-06, + "loss": 0.0001, + "step": 20248 + }, + { + "epoch": 2.195251517779705, + "grad_norm": 2.3891959190368652, + "learning_rate": 5.384447990715219e-06, + "loss": 0.0064, + "step": 20249 + }, + { + "epoch": 2.1953599306157847, + "grad_norm": 0.1015590950846672, + "learning_rate": 5.383722617147832e-06, + "loss": 0.0002, + "step": 20250 + }, + { + "epoch": 2.1954683434518647, + "grad_norm": 4.078512668609619, + "learning_rate": 5.382997243580445e-06, + "loss": 0.0463, + "step": 20251 + }, + { + "epoch": 2.1955767562879447, + "grad_norm": 0.23542122542858124, + "learning_rate": 5.382271870013057e-06, + "loss": 0.0018, + "step": 20252 + }, + { + "epoch": 2.1956851691240242, + "grad_norm": 3.5483407974243164, + "learning_rate": 5.38154649644567e-06, + "loss": 0.0187, + "step": 20253 + }, + { + "epoch": 2.195793581960104, + "grad_norm": 0.05703410133719444, + "learning_rate": 5.380821122878283e-06, + "loss": 0.0008, + "step": 20254 + }, + { + "epoch": 2.1959019947961838, + "grad_norm": 3.5206174850463867, + "learning_rate": 5.380095749310895e-06, + "loss": 0.0115, + "step": 20255 + }, + { + "epoch": 2.1960104076322637, + "grad_norm": 0.1022900715470314, + "learning_rate": 5.379370375743508e-06, + "loss": 0.0004, + "step": 20256 + }, + { + "epoch": 2.1961188204683433, + "grad_norm": 0.10873012989759445, + "learning_rate": 5.378645002176122e-06, + "loss": 0.0004, + "step": 20257 + }, + { + "epoch": 2.1962272333044233, + "grad_norm": 0.10291764885187149, + "learning_rate": 5.377919628608734e-06, + "loss": 0.001, + "step": 20258 + }, + { + "epoch": 2.1963356461405033, + "grad_norm": 0.6077045798301697, + "learning_rate": 5.377194255041347e-06, + "loss": 0.0017, + "step": 20259 + }, + { + "epoch": 2.196444058976583, + "grad_norm": 0.1780065894126892, + "learning_rate": 5.37646888147396e-06, + "loss": 0.0008, + "step": 20260 + }, + { + "epoch": 2.1965524718126628, + "grad_norm": 0.1269473135471344, + "learning_rate": 5.375743507906572e-06, + "loss": 0.0003, + "step": 20261 + }, + { + "epoch": 2.1966608846487423, + "grad_norm": 0.2209872007369995, + "learning_rate": 5.375018134339185e-06, + "loss": 0.0008, + "step": 20262 + }, + { + "epoch": 2.1967692974848223, + "grad_norm": 3.954606056213379, + "learning_rate": 5.374292760771797e-06, + "loss": 0.0114, + "step": 20263 + }, + { + "epoch": 2.196877710320902, + "grad_norm": 0.3626463711261749, + "learning_rate": 5.37356738720441e-06, + "loss": 0.0021, + "step": 20264 + }, + { + "epoch": 2.196986123156982, + "grad_norm": 10.786832809448242, + "learning_rate": 5.372842013637023e-06, + "loss": 0.0119, + "step": 20265 + }, + { + "epoch": 2.1970945359930614, + "grad_norm": 0.28087738156318665, + "learning_rate": 5.372116640069637e-06, + "loss": 0.0011, + "step": 20266 + }, + { + "epoch": 2.1972029488291414, + "grad_norm": 1.8409476280212402, + "learning_rate": 5.371391266502249e-06, + "loss": 0.0055, + "step": 20267 + }, + { + "epoch": 2.1973113616652213, + "grad_norm": 1.6740882396697998, + "learning_rate": 5.370665892934862e-06, + "loss": 0.0063, + "step": 20268 + }, + { + "epoch": 2.197419774501301, + "grad_norm": 0.6577243804931641, + "learning_rate": 5.369940519367474e-06, + "loss": 0.0013, + "step": 20269 + }, + { + "epoch": 2.197528187337381, + "grad_norm": 0.03152558207511902, + "learning_rate": 5.369215145800087e-06, + "loss": 0.0001, + "step": 20270 + }, + { + "epoch": 2.1976366001734604, + "grad_norm": 0.17693448066711426, + "learning_rate": 5.3684897722327e-06, + "loss": 0.0008, + "step": 20271 + }, + { + "epoch": 2.1977450130095404, + "grad_norm": 0.07897219061851501, + "learning_rate": 5.3677643986653125e-06, + "loss": 0.0008, + "step": 20272 + }, + { + "epoch": 2.19785342584562, + "grad_norm": 1.4735362529754639, + "learning_rate": 5.3670390250979255e-06, + "loss": 0.0197, + "step": 20273 + }, + { + "epoch": 2.1979618386817, + "grad_norm": 0.5068982243537903, + "learning_rate": 5.366313651530539e-06, + "loss": 0.0022, + "step": 20274 + }, + { + "epoch": 2.19807025151778, + "grad_norm": 0.12662579119205475, + "learning_rate": 5.3655882779631515e-06, + "loss": 0.0002, + "step": 20275 + }, + { + "epoch": 2.1981786643538594, + "grad_norm": 0.35921794176101685, + "learning_rate": 5.3648629043957645e-06, + "loss": 0.0026, + "step": 20276 + }, + { + "epoch": 2.1982870771899394, + "grad_norm": 2.326761245727539, + "learning_rate": 5.3641375308283775e-06, + "loss": 0.011, + "step": 20277 + }, + { + "epoch": 2.198395490026019, + "grad_norm": 0.2985006272792816, + "learning_rate": 5.36341215726099e-06, + "loss": 0.001, + "step": 20278 + }, + { + "epoch": 2.198503902862099, + "grad_norm": 14.008675575256348, + "learning_rate": 5.362686783693603e-06, + "loss": 0.008, + "step": 20279 + }, + { + "epoch": 2.1986123156981785, + "grad_norm": 0.130026713013649, + "learning_rate": 5.361961410126216e-06, + "loss": 0.0005, + "step": 20280 + }, + { + "epoch": 2.1987207285342585, + "grad_norm": 7.706535339355469, + "learning_rate": 5.361236036558828e-06, + "loss": 0.0127, + "step": 20281 + }, + { + "epoch": 2.1988291413703385, + "grad_norm": 0.004193738102912903, + "learning_rate": 5.360510662991441e-06, + "loss": 0.0, + "step": 20282 + }, + { + "epoch": 2.198937554206418, + "grad_norm": 0.07967368513345718, + "learning_rate": 5.3597852894240546e-06, + "loss": 0.0007, + "step": 20283 + }, + { + "epoch": 2.199045967042498, + "grad_norm": 1.0168931484222412, + "learning_rate": 5.359059915856667e-06, + "loss": 0.0317, + "step": 20284 + }, + { + "epoch": 2.1991543798785775, + "grad_norm": 14.440170288085938, + "learning_rate": 5.35833454228928e-06, + "loss": 0.0084, + "step": 20285 + }, + { + "epoch": 2.1992627927146575, + "grad_norm": 2.8310112953186035, + "learning_rate": 5.357609168721893e-06, + "loss": 0.0199, + "step": 20286 + }, + { + "epoch": 2.199371205550737, + "grad_norm": 7.3400797843933105, + "learning_rate": 5.356883795154505e-06, + "loss": 0.006, + "step": 20287 + }, + { + "epoch": 2.199479618386817, + "grad_norm": 15.480576515197754, + "learning_rate": 5.356158421587118e-06, + "loss": 0.0158, + "step": 20288 + }, + { + "epoch": 2.1995880312228966, + "grad_norm": 0.03609897196292877, + "learning_rate": 5.35543304801973e-06, + "loss": 0.0003, + "step": 20289 + }, + { + "epoch": 2.1996964440589766, + "grad_norm": 1.2550944089889526, + "learning_rate": 5.354707674452343e-06, + "loss": 0.0031, + "step": 20290 + }, + { + "epoch": 2.1998048568950566, + "grad_norm": 2.378652572631836, + "learning_rate": 5.353982300884957e-06, + "loss": 0.0076, + "step": 20291 + }, + { + "epoch": 2.199913269731136, + "grad_norm": 5.834332466125488, + "learning_rate": 5.35325692731757e-06, + "loss": 0.0346, + "step": 20292 + }, + { + "epoch": 2.200021682567216, + "grad_norm": 0.22700412571430206, + "learning_rate": 5.352531553750182e-06, + "loss": 0.0006, + "step": 20293 + }, + { + "epoch": 2.2001300954032956, + "grad_norm": 0.04651922360062599, + "learning_rate": 5.351806180182795e-06, + "loss": 0.0003, + "step": 20294 + }, + { + "epoch": 2.2002385082393756, + "grad_norm": 0.1498880833387375, + "learning_rate": 5.351080806615407e-06, + "loss": 0.0005, + "step": 20295 + }, + { + "epoch": 2.200346921075455, + "grad_norm": 2.086085081100464, + "learning_rate": 5.35035543304802e-06, + "loss": 0.0205, + "step": 20296 + }, + { + "epoch": 2.200455333911535, + "grad_norm": 6.630819320678711, + "learning_rate": 5.349630059480633e-06, + "loss": 0.0249, + "step": 20297 + }, + { + "epoch": 2.200563746747615, + "grad_norm": 1.4979479312896729, + "learning_rate": 5.348904685913245e-06, + "loss": 0.0137, + "step": 20298 + }, + { + "epoch": 2.2006721595836947, + "grad_norm": 0.8511233329772949, + "learning_rate": 5.348179312345858e-06, + "loss": 0.0008, + "step": 20299 + }, + { + "epoch": 2.2007805724197746, + "grad_norm": 0.6847034692764282, + "learning_rate": 5.347453938778472e-06, + "loss": 0.0024, + "step": 20300 + }, + { + "epoch": 2.200888985255854, + "grad_norm": 8.084416389465332, + "learning_rate": 5.346728565211084e-06, + "loss": 0.0296, + "step": 20301 + }, + { + "epoch": 2.200997398091934, + "grad_norm": 0.059720396995544434, + "learning_rate": 5.346003191643697e-06, + "loss": 0.0003, + "step": 20302 + }, + { + "epoch": 2.2011058109280137, + "grad_norm": 0.042314786463975906, + "learning_rate": 5.34527781807631e-06, + "loss": 0.0002, + "step": 20303 + }, + { + "epoch": 2.2012142237640937, + "grad_norm": 6.467336177825928, + "learning_rate": 5.344552444508922e-06, + "loss": 0.0368, + "step": 20304 + }, + { + "epoch": 2.2013226366001737, + "grad_norm": 0.528893768787384, + "learning_rate": 5.343827070941535e-06, + "loss": 0.0009, + "step": 20305 + }, + { + "epoch": 2.2014310494362532, + "grad_norm": 2.6472699642181396, + "learning_rate": 5.343101697374148e-06, + "loss": 0.0047, + "step": 20306 + }, + { + "epoch": 2.201539462272333, + "grad_norm": 0.13170544803142548, + "learning_rate": 5.34237632380676e-06, + "loss": 0.0006, + "step": 20307 + }, + { + "epoch": 2.2016478751084128, + "grad_norm": 4.85050630569458, + "learning_rate": 5.341650950239373e-06, + "loss": 0.0455, + "step": 20308 + }, + { + "epoch": 2.2017562879444927, + "grad_norm": 8.959035873413086, + "learning_rate": 5.340925576671987e-06, + "loss": 0.0647, + "step": 20309 + }, + { + "epoch": 2.2018647007805723, + "grad_norm": 0.21398457884788513, + "learning_rate": 5.340200203104599e-06, + "loss": 0.0002, + "step": 20310 + }, + { + "epoch": 2.2019731136166523, + "grad_norm": 0.45493125915527344, + "learning_rate": 5.339474829537212e-06, + "loss": 0.0103, + "step": 20311 + }, + { + "epoch": 2.202081526452732, + "grad_norm": 1.533519983291626, + "learning_rate": 5.338749455969825e-06, + "loss": 0.0098, + "step": 20312 + }, + { + "epoch": 2.202189939288812, + "grad_norm": 0.3460429012775421, + "learning_rate": 5.3380240824024375e-06, + "loss": 0.0014, + "step": 20313 + }, + { + "epoch": 2.202298352124892, + "grad_norm": 0.09469611197710037, + "learning_rate": 5.3372987088350505e-06, + "loss": 0.0011, + "step": 20314 + }, + { + "epoch": 2.2024067649609713, + "grad_norm": 1.3446881771087646, + "learning_rate": 5.336573335267663e-06, + "loss": 0.0078, + "step": 20315 + }, + { + "epoch": 2.2025151777970513, + "grad_norm": 1.4003431797027588, + "learning_rate": 5.335847961700276e-06, + "loss": 0.0032, + "step": 20316 + }, + { + "epoch": 2.202623590633131, + "grad_norm": 0.8515130281448364, + "learning_rate": 5.3351225881328895e-06, + "loss": 0.0026, + "step": 20317 + }, + { + "epoch": 2.202732003469211, + "grad_norm": 0.361264169216156, + "learning_rate": 5.3343972145655025e-06, + "loss": 0.0016, + "step": 20318 + }, + { + "epoch": 2.2028404163052904, + "grad_norm": 4.396922588348389, + "learning_rate": 5.333671840998115e-06, + "loss": 0.0214, + "step": 20319 + }, + { + "epoch": 2.2029488291413704, + "grad_norm": 0.034387316554784775, + "learning_rate": 5.332946467430728e-06, + "loss": 0.0003, + "step": 20320 + }, + { + "epoch": 2.20305724197745, + "grad_norm": 0.16603347659111023, + "learning_rate": 5.33222109386334e-06, + "loss": 0.0002, + "step": 20321 + }, + { + "epoch": 2.20316565481353, + "grad_norm": 0.2907556891441345, + "learning_rate": 5.331495720295953e-06, + "loss": 0.0012, + "step": 20322 + }, + { + "epoch": 2.20327406764961, + "grad_norm": 0.021445829421281815, + "learning_rate": 5.330770346728566e-06, + "loss": 0.0002, + "step": 20323 + }, + { + "epoch": 2.2033824804856894, + "grad_norm": 0.08893357962369919, + "learning_rate": 5.330044973161178e-06, + "loss": 0.0005, + "step": 20324 + }, + { + "epoch": 2.2034908933217694, + "grad_norm": 0.1265144646167755, + "learning_rate": 5.329319599593791e-06, + "loss": 0.0005, + "step": 20325 + }, + { + "epoch": 2.203599306157849, + "grad_norm": 5.375632286071777, + "learning_rate": 5.328594226026405e-06, + "loss": 0.0494, + "step": 20326 + }, + { + "epoch": 2.203707718993929, + "grad_norm": 0.024612050503492355, + "learning_rate": 5.327868852459017e-06, + "loss": 0.0001, + "step": 20327 + }, + { + "epoch": 2.203816131830009, + "grad_norm": 0.1111566573381424, + "learning_rate": 5.32714347889163e-06, + "loss": 0.0006, + "step": 20328 + }, + { + "epoch": 2.2039245446660884, + "grad_norm": 1.9585777521133423, + "learning_rate": 5.326418105324243e-06, + "loss": 0.0099, + "step": 20329 + }, + { + "epoch": 2.2040329575021684, + "grad_norm": 2.5439252853393555, + "learning_rate": 5.325692731756855e-06, + "loss": 0.0036, + "step": 20330 + }, + { + "epoch": 2.204141370338248, + "grad_norm": 0.8002179861068726, + "learning_rate": 5.324967358189468e-06, + "loss": 0.0019, + "step": 20331 + }, + { + "epoch": 2.204249783174328, + "grad_norm": 0.022427668794989586, + "learning_rate": 5.324241984622081e-06, + "loss": 0.0002, + "step": 20332 + }, + { + "epoch": 2.2043581960104075, + "grad_norm": 2.311854124069214, + "learning_rate": 5.323516611054693e-06, + "loss": 0.0077, + "step": 20333 + }, + { + "epoch": 2.2044666088464875, + "grad_norm": 3.044335126876831, + "learning_rate": 5.322791237487306e-06, + "loss": 0.0037, + "step": 20334 + }, + { + "epoch": 2.204575021682567, + "grad_norm": 0.03853965550661087, + "learning_rate": 5.32206586391992e-06, + "loss": 0.0003, + "step": 20335 + }, + { + "epoch": 2.204683434518647, + "grad_norm": 2.115813970565796, + "learning_rate": 5.321340490352532e-06, + "loss": 0.0036, + "step": 20336 + }, + { + "epoch": 2.204791847354727, + "grad_norm": 4.963855266571045, + "learning_rate": 5.320615116785145e-06, + "loss": 0.0364, + "step": 20337 + }, + { + "epoch": 2.2049002601908065, + "grad_norm": 0.06276386231184006, + "learning_rate": 5.319889743217758e-06, + "loss": 0.0005, + "step": 20338 + }, + { + "epoch": 2.2050086730268865, + "grad_norm": 3.9201841354370117, + "learning_rate": 5.31916436965037e-06, + "loss": 0.0156, + "step": 20339 + }, + { + "epoch": 2.205117085862966, + "grad_norm": 0.013935675844550133, + "learning_rate": 5.318438996082983e-06, + "loss": 0.0002, + "step": 20340 + }, + { + "epoch": 2.205225498699046, + "grad_norm": 0.2748364210128784, + "learning_rate": 5.317713622515595e-06, + "loss": 0.0011, + "step": 20341 + }, + { + "epoch": 2.2053339115351256, + "grad_norm": 0.10114629566669464, + "learning_rate": 5.316988248948208e-06, + "loss": 0.0005, + "step": 20342 + }, + { + "epoch": 2.2054423243712056, + "grad_norm": 4.19965934753418, + "learning_rate": 5.316262875380822e-06, + "loss": 0.0171, + "step": 20343 + }, + { + "epoch": 2.205550737207285, + "grad_norm": 2.2253215312957764, + "learning_rate": 5.315537501813435e-06, + "loss": 0.0169, + "step": 20344 + }, + { + "epoch": 2.205659150043365, + "grad_norm": 0.01741296425461769, + "learning_rate": 5.314812128246047e-06, + "loss": 0.0001, + "step": 20345 + }, + { + "epoch": 2.205767562879445, + "grad_norm": 0.07960563153028488, + "learning_rate": 5.31408675467866e-06, + "loss": 0.0004, + "step": 20346 + }, + { + "epoch": 2.2058759757155246, + "grad_norm": 3.9053921699523926, + "learning_rate": 5.313361381111272e-06, + "loss": 0.0375, + "step": 20347 + }, + { + "epoch": 2.2059843885516046, + "grad_norm": 0.28165048360824585, + "learning_rate": 5.312636007543885e-06, + "loss": 0.0016, + "step": 20348 + }, + { + "epoch": 2.206092801387684, + "grad_norm": 1.4259028434753418, + "learning_rate": 5.311910633976498e-06, + "loss": 0.0022, + "step": 20349 + }, + { + "epoch": 2.206201214223764, + "grad_norm": 0.9126415252685547, + "learning_rate": 5.3111852604091105e-06, + "loss": 0.0428, + "step": 20350 + }, + { + "epoch": 2.2063096270598437, + "grad_norm": 3.0209872722625732, + "learning_rate": 5.3104598868417235e-06, + "loss": 0.0305, + "step": 20351 + }, + { + "epoch": 2.2064180398959237, + "grad_norm": 1.2793258428573608, + "learning_rate": 5.309734513274337e-06, + "loss": 0.0301, + "step": 20352 + }, + { + "epoch": 2.2065264527320037, + "grad_norm": 0.21624448895454407, + "learning_rate": 5.3090091397069495e-06, + "loss": 0.0012, + "step": 20353 + }, + { + "epoch": 2.206634865568083, + "grad_norm": 1.3656657934188843, + "learning_rate": 5.3082837661395625e-06, + "loss": 0.0592, + "step": 20354 + }, + { + "epoch": 2.206743278404163, + "grad_norm": 1.860658049583435, + "learning_rate": 5.3075583925721755e-06, + "loss": 0.0414, + "step": 20355 + }, + { + "epoch": 2.2068516912402427, + "grad_norm": 2.539135456085205, + "learning_rate": 5.306833019004788e-06, + "loss": 0.0416, + "step": 20356 + }, + { + "epoch": 2.2069601040763227, + "grad_norm": 1.4246042966842651, + "learning_rate": 5.306107645437401e-06, + "loss": 0.0327, + "step": 20357 + }, + { + "epoch": 2.2070685169124022, + "grad_norm": 0.021036630496382713, + "learning_rate": 5.305382271870014e-06, + "loss": 0.0003, + "step": 20358 + }, + { + "epoch": 2.2071769297484822, + "grad_norm": 0.6488320231437683, + "learning_rate": 5.304656898302626e-06, + "loss": 0.0009, + "step": 20359 + }, + { + "epoch": 2.207285342584562, + "grad_norm": 1.4470833539962769, + "learning_rate": 5.303931524735239e-06, + "loss": 0.0245, + "step": 20360 + }, + { + "epoch": 2.2073937554206418, + "grad_norm": 0.23579470813274384, + "learning_rate": 5.3032061511678526e-06, + "loss": 0.0007, + "step": 20361 + }, + { + "epoch": 2.2075021682567217, + "grad_norm": 0.012908512726426125, + "learning_rate": 5.302480777600465e-06, + "loss": 0.0001, + "step": 20362 + }, + { + "epoch": 2.2076105810928013, + "grad_norm": 0.3042396903038025, + "learning_rate": 5.301755404033078e-06, + "loss": 0.0022, + "step": 20363 + }, + { + "epoch": 2.2077189939288813, + "grad_norm": 2.592914342880249, + "learning_rate": 5.301030030465691e-06, + "loss": 0.0118, + "step": 20364 + }, + { + "epoch": 2.207827406764961, + "grad_norm": 0.08399046212434769, + "learning_rate": 5.300304656898303e-06, + "loss": 0.0004, + "step": 20365 + }, + { + "epoch": 2.207935819601041, + "grad_norm": 0.8717616200447083, + "learning_rate": 5.299579283330916e-06, + "loss": 0.0016, + "step": 20366 + }, + { + "epoch": 2.2080442324371203, + "grad_norm": 1.258216142654419, + "learning_rate": 5.298853909763528e-06, + "loss": 0.0023, + "step": 20367 + }, + { + "epoch": 2.2081526452732003, + "grad_norm": 0.09946363419294357, + "learning_rate": 5.298128536196141e-06, + "loss": 0.0004, + "step": 20368 + }, + { + "epoch": 2.2082610581092803, + "grad_norm": 0.06629680097103119, + "learning_rate": 5.297403162628755e-06, + "loss": 0.0003, + "step": 20369 + }, + { + "epoch": 2.20836947094536, + "grad_norm": 2.1905322074890137, + "learning_rate": 5.296677789061368e-06, + "loss": 0.0021, + "step": 20370 + }, + { + "epoch": 2.20847788378144, + "grad_norm": 23.50323486328125, + "learning_rate": 5.29595241549398e-06, + "loss": 0.0206, + "step": 20371 + }, + { + "epoch": 2.2085862966175194, + "grad_norm": 0.2480025291442871, + "learning_rate": 5.295227041926593e-06, + "loss": 0.0021, + "step": 20372 + }, + { + "epoch": 2.2086947094535994, + "grad_norm": 0.3105749189853668, + "learning_rate": 5.294501668359205e-06, + "loss": 0.0013, + "step": 20373 + }, + { + "epoch": 2.208803122289679, + "grad_norm": 1.1190217733383179, + "learning_rate": 5.293776294791818e-06, + "loss": 0.0039, + "step": 20374 + }, + { + "epoch": 2.208911535125759, + "grad_norm": 0.36820894479751587, + "learning_rate": 5.293050921224431e-06, + "loss": 0.0009, + "step": 20375 + }, + { + "epoch": 2.209019947961839, + "grad_norm": 0.6541792750358582, + "learning_rate": 5.292325547657043e-06, + "loss": 0.0045, + "step": 20376 + }, + { + "epoch": 2.2091283607979184, + "grad_norm": 0.04176526889204979, + "learning_rate": 5.291600174089656e-06, + "loss": 0.0004, + "step": 20377 + }, + { + "epoch": 2.2092367736339984, + "grad_norm": 2.6477043628692627, + "learning_rate": 5.29087480052227e-06, + "loss": 0.0185, + "step": 20378 + }, + { + "epoch": 2.209345186470078, + "grad_norm": 0.06960268318653107, + "learning_rate": 5.290149426954882e-06, + "loss": 0.0007, + "step": 20379 + }, + { + "epoch": 2.209453599306158, + "grad_norm": 1.4847583770751953, + "learning_rate": 5.289424053387495e-06, + "loss": 0.003, + "step": 20380 + }, + { + "epoch": 2.2095620121422375, + "grad_norm": 0.5702461004257202, + "learning_rate": 5.288698679820108e-06, + "loss": 0.002, + "step": 20381 + }, + { + "epoch": 2.2096704249783174, + "grad_norm": 0.20945489406585693, + "learning_rate": 5.28797330625272e-06, + "loss": 0.0013, + "step": 20382 + }, + { + "epoch": 2.2097788378143974, + "grad_norm": 0.06166110932826996, + "learning_rate": 5.287247932685333e-06, + "loss": 0.0008, + "step": 20383 + }, + { + "epoch": 2.209887250650477, + "grad_norm": 4.135331630706787, + "learning_rate": 5.286522559117946e-06, + "loss": 0.02, + "step": 20384 + }, + { + "epoch": 2.209995663486557, + "grad_norm": 2.4406256675720215, + "learning_rate": 5.285797185550558e-06, + "loss": 0.0273, + "step": 20385 + }, + { + "epoch": 2.2101040763226365, + "grad_norm": 2.503530502319336, + "learning_rate": 5.285071811983172e-06, + "loss": 0.014, + "step": 20386 + }, + { + "epoch": 2.2102124891587165, + "grad_norm": 2.7436983585357666, + "learning_rate": 5.284346438415785e-06, + "loss": 0.0253, + "step": 20387 + }, + { + "epoch": 2.210320901994796, + "grad_norm": 0.14403998851776123, + "learning_rate": 5.283621064848397e-06, + "loss": 0.0016, + "step": 20388 + }, + { + "epoch": 2.210429314830876, + "grad_norm": 4.306475639343262, + "learning_rate": 5.28289569128101e-06, + "loss": 0.0061, + "step": 20389 + }, + { + "epoch": 2.2105377276669556, + "grad_norm": 0.15771393477916718, + "learning_rate": 5.282170317713623e-06, + "loss": 0.0019, + "step": 20390 + }, + { + "epoch": 2.2106461405030355, + "grad_norm": 4.3420257568359375, + "learning_rate": 5.2814449441462355e-06, + "loss": 0.0062, + "step": 20391 + }, + { + "epoch": 2.2107545533391155, + "grad_norm": 0.6879056692123413, + "learning_rate": 5.2807195705788485e-06, + "loss": 0.0037, + "step": 20392 + }, + { + "epoch": 2.210862966175195, + "grad_norm": 0.4557064175605774, + "learning_rate": 5.279994197011461e-06, + "loss": 0.0019, + "step": 20393 + }, + { + "epoch": 2.210971379011275, + "grad_norm": 3.3872475624084473, + "learning_rate": 5.279268823444074e-06, + "loss": 0.0062, + "step": 20394 + }, + { + "epoch": 2.2110797918473546, + "grad_norm": 10.582807540893555, + "learning_rate": 5.2785434498766875e-06, + "loss": 0.005, + "step": 20395 + }, + { + "epoch": 2.2111882046834346, + "grad_norm": 0.6570014953613281, + "learning_rate": 5.2778180763093005e-06, + "loss": 0.0013, + "step": 20396 + }, + { + "epoch": 2.211296617519514, + "grad_norm": 5.49591064453125, + "learning_rate": 5.277092702741913e-06, + "loss": 0.0342, + "step": 20397 + }, + { + "epoch": 2.211405030355594, + "grad_norm": 3.432520866394043, + "learning_rate": 5.276367329174526e-06, + "loss": 0.0676, + "step": 20398 + }, + { + "epoch": 2.211513443191674, + "grad_norm": 0.03902219980955124, + "learning_rate": 5.275641955607138e-06, + "loss": 0.0002, + "step": 20399 + }, + { + "epoch": 2.2116218560277536, + "grad_norm": 0.0044182222336530685, + "learning_rate": 5.274916582039751e-06, + "loss": 0.0, + "step": 20400 + }, + { + "epoch": 2.2117302688638336, + "grad_norm": 3.7306509017944336, + "learning_rate": 5.274191208472364e-06, + "loss": 0.0257, + "step": 20401 + }, + { + "epoch": 2.211838681699913, + "grad_norm": 3.2324588298797607, + "learning_rate": 5.273465834904976e-06, + "loss": 0.0048, + "step": 20402 + }, + { + "epoch": 2.211947094535993, + "grad_norm": 0.5809633731842041, + "learning_rate": 5.272740461337589e-06, + "loss": 0.0013, + "step": 20403 + }, + { + "epoch": 2.2120555073720727, + "grad_norm": 2.785353183746338, + "learning_rate": 5.272015087770203e-06, + "loss": 0.0416, + "step": 20404 + }, + { + "epoch": 2.2121639202081527, + "grad_norm": 0.4423162341117859, + "learning_rate": 5.271289714202815e-06, + "loss": 0.0013, + "step": 20405 + }, + { + "epoch": 2.2122723330442327, + "grad_norm": 0.10281841456890106, + "learning_rate": 5.270564340635428e-06, + "loss": 0.0007, + "step": 20406 + }, + { + "epoch": 2.212380745880312, + "grad_norm": 3.1726250648498535, + "learning_rate": 5.269838967068041e-06, + "loss": 0.0073, + "step": 20407 + }, + { + "epoch": 2.212489158716392, + "grad_norm": 1.2465977668762207, + "learning_rate": 5.269113593500653e-06, + "loss": 0.0034, + "step": 20408 + }, + { + "epoch": 2.2125975715524717, + "grad_norm": 0.27637457847595215, + "learning_rate": 5.268388219933266e-06, + "loss": 0.0005, + "step": 20409 + }, + { + "epoch": 2.2127059843885517, + "grad_norm": 0.6671521067619324, + "learning_rate": 5.267662846365879e-06, + "loss": 0.0049, + "step": 20410 + }, + { + "epoch": 2.2128143972246312, + "grad_norm": 0.1279652714729309, + "learning_rate": 5.266937472798491e-06, + "loss": 0.0013, + "step": 20411 + }, + { + "epoch": 2.2129228100607112, + "grad_norm": 0.12870009243488312, + "learning_rate": 5.266212099231105e-06, + "loss": 0.0002, + "step": 20412 + }, + { + "epoch": 2.2130312228967908, + "grad_norm": 0.24316725134849548, + "learning_rate": 5.265486725663718e-06, + "loss": 0.0016, + "step": 20413 + }, + { + "epoch": 2.2131396357328708, + "grad_norm": 0.042724981904029846, + "learning_rate": 5.26476135209633e-06, + "loss": 0.0005, + "step": 20414 + }, + { + "epoch": 2.2132480485689507, + "grad_norm": 0.040702685713768005, + "learning_rate": 5.264035978528943e-06, + "loss": 0.0002, + "step": 20415 + }, + { + "epoch": 2.2133564614050303, + "grad_norm": 0.33031266927719116, + "learning_rate": 5.263310604961556e-06, + "loss": 0.0027, + "step": 20416 + }, + { + "epoch": 2.2134648742411103, + "grad_norm": 0.04931190237402916, + "learning_rate": 5.262585231394168e-06, + "loss": 0.0003, + "step": 20417 + }, + { + "epoch": 2.21357328707719, + "grad_norm": 1.932546854019165, + "learning_rate": 5.261859857826781e-06, + "loss": 0.0063, + "step": 20418 + }, + { + "epoch": 2.21368169991327, + "grad_norm": 5.384259223937988, + "learning_rate": 5.261134484259393e-06, + "loss": 0.0119, + "step": 20419 + }, + { + "epoch": 2.2137901127493493, + "grad_norm": 0.007778753526508808, + "learning_rate": 5.260409110692006e-06, + "loss": 0.0001, + "step": 20420 + }, + { + "epoch": 2.2138985255854293, + "grad_norm": 0.08766718953847885, + "learning_rate": 5.25968373712462e-06, + "loss": 0.0005, + "step": 20421 + }, + { + "epoch": 2.2140069384215093, + "grad_norm": 2.841550827026367, + "learning_rate": 5.258958363557233e-06, + "loss": 0.0313, + "step": 20422 + }, + { + "epoch": 2.214115351257589, + "grad_norm": 0.5656881332397461, + "learning_rate": 5.258232989989845e-06, + "loss": 0.0017, + "step": 20423 + }, + { + "epoch": 2.214223764093669, + "grad_norm": 6.958686351776123, + "learning_rate": 5.257507616422458e-06, + "loss": 0.0134, + "step": 20424 + }, + { + "epoch": 2.2143321769297484, + "grad_norm": 2.3005220890045166, + "learning_rate": 5.25678224285507e-06, + "loss": 0.0063, + "step": 20425 + }, + { + "epoch": 2.2144405897658284, + "grad_norm": 0.25513094663619995, + "learning_rate": 5.256056869287683e-06, + "loss": 0.0017, + "step": 20426 + }, + { + "epoch": 2.214549002601908, + "grad_norm": 0.2831423878669739, + "learning_rate": 5.255331495720296e-06, + "loss": 0.0016, + "step": 20427 + }, + { + "epoch": 2.214657415437988, + "grad_norm": 2.4614412784576416, + "learning_rate": 5.2546061221529085e-06, + "loss": 0.0125, + "step": 20428 + }, + { + "epoch": 2.214765828274068, + "grad_norm": 7.726816654205322, + "learning_rate": 5.2538807485855215e-06, + "loss": 0.0441, + "step": 20429 + }, + { + "epoch": 2.2148742411101474, + "grad_norm": 0.06292712688446045, + "learning_rate": 5.253155375018135e-06, + "loss": 0.0001, + "step": 20430 + }, + { + "epoch": 2.2149826539462274, + "grad_norm": 0.003972270991653204, + "learning_rate": 5.2524300014507475e-06, + "loss": 0.0001, + "step": 20431 + }, + { + "epoch": 2.215091066782307, + "grad_norm": 0.7347716689109802, + "learning_rate": 5.2517046278833605e-06, + "loss": 0.0049, + "step": 20432 + }, + { + "epoch": 2.215199479618387, + "grad_norm": 0.3300032615661621, + "learning_rate": 5.2509792543159735e-06, + "loss": 0.0015, + "step": 20433 + }, + { + "epoch": 2.2153078924544665, + "grad_norm": 1.4217191934585571, + "learning_rate": 5.250253880748586e-06, + "loss": 0.0105, + "step": 20434 + }, + { + "epoch": 2.2154163052905465, + "grad_norm": 0.7508330941200256, + "learning_rate": 5.249528507181199e-06, + "loss": 0.0034, + "step": 20435 + }, + { + "epoch": 2.215524718126626, + "grad_norm": 0.010394338518381119, + "learning_rate": 5.248803133613812e-06, + "loss": 0.0001, + "step": 20436 + }, + { + "epoch": 2.215633130962706, + "grad_norm": 0.0279079582542181, + "learning_rate": 5.248077760046424e-06, + "loss": 0.0002, + "step": 20437 + }, + { + "epoch": 2.215741543798786, + "grad_norm": 16.036800384521484, + "learning_rate": 5.247352386479038e-06, + "loss": 0.0777, + "step": 20438 + }, + { + "epoch": 2.2158499566348655, + "grad_norm": 0.8590627312660217, + "learning_rate": 5.246627012911651e-06, + "loss": 0.0021, + "step": 20439 + }, + { + "epoch": 2.2159583694709455, + "grad_norm": 0.00974236149340868, + "learning_rate": 5.245901639344263e-06, + "loss": 0.0001, + "step": 20440 + }, + { + "epoch": 2.216066782307025, + "grad_norm": 1.583459734916687, + "learning_rate": 5.245176265776876e-06, + "loss": 0.0071, + "step": 20441 + }, + { + "epoch": 2.216175195143105, + "grad_norm": 0.3928801119327545, + "learning_rate": 5.244450892209489e-06, + "loss": 0.0005, + "step": 20442 + }, + { + "epoch": 2.2162836079791846, + "grad_norm": 0.03982607275247574, + "learning_rate": 5.243725518642101e-06, + "loss": 0.0003, + "step": 20443 + }, + { + "epoch": 2.2163920208152645, + "grad_norm": 3.3060991764068604, + "learning_rate": 5.243000145074714e-06, + "loss": 0.0229, + "step": 20444 + }, + { + "epoch": 2.2165004336513445, + "grad_norm": 0.16252771019935608, + "learning_rate": 5.242274771507326e-06, + "loss": 0.0002, + "step": 20445 + }, + { + "epoch": 2.216608846487424, + "grad_norm": 0.0866975262761116, + "learning_rate": 5.241549397939939e-06, + "loss": 0.0005, + "step": 20446 + }, + { + "epoch": 2.216717259323504, + "grad_norm": 0.25235551595687866, + "learning_rate": 5.240824024372553e-06, + "loss": 0.0011, + "step": 20447 + }, + { + "epoch": 2.2168256721595836, + "grad_norm": 0.07528914511203766, + "learning_rate": 5.240098650805166e-06, + "loss": 0.0007, + "step": 20448 + }, + { + "epoch": 2.2169340849956636, + "grad_norm": 2.438943386077881, + "learning_rate": 5.239373277237778e-06, + "loss": 0.0101, + "step": 20449 + }, + { + "epoch": 2.217042497831743, + "grad_norm": 5.837100028991699, + "learning_rate": 5.238647903670391e-06, + "loss": 0.0057, + "step": 20450 + }, + { + "epoch": 2.217150910667823, + "grad_norm": 0.3226156234741211, + "learning_rate": 5.237922530103003e-06, + "loss": 0.0014, + "step": 20451 + }, + { + "epoch": 2.217259323503903, + "grad_norm": 3.4698212146759033, + "learning_rate": 5.237197156535616e-06, + "loss": 0.0107, + "step": 20452 + }, + { + "epoch": 2.2173677363399826, + "grad_norm": 0.054927077144384384, + "learning_rate": 5.236471782968229e-06, + "loss": 0.0004, + "step": 20453 + }, + { + "epoch": 2.2174761491760626, + "grad_norm": 2.185622453689575, + "learning_rate": 5.235746409400841e-06, + "loss": 0.01, + "step": 20454 + }, + { + "epoch": 2.217584562012142, + "grad_norm": 1.1877413988113403, + "learning_rate": 5.235021035833454e-06, + "loss": 0.0015, + "step": 20455 + }, + { + "epoch": 2.217692974848222, + "grad_norm": 2.8554255962371826, + "learning_rate": 5.234295662266068e-06, + "loss": 0.0102, + "step": 20456 + }, + { + "epoch": 2.2178013876843017, + "grad_norm": 0.01950690895318985, + "learning_rate": 5.23357028869868e-06, + "loss": 0.0001, + "step": 20457 + }, + { + "epoch": 2.2179098005203817, + "grad_norm": 0.3576592803001404, + "learning_rate": 5.232844915131293e-06, + "loss": 0.0016, + "step": 20458 + }, + { + "epoch": 2.218018213356461, + "grad_norm": 4.703495025634766, + "learning_rate": 5.232119541563906e-06, + "loss": 0.0049, + "step": 20459 + }, + { + "epoch": 2.218126626192541, + "grad_norm": 0.10482965409755707, + "learning_rate": 5.231394167996518e-06, + "loss": 0.0004, + "step": 20460 + }, + { + "epoch": 2.218235039028621, + "grad_norm": 0.016326164826750755, + "learning_rate": 5.230668794429131e-06, + "loss": 0.0002, + "step": 20461 + }, + { + "epoch": 2.2183434518647007, + "grad_norm": 1.7411760091781616, + "learning_rate": 5.229943420861744e-06, + "loss": 0.0024, + "step": 20462 + }, + { + "epoch": 2.2184518647007807, + "grad_norm": 0.4097793698310852, + "learning_rate": 5.229218047294356e-06, + "loss": 0.0019, + "step": 20463 + }, + { + "epoch": 2.2185602775368602, + "grad_norm": 2.585247755050659, + "learning_rate": 5.22849267372697e-06, + "loss": 0.0069, + "step": 20464 + }, + { + "epoch": 2.2186686903729402, + "grad_norm": 3.363147497177124, + "learning_rate": 5.227767300159583e-06, + "loss": 0.0212, + "step": 20465 + }, + { + "epoch": 2.2187771032090198, + "grad_norm": 3.331409454345703, + "learning_rate": 5.227041926592195e-06, + "loss": 0.0011, + "step": 20466 + }, + { + "epoch": 2.2188855160450998, + "grad_norm": 5.714230060577393, + "learning_rate": 5.226316553024808e-06, + "loss": 0.0292, + "step": 20467 + }, + { + "epoch": 2.2189939288811793, + "grad_norm": 0.17907877266407013, + "learning_rate": 5.225591179457421e-06, + "loss": 0.0006, + "step": 20468 + }, + { + "epoch": 2.2191023417172593, + "grad_norm": 0.14584384858608246, + "learning_rate": 5.2248658058900335e-06, + "loss": 0.0006, + "step": 20469 + }, + { + "epoch": 2.2192107545533393, + "grad_norm": 0.23617605865001678, + "learning_rate": 5.2241404323226465e-06, + "loss": 0.0015, + "step": 20470 + }, + { + "epoch": 2.219319167389419, + "grad_norm": 7.237552642822266, + "learning_rate": 5.223415058755259e-06, + "loss": 0.0097, + "step": 20471 + }, + { + "epoch": 2.219427580225499, + "grad_norm": 0.3414468765258789, + "learning_rate": 5.222689685187872e-06, + "loss": 0.0011, + "step": 20472 + }, + { + "epoch": 2.2195359930615783, + "grad_norm": 0.09026043117046356, + "learning_rate": 5.2219643116204855e-06, + "loss": 0.0011, + "step": 20473 + }, + { + "epoch": 2.2196444058976583, + "grad_norm": 0.2015397846698761, + "learning_rate": 5.2212389380530985e-06, + "loss": 0.0006, + "step": 20474 + }, + { + "epoch": 2.2197528187337383, + "grad_norm": 37.12307357788086, + "learning_rate": 5.220513564485711e-06, + "loss": 0.0359, + "step": 20475 + }, + { + "epoch": 2.219861231569818, + "grad_norm": 0.04654983431100845, + "learning_rate": 5.219788190918324e-06, + "loss": 0.0005, + "step": 20476 + }, + { + "epoch": 2.219969644405898, + "grad_norm": 0.2833653688430786, + "learning_rate": 5.219062817350936e-06, + "loss": 0.0009, + "step": 20477 + }, + { + "epoch": 2.2200780572419774, + "grad_norm": 0.45351290702819824, + "learning_rate": 5.218337443783549e-06, + "loss": 0.0017, + "step": 20478 + }, + { + "epoch": 2.2201864700780574, + "grad_norm": 0.07246614247560501, + "learning_rate": 5.217612070216162e-06, + "loss": 0.0005, + "step": 20479 + }, + { + "epoch": 2.220294882914137, + "grad_norm": 0.1683587282896042, + "learning_rate": 5.216886696648774e-06, + "loss": 0.0014, + "step": 20480 + }, + { + "epoch": 2.220403295750217, + "grad_norm": 0.09740310907363892, + "learning_rate": 5.216161323081388e-06, + "loss": 0.0005, + "step": 20481 + }, + { + "epoch": 2.2205117085862964, + "grad_norm": 0.22904586791992188, + "learning_rate": 5.215435949514001e-06, + "loss": 0.0006, + "step": 20482 + }, + { + "epoch": 2.2206201214223764, + "grad_norm": 1.8053184747695923, + "learning_rate": 5.214710575946613e-06, + "loss": 0.0234, + "step": 20483 + }, + { + "epoch": 2.2207285342584564, + "grad_norm": 0.06486989557743073, + "learning_rate": 5.213985202379226e-06, + "loss": 0.0002, + "step": 20484 + }, + { + "epoch": 2.220836947094536, + "grad_norm": 0.02620033361017704, + "learning_rate": 5.213259828811839e-06, + "loss": 0.0001, + "step": 20485 + }, + { + "epoch": 2.220945359930616, + "grad_norm": 0.05500289425253868, + "learning_rate": 5.212534455244451e-06, + "loss": 0.0005, + "step": 20486 + }, + { + "epoch": 2.2210537727666955, + "grad_norm": 0.11979050189256668, + "learning_rate": 5.211809081677064e-06, + "loss": 0.0005, + "step": 20487 + }, + { + "epoch": 2.2211621856027755, + "grad_norm": 1.1533024311065674, + "learning_rate": 5.211083708109677e-06, + "loss": 0.0357, + "step": 20488 + }, + { + "epoch": 2.221270598438855, + "grad_norm": 0.4448632001876831, + "learning_rate": 5.210358334542289e-06, + "loss": 0.0028, + "step": 20489 + }, + { + "epoch": 2.221379011274935, + "grad_norm": 0.06525082141160965, + "learning_rate": 5.209632960974903e-06, + "loss": 0.0002, + "step": 20490 + }, + { + "epoch": 2.2214874241110145, + "grad_norm": 0.09200210869312286, + "learning_rate": 5.208907587407516e-06, + "loss": 0.0003, + "step": 20491 + }, + { + "epoch": 2.2215958369470945, + "grad_norm": 1.6051400899887085, + "learning_rate": 5.208182213840128e-06, + "loss": 0.0436, + "step": 20492 + }, + { + "epoch": 2.2217042497831745, + "grad_norm": 0.0378454253077507, + "learning_rate": 5.207456840272741e-06, + "loss": 0.0004, + "step": 20493 + }, + { + "epoch": 2.221812662619254, + "grad_norm": 0.05664600804448128, + "learning_rate": 5.206731466705354e-06, + "loss": 0.0004, + "step": 20494 + }, + { + "epoch": 2.221921075455334, + "grad_norm": 0.010846443474292755, + "learning_rate": 5.206006093137966e-06, + "loss": 0.0001, + "step": 20495 + }, + { + "epoch": 2.2220294882914136, + "grad_norm": 2.5550572872161865, + "learning_rate": 5.205280719570579e-06, + "loss": 0.0052, + "step": 20496 + }, + { + "epoch": 2.2221379011274935, + "grad_norm": 0.22223521769046783, + "learning_rate": 5.204555346003191e-06, + "loss": 0.0007, + "step": 20497 + }, + { + "epoch": 2.222246313963573, + "grad_norm": 3.0335333347320557, + "learning_rate": 5.203829972435804e-06, + "loss": 0.0293, + "step": 20498 + }, + { + "epoch": 2.222354726799653, + "grad_norm": 9.808934211730957, + "learning_rate": 5.203104598868418e-06, + "loss": 0.027, + "step": 20499 + }, + { + "epoch": 2.222463139635733, + "grad_norm": 0.44709092378616333, + "learning_rate": 5.202379225301031e-06, + "loss": 0.0019, + "step": 20500 + }, + { + "epoch": 2.2225715524718126, + "grad_norm": 0.629552960395813, + "learning_rate": 5.201653851733643e-06, + "loss": 0.0019, + "step": 20501 + }, + { + "epoch": 2.2226799653078926, + "grad_norm": 0.1736188679933548, + "learning_rate": 5.200928478166256e-06, + "loss": 0.001, + "step": 20502 + }, + { + "epoch": 2.222788378143972, + "grad_norm": 0.37292206287384033, + "learning_rate": 5.200203104598868e-06, + "loss": 0.0018, + "step": 20503 + }, + { + "epoch": 2.222896790980052, + "grad_norm": 2.4049291610717773, + "learning_rate": 5.199477731031481e-06, + "loss": 0.0311, + "step": 20504 + }, + { + "epoch": 2.2230052038161316, + "grad_norm": 11.240446090698242, + "learning_rate": 5.198752357464094e-06, + "loss": 0.07, + "step": 20505 + }, + { + "epoch": 2.2231136166522116, + "grad_norm": 0.018709268420934677, + "learning_rate": 5.1980269838967065e-06, + "loss": 0.0002, + "step": 20506 + }, + { + "epoch": 2.2232220294882916, + "grad_norm": 0.3029191493988037, + "learning_rate": 5.19730161032932e-06, + "loss": 0.0007, + "step": 20507 + }, + { + "epoch": 2.223330442324371, + "grad_norm": 0.46949151158332825, + "learning_rate": 5.196576236761933e-06, + "loss": 0.0016, + "step": 20508 + }, + { + "epoch": 2.223438855160451, + "grad_norm": 0.00847343634814024, + "learning_rate": 5.1958508631945455e-06, + "loss": 0.0001, + "step": 20509 + }, + { + "epoch": 2.2235472679965307, + "grad_norm": 1.5228043794631958, + "learning_rate": 5.1951254896271585e-06, + "loss": 0.0049, + "step": 20510 + }, + { + "epoch": 2.2236556808326107, + "grad_norm": 6.736849308013916, + "learning_rate": 5.1944001160597715e-06, + "loss": 0.0699, + "step": 20511 + }, + { + "epoch": 2.22376409366869, + "grad_norm": 1.7489246129989624, + "learning_rate": 5.193674742492384e-06, + "loss": 0.0038, + "step": 20512 + }, + { + "epoch": 2.22387250650477, + "grad_norm": 3.3363571166992188, + "learning_rate": 5.192949368924997e-06, + "loss": 0.0196, + "step": 20513 + }, + { + "epoch": 2.2239809193408497, + "grad_norm": 0.26451048254966736, + "learning_rate": 5.19222399535761e-06, + "loss": 0.0023, + "step": 20514 + }, + { + "epoch": 2.2240893321769297, + "grad_norm": 10.025221824645996, + "learning_rate": 5.191498621790222e-06, + "loss": 0.0142, + "step": 20515 + }, + { + "epoch": 2.2241977450130097, + "grad_norm": 0.030975215137004852, + "learning_rate": 5.190773248222836e-06, + "loss": 0.0002, + "step": 20516 + }, + { + "epoch": 2.2243061578490892, + "grad_norm": 0.3689350485801697, + "learning_rate": 5.190047874655449e-06, + "loss": 0.0009, + "step": 20517 + }, + { + "epoch": 2.2244145706851692, + "grad_norm": 0.44826531410217285, + "learning_rate": 5.189322501088061e-06, + "loss": 0.0009, + "step": 20518 + }, + { + "epoch": 2.2245229835212488, + "grad_norm": 4.344545841217041, + "learning_rate": 5.188597127520674e-06, + "loss": 0.015, + "step": 20519 + }, + { + "epoch": 2.2246313963573288, + "grad_norm": 0.8744135499000549, + "learning_rate": 5.187871753953287e-06, + "loss": 0.0012, + "step": 20520 + }, + { + "epoch": 2.2247398091934083, + "grad_norm": 1.0985667705535889, + "learning_rate": 5.187146380385899e-06, + "loss": 0.0039, + "step": 20521 + }, + { + "epoch": 2.2248482220294883, + "grad_norm": 1.1630905866622925, + "learning_rate": 5.186421006818512e-06, + "loss": 0.0285, + "step": 20522 + }, + { + "epoch": 2.2249566348655683, + "grad_norm": 0.9681523442268372, + "learning_rate": 5.185695633251124e-06, + "loss": 0.0042, + "step": 20523 + }, + { + "epoch": 2.225065047701648, + "grad_norm": 0.057158343493938446, + "learning_rate": 5.184970259683737e-06, + "loss": 0.0003, + "step": 20524 + }, + { + "epoch": 2.225173460537728, + "grad_norm": 4.014441967010498, + "learning_rate": 5.184244886116351e-06, + "loss": 0.0137, + "step": 20525 + }, + { + "epoch": 2.2252818733738073, + "grad_norm": 0.10800448060035706, + "learning_rate": 5.183519512548964e-06, + "loss": 0.0003, + "step": 20526 + }, + { + "epoch": 2.2253902862098873, + "grad_norm": 6.931603908538818, + "learning_rate": 5.182794138981576e-06, + "loss": 0.0579, + "step": 20527 + }, + { + "epoch": 2.225498699045967, + "grad_norm": 2.8769912719726562, + "learning_rate": 5.182068765414189e-06, + "loss": 0.0092, + "step": 20528 + }, + { + "epoch": 2.225607111882047, + "grad_norm": 2.2153053283691406, + "learning_rate": 5.181343391846801e-06, + "loss": 0.0015, + "step": 20529 + }, + { + "epoch": 2.225715524718127, + "grad_norm": 6.043540954589844, + "learning_rate": 5.180618018279414e-06, + "loss": 0.0263, + "step": 20530 + }, + { + "epoch": 2.2258239375542064, + "grad_norm": 1.7774187326431274, + "learning_rate": 5.179892644712027e-06, + "loss": 0.0328, + "step": 20531 + }, + { + "epoch": 2.2259323503902864, + "grad_norm": 0.07919991761445999, + "learning_rate": 5.179167271144639e-06, + "loss": 0.0006, + "step": 20532 + }, + { + "epoch": 2.226040763226366, + "grad_norm": 0.0531129464507103, + "learning_rate": 5.178441897577253e-06, + "loss": 0.0003, + "step": 20533 + }, + { + "epoch": 2.226149176062446, + "grad_norm": 0.31138378381729126, + "learning_rate": 5.177716524009866e-06, + "loss": 0.0016, + "step": 20534 + }, + { + "epoch": 2.2262575888985254, + "grad_norm": 0.2809794247150421, + "learning_rate": 5.176991150442478e-06, + "loss": 0.0007, + "step": 20535 + }, + { + "epoch": 2.2263660017346054, + "grad_norm": 4.918141841888428, + "learning_rate": 5.176265776875091e-06, + "loss": 0.015, + "step": 20536 + }, + { + "epoch": 2.226474414570685, + "grad_norm": 4.490921974182129, + "learning_rate": 5.175540403307704e-06, + "loss": 0.0884, + "step": 20537 + }, + { + "epoch": 2.226582827406765, + "grad_norm": 0.17405734956264496, + "learning_rate": 5.174815029740316e-06, + "loss": 0.0009, + "step": 20538 + }, + { + "epoch": 2.226691240242845, + "grad_norm": 0.15458495914936066, + "learning_rate": 5.174089656172929e-06, + "loss": 0.0005, + "step": 20539 + }, + { + "epoch": 2.2267996530789245, + "grad_norm": 0.10063176602125168, + "learning_rate": 5.173364282605542e-06, + "loss": 0.0003, + "step": 20540 + }, + { + "epoch": 2.2269080659150045, + "grad_norm": 0.10395181179046631, + "learning_rate": 5.1726389090381544e-06, + "loss": 0.0004, + "step": 20541 + }, + { + "epoch": 2.227016478751084, + "grad_norm": 2.4889872074127197, + "learning_rate": 5.171913535470768e-06, + "loss": 0.0151, + "step": 20542 + }, + { + "epoch": 2.227124891587164, + "grad_norm": 0.08950360864400864, + "learning_rate": 5.171188161903381e-06, + "loss": 0.0004, + "step": 20543 + }, + { + "epoch": 2.2272333044232435, + "grad_norm": 3.832902431488037, + "learning_rate": 5.170462788335993e-06, + "loss": 0.0046, + "step": 20544 + }, + { + "epoch": 2.2273417172593235, + "grad_norm": 1.328325629234314, + "learning_rate": 5.169737414768606e-06, + "loss": 0.014, + "step": 20545 + }, + { + "epoch": 2.2274501300954035, + "grad_norm": 6.3014349937438965, + "learning_rate": 5.169012041201219e-06, + "loss": 0.0702, + "step": 20546 + }, + { + "epoch": 2.227558542931483, + "grad_norm": 0.14198875427246094, + "learning_rate": 5.1682866676338315e-06, + "loss": 0.0007, + "step": 20547 + }, + { + "epoch": 2.227666955767563, + "grad_norm": 0.027253279462456703, + "learning_rate": 5.1675612940664445e-06, + "loss": 0.0001, + "step": 20548 + }, + { + "epoch": 2.2277753686036426, + "grad_norm": 0.004924834705889225, + "learning_rate": 5.166835920499057e-06, + "loss": 0.0001, + "step": 20549 + }, + { + "epoch": 2.2278837814397225, + "grad_norm": 2.9991917610168457, + "learning_rate": 5.16611054693167e-06, + "loss": 0.0139, + "step": 20550 + }, + { + "epoch": 2.227992194275802, + "grad_norm": 0.09523101896047592, + "learning_rate": 5.1653851733642835e-06, + "loss": 0.0008, + "step": 20551 + }, + { + "epoch": 2.228100607111882, + "grad_norm": 5.183245658874512, + "learning_rate": 5.1646597997968965e-06, + "loss": 0.026, + "step": 20552 + }, + { + "epoch": 2.228209019947962, + "grad_norm": 0.3204308748245239, + "learning_rate": 5.163934426229509e-06, + "loss": 0.0008, + "step": 20553 + }, + { + "epoch": 2.2283174327840416, + "grad_norm": 0.002747653052210808, + "learning_rate": 5.163209052662122e-06, + "loss": 0.0, + "step": 20554 + }, + { + "epoch": 2.2284258456201216, + "grad_norm": 0.4078737199306488, + "learning_rate": 5.162483679094734e-06, + "loss": 0.0016, + "step": 20555 + }, + { + "epoch": 2.228534258456201, + "grad_norm": 2.1731936931610107, + "learning_rate": 5.161758305527347e-06, + "loss": 0.0124, + "step": 20556 + }, + { + "epoch": 2.228642671292281, + "grad_norm": 2.3628010749816895, + "learning_rate": 5.16103293195996e-06, + "loss": 0.0122, + "step": 20557 + }, + { + "epoch": 2.2287510841283606, + "grad_norm": 3.246807336807251, + "learning_rate": 5.160307558392572e-06, + "loss": 0.0142, + "step": 20558 + }, + { + "epoch": 2.2288594969644406, + "grad_norm": 0.37914881110191345, + "learning_rate": 5.159582184825186e-06, + "loss": 0.0018, + "step": 20559 + }, + { + "epoch": 2.22896790980052, + "grad_norm": 0.09259597212076187, + "learning_rate": 5.158856811257799e-06, + "loss": 0.0009, + "step": 20560 + }, + { + "epoch": 2.2290763226366, + "grad_norm": 0.02274460159242153, + "learning_rate": 5.158131437690411e-06, + "loss": 0.0002, + "step": 20561 + }, + { + "epoch": 2.22918473547268, + "grad_norm": 1.4446264505386353, + "learning_rate": 5.157406064123024e-06, + "loss": 0.0063, + "step": 20562 + }, + { + "epoch": 2.2292931483087597, + "grad_norm": 1.6657789945602417, + "learning_rate": 5.156680690555637e-06, + "loss": 0.0116, + "step": 20563 + }, + { + "epoch": 2.2294015611448397, + "grad_norm": 14.333662033081055, + "learning_rate": 5.155955316988249e-06, + "loss": 0.0353, + "step": 20564 + }, + { + "epoch": 2.229509973980919, + "grad_norm": 1.8327620029449463, + "learning_rate": 5.155229943420862e-06, + "loss": 0.001, + "step": 20565 + }, + { + "epoch": 2.229618386816999, + "grad_norm": 0.13378490507602692, + "learning_rate": 5.154504569853475e-06, + "loss": 0.0008, + "step": 20566 + }, + { + "epoch": 2.2297267996530787, + "grad_norm": 0.19696909189224243, + "learning_rate": 5.153779196286087e-06, + "loss": 0.001, + "step": 20567 + }, + { + "epoch": 2.2298352124891587, + "grad_norm": 0.015805747359991074, + "learning_rate": 5.153053822718701e-06, + "loss": 0.0001, + "step": 20568 + }, + { + "epoch": 2.2299436253252387, + "grad_norm": 3.920518636703491, + "learning_rate": 5.152328449151314e-06, + "loss": 0.0097, + "step": 20569 + }, + { + "epoch": 2.2300520381613183, + "grad_norm": 0.08807160705327988, + "learning_rate": 5.151603075583926e-06, + "loss": 0.0004, + "step": 20570 + }, + { + "epoch": 2.2301604509973982, + "grad_norm": 9.831385612487793, + "learning_rate": 5.150877702016539e-06, + "loss": 0.0382, + "step": 20571 + }, + { + "epoch": 2.2302688638334778, + "grad_norm": 0.011959747411310673, + "learning_rate": 5.150152328449152e-06, + "loss": 0.0001, + "step": 20572 + }, + { + "epoch": 2.2303772766695578, + "grad_norm": 0.011944932863116264, + "learning_rate": 5.149426954881764e-06, + "loss": 0.0001, + "step": 20573 + }, + { + "epoch": 2.2304856895056373, + "grad_norm": 1.387839674949646, + "learning_rate": 5.148701581314377e-06, + "loss": 0.002, + "step": 20574 + }, + { + "epoch": 2.2305941023417173, + "grad_norm": 0.09418946504592896, + "learning_rate": 5.147976207746989e-06, + "loss": 0.0007, + "step": 20575 + }, + { + "epoch": 2.2307025151777973, + "grad_norm": 0.009351039305329323, + "learning_rate": 5.147250834179602e-06, + "loss": 0.0001, + "step": 20576 + }, + { + "epoch": 2.230810928013877, + "grad_norm": 1.9661200046539307, + "learning_rate": 5.146525460612216e-06, + "loss": 0.0035, + "step": 20577 + }, + { + "epoch": 2.230919340849957, + "grad_norm": 0.04037236049771309, + "learning_rate": 5.145800087044829e-06, + "loss": 0.0003, + "step": 20578 + }, + { + "epoch": 2.2310277536860363, + "grad_norm": 0.7329587936401367, + "learning_rate": 5.145074713477441e-06, + "loss": 0.003, + "step": 20579 + }, + { + "epoch": 2.2311361665221163, + "grad_norm": 1.1937822103500366, + "learning_rate": 5.144349339910054e-06, + "loss": 0.0023, + "step": 20580 + }, + { + "epoch": 2.231244579358196, + "grad_norm": 2.24347186088562, + "learning_rate": 5.143623966342666e-06, + "loss": 0.0199, + "step": 20581 + }, + { + "epoch": 2.231352992194276, + "grad_norm": 0.3181162476539612, + "learning_rate": 5.142898592775279e-06, + "loss": 0.0017, + "step": 20582 + }, + { + "epoch": 2.2314614050303554, + "grad_norm": 0.3236196041107178, + "learning_rate": 5.142173219207892e-06, + "loss": 0.0003, + "step": 20583 + }, + { + "epoch": 2.2315698178664354, + "grad_norm": 0.34964635968208313, + "learning_rate": 5.1414478456405045e-06, + "loss": 0.002, + "step": 20584 + }, + { + "epoch": 2.2316782307025154, + "grad_norm": 3.771474599838257, + "learning_rate": 5.140722472073118e-06, + "loss": 0.0166, + "step": 20585 + }, + { + "epoch": 2.231786643538595, + "grad_norm": 22.81450080871582, + "learning_rate": 5.139997098505731e-06, + "loss": 0.0037, + "step": 20586 + }, + { + "epoch": 2.231895056374675, + "grad_norm": 0.08432140946388245, + "learning_rate": 5.1392717249383435e-06, + "loss": 0.0002, + "step": 20587 + }, + { + "epoch": 2.2320034692107544, + "grad_norm": 0.735859751701355, + "learning_rate": 5.1385463513709565e-06, + "loss": 0.0006, + "step": 20588 + }, + { + "epoch": 2.2321118820468344, + "grad_norm": 0.6786268353462219, + "learning_rate": 5.1378209778035695e-06, + "loss": 0.0028, + "step": 20589 + }, + { + "epoch": 2.232220294882914, + "grad_norm": 6.182582378387451, + "learning_rate": 5.137095604236182e-06, + "loss": 0.0385, + "step": 20590 + }, + { + "epoch": 2.232328707718994, + "grad_norm": 3.616903781890869, + "learning_rate": 5.136370230668795e-06, + "loss": 0.0108, + "step": 20591 + }, + { + "epoch": 2.232437120555074, + "grad_norm": 0.2119595855474472, + "learning_rate": 5.135644857101408e-06, + "loss": 0.0002, + "step": 20592 + }, + { + "epoch": 2.2325455333911535, + "grad_norm": 9.576674461364746, + "learning_rate": 5.13491948353402e-06, + "loss": 0.0494, + "step": 20593 + }, + { + "epoch": 2.2326539462272335, + "grad_norm": 1.0744727849960327, + "learning_rate": 5.134194109966634e-06, + "loss": 0.0023, + "step": 20594 + }, + { + "epoch": 2.232762359063313, + "grad_norm": 0.08420759439468384, + "learning_rate": 5.133468736399247e-06, + "loss": 0.0002, + "step": 20595 + }, + { + "epoch": 2.232870771899393, + "grad_norm": 3.2059133052825928, + "learning_rate": 5.132743362831859e-06, + "loss": 0.0269, + "step": 20596 + }, + { + "epoch": 2.2329791847354725, + "grad_norm": 0.03999638184905052, + "learning_rate": 5.132017989264472e-06, + "loss": 0.0002, + "step": 20597 + }, + { + "epoch": 2.2330875975715525, + "grad_norm": 1.2666127681732178, + "learning_rate": 5.131292615697085e-06, + "loss": 0.0198, + "step": 20598 + }, + { + "epoch": 2.2331960104076325, + "grad_norm": 4.624757289886475, + "learning_rate": 5.130567242129697e-06, + "loss": 0.0578, + "step": 20599 + }, + { + "epoch": 2.233304423243712, + "grad_norm": 0.2930404841899872, + "learning_rate": 5.12984186856231e-06, + "loss": 0.0014, + "step": 20600 + }, + { + "epoch": 2.233412836079792, + "grad_norm": 1.764647126197815, + "learning_rate": 5.129116494994922e-06, + "loss": 0.0093, + "step": 20601 + }, + { + "epoch": 2.2335212489158716, + "grad_norm": 4.367203712463379, + "learning_rate": 5.128391121427536e-06, + "loss": 0.0667, + "step": 20602 + }, + { + "epoch": 2.2336296617519515, + "grad_norm": 7.134408950805664, + "learning_rate": 5.127665747860149e-06, + "loss": 0.0031, + "step": 20603 + }, + { + "epoch": 2.233738074588031, + "grad_norm": 5.638484954833984, + "learning_rate": 5.126940374292762e-06, + "loss": 0.0244, + "step": 20604 + }, + { + "epoch": 2.233846487424111, + "grad_norm": 0.3462599515914917, + "learning_rate": 5.126215000725374e-06, + "loss": 0.0023, + "step": 20605 + }, + { + "epoch": 2.2339549002601906, + "grad_norm": 2.064711093902588, + "learning_rate": 5.125489627157987e-06, + "loss": 0.0567, + "step": 20606 + }, + { + "epoch": 2.2340633130962706, + "grad_norm": 0.027259157970547676, + "learning_rate": 5.124764253590599e-06, + "loss": 0.0002, + "step": 20607 + }, + { + "epoch": 2.2341717259323506, + "grad_norm": 1.6674168109893799, + "learning_rate": 5.124038880023212e-06, + "loss": 0.0434, + "step": 20608 + }, + { + "epoch": 2.23428013876843, + "grad_norm": 8.820717811584473, + "learning_rate": 5.123313506455825e-06, + "loss": 0.0437, + "step": 20609 + }, + { + "epoch": 2.23438855160451, + "grad_norm": 2.2726476192474365, + "learning_rate": 5.122588132888437e-06, + "loss": 0.0097, + "step": 20610 + }, + { + "epoch": 2.2344969644405897, + "grad_norm": 0.5479363203048706, + "learning_rate": 5.121862759321051e-06, + "loss": 0.0009, + "step": 20611 + }, + { + "epoch": 2.2346053772766696, + "grad_norm": 2.0246763229370117, + "learning_rate": 5.121137385753664e-06, + "loss": 0.0207, + "step": 20612 + }, + { + "epoch": 2.234713790112749, + "grad_norm": 0.07160185277462006, + "learning_rate": 5.120412012186276e-06, + "loss": 0.0002, + "step": 20613 + }, + { + "epoch": 2.234822202948829, + "grad_norm": 0.16015242040157318, + "learning_rate": 5.119686638618889e-06, + "loss": 0.0007, + "step": 20614 + }, + { + "epoch": 2.2349306157849087, + "grad_norm": 16.4344482421875, + "learning_rate": 5.118961265051502e-06, + "loss": 0.0125, + "step": 20615 + }, + { + "epoch": 2.2350390286209887, + "grad_norm": 0.00997848343104124, + "learning_rate": 5.118235891484114e-06, + "loss": 0.0001, + "step": 20616 + }, + { + "epoch": 2.2351474414570687, + "grad_norm": 0.07820212095975876, + "learning_rate": 5.117510517916727e-06, + "loss": 0.0002, + "step": 20617 + }, + { + "epoch": 2.235255854293148, + "grad_norm": 1.6526793241500854, + "learning_rate": 5.11678514434934e-06, + "loss": 0.0036, + "step": 20618 + }, + { + "epoch": 2.235364267129228, + "grad_norm": 0.11293226480484009, + "learning_rate": 5.1160597707819524e-06, + "loss": 0.0008, + "step": 20619 + }, + { + "epoch": 2.2354726799653077, + "grad_norm": 0.01613462343811989, + "learning_rate": 5.115334397214566e-06, + "loss": 0.0001, + "step": 20620 + }, + { + "epoch": 2.2355810928013877, + "grad_norm": 2.9714653491973877, + "learning_rate": 5.114609023647179e-06, + "loss": 0.0069, + "step": 20621 + }, + { + "epoch": 2.2356895056374677, + "grad_norm": 4.837397575378418, + "learning_rate": 5.113883650079791e-06, + "loss": 0.0243, + "step": 20622 + }, + { + "epoch": 2.2357979184735473, + "grad_norm": 5.409450531005859, + "learning_rate": 5.113158276512404e-06, + "loss": 0.0117, + "step": 20623 + }, + { + "epoch": 2.2359063313096272, + "grad_norm": 0.9735552072525024, + "learning_rate": 5.112432902945017e-06, + "loss": 0.0046, + "step": 20624 + }, + { + "epoch": 2.236014744145707, + "grad_norm": 11.112704277038574, + "learning_rate": 5.1117075293776295e-06, + "loss": 0.0151, + "step": 20625 + }, + { + "epoch": 2.2361231569817868, + "grad_norm": 2.4463255405426025, + "learning_rate": 5.1109821558102425e-06, + "loss": 0.0532, + "step": 20626 + }, + { + "epoch": 2.2362315698178663, + "grad_norm": 0.10432151705026627, + "learning_rate": 5.110256782242855e-06, + "loss": 0.0009, + "step": 20627 + }, + { + "epoch": 2.2363399826539463, + "grad_norm": 4.027296543121338, + "learning_rate": 5.1095314086754685e-06, + "loss": 0.0158, + "step": 20628 + }, + { + "epoch": 2.236448395490026, + "grad_norm": 0.38672059774398804, + "learning_rate": 5.1088060351080815e-06, + "loss": 0.0006, + "step": 20629 + }, + { + "epoch": 2.236556808326106, + "grad_norm": 13.007914543151855, + "learning_rate": 5.1080806615406945e-06, + "loss": 0.0444, + "step": 20630 + }, + { + "epoch": 2.236665221162186, + "grad_norm": 0.0671967938542366, + "learning_rate": 5.107355287973307e-06, + "loss": 0.0005, + "step": 20631 + }, + { + "epoch": 2.2367736339982653, + "grad_norm": 2.982621431350708, + "learning_rate": 5.10662991440592e-06, + "loss": 0.0177, + "step": 20632 + }, + { + "epoch": 2.2368820468343453, + "grad_norm": 0.08373327553272247, + "learning_rate": 5.105904540838532e-06, + "loss": 0.0004, + "step": 20633 + }, + { + "epoch": 2.236990459670425, + "grad_norm": 0.08000224828720093, + "learning_rate": 5.105179167271145e-06, + "loss": 0.0006, + "step": 20634 + }, + { + "epoch": 2.237098872506505, + "grad_norm": 3.7094204425811768, + "learning_rate": 5.104453793703758e-06, + "loss": 0.0078, + "step": 20635 + }, + { + "epoch": 2.2372072853425844, + "grad_norm": 0.040113817900419235, + "learning_rate": 5.10372842013637e-06, + "loss": 0.0004, + "step": 20636 + }, + { + "epoch": 2.2373156981786644, + "grad_norm": 0.027393054217100143, + "learning_rate": 5.103003046568984e-06, + "loss": 0.0001, + "step": 20637 + }, + { + "epoch": 2.237424111014744, + "grad_norm": 0.3885439336299896, + "learning_rate": 5.102277673001597e-06, + "loss": 0.0019, + "step": 20638 + }, + { + "epoch": 2.237532523850824, + "grad_norm": 0.6547964215278625, + "learning_rate": 5.101552299434209e-06, + "loss": 0.0025, + "step": 20639 + }, + { + "epoch": 2.237640936686904, + "grad_norm": 1.4224258661270142, + "learning_rate": 5.100826925866822e-06, + "loss": 0.0288, + "step": 20640 + }, + { + "epoch": 2.2377493495229834, + "grad_norm": 0.5519512295722961, + "learning_rate": 5.100101552299435e-06, + "loss": 0.0029, + "step": 20641 + }, + { + "epoch": 2.2378577623590634, + "grad_norm": 0.04106444492936134, + "learning_rate": 5.099376178732047e-06, + "loss": 0.0004, + "step": 20642 + }, + { + "epoch": 2.237966175195143, + "grad_norm": 0.08960753679275513, + "learning_rate": 5.09865080516466e-06, + "loss": 0.0008, + "step": 20643 + }, + { + "epoch": 2.238074588031223, + "grad_norm": 4.523474216461182, + "learning_rate": 5.097925431597273e-06, + "loss": 0.009, + "step": 20644 + }, + { + "epoch": 2.2381830008673025, + "grad_norm": 4.334554195404053, + "learning_rate": 5.097200058029885e-06, + "loss": 0.0131, + "step": 20645 + }, + { + "epoch": 2.2382914137033825, + "grad_norm": 1.474791407585144, + "learning_rate": 5.096474684462499e-06, + "loss": 0.0079, + "step": 20646 + }, + { + "epoch": 2.2383998265394625, + "grad_norm": 2.6974728107452393, + "learning_rate": 5.095749310895112e-06, + "loss": 0.0124, + "step": 20647 + }, + { + "epoch": 2.238508239375542, + "grad_norm": 0.5280365347862244, + "learning_rate": 5.095023937327724e-06, + "loss": 0.0014, + "step": 20648 + }, + { + "epoch": 2.238616652211622, + "grad_norm": 0.14867950975894928, + "learning_rate": 5.094298563760337e-06, + "loss": 0.0007, + "step": 20649 + }, + { + "epoch": 2.2387250650477015, + "grad_norm": 0.019780201837420464, + "learning_rate": 5.09357319019295e-06, + "loss": 0.0002, + "step": 20650 + }, + { + "epoch": 2.2388334778837815, + "grad_norm": 1.0558561086654663, + "learning_rate": 5.092847816625562e-06, + "loss": 0.0039, + "step": 20651 + }, + { + "epoch": 2.238941890719861, + "grad_norm": 0.28343647718429565, + "learning_rate": 5.092122443058175e-06, + "loss": 0.0008, + "step": 20652 + }, + { + "epoch": 2.239050303555941, + "grad_norm": 0.0523095540702343, + "learning_rate": 5.091397069490787e-06, + "loss": 0.0004, + "step": 20653 + }, + { + "epoch": 2.239158716392021, + "grad_norm": 0.0902155190706253, + "learning_rate": 5.090671695923401e-06, + "loss": 0.0005, + "step": 20654 + }, + { + "epoch": 2.2392671292281006, + "grad_norm": 0.20289574563503265, + "learning_rate": 5.089946322356014e-06, + "loss": 0.0007, + "step": 20655 + }, + { + "epoch": 2.2393755420641805, + "grad_norm": 0.021770648658275604, + "learning_rate": 5.089220948788627e-06, + "loss": 0.0001, + "step": 20656 + }, + { + "epoch": 2.23948395490026, + "grad_norm": 0.20865878462791443, + "learning_rate": 5.088495575221239e-06, + "loss": 0.0008, + "step": 20657 + }, + { + "epoch": 2.23959236773634, + "grad_norm": 1.1065313816070557, + "learning_rate": 5.087770201653852e-06, + "loss": 0.0024, + "step": 20658 + }, + { + "epoch": 2.2397007805724196, + "grad_norm": 0.37194040417671204, + "learning_rate": 5.0870448280864644e-06, + "loss": 0.0012, + "step": 20659 + }, + { + "epoch": 2.2398091934084996, + "grad_norm": 0.9867643713951111, + "learning_rate": 5.086319454519077e-06, + "loss": 0.0023, + "step": 20660 + }, + { + "epoch": 2.239917606244579, + "grad_norm": 1.4734224081039429, + "learning_rate": 5.08559408095169e-06, + "loss": 0.0022, + "step": 20661 + }, + { + "epoch": 2.240026019080659, + "grad_norm": 0.02920421212911606, + "learning_rate": 5.0848687073843026e-06, + "loss": 0.0002, + "step": 20662 + }, + { + "epoch": 2.240134431916739, + "grad_norm": 2.8902299404144287, + "learning_rate": 5.084143333816916e-06, + "loss": 0.0049, + "step": 20663 + }, + { + "epoch": 2.2402428447528187, + "grad_norm": 0.052564337849617004, + "learning_rate": 5.083417960249529e-06, + "loss": 0.0002, + "step": 20664 + }, + { + "epoch": 2.2403512575888986, + "grad_norm": 1.7857518196105957, + "learning_rate": 5.0826925866821415e-06, + "loss": 0.0124, + "step": 20665 + }, + { + "epoch": 2.240459670424978, + "grad_norm": 0.08147778362035751, + "learning_rate": 5.0819672131147545e-06, + "loss": 0.0008, + "step": 20666 + }, + { + "epoch": 2.240568083261058, + "grad_norm": 3.384397268295288, + "learning_rate": 5.0812418395473675e-06, + "loss": 0.0392, + "step": 20667 + }, + { + "epoch": 2.2406764960971377, + "grad_norm": 2.095155954360962, + "learning_rate": 5.08051646597998e-06, + "loss": 0.0351, + "step": 20668 + }, + { + "epoch": 2.2407849089332177, + "grad_norm": 0.4230218231678009, + "learning_rate": 5.079791092412593e-06, + "loss": 0.0026, + "step": 20669 + }, + { + "epoch": 2.2408933217692977, + "grad_norm": 0.24491016566753387, + "learning_rate": 5.079065718845206e-06, + "loss": 0.0012, + "step": 20670 + }, + { + "epoch": 2.241001734605377, + "grad_norm": 3.9348111152648926, + "learning_rate": 5.078340345277818e-06, + "loss": 0.0115, + "step": 20671 + }, + { + "epoch": 2.241110147441457, + "grad_norm": 0.5458253622055054, + "learning_rate": 5.077614971710432e-06, + "loss": 0.0036, + "step": 20672 + }, + { + "epoch": 2.2412185602775367, + "grad_norm": 0.03980598598718643, + "learning_rate": 5.076889598143045e-06, + "loss": 0.0003, + "step": 20673 + }, + { + "epoch": 2.2413269731136167, + "grad_norm": 0.04602118209004402, + "learning_rate": 5.076164224575657e-06, + "loss": 0.0003, + "step": 20674 + }, + { + "epoch": 2.2414353859496963, + "grad_norm": 2.767721176147461, + "learning_rate": 5.07543885100827e-06, + "loss": 0.0093, + "step": 20675 + }, + { + "epoch": 2.2415437987857763, + "grad_norm": 3.7041075229644775, + "learning_rate": 5.074713477440883e-06, + "loss": 0.0148, + "step": 20676 + }, + { + "epoch": 2.2416522116218562, + "grad_norm": 0.181002676486969, + "learning_rate": 5.073988103873495e-06, + "loss": 0.0007, + "step": 20677 + }, + { + "epoch": 2.241760624457936, + "grad_norm": 2.615703582763672, + "learning_rate": 5.073262730306108e-06, + "loss": 0.0316, + "step": 20678 + }, + { + "epoch": 2.2418690372940158, + "grad_norm": 12.157721519470215, + "learning_rate": 5.07253735673872e-06, + "loss": 0.0146, + "step": 20679 + }, + { + "epoch": 2.2419774501300953, + "grad_norm": 0.12783823907375336, + "learning_rate": 5.071811983171334e-06, + "loss": 0.0006, + "step": 20680 + }, + { + "epoch": 2.2420858629661753, + "grad_norm": 2.99904203414917, + "learning_rate": 5.071086609603947e-06, + "loss": 0.0128, + "step": 20681 + }, + { + "epoch": 2.242194275802255, + "grad_norm": 3.742804765701294, + "learning_rate": 5.07036123603656e-06, + "loss": 0.0599, + "step": 20682 + }, + { + "epoch": 2.242302688638335, + "grad_norm": 0.16625139117240906, + "learning_rate": 5.069635862469172e-06, + "loss": 0.0004, + "step": 20683 + }, + { + "epoch": 2.2424111014744144, + "grad_norm": 3.192187547683716, + "learning_rate": 5.068910488901785e-06, + "loss": 0.006, + "step": 20684 + }, + { + "epoch": 2.2425195143104943, + "grad_norm": 4.217144966125488, + "learning_rate": 5.068185115334397e-06, + "loss": 0.0082, + "step": 20685 + }, + { + "epoch": 2.2426279271465743, + "grad_norm": 0.1681399643421173, + "learning_rate": 5.06745974176701e-06, + "loss": 0.0011, + "step": 20686 + }, + { + "epoch": 2.242736339982654, + "grad_norm": 0.10790230333805084, + "learning_rate": 5.066734368199623e-06, + "loss": 0.0007, + "step": 20687 + }, + { + "epoch": 2.242844752818734, + "grad_norm": 0.5159698128700256, + "learning_rate": 5.066008994632235e-06, + "loss": 0.0028, + "step": 20688 + }, + { + "epoch": 2.2429531656548134, + "grad_norm": 1.4026833772659302, + "learning_rate": 5.065283621064849e-06, + "loss": 0.002, + "step": 20689 + }, + { + "epoch": 2.2430615784908934, + "grad_norm": 0.5245637893676758, + "learning_rate": 5.064558247497462e-06, + "loss": 0.0019, + "step": 20690 + }, + { + "epoch": 2.243169991326973, + "grad_norm": 1.1644855737686157, + "learning_rate": 5.063832873930074e-06, + "loss": 0.0018, + "step": 20691 + }, + { + "epoch": 2.243278404163053, + "grad_norm": 0.14994464814662933, + "learning_rate": 5.063107500362687e-06, + "loss": 0.0008, + "step": 20692 + }, + { + "epoch": 2.243386816999133, + "grad_norm": 0.06784595549106598, + "learning_rate": 5.0623821267953e-06, + "loss": 0.0004, + "step": 20693 + }, + { + "epoch": 2.2434952298352124, + "grad_norm": 1.9648420810699463, + "learning_rate": 5.061656753227912e-06, + "loss": 0.0026, + "step": 20694 + }, + { + "epoch": 2.2436036426712924, + "grad_norm": 2.463862657546997, + "learning_rate": 5.060931379660525e-06, + "loss": 0.0373, + "step": 20695 + }, + { + "epoch": 2.243712055507372, + "grad_norm": 0.06147833168506622, + "learning_rate": 5.060206006093138e-06, + "loss": 0.0005, + "step": 20696 + }, + { + "epoch": 2.243820468343452, + "grad_norm": 1.83074951171875, + "learning_rate": 5.059480632525752e-06, + "loss": 0.0049, + "step": 20697 + }, + { + "epoch": 2.2439288811795315, + "grad_norm": 1.3143646717071533, + "learning_rate": 5.058755258958364e-06, + "loss": 0.0022, + "step": 20698 + }, + { + "epoch": 2.2440372940156115, + "grad_norm": 0.12374962866306305, + "learning_rate": 5.058029885390977e-06, + "loss": 0.0009, + "step": 20699 + }, + { + "epoch": 2.2441457068516915, + "grad_norm": 2.4691340923309326, + "learning_rate": 5.057304511823589e-06, + "loss": 0.0132, + "step": 20700 + }, + { + "epoch": 2.244254119687771, + "grad_norm": 3.9849331378936768, + "learning_rate": 5.056579138256202e-06, + "loss": 0.0146, + "step": 20701 + }, + { + "epoch": 2.244362532523851, + "grad_norm": 0.047613196074962616, + "learning_rate": 5.055853764688815e-06, + "loss": 0.0005, + "step": 20702 + }, + { + "epoch": 2.2444709453599305, + "grad_norm": 0.4214661717414856, + "learning_rate": 5.0551283911214275e-06, + "loss": 0.0018, + "step": 20703 + }, + { + "epoch": 2.2445793581960105, + "grad_norm": 1.22254478931427, + "learning_rate": 5.0544030175540405e-06, + "loss": 0.0011, + "step": 20704 + }, + { + "epoch": 2.24468777103209, + "grad_norm": 0.0358608178794384, + "learning_rate": 5.053677643986653e-06, + "loss": 0.0004, + "step": 20705 + }, + { + "epoch": 2.24479618386817, + "grad_norm": 0.05789504572749138, + "learning_rate": 5.0529522704192665e-06, + "loss": 0.0002, + "step": 20706 + }, + { + "epoch": 2.2449045967042496, + "grad_norm": 1.3933132886886597, + "learning_rate": 5.0522268968518795e-06, + "loss": 0.0185, + "step": 20707 + }, + { + "epoch": 2.2450130095403296, + "grad_norm": 7.510309219360352, + "learning_rate": 5.0515015232844925e-06, + "loss": 0.0167, + "step": 20708 + }, + { + "epoch": 2.2451214223764095, + "grad_norm": 0.6401904225349426, + "learning_rate": 5.050776149717105e-06, + "loss": 0.002, + "step": 20709 + }, + { + "epoch": 2.245229835212489, + "grad_norm": 0.018565095961093903, + "learning_rate": 5.050050776149718e-06, + "loss": 0.0002, + "step": 20710 + }, + { + "epoch": 2.245338248048569, + "grad_norm": 9.047266006469727, + "learning_rate": 5.04932540258233e-06, + "loss": 0.0168, + "step": 20711 + }, + { + "epoch": 2.2454466608846486, + "grad_norm": 0.6711816191673279, + "learning_rate": 5.048600029014943e-06, + "loss": 0.0014, + "step": 20712 + }, + { + "epoch": 2.2455550737207286, + "grad_norm": 0.07727126777172089, + "learning_rate": 5.047874655447556e-06, + "loss": 0.0006, + "step": 20713 + }, + { + "epoch": 2.245663486556808, + "grad_norm": 2.742661952972412, + "learning_rate": 5.047149281880168e-06, + "loss": 0.0084, + "step": 20714 + }, + { + "epoch": 2.245771899392888, + "grad_norm": 0.2367740273475647, + "learning_rate": 5.046423908312782e-06, + "loss": 0.0007, + "step": 20715 + }, + { + "epoch": 2.245880312228968, + "grad_norm": 0.06621779501438141, + "learning_rate": 5.045698534745395e-06, + "loss": 0.0005, + "step": 20716 + }, + { + "epoch": 2.2459887250650477, + "grad_norm": 6.073638916015625, + "learning_rate": 5.044973161178007e-06, + "loss": 0.0833, + "step": 20717 + }, + { + "epoch": 2.2460971379011276, + "grad_norm": 0.11841356009244919, + "learning_rate": 5.04424778761062e-06, + "loss": 0.0004, + "step": 20718 + }, + { + "epoch": 2.246205550737207, + "grad_norm": 2.8981144428253174, + "learning_rate": 5.043522414043233e-06, + "loss": 0.0061, + "step": 20719 + }, + { + "epoch": 2.246313963573287, + "grad_norm": 0.8267769813537598, + "learning_rate": 5.042797040475845e-06, + "loss": 0.0019, + "step": 20720 + }, + { + "epoch": 2.2464223764093667, + "grad_norm": 2.893357992172241, + "learning_rate": 5.042071666908458e-06, + "loss": 0.0202, + "step": 20721 + }, + { + "epoch": 2.2465307892454467, + "grad_norm": 0.15632222592830658, + "learning_rate": 5.041346293341071e-06, + "loss": 0.0011, + "step": 20722 + }, + { + "epoch": 2.2466392020815267, + "grad_norm": 2.816053628921509, + "learning_rate": 5.040620919773685e-06, + "loss": 0.0595, + "step": 20723 + }, + { + "epoch": 2.246747614917606, + "grad_norm": 3.262096643447876, + "learning_rate": 5.039895546206297e-06, + "loss": 0.0135, + "step": 20724 + }, + { + "epoch": 2.246856027753686, + "grad_norm": 4.1210455894470215, + "learning_rate": 5.03917017263891e-06, + "loss": 0.0473, + "step": 20725 + }, + { + "epoch": 2.2469644405897657, + "grad_norm": 0.8284714221954346, + "learning_rate": 5.038444799071522e-06, + "loss": 0.0041, + "step": 20726 + }, + { + "epoch": 2.2470728534258457, + "grad_norm": 6.847797870635986, + "learning_rate": 5.037719425504135e-06, + "loss": 0.0079, + "step": 20727 + }, + { + "epoch": 2.2471812662619253, + "grad_norm": 0.43462085723876953, + "learning_rate": 5.036994051936748e-06, + "loss": 0.0018, + "step": 20728 + }, + { + "epoch": 2.2472896790980053, + "grad_norm": 2.733783721923828, + "learning_rate": 5.03626867836936e-06, + "loss": 0.0111, + "step": 20729 + }, + { + "epoch": 2.247398091934085, + "grad_norm": 5.931308269500732, + "learning_rate": 5.035543304801973e-06, + "loss": 0.0635, + "step": 20730 + }, + { + "epoch": 2.247506504770165, + "grad_norm": 2.7582855224609375, + "learning_rate": 5.034817931234585e-06, + "loss": 0.0063, + "step": 20731 + }, + { + "epoch": 2.2476149176062448, + "grad_norm": 2.4252243041992188, + "learning_rate": 5.034092557667199e-06, + "loss": 0.0097, + "step": 20732 + }, + { + "epoch": 2.2477233304423243, + "grad_norm": 0.24760611355304718, + "learning_rate": 5.033367184099812e-06, + "loss": 0.0005, + "step": 20733 + }, + { + "epoch": 2.2478317432784043, + "grad_norm": 0.10202886164188385, + "learning_rate": 5.032641810532425e-06, + "loss": 0.0008, + "step": 20734 + }, + { + "epoch": 2.247940156114484, + "grad_norm": 7.603097438812256, + "learning_rate": 5.031916436965037e-06, + "loss": 0.019, + "step": 20735 + }, + { + "epoch": 2.248048568950564, + "grad_norm": 0.6493788361549377, + "learning_rate": 5.03119106339765e-06, + "loss": 0.0019, + "step": 20736 + }, + { + "epoch": 2.2481569817866434, + "grad_norm": 0.44164127111434937, + "learning_rate": 5.0304656898302624e-06, + "loss": 0.0013, + "step": 20737 + }, + { + "epoch": 2.2482653946227233, + "grad_norm": 0.11188197135925293, + "learning_rate": 5.0297403162628754e-06, + "loss": 0.0007, + "step": 20738 + }, + { + "epoch": 2.248373807458803, + "grad_norm": 0.1512579619884491, + "learning_rate": 5.029014942695488e-06, + "loss": 0.0004, + "step": 20739 + }, + { + "epoch": 2.248482220294883, + "grad_norm": 2.4648423194885254, + "learning_rate": 5.0282895691281006e-06, + "loss": 0.0093, + "step": 20740 + }, + { + "epoch": 2.248590633130963, + "grad_norm": 4.917675971984863, + "learning_rate": 5.027564195560714e-06, + "loss": 0.0255, + "step": 20741 + }, + { + "epoch": 2.2486990459670424, + "grad_norm": 1.8745580911636353, + "learning_rate": 5.026838821993327e-06, + "loss": 0.0022, + "step": 20742 + }, + { + "epoch": 2.2488074588031224, + "grad_norm": 0.029432082548737526, + "learning_rate": 5.02611344842594e-06, + "loss": 0.0002, + "step": 20743 + }, + { + "epoch": 2.248915871639202, + "grad_norm": 1.5854905843734741, + "learning_rate": 5.0253880748585525e-06, + "loss": 0.0017, + "step": 20744 + }, + { + "epoch": 2.249024284475282, + "grad_norm": 19.06135368347168, + "learning_rate": 5.0246627012911655e-06, + "loss": 0.0274, + "step": 20745 + }, + { + "epoch": 2.249132697311362, + "grad_norm": 0.6331288814544678, + "learning_rate": 5.023937327723778e-06, + "loss": 0.0019, + "step": 20746 + }, + { + "epoch": 2.2492411101474414, + "grad_norm": 8.108207702636719, + "learning_rate": 5.023211954156391e-06, + "loss": 0.0131, + "step": 20747 + }, + { + "epoch": 2.2493495229835214, + "grad_norm": 4.362578868865967, + "learning_rate": 5.022486580589004e-06, + "loss": 0.0238, + "step": 20748 + }, + { + "epoch": 2.249457935819601, + "grad_norm": 3.479875087738037, + "learning_rate": 5.0217612070216175e-06, + "loss": 0.0071, + "step": 20749 + }, + { + "epoch": 2.249566348655681, + "grad_norm": 0.09731383621692657, + "learning_rate": 5.02103583345423e-06, + "loss": 0.0007, + "step": 20750 + }, + { + "epoch": 2.2496747614917605, + "grad_norm": 2.5974185466766357, + "learning_rate": 5.020310459886843e-06, + "loss": 0.0239, + "step": 20751 + }, + { + "epoch": 2.2497831743278405, + "grad_norm": 1.2586309909820557, + "learning_rate": 5.019585086319455e-06, + "loss": 0.0038, + "step": 20752 + }, + { + "epoch": 2.24989158716392, + "grad_norm": 0.015691706910729408, + "learning_rate": 5.018859712752068e-06, + "loss": 0.0001, + "step": 20753 + }, + { + "epoch": 2.25, + "grad_norm": 3.6274564266204834, + "learning_rate": 5.018134339184681e-06, + "loss": 0.0182, + "step": 20754 + }, + { + "epoch": 2.25010841283608, + "grad_norm": 0.4588405191898346, + "learning_rate": 5.017408965617293e-06, + "loss": 0.003, + "step": 20755 + }, + { + "epoch": 2.2502168256721595, + "grad_norm": 2.69832706451416, + "learning_rate": 5.016683592049906e-06, + "loss": 0.0234, + "step": 20756 + }, + { + "epoch": 2.2503252385082395, + "grad_norm": 0.033740658313035965, + "learning_rate": 5.015958218482518e-06, + "loss": 0.0002, + "step": 20757 + }, + { + "epoch": 2.250433651344319, + "grad_norm": 2.6826930046081543, + "learning_rate": 5.015232844915132e-06, + "loss": 0.0041, + "step": 20758 + }, + { + "epoch": 2.250542064180399, + "grad_norm": 0.8755128383636475, + "learning_rate": 5.014507471347745e-06, + "loss": 0.0027, + "step": 20759 + }, + { + "epoch": 2.2506504770164786, + "grad_norm": 3.133312940597534, + "learning_rate": 5.013782097780358e-06, + "loss": 0.0137, + "step": 20760 + }, + { + "epoch": 2.2507588898525586, + "grad_norm": 17.60947036743164, + "learning_rate": 5.01305672421297e-06, + "loss": 0.0205, + "step": 20761 + }, + { + "epoch": 2.250867302688638, + "grad_norm": 5.300087928771973, + "learning_rate": 5.012331350645583e-06, + "loss": 0.0188, + "step": 20762 + }, + { + "epoch": 2.250975715524718, + "grad_norm": 3.1235897541046143, + "learning_rate": 5.011605977078195e-06, + "loss": 0.0076, + "step": 20763 + }, + { + "epoch": 2.251084128360798, + "grad_norm": 0.13099592924118042, + "learning_rate": 5.010880603510808e-06, + "loss": 0.0006, + "step": 20764 + }, + { + "epoch": 2.2511925411968776, + "grad_norm": 2.579253673553467, + "learning_rate": 5.010155229943421e-06, + "loss": 0.0047, + "step": 20765 + }, + { + "epoch": 2.2513009540329576, + "grad_norm": 5.8708696365356445, + "learning_rate": 5.009429856376033e-06, + "loss": 0.0181, + "step": 20766 + }, + { + "epoch": 2.251409366869037, + "grad_norm": 0.04955681785941124, + "learning_rate": 5.008704482808647e-06, + "loss": 0.0005, + "step": 20767 + }, + { + "epoch": 2.251517779705117, + "grad_norm": 0.02172638103365898, + "learning_rate": 5.00797910924126e-06, + "loss": 0.0001, + "step": 20768 + }, + { + "epoch": 2.251626192541197, + "grad_norm": 0.030894620344042778, + "learning_rate": 5.007253735673873e-06, + "loss": 0.0002, + "step": 20769 + }, + { + "epoch": 2.2517346053772767, + "grad_norm": 2.0302329063415527, + "learning_rate": 5.006528362106485e-06, + "loss": 0.0033, + "step": 20770 + }, + { + "epoch": 2.2518430182133566, + "grad_norm": 0.07157522439956665, + "learning_rate": 5.005802988539098e-06, + "loss": 0.0003, + "step": 20771 + }, + { + "epoch": 2.251951431049436, + "grad_norm": 2.6761975288391113, + "learning_rate": 5.00507761497171e-06, + "loss": 0.0049, + "step": 20772 + }, + { + "epoch": 2.252059843885516, + "grad_norm": 1.3338589668273926, + "learning_rate": 5.004352241404323e-06, + "loss": 0.0045, + "step": 20773 + }, + { + "epoch": 2.2521682567215957, + "grad_norm": 10.455256462097168, + "learning_rate": 5.003626867836936e-06, + "loss": 0.0334, + "step": 20774 + }, + { + "epoch": 2.2522766695576757, + "grad_norm": 2.6180591583251953, + "learning_rate": 5.00290149426955e-06, + "loss": 0.0322, + "step": 20775 + }, + { + "epoch": 2.2523850823937552, + "grad_norm": 0.11527751386165619, + "learning_rate": 5.002176120702162e-06, + "loss": 0.0003, + "step": 20776 + }, + { + "epoch": 2.252493495229835, + "grad_norm": 0.9419148564338684, + "learning_rate": 5.001450747134775e-06, + "loss": 0.0015, + "step": 20777 + }, + { + "epoch": 2.252601908065915, + "grad_norm": 2.060910224914551, + "learning_rate": 5.000725373567387e-06, + "loss": 0.0092, + "step": 20778 + }, + { + "epoch": 2.2527103209019947, + "grad_norm": 0.018372856080532074, + "learning_rate": 5e-06, + "loss": 0.0001, + "step": 20779 + }, + { + "epoch": 2.2528187337380747, + "grad_norm": 5.633228778839111, + "learning_rate": 4.999274626432613e-06, + "loss": 0.0214, + "step": 20780 + }, + { + "epoch": 2.2529271465741543, + "grad_norm": 1.7319000959396362, + "learning_rate": 4.9985492528652255e-06, + "loss": 0.0064, + "step": 20781 + }, + { + "epoch": 2.2530355594102343, + "grad_norm": 0.12682953476905823, + "learning_rate": 4.997823879297839e-06, + "loss": 0.0007, + "step": 20782 + }, + { + "epoch": 2.253143972246314, + "grad_norm": 1.5755101442337036, + "learning_rate": 4.9970985057304515e-06, + "loss": 0.0166, + "step": 20783 + }, + { + "epoch": 2.253252385082394, + "grad_norm": 0.01085817813873291, + "learning_rate": 4.9963731321630645e-06, + "loss": 0.0001, + "step": 20784 + }, + { + "epoch": 2.2533607979184733, + "grad_norm": 5.287868499755859, + "learning_rate": 4.995647758595677e-06, + "loss": 0.0136, + "step": 20785 + }, + { + "epoch": 2.2534692107545533, + "grad_norm": 1.8613276481628418, + "learning_rate": 4.9949223850282905e-06, + "loss": 0.0105, + "step": 20786 + }, + { + "epoch": 2.2535776235906333, + "grad_norm": 5.480228900909424, + "learning_rate": 4.994197011460903e-06, + "loss": 0.0043, + "step": 20787 + }, + { + "epoch": 2.253686036426713, + "grad_norm": 0.12195557355880737, + "learning_rate": 4.993471637893516e-06, + "loss": 0.0002, + "step": 20788 + }, + { + "epoch": 2.253794449262793, + "grad_norm": 0.07702472805976868, + "learning_rate": 4.992746264326128e-06, + "loss": 0.0006, + "step": 20789 + }, + { + "epoch": 2.2539028620988724, + "grad_norm": 0.03574967011809349, + "learning_rate": 4.992020890758742e-06, + "loss": 0.0001, + "step": 20790 + }, + { + "epoch": 2.2540112749349523, + "grad_norm": 0.7472628951072693, + "learning_rate": 4.991295517191354e-06, + "loss": 0.0004, + "step": 20791 + }, + { + "epoch": 2.2541196877710323, + "grad_norm": 0.03467756137251854, + "learning_rate": 4.990570143623967e-06, + "loss": 0.0003, + "step": 20792 + }, + { + "epoch": 2.254228100607112, + "grad_norm": 0.04650304093956947, + "learning_rate": 4.98984477005658e-06, + "loss": 0.0003, + "step": 20793 + }, + { + "epoch": 2.254336513443192, + "grad_norm": 3.3135292530059814, + "learning_rate": 4.989119396489192e-06, + "loss": 0.0239, + "step": 20794 + }, + { + "epoch": 2.2544449262792714, + "grad_norm": 0.03244015946984291, + "learning_rate": 4.988394022921806e-06, + "loss": 0.0003, + "step": 20795 + }, + { + "epoch": 2.2545533391153514, + "grad_norm": 2.1547439098358154, + "learning_rate": 4.987668649354418e-06, + "loss": 0.0395, + "step": 20796 + }, + { + "epoch": 2.254661751951431, + "grad_norm": 0.013813270255923271, + "learning_rate": 4.986943275787031e-06, + "loss": 0.0002, + "step": 20797 + }, + { + "epoch": 2.254770164787511, + "grad_norm": 0.22480769455432892, + "learning_rate": 4.986217902219643e-06, + "loss": 0.0016, + "step": 20798 + }, + { + "epoch": 2.2548785776235905, + "grad_norm": 3.615914821624756, + "learning_rate": 4.985492528652257e-06, + "loss": 0.0418, + "step": 20799 + }, + { + "epoch": 2.2549869904596704, + "grad_norm": 0.5979761481285095, + "learning_rate": 4.984767155084869e-06, + "loss": 0.0022, + "step": 20800 + }, + { + "epoch": 2.2550954032957504, + "grad_norm": 1.5372108221054077, + "learning_rate": 4.984041781517482e-06, + "loss": 0.0048, + "step": 20801 + }, + { + "epoch": 2.25520381613183, + "grad_norm": 6.719300746917725, + "learning_rate": 4.983316407950094e-06, + "loss": 0.004, + "step": 20802 + }, + { + "epoch": 2.25531222896791, + "grad_norm": 0.47748926281929016, + "learning_rate": 4.982591034382708e-06, + "loss": 0.0009, + "step": 20803 + }, + { + "epoch": 2.2554206418039895, + "grad_norm": 1.3686808347702026, + "learning_rate": 4.98186566081532e-06, + "loss": 0.0035, + "step": 20804 + }, + { + "epoch": 2.2555290546400695, + "grad_norm": 0.8365625739097595, + "learning_rate": 4.981140287247933e-06, + "loss": 0.0027, + "step": 20805 + }, + { + "epoch": 2.255637467476149, + "grad_norm": 1.6168222427368164, + "learning_rate": 4.980414913680546e-06, + "loss": 0.0501, + "step": 20806 + }, + { + "epoch": 2.255745880312229, + "grad_norm": 4.283636569976807, + "learning_rate": 4.979689540113158e-06, + "loss": 0.0032, + "step": 20807 + }, + { + "epoch": 2.2558542931483085, + "grad_norm": 1.4942409992218018, + "learning_rate": 4.978964166545772e-06, + "loss": 0.053, + "step": 20808 + }, + { + "epoch": 2.2559627059843885, + "grad_norm": 0.8831513524055481, + "learning_rate": 4.978238792978384e-06, + "loss": 0.0019, + "step": 20809 + }, + { + "epoch": 2.2560711188204685, + "grad_norm": 0.23507820069789886, + "learning_rate": 4.977513419410997e-06, + "loss": 0.0019, + "step": 20810 + }, + { + "epoch": 2.256179531656548, + "grad_norm": 1.074774146080017, + "learning_rate": 4.976788045843609e-06, + "loss": 0.002, + "step": 20811 + }, + { + "epoch": 2.256287944492628, + "grad_norm": 1.4140621423721313, + "learning_rate": 4.976062672276223e-06, + "loss": 0.002, + "step": 20812 + }, + { + "epoch": 2.2563963573287076, + "grad_norm": 0.09746062755584717, + "learning_rate": 4.975337298708835e-06, + "loss": 0.0005, + "step": 20813 + }, + { + "epoch": 2.2565047701647876, + "grad_norm": 0.08063157647848129, + "learning_rate": 4.974611925141448e-06, + "loss": 0.0006, + "step": 20814 + }, + { + "epoch": 2.2566131830008676, + "grad_norm": 0.35656413435935974, + "learning_rate": 4.9738865515740604e-06, + "loss": 0.0023, + "step": 20815 + }, + { + "epoch": 2.256721595836947, + "grad_norm": 6.224785327911377, + "learning_rate": 4.973161178006674e-06, + "loss": 0.0228, + "step": 20816 + }, + { + "epoch": 2.256830008673027, + "grad_norm": 1.2021009922027588, + "learning_rate": 4.972435804439286e-06, + "loss": 0.0033, + "step": 20817 + }, + { + "epoch": 2.2569384215091066, + "grad_norm": 0.11957857757806778, + "learning_rate": 4.971710430871899e-06, + "loss": 0.0008, + "step": 20818 + }, + { + "epoch": 2.2570468343451866, + "grad_norm": 0.10301321744918823, + "learning_rate": 4.970985057304512e-06, + "loss": 0.0006, + "step": 20819 + }, + { + "epoch": 2.257155247181266, + "grad_norm": 0.1316286325454712, + "learning_rate": 4.9702596837371245e-06, + "loss": 0.0009, + "step": 20820 + }, + { + "epoch": 2.257263660017346, + "grad_norm": 3.8921234607696533, + "learning_rate": 4.969534310169738e-06, + "loss": 0.0031, + "step": 20821 + }, + { + "epoch": 2.2573720728534257, + "grad_norm": 0.2682565450668335, + "learning_rate": 4.9688089366023505e-06, + "loss": 0.0015, + "step": 20822 + }, + { + "epoch": 2.2574804856895057, + "grad_norm": 0.2534610331058502, + "learning_rate": 4.9680835630349635e-06, + "loss": 0.0017, + "step": 20823 + }, + { + "epoch": 2.2575888985255856, + "grad_norm": 0.5861396193504333, + "learning_rate": 4.967358189467576e-06, + "loss": 0.0024, + "step": 20824 + }, + { + "epoch": 2.257697311361665, + "grad_norm": 0.0062347641214728355, + "learning_rate": 4.9666328159001895e-06, + "loss": 0.0001, + "step": 20825 + }, + { + "epoch": 2.257805724197745, + "grad_norm": 2.78963303565979, + "learning_rate": 4.965907442332802e-06, + "loss": 0.0026, + "step": 20826 + }, + { + "epoch": 2.2579141370338247, + "grad_norm": 0.9261437654495239, + "learning_rate": 4.965182068765415e-06, + "loss": 0.0014, + "step": 20827 + }, + { + "epoch": 2.2580225498699047, + "grad_norm": 0.8501366972923279, + "learning_rate": 4.964456695198027e-06, + "loss": 0.0028, + "step": 20828 + }, + { + "epoch": 2.2581309627059842, + "grad_norm": 0.28113359212875366, + "learning_rate": 4.963731321630641e-06, + "loss": 0.0017, + "step": 20829 + }, + { + "epoch": 2.258239375542064, + "grad_norm": 3.432515859603882, + "learning_rate": 4.963005948063253e-06, + "loss": 0.016, + "step": 20830 + }, + { + "epoch": 2.2583477883781438, + "grad_norm": 5.456948280334473, + "learning_rate": 4.962280574495866e-06, + "loss": 0.0089, + "step": 20831 + }, + { + "epoch": 2.2584562012142237, + "grad_norm": 0.6712867617607117, + "learning_rate": 4.961555200928479e-06, + "loss": 0.0007, + "step": 20832 + }, + { + "epoch": 2.2585646140503037, + "grad_norm": 9.013051986694336, + "learning_rate": 4.960829827361092e-06, + "loss": 0.0135, + "step": 20833 + }, + { + "epoch": 2.2586730268863833, + "grad_norm": 6.904846668243408, + "learning_rate": 4.960104453793705e-06, + "loss": 0.0128, + "step": 20834 + }, + { + "epoch": 2.2587814397224633, + "grad_norm": 0.3224292993545532, + "learning_rate": 4.959379080226317e-06, + "loss": 0.0017, + "step": 20835 + }, + { + "epoch": 2.258889852558543, + "grad_norm": 1.297669768333435, + "learning_rate": 4.95865370665893e-06, + "loss": 0.0034, + "step": 20836 + }, + { + "epoch": 2.258998265394623, + "grad_norm": 1.8447086811065674, + "learning_rate": 4.957928333091542e-06, + "loss": 0.0295, + "step": 20837 + }, + { + "epoch": 2.2591066782307023, + "grad_norm": 0.029423370957374573, + "learning_rate": 4.957202959524156e-06, + "loss": 0.0003, + "step": 20838 + }, + { + "epoch": 2.2592150910667823, + "grad_norm": 1.5854562520980835, + "learning_rate": 4.956477585956768e-06, + "loss": 0.0207, + "step": 20839 + }, + { + "epoch": 2.259323503902862, + "grad_norm": 0.6947726607322693, + "learning_rate": 4.955752212389381e-06, + "loss": 0.0012, + "step": 20840 + }, + { + "epoch": 2.259431916738942, + "grad_norm": 0.36962881684303284, + "learning_rate": 4.955026838821994e-06, + "loss": 0.0021, + "step": 20841 + }, + { + "epoch": 2.259540329575022, + "grad_norm": 0.22107143700122833, + "learning_rate": 4.954301465254607e-06, + "loss": 0.0009, + "step": 20842 + }, + { + "epoch": 2.2596487424111014, + "grad_norm": 0.5108480453491211, + "learning_rate": 4.953576091687219e-06, + "loss": 0.0019, + "step": 20843 + }, + { + "epoch": 2.2597571552471813, + "grad_norm": 0.0510033518075943, + "learning_rate": 4.952850718119832e-06, + "loss": 0.0005, + "step": 20844 + }, + { + "epoch": 2.259865568083261, + "grad_norm": 2.7798705101013184, + "learning_rate": 4.952125344552445e-06, + "loss": 0.0255, + "step": 20845 + }, + { + "epoch": 2.259973980919341, + "grad_norm": 2.742985725402832, + "learning_rate": 4.951399970985058e-06, + "loss": 0.0228, + "step": 20846 + }, + { + "epoch": 2.260082393755421, + "grad_norm": 0.049859825521707535, + "learning_rate": 4.950674597417671e-06, + "loss": 0.0004, + "step": 20847 + }, + { + "epoch": 2.2601908065915004, + "grad_norm": 8.575931549072266, + "learning_rate": 4.949949223850283e-06, + "loss": 0.0104, + "step": 20848 + }, + { + "epoch": 2.2602992194275804, + "grad_norm": 0.6597049236297607, + "learning_rate": 4.949223850282896e-06, + "loss": 0.0018, + "step": 20849 + }, + { + "epoch": 2.26040763226366, + "grad_norm": 0.08963344991207123, + "learning_rate": 4.948498476715508e-06, + "loss": 0.0005, + "step": 20850 + }, + { + "epoch": 2.26051604509974, + "grad_norm": 0.20377543568611145, + "learning_rate": 4.947773103148122e-06, + "loss": 0.0004, + "step": 20851 + }, + { + "epoch": 2.2606244579358195, + "grad_norm": 5.063832759857178, + "learning_rate": 4.947047729580734e-06, + "loss": 0.0456, + "step": 20852 + }, + { + "epoch": 2.2607328707718994, + "grad_norm": 0.657673180103302, + "learning_rate": 4.946322356013347e-06, + "loss": 0.0023, + "step": 20853 + }, + { + "epoch": 2.260841283607979, + "grad_norm": 0.10151167958974838, + "learning_rate": 4.94559698244596e-06, + "loss": 0.0006, + "step": 20854 + }, + { + "epoch": 2.260949696444059, + "grad_norm": 0.10725279152393341, + "learning_rate": 4.944871608878573e-06, + "loss": 0.0007, + "step": 20855 + }, + { + "epoch": 2.261058109280139, + "grad_norm": 3.5395164489746094, + "learning_rate": 4.9441462353111854e-06, + "loss": 0.0108, + "step": 20856 + }, + { + "epoch": 2.2611665221162185, + "grad_norm": 0.337627112865448, + "learning_rate": 4.943420861743798e-06, + "loss": 0.0008, + "step": 20857 + }, + { + "epoch": 2.2612749349522985, + "grad_norm": 0.09630410373210907, + "learning_rate": 4.942695488176411e-06, + "loss": 0.001, + "step": 20858 + }, + { + "epoch": 2.261383347788378, + "grad_norm": 0.5065947771072388, + "learning_rate": 4.941970114609024e-06, + "loss": 0.0021, + "step": 20859 + }, + { + "epoch": 2.261491760624458, + "grad_norm": 0.07692786306142807, + "learning_rate": 4.941244741041637e-06, + "loss": 0.0005, + "step": 20860 + }, + { + "epoch": 2.2616001734605375, + "grad_norm": 0.17519792914390564, + "learning_rate": 4.9405193674742495e-06, + "loss": 0.0011, + "step": 20861 + }, + { + "epoch": 2.2617085862966175, + "grad_norm": 5.009507656097412, + "learning_rate": 4.9397939939068625e-06, + "loss": 0.0369, + "step": 20862 + }, + { + "epoch": 2.261816999132697, + "grad_norm": 5.153097629547119, + "learning_rate": 4.939068620339475e-06, + "loss": 0.0058, + "step": 20863 + }, + { + "epoch": 2.261925411968777, + "grad_norm": 0.043743718415498734, + "learning_rate": 4.9383432467720885e-06, + "loss": 0.0004, + "step": 20864 + }, + { + "epoch": 2.262033824804857, + "grad_norm": 11.955217361450195, + "learning_rate": 4.937617873204701e-06, + "loss": 0.0574, + "step": 20865 + }, + { + "epoch": 2.2621422376409366, + "grad_norm": 1.7506965398788452, + "learning_rate": 4.936892499637314e-06, + "loss": 0.0033, + "step": 20866 + }, + { + "epoch": 2.2622506504770166, + "grad_norm": 0.22957859933376312, + "learning_rate": 4.936167126069927e-06, + "loss": 0.0008, + "step": 20867 + }, + { + "epoch": 2.262359063313096, + "grad_norm": 5.880843162536621, + "learning_rate": 4.93544175250254e-06, + "loss": 0.0026, + "step": 20868 + }, + { + "epoch": 2.262467476149176, + "grad_norm": 0.050294239073991776, + "learning_rate": 4.934716378935152e-06, + "loss": 0.0004, + "step": 20869 + }, + { + "epoch": 2.262575888985256, + "grad_norm": 0.04947029799222946, + "learning_rate": 4.933991005367765e-06, + "loss": 0.0003, + "step": 20870 + }, + { + "epoch": 2.2626843018213356, + "grad_norm": 3.2835302352905273, + "learning_rate": 4.933265631800378e-06, + "loss": 0.0097, + "step": 20871 + }, + { + "epoch": 2.2627927146574156, + "grad_norm": 0.2882920503616333, + "learning_rate": 4.932540258232991e-06, + "loss": 0.0007, + "step": 20872 + }, + { + "epoch": 2.262901127493495, + "grad_norm": 0.7778895497322083, + "learning_rate": 4.931814884665604e-06, + "loss": 0.0014, + "step": 20873 + }, + { + "epoch": 2.263009540329575, + "grad_norm": 3.8423702716827393, + "learning_rate": 4.931089511098216e-06, + "loss": 0.0421, + "step": 20874 + }, + { + "epoch": 2.2631179531656547, + "grad_norm": 0.20316098630428314, + "learning_rate": 4.930364137530829e-06, + "loss": 0.0008, + "step": 20875 + }, + { + "epoch": 2.2632263660017347, + "grad_norm": 0.03766325116157532, + "learning_rate": 4.929638763963441e-06, + "loss": 0.0003, + "step": 20876 + }, + { + "epoch": 2.263334778837814, + "grad_norm": 0.06038046255707741, + "learning_rate": 4.928913390396055e-06, + "loss": 0.0003, + "step": 20877 + }, + { + "epoch": 2.263443191673894, + "grad_norm": 7.407568454742432, + "learning_rate": 4.928188016828667e-06, + "loss": 0.0431, + "step": 20878 + }, + { + "epoch": 2.263551604509974, + "grad_norm": 11.340458869934082, + "learning_rate": 4.92746264326128e-06, + "loss": 0.0089, + "step": 20879 + }, + { + "epoch": 2.2636600173460537, + "grad_norm": 1.316179633140564, + "learning_rate": 4.926737269693893e-06, + "loss": 0.0014, + "step": 20880 + }, + { + "epoch": 2.2637684301821337, + "grad_norm": 0.053923219442367554, + "learning_rate": 4.926011896126506e-06, + "loss": 0.0005, + "step": 20881 + }, + { + "epoch": 2.2638768430182132, + "grad_norm": 0.033409539610147476, + "learning_rate": 4.925286522559118e-06, + "loss": 0.0003, + "step": 20882 + }, + { + "epoch": 2.263985255854293, + "grad_norm": 1.1263506412506104, + "learning_rate": 4.924561148991731e-06, + "loss": 0.0004, + "step": 20883 + }, + { + "epoch": 2.2640936686903728, + "grad_norm": 2.7534453868865967, + "learning_rate": 4.923835775424344e-06, + "loss": 0.0209, + "step": 20884 + }, + { + "epoch": 2.2642020815264527, + "grad_norm": 0.039327073842287064, + "learning_rate": 4.923110401856957e-06, + "loss": 0.0002, + "step": 20885 + }, + { + "epoch": 2.2643104943625323, + "grad_norm": 3.4832136631011963, + "learning_rate": 4.92238502828957e-06, + "loss": 0.0059, + "step": 20886 + }, + { + "epoch": 2.2644189071986123, + "grad_norm": 0.5292734503746033, + "learning_rate": 4.921659654722182e-06, + "loss": 0.0014, + "step": 20887 + }, + { + "epoch": 2.2645273200346923, + "grad_norm": 0.8653016090393066, + "learning_rate": 4.920934281154795e-06, + "loss": 0.0018, + "step": 20888 + }, + { + "epoch": 2.264635732870772, + "grad_norm": 4.177196979522705, + "learning_rate": 4.920208907587407e-06, + "loss": 0.0412, + "step": 20889 + }, + { + "epoch": 2.264744145706852, + "grad_norm": 0.02980128489434719, + "learning_rate": 4.919483534020021e-06, + "loss": 0.0002, + "step": 20890 + }, + { + "epoch": 2.2648525585429313, + "grad_norm": 0.11806151270866394, + "learning_rate": 4.918758160452633e-06, + "loss": 0.0003, + "step": 20891 + }, + { + "epoch": 2.2649609713790113, + "grad_norm": 4.5971879959106445, + "learning_rate": 4.918032786885246e-06, + "loss": 0.0311, + "step": 20892 + }, + { + "epoch": 2.2650693842150913, + "grad_norm": 0.05141831561923027, + "learning_rate": 4.917307413317859e-06, + "loss": 0.0005, + "step": 20893 + }, + { + "epoch": 2.265177797051171, + "grad_norm": 8.909313201904297, + "learning_rate": 4.916582039750472e-06, + "loss": 0.03, + "step": 20894 + }, + { + "epoch": 2.265286209887251, + "grad_norm": 0.021012019366025925, + "learning_rate": 4.9158566661830844e-06, + "loss": 0.0001, + "step": 20895 + }, + { + "epoch": 2.2653946227233304, + "grad_norm": 3.659806489944458, + "learning_rate": 4.915131292615697e-06, + "loss": 0.0717, + "step": 20896 + }, + { + "epoch": 2.2655030355594103, + "grad_norm": 3.1863694190979004, + "learning_rate": 4.91440591904831e-06, + "loss": 0.0069, + "step": 20897 + }, + { + "epoch": 2.26561144839549, + "grad_norm": 0.10363848507404327, + "learning_rate": 4.913680545480923e-06, + "loss": 0.0002, + "step": 20898 + }, + { + "epoch": 2.26571986123157, + "grad_norm": 0.5706408023834229, + "learning_rate": 4.912955171913536e-06, + "loss": 0.0015, + "step": 20899 + }, + { + "epoch": 2.2658282740676494, + "grad_norm": 0.9851208329200745, + "learning_rate": 4.9122297983461485e-06, + "loss": 0.0401, + "step": 20900 + }, + { + "epoch": 2.2659366869037294, + "grad_norm": 9.310181617736816, + "learning_rate": 4.9115044247787615e-06, + "loss": 0.0087, + "step": 20901 + }, + { + "epoch": 2.2660450997398094, + "grad_norm": 13.763266563415527, + "learning_rate": 4.910779051211374e-06, + "loss": 0.0335, + "step": 20902 + }, + { + "epoch": 2.266153512575889, + "grad_norm": 0.9179865121841431, + "learning_rate": 4.9100536776439875e-06, + "loss": 0.0024, + "step": 20903 + }, + { + "epoch": 2.266261925411969, + "grad_norm": 0.056764692068099976, + "learning_rate": 4.9093283040766e-06, + "loss": 0.0002, + "step": 20904 + }, + { + "epoch": 2.2663703382480485, + "grad_norm": 2.4895496368408203, + "learning_rate": 4.908602930509213e-06, + "loss": 0.0051, + "step": 20905 + }, + { + "epoch": 2.2664787510841284, + "grad_norm": 0.08321560174226761, + "learning_rate": 4.907877556941826e-06, + "loss": 0.0007, + "step": 20906 + }, + { + "epoch": 2.266587163920208, + "grad_norm": 2.6573100090026855, + "learning_rate": 4.907152183374439e-06, + "loss": 0.0112, + "step": 20907 + }, + { + "epoch": 2.266695576756288, + "grad_norm": 4.870460033416748, + "learning_rate": 4.906426809807051e-06, + "loss": 0.0039, + "step": 20908 + }, + { + "epoch": 2.2668039895923675, + "grad_norm": 0.2564341723918915, + "learning_rate": 4.905701436239664e-06, + "loss": 0.0013, + "step": 20909 + }, + { + "epoch": 2.2669124024284475, + "grad_norm": 5.17692756652832, + "learning_rate": 4.904976062672277e-06, + "loss": 0.033, + "step": 20910 + }, + { + "epoch": 2.2670208152645275, + "grad_norm": 8.896367073059082, + "learning_rate": 4.90425068910489e-06, + "loss": 0.0223, + "step": 20911 + }, + { + "epoch": 2.267129228100607, + "grad_norm": 0.029098786413669586, + "learning_rate": 4.903525315537503e-06, + "loss": 0.0004, + "step": 20912 + }, + { + "epoch": 2.267237640936687, + "grad_norm": 2.4930949211120605, + "learning_rate": 4.902799941970115e-06, + "loss": 0.0478, + "step": 20913 + }, + { + "epoch": 2.2673460537727665, + "grad_norm": 2.6852025985717773, + "learning_rate": 4.902074568402728e-06, + "loss": 0.025, + "step": 20914 + }, + { + "epoch": 2.2674544666088465, + "grad_norm": 0.3594805896282196, + "learning_rate": 4.90134919483534e-06, + "loss": 0.0017, + "step": 20915 + }, + { + "epoch": 2.2675628794449265, + "grad_norm": 2.051730155944824, + "learning_rate": 4.900623821267954e-06, + "loss": 0.0089, + "step": 20916 + }, + { + "epoch": 2.267671292281006, + "grad_norm": 1.7099462747573853, + "learning_rate": 4.899898447700566e-06, + "loss": 0.0085, + "step": 20917 + }, + { + "epoch": 2.267779705117086, + "grad_norm": 0.06232357770204544, + "learning_rate": 4.899173074133179e-06, + "loss": 0.0004, + "step": 20918 + }, + { + "epoch": 2.2678881179531656, + "grad_norm": 0.8791376352310181, + "learning_rate": 4.898447700565792e-06, + "loss": 0.0028, + "step": 20919 + }, + { + "epoch": 2.2679965307892456, + "grad_norm": 0.07398875057697296, + "learning_rate": 4.897722326998405e-06, + "loss": 0.0006, + "step": 20920 + }, + { + "epoch": 2.268104943625325, + "grad_norm": 5.818215370178223, + "learning_rate": 4.896996953431017e-06, + "loss": 0.019, + "step": 20921 + }, + { + "epoch": 2.268213356461405, + "grad_norm": 1.8553911447525024, + "learning_rate": 4.89627157986363e-06, + "loss": 0.0191, + "step": 20922 + }, + { + "epoch": 2.2683217692974846, + "grad_norm": 0.30739912390708923, + "learning_rate": 4.895546206296243e-06, + "loss": 0.0015, + "step": 20923 + }, + { + "epoch": 2.2684301821335646, + "grad_norm": 0.44555601477622986, + "learning_rate": 4.894820832728856e-06, + "loss": 0.0012, + "step": 20924 + }, + { + "epoch": 2.2685385949696446, + "grad_norm": 16.173734664916992, + "learning_rate": 4.894095459161469e-06, + "loss": 0.0185, + "step": 20925 + }, + { + "epoch": 2.268647007805724, + "grad_norm": 0.30974772572517395, + "learning_rate": 4.893370085594081e-06, + "loss": 0.002, + "step": 20926 + }, + { + "epoch": 2.268755420641804, + "grad_norm": 3.2573583126068115, + "learning_rate": 4.892644712026694e-06, + "loss": 0.0323, + "step": 20927 + }, + { + "epoch": 2.2688638334778837, + "grad_norm": 2.2587037086486816, + "learning_rate": 4.891919338459306e-06, + "loss": 0.0188, + "step": 20928 + }, + { + "epoch": 2.2689722463139637, + "grad_norm": 0.1407390832901001, + "learning_rate": 4.89119396489192e-06, + "loss": 0.0007, + "step": 20929 + }, + { + "epoch": 2.269080659150043, + "grad_norm": 0.04649835452437401, + "learning_rate": 4.890468591324532e-06, + "loss": 0.0003, + "step": 20930 + }, + { + "epoch": 2.269189071986123, + "grad_norm": 3.41033935546875, + "learning_rate": 4.889743217757145e-06, + "loss": 0.0111, + "step": 20931 + }, + { + "epoch": 2.2692974848222027, + "grad_norm": 2.021782636642456, + "learning_rate": 4.889017844189758e-06, + "loss": 0.0228, + "step": 20932 + }, + { + "epoch": 2.2694058976582827, + "grad_norm": 0.1542740762233734, + "learning_rate": 4.888292470622371e-06, + "loss": 0.0005, + "step": 20933 + }, + { + "epoch": 2.2695143104943627, + "grad_norm": 1.541758418083191, + "learning_rate": 4.8875670970549834e-06, + "loss": 0.0014, + "step": 20934 + }, + { + "epoch": 2.2696227233304422, + "grad_norm": 3.2501823902130127, + "learning_rate": 4.886841723487596e-06, + "loss": 0.005, + "step": 20935 + }, + { + "epoch": 2.2697311361665222, + "grad_norm": 0.5553052425384521, + "learning_rate": 4.886116349920209e-06, + "loss": 0.0022, + "step": 20936 + }, + { + "epoch": 2.2698395490026018, + "grad_norm": 0.741507351398468, + "learning_rate": 4.885390976352822e-06, + "loss": 0.0017, + "step": 20937 + }, + { + "epoch": 2.2699479618386817, + "grad_norm": 0.09819087386131287, + "learning_rate": 4.884665602785435e-06, + "loss": 0.0003, + "step": 20938 + }, + { + "epoch": 2.2700563746747617, + "grad_norm": 0.06345465034246445, + "learning_rate": 4.8839402292180475e-06, + "loss": 0.0004, + "step": 20939 + }, + { + "epoch": 2.2701647875108413, + "grad_norm": 8.880714416503906, + "learning_rate": 4.8832148556506605e-06, + "loss": 0.0234, + "step": 20940 + }, + { + "epoch": 2.2702732003469213, + "grad_norm": 2.3472115993499756, + "learning_rate": 4.8824894820832735e-06, + "loss": 0.0141, + "step": 20941 + }, + { + "epoch": 2.270381613183001, + "grad_norm": 1.001044511795044, + "learning_rate": 4.8817641085158865e-06, + "loss": 0.0051, + "step": 20942 + }, + { + "epoch": 2.270490026019081, + "grad_norm": 1.2876155376434326, + "learning_rate": 4.881038734948499e-06, + "loss": 0.0055, + "step": 20943 + }, + { + "epoch": 2.2705984388551603, + "grad_norm": 0.4755541682243347, + "learning_rate": 4.880313361381112e-06, + "loss": 0.0013, + "step": 20944 + }, + { + "epoch": 2.2707068516912403, + "grad_norm": 1.5555777549743652, + "learning_rate": 4.879587987813725e-06, + "loss": 0.0368, + "step": 20945 + }, + { + "epoch": 2.27081526452732, + "grad_norm": 5.215332984924316, + "learning_rate": 4.878862614246338e-06, + "loss": 0.0256, + "step": 20946 + }, + { + "epoch": 2.2709236773634, + "grad_norm": 1.6489942073822021, + "learning_rate": 4.87813724067895e-06, + "loss": 0.027, + "step": 20947 + }, + { + "epoch": 2.27103209019948, + "grad_norm": 3.192476511001587, + "learning_rate": 4.877411867111563e-06, + "loss": 0.0087, + "step": 20948 + }, + { + "epoch": 2.2711405030355594, + "grad_norm": 0.02509422041475773, + "learning_rate": 4.876686493544176e-06, + "loss": 0.0002, + "step": 20949 + }, + { + "epoch": 2.2712489158716394, + "grad_norm": 0.08570989966392517, + "learning_rate": 4.875961119976789e-06, + "loss": 0.0005, + "step": 20950 + }, + { + "epoch": 2.271357328707719, + "grad_norm": 0.4048610329627991, + "learning_rate": 4.875235746409402e-06, + "loss": 0.0012, + "step": 20951 + }, + { + "epoch": 2.271465741543799, + "grad_norm": 0.3956068754196167, + "learning_rate": 4.874510372842014e-06, + "loss": 0.001, + "step": 20952 + }, + { + "epoch": 2.2715741543798784, + "grad_norm": 8.491100311279297, + "learning_rate": 4.873784999274627e-06, + "loss": 0.0533, + "step": 20953 + }, + { + "epoch": 2.2716825672159584, + "grad_norm": 0.6665157079696655, + "learning_rate": 4.87305962570724e-06, + "loss": 0.0023, + "step": 20954 + }, + { + "epoch": 2.271790980052038, + "grad_norm": 2.4007728099823, + "learning_rate": 4.872334252139853e-06, + "loss": 0.0311, + "step": 20955 + }, + { + "epoch": 2.271899392888118, + "grad_norm": 2.0834038257598877, + "learning_rate": 4.871608878572465e-06, + "loss": 0.0105, + "step": 20956 + }, + { + "epoch": 2.272007805724198, + "grad_norm": 6.618618488311768, + "learning_rate": 4.870883505005078e-06, + "loss": 0.0166, + "step": 20957 + }, + { + "epoch": 2.2721162185602775, + "grad_norm": 10.350756645202637, + "learning_rate": 4.870158131437691e-06, + "loss": 0.0212, + "step": 20958 + }, + { + "epoch": 2.2722246313963574, + "grad_norm": 0.3609418570995331, + "learning_rate": 4.869432757870304e-06, + "loss": 0.0023, + "step": 20959 + }, + { + "epoch": 2.272333044232437, + "grad_norm": 1.240822672843933, + "learning_rate": 4.868707384302916e-06, + "loss": 0.0301, + "step": 20960 + }, + { + "epoch": 2.272441457068517, + "grad_norm": 0.023935547098517418, + "learning_rate": 4.867982010735529e-06, + "loss": 0.0001, + "step": 20961 + }, + { + "epoch": 2.272549869904597, + "grad_norm": 0.12282499670982361, + "learning_rate": 4.867256637168142e-06, + "loss": 0.0008, + "step": 20962 + }, + { + "epoch": 2.2726582827406765, + "grad_norm": 0.27368447184562683, + "learning_rate": 4.866531263600755e-06, + "loss": 0.001, + "step": 20963 + }, + { + "epoch": 2.2727666955767565, + "grad_norm": 4.746438980102539, + "learning_rate": 4.865805890033368e-06, + "loss": 0.0368, + "step": 20964 + }, + { + "epoch": 2.272875108412836, + "grad_norm": 3.5121049880981445, + "learning_rate": 4.86508051646598e-06, + "loss": 0.0214, + "step": 20965 + }, + { + "epoch": 2.272983521248916, + "grad_norm": 0.0411398783326149, + "learning_rate": 4.864355142898593e-06, + "loss": 0.0003, + "step": 20966 + }, + { + "epoch": 2.2730919340849955, + "grad_norm": 1.4302583932876587, + "learning_rate": 4.863629769331206e-06, + "loss": 0.0074, + "step": 20967 + }, + { + "epoch": 2.2732003469210755, + "grad_norm": 2.1764581203460693, + "learning_rate": 4.862904395763819e-06, + "loss": 0.0341, + "step": 20968 + }, + { + "epoch": 2.273308759757155, + "grad_norm": 0.5040300488471985, + "learning_rate": 4.862179022196431e-06, + "loss": 0.0007, + "step": 20969 + }, + { + "epoch": 2.273417172593235, + "grad_norm": 1.257056713104248, + "learning_rate": 4.861453648629044e-06, + "loss": 0.0093, + "step": 20970 + }, + { + "epoch": 2.273525585429315, + "grad_norm": 1.4211812019348145, + "learning_rate": 4.860728275061657e-06, + "loss": 0.0027, + "step": 20971 + }, + { + "epoch": 2.2736339982653946, + "grad_norm": 2.2063164710998535, + "learning_rate": 4.86000290149427e-06, + "loss": 0.0055, + "step": 20972 + }, + { + "epoch": 2.2737424111014746, + "grad_norm": 0.5656431317329407, + "learning_rate": 4.8592775279268824e-06, + "loss": 0.0013, + "step": 20973 + }, + { + "epoch": 2.273850823937554, + "grad_norm": 0.22077961266040802, + "learning_rate": 4.8585521543594954e-06, + "loss": 0.0009, + "step": 20974 + }, + { + "epoch": 2.273959236773634, + "grad_norm": 0.767295241355896, + "learning_rate": 4.857826780792108e-06, + "loss": 0.0019, + "step": 20975 + }, + { + "epoch": 2.2740676496097136, + "grad_norm": 1.277205228805542, + "learning_rate": 4.857101407224721e-06, + "loss": 0.0045, + "step": 20976 + }, + { + "epoch": 2.2741760624457936, + "grad_norm": 0.09902076423168182, + "learning_rate": 4.856376033657334e-06, + "loss": 0.0005, + "step": 20977 + }, + { + "epoch": 2.274284475281873, + "grad_norm": 1.201471209526062, + "learning_rate": 4.8556506600899465e-06, + "loss": 0.0024, + "step": 20978 + }, + { + "epoch": 2.274392888117953, + "grad_norm": 4.35164737701416, + "learning_rate": 4.8549252865225595e-06, + "loss": 0.0271, + "step": 20979 + }, + { + "epoch": 2.274501300954033, + "grad_norm": 1.3884248733520508, + "learning_rate": 4.8541999129551725e-06, + "loss": 0.0049, + "step": 20980 + }, + { + "epoch": 2.2746097137901127, + "grad_norm": 3.0268259048461914, + "learning_rate": 4.8534745393877855e-06, + "loss": 0.031, + "step": 20981 + }, + { + "epoch": 2.2747181266261927, + "grad_norm": 0.17595885694026947, + "learning_rate": 4.852749165820398e-06, + "loss": 0.0009, + "step": 20982 + }, + { + "epoch": 2.274826539462272, + "grad_norm": 4.91500997543335, + "learning_rate": 4.852023792253011e-06, + "loss": 0.029, + "step": 20983 + }, + { + "epoch": 2.274934952298352, + "grad_norm": 2.1990251541137695, + "learning_rate": 4.851298418685624e-06, + "loss": 0.0215, + "step": 20984 + }, + { + "epoch": 2.2750433651344317, + "grad_norm": 0.01397232711315155, + "learning_rate": 4.850573045118237e-06, + "loss": 0.0001, + "step": 20985 + }, + { + "epoch": 2.2751517779705117, + "grad_norm": 0.0885591134428978, + "learning_rate": 4.849847671550849e-06, + "loss": 0.0003, + "step": 20986 + }, + { + "epoch": 2.2752601908065913, + "grad_norm": 0.13983044028282166, + "learning_rate": 4.849122297983462e-06, + "loss": 0.0006, + "step": 20987 + }, + { + "epoch": 2.2753686036426712, + "grad_norm": 16.820594787597656, + "learning_rate": 4.848396924416075e-06, + "loss": 0.0395, + "step": 20988 + }, + { + "epoch": 2.2754770164787512, + "grad_norm": 0.19345052540302277, + "learning_rate": 4.847671550848688e-06, + "loss": 0.0018, + "step": 20989 + }, + { + "epoch": 2.2755854293148308, + "grad_norm": 0.033899519592523575, + "learning_rate": 4.846946177281301e-06, + "loss": 0.0002, + "step": 20990 + }, + { + "epoch": 2.2756938421509108, + "grad_norm": 1.0528533458709717, + "learning_rate": 4.846220803713913e-06, + "loss": 0.0033, + "step": 20991 + }, + { + "epoch": 2.2758022549869903, + "grad_norm": 0.45289042592048645, + "learning_rate": 4.845495430146526e-06, + "loss": 0.0032, + "step": 20992 + }, + { + "epoch": 2.2759106678230703, + "grad_norm": 1.960645079612732, + "learning_rate": 4.844770056579139e-06, + "loss": 0.0035, + "step": 20993 + }, + { + "epoch": 2.2760190806591503, + "grad_norm": 2.3928279876708984, + "learning_rate": 4.844044683011752e-06, + "loss": 0.0182, + "step": 20994 + }, + { + "epoch": 2.27612749349523, + "grad_norm": 0.8124071359634399, + "learning_rate": 4.843319309444364e-06, + "loss": 0.0028, + "step": 20995 + }, + { + "epoch": 2.27623590633131, + "grad_norm": 2.487658977508545, + "learning_rate": 4.842593935876977e-06, + "loss": 0.0268, + "step": 20996 + }, + { + "epoch": 2.2763443191673893, + "grad_norm": 19.18356704711914, + "learning_rate": 4.84186856230959e-06, + "loss": 0.0144, + "step": 20997 + }, + { + "epoch": 2.2764527320034693, + "grad_norm": 2.524158000946045, + "learning_rate": 4.841143188742203e-06, + "loss": 0.0215, + "step": 20998 + }, + { + "epoch": 2.276561144839549, + "grad_norm": 7.2973833084106445, + "learning_rate": 4.840417815174815e-06, + "loss": 0.0035, + "step": 20999 + }, + { + "epoch": 2.276669557675629, + "grad_norm": 0.049114421010017395, + "learning_rate": 4.839692441607428e-06, + "loss": 0.0004, + "step": 21000 + }, + { + "epoch": 2.2767779705117084, + "grad_norm": 0.8463658094406128, + "learning_rate": 4.838967068040041e-06, + "loss": 0.0036, + "step": 21001 + }, + { + "epoch": 2.2768863833477884, + "grad_norm": 0.17156867682933807, + "learning_rate": 4.838241694472654e-06, + "loss": 0.0012, + "step": 21002 + }, + { + "epoch": 2.2769947961838684, + "grad_norm": 0.014151417650282383, + "learning_rate": 4.837516320905267e-06, + "loss": 0.0002, + "step": 21003 + }, + { + "epoch": 2.277103209019948, + "grad_norm": 6.091987133026123, + "learning_rate": 4.836790947337879e-06, + "loss": 0.0639, + "step": 21004 + }, + { + "epoch": 2.277211621856028, + "grad_norm": 0.1451556533575058, + "learning_rate": 4.836065573770492e-06, + "loss": 0.0009, + "step": 21005 + }, + { + "epoch": 2.2773200346921074, + "grad_norm": 3.3821287155151367, + "learning_rate": 4.835340200203105e-06, + "loss": 0.0088, + "step": 21006 + }, + { + "epoch": 2.2774284475281874, + "grad_norm": 0.8151775598526001, + "learning_rate": 4.834614826635718e-06, + "loss": 0.0018, + "step": 21007 + }, + { + "epoch": 2.277536860364267, + "grad_norm": 2.29338002204895, + "learning_rate": 4.83388945306833e-06, + "loss": 0.0425, + "step": 21008 + }, + { + "epoch": 2.277645273200347, + "grad_norm": 3.278566360473633, + "learning_rate": 4.833164079500943e-06, + "loss": 0.0287, + "step": 21009 + }, + { + "epoch": 2.2777536860364265, + "grad_norm": 16.472904205322266, + "learning_rate": 4.832438705933556e-06, + "loss": 0.0307, + "step": 21010 + }, + { + "epoch": 2.2778620988725065, + "grad_norm": 2.4214775562286377, + "learning_rate": 4.831713332366169e-06, + "loss": 0.0036, + "step": 21011 + }, + { + "epoch": 2.2779705117085864, + "grad_norm": 5.943185329437256, + "learning_rate": 4.8309879587987814e-06, + "loss": 0.0384, + "step": 21012 + }, + { + "epoch": 2.278078924544666, + "grad_norm": 2.9500973224639893, + "learning_rate": 4.8302625852313944e-06, + "loss": 0.0172, + "step": 21013 + }, + { + "epoch": 2.278187337380746, + "grad_norm": 0.06310834735631943, + "learning_rate": 4.829537211664007e-06, + "loss": 0.0004, + "step": 21014 + }, + { + "epoch": 2.2782957502168255, + "grad_norm": 1.403910517692566, + "learning_rate": 4.82881183809662e-06, + "loss": 0.001, + "step": 21015 + }, + { + "epoch": 2.2784041630529055, + "grad_norm": 4.697492599487305, + "learning_rate": 4.828086464529233e-06, + "loss": 0.0122, + "step": 21016 + }, + { + "epoch": 2.2785125758889855, + "grad_norm": 1.04887056350708, + "learning_rate": 4.8273610909618455e-06, + "loss": 0.0051, + "step": 21017 + }, + { + "epoch": 2.278620988725065, + "grad_norm": 1.6483063697814941, + "learning_rate": 4.8266357173944585e-06, + "loss": 0.0061, + "step": 21018 + }, + { + "epoch": 2.278729401561145, + "grad_norm": 0.22169730067253113, + "learning_rate": 4.8259103438270715e-06, + "loss": 0.0005, + "step": 21019 + }, + { + "epoch": 2.2788378143972245, + "grad_norm": 0.9822989702224731, + "learning_rate": 4.8251849702596845e-06, + "loss": 0.0053, + "step": 21020 + }, + { + "epoch": 2.2789462272333045, + "grad_norm": 0.7184816598892212, + "learning_rate": 4.824459596692297e-06, + "loss": 0.0019, + "step": 21021 + }, + { + "epoch": 2.279054640069384, + "grad_norm": 0.0913492888212204, + "learning_rate": 4.82373422312491e-06, + "loss": 0.0003, + "step": 21022 + }, + { + "epoch": 2.279163052905464, + "grad_norm": 5.607392311096191, + "learning_rate": 4.823008849557523e-06, + "loss": 0.0084, + "step": 21023 + }, + { + "epoch": 2.2792714657415436, + "grad_norm": 2.635077476501465, + "learning_rate": 4.822283475990136e-06, + "loss": 0.0069, + "step": 21024 + }, + { + "epoch": 2.2793798785776236, + "grad_norm": 0.2638075351715088, + "learning_rate": 4.821558102422748e-06, + "loss": 0.0009, + "step": 21025 + }, + { + "epoch": 2.2794882914137036, + "grad_norm": 2.92889404296875, + "learning_rate": 4.820832728855361e-06, + "loss": 0.0294, + "step": 21026 + }, + { + "epoch": 2.279596704249783, + "grad_norm": 4.743237018585205, + "learning_rate": 4.820107355287974e-06, + "loss": 0.0254, + "step": 21027 + }, + { + "epoch": 2.279705117085863, + "grad_norm": 1.2476718425750732, + "learning_rate": 4.819381981720587e-06, + "loss": 0.0284, + "step": 21028 + }, + { + "epoch": 2.2798135299219426, + "grad_norm": 11.52227783203125, + "learning_rate": 4.8186566081532e-06, + "loss": 0.0143, + "step": 21029 + }, + { + "epoch": 2.2799219427580226, + "grad_norm": 0.3086039423942566, + "learning_rate": 4.817931234585812e-06, + "loss": 0.0005, + "step": 21030 + }, + { + "epoch": 2.280030355594102, + "grad_norm": 0.8295626640319824, + "learning_rate": 4.817205861018425e-06, + "loss": 0.002, + "step": 21031 + }, + { + "epoch": 2.280138768430182, + "grad_norm": 0.7291170954704285, + "learning_rate": 4.816480487451038e-06, + "loss": 0.0023, + "step": 21032 + }, + { + "epoch": 2.2802471812662617, + "grad_norm": 1.8370652198791504, + "learning_rate": 4.815755113883651e-06, + "loss": 0.0078, + "step": 21033 + }, + { + "epoch": 2.2803555941023417, + "grad_norm": 0.12477853894233704, + "learning_rate": 4.815029740316263e-06, + "loss": 0.0006, + "step": 21034 + }, + { + "epoch": 2.2804640069384217, + "grad_norm": 7.670587539672852, + "learning_rate": 4.814304366748876e-06, + "loss": 0.0136, + "step": 21035 + }, + { + "epoch": 2.280572419774501, + "grad_norm": 9.07516860961914, + "learning_rate": 4.813578993181489e-06, + "loss": 0.0286, + "step": 21036 + }, + { + "epoch": 2.280680832610581, + "grad_norm": 0.6379855871200562, + "learning_rate": 4.812853619614102e-06, + "loss": 0.004, + "step": 21037 + }, + { + "epoch": 2.2807892454466607, + "grad_norm": 0.31468072533607483, + "learning_rate": 4.812128246046714e-06, + "loss": 0.0007, + "step": 21038 + }, + { + "epoch": 2.2808976582827407, + "grad_norm": 0.2922191321849823, + "learning_rate": 4.811402872479327e-06, + "loss": 0.0006, + "step": 21039 + }, + { + "epoch": 2.2810060711188207, + "grad_norm": 4.444649696350098, + "learning_rate": 4.81067749891194e-06, + "loss": 0.0151, + "step": 21040 + }, + { + "epoch": 2.2811144839549002, + "grad_norm": 0.5671833157539368, + "learning_rate": 4.809952125344553e-06, + "loss": 0.0019, + "step": 21041 + }, + { + "epoch": 2.2812228967909802, + "grad_norm": 32.24856948852539, + "learning_rate": 4.809226751777166e-06, + "loss": 0.0115, + "step": 21042 + }, + { + "epoch": 2.2813313096270598, + "grad_norm": 0.17329268157482147, + "learning_rate": 4.808501378209778e-06, + "loss": 0.0008, + "step": 21043 + }, + { + "epoch": 2.2814397224631398, + "grad_norm": 27.58274269104004, + "learning_rate": 4.807776004642391e-06, + "loss": 0.0226, + "step": 21044 + }, + { + "epoch": 2.2815481352992193, + "grad_norm": 0.6361090540885925, + "learning_rate": 4.807050631075004e-06, + "loss": 0.0028, + "step": 21045 + }, + { + "epoch": 2.2816565481352993, + "grad_norm": 0.08266075700521469, + "learning_rate": 4.806325257507617e-06, + "loss": 0.0007, + "step": 21046 + }, + { + "epoch": 2.281764960971379, + "grad_norm": 5.872646331787109, + "learning_rate": 4.805599883940229e-06, + "loss": 0.0229, + "step": 21047 + }, + { + "epoch": 2.281873373807459, + "grad_norm": 0.9352372288703918, + "learning_rate": 4.804874510372842e-06, + "loss": 0.0013, + "step": 21048 + }, + { + "epoch": 2.281981786643539, + "grad_norm": 0.10018880665302277, + "learning_rate": 4.804149136805455e-06, + "loss": 0.0009, + "step": 21049 + }, + { + "epoch": 2.2820901994796183, + "grad_norm": 0.04592815414071083, + "learning_rate": 4.803423763238068e-06, + "loss": 0.0003, + "step": 21050 + }, + { + "epoch": 2.2821986123156983, + "grad_norm": 5.628114700317383, + "learning_rate": 4.8026983896706804e-06, + "loss": 0.0183, + "step": 21051 + }, + { + "epoch": 2.282307025151778, + "grad_norm": 0.12752369046211243, + "learning_rate": 4.8019730161032934e-06, + "loss": 0.001, + "step": 21052 + }, + { + "epoch": 2.282415437987858, + "grad_norm": 2.789569616317749, + "learning_rate": 4.801247642535906e-06, + "loss": 0.0228, + "step": 21053 + }, + { + "epoch": 2.2825238508239374, + "grad_norm": 0.1501694619655609, + "learning_rate": 4.800522268968519e-06, + "loss": 0.0006, + "step": 21054 + }, + { + "epoch": 2.2826322636600174, + "grad_norm": 3.1311488151550293, + "learning_rate": 4.799796895401132e-06, + "loss": 0.0276, + "step": 21055 + }, + { + "epoch": 2.282740676496097, + "grad_norm": 0.00587061420083046, + "learning_rate": 4.7990715218337445e-06, + "loss": 0.0, + "step": 21056 + }, + { + "epoch": 2.282849089332177, + "grad_norm": 0.3087410628795624, + "learning_rate": 4.7983461482663575e-06, + "loss": 0.0006, + "step": 21057 + }, + { + "epoch": 2.282957502168257, + "grad_norm": 0.2316744029521942, + "learning_rate": 4.7976207746989705e-06, + "loss": 0.0008, + "step": 21058 + }, + { + "epoch": 2.2830659150043364, + "grad_norm": 4.116745471954346, + "learning_rate": 4.7968954011315835e-06, + "loss": 0.0111, + "step": 21059 + }, + { + "epoch": 2.2831743278404164, + "grad_norm": 6.507255554199219, + "learning_rate": 4.796170027564196e-06, + "loss": 0.0078, + "step": 21060 + }, + { + "epoch": 2.283282740676496, + "grad_norm": 2.30507230758667, + "learning_rate": 4.795444653996809e-06, + "loss": 0.0039, + "step": 21061 + }, + { + "epoch": 2.283391153512576, + "grad_norm": 2.8048977851867676, + "learning_rate": 4.794719280429422e-06, + "loss": 0.0299, + "step": 21062 + }, + { + "epoch": 2.283499566348656, + "grad_norm": 0.20725823938846588, + "learning_rate": 4.793993906862035e-06, + "loss": 0.0006, + "step": 21063 + }, + { + "epoch": 2.2836079791847355, + "grad_norm": 0.016825495287775993, + "learning_rate": 4.793268533294647e-06, + "loss": 0.0002, + "step": 21064 + }, + { + "epoch": 2.2837163920208154, + "grad_norm": 1.182849645614624, + "learning_rate": 4.79254315972726e-06, + "loss": 0.0024, + "step": 21065 + }, + { + "epoch": 2.283824804856895, + "grad_norm": 0.033102847635746, + "learning_rate": 4.791817786159873e-06, + "loss": 0.0002, + "step": 21066 + }, + { + "epoch": 2.283933217692975, + "grad_norm": 2.0134880542755127, + "learning_rate": 4.791092412592486e-06, + "loss": 0.0019, + "step": 21067 + }, + { + "epoch": 2.2840416305290545, + "grad_norm": 0.08422952145338058, + "learning_rate": 4.790367039025099e-06, + "loss": 0.0003, + "step": 21068 + }, + { + "epoch": 2.2841500433651345, + "grad_norm": 0.010587132535874844, + "learning_rate": 4.789641665457711e-06, + "loss": 0.0001, + "step": 21069 + }, + { + "epoch": 2.284258456201214, + "grad_norm": 0.01748020201921463, + "learning_rate": 4.788916291890324e-06, + "loss": 0.0001, + "step": 21070 + }, + { + "epoch": 2.284366869037294, + "grad_norm": 0.33298638463020325, + "learning_rate": 4.788190918322937e-06, + "loss": 0.0005, + "step": 21071 + }, + { + "epoch": 2.284475281873374, + "grad_norm": 0.06362254172563553, + "learning_rate": 4.78746554475555e-06, + "loss": 0.0003, + "step": 21072 + }, + { + "epoch": 2.2845836947094535, + "grad_norm": 5.859216213226318, + "learning_rate": 4.786740171188162e-06, + "loss": 0.011, + "step": 21073 + }, + { + "epoch": 2.2846921075455335, + "grad_norm": 3.0605573654174805, + "learning_rate": 4.786014797620775e-06, + "loss": 0.0114, + "step": 21074 + }, + { + "epoch": 2.284800520381613, + "grad_norm": 4.047136306762695, + "learning_rate": 4.785289424053388e-06, + "loss": 0.0523, + "step": 21075 + }, + { + "epoch": 2.284908933217693, + "grad_norm": 13.205047607421875, + "learning_rate": 4.784564050486001e-06, + "loss": 0.0296, + "step": 21076 + }, + { + "epoch": 2.2850173460537726, + "grad_norm": 2.4033939838409424, + "learning_rate": 4.783838676918613e-06, + "loss": 0.0306, + "step": 21077 + }, + { + "epoch": 2.2851257588898526, + "grad_norm": 0.48498108983039856, + "learning_rate": 4.783113303351226e-06, + "loss": 0.0011, + "step": 21078 + }, + { + "epoch": 2.285234171725932, + "grad_norm": 1.156085729598999, + "learning_rate": 4.782387929783839e-06, + "loss": 0.005, + "step": 21079 + }, + { + "epoch": 2.285342584562012, + "grad_norm": 0.11075523495674133, + "learning_rate": 4.781662556216452e-06, + "loss": 0.0006, + "step": 21080 + }, + { + "epoch": 2.285450997398092, + "grad_norm": 2.937028408050537, + "learning_rate": 4.780937182649065e-06, + "loss": 0.0696, + "step": 21081 + }, + { + "epoch": 2.2855594102341716, + "grad_norm": 0.012968792580068111, + "learning_rate": 4.780211809081677e-06, + "loss": 0.0001, + "step": 21082 + }, + { + "epoch": 2.2856678230702516, + "grad_norm": 1.3110289573669434, + "learning_rate": 4.77948643551429e-06, + "loss": 0.0019, + "step": 21083 + }, + { + "epoch": 2.285776235906331, + "grad_norm": 0.004308110103011131, + "learning_rate": 4.778761061946903e-06, + "loss": 0.0, + "step": 21084 + }, + { + "epoch": 2.285884648742411, + "grad_norm": 0.6791173219680786, + "learning_rate": 4.778035688379516e-06, + "loss": 0.0023, + "step": 21085 + }, + { + "epoch": 2.285993061578491, + "grad_norm": 0.07652822881937027, + "learning_rate": 4.777310314812128e-06, + "loss": 0.0004, + "step": 21086 + }, + { + "epoch": 2.2861014744145707, + "grad_norm": 0.06877472251653671, + "learning_rate": 4.776584941244741e-06, + "loss": 0.0003, + "step": 21087 + }, + { + "epoch": 2.2862098872506507, + "grad_norm": 13.270882606506348, + "learning_rate": 4.775859567677354e-06, + "loss": 0.0138, + "step": 21088 + }, + { + "epoch": 2.28631830008673, + "grad_norm": 1.2215009927749634, + "learning_rate": 4.775134194109967e-06, + "loss": 0.0031, + "step": 21089 + }, + { + "epoch": 2.28642671292281, + "grad_norm": 0.16042087972164154, + "learning_rate": 4.7744088205425794e-06, + "loss": 0.0005, + "step": 21090 + }, + { + "epoch": 2.2865351257588897, + "grad_norm": 0.8112508058547974, + "learning_rate": 4.7736834469751924e-06, + "loss": 0.0021, + "step": 21091 + }, + { + "epoch": 2.2866435385949697, + "grad_norm": 16.006729125976562, + "learning_rate": 4.7729580734078054e-06, + "loss": 0.0386, + "step": 21092 + }, + { + "epoch": 2.2867519514310493, + "grad_norm": 11.531128883361816, + "learning_rate": 4.772232699840418e-06, + "loss": 0.0264, + "step": 21093 + }, + { + "epoch": 2.2868603642671292, + "grad_norm": 0.08887697756290436, + "learning_rate": 4.771507326273031e-06, + "loss": 0.0005, + "step": 21094 + }, + { + "epoch": 2.2869687771032092, + "grad_norm": 0.1814378798007965, + "learning_rate": 4.7707819527056435e-06, + "loss": 0.0006, + "step": 21095 + }, + { + "epoch": 2.2870771899392888, + "grad_norm": 0.03133756294846535, + "learning_rate": 4.7700565791382565e-06, + "loss": 0.0004, + "step": 21096 + }, + { + "epoch": 2.2871856027753688, + "grad_norm": 0.09043312817811966, + "learning_rate": 4.7693312055708695e-06, + "loss": 0.0005, + "step": 21097 + }, + { + "epoch": 2.2872940156114483, + "grad_norm": 2.7979981899261475, + "learning_rate": 4.7686058320034825e-06, + "loss": 0.0228, + "step": 21098 + }, + { + "epoch": 2.2874024284475283, + "grad_norm": 0.029973004013299942, + "learning_rate": 4.767880458436095e-06, + "loss": 0.0003, + "step": 21099 + }, + { + "epoch": 2.287510841283608, + "grad_norm": 0.08345706760883331, + "learning_rate": 4.767155084868708e-06, + "loss": 0.0005, + "step": 21100 + }, + { + "epoch": 2.287619254119688, + "grad_norm": 0.3683091104030609, + "learning_rate": 4.766429711301321e-06, + "loss": 0.0011, + "step": 21101 + }, + { + "epoch": 2.2877276669557673, + "grad_norm": 0.6681783199310303, + "learning_rate": 4.765704337733934e-06, + "loss": 0.0026, + "step": 21102 + }, + { + "epoch": 2.2878360797918473, + "grad_norm": 12.749382019042969, + "learning_rate": 4.764978964166546e-06, + "loss": 0.0144, + "step": 21103 + }, + { + "epoch": 2.2879444926279273, + "grad_norm": 1.2224096059799194, + "learning_rate": 4.764253590599159e-06, + "loss": 0.0063, + "step": 21104 + }, + { + "epoch": 2.288052905464007, + "grad_norm": 5.958838939666748, + "learning_rate": 4.763528217031772e-06, + "loss": 0.0246, + "step": 21105 + }, + { + "epoch": 2.288161318300087, + "grad_norm": 5.817648410797119, + "learning_rate": 4.762802843464385e-06, + "loss": 0.0259, + "step": 21106 + }, + { + "epoch": 2.2882697311361664, + "grad_norm": 0.07045411318540573, + "learning_rate": 4.762077469896998e-06, + "loss": 0.0003, + "step": 21107 + }, + { + "epoch": 2.2883781439722464, + "grad_norm": 0.03531984984874725, + "learning_rate": 4.76135209632961e-06, + "loss": 0.0003, + "step": 21108 + }, + { + "epoch": 2.2884865568083264, + "grad_norm": 3.7866721153259277, + "learning_rate": 4.760626722762223e-06, + "loss": 0.0211, + "step": 21109 + }, + { + "epoch": 2.288594969644406, + "grad_norm": 0.510244607925415, + "learning_rate": 4.759901349194836e-06, + "loss": 0.0006, + "step": 21110 + }, + { + "epoch": 2.288703382480486, + "grad_norm": 0.2669422924518585, + "learning_rate": 4.759175975627449e-06, + "loss": 0.0008, + "step": 21111 + }, + { + "epoch": 2.2888117953165654, + "grad_norm": 0.006754592061042786, + "learning_rate": 4.758450602060061e-06, + "loss": 0.0001, + "step": 21112 + }, + { + "epoch": 2.2889202081526454, + "grad_norm": 4.914307594299316, + "learning_rate": 4.757725228492674e-06, + "loss": 0.0143, + "step": 21113 + }, + { + "epoch": 2.289028620988725, + "grad_norm": 3.1885595321655273, + "learning_rate": 4.756999854925287e-06, + "loss": 0.0025, + "step": 21114 + }, + { + "epoch": 2.289137033824805, + "grad_norm": 3.102494955062866, + "learning_rate": 4.7562744813579e-06, + "loss": 0.0089, + "step": 21115 + }, + { + "epoch": 2.2892454466608845, + "grad_norm": 6.615681171417236, + "learning_rate": 4.755549107790512e-06, + "loss": 0.0378, + "step": 21116 + }, + { + "epoch": 2.2893538594969645, + "grad_norm": 4.070464134216309, + "learning_rate": 4.754823734223125e-06, + "loss": 0.0405, + "step": 21117 + }, + { + "epoch": 2.2894622723330444, + "grad_norm": 0.44947493076324463, + "learning_rate": 4.754098360655738e-06, + "loss": 0.0014, + "step": 21118 + }, + { + "epoch": 2.289570685169124, + "grad_norm": 3.8539345264434814, + "learning_rate": 4.753372987088351e-06, + "loss": 0.0308, + "step": 21119 + }, + { + "epoch": 2.289679098005204, + "grad_norm": 0.5244950652122498, + "learning_rate": 4.752647613520964e-06, + "loss": 0.0016, + "step": 21120 + }, + { + "epoch": 2.2897875108412835, + "grad_norm": 3.713754892349243, + "learning_rate": 4.751922239953576e-06, + "loss": 0.0213, + "step": 21121 + }, + { + "epoch": 2.2898959236773635, + "grad_norm": 0.5713268518447876, + "learning_rate": 4.751196866386189e-06, + "loss": 0.0013, + "step": 21122 + }, + { + "epoch": 2.290004336513443, + "grad_norm": 0.4765262305736542, + "learning_rate": 4.750471492818802e-06, + "loss": 0.0026, + "step": 21123 + }, + { + "epoch": 2.290112749349523, + "grad_norm": 4.3940863609313965, + "learning_rate": 4.749746119251415e-06, + "loss": 0.0175, + "step": 21124 + }, + { + "epoch": 2.2902211621856026, + "grad_norm": 0.17830874025821686, + "learning_rate": 4.749020745684027e-06, + "loss": 0.0005, + "step": 21125 + }, + { + "epoch": 2.2903295750216826, + "grad_norm": 3.823650598526001, + "learning_rate": 4.74829537211664e-06, + "loss": 0.0132, + "step": 21126 + }, + { + "epoch": 2.2904379878577625, + "grad_norm": 3.3440136909484863, + "learning_rate": 4.747569998549253e-06, + "loss": 0.0498, + "step": 21127 + }, + { + "epoch": 2.290546400693842, + "grad_norm": 0.015114272013306618, + "learning_rate": 4.746844624981866e-06, + "loss": 0.0001, + "step": 21128 + }, + { + "epoch": 2.290654813529922, + "grad_norm": 1.0727423429489136, + "learning_rate": 4.7461192514144784e-06, + "loss": 0.0316, + "step": 21129 + }, + { + "epoch": 2.2907632263660016, + "grad_norm": 0.03297589719295502, + "learning_rate": 4.7453938778470914e-06, + "loss": 0.0002, + "step": 21130 + }, + { + "epoch": 2.2908716392020816, + "grad_norm": 0.26596835255622864, + "learning_rate": 4.7446685042797044e-06, + "loss": 0.0007, + "step": 21131 + }, + { + "epoch": 2.290980052038161, + "grad_norm": 0.13084548711776733, + "learning_rate": 4.743943130712317e-06, + "loss": 0.0005, + "step": 21132 + }, + { + "epoch": 2.291088464874241, + "grad_norm": 0.8747623562812805, + "learning_rate": 4.74321775714493e-06, + "loss": 0.0032, + "step": 21133 + }, + { + "epoch": 2.2911968777103207, + "grad_norm": 0.009312202222645283, + "learning_rate": 4.7424923835775426e-06, + "loss": 0.0001, + "step": 21134 + }, + { + "epoch": 2.2913052905464006, + "grad_norm": 1.099125623703003, + "learning_rate": 4.7417670100101555e-06, + "loss": 0.0038, + "step": 21135 + }, + { + "epoch": 2.2914137033824806, + "grad_norm": 1.943692684173584, + "learning_rate": 4.7410416364427685e-06, + "loss": 0.0294, + "step": 21136 + }, + { + "epoch": 2.29152211621856, + "grad_norm": 0.6956678032875061, + "learning_rate": 4.7403162628753815e-06, + "loss": 0.0012, + "step": 21137 + }, + { + "epoch": 2.29163052905464, + "grad_norm": 1.0132489204406738, + "learning_rate": 4.739590889307994e-06, + "loss": 0.0052, + "step": 21138 + }, + { + "epoch": 2.2917389418907197, + "grad_norm": 1.4899617433547974, + "learning_rate": 4.738865515740607e-06, + "loss": 0.0082, + "step": 21139 + }, + { + "epoch": 2.2918473547267997, + "grad_norm": 2.3859925270080566, + "learning_rate": 4.73814014217322e-06, + "loss": 0.006, + "step": 21140 + }, + { + "epoch": 2.2919557675628797, + "grad_norm": 3.4218130111694336, + "learning_rate": 4.737414768605833e-06, + "loss": 0.0233, + "step": 21141 + }, + { + "epoch": 2.292064180398959, + "grad_norm": 1.3652491569519043, + "learning_rate": 4.736689395038445e-06, + "loss": 0.0257, + "step": 21142 + }, + { + "epoch": 2.292172593235039, + "grad_norm": 0.03147411718964577, + "learning_rate": 4.735964021471058e-06, + "loss": 0.0002, + "step": 21143 + }, + { + "epoch": 2.2922810060711187, + "grad_norm": 0.17222070693969727, + "learning_rate": 4.735238647903671e-06, + "loss": 0.0007, + "step": 21144 + }, + { + "epoch": 2.2923894189071987, + "grad_norm": 1.176808476448059, + "learning_rate": 4.734513274336284e-06, + "loss": 0.0034, + "step": 21145 + }, + { + "epoch": 2.2924978317432783, + "grad_norm": 1.333185076713562, + "learning_rate": 4.733787900768897e-06, + "loss": 0.0278, + "step": 21146 + }, + { + "epoch": 2.2926062445793582, + "grad_norm": 0.03998640179634094, + "learning_rate": 4.733062527201509e-06, + "loss": 0.0006, + "step": 21147 + }, + { + "epoch": 2.292714657415438, + "grad_norm": 9.603440284729004, + "learning_rate": 4.732337153634122e-06, + "loss": 0.0171, + "step": 21148 + }, + { + "epoch": 2.2928230702515178, + "grad_norm": 2.0200610160827637, + "learning_rate": 4.731611780066735e-06, + "loss": 0.0137, + "step": 21149 + }, + { + "epoch": 2.2929314830875978, + "grad_norm": 0.14406037330627441, + "learning_rate": 4.730886406499348e-06, + "loss": 0.0007, + "step": 21150 + }, + { + "epoch": 2.2930398959236773, + "grad_norm": 0.38778039813041687, + "learning_rate": 4.73016103293196e-06, + "loss": 0.0008, + "step": 21151 + }, + { + "epoch": 2.2931483087597573, + "grad_norm": 0.6752896308898926, + "learning_rate": 4.729435659364573e-06, + "loss": 0.0016, + "step": 21152 + }, + { + "epoch": 2.293256721595837, + "grad_norm": 0.6566492915153503, + "learning_rate": 4.728710285797186e-06, + "loss": 0.0021, + "step": 21153 + }, + { + "epoch": 2.293365134431917, + "grad_norm": 0.32476407289505005, + "learning_rate": 4.727984912229799e-06, + "loss": 0.0017, + "step": 21154 + }, + { + "epoch": 2.2934735472679963, + "grad_norm": 0.2565900683403015, + "learning_rate": 4.727259538662411e-06, + "loss": 0.0024, + "step": 21155 + }, + { + "epoch": 2.2935819601040763, + "grad_norm": 0.019918251782655716, + "learning_rate": 4.726534165095024e-06, + "loss": 0.0002, + "step": 21156 + }, + { + "epoch": 2.293690372940156, + "grad_norm": 4.52682638168335, + "learning_rate": 4.725808791527637e-06, + "loss": 0.0175, + "step": 21157 + }, + { + "epoch": 2.293798785776236, + "grad_norm": 0.12169571965932846, + "learning_rate": 4.72508341796025e-06, + "loss": 0.0012, + "step": 21158 + }, + { + "epoch": 2.293907198612316, + "grad_norm": 5.3243865966796875, + "learning_rate": 4.724358044392863e-06, + "loss": 0.0305, + "step": 21159 + }, + { + "epoch": 2.2940156114483954, + "grad_norm": 0.42470666766166687, + "learning_rate": 4.723632670825475e-06, + "loss": 0.0005, + "step": 21160 + }, + { + "epoch": 2.2941240242844754, + "grad_norm": 2.2726573944091797, + "learning_rate": 4.722907297258088e-06, + "loss": 0.0402, + "step": 21161 + }, + { + "epoch": 2.294232437120555, + "grad_norm": 0.42088863253593445, + "learning_rate": 4.722181923690701e-06, + "loss": 0.0008, + "step": 21162 + }, + { + "epoch": 2.294340849956635, + "grad_norm": 0.028576698154211044, + "learning_rate": 4.721456550123314e-06, + "loss": 0.0003, + "step": 21163 + }, + { + "epoch": 2.294449262792715, + "grad_norm": 0.019141661003232002, + "learning_rate": 4.720731176555926e-06, + "loss": 0.0001, + "step": 21164 + }, + { + "epoch": 2.2945576756287944, + "grad_norm": 0.07281162589788437, + "learning_rate": 4.720005802988539e-06, + "loss": 0.0003, + "step": 21165 + }, + { + "epoch": 2.2946660884648744, + "grad_norm": 1.631447672843933, + "learning_rate": 4.719280429421152e-06, + "loss": 0.0008, + "step": 21166 + }, + { + "epoch": 2.294774501300954, + "grad_norm": 0.0819452777504921, + "learning_rate": 4.718555055853765e-06, + "loss": 0.0004, + "step": 21167 + }, + { + "epoch": 2.294882914137034, + "grad_norm": 0.5332508683204651, + "learning_rate": 4.7178296822863775e-06, + "loss": 0.0011, + "step": 21168 + }, + { + "epoch": 2.2949913269731135, + "grad_norm": 0.7070731520652771, + "learning_rate": 4.7171043087189904e-06, + "loss": 0.001, + "step": 21169 + }, + { + "epoch": 2.2950997398091935, + "grad_norm": 0.16055242717266083, + "learning_rate": 4.7163789351516034e-06, + "loss": 0.0004, + "step": 21170 + }, + { + "epoch": 2.295208152645273, + "grad_norm": 0.17841459810733795, + "learning_rate": 4.715653561584216e-06, + "loss": 0.0007, + "step": 21171 + }, + { + "epoch": 2.295316565481353, + "grad_norm": 0.2337292730808258, + "learning_rate": 4.714928188016829e-06, + "loss": 0.0009, + "step": 21172 + }, + { + "epoch": 2.295424978317433, + "grad_norm": 0.07820835709571838, + "learning_rate": 4.7142028144494416e-06, + "loss": 0.0007, + "step": 21173 + }, + { + "epoch": 2.2955333911535125, + "grad_norm": 0.9678328037261963, + "learning_rate": 4.7134774408820545e-06, + "loss": 0.0066, + "step": 21174 + }, + { + "epoch": 2.2956418039895925, + "grad_norm": 4.452573299407959, + "learning_rate": 4.7127520673146675e-06, + "loss": 0.0161, + "step": 21175 + }, + { + "epoch": 2.295750216825672, + "grad_norm": 4.811293601989746, + "learning_rate": 4.7120266937472805e-06, + "loss": 0.0307, + "step": 21176 + }, + { + "epoch": 2.295858629661752, + "grad_norm": 0.06731881946325302, + "learning_rate": 4.711301320179893e-06, + "loss": 0.0004, + "step": 21177 + }, + { + "epoch": 2.2959670424978316, + "grad_norm": 1.1999119520187378, + "learning_rate": 4.710575946612506e-06, + "loss": 0.003, + "step": 21178 + }, + { + "epoch": 2.2960754553339116, + "grad_norm": 7.971529006958008, + "learning_rate": 4.709850573045119e-06, + "loss": 0.0296, + "step": 21179 + }, + { + "epoch": 2.296183868169991, + "grad_norm": 0.1031108871102333, + "learning_rate": 4.709125199477732e-06, + "loss": 0.0006, + "step": 21180 + }, + { + "epoch": 2.296292281006071, + "grad_norm": 0.4695117771625519, + "learning_rate": 4.708399825910344e-06, + "loss": 0.0014, + "step": 21181 + }, + { + "epoch": 2.296400693842151, + "grad_norm": 3.6294033527374268, + "learning_rate": 4.707674452342957e-06, + "loss": 0.0101, + "step": 21182 + }, + { + "epoch": 2.2965091066782306, + "grad_norm": 3.804370641708374, + "learning_rate": 4.70694907877557e-06, + "loss": 0.0047, + "step": 21183 + }, + { + "epoch": 2.2966175195143106, + "grad_norm": 0.008103230968117714, + "learning_rate": 4.706223705208183e-06, + "loss": 0.0001, + "step": 21184 + }, + { + "epoch": 2.29672593235039, + "grad_norm": 3.3110079765319824, + "learning_rate": 4.705498331640796e-06, + "loss": 0.0052, + "step": 21185 + }, + { + "epoch": 2.29683434518647, + "grad_norm": 0.0797216072678566, + "learning_rate": 4.704772958073408e-06, + "loss": 0.0004, + "step": 21186 + }, + { + "epoch": 2.29694275802255, + "grad_norm": 0.052340853959321976, + "learning_rate": 4.704047584506021e-06, + "loss": 0.0008, + "step": 21187 + }, + { + "epoch": 2.2970511708586296, + "grad_norm": 0.2892260253429413, + "learning_rate": 4.703322210938634e-06, + "loss": 0.0022, + "step": 21188 + }, + { + "epoch": 2.2971595836947096, + "grad_norm": 0.09271744638681412, + "learning_rate": 4.702596837371247e-06, + "loss": 0.0006, + "step": 21189 + }, + { + "epoch": 2.297267996530789, + "grad_norm": 0.2668484151363373, + "learning_rate": 4.701871463803859e-06, + "loss": 0.0002, + "step": 21190 + }, + { + "epoch": 2.297376409366869, + "grad_norm": 0.37709110975265503, + "learning_rate": 4.701146090236472e-06, + "loss": 0.0019, + "step": 21191 + }, + { + "epoch": 2.2974848222029487, + "grad_norm": 0.04369591921567917, + "learning_rate": 4.700420716669085e-06, + "loss": 0.0002, + "step": 21192 + }, + { + "epoch": 2.2975932350390287, + "grad_norm": 0.2449491024017334, + "learning_rate": 4.699695343101698e-06, + "loss": 0.0013, + "step": 21193 + }, + { + "epoch": 2.297701647875108, + "grad_norm": 0.17014755308628082, + "learning_rate": 4.69896996953431e-06, + "loss": 0.0006, + "step": 21194 + }, + { + "epoch": 2.297810060711188, + "grad_norm": 0.14253653585910797, + "learning_rate": 4.698244595966923e-06, + "loss": 0.0003, + "step": 21195 + }, + { + "epoch": 2.297918473547268, + "grad_norm": 1.3524037599563599, + "learning_rate": 4.697519222399536e-06, + "loss": 0.0217, + "step": 21196 + }, + { + "epoch": 2.2980268863833477, + "grad_norm": 0.07442425936460495, + "learning_rate": 4.696793848832149e-06, + "loss": 0.0003, + "step": 21197 + }, + { + "epoch": 2.2981352992194277, + "grad_norm": 1.6445189714431763, + "learning_rate": 4.696068475264762e-06, + "loss": 0.0036, + "step": 21198 + }, + { + "epoch": 2.2982437120555073, + "grad_norm": 0.0785488709807396, + "learning_rate": 4.695343101697374e-06, + "loss": 0.0005, + "step": 21199 + }, + { + "epoch": 2.2983521248915872, + "grad_norm": 2.680081605911255, + "learning_rate": 4.694617728129987e-06, + "loss": 0.042, + "step": 21200 + }, + { + "epoch": 2.298460537727667, + "grad_norm": 2.960249185562134, + "learning_rate": 4.6938923545626e-06, + "loss": 0.0035, + "step": 21201 + }, + { + "epoch": 2.2985689505637468, + "grad_norm": 0.046799272298812866, + "learning_rate": 4.693166980995213e-06, + "loss": 0.0003, + "step": 21202 + }, + { + "epoch": 2.2986773633998263, + "grad_norm": 0.061705779284238815, + "learning_rate": 4.692441607427825e-06, + "loss": 0.0005, + "step": 21203 + }, + { + "epoch": 2.2987857762359063, + "grad_norm": 8.237283706665039, + "learning_rate": 4.691716233860438e-06, + "loss": 0.0545, + "step": 21204 + }, + { + "epoch": 2.2988941890719863, + "grad_norm": 0.16198647022247314, + "learning_rate": 4.690990860293051e-06, + "loss": 0.0004, + "step": 21205 + }, + { + "epoch": 2.299002601908066, + "grad_norm": 0.16901573538780212, + "learning_rate": 4.690265486725664e-06, + "loss": 0.0014, + "step": 21206 + }, + { + "epoch": 2.299111014744146, + "grad_norm": 4.483229160308838, + "learning_rate": 4.6895401131582765e-06, + "loss": 0.0331, + "step": 21207 + }, + { + "epoch": 2.2992194275802254, + "grad_norm": 0.2316264659166336, + "learning_rate": 4.6888147395908894e-06, + "loss": 0.0007, + "step": 21208 + }, + { + "epoch": 2.2993278404163053, + "grad_norm": 11.574478149414062, + "learning_rate": 4.6880893660235024e-06, + "loss": 0.016, + "step": 21209 + }, + { + "epoch": 2.2994362532523853, + "grad_norm": 0.9394055604934692, + "learning_rate": 4.6873639924561154e-06, + "loss": 0.0009, + "step": 21210 + }, + { + "epoch": 2.299544666088465, + "grad_norm": 0.9468189477920532, + "learning_rate": 4.686638618888728e-06, + "loss": 0.0025, + "step": 21211 + }, + { + "epoch": 2.299653078924545, + "grad_norm": 0.04269174113869667, + "learning_rate": 4.6859132453213406e-06, + "loss": 0.0001, + "step": 21212 + }, + { + "epoch": 2.2997614917606244, + "grad_norm": 0.015656083822250366, + "learning_rate": 4.6851878717539535e-06, + "loss": 0.0002, + "step": 21213 + }, + { + "epoch": 2.2998699045967044, + "grad_norm": 2.1453030109405518, + "learning_rate": 4.6844624981865665e-06, + "loss": 0.0058, + "step": 21214 + }, + { + "epoch": 2.299978317432784, + "grad_norm": 0.024613672867417336, + "learning_rate": 4.6837371246191795e-06, + "loss": 0.0003, + "step": 21215 + }, + { + "epoch": 2.300086730268864, + "grad_norm": 0.46171438694000244, + "learning_rate": 4.683011751051792e-06, + "loss": 0.0019, + "step": 21216 + }, + { + "epoch": 2.3001951431049434, + "grad_norm": 0.7068085074424744, + "learning_rate": 4.682286377484405e-06, + "loss": 0.0011, + "step": 21217 + }, + { + "epoch": 2.3003035559410234, + "grad_norm": 0.3273686468601227, + "learning_rate": 4.681561003917018e-06, + "loss": 0.001, + "step": 21218 + }, + { + "epoch": 2.3004119687771034, + "grad_norm": 3.081332206726074, + "learning_rate": 4.680835630349631e-06, + "loss": 0.0032, + "step": 21219 + }, + { + "epoch": 2.300520381613183, + "grad_norm": 2.472018003463745, + "learning_rate": 4.680110256782243e-06, + "loss": 0.0219, + "step": 21220 + }, + { + "epoch": 2.300628794449263, + "grad_norm": 0.09666608273983002, + "learning_rate": 4.679384883214856e-06, + "loss": 0.0004, + "step": 21221 + }, + { + "epoch": 2.3007372072853425, + "grad_norm": 1.2701319456100464, + "learning_rate": 4.678659509647469e-06, + "loss": 0.0262, + "step": 21222 + }, + { + "epoch": 2.3008456201214225, + "grad_norm": 0.26642030477523804, + "learning_rate": 4.677934136080082e-06, + "loss": 0.0016, + "step": 21223 + }, + { + "epoch": 2.300954032957502, + "grad_norm": 2.227302074432373, + "learning_rate": 4.677208762512695e-06, + "loss": 0.0056, + "step": 21224 + }, + { + "epoch": 2.301062445793582, + "grad_norm": 0.38615503907203674, + "learning_rate": 4.676483388945307e-06, + "loss": 0.001, + "step": 21225 + }, + { + "epoch": 2.3011708586296615, + "grad_norm": 2.798574686050415, + "learning_rate": 4.67575801537792e-06, + "loss": 0.011, + "step": 21226 + }, + { + "epoch": 2.3012792714657415, + "grad_norm": 0.09285274147987366, + "learning_rate": 4.675032641810533e-06, + "loss": 0.0008, + "step": 21227 + }, + { + "epoch": 2.3013876843018215, + "grad_norm": 0.03415360674262047, + "learning_rate": 4.674307268243146e-06, + "loss": 0.0002, + "step": 21228 + }, + { + "epoch": 2.301496097137901, + "grad_norm": 2.968369245529175, + "learning_rate": 4.673581894675758e-06, + "loss": 0.0731, + "step": 21229 + }, + { + "epoch": 2.301604509973981, + "grad_norm": 0.012586239725351334, + "learning_rate": 4.672856521108371e-06, + "loss": 0.0001, + "step": 21230 + }, + { + "epoch": 2.3017129228100606, + "grad_norm": 0.2783941328525543, + "learning_rate": 4.672131147540984e-06, + "loss": 0.0007, + "step": 21231 + }, + { + "epoch": 2.3018213356461406, + "grad_norm": 0.36743617057800293, + "learning_rate": 4.671405773973597e-06, + "loss": 0.0013, + "step": 21232 + }, + { + "epoch": 2.3019297484822205, + "grad_norm": 1.2810308933258057, + "learning_rate": 4.670680400406209e-06, + "loss": 0.0086, + "step": 21233 + }, + { + "epoch": 2.3020381613183, + "grad_norm": 0.04892832413315773, + "learning_rate": 4.669955026838822e-06, + "loss": 0.0003, + "step": 21234 + }, + { + "epoch": 2.30214657415438, + "grad_norm": 1.0775283575057983, + "learning_rate": 4.669229653271435e-06, + "loss": 0.0013, + "step": 21235 + }, + { + "epoch": 2.3022549869904596, + "grad_norm": 4.5456743240356445, + "learning_rate": 4.668504279704048e-06, + "loss": 0.0194, + "step": 21236 + }, + { + "epoch": 2.3023633998265396, + "grad_norm": 0.05900489166378975, + "learning_rate": 4.667778906136661e-06, + "loss": 0.0003, + "step": 21237 + }, + { + "epoch": 2.302471812662619, + "grad_norm": 5.819469451904297, + "learning_rate": 4.667053532569273e-06, + "loss": 0.0345, + "step": 21238 + }, + { + "epoch": 2.302580225498699, + "grad_norm": 0.8797693848609924, + "learning_rate": 4.666328159001886e-06, + "loss": 0.0027, + "step": 21239 + }, + { + "epoch": 2.3026886383347787, + "grad_norm": 3.8380045890808105, + "learning_rate": 4.665602785434499e-06, + "loss": 0.0084, + "step": 21240 + }, + { + "epoch": 2.3027970511708586, + "grad_norm": 0.9103559255599976, + "learning_rate": 4.664877411867112e-06, + "loss": 0.0027, + "step": 21241 + }, + { + "epoch": 2.3029054640069386, + "grad_norm": 0.41400012373924255, + "learning_rate": 4.664152038299724e-06, + "loss": 0.001, + "step": 21242 + }, + { + "epoch": 2.303013876843018, + "grad_norm": 1.5547815561294556, + "learning_rate": 4.663426664732337e-06, + "loss": 0.0033, + "step": 21243 + }, + { + "epoch": 2.303122289679098, + "grad_norm": 2.942837953567505, + "learning_rate": 4.66270129116495e-06, + "loss": 0.0252, + "step": 21244 + }, + { + "epoch": 2.3032307025151777, + "grad_norm": 0.024899156764149666, + "learning_rate": 4.661975917597563e-06, + "loss": 0.0003, + "step": 21245 + }, + { + "epoch": 2.3033391153512577, + "grad_norm": 0.07179410010576248, + "learning_rate": 4.6612505440301755e-06, + "loss": 0.0005, + "step": 21246 + }, + { + "epoch": 2.3034475281873372, + "grad_norm": 3.1286842823028564, + "learning_rate": 4.6605251704627884e-06, + "loss": 0.0583, + "step": 21247 + }, + { + "epoch": 2.303555941023417, + "grad_norm": 0.00268018851056695, + "learning_rate": 4.6597997968954014e-06, + "loss": 0.0, + "step": 21248 + }, + { + "epoch": 2.3036643538594967, + "grad_norm": 2.111201286315918, + "learning_rate": 4.6590744233280144e-06, + "loss": 0.0075, + "step": 21249 + }, + { + "epoch": 2.3037727666955767, + "grad_norm": 1.2531858682632446, + "learning_rate": 4.658349049760627e-06, + "loss": 0.0043, + "step": 21250 + }, + { + "epoch": 2.3038811795316567, + "grad_norm": 3.616950273513794, + "learning_rate": 4.6576236761932396e-06, + "loss": 0.014, + "step": 21251 + }, + { + "epoch": 2.3039895923677363, + "grad_norm": 4.5561041831970215, + "learning_rate": 4.6568983026258526e-06, + "loss": 0.0121, + "step": 21252 + }, + { + "epoch": 2.3040980052038162, + "grad_norm": 0.10775182396173477, + "learning_rate": 4.6561729290584655e-06, + "loss": 0.0008, + "step": 21253 + }, + { + "epoch": 2.304206418039896, + "grad_norm": 4.1816182136535645, + "learning_rate": 4.6554475554910785e-06, + "loss": 0.0564, + "step": 21254 + }, + { + "epoch": 2.3043148308759758, + "grad_norm": 7.957067489624023, + "learning_rate": 4.654722181923691e-06, + "loss": 0.0218, + "step": 21255 + }, + { + "epoch": 2.3044232437120553, + "grad_norm": 0.03748658299446106, + "learning_rate": 4.653996808356304e-06, + "loss": 0.0003, + "step": 21256 + }, + { + "epoch": 2.3045316565481353, + "grad_norm": 6.2580037117004395, + "learning_rate": 4.653271434788917e-06, + "loss": 0.0044, + "step": 21257 + }, + { + "epoch": 2.3046400693842153, + "grad_norm": 0.010126656852662563, + "learning_rate": 4.65254606122153e-06, + "loss": 0.0001, + "step": 21258 + }, + { + "epoch": 2.304748482220295, + "grad_norm": 0.09077673405408859, + "learning_rate": 4.651820687654142e-06, + "loss": 0.0005, + "step": 21259 + }, + { + "epoch": 2.304856895056375, + "grad_norm": 0.03420598804950714, + "learning_rate": 4.651095314086755e-06, + "loss": 0.0003, + "step": 21260 + }, + { + "epoch": 2.3049653078924544, + "grad_norm": 0.35073208808898926, + "learning_rate": 4.650369940519368e-06, + "loss": 0.0021, + "step": 21261 + }, + { + "epoch": 2.3050737207285343, + "grad_norm": 1.8488569259643555, + "learning_rate": 4.649644566951981e-06, + "loss": 0.0128, + "step": 21262 + }, + { + "epoch": 2.305182133564614, + "grad_norm": 3.2578165531158447, + "learning_rate": 4.648919193384594e-06, + "loss": 0.037, + "step": 21263 + }, + { + "epoch": 2.305290546400694, + "grad_norm": 0.06756089627742767, + "learning_rate": 4.648193819817206e-06, + "loss": 0.0002, + "step": 21264 + }, + { + "epoch": 2.305398959236774, + "grad_norm": 3.8690199851989746, + "learning_rate": 4.647468446249819e-06, + "loss": 0.0016, + "step": 21265 + }, + { + "epoch": 2.3055073720728534, + "grad_norm": 0.7638605833053589, + "learning_rate": 4.646743072682432e-06, + "loss": 0.0007, + "step": 21266 + }, + { + "epoch": 2.3056157849089334, + "grad_norm": 0.49147313833236694, + "learning_rate": 4.646017699115045e-06, + "loss": 0.0007, + "step": 21267 + }, + { + "epoch": 2.305724197745013, + "grad_norm": 0.15242113173007965, + "learning_rate": 4.645292325547657e-06, + "loss": 0.0006, + "step": 21268 + }, + { + "epoch": 2.305832610581093, + "grad_norm": 0.38286465406417847, + "learning_rate": 4.64456695198027e-06, + "loss": 0.0022, + "step": 21269 + }, + { + "epoch": 2.3059410234171724, + "grad_norm": 0.03898223862051964, + "learning_rate": 4.643841578412883e-06, + "loss": 0.0004, + "step": 21270 + }, + { + "epoch": 2.3060494362532524, + "grad_norm": 0.20034867525100708, + "learning_rate": 4.643116204845496e-06, + "loss": 0.0008, + "step": 21271 + }, + { + "epoch": 2.306157849089332, + "grad_norm": 1.347429633140564, + "learning_rate": 4.642390831278108e-06, + "loss": 0.0025, + "step": 21272 + }, + { + "epoch": 2.306266261925412, + "grad_norm": 4.759328842163086, + "learning_rate": 4.641665457710721e-06, + "loss": 0.0186, + "step": 21273 + }, + { + "epoch": 2.306374674761492, + "grad_norm": 0.497440904378891, + "learning_rate": 4.640940084143334e-06, + "loss": 0.0013, + "step": 21274 + }, + { + "epoch": 2.3064830875975715, + "grad_norm": 0.8819469213485718, + "learning_rate": 4.640214710575947e-06, + "loss": 0.0031, + "step": 21275 + }, + { + "epoch": 2.3065915004336515, + "grad_norm": 2.089996337890625, + "learning_rate": 4.63948933700856e-06, + "loss": 0.0173, + "step": 21276 + }, + { + "epoch": 2.306699913269731, + "grad_norm": 1.9485770463943481, + "learning_rate": 4.638763963441172e-06, + "loss": 0.0293, + "step": 21277 + }, + { + "epoch": 2.306808326105811, + "grad_norm": 0.044885922223329544, + "learning_rate": 4.638038589873785e-06, + "loss": 0.0004, + "step": 21278 + }, + { + "epoch": 2.3069167389418905, + "grad_norm": 0.13869239389896393, + "learning_rate": 4.637313216306398e-06, + "loss": 0.0002, + "step": 21279 + }, + { + "epoch": 2.3070251517779705, + "grad_norm": 0.012517066672444344, + "learning_rate": 4.636587842739011e-06, + "loss": 0.0001, + "step": 21280 + }, + { + "epoch": 2.30713356461405, + "grad_norm": 1.2867192029953003, + "learning_rate": 4.635862469171623e-06, + "loss": 0.0013, + "step": 21281 + }, + { + "epoch": 2.30724197745013, + "grad_norm": 6.504703521728516, + "learning_rate": 4.635137095604236e-06, + "loss": 0.0436, + "step": 21282 + }, + { + "epoch": 2.30735039028621, + "grad_norm": 0.017061125487089157, + "learning_rate": 4.634411722036849e-06, + "loss": 0.0002, + "step": 21283 + }, + { + "epoch": 2.3074588031222896, + "grad_norm": 0.12237044423818588, + "learning_rate": 4.633686348469462e-06, + "loss": 0.0004, + "step": 21284 + }, + { + "epoch": 2.3075672159583696, + "grad_norm": 0.2992326319217682, + "learning_rate": 4.6329609749020745e-06, + "loss": 0.0014, + "step": 21285 + }, + { + "epoch": 2.307675628794449, + "grad_norm": 0.010507576167583466, + "learning_rate": 4.6322356013346875e-06, + "loss": 0.0, + "step": 21286 + }, + { + "epoch": 2.307784041630529, + "grad_norm": 0.44788357615470886, + "learning_rate": 4.6315102277673004e-06, + "loss": 0.0022, + "step": 21287 + }, + { + "epoch": 2.307892454466609, + "grad_norm": 1.21443510055542, + "learning_rate": 4.6307848541999134e-06, + "loss": 0.002, + "step": 21288 + }, + { + "epoch": 2.3080008673026886, + "grad_norm": 7.974324703216553, + "learning_rate": 4.630059480632526e-06, + "loss": 0.0086, + "step": 21289 + }, + { + "epoch": 2.3081092801387686, + "grad_norm": 3.140104293823242, + "learning_rate": 4.6293341070651386e-06, + "loss": 0.0467, + "step": 21290 + }, + { + "epoch": 2.308217692974848, + "grad_norm": 7.663329601287842, + "learning_rate": 4.6286087334977516e-06, + "loss": 0.0349, + "step": 21291 + }, + { + "epoch": 2.308326105810928, + "grad_norm": 3.241760730743408, + "learning_rate": 4.6278833599303645e-06, + "loss": 0.0113, + "step": 21292 + }, + { + "epoch": 2.3084345186470077, + "grad_norm": 4.082880973815918, + "learning_rate": 4.6271579863629775e-06, + "loss": 0.0091, + "step": 21293 + }, + { + "epoch": 2.3085429314830876, + "grad_norm": 0.7883156538009644, + "learning_rate": 4.62643261279559e-06, + "loss": 0.001, + "step": 21294 + }, + { + "epoch": 2.308651344319167, + "grad_norm": 0.6406514644622803, + "learning_rate": 4.625707239228203e-06, + "loss": 0.0014, + "step": 21295 + }, + { + "epoch": 2.308759757155247, + "grad_norm": 0.12501133978366852, + "learning_rate": 4.624981865660816e-06, + "loss": 0.0008, + "step": 21296 + }, + { + "epoch": 2.308868169991327, + "grad_norm": 3.416036367416382, + "learning_rate": 4.624256492093429e-06, + "loss": 0.0036, + "step": 21297 + }, + { + "epoch": 2.3089765828274067, + "grad_norm": 0.2979544401168823, + "learning_rate": 4.623531118526041e-06, + "loss": 0.0008, + "step": 21298 + }, + { + "epoch": 2.3090849956634867, + "grad_norm": 0.011725085787475109, + "learning_rate": 4.622805744958654e-06, + "loss": 0.0001, + "step": 21299 + }, + { + "epoch": 2.3091934084995662, + "grad_norm": 0.04277755692601204, + "learning_rate": 4.622080371391267e-06, + "loss": 0.0003, + "step": 21300 + }, + { + "epoch": 2.309301821335646, + "grad_norm": 1.8805383443832397, + "learning_rate": 4.62135499782388e-06, + "loss": 0.0073, + "step": 21301 + }, + { + "epoch": 2.3094102341717258, + "grad_norm": 0.8017145991325378, + "learning_rate": 4.620629624256493e-06, + "loss": 0.0015, + "step": 21302 + }, + { + "epoch": 2.3095186470078057, + "grad_norm": 0.20188534259796143, + "learning_rate": 4.619904250689105e-06, + "loss": 0.001, + "step": 21303 + }, + { + "epoch": 2.3096270598438853, + "grad_norm": 10.994192123413086, + "learning_rate": 4.619178877121718e-06, + "loss": 0.0228, + "step": 21304 + }, + { + "epoch": 2.3097354726799653, + "grad_norm": 0.06301537156105042, + "learning_rate": 4.618453503554331e-06, + "loss": 0.0006, + "step": 21305 + }, + { + "epoch": 2.3098438855160452, + "grad_norm": 0.02248728647828102, + "learning_rate": 4.617728129986944e-06, + "loss": 0.0, + "step": 21306 + }, + { + "epoch": 2.309952298352125, + "grad_norm": 0.7208925485610962, + "learning_rate": 4.617002756419556e-06, + "loss": 0.004, + "step": 21307 + }, + { + "epoch": 2.3100607111882048, + "grad_norm": 0.13855016231536865, + "learning_rate": 4.616277382852169e-06, + "loss": 0.0006, + "step": 21308 + }, + { + "epoch": 2.3101691240242843, + "grad_norm": 1.7019844055175781, + "learning_rate": 4.615552009284782e-06, + "loss": 0.0114, + "step": 21309 + }, + { + "epoch": 2.3102775368603643, + "grad_norm": 3.141496419906616, + "learning_rate": 4.614826635717395e-06, + "loss": 0.0335, + "step": 21310 + }, + { + "epoch": 2.3103859496964443, + "grad_norm": 3.676319122314453, + "learning_rate": 4.614101262150007e-06, + "loss": 0.0208, + "step": 21311 + }, + { + "epoch": 2.310494362532524, + "grad_norm": 1.9166043996810913, + "learning_rate": 4.61337588858262e-06, + "loss": 0.0038, + "step": 21312 + }, + { + "epoch": 2.310602775368604, + "grad_norm": 3.166504383087158, + "learning_rate": 4.612650515015233e-06, + "loss": 0.0126, + "step": 21313 + }, + { + "epoch": 2.3107111882046834, + "grad_norm": 0.1728903353214264, + "learning_rate": 4.611925141447846e-06, + "loss": 0.0011, + "step": 21314 + }, + { + "epoch": 2.3108196010407633, + "grad_norm": 0.03802941367030144, + "learning_rate": 4.611199767880459e-06, + "loss": 0.0002, + "step": 21315 + }, + { + "epoch": 2.310928013876843, + "grad_norm": 3.8479645252227783, + "learning_rate": 4.610474394313071e-06, + "loss": 0.0048, + "step": 21316 + }, + { + "epoch": 2.311036426712923, + "grad_norm": 1.8916237354278564, + "learning_rate": 4.609749020745684e-06, + "loss": 0.0106, + "step": 21317 + }, + { + "epoch": 2.3111448395490024, + "grad_norm": 2.118917226791382, + "learning_rate": 4.609023647178297e-06, + "loss": 0.0071, + "step": 21318 + }, + { + "epoch": 2.3112532523850824, + "grad_norm": 1.8674544095993042, + "learning_rate": 4.60829827361091e-06, + "loss": 0.0249, + "step": 21319 + }, + { + "epoch": 2.3113616652211624, + "grad_norm": 5.316089153289795, + "learning_rate": 4.607572900043522e-06, + "loss": 0.0178, + "step": 21320 + }, + { + "epoch": 2.311470078057242, + "grad_norm": 2.9504830837249756, + "learning_rate": 4.606847526476135e-06, + "loss": 0.0149, + "step": 21321 + }, + { + "epoch": 2.311578490893322, + "grad_norm": 0.32519546151161194, + "learning_rate": 4.606122152908748e-06, + "loss": 0.0006, + "step": 21322 + }, + { + "epoch": 2.3116869037294014, + "grad_norm": 0.9107850193977356, + "learning_rate": 4.605396779341361e-06, + "loss": 0.0008, + "step": 21323 + }, + { + "epoch": 2.3117953165654814, + "grad_norm": 0.012503836303949356, + "learning_rate": 4.6046714057739735e-06, + "loss": 0.0001, + "step": 21324 + }, + { + "epoch": 2.311903729401561, + "grad_norm": 0.00489064771682024, + "learning_rate": 4.6039460322065865e-06, + "loss": 0.0, + "step": 21325 + }, + { + "epoch": 2.312012142237641, + "grad_norm": 0.023658230900764465, + "learning_rate": 4.6032206586391994e-06, + "loss": 0.0002, + "step": 21326 + }, + { + "epoch": 2.3121205550737205, + "grad_norm": 0.10133396089076996, + "learning_rate": 4.6024952850718124e-06, + "loss": 0.0008, + "step": 21327 + }, + { + "epoch": 2.3122289679098005, + "grad_norm": 2.116589069366455, + "learning_rate": 4.6017699115044254e-06, + "loss": 0.0093, + "step": 21328 + }, + { + "epoch": 2.3123373807458805, + "grad_norm": 0.05013439431786537, + "learning_rate": 4.6010445379370376e-06, + "loss": 0.0003, + "step": 21329 + }, + { + "epoch": 2.31244579358196, + "grad_norm": 0.3769585192203522, + "learning_rate": 4.6003191643696506e-06, + "loss": 0.0018, + "step": 21330 + }, + { + "epoch": 2.31255420641804, + "grad_norm": 2.9659528732299805, + "learning_rate": 4.5995937908022635e-06, + "loss": 0.0121, + "step": 21331 + }, + { + "epoch": 2.3126626192541195, + "grad_norm": 0.23507408797740936, + "learning_rate": 4.5988684172348765e-06, + "loss": 0.0005, + "step": 21332 + }, + { + "epoch": 2.3127710320901995, + "grad_norm": 3.0387790203094482, + "learning_rate": 4.598143043667489e-06, + "loss": 0.0015, + "step": 21333 + }, + { + "epoch": 2.3128794449262795, + "grad_norm": 0.06433417648077011, + "learning_rate": 4.5974176701001025e-06, + "loss": 0.0002, + "step": 21334 + }, + { + "epoch": 2.312987857762359, + "grad_norm": 0.40112587809562683, + "learning_rate": 4.596692296532715e-06, + "loss": 0.0024, + "step": 21335 + }, + { + "epoch": 2.313096270598439, + "grad_norm": 4.968488693237305, + "learning_rate": 4.595966922965328e-06, + "loss": 0.0135, + "step": 21336 + }, + { + "epoch": 2.3132046834345186, + "grad_norm": 28.799421310424805, + "learning_rate": 4.59524154939794e-06, + "loss": 0.0197, + "step": 21337 + }, + { + "epoch": 2.3133130962705986, + "grad_norm": 1.1725516319274902, + "learning_rate": 4.594516175830553e-06, + "loss": 0.0247, + "step": 21338 + }, + { + "epoch": 2.313421509106678, + "grad_norm": 1.1846556663513184, + "learning_rate": 4.593790802263166e-06, + "loss": 0.0022, + "step": 21339 + }, + { + "epoch": 2.313529921942758, + "grad_norm": 0.009036262519657612, + "learning_rate": 4.593065428695779e-06, + "loss": 0.0001, + "step": 21340 + }, + { + "epoch": 2.3136383347788376, + "grad_norm": 0.4548817276954651, + "learning_rate": 4.592340055128392e-06, + "loss": 0.0011, + "step": 21341 + }, + { + "epoch": 2.3137467476149176, + "grad_norm": 0.06863351166248322, + "learning_rate": 4.591614681561004e-06, + "loss": 0.0006, + "step": 21342 + }, + { + "epoch": 2.3138551604509976, + "grad_norm": 0.07275541871786118, + "learning_rate": 4.590889307993618e-06, + "loss": 0.0008, + "step": 21343 + }, + { + "epoch": 2.313963573287077, + "grad_norm": 8.986686706542969, + "learning_rate": 4.59016393442623e-06, + "loss": 0.0179, + "step": 21344 + }, + { + "epoch": 2.314071986123157, + "grad_norm": 0.03346438705921173, + "learning_rate": 4.589438560858843e-06, + "loss": 0.0002, + "step": 21345 + }, + { + "epoch": 2.3141803989592367, + "grad_norm": 3.4886016845703125, + "learning_rate": 4.588713187291455e-06, + "loss": 0.0441, + "step": 21346 + }, + { + "epoch": 2.3142888117953166, + "grad_norm": 2.6092710494995117, + "learning_rate": 4.587987813724069e-06, + "loss": 0.0261, + "step": 21347 + }, + { + "epoch": 2.314397224631396, + "grad_norm": 0.19187894463539124, + "learning_rate": 4.587262440156681e-06, + "loss": 0.0011, + "step": 21348 + }, + { + "epoch": 2.314505637467476, + "grad_norm": 0.05184296891093254, + "learning_rate": 4.586537066589294e-06, + "loss": 0.0005, + "step": 21349 + }, + { + "epoch": 2.3146140503035557, + "grad_norm": 0.06630279868841171, + "learning_rate": 4.585811693021906e-06, + "loss": 0.0005, + "step": 21350 + }, + { + "epoch": 2.3147224631396357, + "grad_norm": 0.1712905615568161, + "learning_rate": 4.585086319454519e-06, + "loss": 0.001, + "step": 21351 + }, + { + "epoch": 2.3148308759757157, + "grad_norm": 0.23971286416053772, + "learning_rate": 4.584360945887132e-06, + "loss": 0.0011, + "step": 21352 + }, + { + "epoch": 2.3149392888117952, + "grad_norm": 3.108191967010498, + "learning_rate": 4.583635572319745e-06, + "loss": 0.0306, + "step": 21353 + }, + { + "epoch": 2.315047701647875, + "grad_norm": 0.5398393273353577, + "learning_rate": 4.582910198752358e-06, + "loss": 0.0008, + "step": 21354 + }, + { + "epoch": 2.3151561144839548, + "grad_norm": 0.1431492567062378, + "learning_rate": 4.58218482518497e-06, + "loss": 0.0012, + "step": 21355 + }, + { + "epoch": 2.3152645273200347, + "grad_norm": 0.28538596630096436, + "learning_rate": 4.581459451617584e-06, + "loss": 0.0004, + "step": 21356 + }, + { + "epoch": 2.3153729401561147, + "grad_norm": 0.038795072585344315, + "learning_rate": 4.580734078050196e-06, + "loss": 0.0004, + "step": 21357 + }, + { + "epoch": 2.3154813529921943, + "grad_norm": 0.03581944480538368, + "learning_rate": 4.580008704482809e-06, + "loss": 0.0001, + "step": 21358 + }, + { + "epoch": 2.3155897658282742, + "grad_norm": 0.016041038557887077, + "learning_rate": 4.579283330915421e-06, + "loss": 0.0001, + "step": 21359 + }, + { + "epoch": 2.315698178664354, + "grad_norm": 4.51008939743042, + "learning_rate": 4.578557957348035e-06, + "loss": 0.0052, + "step": 21360 + }, + { + "epoch": 2.3158065915004338, + "grad_norm": 0.20427127182483673, + "learning_rate": 4.577832583780647e-06, + "loss": 0.0012, + "step": 21361 + }, + { + "epoch": 2.3159150043365133, + "grad_norm": 0.44422927498817444, + "learning_rate": 4.57710721021326e-06, + "loss": 0.0006, + "step": 21362 + }, + { + "epoch": 2.3160234171725933, + "grad_norm": 4.038909435272217, + "learning_rate": 4.5763818366458725e-06, + "loss": 0.0245, + "step": 21363 + }, + { + "epoch": 2.316131830008673, + "grad_norm": 3.060559034347534, + "learning_rate": 4.5756564630784855e-06, + "loss": 0.0084, + "step": 21364 + }, + { + "epoch": 2.316240242844753, + "grad_norm": 0.014564885757863522, + "learning_rate": 4.5749310895110984e-06, + "loss": 0.0001, + "step": 21365 + }, + { + "epoch": 2.316348655680833, + "grad_norm": 1.2578001022338867, + "learning_rate": 4.5742057159437114e-06, + "loss": 0.0036, + "step": 21366 + }, + { + "epoch": 2.3164570685169124, + "grad_norm": 0.018637895584106445, + "learning_rate": 4.5734803423763244e-06, + "loss": 0.0001, + "step": 21367 + }, + { + "epoch": 2.3165654813529923, + "grad_norm": 1.643599033355713, + "learning_rate": 4.5727549688089366e-06, + "loss": 0.0111, + "step": 21368 + }, + { + "epoch": 2.316673894189072, + "grad_norm": 2.6374106407165527, + "learning_rate": 4.57202959524155e-06, + "loss": 0.0143, + "step": 21369 + }, + { + "epoch": 2.316782307025152, + "grad_norm": 2.4022037982940674, + "learning_rate": 4.5713042216741626e-06, + "loss": 0.0374, + "step": 21370 + }, + { + "epoch": 2.3168907198612314, + "grad_norm": 1.0731395483016968, + "learning_rate": 4.5705788481067755e-06, + "loss": 0.0037, + "step": 21371 + }, + { + "epoch": 2.3169991326973114, + "grad_norm": 1.8327903747558594, + "learning_rate": 4.569853474539388e-06, + "loss": 0.0167, + "step": 21372 + }, + { + "epoch": 2.317107545533391, + "grad_norm": 4.5787224769592285, + "learning_rate": 4.5691281009720015e-06, + "loss": 0.0363, + "step": 21373 + }, + { + "epoch": 2.317215958369471, + "grad_norm": 0.08792487531900406, + "learning_rate": 4.568402727404614e-06, + "loss": 0.0004, + "step": 21374 + }, + { + "epoch": 2.317324371205551, + "grad_norm": 0.0018396644154563546, + "learning_rate": 4.567677353837227e-06, + "loss": 0.0, + "step": 21375 + }, + { + "epoch": 2.3174327840416304, + "grad_norm": 12.640652656555176, + "learning_rate": 4.566951980269839e-06, + "loss": 0.0201, + "step": 21376 + }, + { + "epoch": 2.3175411968777104, + "grad_norm": 0.1392795741558075, + "learning_rate": 4.566226606702452e-06, + "loss": 0.0006, + "step": 21377 + }, + { + "epoch": 2.31764960971379, + "grad_norm": 0.12259475141763687, + "learning_rate": 4.565501233135065e-06, + "loss": 0.0007, + "step": 21378 + }, + { + "epoch": 2.31775802254987, + "grad_norm": 1.7831487655639648, + "learning_rate": 4.564775859567678e-06, + "loss": 0.003, + "step": 21379 + }, + { + "epoch": 2.31786643538595, + "grad_norm": 3.371434211730957, + "learning_rate": 4.564050486000291e-06, + "loss": 0.0511, + "step": 21380 + }, + { + "epoch": 2.3179748482220295, + "grad_norm": 0.6635599732398987, + "learning_rate": 4.563325112432903e-06, + "loss": 0.0007, + "step": 21381 + }, + { + "epoch": 2.3180832610581095, + "grad_norm": 4.53303337097168, + "learning_rate": 4.562599738865517e-06, + "loss": 0.0259, + "step": 21382 + }, + { + "epoch": 2.318191673894189, + "grad_norm": 0.1431981772184372, + "learning_rate": 4.561874365298129e-06, + "loss": 0.0006, + "step": 21383 + }, + { + "epoch": 2.318300086730269, + "grad_norm": 6.6341657638549805, + "learning_rate": 4.561148991730742e-06, + "loss": 0.0309, + "step": 21384 + }, + { + "epoch": 2.3184084995663485, + "grad_norm": 0.07186675816774368, + "learning_rate": 4.560423618163354e-06, + "loss": 0.0006, + "step": 21385 + }, + { + "epoch": 2.3185169124024285, + "grad_norm": 0.1004057377576828, + "learning_rate": 4.559698244595968e-06, + "loss": 0.0008, + "step": 21386 + }, + { + "epoch": 2.318625325238508, + "grad_norm": 0.7782362699508667, + "learning_rate": 4.55897287102858e-06, + "loss": 0.003, + "step": 21387 + }, + { + "epoch": 2.318733738074588, + "grad_norm": 0.029866095632314682, + "learning_rate": 4.558247497461193e-06, + "loss": 0.0001, + "step": 21388 + }, + { + "epoch": 2.318842150910668, + "grad_norm": 6.800045490264893, + "learning_rate": 4.557522123893805e-06, + "loss": 0.0337, + "step": 21389 + }, + { + "epoch": 2.3189505637467476, + "grad_norm": 5.0921783447265625, + "learning_rate": 4.556796750326418e-06, + "loss": 0.0069, + "step": 21390 + }, + { + "epoch": 2.3190589765828276, + "grad_norm": 2.860034465789795, + "learning_rate": 4.556071376759031e-06, + "loss": 0.0276, + "step": 21391 + }, + { + "epoch": 2.319167389418907, + "grad_norm": 0.16487948596477509, + "learning_rate": 4.555346003191644e-06, + "loss": 0.0011, + "step": 21392 + }, + { + "epoch": 2.319275802254987, + "grad_norm": 1.7431939840316772, + "learning_rate": 4.554620629624257e-06, + "loss": 0.002, + "step": 21393 + }, + { + "epoch": 2.3193842150910666, + "grad_norm": 0.6546019911766052, + "learning_rate": 4.553895256056869e-06, + "loss": 0.0018, + "step": 21394 + }, + { + "epoch": 2.3194926279271466, + "grad_norm": 0.1277443915605545, + "learning_rate": 4.553169882489483e-06, + "loss": 0.0003, + "step": 21395 + }, + { + "epoch": 2.319601040763226, + "grad_norm": 0.1623295247554779, + "learning_rate": 4.552444508922095e-06, + "loss": 0.0012, + "step": 21396 + }, + { + "epoch": 2.319709453599306, + "grad_norm": 0.0272494088858366, + "learning_rate": 4.551719135354708e-06, + "loss": 0.0003, + "step": 21397 + }, + { + "epoch": 2.319817866435386, + "grad_norm": 1.0847601890563965, + "learning_rate": 4.55099376178732e-06, + "loss": 0.0029, + "step": 21398 + }, + { + "epoch": 2.3199262792714657, + "grad_norm": 0.04390662908554077, + "learning_rate": 4.550268388219934e-06, + "loss": 0.0004, + "step": 21399 + }, + { + "epoch": 2.3200346921075456, + "grad_norm": 11.442466735839844, + "learning_rate": 4.549543014652546e-06, + "loss": 0.0197, + "step": 21400 + }, + { + "epoch": 2.320143104943625, + "grad_norm": 0.5814111828804016, + "learning_rate": 4.548817641085159e-06, + "loss": 0.0007, + "step": 21401 + }, + { + "epoch": 2.320251517779705, + "grad_norm": 0.052667275071144104, + "learning_rate": 4.5480922675177715e-06, + "loss": 0.0005, + "step": 21402 + }, + { + "epoch": 2.3203599306157847, + "grad_norm": 12.741473197937012, + "learning_rate": 4.5473668939503845e-06, + "loss": 0.0393, + "step": 21403 + }, + { + "epoch": 2.3204683434518647, + "grad_norm": 0.053654130548238754, + "learning_rate": 4.5466415203829975e-06, + "loss": 0.0004, + "step": 21404 + }, + { + "epoch": 2.3205767562879447, + "grad_norm": 2.547722101211548, + "learning_rate": 4.5459161468156104e-06, + "loss": 0.0256, + "step": 21405 + }, + { + "epoch": 2.3206851691240242, + "grad_norm": 0.009152371436357498, + "learning_rate": 4.5451907732482234e-06, + "loss": 0.0001, + "step": 21406 + }, + { + "epoch": 2.320793581960104, + "grad_norm": 5.951592445373535, + "learning_rate": 4.5444653996808356e-06, + "loss": 0.0156, + "step": 21407 + }, + { + "epoch": 2.3209019947961838, + "grad_norm": 1.6241413354873657, + "learning_rate": 4.543740026113449e-06, + "loss": 0.0236, + "step": 21408 + }, + { + "epoch": 2.3210104076322637, + "grad_norm": 0.027489108964800835, + "learning_rate": 4.5430146525460616e-06, + "loss": 0.0001, + "step": 21409 + }, + { + "epoch": 2.3211188204683433, + "grad_norm": 3.19002366065979, + "learning_rate": 4.5422892789786745e-06, + "loss": 0.0175, + "step": 21410 + }, + { + "epoch": 2.3212272333044233, + "grad_norm": 2.3900420665740967, + "learning_rate": 4.541563905411287e-06, + "loss": 0.0071, + "step": 21411 + }, + { + "epoch": 2.3213356461405033, + "grad_norm": 0.08726273477077484, + "learning_rate": 4.5408385318439005e-06, + "loss": 0.0004, + "step": 21412 + }, + { + "epoch": 2.321444058976583, + "grad_norm": 0.3620941638946533, + "learning_rate": 4.540113158276513e-06, + "loss": 0.0019, + "step": 21413 + }, + { + "epoch": 2.3215524718126628, + "grad_norm": 2.002356767654419, + "learning_rate": 4.539387784709126e-06, + "loss": 0.0033, + "step": 21414 + }, + { + "epoch": 2.3216608846487423, + "grad_norm": 0.0787738487124443, + "learning_rate": 4.538662411141738e-06, + "loss": 0.0006, + "step": 21415 + }, + { + "epoch": 2.3217692974848223, + "grad_norm": 1.34762704372406, + "learning_rate": 4.537937037574351e-06, + "loss": 0.0086, + "step": 21416 + }, + { + "epoch": 2.321877710320902, + "grad_norm": 3.7939929962158203, + "learning_rate": 4.537211664006964e-06, + "loss": 0.0161, + "step": 21417 + }, + { + "epoch": 2.321986123156982, + "grad_norm": 1.0811737775802612, + "learning_rate": 4.536486290439577e-06, + "loss": 0.0011, + "step": 21418 + }, + { + "epoch": 2.3220945359930614, + "grad_norm": 0.02387145161628723, + "learning_rate": 4.53576091687219e-06, + "loss": 0.0002, + "step": 21419 + }, + { + "epoch": 2.3222029488291414, + "grad_norm": 0.12299978733062744, + "learning_rate": 4.535035543304802e-06, + "loss": 0.0005, + "step": 21420 + }, + { + "epoch": 2.3223113616652213, + "grad_norm": 4.26116418838501, + "learning_rate": 4.534310169737416e-06, + "loss": 0.0234, + "step": 21421 + }, + { + "epoch": 2.322419774501301, + "grad_norm": 0.08980921655893326, + "learning_rate": 4.533584796170028e-06, + "loss": 0.0007, + "step": 21422 + }, + { + "epoch": 2.322528187337381, + "grad_norm": 5.751713752746582, + "learning_rate": 4.532859422602641e-06, + "loss": 0.0183, + "step": 21423 + }, + { + "epoch": 2.3226366001734604, + "grad_norm": 1.4451591968536377, + "learning_rate": 4.532134049035253e-06, + "loss": 0.0048, + "step": 21424 + }, + { + "epoch": 2.3227450130095404, + "grad_norm": 2.4874885082244873, + "learning_rate": 4.531408675467867e-06, + "loss": 0.0061, + "step": 21425 + }, + { + "epoch": 2.32285342584562, + "grad_norm": 4.946592330932617, + "learning_rate": 4.530683301900479e-06, + "loss": 0.0368, + "step": 21426 + }, + { + "epoch": 2.3229618386817, + "grad_norm": 0.13800549507141113, + "learning_rate": 4.529957928333092e-06, + "loss": 0.0006, + "step": 21427 + }, + { + "epoch": 2.3230702515177795, + "grad_norm": 4.15167236328125, + "learning_rate": 4.529232554765704e-06, + "loss": 0.0079, + "step": 21428 + }, + { + "epoch": 2.3231786643538594, + "grad_norm": 0.2169954925775528, + "learning_rate": 4.528507181198318e-06, + "loss": 0.001, + "step": 21429 + }, + { + "epoch": 2.3232870771899394, + "grad_norm": 0.23163436353206635, + "learning_rate": 4.52778180763093e-06, + "loss": 0.001, + "step": 21430 + }, + { + "epoch": 2.323395490026019, + "grad_norm": 0.22406287491321564, + "learning_rate": 4.527056434063543e-06, + "loss": 0.0007, + "step": 21431 + }, + { + "epoch": 2.323503902862099, + "grad_norm": 0.05829164758324623, + "learning_rate": 4.526331060496156e-06, + "loss": 0.0002, + "step": 21432 + }, + { + "epoch": 2.3236123156981785, + "grad_norm": 1.3909178972244263, + "learning_rate": 4.525605686928768e-06, + "loss": 0.0058, + "step": 21433 + }, + { + "epoch": 2.3237207285342585, + "grad_norm": 0.3322831392288208, + "learning_rate": 4.524880313361382e-06, + "loss": 0.0011, + "step": 21434 + }, + { + "epoch": 2.3238291413703385, + "grad_norm": 5.044602394104004, + "learning_rate": 4.524154939793994e-06, + "loss": 0.0218, + "step": 21435 + }, + { + "epoch": 2.323937554206418, + "grad_norm": 3.840726375579834, + "learning_rate": 4.523429566226607e-06, + "loss": 0.0085, + "step": 21436 + }, + { + "epoch": 2.324045967042498, + "grad_norm": 6.641668319702148, + "learning_rate": 4.522704192659219e-06, + "loss": 0.0043, + "step": 21437 + }, + { + "epoch": 2.3241543798785775, + "grad_norm": 0.9876584410667419, + "learning_rate": 4.521978819091833e-06, + "loss": 0.0043, + "step": 21438 + }, + { + "epoch": 2.3242627927146575, + "grad_norm": 0.053657159209251404, + "learning_rate": 4.521253445524445e-06, + "loss": 0.0003, + "step": 21439 + }, + { + "epoch": 2.324371205550737, + "grad_norm": 4.980261325836182, + "learning_rate": 4.520528071957058e-06, + "loss": 0.0025, + "step": 21440 + }, + { + "epoch": 2.324479618386817, + "grad_norm": 0.022369567304849625, + "learning_rate": 4.5198026983896705e-06, + "loss": 0.0001, + "step": 21441 + }, + { + "epoch": 2.3245880312228966, + "grad_norm": 0.26208120584487915, + "learning_rate": 4.519077324822284e-06, + "loss": 0.0008, + "step": 21442 + }, + { + "epoch": 2.3246964440589766, + "grad_norm": 1.6519508361816406, + "learning_rate": 4.5183519512548965e-06, + "loss": 0.0049, + "step": 21443 + }, + { + "epoch": 2.3248048568950566, + "grad_norm": 2.653102397918701, + "learning_rate": 4.5176265776875094e-06, + "loss": 0.0243, + "step": 21444 + }, + { + "epoch": 2.324913269731136, + "grad_norm": 0.5386205315589905, + "learning_rate": 4.5169012041201224e-06, + "loss": 0.0027, + "step": 21445 + }, + { + "epoch": 2.325021682567216, + "grad_norm": 0.032802075147628784, + "learning_rate": 4.516175830552735e-06, + "loss": 0.0004, + "step": 21446 + }, + { + "epoch": 2.3251300954032956, + "grad_norm": 24.122289657592773, + "learning_rate": 4.515450456985348e-06, + "loss": 0.0706, + "step": 21447 + }, + { + "epoch": 2.3252385082393756, + "grad_norm": 0.04349656403064728, + "learning_rate": 4.5147250834179606e-06, + "loss": 0.0002, + "step": 21448 + }, + { + "epoch": 2.325346921075455, + "grad_norm": 3.920844316482544, + "learning_rate": 4.5139997098505735e-06, + "loss": 0.0222, + "step": 21449 + }, + { + "epoch": 2.325455333911535, + "grad_norm": 1.4593313932418823, + "learning_rate": 4.513274336283186e-06, + "loss": 0.0068, + "step": 21450 + }, + { + "epoch": 2.3255637467476147, + "grad_norm": 0.5539354681968689, + "learning_rate": 4.5125489627157995e-06, + "loss": 0.0021, + "step": 21451 + }, + { + "epoch": 2.3256721595836947, + "grad_norm": 1.550335168838501, + "learning_rate": 4.511823589148412e-06, + "loss": 0.0011, + "step": 21452 + }, + { + "epoch": 2.3257805724197746, + "grad_norm": 1.2570736408233643, + "learning_rate": 4.511098215581025e-06, + "loss": 0.0005, + "step": 21453 + }, + { + "epoch": 2.325888985255854, + "grad_norm": 3.094566822052002, + "learning_rate": 4.510372842013638e-06, + "loss": 0.0059, + "step": 21454 + }, + { + "epoch": 2.325997398091934, + "grad_norm": 3.81596302986145, + "learning_rate": 4.509647468446251e-06, + "loss": 0.0206, + "step": 21455 + }, + { + "epoch": 2.3261058109280137, + "grad_norm": 4.319004535675049, + "learning_rate": 4.508922094878863e-06, + "loss": 0.0031, + "step": 21456 + }, + { + "epoch": 2.3262142237640937, + "grad_norm": 3.8166165351867676, + "learning_rate": 4.508196721311476e-06, + "loss": 0.0036, + "step": 21457 + }, + { + "epoch": 2.3263226366001737, + "grad_norm": 7.484851837158203, + "learning_rate": 4.507471347744089e-06, + "loss": 0.0055, + "step": 21458 + }, + { + "epoch": 2.3264310494362532, + "grad_norm": 0.15176941454410553, + "learning_rate": 4.506745974176701e-06, + "loss": 0.0004, + "step": 21459 + }, + { + "epoch": 2.326539462272333, + "grad_norm": 2.841233253479004, + "learning_rate": 4.506020600609315e-06, + "loss": 0.0238, + "step": 21460 + }, + { + "epoch": 2.3266478751084128, + "grad_norm": 11.681811332702637, + "learning_rate": 4.505295227041927e-06, + "loss": 0.0833, + "step": 21461 + }, + { + "epoch": 2.3267562879444927, + "grad_norm": 1.4754496812820435, + "learning_rate": 4.50456985347454e-06, + "loss": 0.0018, + "step": 21462 + }, + { + "epoch": 2.3268647007805723, + "grad_norm": 2.816260814666748, + "learning_rate": 4.503844479907152e-06, + "loss": 0.0352, + "step": 21463 + }, + { + "epoch": 2.3269731136166523, + "grad_norm": 0.1116771399974823, + "learning_rate": 4.503119106339766e-06, + "loss": 0.0007, + "step": 21464 + }, + { + "epoch": 2.327081526452732, + "grad_norm": 4.325491428375244, + "learning_rate": 4.502393732772378e-06, + "loss": 0.0096, + "step": 21465 + }, + { + "epoch": 2.327189939288812, + "grad_norm": 1.881795048713684, + "learning_rate": 4.501668359204991e-06, + "loss": 0.0211, + "step": 21466 + }, + { + "epoch": 2.327298352124892, + "grad_norm": 0.1266607642173767, + "learning_rate": 4.500942985637604e-06, + "loss": 0.0007, + "step": 21467 + }, + { + "epoch": 2.3274067649609713, + "grad_norm": 1.872301459312439, + "learning_rate": 4.500217612070217e-06, + "loss": 0.0231, + "step": 21468 + }, + { + "epoch": 2.3275151777970513, + "grad_norm": 4.947068214416504, + "learning_rate": 4.499492238502829e-06, + "loss": 0.0305, + "step": 21469 + }, + { + "epoch": 2.327623590633131, + "grad_norm": 1.2912319898605347, + "learning_rate": 4.498766864935442e-06, + "loss": 0.0037, + "step": 21470 + }, + { + "epoch": 2.327732003469211, + "grad_norm": 2.6534392833709717, + "learning_rate": 4.498041491368055e-06, + "loss": 0.0084, + "step": 21471 + }, + { + "epoch": 2.3278404163052904, + "grad_norm": 6.6687421798706055, + "learning_rate": 4.497316117800667e-06, + "loss": 0.0279, + "step": 21472 + }, + { + "epoch": 2.3279488291413704, + "grad_norm": 0.014555800706148148, + "learning_rate": 4.496590744233281e-06, + "loss": 0.0001, + "step": 21473 + }, + { + "epoch": 2.32805724197745, + "grad_norm": 1.8259735107421875, + "learning_rate": 4.495865370665893e-06, + "loss": 0.0117, + "step": 21474 + }, + { + "epoch": 2.32816565481353, + "grad_norm": 3.1847362518310547, + "learning_rate": 4.495139997098506e-06, + "loss": 0.0038, + "step": 21475 + }, + { + "epoch": 2.32827406764961, + "grad_norm": 0.25908026099205017, + "learning_rate": 4.494414623531118e-06, + "loss": 0.0012, + "step": 21476 + }, + { + "epoch": 2.3283824804856894, + "grad_norm": 1.6506242752075195, + "learning_rate": 4.493689249963732e-06, + "loss": 0.0057, + "step": 21477 + }, + { + "epoch": 2.3284908933217694, + "grad_norm": 2.724485397338867, + "learning_rate": 4.492963876396344e-06, + "loss": 0.0056, + "step": 21478 + }, + { + "epoch": 2.328599306157849, + "grad_norm": 0.5470009446144104, + "learning_rate": 4.492238502828957e-06, + "loss": 0.0012, + "step": 21479 + }, + { + "epoch": 2.328707718993929, + "grad_norm": 2.5771758556365967, + "learning_rate": 4.49151312926157e-06, + "loss": 0.048, + "step": 21480 + }, + { + "epoch": 2.328816131830009, + "grad_norm": 0.12107736617326736, + "learning_rate": 4.490787755694183e-06, + "loss": 0.0005, + "step": 21481 + }, + { + "epoch": 2.3289245446660884, + "grad_norm": 0.09612388908863068, + "learning_rate": 4.4900623821267955e-06, + "loss": 0.0005, + "step": 21482 + }, + { + "epoch": 2.3290329575021684, + "grad_norm": 1.0730448961257935, + "learning_rate": 4.4893370085594084e-06, + "loss": 0.0057, + "step": 21483 + }, + { + "epoch": 2.329141370338248, + "grad_norm": 1.7404240369796753, + "learning_rate": 4.4886116349920214e-06, + "loss": 0.0008, + "step": 21484 + }, + { + "epoch": 2.329249783174328, + "grad_norm": 0.40540018677711487, + "learning_rate": 4.487886261424634e-06, + "loss": 0.0009, + "step": 21485 + }, + { + "epoch": 2.3293581960104075, + "grad_norm": 0.4249764084815979, + "learning_rate": 4.487160887857247e-06, + "loss": 0.0006, + "step": 21486 + }, + { + "epoch": 2.3294666088464875, + "grad_norm": 0.09572556614875793, + "learning_rate": 4.4864355142898596e-06, + "loss": 0.001, + "step": 21487 + }, + { + "epoch": 2.329575021682567, + "grad_norm": 0.03229421749711037, + "learning_rate": 4.4857101407224726e-06, + "loss": 0.0003, + "step": 21488 + }, + { + "epoch": 2.329683434518647, + "grad_norm": 0.03732183948159218, + "learning_rate": 4.484984767155085e-06, + "loss": 0.0005, + "step": 21489 + }, + { + "epoch": 2.329791847354727, + "grad_norm": 8.799104690551758, + "learning_rate": 4.4842593935876985e-06, + "loss": 0.0074, + "step": 21490 + }, + { + "epoch": 2.3299002601908065, + "grad_norm": 0.3453771770000458, + "learning_rate": 4.483534020020311e-06, + "loss": 0.0017, + "step": 21491 + }, + { + "epoch": 2.3300086730268865, + "grad_norm": 0.07174618542194366, + "learning_rate": 4.482808646452924e-06, + "loss": 0.0003, + "step": 21492 + }, + { + "epoch": 2.330117085862966, + "grad_norm": 8.271306037902832, + "learning_rate": 4.482083272885537e-06, + "loss": 0.0142, + "step": 21493 + }, + { + "epoch": 2.330225498699046, + "grad_norm": 17.84942054748535, + "learning_rate": 4.48135789931815e-06, + "loss": 0.0104, + "step": 21494 + }, + { + "epoch": 2.3303339115351256, + "grad_norm": 0.01328473724424839, + "learning_rate": 4.480632525750762e-06, + "loss": 0.0001, + "step": 21495 + }, + { + "epoch": 2.3304423243712056, + "grad_norm": 0.7540588974952698, + "learning_rate": 4.479907152183375e-06, + "loss": 0.0007, + "step": 21496 + }, + { + "epoch": 2.330550737207285, + "grad_norm": 28.633981704711914, + "learning_rate": 4.479181778615988e-06, + "loss": 0.0179, + "step": 21497 + }, + { + "epoch": 2.330659150043365, + "grad_norm": 4.333614826202393, + "learning_rate": 4.4784564050486e-06, + "loss": 0.0085, + "step": 21498 + }, + { + "epoch": 2.330767562879445, + "grad_norm": 0.2848246097564697, + "learning_rate": 4.477731031481214e-06, + "loss": 0.0019, + "step": 21499 + }, + { + "epoch": 2.3308759757155246, + "grad_norm": 0.9890261888504028, + "learning_rate": 4.477005657913826e-06, + "loss": 0.0042, + "step": 21500 + }, + { + "epoch": 2.3309843885516046, + "grad_norm": 2.7882516384124756, + "learning_rate": 4.476280284346439e-06, + "loss": 0.0267, + "step": 21501 + }, + { + "epoch": 2.331092801387684, + "grad_norm": 0.9060301184654236, + "learning_rate": 4.475554910779051e-06, + "loss": 0.0037, + "step": 21502 + }, + { + "epoch": 2.331201214223764, + "grad_norm": 0.8040800094604492, + "learning_rate": 4.474829537211665e-06, + "loss": 0.001, + "step": 21503 + }, + { + "epoch": 2.331309627059844, + "grad_norm": 0.19916188716888428, + "learning_rate": 4.474104163644277e-06, + "loss": 0.0011, + "step": 21504 + }, + { + "epoch": 2.3314180398959237, + "grad_norm": 1.7087994813919067, + "learning_rate": 4.47337879007689e-06, + "loss": 0.0296, + "step": 21505 + }, + { + "epoch": 2.3315264527320037, + "grad_norm": 0.027050668373703957, + "learning_rate": 4.472653416509503e-06, + "loss": 0.0001, + "step": 21506 + }, + { + "epoch": 2.331634865568083, + "grad_norm": 0.00506388396024704, + "learning_rate": 4.471928042942116e-06, + "loss": 0.0, + "step": 21507 + }, + { + "epoch": 2.331743278404163, + "grad_norm": 0.0992770865559578, + "learning_rate": 4.471202669374728e-06, + "loss": 0.0009, + "step": 21508 + }, + { + "epoch": 2.3318516912402427, + "grad_norm": 0.10551994293928146, + "learning_rate": 4.470477295807341e-06, + "loss": 0.0007, + "step": 21509 + }, + { + "epoch": 2.3319601040763227, + "grad_norm": 0.013390161097049713, + "learning_rate": 4.469751922239954e-06, + "loss": 0.0001, + "step": 21510 + }, + { + "epoch": 2.3320685169124022, + "grad_norm": 4.6707305908203125, + "learning_rate": 4.469026548672566e-06, + "loss": 0.0494, + "step": 21511 + }, + { + "epoch": 2.3321769297484822, + "grad_norm": 0.3232259750366211, + "learning_rate": 4.46830117510518e-06, + "loss": 0.0005, + "step": 21512 + }, + { + "epoch": 2.332285342584562, + "grad_norm": 0.40626370906829834, + "learning_rate": 4.467575801537792e-06, + "loss": 0.0015, + "step": 21513 + }, + { + "epoch": 2.3323937554206418, + "grad_norm": 0.9650663733482361, + "learning_rate": 4.466850427970405e-06, + "loss": 0.003, + "step": 21514 + }, + { + "epoch": 2.3325021682567217, + "grad_norm": 1.6851987838745117, + "learning_rate": 4.466125054403017e-06, + "loss": 0.0035, + "step": 21515 + }, + { + "epoch": 2.3326105810928013, + "grad_norm": 0.05494380369782448, + "learning_rate": 4.465399680835631e-06, + "loss": 0.0002, + "step": 21516 + }, + { + "epoch": 2.3327189939288813, + "grad_norm": 4.983859062194824, + "learning_rate": 4.464674307268243e-06, + "loss": 0.0146, + "step": 21517 + }, + { + "epoch": 2.332827406764961, + "grad_norm": 3.1710336208343506, + "learning_rate": 4.463948933700856e-06, + "loss": 0.0098, + "step": 21518 + }, + { + "epoch": 2.332935819601041, + "grad_norm": 0.14325349032878876, + "learning_rate": 4.463223560133469e-06, + "loss": 0.001, + "step": 21519 + }, + { + "epoch": 2.3330442324371203, + "grad_norm": 3.6503000259399414, + "learning_rate": 4.462498186566082e-06, + "loss": 0.0751, + "step": 21520 + }, + { + "epoch": 2.3331526452732003, + "grad_norm": 0.045499831438064575, + "learning_rate": 4.4617728129986945e-06, + "loss": 0.0004, + "step": 21521 + }, + { + "epoch": 2.3332610581092803, + "grad_norm": 12.01272201538086, + "learning_rate": 4.4610474394313075e-06, + "loss": 0.0123, + "step": 21522 + }, + { + "epoch": 2.33336947094536, + "grad_norm": 0.018665216863155365, + "learning_rate": 4.4603220658639204e-06, + "loss": 0.0001, + "step": 21523 + }, + { + "epoch": 2.33347788378144, + "grad_norm": 0.6255878806114197, + "learning_rate": 4.459596692296533e-06, + "loss": 0.0011, + "step": 21524 + }, + { + "epoch": 2.3335862966175194, + "grad_norm": 4.684072017669678, + "learning_rate": 4.458871318729146e-06, + "loss": 0.0209, + "step": 21525 + }, + { + "epoch": 2.3336947094535994, + "grad_norm": 0.043999120593070984, + "learning_rate": 4.4581459451617586e-06, + "loss": 0.0003, + "step": 21526 + }, + { + "epoch": 2.3338031222896793, + "grad_norm": 0.03930012509226799, + "learning_rate": 4.4574205715943716e-06, + "loss": 0.0001, + "step": 21527 + }, + { + "epoch": 2.333911535125759, + "grad_norm": 0.4366627037525177, + "learning_rate": 4.456695198026984e-06, + "loss": 0.0024, + "step": 21528 + }, + { + "epoch": 2.334019947961839, + "grad_norm": 2.8736684322357178, + "learning_rate": 4.4559698244595975e-06, + "loss": 0.0049, + "step": 21529 + }, + { + "epoch": 2.3341283607979184, + "grad_norm": 1.6374781131744385, + "learning_rate": 4.45524445089221e-06, + "loss": 0.005, + "step": 21530 + }, + { + "epoch": 2.3342367736339984, + "grad_norm": 0.7158272862434387, + "learning_rate": 4.454519077324823e-06, + "loss": 0.0019, + "step": 21531 + }, + { + "epoch": 2.334345186470078, + "grad_norm": 10.22960376739502, + "learning_rate": 4.453793703757436e-06, + "loss": 0.0382, + "step": 21532 + }, + { + "epoch": 2.334453599306158, + "grad_norm": 1.6664482355117798, + "learning_rate": 4.453068330190049e-06, + "loss": 0.0072, + "step": 21533 + }, + { + "epoch": 2.3345620121422375, + "grad_norm": 4.09210729598999, + "learning_rate": 4.452342956622661e-06, + "loss": 0.0178, + "step": 21534 + }, + { + "epoch": 2.3346704249783174, + "grad_norm": 0.16374830901622772, + "learning_rate": 4.451617583055274e-06, + "loss": 0.0007, + "step": 21535 + }, + { + "epoch": 2.3347788378143974, + "grad_norm": 0.00983364973217249, + "learning_rate": 4.450892209487887e-06, + "loss": 0.0001, + "step": 21536 + }, + { + "epoch": 2.334887250650477, + "grad_norm": 5.4296650886535645, + "learning_rate": 4.4501668359205e-06, + "loss": 0.0127, + "step": 21537 + }, + { + "epoch": 2.334995663486557, + "grad_norm": 0.10825138539075851, + "learning_rate": 4.449441462353113e-06, + "loss": 0.0005, + "step": 21538 + }, + { + "epoch": 2.3351040763226365, + "grad_norm": 2.6354122161865234, + "learning_rate": 4.448716088785725e-06, + "loss": 0.0073, + "step": 21539 + }, + { + "epoch": 2.3352124891587165, + "grad_norm": 6.354498863220215, + "learning_rate": 4.447990715218338e-06, + "loss": 0.0217, + "step": 21540 + }, + { + "epoch": 2.335320901994796, + "grad_norm": 0.5855889916419983, + "learning_rate": 4.44726534165095e-06, + "loss": 0.0022, + "step": 21541 + }, + { + "epoch": 2.335429314830876, + "grad_norm": 0.9780202507972717, + "learning_rate": 4.446539968083564e-06, + "loss": 0.0012, + "step": 21542 + }, + { + "epoch": 2.3355377276669556, + "grad_norm": 2.343123435974121, + "learning_rate": 4.445814594516176e-06, + "loss": 0.0257, + "step": 21543 + }, + { + "epoch": 2.3356461405030355, + "grad_norm": 0.029216155409812927, + "learning_rate": 4.445089220948789e-06, + "loss": 0.0002, + "step": 21544 + }, + { + "epoch": 2.3357545533391155, + "grad_norm": 1.8260002136230469, + "learning_rate": 4.444363847381402e-06, + "loss": 0.0256, + "step": 21545 + }, + { + "epoch": 2.335862966175195, + "grad_norm": 1.2967913150787354, + "learning_rate": 4.443638473814015e-06, + "loss": 0.0162, + "step": 21546 + }, + { + "epoch": 2.335971379011275, + "grad_norm": 0.09858302772045135, + "learning_rate": 4.442913100246627e-06, + "loss": 0.0003, + "step": 21547 + }, + { + "epoch": 2.3360797918473546, + "grad_norm": 3.6625611782073975, + "learning_rate": 4.44218772667924e-06, + "loss": 0.0084, + "step": 21548 + }, + { + "epoch": 2.3361882046834346, + "grad_norm": 0.08540890365839005, + "learning_rate": 4.441462353111853e-06, + "loss": 0.0004, + "step": 21549 + }, + { + "epoch": 2.336296617519514, + "grad_norm": 2.3188180923461914, + "learning_rate": 4.440736979544466e-06, + "loss": 0.0013, + "step": 21550 + }, + { + "epoch": 2.336405030355594, + "grad_norm": 0.02428979054093361, + "learning_rate": 4.440011605977079e-06, + "loss": 0.0002, + "step": 21551 + }, + { + "epoch": 2.3365134431916736, + "grad_norm": 2.431767225265503, + "learning_rate": 4.439286232409691e-06, + "loss": 0.0054, + "step": 21552 + }, + { + "epoch": 2.3366218560277536, + "grad_norm": 1.0635199546813965, + "learning_rate": 4.438560858842304e-06, + "loss": 0.0048, + "step": 21553 + }, + { + "epoch": 2.3367302688638336, + "grad_norm": 4.385517120361328, + "learning_rate": 4.437835485274916e-06, + "loss": 0.0122, + "step": 21554 + }, + { + "epoch": 2.336838681699913, + "grad_norm": 2.9585559368133545, + "learning_rate": 4.43711011170753e-06, + "loss": 0.0368, + "step": 21555 + }, + { + "epoch": 2.336947094535993, + "grad_norm": 1.1648179292678833, + "learning_rate": 4.436384738140142e-06, + "loss": 0.0062, + "step": 21556 + }, + { + "epoch": 2.3370555073720727, + "grad_norm": 7.8358612060546875, + "learning_rate": 4.435659364572755e-06, + "loss": 0.0484, + "step": 21557 + }, + { + "epoch": 2.3371639202081527, + "grad_norm": 6.359304904937744, + "learning_rate": 4.434933991005368e-06, + "loss": 0.0127, + "step": 21558 + }, + { + "epoch": 2.3372723330442327, + "grad_norm": 1.1548906564712524, + "learning_rate": 4.434208617437981e-06, + "loss": 0.0031, + "step": 21559 + }, + { + "epoch": 2.337380745880312, + "grad_norm": 10.969192504882812, + "learning_rate": 4.4334832438705935e-06, + "loss": 0.0249, + "step": 21560 + }, + { + "epoch": 2.337489158716392, + "grad_norm": 0.39716970920562744, + "learning_rate": 4.4327578703032065e-06, + "loss": 0.0018, + "step": 21561 + }, + { + "epoch": 2.3375975715524717, + "grad_norm": 4.585058689117432, + "learning_rate": 4.4320324967358194e-06, + "loss": 0.0377, + "step": 21562 + }, + { + "epoch": 2.3377059843885517, + "grad_norm": 2.444317102432251, + "learning_rate": 4.4313071231684324e-06, + "loss": 0.0038, + "step": 21563 + }, + { + "epoch": 2.3378143972246312, + "grad_norm": 0.015391245484352112, + "learning_rate": 4.4305817496010454e-06, + "loss": 0.0001, + "step": 21564 + }, + { + "epoch": 2.3379228100607112, + "grad_norm": 1.5040483474731445, + "learning_rate": 4.4298563760336576e-06, + "loss": 0.0356, + "step": 21565 + }, + { + "epoch": 2.3380312228967908, + "grad_norm": 0.7942842841148376, + "learning_rate": 4.4291310024662706e-06, + "loss": 0.0011, + "step": 21566 + }, + { + "epoch": 2.3381396357328708, + "grad_norm": 3.2111878395080566, + "learning_rate": 4.428405628898883e-06, + "loss": 0.048, + "step": 21567 + }, + { + "epoch": 2.3382480485689507, + "grad_norm": 0.016790127381682396, + "learning_rate": 4.4276802553314965e-06, + "loss": 0.0001, + "step": 21568 + }, + { + "epoch": 2.3383564614050303, + "grad_norm": 0.04895789921283722, + "learning_rate": 4.426954881764109e-06, + "loss": 0.0003, + "step": 21569 + }, + { + "epoch": 2.3384648742411103, + "grad_norm": 4.066097736358643, + "learning_rate": 4.426229508196722e-06, + "loss": 0.0339, + "step": 21570 + }, + { + "epoch": 2.33857328707719, + "grad_norm": 0.30127671360969543, + "learning_rate": 4.425504134629335e-06, + "loss": 0.0012, + "step": 21571 + }, + { + "epoch": 2.33868169991327, + "grad_norm": 0.2863726317882538, + "learning_rate": 4.424778761061948e-06, + "loss": 0.0008, + "step": 21572 + }, + { + "epoch": 2.3387901127493493, + "grad_norm": 0.9681770205497742, + "learning_rate": 4.42405338749456e-06, + "loss": 0.0022, + "step": 21573 + }, + { + "epoch": 2.3388985255854293, + "grad_norm": 0.02287961356341839, + "learning_rate": 4.423328013927173e-06, + "loss": 0.0001, + "step": 21574 + }, + { + "epoch": 2.339006938421509, + "grad_norm": 0.01679958403110504, + "learning_rate": 4.422602640359786e-06, + "loss": 0.0001, + "step": 21575 + }, + { + "epoch": 2.339115351257589, + "grad_norm": 3.0982320308685303, + "learning_rate": 4.421877266792399e-06, + "loss": 0.0059, + "step": 21576 + }, + { + "epoch": 2.339223764093669, + "grad_norm": 0.013167551718652248, + "learning_rate": 4.421151893225012e-06, + "loss": 0.0001, + "step": 21577 + }, + { + "epoch": 2.3393321769297484, + "grad_norm": 3.976243019104004, + "learning_rate": 4.420426519657624e-06, + "loss": 0.0017, + "step": 21578 + }, + { + "epoch": 2.3394405897658284, + "grad_norm": 3.830092430114746, + "learning_rate": 4.419701146090237e-06, + "loss": 0.0336, + "step": 21579 + }, + { + "epoch": 2.339549002601908, + "grad_norm": 0.051034800708293915, + "learning_rate": 4.418975772522849e-06, + "loss": 0.0003, + "step": 21580 + }, + { + "epoch": 2.339657415437988, + "grad_norm": 0.27647730708122253, + "learning_rate": 4.418250398955463e-06, + "loss": 0.0013, + "step": 21581 + }, + { + "epoch": 2.339765828274068, + "grad_norm": 0.8771579265594482, + "learning_rate": 4.417525025388075e-06, + "loss": 0.0015, + "step": 21582 + }, + { + "epoch": 2.3398742411101474, + "grad_norm": 10.609027862548828, + "learning_rate": 4.416799651820688e-06, + "loss": 0.0261, + "step": 21583 + }, + { + "epoch": 2.3399826539462274, + "grad_norm": 12.820513725280762, + "learning_rate": 4.416074278253301e-06, + "loss": 0.0223, + "step": 21584 + }, + { + "epoch": 2.340091066782307, + "grad_norm": 2.0199897289276123, + "learning_rate": 4.415348904685914e-06, + "loss": 0.0029, + "step": 21585 + }, + { + "epoch": 2.340199479618387, + "grad_norm": 0.13812988996505737, + "learning_rate": 4.414623531118526e-06, + "loss": 0.0011, + "step": 21586 + }, + { + "epoch": 2.3403078924544665, + "grad_norm": 7.766622543334961, + "learning_rate": 4.413898157551139e-06, + "loss": 0.003, + "step": 21587 + }, + { + "epoch": 2.3404163052905465, + "grad_norm": 4.566596031188965, + "learning_rate": 4.413172783983752e-06, + "loss": 0.0083, + "step": 21588 + }, + { + "epoch": 2.340524718126626, + "grad_norm": 2.025010108947754, + "learning_rate": 4.412447410416365e-06, + "loss": 0.0214, + "step": 21589 + }, + { + "epoch": 2.340633130962706, + "grad_norm": 0.48585957288742065, + "learning_rate": 4.411722036848978e-06, + "loss": 0.0015, + "step": 21590 + }, + { + "epoch": 2.340741543798786, + "grad_norm": 0.7422318458557129, + "learning_rate": 4.41099666328159e-06, + "loss": 0.003, + "step": 21591 + }, + { + "epoch": 2.3408499566348655, + "grad_norm": 1.1168714761734009, + "learning_rate": 4.410271289714203e-06, + "loss": 0.0008, + "step": 21592 + }, + { + "epoch": 2.3409583694709455, + "grad_norm": 0.033228788524866104, + "learning_rate": 4.409545916146815e-06, + "loss": 0.0004, + "step": 21593 + }, + { + "epoch": 2.341066782307025, + "grad_norm": 0.9593937993049622, + "learning_rate": 4.408820542579429e-06, + "loss": 0.0017, + "step": 21594 + }, + { + "epoch": 2.341175195143105, + "grad_norm": 4.847328186035156, + "learning_rate": 4.408095169012041e-06, + "loss": 0.0033, + "step": 21595 + }, + { + "epoch": 2.3412836079791846, + "grad_norm": 7.713535785675049, + "learning_rate": 4.407369795444654e-06, + "loss": 0.0279, + "step": 21596 + }, + { + "epoch": 2.3413920208152645, + "grad_norm": 0.42931443452835083, + "learning_rate": 4.406644421877267e-06, + "loss": 0.0024, + "step": 21597 + }, + { + "epoch": 2.341500433651344, + "grad_norm": 0.5630757808685303, + "learning_rate": 4.40591904830988e-06, + "loss": 0.0016, + "step": 21598 + }, + { + "epoch": 2.341608846487424, + "grad_norm": 0.05448989197611809, + "learning_rate": 4.4051936747424925e-06, + "loss": 0.0005, + "step": 21599 + }, + { + "epoch": 2.341717259323504, + "grad_norm": 0.1352747529745102, + "learning_rate": 4.4044683011751055e-06, + "loss": 0.0008, + "step": 21600 + }, + { + "epoch": 2.3418256721595836, + "grad_norm": 2.4252703189849854, + "learning_rate": 4.4037429276077184e-06, + "loss": 0.016, + "step": 21601 + }, + { + "epoch": 2.3419340849956636, + "grad_norm": 11.428044319152832, + "learning_rate": 4.4030175540403314e-06, + "loss": 0.0549, + "step": 21602 + }, + { + "epoch": 2.342042497831743, + "grad_norm": 0.25728607177734375, + "learning_rate": 4.4022921804729444e-06, + "loss": 0.0016, + "step": 21603 + }, + { + "epoch": 2.342150910667823, + "grad_norm": 0.0626920834183693, + "learning_rate": 4.4015668069055566e-06, + "loss": 0.0005, + "step": 21604 + }, + { + "epoch": 2.342259323503903, + "grad_norm": 9.737992286682129, + "learning_rate": 4.4008414333381696e-06, + "loss": 0.0055, + "step": 21605 + }, + { + "epoch": 2.3423677363399826, + "grad_norm": 2.0208613872528076, + "learning_rate": 4.400116059770782e-06, + "loss": 0.0295, + "step": 21606 + }, + { + "epoch": 2.3424761491760626, + "grad_norm": 0.7727150321006775, + "learning_rate": 4.3993906862033955e-06, + "loss": 0.0048, + "step": 21607 + }, + { + "epoch": 2.342584562012142, + "grad_norm": 0.8753328323364258, + "learning_rate": 4.398665312636008e-06, + "loss": 0.0027, + "step": 21608 + }, + { + "epoch": 2.342692974848222, + "grad_norm": 0.42143675684928894, + "learning_rate": 4.397939939068621e-06, + "loss": 0.0006, + "step": 21609 + }, + { + "epoch": 2.3428013876843017, + "grad_norm": 0.17321816086769104, + "learning_rate": 4.397214565501234e-06, + "loss": 0.0003, + "step": 21610 + }, + { + "epoch": 2.3429098005203817, + "grad_norm": 1.4220585823059082, + "learning_rate": 4.396489191933847e-06, + "loss": 0.0028, + "step": 21611 + }, + { + "epoch": 2.343018213356461, + "grad_norm": 0.09987538307905197, + "learning_rate": 4.395763818366459e-06, + "loss": 0.001, + "step": 21612 + }, + { + "epoch": 2.343126626192541, + "grad_norm": 0.7542576789855957, + "learning_rate": 4.395038444799072e-06, + "loss": 0.0018, + "step": 21613 + }, + { + "epoch": 2.343235039028621, + "grad_norm": 0.022785481065511703, + "learning_rate": 4.394313071231685e-06, + "loss": 0.0002, + "step": 21614 + }, + { + "epoch": 2.3433434518647007, + "grad_norm": 2.2387988567352295, + "learning_rate": 4.393587697664298e-06, + "loss": 0.014, + "step": 21615 + }, + { + "epoch": 2.3434518647007807, + "grad_norm": 2.1308796405792236, + "learning_rate": 4.392862324096911e-06, + "loss": 0.0049, + "step": 21616 + }, + { + "epoch": 2.3435602775368602, + "grad_norm": 0.03384166583418846, + "learning_rate": 4.392136950529523e-06, + "loss": 0.0002, + "step": 21617 + }, + { + "epoch": 2.3436686903729402, + "grad_norm": 0.19953016936779022, + "learning_rate": 4.391411576962136e-06, + "loss": 0.0005, + "step": 21618 + }, + { + "epoch": 2.3437771032090198, + "grad_norm": 0.9105194807052612, + "learning_rate": 4.390686203394748e-06, + "loss": 0.0045, + "step": 21619 + }, + { + "epoch": 2.3438855160450998, + "grad_norm": 3.765505313873291, + "learning_rate": 4.389960829827362e-06, + "loss": 0.0158, + "step": 21620 + }, + { + "epoch": 2.3439939288811793, + "grad_norm": 2.1181440353393555, + "learning_rate": 4.389235456259974e-06, + "loss": 0.0064, + "step": 21621 + }, + { + "epoch": 2.3441023417172593, + "grad_norm": 0.169633686542511, + "learning_rate": 4.388510082692587e-06, + "loss": 0.0008, + "step": 21622 + }, + { + "epoch": 2.3442107545533393, + "grad_norm": 3.8988256454467773, + "learning_rate": 4.3877847091252e-06, + "loss": 0.0164, + "step": 21623 + }, + { + "epoch": 2.344319167389419, + "grad_norm": 2.5818426609039307, + "learning_rate": 4.387059335557813e-06, + "loss": 0.029, + "step": 21624 + }, + { + "epoch": 2.344427580225499, + "grad_norm": 3.475830078125, + "learning_rate": 4.386333961990425e-06, + "loss": 0.027, + "step": 21625 + }, + { + "epoch": 2.3445359930615783, + "grad_norm": 0.05169922858476639, + "learning_rate": 4.385608588423038e-06, + "loss": 0.0006, + "step": 21626 + }, + { + "epoch": 2.3446444058976583, + "grad_norm": 0.12248076498508453, + "learning_rate": 4.384883214855651e-06, + "loss": 0.0003, + "step": 21627 + }, + { + "epoch": 2.3447528187337383, + "grad_norm": 5.390374183654785, + "learning_rate": 4.384157841288264e-06, + "loss": 0.0035, + "step": 21628 + }, + { + "epoch": 2.344861231569818, + "grad_norm": 0.5094797015190125, + "learning_rate": 4.383432467720877e-06, + "loss": 0.0019, + "step": 21629 + }, + { + "epoch": 2.344969644405898, + "grad_norm": 1.9591479301452637, + "learning_rate": 4.382707094153489e-06, + "loss": 0.0033, + "step": 21630 + }, + { + "epoch": 2.3450780572419774, + "grad_norm": 3.6813085079193115, + "learning_rate": 4.381981720586102e-06, + "loss": 0.0027, + "step": 21631 + }, + { + "epoch": 2.3451864700780574, + "grad_norm": 0.13927815854549408, + "learning_rate": 4.381256347018715e-06, + "loss": 0.001, + "step": 21632 + }, + { + "epoch": 2.345294882914137, + "grad_norm": 0.022743133828043938, + "learning_rate": 4.380530973451328e-06, + "loss": 0.0001, + "step": 21633 + }, + { + "epoch": 2.345403295750217, + "grad_norm": 0.08785196393728256, + "learning_rate": 4.37980559988394e-06, + "loss": 0.0004, + "step": 21634 + }, + { + "epoch": 2.3455117085862964, + "grad_norm": 0.04188396409153938, + "learning_rate": 4.379080226316553e-06, + "loss": 0.0002, + "step": 21635 + }, + { + "epoch": 2.3456201214223764, + "grad_norm": 1.4530216455459595, + "learning_rate": 4.378354852749166e-06, + "loss": 0.0515, + "step": 21636 + }, + { + "epoch": 2.3457285342584564, + "grad_norm": 0.09555575996637344, + "learning_rate": 4.377629479181779e-06, + "loss": 0.0001, + "step": 21637 + }, + { + "epoch": 2.345836947094536, + "grad_norm": 3.0713279247283936, + "learning_rate": 4.3769041056143915e-06, + "loss": 0.0228, + "step": 21638 + }, + { + "epoch": 2.345945359930616, + "grad_norm": 17.582916259765625, + "learning_rate": 4.3761787320470045e-06, + "loss": 0.0518, + "step": 21639 + }, + { + "epoch": 2.3460537727666955, + "grad_norm": 0.09327396750450134, + "learning_rate": 4.3754533584796175e-06, + "loss": 0.0005, + "step": 21640 + }, + { + "epoch": 2.3461621856027755, + "grad_norm": 11.658611297607422, + "learning_rate": 4.3747279849122304e-06, + "loss": 0.0347, + "step": 21641 + }, + { + "epoch": 2.346270598438855, + "grad_norm": 0.05887081101536751, + "learning_rate": 4.3740026113448434e-06, + "loss": 0.0002, + "step": 21642 + }, + { + "epoch": 2.346379011274935, + "grad_norm": 1.6147116422653198, + "learning_rate": 4.3732772377774556e-06, + "loss": 0.0384, + "step": 21643 + }, + { + "epoch": 2.3464874241110145, + "grad_norm": 3.983933925628662, + "learning_rate": 4.3725518642100686e-06, + "loss": 0.007, + "step": 21644 + }, + { + "epoch": 2.3465958369470945, + "grad_norm": 1.4478424787521362, + "learning_rate": 4.3718264906426816e-06, + "loss": 0.0323, + "step": 21645 + }, + { + "epoch": 2.3467042497831745, + "grad_norm": 0.0022546956315636635, + "learning_rate": 4.3711011170752945e-06, + "loss": 0.0, + "step": 21646 + }, + { + "epoch": 2.346812662619254, + "grad_norm": 0.10721028596162796, + "learning_rate": 4.370375743507907e-06, + "loss": 0.0005, + "step": 21647 + }, + { + "epoch": 2.346921075455334, + "grad_norm": 1.8165760040283203, + "learning_rate": 4.36965036994052e-06, + "loss": 0.0039, + "step": 21648 + }, + { + "epoch": 2.3470294882914136, + "grad_norm": 4.279516696929932, + "learning_rate": 4.368924996373133e-06, + "loss": 0.0296, + "step": 21649 + }, + { + "epoch": 2.3471379011274935, + "grad_norm": 7.484470844268799, + "learning_rate": 4.368199622805746e-06, + "loss": 0.0445, + "step": 21650 + }, + { + "epoch": 2.3472463139635735, + "grad_norm": 13.126590728759766, + "learning_rate": 4.367474249238358e-06, + "loss": 0.0202, + "step": 21651 + }, + { + "epoch": 2.347354726799653, + "grad_norm": 0.13917970657348633, + "learning_rate": 4.366748875670971e-06, + "loss": 0.0005, + "step": 21652 + }, + { + "epoch": 2.347463139635733, + "grad_norm": 2.6992759704589844, + "learning_rate": 4.366023502103584e-06, + "loss": 0.002, + "step": 21653 + }, + { + "epoch": 2.3475715524718126, + "grad_norm": 1.4389779567718506, + "learning_rate": 4.365298128536197e-06, + "loss": 0.0042, + "step": 21654 + }, + { + "epoch": 2.3476799653078926, + "grad_norm": 0.6294470429420471, + "learning_rate": 4.36457275496881e-06, + "loss": 0.0021, + "step": 21655 + }, + { + "epoch": 2.347788378143972, + "grad_norm": 1.2269600629806519, + "learning_rate": 4.363847381401422e-06, + "loss": 0.0148, + "step": 21656 + }, + { + "epoch": 2.347896790980052, + "grad_norm": 1.690416693687439, + "learning_rate": 4.363122007834035e-06, + "loss": 0.0074, + "step": 21657 + }, + { + "epoch": 2.3480052038161316, + "grad_norm": 4.917256832122803, + "learning_rate": 4.362396634266648e-06, + "loss": 0.012, + "step": 21658 + }, + { + "epoch": 2.3481136166522116, + "grad_norm": 0.16073870658874512, + "learning_rate": 4.361671260699261e-06, + "loss": 0.0006, + "step": 21659 + }, + { + "epoch": 2.3482220294882916, + "grad_norm": 4.982423305511475, + "learning_rate": 4.360945887131873e-06, + "loss": 0.0292, + "step": 21660 + }, + { + "epoch": 2.348330442324371, + "grad_norm": 0.4686235785484314, + "learning_rate": 4.360220513564486e-06, + "loss": 0.0018, + "step": 21661 + }, + { + "epoch": 2.348438855160451, + "grad_norm": 1.9749325513839722, + "learning_rate": 4.359495139997099e-06, + "loss": 0.0167, + "step": 21662 + }, + { + "epoch": 2.3485472679965307, + "grad_norm": 1.899329662322998, + "learning_rate": 4.358769766429712e-06, + "loss": 0.006, + "step": 21663 + }, + { + "epoch": 2.3486556808326107, + "grad_norm": 0.12698283791542053, + "learning_rate": 4.358044392862324e-06, + "loss": 0.0006, + "step": 21664 + }, + { + "epoch": 2.34876409366869, + "grad_norm": 0.07287024706602097, + "learning_rate": 4.357319019294937e-06, + "loss": 0.0004, + "step": 21665 + }, + { + "epoch": 2.34887250650477, + "grad_norm": 0.09876815229654312, + "learning_rate": 4.35659364572755e-06, + "loss": 0.0005, + "step": 21666 + }, + { + "epoch": 2.3489809193408497, + "grad_norm": 0.1250690072774887, + "learning_rate": 4.355868272160163e-06, + "loss": 0.0004, + "step": 21667 + }, + { + "epoch": 2.3490893321769297, + "grad_norm": 3.257241725921631, + "learning_rate": 4.355142898592776e-06, + "loss": 0.0322, + "step": 21668 + }, + { + "epoch": 2.3491977450130097, + "grad_norm": 4.646300792694092, + "learning_rate": 4.354417525025388e-06, + "loss": 0.0209, + "step": 21669 + }, + { + "epoch": 2.3493061578490892, + "grad_norm": 0.06718457490205765, + "learning_rate": 4.353692151458001e-06, + "loss": 0.0004, + "step": 21670 + }, + { + "epoch": 2.3494145706851692, + "grad_norm": 1.3998851776123047, + "learning_rate": 4.352966777890614e-06, + "loss": 0.0438, + "step": 21671 + }, + { + "epoch": 2.3495229835212488, + "grad_norm": 1.02028226852417, + "learning_rate": 4.352241404323227e-06, + "loss": 0.0237, + "step": 21672 + }, + { + "epoch": 2.3496313963573288, + "grad_norm": 0.06803061813116074, + "learning_rate": 4.351516030755839e-06, + "loss": 0.0005, + "step": 21673 + }, + { + "epoch": 2.3497398091934087, + "grad_norm": 0.021114278584718704, + "learning_rate": 4.350790657188452e-06, + "loss": 0.0002, + "step": 21674 + }, + { + "epoch": 2.3498482220294883, + "grad_norm": 0.113225556910038, + "learning_rate": 4.350065283621065e-06, + "loss": 0.0005, + "step": 21675 + }, + { + "epoch": 2.3499566348655683, + "grad_norm": 2.085616111755371, + "learning_rate": 4.349339910053678e-06, + "loss": 0.0058, + "step": 21676 + }, + { + "epoch": 2.350065047701648, + "grad_norm": 0.9061150550842285, + "learning_rate": 4.3486145364862905e-06, + "loss": 0.0039, + "step": 21677 + }, + { + "epoch": 2.350173460537728, + "grad_norm": 0.21377159655094147, + "learning_rate": 4.3478891629189035e-06, + "loss": 0.0015, + "step": 21678 + }, + { + "epoch": 2.3502818733738073, + "grad_norm": 0.6131760478019714, + "learning_rate": 4.3471637893515165e-06, + "loss": 0.0031, + "step": 21679 + }, + { + "epoch": 2.3503902862098873, + "grad_norm": 4.900815010070801, + "learning_rate": 4.3464384157841294e-06, + "loss": 0.0483, + "step": 21680 + }, + { + "epoch": 2.350498699045967, + "grad_norm": 6.904674053192139, + "learning_rate": 4.3457130422167424e-06, + "loss": 0.0138, + "step": 21681 + }, + { + "epoch": 2.350607111882047, + "grad_norm": 0.05466297268867493, + "learning_rate": 4.344987668649355e-06, + "loss": 0.0004, + "step": 21682 + }, + { + "epoch": 2.350715524718127, + "grad_norm": 6.335350513458252, + "learning_rate": 4.3442622950819676e-06, + "loss": 0.0553, + "step": 21683 + }, + { + "epoch": 2.3508239375542064, + "grad_norm": 2.035081624984741, + "learning_rate": 4.3435369215145806e-06, + "loss": 0.012, + "step": 21684 + }, + { + "epoch": 2.3509323503902864, + "grad_norm": 0.05578033998608589, + "learning_rate": 4.3428115479471935e-06, + "loss": 0.0004, + "step": 21685 + }, + { + "epoch": 2.351040763226366, + "grad_norm": 2.7415623664855957, + "learning_rate": 4.342086174379806e-06, + "loss": 0.0197, + "step": 21686 + }, + { + "epoch": 2.351149176062446, + "grad_norm": 0.051589831709861755, + "learning_rate": 4.341360800812419e-06, + "loss": 0.0002, + "step": 21687 + }, + { + "epoch": 2.3512575888985254, + "grad_norm": 5.79502010345459, + "learning_rate": 4.340635427245032e-06, + "loss": 0.0146, + "step": 21688 + }, + { + "epoch": 2.3513660017346054, + "grad_norm": 4.0731682777404785, + "learning_rate": 4.339910053677645e-06, + "loss": 0.0182, + "step": 21689 + }, + { + "epoch": 2.351474414570685, + "grad_norm": 0.05626203864812851, + "learning_rate": 4.339184680110257e-06, + "loss": 0.0005, + "step": 21690 + }, + { + "epoch": 2.351582827406765, + "grad_norm": 0.033968038856983185, + "learning_rate": 4.33845930654287e-06, + "loss": 0.0003, + "step": 21691 + }, + { + "epoch": 2.351691240242845, + "grad_norm": 0.5766446590423584, + "learning_rate": 4.337733932975483e-06, + "loss": 0.0007, + "step": 21692 + }, + { + "epoch": 2.3517996530789245, + "grad_norm": 0.34062471985816956, + "learning_rate": 4.337008559408096e-06, + "loss": 0.001, + "step": 21693 + }, + { + "epoch": 2.3519080659150045, + "grad_norm": 2.926469326019287, + "learning_rate": 4.336283185840709e-06, + "loss": 0.0221, + "step": 21694 + }, + { + "epoch": 2.352016478751084, + "grad_norm": 5.969334602355957, + "learning_rate": 4.335557812273321e-06, + "loss": 0.006, + "step": 21695 + }, + { + "epoch": 2.352124891587164, + "grad_norm": 0.32876503467559814, + "learning_rate": 4.334832438705934e-06, + "loss": 0.0007, + "step": 21696 + }, + { + "epoch": 2.3522333044232435, + "grad_norm": 1.727055311203003, + "learning_rate": 4.334107065138547e-06, + "loss": 0.0072, + "step": 21697 + }, + { + "epoch": 2.3523417172593235, + "grad_norm": 1.0016107559204102, + "learning_rate": 4.33338169157116e-06, + "loss": 0.0064, + "step": 21698 + }, + { + "epoch": 2.352450130095403, + "grad_norm": 4.749090194702148, + "learning_rate": 4.332656318003772e-06, + "loss": 0.0315, + "step": 21699 + }, + { + "epoch": 2.352558542931483, + "grad_norm": 3.063812494277954, + "learning_rate": 4.331930944436385e-06, + "loss": 0.0119, + "step": 21700 + }, + { + "epoch": 2.352666955767563, + "grad_norm": 3.163670778274536, + "learning_rate": 4.331205570868998e-06, + "loss": 0.0079, + "step": 21701 + }, + { + "epoch": 2.3527753686036426, + "grad_norm": 4.206569194793701, + "learning_rate": 4.330480197301611e-06, + "loss": 0.0178, + "step": 21702 + }, + { + "epoch": 2.3528837814397225, + "grad_norm": 1.788413643836975, + "learning_rate": 4.329754823734223e-06, + "loss": 0.0195, + "step": 21703 + }, + { + "epoch": 2.352992194275802, + "grad_norm": 0.7496726512908936, + "learning_rate": 4.329029450166836e-06, + "loss": 0.002, + "step": 21704 + }, + { + "epoch": 2.353100607111882, + "grad_norm": 0.3023413419723511, + "learning_rate": 4.328304076599449e-06, + "loss": 0.0007, + "step": 21705 + }, + { + "epoch": 2.353209019947962, + "grad_norm": 0.2885648310184479, + "learning_rate": 4.327578703032062e-06, + "loss": 0.0008, + "step": 21706 + }, + { + "epoch": 2.3533174327840416, + "grad_norm": 0.6946473717689514, + "learning_rate": 4.326853329464675e-06, + "loss": 0.0033, + "step": 21707 + }, + { + "epoch": 2.3534258456201216, + "grad_norm": 2.2257819175720215, + "learning_rate": 4.326127955897287e-06, + "loss": 0.0034, + "step": 21708 + }, + { + "epoch": 2.353534258456201, + "grad_norm": 1.4631285667419434, + "learning_rate": 4.3254025823299e-06, + "loss": 0.0029, + "step": 21709 + }, + { + "epoch": 2.353642671292281, + "grad_norm": 0.02658224105834961, + "learning_rate": 4.324677208762513e-06, + "loss": 0.0001, + "step": 21710 + }, + { + "epoch": 2.3537510841283606, + "grad_norm": 1.5380189418792725, + "learning_rate": 4.323951835195126e-06, + "loss": 0.0034, + "step": 21711 + }, + { + "epoch": 2.3538594969644406, + "grad_norm": 1.4887372255325317, + "learning_rate": 4.323226461627738e-06, + "loss": 0.006, + "step": 21712 + }, + { + "epoch": 2.35396790980052, + "grad_norm": 19.515871047973633, + "learning_rate": 4.322501088060351e-06, + "loss": 0.0262, + "step": 21713 + }, + { + "epoch": 2.3540763226366, + "grad_norm": 0.3993988335132599, + "learning_rate": 4.321775714492964e-06, + "loss": 0.001, + "step": 21714 + }, + { + "epoch": 2.35418473547268, + "grad_norm": 1.9203839302062988, + "learning_rate": 4.321050340925577e-06, + "loss": 0.0105, + "step": 21715 + }, + { + "epoch": 2.3542931483087597, + "grad_norm": 3.9148051738739014, + "learning_rate": 4.3203249673581895e-06, + "loss": 0.0037, + "step": 21716 + }, + { + "epoch": 2.3544015611448397, + "grad_norm": 5.526285171508789, + "learning_rate": 4.3195995937908025e-06, + "loss": 0.0187, + "step": 21717 + }, + { + "epoch": 2.354509973980919, + "grad_norm": 0.2588311731815338, + "learning_rate": 4.3188742202234155e-06, + "loss": 0.0015, + "step": 21718 + }, + { + "epoch": 2.354618386816999, + "grad_norm": 13.528853416442871, + "learning_rate": 4.3181488466560284e-06, + "loss": 0.0299, + "step": 21719 + }, + { + "epoch": 2.3547267996530787, + "grad_norm": 3.647296905517578, + "learning_rate": 4.3174234730886414e-06, + "loss": 0.0043, + "step": 21720 + }, + { + "epoch": 2.3548352124891587, + "grad_norm": 2.8802905082702637, + "learning_rate": 4.316698099521254e-06, + "loss": 0.0502, + "step": 21721 + }, + { + "epoch": 2.3549436253252383, + "grad_norm": 4.653160572052002, + "learning_rate": 4.3159727259538666e-06, + "loss": 0.009, + "step": 21722 + }, + { + "epoch": 2.3550520381613183, + "grad_norm": 0.010515337809920311, + "learning_rate": 4.3152473523864796e-06, + "loss": 0.0001, + "step": 21723 + }, + { + "epoch": 2.3551604509973982, + "grad_norm": 4.084019660949707, + "learning_rate": 4.3145219788190926e-06, + "loss": 0.0152, + "step": 21724 + }, + { + "epoch": 2.3552688638334778, + "grad_norm": 4.1479010581970215, + "learning_rate": 4.313796605251705e-06, + "loss": 0.0073, + "step": 21725 + }, + { + "epoch": 2.3553772766695578, + "grad_norm": 0.28704339265823364, + "learning_rate": 4.313071231684318e-06, + "loss": 0.0014, + "step": 21726 + }, + { + "epoch": 2.3554856895056373, + "grad_norm": 1.067956805229187, + "learning_rate": 4.312345858116931e-06, + "loss": 0.0041, + "step": 21727 + }, + { + "epoch": 2.3555941023417173, + "grad_norm": 2.493360757827759, + "learning_rate": 4.311620484549544e-06, + "loss": 0.0049, + "step": 21728 + }, + { + "epoch": 2.3557025151777973, + "grad_norm": 1.944749116897583, + "learning_rate": 4.310895110982156e-06, + "loss": 0.0209, + "step": 21729 + }, + { + "epoch": 2.355810928013877, + "grad_norm": 0.10775413364171982, + "learning_rate": 4.310169737414769e-06, + "loss": 0.0007, + "step": 21730 + }, + { + "epoch": 2.355919340849957, + "grad_norm": 3.7496211528778076, + "learning_rate": 4.309444363847382e-06, + "loss": 0.0169, + "step": 21731 + }, + { + "epoch": 2.3560277536860363, + "grad_norm": 0.13566096127033234, + "learning_rate": 4.308718990279995e-06, + "loss": 0.0006, + "step": 21732 + }, + { + "epoch": 2.3561361665221163, + "grad_norm": 0.40435144305229187, + "learning_rate": 4.307993616712608e-06, + "loss": 0.0017, + "step": 21733 + }, + { + "epoch": 2.356244579358196, + "grad_norm": 0.016795683652162552, + "learning_rate": 4.30726824314522e-06, + "loss": 0.0001, + "step": 21734 + }, + { + "epoch": 2.356352992194276, + "grad_norm": 2.1910006999969482, + "learning_rate": 4.306542869577833e-06, + "loss": 0.0056, + "step": 21735 + }, + { + "epoch": 2.3564614050303554, + "grad_norm": 4.7481513023376465, + "learning_rate": 4.305817496010446e-06, + "loss": 0.0153, + "step": 21736 + }, + { + "epoch": 2.3565698178664354, + "grad_norm": 1.4728420972824097, + "learning_rate": 4.305092122443059e-06, + "loss": 0.0351, + "step": 21737 + }, + { + "epoch": 2.3566782307025154, + "grad_norm": 3.381423234939575, + "learning_rate": 4.304366748875671e-06, + "loss": 0.0099, + "step": 21738 + }, + { + "epoch": 2.356786643538595, + "grad_norm": 3.1400039196014404, + "learning_rate": 4.303641375308284e-06, + "loss": 0.0156, + "step": 21739 + }, + { + "epoch": 2.356895056374675, + "grad_norm": 8.08078384399414, + "learning_rate": 4.302916001740897e-06, + "loss": 0.0285, + "step": 21740 + }, + { + "epoch": 2.3570034692107544, + "grad_norm": 11.925808906555176, + "learning_rate": 4.30219062817351e-06, + "loss": 0.0083, + "step": 21741 + }, + { + "epoch": 2.3571118820468344, + "grad_norm": 0.05684764310717583, + "learning_rate": 4.301465254606122e-06, + "loss": 0.0004, + "step": 21742 + }, + { + "epoch": 2.357220294882914, + "grad_norm": 0.1891576498746872, + "learning_rate": 4.300739881038735e-06, + "loss": 0.0006, + "step": 21743 + }, + { + "epoch": 2.357328707718994, + "grad_norm": 3.319749593734741, + "learning_rate": 4.300014507471348e-06, + "loss": 0.007, + "step": 21744 + }, + { + "epoch": 2.3574371205550735, + "grad_norm": 1.710488200187683, + "learning_rate": 4.299289133903961e-06, + "loss": 0.0629, + "step": 21745 + }, + { + "epoch": 2.3575455333911535, + "grad_norm": 5.135892868041992, + "learning_rate": 4.298563760336574e-06, + "loss": 0.0133, + "step": 21746 + }, + { + "epoch": 2.3576539462272335, + "grad_norm": 0.12041378766298294, + "learning_rate": 4.297838386769186e-06, + "loss": 0.0006, + "step": 21747 + }, + { + "epoch": 2.357762359063313, + "grad_norm": 2.5794382095336914, + "learning_rate": 4.297113013201799e-06, + "loss": 0.0237, + "step": 21748 + }, + { + "epoch": 2.357870771899393, + "grad_norm": 9.058320999145508, + "learning_rate": 4.296387639634412e-06, + "loss": 0.0166, + "step": 21749 + }, + { + "epoch": 2.3579791847354725, + "grad_norm": 1.6487513780593872, + "learning_rate": 4.295662266067025e-06, + "loss": 0.0019, + "step": 21750 + }, + { + "epoch": 2.3580875975715525, + "grad_norm": 0.157064288854599, + "learning_rate": 4.294936892499637e-06, + "loss": 0.0004, + "step": 21751 + }, + { + "epoch": 2.3581960104076325, + "grad_norm": 0.1953285187482834, + "learning_rate": 4.29421151893225e-06, + "loss": 0.0008, + "step": 21752 + }, + { + "epoch": 2.358304423243712, + "grad_norm": 1.9746966361999512, + "learning_rate": 4.293486145364863e-06, + "loss": 0.0293, + "step": 21753 + }, + { + "epoch": 2.358412836079792, + "grad_norm": 0.29612836241722107, + "learning_rate": 4.292760771797476e-06, + "loss": 0.0014, + "step": 21754 + }, + { + "epoch": 2.3585212489158716, + "grad_norm": 1.0392979383468628, + "learning_rate": 4.2920353982300885e-06, + "loss": 0.0063, + "step": 21755 + }, + { + "epoch": 2.3586296617519515, + "grad_norm": 1.682081937789917, + "learning_rate": 4.2913100246627015e-06, + "loss": 0.005, + "step": 21756 + }, + { + "epoch": 2.358738074588031, + "grad_norm": 5.143861770629883, + "learning_rate": 4.2905846510953145e-06, + "loss": 0.0225, + "step": 21757 + }, + { + "epoch": 2.358846487424111, + "grad_norm": 8.531697273254395, + "learning_rate": 4.2898592775279275e-06, + "loss": 0.0367, + "step": 21758 + }, + { + "epoch": 2.3589549002601906, + "grad_norm": 0.1490706205368042, + "learning_rate": 4.2891339039605404e-06, + "loss": 0.0008, + "step": 21759 + }, + { + "epoch": 2.3590633130962706, + "grad_norm": 0.22868500649929047, + "learning_rate": 4.288408530393153e-06, + "loss": 0.0006, + "step": 21760 + }, + { + "epoch": 2.3591717259323506, + "grad_norm": 0.02671905979514122, + "learning_rate": 4.2876831568257656e-06, + "loss": 0.0002, + "step": 21761 + }, + { + "epoch": 2.35928013876843, + "grad_norm": 0.014604943804442883, + "learning_rate": 4.2869577832583786e-06, + "loss": 0.0001, + "step": 21762 + }, + { + "epoch": 2.35938855160451, + "grad_norm": 1.6381807327270508, + "learning_rate": 4.2862324096909916e-06, + "loss": 0.0047, + "step": 21763 + }, + { + "epoch": 2.3594969644405897, + "grad_norm": 1.4233412742614746, + "learning_rate": 4.285507036123604e-06, + "loss": 0.0044, + "step": 21764 + }, + { + "epoch": 2.3596053772766696, + "grad_norm": 11.884438514709473, + "learning_rate": 4.284781662556217e-06, + "loss": 0.0177, + "step": 21765 + }, + { + "epoch": 2.359713790112749, + "grad_norm": 0.4417029321193695, + "learning_rate": 4.28405628898883e-06, + "loss": 0.0011, + "step": 21766 + }, + { + "epoch": 2.359822202948829, + "grad_norm": 4.998505115509033, + "learning_rate": 4.283330915421443e-06, + "loss": 0.0475, + "step": 21767 + }, + { + "epoch": 2.3599306157849087, + "grad_norm": 0.07567286491394043, + "learning_rate": 4.282605541854055e-06, + "loss": 0.0004, + "step": 21768 + }, + { + "epoch": 2.3600390286209887, + "grad_norm": 3.020765542984009, + "learning_rate": 4.281880168286668e-06, + "loss": 0.0037, + "step": 21769 + }, + { + "epoch": 2.3601474414570687, + "grad_norm": 0.013228720054030418, + "learning_rate": 4.281154794719281e-06, + "loss": 0.0001, + "step": 21770 + }, + { + "epoch": 2.360255854293148, + "grad_norm": 0.9064100980758667, + "learning_rate": 4.280429421151894e-06, + "loss": 0.0028, + "step": 21771 + }, + { + "epoch": 2.360364267129228, + "grad_norm": 0.24516561627388, + "learning_rate": 4.279704047584507e-06, + "loss": 0.0012, + "step": 21772 + }, + { + "epoch": 2.3604726799653077, + "grad_norm": 1.0529534816741943, + "learning_rate": 4.278978674017119e-06, + "loss": 0.0052, + "step": 21773 + }, + { + "epoch": 2.3605810928013877, + "grad_norm": 6.174117565155029, + "learning_rate": 4.278253300449732e-06, + "loss": 0.0817, + "step": 21774 + }, + { + "epoch": 2.3606895056374677, + "grad_norm": 0.04268565773963928, + "learning_rate": 4.277527926882345e-06, + "loss": 0.0003, + "step": 21775 + }, + { + "epoch": 2.3607979184735473, + "grad_norm": 6.316574573516846, + "learning_rate": 4.276802553314958e-06, + "loss": 0.0501, + "step": 21776 + }, + { + "epoch": 2.3609063313096272, + "grad_norm": 0.0251736119389534, + "learning_rate": 4.27607717974757e-06, + "loss": 0.0002, + "step": 21777 + }, + { + "epoch": 2.361014744145707, + "grad_norm": 0.08733291178941727, + "learning_rate": 4.275351806180183e-06, + "loss": 0.0003, + "step": 21778 + }, + { + "epoch": 2.3611231569817868, + "grad_norm": 2.080810785293579, + "learning_rate": 4.274626432612796e-06, + "loss": 0.0063, + "step": 21779 + }, + { + "epoch": 2.3612315698178663, + "grad_norm": 1.6434717178344727, + "learning_rate": 4.273901059045409e-06, + "loss": 0.0017, + "step": 21780 + }, + { + "epoch": 2.3613399826539463, + "grad_norm": 38.139488220214844, + "learning_rate": 4.273175685478021e-06, + "loss": 0.0365, + "step": 21781 + }, + { + "epoch": 2.361448395490026, + "grad_norm": 3.167226791381836, + "learning_rate": 4.272450311910634e-06, + "loss": 0.0605, + "step": 21782 + }, + { + "epoch": 2.361556808326106, + "grad_norm": 13.033670425415039, + "learning_rate": 4.271724938343247e-06, + "loss": 0.0104, + "step": 21783 + }, + { + "epoch": 2.361665221162186, + "grad_norm": 0.5909519791603088, + "learning_rate": 4.27099956477586e-06, + "loss": 0.0024, + "step": 21784 + }, + { + "epoch": 2.3617736339982653, + "grad_norm": 0.0062330602668225765, + "learning_rate": 4.270274191208473e-06, + "loss": 0.0001, + "step": 21785 + }, + { + "epoch": 2.3618820468343453, + "grad_norm": 1.927908182144165, + "learning_rate": 4.269548817641085e-06, + "loss": 0.0306, + "step": 21786 + }, + { + "epoch": 2.361990459670425, + "grad_norm": 1.3637434244155884, + "learning_rate": 4.268823444073698e-06, + "loss": 0.0032, + "step": 21787 + }, + { + "epoch": 2.362098872506505, + "grad_norm": 0.30328190326690674, + "learning_rate": 4.268098070506311e-06, + "loss": 0.0011, + "step": 21788 + }, + { + "epoch": 2.3622072853425844, + "grad_norm": 0.04225486144423485, + "learning_rate": 4.267372696938924e-06, + "loss": 0.0002, + "step": 21789 + }, + { + "epoch": 2.3623156981786644, + "grad_norm": 11.492191314697266, + "learning_rate": 4.266647323371536e-06, + "loss": 0.0186, + "step": 21790 + }, + { + "epoch": 2.362424111014744, + "grad_norm": 0.3557821810245514, + "learning_rate": 4.265921949804149e-06, + "loss": 0.0005, + "step": 21791 + }, + { + "epoch": 2.362532523850824, + "grad_norm": 0.0783335492014885, + "learning_rate": 4.265196576236762e-06, + "loss": 0.0004, + "step": 21792 + }, + { + "epoch": 2.362640936686904, + "grad_norm": 37.82839584350586, + "learning_rate": 4.264471202669375e-06, + "loss": 0.0393, + "step": 21793 + }, + { + "epoch": 2.3627493495229834, + "grad_norm": 0.02883327007293701, + "learning_rate": 4.2637458291019875e-06, + "loss": 0.0002, + "step": 21794 + }, + { + "epoch": 2.3628577623590634, + "grad_norm": 2.496661424636841, + "learning_rate": 4.2630204555346005e-06, + "loss": 0.0184, + "step": 21795 + }, + { + "epoch": 2.362966175195143, + "grad_norm": 6.492043495178223, + "learning_rate": 4.2622950819672135e-06, + "loss": 0.0793, + "step": 21796 + }, + { + "epoch": 2.363074588031223, + "grad_norm": 1.367510437965393, + "learning_rate": 4.2615697083998265e-06, + "loss": 0.008, + "step": 21797 + }, + { + "epoch": 2.363183000867303, + "grad_norm": 0.1380883753299713, + "learning_rate": 4.2608443348324394e-06, + "loss": 0.0007, + "step": 21798 + }, + { + "epoch": 2.3632914137033825, + "grad_norm": 0.174421027302742, + "learning_rate": 4.260118961265052e-06, + "loss": 0.0012, + "step": 21799 + }, + { + "epoch": 2.3633998265394625, + "grad_norm": 0.1230727955698967, + "learning_rate": 4.259393587697665e-06, + "loss": 0.0009, + "step": 21800 + }, + { + "epoch": 2.363508239375542, + "grad_norm": 0.06916074454784393, + "learning_rate": 4.2586682141302776e-06, + "loss": 0.0004, + "step": 21801 + }, + { + "epoch": 2.363616652211622, + "grad_norm": 5.878494739532471, + "learning_rate": 4.2579428405628906e-06, + "loss": 0.0252, + "step": 21802 + }, + { + "epoch": 2.3637250650477015, + "grad_norm": 0.028066974133253098, + "learning_rate": 4.257217466995503e-06, + "loss": 0.0002, + "step": 21803 + }, + { + "epoch": 2.3638334778837815, + "grad_norm": 0.016132520511746407, + "learning_rate": 4.256492093428116e-06, + "loss": 0.0001, + "step": 21804 + }, + { + "epoch": 2.363941890719861, + "grad_norm": 3.8576440811157227, + "learning_rate": 4.255766719860729e-06, + "loss": 0.0399, + "step": 21805 + }, + { + "epoch": 2.364050303555941, + "grad_norm": 0.2318040132522583, + "learning_rate": 4.255041346293342e-06, + "loss": 0.0007, + "step": 21806 + }, + { + "epoch": 2.364158716392021, + "grad_norm": 1.4298713207244873, + "learning_rate": 4.254315972725954e-06, + "loss": 0.0034, + "step": 21807 + }, + { + "epoch": 2.3642671292281006, + "grad_norm": 1.3741235733032227, + "learning_rate": 4.253590599158567e-06, + "loss": 0.0215, + "step": 21808 + }, + { + "epoch": 2.3643755420641805, + "grad_norm": 0.13614217936992645, + "learning_rate": 4.25286522559118e-06, + "loss": 0.0006, + "step": 21809 + }, + { + "epoch": 2.36448395490026, + "grad_norm": 1.5490003824234009, + "learning_rate": 4.252139852023793e-06, + "loss": 0.0029, + "step": 21810 + }, + { + "epoch": 2.36459236773634, + "grad_norm": 2.8566389083862305, + "learning_rate": 4.251414478456406e-06, + "loss": 0.0273, + "step": 21811 + }, + { + "epoch": 2.3647007805724196, + "grad_norm": 1.2763477563858032, + "learning_rate": 4.250689104889018e-06, + "loss": 0.003, + "step": 21812 + }, + { + "epoch": 2.3648091934084996, + "grad_norm": 0.32563209533691406, + "learning_rate": 4.249963731321631e-06, + "loss": 0.0008, + "step": 21813 + }, + { + "epoch": 2.364917606244579, + "grad_norm": 3.515479564666748, + "learning_rate": 4.249238357754244e-06, + "loss": 0.0201, + "step": 21814 + }, + { + "epoch": 2.365026019080659, + "grad_norm": 0.6193768978118896, + "learning_rate": 4.248512984186857e-06, + "loss": 0.0031, + "step": 21815 + }, + { + "epoch": 2.365134431916739, + "grad_norm": 12.585679054260254, + "learning_rate": 4.247787610619469e-06, + "loss": 0.0453, + "step": 21816 + }, + { + "epoch": 2.3652428447528187, + "grad_norm": 2.195559024810791, + "learning_rate": 4.247062237052082e-06, + "loss": 0.0235, + "step": 21817 + }, + { + "epoch": 2.3653512575888986, + "grad_norm": 8.545373916625977, + "learning_rate": 4.246336863484695e-06, + "loss": 0.0232, + "step": 21818 + }, + { + "epoch": 2.365459670424978, + "grad_norm": 8.552213668823242, + "learning_rate": 4.245611489917308e-06, + "loss": 0.0668, + "step": 21819 + }, + { + "epoch": 2.365568083261058, + "grad_norm": 11.46872615814209, + "learning_rate": 4.24488611634992e-06, + "loss": 0.0174, + "step": 21820 + }, + { + "epoch": 2.365676496097138, + "grad_norm": 4.746603965759277, + "learning_rate": 4.244160742782533e-06, + "loss": 0.0322, + "step": 21821 + }, + { + "epoch": 2.3657849089332177, + "grad_norm": 0.011741413734853268, + "learning_rate": 4.243435369215146e-06, + "loss": 0.0001, + "step": 21822 + }, + { + "epoch": 2.3658933217692977, + "grad_norm": 1.6493834257125854, + "learning_rate": 4.242709995647759e-06, + "loss": 0.0043, + "step": 21823 + }, + { + "epoch": 2.366001734605377, + "grad_norm": 2.058330774307251, + "learning_rate": 4.241984622080372e-06, + "loss": 0.0035, + "step": 21824 + }, + { + "epoch": 2.366110147441457, + "grad_norm": 0.07568643242120743, + "learning_rate": 4.241259248512984e-06, + "loss": 0.0005, + "step": 21825 + }, + { + "epoch": 2.3662185602775367, + "grad_norm": 0.007584450766444206, + "learning_rate": 4.240533874945597e-06, + "loss": 0.0001, + "step": 21826 + }, + { + "epoch": 2.3663269731136167, + "grad_norm": 0.13639222085475922, + "learning_rate": 4.23980850137821e-06, + "loss": 0.0009, + "step": 21827 + }, + { + "epoch": 2.3664353859496963, + "grad_norm": 0.4082978665828705, + "learning_rate": 4.239083127810823e-06, + "loss": 0.0022, + "step": 21828 + }, + { + "epoch": 2.3665437987857763, + "grad_norm": 0.11177975684404373, + "learning_rate": 4.238357754243435e-06, + "loss": 0.0007, + "step": 21829 + }, + { + "epoch": 2.3666522116218562, + "grad_norm": 0.025906022638082504, + "learning_rate": 4.237632380676048e-06, + "loss": 0.0001, + "step": 21830 + }, + { + "epoch": 2.366760624457936, + "grad_norm": 1.1117876768112183, + "learning_rate": 4.236907007108661e-06, + "loss": 0.0045, + "step": 21831 + }, + { + "epoch": 2.3668690372940158, + "grad_norm": 0.14732331037521362, + "learning_rate": 4.236181633541274e-06, + "loss": 0.0005, + "step": 21832 + }, + { + "epoch": 2.3669774501300953, + "grad_norm": 2.2806363105773926, + "learning_rate": 4.2354562599738865e-06, + "loss": 0.0183, + "step": 21833 + }, + { + "epoch": 2.3670858629661753, + "grad_norm": 0.26758670806884766, + "learning_rate": 4.2347308864064995e-06, + "loss": 0.0017, + "step": 21834 + }, + { + "epoch": 2.367194275802255, + "grad_norm": 1.0334872007369995, + "learning_rate": 4.2340055128391125e-06, + "loss": 0.004, + "step": 21835 + }, + { + "epoch": 2.367302688638335, + "grad_norm": 8.232808113098145, + "learning_rate": 4.2332801392717255e-06, + "loss": 0.0077, + "step": 21836 + }, + { + "epoch": 2.3674111014744144, + "grad_norm": 0.6652202010154724, + "learning_rate": 4.2325547657043384e-06, + "loss": 0.0045, + "step": 21837 + }, + { + "epoch": 2.3675195143104943, + "grad_norm": 0.9154955744743347, + "learning_rate": 4.231829392136951e-06, + "loss": 0.0024, + "step": 21838 + }, + { + "epoch": 2.3676279271465743, + "grad_norm": 0.2213709056377411, + "learning_rate": 4.231104018569564e-06, + "loss": 0.0005, + "step": 21839 + }, + { + "epoch": 2.367736339982654, + "grad_norm": 0.9200167655944824, + "learning_rate": 4.2303786450021766e-06, + "loss": 0.0465, + "step": 21840 + }, + { + "epoch": 2.367844752818734, + "grad_norm": 1.6059248447418213, + "learning_rate": 4.2296532714347896e-06, + "loss": 0.0038, + "step": 21841 + }, + { + "epoch": 2.3679531656548134, + "grad_norm": 11.944820404052734, + "learning_rate": 4.228927897867402e-06, + "loss": 0.0075, + "step": 21842 + }, + { + "epoch": 2.3680615784908934, + "grad_norm": 1.155167579650879, + "learning_rate": 4.228202524300015e-06, + "loss": 0.0022, + "step": 21843 + }, + { + "epoch": 2.368169991326973, + "grad_norm": 0.6778224110603333, + "learning_rate": 4.227477150732628e-06, + "loss": 0.0027, + "step": 21844 + }, + { + "epoch": 2.368278404163053, + "grad_norm": 0.12824110686779022, + "learning_rate": 4.226751777165241e-06, + "loss": 0.0005, + "step": 21845 + }, + { + "epoch": 2.3683868169991324, + "grad_norm": 0.9144092798233032, + "learning_rate": 4.226026403597853e-06, + "loss": 0.0026, + "step": 21846 + }, + { + "epoch": 2.3684952298352124, + "grad_norm": 0.10090793669223785, + "learning_rate": 4.225301030030466e-06, + "loss": 0.0006, + "step": 21847 + }, + { + "epoch": 2.3686036426712924, + "grad_norm": 0.40130674839019775, + "learning_rate": 4.224575656463079e-06, + "loss": 0.0009, + "step": 21848 + }, + { + "epoch": 2.368712055507372, + "grad_norm": 5.86325216293335, + "learning_rate": 4.223850282895692e-06, + "loss": 0.0084, + "step": 21849 + }, + { + "epoch": 2.368820468343452, + "grad_norm": 7.258860111236572, + "learning_rate": 4.223124909328305e-06, + "loss": 0.0544, + "step": 21850 + }, + { + "epoch": 2.3689288811795315, + "grad_norm": 0.062406767159700394, + "learning_rate": 4.222399535760917e-06, + "loss": 0.0005, + "step": 21851 + }, + { + "epoch": 2.3690372940156115, + "grad_norm": 0.037793874740600586, + "learning_rate": 4.22167416219353e-06, + "loss": 0.0005, + "step": 21852 + }, + { + "epoch": 2.3691457068516915, + "grad_norm": 1.8372725248336792, + "learning_rate": 4.220948788626143e-06, + "loss": 0.0008, + "step": 21853 + }, + { + "epoch": 2.369254119687771, + "grad_norm": 8.862432479858398, + "learning_rate": 4.220223415058756e-06, + "loss": 0.0327, + "step": 21854 + }, + { + "epoch": 2.369362532523851, + "grad_norm": 0.28019288182258606, + "learning_rate": 4.219498041491368e-06, + "loss": 0.0016, + "step": 21855 + }, + { + "epoch": 2.3694709453599305, + "grad_norm": 0.02211993932723999, + "learning_rate": 4.218772667923981e-06, + "loss": 0.0002, + "step": 21856 + }, + { + "epoch": 2.3695793581960105, + "grad_norm": 0.20006002485752106, + "learning_rate": 4.218047294356594e-06, + "loss": 0.0013, + "step": 21857 + }, + { + "epoch": 2.36968777103209, + "grad_norm": 0.1465035378932953, + "learning_rate": 4.217321920789207e-06, + "loss": 0.0007, + "step": 21858 + }, + { + "epoch": 2.36979618386817, + "grad_norm": 0.11053509265184402, + "learning_rate": 4.216596547221819e-06, + "loss": 0.0005, + "step": 21859 + }, + { + "epoch": 2.3699045967042496, + "grad_norm": 0.08536771684885025, + "learning_rate": 4.215871173654432e-06, + "loss": 0.0003, + "step": 21860 + }, + { + "epoch": 2.3700130095403296, + "grad_norm": 8.62662124633789, + "learning_rate": 4.215145800087045e-06, + "loss": 0.0152, + "step": 21861 + }, + { + "epoch": 2.3701214223764095, + "grad_norm": 0.06700047105550766, + "learning_rate": 4.214420426519658e-06, + "loss": 0.0009, + "step": 21862 + }, + { + "epoch": 2.370229835212489, + "grad_norm": 2.1697089672088623, + "learning_rate": 4.213695052952271e-06, + "loss": 0.0073, + "step": 21863 + }, + { + "epoch": 2.370338248048569, + "grad_norm": 3.6746304035186768, + "learning_rate": 4.212969679384883e-06, + "loss": 0.019, + "step": 21864 + }, + { + "epoch": 2.3704466608846486, + "grad_norm": 0.12788090109825134, + "learning_rate": 4.212244305817496e-06, + "loss": 0.0006, + "step": 21865 + }, + { + "epoch": 2.3705550737207286, + "grad_norm": 0.3249266743659973, + "learning_rate": 4.211518932250109e-06, + "loss": 0.0015, + "step": 21866 + }, + { + "epoch": 2.370663486556808, + "grad_norm": 1.7202950716018677, + "learning_rate": 4.210793558682722e-06, + "loss": 0.0026, + "step": 21867 + }, + { + "epoch": 2.370771899392888, + "grad_norm": 3.2426891326904297, + "learning_rate": 4.210068185115334e-06, + "loss": 0.0212, + "step": 21868 + }, + { + "epoch": 2.3708803122289677, + "grad_norm": 0.04822879657149315, + "learning_rate": 4.209342811547947e-06, + "loss": 0.0004, + "step": 21869 + }, + { + "epoch": 2.3709887250650477, + "grad_norm": 4.542490005493164, + "learning_rate": 4.20861743798056e-06, + "loss": 0.0063, + "step": 21870 + }, + { + "epoch": 2.3710971379011276, + "grad_norm": 0.09724685549736023, + "learning_rate": 4.207892064413173e-06, + "loss": 0.0003, + "step": 21871 + }, + { + "epoch": 2.371205550737207, + "grad_norm": 8.653314590454102, + "learning_rate": 4.2071666908457855e-06, + "loss": 0.048, + "step": 21872 + }, + { + "epoch": 2.371313963573287, + "grad_norm": 1.0788649320602417, + "learning_rate": 4.2064413172783985e-06, + "loss": 0.0027, + "step": 21873 + }, + { + "epoch": 2.3714223764093667, + "grad_norm": 0.02463807538151741, + "learning_rate": 4.2057159437110115e-06, + "loss": 0.0002, + "step": 21874 + }, + { + "epoch": 2.3715307892454467, + "grad_norm": 8.994820594787598, + "learning_rate": 4.2049905701436245e-06, + "loss": 0.0198, + "step": 21875 + }, + { + "epoch": 2.3716392020815267, + "grad_norm": 0.11843205988407135, + "learning_rate": 4.2042651965762375e-06, + "loss": 0.0007, + "step": 21876 + }, + { + "epoch": 2.371747614917606, + "grad_norm": 4.220954418182373, + "learning_rate": 4.20353982300885e-06, + "loss": 0.0108, + "step": 21877 + }, + { + "epoch": 2.371856027753686, + "grad_norm": 0.007927275262773037, + "learning_rate": 4.202814449441463e-06, + "loss": 0.0001, + "step": 21878 + }, + { + "epoch": 2.3719644405897657, + "grad_norm": 0.09382645785808563, + "learning_rate": 4.2020890758740756e-06, + "loss": 0.0009, + "step": 21879 + }, + { + "epoch": 2.3720728534258457, + "grad_norm": 3.683749198913574, + "learning_rate": 4.2013637023066886e-06, + "loss": 0.0086, + "step": 21880 + }, + { + "epoch": 2.3721812662619253, + "grad_norm": 0.014516834169626236, + "learning_rate": 4.200638328739301e-06, + "loss": 0.0002, + "step": 21881 + }, + { + "epoch": 2.3722896790980053, + "grad_norm": 0.3442015051841736, + "learning_rate": 4.199912955171914e-06, + "loss": 0.0024, + "step": 21882 + }, + { + "epoch": 2.372398091934085, + "grad_norm": 3.6923205852508545, + "learning_rate": 4.199187581604527e-06, + "loss": 0.0193, + "step": 21883 + }, + { + "epoch": 2.372506504770165, + "grad_norm": 3.282733917236328, + "learning_rate": 4.19846220803714e-06, + "loss": 0.0062, + "step": 21884 + }, + { + "epoch": 2.3726149176062448, + "grad_norm": 4.357016086578369, + "learning_rate": 4.197736834469752e-06, + "loss": 0.0315, + "step": 21885 + }, + { + "epoch": 2.3727233304423243, + "grad_norm": 0.39291995763778687, + "learning_rate": 4.197011460902365e-06, + "loss": 0.002, + "step": 21886 + }, + { + "epoch": 2.3728317432784043, + "grad_norm": 13.741382598876953, + "learning_rate": 4.196286087334978e-06, + "loss": 0.0161, + "step": 21887 + }, + { + "epoch": 2.372940156114484, + "grad_norm": 3.0063276290893555, + "learning_rate": 4.195560713767591e-06, + "loss": 0.0028, + "step": 21888 + }, + { + "epoch": 2.373048568950564, + "grad_norm": 0.008275742642581463, + "learning_rate": 4.194835340200204e-06, + "loss": 0.0001, + "step": 21889 + }, + { + "epoch": 2.3731569817866434, + "grad_norm": 0.12195456773042679, + "learning_rate": 4.194109966632816e-06, + "loss": 0.0003, + "step": 21890 + }, + { + "epoch": 2.3732653946227233, + "grad_norm": 0.175845667719841, + "learning_rate": 4.193384593065429e-06, + "loss": 0.0011, + "step": 21891 + }, + { + "epoch": 2.373373807458803, + "grad_norm": 0.4769565463066101, + "learning_rate": 4.192659219498042e-06, + "loss": 0.0014, + "step": 21892 + }, + { + "epoch": 2.373482220294883, + "grad_norm": 0.03364205360412598, + "learning_rate": 4.191933845930655e-06, + "loss": 0.0003, + "step": 21893 + }, + { + "epoch": 2.373590633130963, + "grad_norm": 0.05203619971871376, + "learning_rate": 4.191208472363267e-06, + "loss": 0.0002, + "step": 21894 + }, + { + "epoch": 2.3736990459670424, + "grad_norm": 0.6082697510719299, + "learning_rate": 4.19048309879588e-06, + "loss": 0.0007, + "step": 21895 + }, + { + "epoch": 2.3738074588031224, + "grad_norm": 0.758232831954956, + "learning_rate": 4.189757725228493e-06, + "loss": 0.0012, + "step": 21896 + }, + { + "epoch": 2.373915871639202, + "grad_norm": 0.20501042902469635, + "learning_rate": 4.189032351661106e-06, + "loss": 0.0015, + "step": 21897 + }, + { + "epoch": 2.374024284475282, + "grad_norm": 0.036228109151124954, + "learning_rate": 4.188306978093718e-06, + "loss": 0.0003, + "step": 21898 + }, + { + "epoch": 2.374132697311362, + "grad_norm": 0.07035494595766068, + "learning_rate": 4.187581604526331e-06, + "loss": 0.0004, + "step": 21899 + }, + { + "epoch": 2.3742411101474414, + "grad_norm": 4.029742240905762, + "learning_rate": 4.186856230958944e-06, + "loss": 0.0456, + "step": 21900 + }, + { + "epoch": 2.3743495229835214, + "grad_norm": 3.604903221130371, + "learning_rate": 4.186130857391557e-06, + "loss": 0.0493, + "step": 21901 + }, + { + "epoch": 2.374457935819601, + "grad_norm": 5.719594955444336, + "learning_rate": 4.18540548382417e-06, + "loss": 0.0014, + "step": 21902 + }, + { + "epoch": 2.374566348655681, + "grad_norm": 0.5979875922203064, + "learning_rate": 4.184680110256782e-06, + "loss": 0.0019, + "step": 21903 + }, + { + "epoch": 2.3746747614917605, + "grad_norm": 2.4223365783691406, + "learning_rate": 4.183954736689395e-06, + "loss": 0.004, + "step": 21904 + }, + { + "epoch": 2.3747831743278405, + "grad_norm": 0.15005318820476532, + "learning_rate": 4.183229363122008e-06, + "loss": 0.0011, + "step": 21905 + }, + { + "epoch": 2.37489158716392, + "grad_norm": 0.09834928065538406, + "learning_rate": 4.182503989554621e-06, + "loss": 0.0004, + "step": 21906 + }, + { + "epoch": 2.375, + "grad_norm": 4.745009899139404, + "learning_rate": 4.181778615987233e-06, + "loss": 0.0334, + "step": 21907 + }, + { + "epoch": 2.37510841283608, + "grad_norm": 0.05782472714781761, + "learning_rate": 4.181053242419846e-06, + "loss": 0.0006, + "step": 21908 + }, + { + "epoch": 2.3752168256721595, + "grad_norm": 0.04063808545470238, + "learning_rate": 4.180327868852459e-06, + "loss": 0.0003, + "step": 21909 + }, + { + "epoch": 2.3753252385082395, + "grad_norm": 4.810581684112549, + "learning_rate": 4.179602495285072e-06, + "loss": 0.0384, + "step": 21910 + }, + { + "epoch": 2.375433651344319, + "grad_norm": 3.879394292831421, + "learning_rate": 4.1788771217176845e-06, + "loss": 0.0151, + "step": 21911 + }, + { + "epoch": 2.375542064180399, + "grad_norm": 0.40252169966697693, + "learning_rate": 4.1781517481502975e-06, + "loss": 0.0011, + "step": 21912 + }, + { + "epoch": 2.3756504770164786, + "grad_norm": 0.22654768824577332, + "learning_rate": 4.1774263745829105e-06, + "loss": 0.0003, + "step": 21913 + }, + { + "epoch": 2.3757588898525586, + "grad_norm": 0.020052220672369003, + "learning_rate": 4.1767010010155235e-06, + "loss": 0.0001, + "step": 21914 + }, + { + "epoch": 2.375867302688638, + "grad_norm": 1.2299474477767944, + "learning_rate": 4.1759756274481365e-06, + "loss": 0.002, + "step": 21915 + }, + { + "epoch": 2.375975715524718, + "grad_norm": 0.17914237082004547, + "learning_rate": 4.175250253880749e-06, + "loss": 0.0006, + "step": 21916 + }, + { + "epoch": 2.376084128360798, + "grad_norm": 0.5768160820007324, + "learning_rate": 4.174524880313362e-06, + "loss": 0.0023, + "step": 21917 + }, + { + "epoch": 2.3761925411968776, + "grad_norm": 2.026620864868164, + "learning_rate": 4.173799506745975e-06, + "loss": 0.0037, + "step": 21918 + }, + { + "epoch": 2.3763009540329576, + "grad_norm": 1.7255357503890991, + "learning_rate": 4.1730741331785876e-06, + "loss": 0.006, + "step": 21919 + }, + { + "epoch": 2.376409366869037, + "grad_norm": 4.3375773429870605, + "learning_rate": 4.1723487596112e-06, + "loss": 0.0122, + "step": 21920 + }, + { + "epoch": 2.376517779705117, + "grad_norm": 5.7527971267700195, + "learning_rate": 4.171623386043813e-06, + "loss": 0.0177, + "step": 21921 + }, + { + "epoch": 2.376626192541197, + "grad_norm": 0.2298743575811386, + "learning_rate": 4.170898012476426e-06, + "loss": 0.001, + "step": 21922 + }, + { + "epoch": 2.3767346053772767, + "grad_norm": 0.6913186311721802, + "learning_rate": 4.170172638909039e-06, + "loss": 0.003, + "step": 21923 + }, + { + "epoch": 2.3768430182133566, + "grad_norm": 0.16346025466918945, + "learning_rate": 4.169447265341651e-06, + "loss": 0.0009, + "step": 21924 + }, + { + "epoch": 2.376951431049436, + "grad_norm": 0.12582994997501373, + "learning_rate": 4.168721891774264e-06, + "loss": 0.0004, + "step": 21925 + }, + { + "epoch": 2.377059843885516, + "grad_norm": 0.020865069702267647, + "learning_rate": 4.167996518206877e-06, + "loss": 0.0002, + "step": 21926 + }, + { + "epoch": 2.3771682567215957, + "grad_norm": 0.9972116351127625, + "learning_rate": 4.16727114463949e-06, + "loss": 0.0283, + "step": 21927 + }, + { + "epoch": 2.3772766695576757, + "grad_norm": 0.10449742525815964, + "learning_rate": 4.166545771072103e-06, + "loss": 0.0005, + "step": 21928 + }, + { + "epoch": 2.3773850823937552, + "grad_norm": 0.11091809719800949, + "learning_rate": 4.165820397504715e-06, + "loss": 0.0006, + "step": 21929 + }, + { + "epoch": 2.377493495229835, + "grad_norm": 0.0563964806497097, + "learning_rate": 4.165095023937328e-06, + "loss": 0.0003, + "step": 21930 + }, + { + "epoch": 2.377601908065915, + "grad_norm": 2.055243492126465, + "learning_rate": 4.164369650369941e-06, + "loss": 0.0368, + "step": 21931 + }, + { + "epoch": 2.3777103209019947, + "grad_norm": 1.6992597579956055, + "learning_rate": 4.163644276802554e-06, + "loss": 0.0169, + "step": 21932 + }, + { + "epoch": 2.3778187337380747, + "grad_norm": 3.51605224609375, + "learning_rate": 4.162918903235166e-06, + "loss": 0.0088, + "step": 21933 + }, + { + "epoch": 2.3779271465741543, + "grad_norm": 0.06709721684455872, + "learning_rate": 4.162193529667779e-06, + "loss": 0.0003, + "step": 21934 + }, + { + "epoch": 2.3780355594102343, + "grad_norm": 8.418091773986816, + "learning_rate": 4.161468156100392e-06, + "loss": 0.0092, + "step": 21935 + }, + { + "epoch": 2.378143972246314, + "grad_norm": 2.3302979469299316, + "learning_rate": 4.160742782533005e-06, + "loss": 0.0077, + "step": 21936 + }, + { + "epoch": 2.378252385082394, + "grad_norm": 0.010792205110192299, + "learning_rate": 4.160017408965617e-06, + "loss": 0.0001, + "step": 21937 + }, + { + "epoch": 2.3783607979184733, + "grad_norm": 1.478541612625122, + "learning_rate": 4.15929203539823e-06, + "loss": 0.0037, + "step": 21938 + }, + { + "epoch": 2.3784692107545533, + "grad_norm": 1.168824315071106, + "learning_rate": 4.158566661830843e-06, + "loss": 0.0026, + "step": 21939 + }, + { + "epoch": 2.3785776235906333, + "grad_norm": 6.964822292327881, + "learning_rate": 4.157841288263456e-06, + "loss": 0.068, + "step": 21940 + }, + { + "epoch": 2.378686036426713, + "grad_norm": 1.5994784832000732, + "learning_rate": 4.157115914696069e-06, + "loss": 0.0046, + "step": 21941 + }, + { + "epoch": 2.378794449262793, + "grad_norm": 0.0747908353805542, + "learning_rate": 4.156390541128681e-06, + "loss": 0.0004, + "step": 21942 + }, + { + "epoch": 2.3789028620988724, + "grad_norm": 1.1419072151184082, + "learning_rate": 4.155665167561294e-06, + "loss": 0.002, + "step": 21943 + }, + { + "epoch": 2.3790112749349523, + "grad_norm": 0.21452710032463074, + "learning_rate": 4.154939793993907e-06, + "loss": 0.001, + "step": 21944 + }, + { + "epoch": 2.3791196877710323, + "grad_norm": 1.7816762924194336, + "learning_rate": 4.15421442042652e-06, + "loss": 0.0129, + "step": 21945 + }, + { + "epoch": 2.379228100607112, + "grad_norm": 1.2413872480392456, + "learning_rate": 4.153489046859132e-06, + "loss": 0.0027, + "step": 21946 + }, + { + "epoch": 2.379336513443192, + "grad_norm": 0.04091238975524902, + "learning_rate": 4.152763673291745e-06, + "loss": 0.0004, + "step": 21947 + }, + { + "epoch": 2.3794449262792714, + "grad_norm": 0.13911794126033783, + "learning_rate": 4.152038299724358e-06, + "loss": 0.0012, + "step": 21948 + }, + { + "epoch": 2.3795533391153514, + "grad_norm": 10.252177238464355, + "learning_rate": 4.151312926156971e-06, + "loss": 0.0089, + "step": 21949 + }, + { + "epoch": 2.379661751951431, + "grad_norm": 0.07837578654289246, + "learning_rate": 4.1505875525895835e-06, + "loss": 0.0004, + "step": 21950 + }, + { + "epoch": 2.379770164787511, + "grad_norm": 0.02928462252020836, + "learning_rate": 4.1498621790221965e-06, + "loss": 0.0002, + "step": 21951 + }, + { + "epoch": 2.3798785776235905, + "grad_norm": 0.32491904497146606, + "learning_rate": 4.1491368054548095e-06, + "loss": 0.0011, + "step": 21952 + }, + { + "epoch": 2.3799869904596704, + "grad_norm": 2.0874409675598145, + "learning_rate": 4.1484114318874225e-06, + "loss": 0.0145, + "step": 21953 + }, + { + "epoch": 2.3800954032957504, + "grad_norm": 1.9269495010375977, + "learning_rate": 4.1476860583200355e-06, + "loss": 0.0032, + "step": 21954 + }, + { + "epoch": 2.38020381613183, + "grad_norm": 5.752504348754883, + "learning_rate": 4.146960684752648e-06, + "loss": 0.0099, + "step": 21955 + }, + { + "epoch": 2.38031222896791, + "grad_norm": 0.009374193847179413, + "learning_rate": 4.146235311185261e-06, + "loss": 0.0001, + "step": 21956 + }, + { + "epoch": 2.3804206418039895, + "grad_norm": 14.276299476623535, + "learning_rate": 4.145509937617874e-06, + "loss": 0.0132, + "step": 21957 + }, + { + "epoch": 2.3805290546400695, + "grad_norm": 1.5216102600097656, + "learning_rate": 4.1447845640504866e-06, + "loss": 0.0048, + "step": 21958 + }, + { + "epoch": 2.380637467476149, + "grad_norm": 2.164442539215088, + "learning_rate": 4.144059190483099e-06, + "loss": 0.0063, + "step": 21959 + }, + { + "epoch": 2.380745880312229, + "grad_norm": 5.634377479553223, + "learning_rate": 4.143333816915712e-06, + "loss": 0.0108, + "step": 21960 + }, + { + "epoch": 2.3808542931483085, + "grad_norm": 0.10337977856397629, + "learning_rate": 4.142608443348325e-06, + "loss": 0.0007, + "step": 21961 + }, + { + "epoch": 2.3809627059843885, + "grad_norm": 0.06361441314220428, + "learning_rate": 4.141883069780938e-06, + "loss": 0.0005, + "step": 21962 + }, + { + "epoch": 2.3810711188204685, + "grad_norm": 0.07802256941795349, + "learning_rate": 4.14115769621355e-06, + "loss": 0.0003, + "step": 21963 + }, + { + "epoch": 2.381179531656548, + "grad_norm": 1.2087090015411377, + "learning_rate": 4.140432322646163e-06, + "loss": 0.0052, + "step": 21964 + }, + { + "epoch": 2.381287944492628, + "grad_norm": 0.025325529277324677, + "learning_rate": 4.139706949078776e-06, + "loss": 0.0002, + "step": 21965 + }, + { + "epoch": 2.3813963573287076, + "grad_norm": 0.0699634999036789, + "learning_rate": 4.138981575511389e-06, + "loss": 0.0006, + "step": 21966 + }, + { + "epoch": 2.3815047701647876, + "grad_norm": 0.11523748934268951, + "learning_rate": 4.138256201944002e-06, + "loss": 0.0005, + "step": 21967 + }, + { + "epoch": 2.3816131830008676, + "grad_norm": 7.8963212966918945, + "learning_rate": 4.137530828376614e-06, + "loss": 0.0143, + "step": 21968 + }, + { + "epoch": 2.381721595836947, + "grad_norm": 0.015657678246498108, + "learning_rate": 4.136805454809228e-06, + "loss": 0.0001, + "step": 21969 + }, + { + "epoch": 2.381830008673027, + "grad_norm": 0.2843862771987915, + "learning_rate": 4.13608008124184e-06, + "loss": 0.0014, + "step": 21970 + }, + { + "epoch": 2.3819384215091066, + "grad_norm": 0.1928999274969101, + "learning_rate": 4.135354707674453e-06, + "loss": 0.0006, + "step": 21971 + }, + { + "epoch": 2.3820468343451866, + "grad_norm": 0.03493855521082878, + "learning_rate": 4.134629334107065e-06, + "loss": 0.0002, + "step": 21972 + }, + { + "epoch": 2.382155247181266, + "grad_norm": 4.596784591674805, + "learning_rate": 4.133903960539678e-06, + "loss": 0.0123, + "step": 21973 + }, + { + "epoch": 2.382263660017346, + "grad_norm": 0.003957820124924183, + "learning_rate": 4.133178586972291e-06, + "loss": 0.0, + "step": 21974 + }, + { + "epoch": 2.3823720728534257, + "grad_norm": 0.022266285493969917, + "learning_rate": 4.132453213404904e-06, + "loss": 0.0001, + "step": 21975 + }, + { + "epoch": 2.3824804856895057, + "grad_norm": 0.10023657977581024, + "learning_rate": 4.131727839837516e-06, + "loss": 0.0005, + "step": 21976 + }, + { + "epoch": 2.3825888985255856, + "grad_norm": 8.994285583496094, + "learning_rate": 4.131002466270129e-06, + "loss": 0.0132, + "step": 21977 + }, + { + "epoch": 2.382697311361665, + "grad_norm": 0.0991053655743599, + "learning_rate": 4.130277092702742e-06, + "loss": 0.0004, + "step": 21978 + }, + { + "epoch": 2.382805724197745, + "grad_norm": 0.11863864958286285, + "learning_rate": 4.129551719135355e-06, + "loss": 0.0007, + "step": 21979 + }, + { + "epoch": 2.3829141370338247, + "grad_norm": 0.12613174319267273, + "learning_rate": 4.128826345567968e-06, + "loss": 0.0004, + "step": 21980 + }, + { + "epoch": 2.3830225498699047, + "grad_norm": 0.028156369924545288, + "learning_rate": 4.12810097200058e-06, + "loss": 0.0002, + "step": 21981 + }, + { + "epoch": 2.3831309627059842, + "grad_norm": 0.022031620144844055, + "learning_rate": 4.127375598433194e-06, + "loss": 0.0002, + "step": 21982 + }, + { + "epoch": 2.383239375542064, + "grad_norm": 0.0058936914429068565, + "learning_rate": 4.126650224865806e-06, + "loss": 0.0001, + "step": 21983 + }, + { + "epoch": 2.3833477883781438, + "grad_norm": 0.2656874358654022, + "learning_rate": 4.125924851298419e-06, + "loss": 0.0014, + "step": 21984 + }, + { + "epoch": 2.3834562012142237, + "grad_norm": 1.739073395729065, + "learning_rate": 4.125199477731031e-06, + "loss": 0.0336, + "step": 21985 + }, + { + "epoch": 2.3835646140503037, + "grad_norm": 5.172744274139404, + "learning_rate": 4.124474104163644e-06, + "loss": 0.0202, + "step": 21986 + }, + { + "epoch": 2.3836730268863833, + "grad_norm": 0.3244609236717224, + "learning_rate": 4.123748730596257e-06, + "loss": 0.0012, + "step": 21987 + }, + { + "epoch": 2.3837814397224633, + "grad_norm": 2.487837791442871, + "learning_rate": 4.12302335702887e-06, + "loss": 0.0118, + "step": 21988 + }, + { + "epoch": 2.383889852558543, + "grad_norm": 0.12415517866611481, + "learning_rate": 4.1222979834614825e-06, + "loss": 0.0005, + "step": 21989 + }, + { + "epoch": 2.383998265394623, + "grad_norm": 2.882983684539795, + "learning_rate": 4.1215726098940955e-06, + "loss": 0.0233, + "step": 21990 + }, + { + "epoch": 2.3841066782307023, + "grad_norm": 0.3386293649673462, + "learning_rate": 4.1208472363267085e-06, + "loss": 0.0011, + "step": 21991 + }, + { + "epoch": 2.3842150910667823, + "grad_norm": 2.3221492767333984, + "learning_rate": 4.1201218627593215e-06, + "loss": 0.019, + "step": 21992 + }, + { + "epoch": 2.384323503902862, + "grad_norm": 0.4402681589126587, + "learning_rate": 4.1193964891919345e-06, + "loss": 0.0011, + "step": 21993 + }, + { + "epoch": 2.384431916738942, + "grad_norm": 0.022125007584691048, + "learning_rate": 4.118671115624547e-06, + "loss": 0.0001, + "step": 21994 + }, + { + "epoch": 2.384540329575022, + "grad_norm": 6.535558223724365, + "learning_rate": 4.1179457420571604e-06, + "loss": 0.0179, + "step": 21995 + }, + { + "epoch": 2.3846487424111014, + "grad_norm": 0.8953417539596558, + "learning_rate": 4.117220368489773e-06, + "loss": 0.0026, + "step": 21996 + }, + { + "epoch": 2.3847571552471813, + "grad_norm": 10.632055282592773, + "learning_rate": 4.1164949949223856e-06, + "loss": 0.0099, + "step": 21997 + }, + { + "epoch": 2.384865568083261, + "grad_norm": 0.5697341561317444, + "learning_rate": 4.115769621354998e-06, + "loss": 0.0029, + "step": 21998 + }, + { + "epoch": 2.384973980919341, + "grad_norm": 0.4954107701778412, + "learning_rate": 4.115044247787611e-06, + "loss": 0.0013, + "step": 21999 + }, + { + "epoch": 2.385082393755421, + "grad_norm": 0.10077734291553497, + "learning_rate": 4.114318874220224e-06, + "loss": 0.0008, + "step": 22000 + }, + { + "epoch": 2.3851908065915004, + "grad_norm": 1.3419142961502075, + "learning_rate": 4.113593500652837e-06, + "loss": 0.0065, + "step": 22001 + }, + { + "epoch": 2.3852992194275804, + "grad_norm": 0.013960707001388073, + "learning_rate": 4.112868127085449e-06, + "loss": 0.0001, + "step": 22002 + }, + { + "epoch": 2.38540763226366, + "grad_norm": 0.6936956644058228, + "learning_rate": 4.112142753518062e-06, + "loss": 0.0021, + "step": 22003 + }, + { + "epoch": 2.38551604509974, + "grad_norm": 0.3019002079963684, + "learning_rate": 4.111417379950675e-06, + "loss": 0.0008, + "step": 22004 + }, + { + "epoch": 2.3856244579358195, + "grad_norm": 1.6444164514541626, + "learning_rate": 4.110692006383288e-06, + "loss": 0.0052, + "step": 22005 + }, + { + "epoch": 2.3857328707718994, + "grad_norm": 1.3782755136489868, + "learning_rate": 4.109966632815901e-06, + "loss": 0.002, + "step": 22006 + }, + { + "epoch": 2.385841283607979, + "grad_norm": 15.320838928222656, + "learning_rate": 4.109241259248513e-06, + "loss": 0.0177, + "step": 22007 + }, + { + "epoch": 2.385949696444059, + "grad_norm": 3.4042012691497803, + "learning_rate": 4.108515885681127e-06, + "loss": 0.0327, + "step": 22008 + }, + { + "epoch": 2.386058109280139, + "grad_norm": 0.007533587981015444, + "learning_rate": 4.107790512113739e-06, + "loss": 0.0, + "step": 22009 + }, + { + "epoch": 2.3861665221162185, + "grad_norm": 0.16320550441741943, + "learning_rate": 4.107065138546352e-06, + "loss": 0.0006, + "step": 22010 + }, + { + "epoch": 2.3862749349522985, + "grad_norm": 0.8260809183120728, + "learning_rate": 4.106339764978964e-06, + "loss": 0.0029, + "step": 22011 + }, + { + "epoch": 2.386383347788378, + "grad_norm": 2.927255153656006, + "learning_rate": 4.105614391411577e-06, + "loss": 0.0127, + "step": 22012 + }, + { + "epoch": 2.386491760624458, + "grad_norm": 0.5784169435501099, + "learning_rate": 4.10488901784419e-06, + "loss": 0.0022, + "step": 22013 + }, + { + "epoch": 2.3866001734605375, + "grad_norm": 0.7856289148330688, + "learning_rate": 4.104163644276803e-06, + "loss": 0.0027, + "step": 22014 + }, + { + "epoch": 2.3867085862966175, + "grad_norm": 1.4109423160552979, + "learning_rate": 4.103438270709415e-06, + "loss": 0.0357, + "step": 22015 + }, + { + "epoch": 2.386816999132697, + "grad_norm": 0.11402394622564316, + "learning_rate": 4.102712897142028e-06, + "loss": 0.0007, + "step": 22016 + }, + { + "epoch": 2.386925411968777, + "grad_norm": 0.12081168591976166, + "learning_rate": 4.101987523574641e-06, + "loss": 0.0005, + "step": 22017 + }, + { + "epoch": 2.387033824804857, + "grad_norm": 1.9660793542861938, + "learning_rate": 4.101262150007254e-06, + "loss": 0.0024, + "step": 22018 + }, + { + "epoch": 2.3871422376409366, + "grad_norm": 0.24219746887683868, + "learning_rate": 4.100536776439867e-06, + "loss": 0.0009, + "step": 22019 + }, + { + "epoch": 2.3872506504770166, + "grad_norm": 0.09392517060041428, + "learning_rate": 4.099811402872479e-06, + "loss": 0.0004, + "step": 22020 + }, + { + "epoch": 2.387359063313096, + "grad_norm": 0.7089408040046692, + "learning_rate": 4.099086029305093e-06, + "loss": 0.0021, + "step": 22021 + }, + { + "epoch": 2.387467476149176, + "grad_norm": 1.087814211845398, + "learning_rate": 4.098360655737705e-06, + "loss": 0.0394, + "step": 22022 + }, + { + "epoch": 2.387575888985256, + "grad_norm": 0.025827594101428986, + "learning_rate": 4.097635282170318e-06, + "loss": 0.0001, + "step": 22023 + }, + { + "epoch": 2.3876843018213356, + "grad_norm": 1.3841220140457153, + "learning_rate": 4.09690990860293e-06, + "loss": 0.0059, + "step": 22024 + }, + { + "epoch": 2.3877927146574156, + "grad_norm": 0.03292837738990784, + "learning_rate": 4.096184535035543e-06, + "loss": 0.0003, + "step": 22025 + }, + { + "epoch": 2.387901127493495, + "grad_norm": 0.7581271529197693, + "learning_rate": 4.095459161468156e-06, + "loss": 0.0041, + "step": 22026 + }, + { + "epoch": 2.388009540329575, + "grad_norm": 0.07461491972208023, + "learning_rate": 4.094733787900769e-06, + "loss": 0.0003, + "step": 22027 + }, + { + "epoch": 2.3881179531656547, + "grad_norm": 0.03334081918001175, + "learning_rate": 4.0940084143333815e-06, + "loss": 0.0002, + "step": 22028 + }, + { + "epoch": 2.3882263660017347, + "grad_norm": 0.1437508761882782, + "learning_rate": 4.0932830407659945e-06, + "loss": 0.0003, + "step": 22029 + }, + { + "epoch": 2.388334778837814, + "grad_norm": 0.13786499202251434, + "learning_rate": 4.0925576671986075e-06, + "loss": 0.0004, + "step": 22030 + }, + { + "epoch": 2.388443191673894, + "grad_norm": 14.804088592529297, + "learning_rate": 4.0918322936312205e-06, + "loss": 0.0887, + "step": 22031 + }, + { + "epoch": 2.388551604509974, + "grad_norm": 0.010585430078208447, + "learning_rate": 4.0911069200638335e-06, + "loss": 0.0001, + "step": 22032 + }, + { + "epoch": 2.3886600173460537, + "grad_norm": 0.005937401205301285, + "learning_rate": 4.090381546496446e-06, + "loss": 0.0001, + "step": 22033 + }, + { + "epoch": 2.3887684301821337, + "grad_norm": 0.042931731790304184, + "learning_rate": 4.0896561729290594e-06, + "loss": 0.0003, + "step": 22034 + }, + { + "epoch": 2.3888768430182132, + "grad_norm": 6.106297492980957, + "learning_rate": 4.088930799361672e-06, + "loss": 0.0303, + "step": 22035 + }, + { + "epoch": 2.388985255854293, + "grad_norm": 2.830995559692383, + "learning_rate": 4.088205425794285e-06, + "loss": 0.0202, + "step": 22036 + }, + { + "epoch": 2.3890936686903728, + "grad_norm": 0.6930083632469177, + "learning_rate": 4.087480052226897e-06, + "loss": 0.0262, + "step": 22037 + }, + { + "epoch": 2.3892020815264527, + "grad_norm": 2.170713424682617, + "learning_rate": 4.0867546786595106e-06, + "loss": 0.0137, + "step": 22038 + }, + { + "epoch": 2.3893104943625323, + "grad_norm": 2.851738929748535, + "learning_rate": 4.086029305092123e-06, + "loss": 0.0129, + "step": 22039 + }, + { + "epoch": 2.3894189071986123, + "grad_norm": 3.75089955329895, + "learning_rate": 4.085303931524736e-06, + "loss": 0.006, + "step": 22040 + }, + { + "epoch": 2.3895273200346923, + "grad_norm": 4.27709436416626, + "learning_rate": 4.084578557957348e-06, + "loss": 0.02, + "step": 22041 + }, + { + "epoch": 2.389635732870772, + "grad_norm": 1.172811508178711, + "learning_rate": 4.083853184389961e-06, + "loss": 0.0278, + "step": 22042 + }, + { + "epoch": 2.389744145706852, + "grad_norm": 11.73270034790039, + "learning_rate": 4.083127810822574e-06, + "loss": 0.0446, + "step": 22043 + }, + { + "epoch": 2.3898525585429313, + "grad_norm": 2.7642242908477783, + "learning_rate": 4.082402437255187e-06, + "loss": 0.0406, + "step": 22044 + }, + { + "epoch": 2.3899609713790113, + "grad_norm": 0.05233188346028328, + "learning_rate": 4.0816770636878e-06, + "loss": 0.0004, + "step": 22045 + }, + { + "epoch": 2.3900693842150913, + "grad_norm": 1.72965669631958, + "learning_rate": 4.080951690120412e-06, + "loss": 0.0335, + "step": 22046 + }, + { + "epoch": 2.390177797051171, + "grad_norm": 0.1822883039712906, + "learning_rate": 4.080226316553026e-06, + "loss": 0.0005, + "step": 22047 + }, + { + "epoch": 2.390286209887251, + "grad_norm": 0.10579103231430054, + "learning_rate": 4.079500942985638e-06, + "loss": 0.0002, + "step": 22048 + }, + { + "epoch": 2.3903946227233304, + "grad_norm": 0.14578212797641754, + "learning_rate": 4.078775569418251e-06, + "loss": 0.0003, + "step": 22049 + }, + { + "epoch": 2.3905030355594103, + "grad_norm": 4.473652362823486, + "learning_rate": 4.078050195850863e-06, + "loss": 0.0016, + "step": 22050 + }, + { + "epoch": 2.39061144839549, + "grad_norm": 0.5571759939193726, + "learning_rate": 4.077324822283477e-06, + "loss": 0.0005, + "step": 22051 + }, + { + "epoch": 2.39071986123157, + "grad_norm": 0.05016952380537987, + "learning_rate": 4.076599448716089e-06, + "loss": 0.0003, + "step": 22052 + }, + { + "epoch": 2.3908282740676494, + "grad_norm": 0.13867135345935822, + "learning_rate": 4.075874075148702e-06, + "loss": 0.0005, + "step": 22053 + }, + { + "epoch": 2.3909366869037294, + "grad_norm": 1.2272893190383911, + "learning_rate": 4.075148701581314e-06, + "loss": 0.0041, + "step": 22054 + }, + { + "epoch": 2.3910450997398094, + "grad_norm": 4.8773674964904785, + "learning_rate": 4.074423328013927e-06, + "loss": 0.1154, + "step": 22055 + }, + { + "epoch": 2.391153512575889, + "grad_norm": 3.252366065979004, + "learning_rate": 4.07369795444654e-06, + "loss": 0.0134, + "step": 22056 + }, + { + "epoch": 2.391261925411969, + "grad_norm": 2.543182849884033, + "learning_rate": 4.072972580879153e-06, + "loss": 0.0143, + "step": 22057 + }, + { + "epoch": 2.3913703382480485, + "grad_norm": 2.4905505180358887, + "learning_rate": 4.072247207311766e-06, + "loss": 0.0076, + "step": 22058 + }, + { + "epoch": 2.3914787510841284, + "grad_norm": 4.368950843811035, + "learning_rate": 4.071521833744378e-06, + "loss": 0.0127, + "step": 22059 + }, + { + "epoch": 2.391587163920208, + "grad_norm": 3.6892004013061523, + "learning_rate": 4.070796460176992e-06, + "loss": 0.0021, + "step": 22060 + }, + { + "epoch": 2.391695576756288, + "grad_norm": 0.08634624630212784, + "learning_rate": 4.070071086609604e-06, + "loss": 0.0013, + "step": 22061 + }, + { + "epoch": 2.3918039895923675, + "grad_norm": 0.1404210329055786, + "learning_rate": 4.069345713042217e-06, + "loss": 0.0013, + "step": 22062 + }, + { + "epoch": 2.3919124024284475, + "grad_norm": 0.15856695175170898, + "learning_rate": 4.068620339474829e-06, + "loss": 0.0013, + "step": 22063 + }, + { + "epoch": 2.3920208152645275, + "grad_norm": 2.010045051574707, + "learning_rate": 4.067894965907443e-06, + "loss": 0.0374, + "step": 22064 + }, + { + "epoch": 2.392129228100607, + "grad_norm": 5.043254852294922, + "learning_rate": 4.067169592340055e-06, + "loss": 0.0177, + "step": 22065 + }, + { + "epoch": 2.392237640936687, + "grad_norm": 0.2047336995601654, + "learning_rate": 4.066444218772668e-06, + "loss": 0.0017, + "step": 22066 + }, + { + "epoch": 2.3923460537727665, + "grad_norm": 4.423497676849365, + "learning_rate": 4.065718845205281e-06, + "loss": 0.004, + "step": 22067 + }, + { + "epoch": 2.3924544666088465, + "grad_norm": 0.004140147008001804, + "learning_rate": 4.0649934716378935e-06, + "loss": 0.0001, + "step": 22068 + }, + { + "epoch": 2.3925628794449265, + "grad_norm": 2.3802974224090576, + "learning_rate": 4.0642680980705065e-06, + "loss": 0.0041, + "step": 22069 + }, + { + "epoch": 2.392671292281006, + "grad_norm": 3.0296809673309326, + "learning_rate": 4.0635427245031195e-06, + "loss": 0.0087, + "step": 22070 + }, + { + "epoch": 2.392779705117086, + "grad_norm": 0.030723916366696358, + "learning_rate": 4.0628173509357325e-06, + "loss": 0.0002, + "step": 22071 + }, + { + "epoch": 2.3928881179531656, + "grad_norm": 2.6554994583129883, + "learning_rate": 4.062091977368345e-06, + "loss": 0.0041, + "step": 22072 + }, + { + "epoch": 2.3929965307892456, + "grad_norm": 0.5102653503417969, + "learning_rate": 4.0613666038009584e-06, + "loss": 0.0021, + "step": 22073 + }, + { + "epoch": 2.393104943625325, + "grad_norm": 8.167162895202637, + "learning_rate": 4.060641230233571e-06, + "loss": 0.0366, + "step": 22074 + }, + { + "epoch": 2.393213356461405, + "grad_norm": 4.427125453948975, + "learning_rate": 4.059915856666184e-06, + "loss": 0.0276, + "step": 22075 + }, + { + "epoch": 2.3933217692974846, + "grad_norm": 0.20039115846157074, + "learning_rate": 4.059190483098796e-06, + "loss": 0.0018, + "step": 22076 + }, + { + "epoch": 2.3934301821335646, + "grad_norm": 1.2560862302780151, + "learning_rate": 4.0584651095314096e-06, + "loss": 0.0015, + "step": 22077 + }, + { + "epoch": 2.3935385949696446, + "grad_norm": 1.5155390501022339, + "learning_rate": 4.057739735964022e-06, + "loss": 0.0057, + "step": 22078 + }, + { + "epoch": 2.393647007805724, + "grad_norm": 5.021583557128906, + "learning_rate": 4.057014362396635e-06, + "loss": 0.0437, + "step": 22079 + }, + { + "epoch": 2.393755420641804, + "grad_norm": 0.08056825399398804, + "learning_rate": 4.056288988829248e-06, + "loss": 0.0003, + "step": 22080 + }, + { + "epoch": 2.3938638334778837, + "grad_norm": 3.38663649559021, + "learning_rate": 4.05556361526186e-06, + "loss": 0.0041, + "step": 22081 + }, + { + "epoch": 2.3939722463139637, + "grad_norm": 2.9798784255981445, + "learning_rate": 4.054838241694473e-06, + "loss": 0.025, + "step": 22082 + }, + { + "epoch": 2.394080659150043, + "grad_norm": 0.22598621249198914, + "learning_rate": 4.054112868127086e-06, + "loss": 0.0005, + "step": 22083 + }, + { + "epoch": 2.394189071986123, + "grad_norm": 0.054730407893657684, + "learning_rate": 4.053387494559699e-06, + "loss": 0.0004, + "step": 22084 + }, + { + "epoch": 2.3942974848222027, + "grad_norm": 0.1849924772977829, + "learning_rate": 4.052662120992311e-06, + "loss": 0.0007, + "step": 22085 + }, + { + "epoch": 2.3944058976582827, + "grad_norm": 0.058503441512584686, + "learning_rate": 4.051936747424925e-06, + "loss": 0.0006, + "step": 22086 + }, + { + "epoch": 2.3945143104943627, + "grad_norm": 2.887371063232422, + "learning_rate": 4.051211373857537e-06, + "loss": 0.005, + "step": 22087 + }, + { + "epoch": 2.3946227233304422, + "grad_norm": 0.19566524028778076, + "learning_rate": 4.05048600029015e-06, + "loss": 0.0021, + "step": 22088 + }, + { + "epoch": 2.3947311361665222, + "grad_norm": 0.10011337697505951, + "learning_rate": 4.049760626722762e-06, + "loss": 0.0002, + "step": 22089 + }, + { + "epoch": 2.3948395490026018, + "grad_norm": 0.06756134331226349, + "learning_rate": 4.049035253155376e-06, + "loss": 0.0004, + "step": 22090 + }, + { + "epoch": 2.3949479618386817, + "grad_norm": 2.258943796157837, + "learning_rate": 4.048309879587988e-06, + "loss": 0.0024, + "step": 22091 + }, + { + "epoch": 2.3950563746747617, + "grad_norm": 3.1762781143188477, + "learning_rate": 4.047584506020601e-06, + "loss": 0.0109, + "step": 22092 + }, + { + "epoch": 2.3951647875108413, + "grad_norm": 0.018706470727920532, + "learning_rate": 4.046859132453214e-06, + "loss": 0.0001, + "step": 22093 + }, + { + "epoch": 2.3952732003469213, + "grad_norm": 0.09711265563964844, + "learning_rate": 4.046133758885826e-06, + "loss": 0.0003, + "step": 22094 + }, + { + "epoch": 2.395381613183001, + "grad_norm": 2.459388494491577, + "learning_rate": 4.045408385318439e-06, + "loss": 0.0176, + "step": 22095 + }, + { + "epoch": 2.395490026019081, + "grad_norm": 0.08837798982858658, + "learning_rate": 4.044683011751052e-06, + "loss": 0.001, + "step": 22096 + }, + { + "epoch": 2.3955984388551603, + "grad_norm": 0.2170228362083435, + "learning_rate": 4.043957638183665e-06, + "loss": 0.0017, + "step": 22097 + }, + { + "epoch": 2.3957068516912403, + "grad_norm": 8.50360107421875, + "learning_rate": 4.043232264616277e-06, + "loss": 0.0394, + "step": 22098 + }, + { + "epoch": 2.39581526452732, + "grad_norm": 0.6800627112388611, + "learning_rate": 4.042506891048891e-06, + "loss": 0.001, + "step": 22099 + }, + { + "epoch": 2.3959236773634, + "grad_norm": 0.10273579508066177, + "learning_rate": 4.041781517481503e-06, + "loss": 0.0005, + "step": 22100 + }, + { + "epoch": 2.39603209019948, + "grad_norm": 0.08016617596149445, + "learning_rate": 4.041056143914116e-06, + "loss": 0.0004, + "step": 22101 + }, + { + "epoch": 2.3961405030355594, + "grad_norm": 2.120840549468994, + "learning_rate": 4.040330770346728e-06, + "loss": 0.0083, + "step": 22102 + }, + { + "epoch": 2.3962489158716394, + "grad_norm": 1.7471308708190918, + "learning_rate": 4.039605396779342e-06, + "loss": 0.0128, + "step": 22103 + }, + { + "epoch": 2.396357328707719, + "grad_norm": 3.4314234256744385, + "learning_rate": 4.038880023211954e-06, + "loss": 0.0014, + "step": 22104 + }, + { + "epoch": 2.396465741543799, + "grad_norm": 2.651684045791626, + "learning_rate": 4.038154649644567e-06, + "loss": 0.0141, + "step": 22105 + }, + { + "epoch": 2.3965741543798784, + "grad_norm": 0.061231303960084915, + "learning_rate": 4.03742927607718e-06, + "loss": 0.0002, + "step": 22106 + }, + { + "epoch": 2.3966825672159584, + "grad_norm": 0.31944331526756287, + "learning_rate": 4.0367039025097925e-06, + "loss": 0.0015, + "step": 22107 + }, + { + "epoch": 2.396790980052038, + "grad_norm": 3.2144808769226074, + "learning_rate": 4.0359785289424055e-06, + "loss": 0.0401, + "step": 22108 + }, + { + "epoch": 2.396899392888118, + "grad_norm": 10.156523704528809, + "learning_rate": 4.0352531553750185e-06, + "loss": 0.0269, + "step": 22109 + }, + { + "epoch": 2.397007805724198, + "grad_norm": 7.003114223480225, + "learning_rate": 4.0345277818076315e-06, + "loss": 0.0355, + "step": 22110 + }, + { + "epoch": 2.3971162185602775, + "grad_norm": 3.731895685195923, + "learning_rate": 4.033802408240244e-06, + "loss": 0.0571, + "step": 22111 + }, + { + "epoch": 2.3972246313963574, + "grad_norm": 2.471867561340332, + "learning_rate": 4.0330770346728575e-06, + "loss": 0.0124, + "step": 22112 + }, + { + "epoch": 2.397333044232437, + "grad_norm": 3.09057879447937, + "learning_rate": 4.03235166110547e-06, + "loss": 0.0275, + "step": 22113 + }, + { + "epoch": 2.397441457068517, + "grad_norm": 0.0416627936065197, + "learning_rate": 4.031626287538083e-06, + "loss": 0.0001, + "step": 22114 + }, + { + "epoch": 2.397549869904597, + "grad_norm": 0.3840208649635315, + "learning_rate": 4.030900913970695e-06, + "loss": 0.0012, + "step": 22115 + }, + { + "epoch": 2.3976582827406765, + "grad_norm": 3.620661973953247, + "learning_rate": 4.0301755404033086e-06, + "loss": 0.0205, + "step": 22116 + }, + { + "epoch": 2.3977666955767565, + "grad_norm": 1.2132207155227661, + "learning_rate": 4.029450166835921e-06, + "loss": 0.0036, + "step": 22117 + }, + { + "epoch": 2.397875108412836, + "grad_norm": 0.2524184286594391, + "learning_rate": 4.028724793268534e-06, + "loss": 0.0026, + "step": 22118 + }, + { + "epoch": 2.397983521248916, + "grad_norm": 0.024455105885863304, + "learning_rate": 4.027999419701147e-06, + "loss": 0.0002, + "step": 22119 + }, + { + "epoch": 2.3980919340849955, + "grad_norm": 0.8164154291152954, + "learning_rate": 4.027274046133759e-06, + "loss": 0.0024, + "step": 22120 + }, + { + "epoch": 2.3982003469210755, + "grad_norm": 0.9796938300132751, + "learning_rate": 4.026548672566372e-06, + "loss": 0.0035, + "step": 22121 + }, + { + "epoch": 2.398308759757155, + "grad_norm": 6.9606099128723145, + "learning_rate": 4.025823298998985e-06, + "loss": 0.0256, + "step": 22122 + }, + { + "epoch": 2.398417172593235, + "grad_norm": 0.701344907283783, + "learning_rate": 4.025097925431598e-06, + "loss": 0.0017, + "step": 22123 + }, + { + "epoch": 2.398525585429315, + "grad_norm": 0.038012176752090454, + "learning_rate": 4.02437255186421e-06, + "loss": 0.0002, + "step": 22124 + }, + { + "epoch": 2.3986339982653946, + "grad_norm": 1.3974825143814087, + "learning_rate": 4.023647178296824e-06, + "loss": 0.029, + "step": 22125 + }, + { + "epoch": 2.3987424111014746, + "grad_norm": 0.5926151871681213, + "learning_rate": 4.022921804729436e-06, + "loss": 0.0019, + "step": 22126 + }, + { + "epoch": 2.398850823937554, + "grad_norm": 4.271730899810791, + "learning_rate": 4.022196431162049e-06, + "loss": 0.041, + "step": 22127 + }, + { + "epoch": 2.398959236773634, + "grad_norm": 3.0886404514312744, + "learning_rate": 4.021471057594661e-06, + "loss": 0.0082, + "step": 22128 + }, + { + "epoch": 2.3990676496097136, + "grad_norm": 0.16925948858261108, + "learning_rate": 4.020745684027275e-06, + "loss": 0.0007, + "step": 22129 + }, + { + "epoch": 2.3991760624457936, + "grad_norm": 13.41009521484375, + "learning_rate": 4.020020310459887e-06, + "loss": 0.0228, + "step": 22130 + }, + { + "epoch": 2.399284475281873, + "grad_norm": 0.048211731016635895, + "learning_rate": 4.0192949368925e-06, + "loss": 0.0005, + "step": 22131 + }, + { + "epoch": 2.399392888117953, + "grad_norm": 3.8395001888275146, + "learning_rate": 4.018569563325113e-06, + "loss": 0.023, + "step": 22132 + }, + { + "epoch": 2.399501300954033, + "grad_norm": 0.19702790677547455, + "learning_rate": 4.017844189757726e-06, + "loss": 0.0015, + "step": 22133 + }, + { + "epoch": 2.3996097137901127, + "grad_norm": 3.1800267696380615, + "learning_rate": 4.017118816190338e-06, + "loss": 0.0066, + "step": 22134 + }, + { + "epoch": 2.3997181266261927, + "grad_norm": 2.448679208755493, + "learning_rate": 4.016393442622951e-06, + "loss": 0.0045, + "step": 22135 + }, + { + "epoch": 2.399826539462272, + "grad_norm": 6.2904815673828125, + "learning_rate": 4.015668069055564e-06, + "loss": 0.0113, + "step": 22136 + }, + { + "epoch": 2.399934952298352, + "grad_norm": 0.019502056762576103, + "learning_rate": 4.014942695488176e-06, + "loss": 0.0002, + "step": 22137 + }, + { + "epoch": 2.4000433651344317, + "grad_norm": 0.6582026481628418, + "learning_rate": 4.01421732192079e-06, + "loss": 0.0021, + "step": 22138 + }, + { + "epoch": 2.4001517779705117, + "grad_norm": 2.6550559997558594, + "learning_rate": 4.013491948353402e-06, + "loss": 0.0101, + "step": 22139 + }, + { + "epoch": 2.4002601908065913, + "grad_norm": 1.7424236536026, + "learning_rate": 4.012766574786015e-06, + "loss": 0.0112, + "step": 22140 + }, + { + "epoch": 2.4003686036426712, + "grad_norm": 3.295632839202881, + "learning_rate": 4.012041201218627e-06, + "loss": 0.0299, + "step": 22141 + }, + { + "epoch": 2.4004770164787512, + "grad_norm": 3.4548964500427246, + "learning_rate": 4.011315827651241e-06, + "loss": 0.0145, + "step": 22142 + }, + { + "epoch": 2.4005854293148308, + "grad_norm": 2.996279716491699, + "learning_rate": 4.010590454083853e-06, + "loss": 0.0072, + "step": 22143 + }, + { + "epoch": 2.4006938421509108, + "grad_norm": 11.443940162658691, + "learning_rate": 4.009865080516466e-06, + "loss": 0.0076, + "step": 22144 + }, + { + "epoch": 2.4008022549869903, + "grad_norm": 5.569387435913086, + "learning_rate": 4.009139706949079e-06, + "loss": 0.0149, + "step": 22145 + }, + { + "epoch": 2.4009106678230703, + "grad_norm": 0.05080648884177208, + "learning_rate": 4.008414333381692e-06, + "loss": 0.0003, + "step": 22146 + }, + { + "epoch": 2.4010190806591503, + "grad_norm": 0.0759713351726532, + "learning_rate": 4.0076889598143045e-06, + "loss": 0.0007, + "step": 22147 + }, + { + "epoch": 2.40112749349523, + "grad_norm": 2.427647829055786, + "learning_rate": 4.0069635862469175e-06, + "loss": 0.0037, + "step": 22148 + }, + { + "epoch": 2.40123590633131, + "grad_norm": 0.20324549078941345, + "learning_rate": 4.0062382126795305e-06, + "loss": 0.0014, + "step": 22149 + }, + { + "epoch": 2.4013443191673893, + "grad_norm": 0.05112941935658455, + "learning_rate": 4.005512839112143e-06, + "loss": 0.0002, + "step": 22150 + }, + { + "epoch": 2.4014527320034693, + "grad_norm": 0.24565286934375763, + "learning_rate": 4.0047874655447565e-06, + "loss": 0.0011, + "step": 22151 + }, + { + "epoch": 2.401561144839549, + "grad_norm": 20.343948364257812, + "learning_rate": 4.004062091977369e-06, + "loss": 0.0185, + "step": 22152 + }, + { + "epoch": 2.401669557675629, + "grad_norm": 1.334948182106018, + "learning_rate": 4.003336718409982e-06, + "loss": 0.0061, + "step": 22153 + }, + { + "epoch": 2.4017779705117084, + "grad_norm": 0.2212228924036026, + "learning_rate": 4.002611344842594e-06, + "loss": 0.0008, + "step": 22154 + }, + { + "epoch": 2.4018863833477884, + "grad_norm": 0.9993439316749573, + "learning_rate": 4.0018859712752076e-06, + "loss": 0.0057, + "step": 22155 + }, + { + "epoch": 2.4019947961838684, + "grad_norm": 3.4006261825561523, + "learning_rate": 4.00116059770782e-06, + "loss": 0.0327, + "step": 22156 + }, + { + "epoch": 2.402103209019948, + "grad_norm": 0.08089546114206314, + "learning_rate": 4.000435224140433e-06, + "loss": 0.0006, + "step": 22157 + }, + { + "epoch": 2.402211621856028, + "grad_norm": 0.9995732307434082, + "learning_rate": 3.999709850573046e-06, + "loss": 0.0089, + "step": 22158 + }, + { + "epoch": 2.4023200346921074, + "grad_norm": 0.40461140871047974, + "learning_rate": 3.998984477005659e-06, + "loss": 0.0011, + "step": 22159 + }, + { + "epoch": 2.4024284475281874, + "grad_norm": 1.8029835224151611, + "learning_rate": 3.998259103438271e-06, + "loss": 0.0091, + "step": 22160 + }, + { + "epoch": 2.402536860364267, + "grad_norm": 0.10787609219551086, + "learning_rate": 3.997533729870884e-06, + "loss": 0.0007, + "step": 22161 + }, + { + "epoch": 2.402645273200347, + "grad_norm": 1.6295723915100098, + "learning_rate": 3.996808356303497e-06, + "loss": 0.0306, + "step": 22162 + }, + { + "epoch": 2.4027536860364265, + "grad_norm": 0.2062264084815979, + "learning_rate": 3.996082982736109e-06, + "loss": 0.0019, + "step": 22163 + }, + { + "epoch": 2.4028620988725065, + "grad_norm": 0.07004088908433914, + "learning_rate": 3.995357609168723e-06, + "loss": 0.0005, + "step": 22164 + }, + { + "epoch": 2.4029705117085864, + "grad_norm": 0.937610387802124, + "learning_rate": 3.994632235601335e-06, + "loss": 0.0029, + "step": 22165 + }, + { + "epoch": 2.403078924544666, + "grad_norm": 0.035370420664548874, + "learning_rate": 3.993906862033948e-06, + "loss": 0.0003, + "step": 22166 + }, + { + "epoch": 2.403187337380746, + "grad_norm": 3.542073965072632, + "learning_rate": 3.99318148846656e-06, + "loss": 0.0069, + "step": 22167 + }, + { + "epoch": 2.4032957502168255, + "grad_norm": 0.23377476632595062, + "learning_rate": 3.992456114899174e-06, + "loss": 0.0012, + "step": 22168 + }, + { + "epoch": 2.4034041630529055, + "grad_norm": 2.7063088417053223, + "learning_rate": 3.991730741331786e-06, + "loss": 0.0277, + "step": 22169 + }, + { + "epoch": 2.4035125758889855, + "grad_norm": 1.2986729145050049, + "learning_rate": 3.991005367764399e-06, + "loss": 0.0176, + "step": 22170 + }, + { + "epoch": 2.403620988725065, + "grad_norm": 0.04800191894173622, + "learning_rate": 3.990279994197012e-06, + "loss": 0.0004, + "step": 22171 + }, + { + "epoch": 2.403729401561145, + "grad_norm": 0.0024811523035168648, + "learning_rate": 3.989554620629625e-06, + "loss": 0.0, + "step": 22172 + }, + { + "epoch": 2.4038378143972245, + "grad_norm": 0.39125677943229675, + "learning_rate": 3.988829247062237e-06, + "loss": 0.0026, + "step": 22173 + }, + { + "epoch": 2.4039462272333045, + "grad_norm": 1.6440391540527344, + "learning_rate": 3.98810387349485e-06, + "loss": 0.0217, + "step": 22174 + }, + { + "epoch": 2.404054640069384, + "grad_norm": 2.1258671283721924, + "learning_rate": 3.987378499927463e-06, + "loss": 0.0029, + "step": 22175 + }, + { + "epoch": 2.404163052905464, + "grad_norm": 3.845950126647949, + "learning_rate": 3.986653126360075e-06, + "loss": 0.0226, + "step": 22176 + }, + { + "epoch": 2.4042714657415436, + "grad_norm": 0.0744309052824974, + "learning_rate": 3.985927752792689e-06, + "loss": 0.0006, + "step": 22177 + }, + { + "epoch": 2.4043798785776236, + "grad_norm": 0.285629540681839, + "learning_rate": 3.985202379225301e-06, + "loss": 0.0019, + "step": 22178 + }, + { + "epoch": 2.4044882914137036, + "grad_norm": 0.014293832704424858, + "learning_rate": 3.984477005657914e-06, + "loss": 0.0001, + "step": 22179 + }, + { + "epoch": 2.404596704249783, + "grad_norm": 2.002959966659546, + "learning_rate": 3.983751632090526e-06, + "loss": 0.0245, + "step": 22180 + }, + { + "epoch": 2.404705117085863, + "grad_norm": 0.4915343225002289, + "learning_rate": 3.98302625852314e-06, + "loss": 0.0015, + "step": 22181 + }, + { + "epoch": 2.4048135299219426, + "grad_norm": 4.286152362823486, + "learning_rate": 3.982300884955752e-06, + "loss": 0.0088, + "step": 22182 + }, + { + "epoch": 2.4049219427580226, + "grad_norm": 2.377613067626953, + "learning_rate": 3.981575511388365e-06, + "loss": 0.0039, + "step": 22183 + }, + { + "epoch": 2.405030355594102, + "grad_norm": 0.009317515417933464, + "learning_rate": 3.980850137820978e-06, + "loss": 0.0, + "step": 22184 + }, + { + "epoch": 2.405138768430182, + "grad_norm": 0.05216510221362114, + "learning_rate": 3.980124764253591e-06, + "loss": 0.0005, + "step": 22185 + }, + { + "epoch": 2.4052471812662617, + "grad_norm": 0.0397784486413002, + "learning_rate": 3.9793993906862035e-06, + "loss": 0.0002, + "step": 22186 + }, + { + "epoch": 2.4053555941023417, + "grad_norm": 0.3696039617061615, + "learning_rate": 3.9786740171188165e-06, + "loss": 0.0016, + "step": 22187 + }, + { + "epoch": 2.4054640069384217, + "grad_norm": 0.02527155540883541, + "learning_rate": 3.9779486435514295e-06, + "loss": 0.0001, + "step": 22188 + }, + { + "epoch": 2.405572419774501, + "grad_norm": 1.7844696044921875, + "learning_rate": 3.977223269984042e-06, + "loss": 0.0037, + "step": 22189 + }, + { + "epoch": 2.405680832610581, + "grad_norm": 2.125197649002075, + "learning_rate": 3.9764978964166555e-06, + "loss": 0.0038, + "step": 22190 + }, + { + "epoch": 2.4057892454466607, + "grad_norm": 0.039729200303554535, + "learning_rate": 3.975772522849268e-06, + "loss": 0.0004, + "step": 22191 + }, + { + "epoch": 2.4058976582827407, + "grad_norm": 4.9385986328125, + "learning_rate": 3.975047149281881e-06, + "loss": 0.0366, + "step": 22192 + }, + { + "epoch": 2.4060060711188207, + "grad_norm": 0.2172917276620865, + "learning_rate": 3.974321775714493e-06, + "loss": 0.0012, + "step": 22193 + }, + { + "epoch": 2.4061144839549002, + "grad_norm": 1.806091547012329, + "learning_rate": 3.9735964021471066e-06, + "loss": 0.0029, + "step": 22194 + }, + { + "epoch": 2.4062228967909802, + "grad_norm": 2.9737977981567383, + "learning_rate": 3.972871028579719e-06, + "loss": 0.0009, + "step": 22195 + }, + { + "epoch": 2.4063313096270598, + "grad_norm": 1.733555555343628, + "learning_rate": 3.972145655012332e-06, + "loss": 0.0023, + "step": 22196 + }, + { + "epoch": 2.4064397224631398, + "grad_norm": 2.376164674758911, + "learning_rate": 3.971420281444945e-06, + "loss": 0.0261, + "step": 22197 + }, + { + "epoch": 2.4065481352992193, + "grad_norm": 0.7712901830673218, + "learning_rate": 3.970694907877558e-06, + "loss": 0.0037, + "step": 22198 + }, + { + "epoch": 2.4066565481352993, + "grad_norm": 3.5062403678894043, + "learning_rate": 3.96996953431017e-06, + "loss": 0.0048, + "step": 22199 + }, + { + "epoch": 2.406764960971379, + "grad_norm": 0.18036451935768127, + "learning_rate": 3.969244160742783e-06, + "loss": 0.0011, + "step": 22200 + }, + { + "epoch": 2.406873373807459, + "grad_norm": 0.5483382344245911, + "learning_rate": 3.968518787175396e-06, + "loss": 0.0021, + "step": 22201 + }, + { + "epoch": 2.406981786643539, + "grad_norm": 0.7242801189422607, + "learning_rate": 3.967793413608008e-06, + "loss": 0.0016, + "step": 22202 + }, + { + "epoch": 2.4070901994796183, + "grad_norm": 0.03891483321785927, + "learning_rate": 3.967068040040622e-06, + "loss": 0.0003, + "step": 22203 + }, + { + "epoch": 2.4071986123156983, + "grad_norm": 2.8324499130249023, + "learning_rate": 3.966342666473234e-06, + "loss": 0.03, + "step": 22204 + }, + { + "epoch": 2.407307025151778, + "grad_norm": 0.09489352256059647, + "learning_rate": 3.965617292905847e-06, + "loss": 0.0007, + "step": 22205 + }, + { + "epoch": 2.407415437987858, + "grad_norm": 0.8337063193321228, + "learning_rate": 3.964891919338459e-06, + "loss": 0.028, + "step": 22206 + }, + { + "epoch": 2.4075238508239374, + "grad_norm": 0.35277706384658813, + "learning_rate": 3.964166545771073e-06, + "loss": 0.0004, + "step": 22207 + }, + { + "epoch": 2.4076322636600174, + "grad_norm": 1.6514207124710083, + "learning_rate": 3.963441172203685e-06, + "loss": 0.0072, + "step": 22208 + }, + { + "epoch": 2.407740676496097, + "grad_norm": 0.1385488510131836, + "learning_rate": 3.962715798636298e-06, + "loss": 0.0009, + "step": 22209 + }, + { + "epoch": 2.407849089332177, + "grad_norm": 5.842390060424805, + "learning_rate": 3.961990425068911e-06, + "loss": 0.0076, + "step": 22210 + }, + { + "epoch": 2.407957502168257, + "grad_norm": 0.10929542034864426, + "learning_rate": 3.961265051501524e-06, + "loss": 0.0004, + "step": 22211 + }, + { + "epoch": 2.4080659150043364, + "grad_norm": 25.759790420532227, + "learning_rate": 3.960539677934136e-06, + "loss": 0.0059, + "step": 22212 + }, + { + "epoch": 2.4081743278404164, + "grad_norm": 1.133802056312561, + "learning_rate": 3.959814304366749e-06, + "loss": 0.0032, + "step": 22213 + }, + { + "epoch": 2.408282740676496, + "grad_norm": 0.1618087738752365, + "learning_rate": 3.959088930799362e-06, + "loss": 0.001, + "step": 22214 + }, + { + "epoch": 2.408391153512576, + "grad_norm": 2.915104389190674, + "learning_rate": 3.958363557231974e-06, + "loss": 0.0134, + "step": 22215 + }, + { + "epoch": 2.408499566348656, + "grad_norm": 3.786130905151367, + "learning_rate": 3.957638183664588e-06, + "loss": 0.0105, + "step": 22216 + }, + { + "epoch": 2.4086079791847355, + "grad_norm": 3.455125093460083, + "learning_rate": 3.9569128100972e-06, + "loss": 0.0106, + "step": 22217 + }, + { + "epoch": 2.4087163920208154, + "grad_norm": 3.428865671157837, + "learning_rate": 3.956187436529813e-06, + "loss": 0.0418, + "step": 22218 + }, + { + "epoch": 2.408824804856895, + "grad_norm": 5.42294454574585, + "learning_rate": 3.955462062962425e-06, + "loss": 0.0126, + "step": 22219 + }, + { + "epoch": 2.408933217692975, + "grad_norm": 0.6338821649551392, + "learning_rate": 3.954736689395039e-06, + "loss": 0.0011, + "step": 22220 + }, + { + "epoch": 2.4090416305290545, + "grad_norm": 0.1264972686767578, + "learning_rate": 3.954011315827651e-06, + "loss": 0.0006, + "step": 22221 + }, + { + "epoch": 2.4091500433651345, + "grad_norm": 0.031167732551693916, + "learning_rate": 3.953285942260264e-06, + "loss": 0.0002, + "step": 22222 + }, + { + "epoch": 2.409258456201214, + "grad_norm": 5.288792610168457, + "learning_rate": 3.952560568692877e-06, + "loss": 0.0087, + "step": 22223 + }, + { + "epoch": 2.409366869037294, + "grad_norm": 0.05984394624829292, + "learning_rate": 3.95183519512549e-06, + "loss": 0.0001, + "step": 22224 + }, + { + "epoch": 2.409475281873374, + "grad_norm": 0.2888372242450714, + "learning_rate": 3.9511098215581025e-06, + "loss": 0.0021, + "step": 22225 + }, + { + "epoch": 2.4095836947094535, + "grad_norm": 1.3003926277160645, + "learning_rate": 3.9503844479907155e-06, + "loss": 0.0027, + "step": 22226 + }, + { + "epoch": 2.4096921075455335, + "grad_norm": 0.02417776733636856, + "learning_rate": 3.9496590744233285e-06, + "loss": 0.0001, + "step": 22227 + }, + { + "epoch": 2.409800520381613, + "grad_norm": 2.7448394298553467, + "learning_rate": 3.9489337008559415e-06, + "loss": 0.01, + "step": 22228 + }, + { + "epoch": 2.409908933217693, + "grad_norm": 1.9262821674346924, + "learning_rate": 3.9482083272885545e-06, + "loss": 0.0137, + "step": 22229 + }, + { + "epoch": 2.4100173460537726, + "grad_norm": 1.1993107795715332, + "learning_rate": 3.947482953721167e-06, + "loss": 0.0034, + "step": 22230 + }, + { + "epoch": 2.4101257588898526, + "grad_norm": 0.09667262434959412, + "learning_rate": 3.94675758015378e-06, + "loss": 0.0004, + "step": 22231 + }, + { + "epoch": 2.410234171725932, + "grad_norm": 0.7125680446624756, + "learning_rate": 3.946032206586392e-06, + "loss": 0.0064, + "step": 22232 + }, + { + "epoch": 2.410342584562012, + "grad_norm": 1.8185930252075195, + "learning_rate": 3.9453068330190056e-06, + "loss": 0.0068, + "step": 22233 + }, + { + "epoch": 2.410450997398092, + "grad_norm": 2.852306365966797, + "learning_rate": 3.944581459451618e-06, + "loss": 0.0026, + "step": 22234 + }, + { + "epoch": 2.4105594102341716, + "grad_norm": 1.8324787616729736, + "learning_rate": 3.943856085884231e-06, + "loss": 0.0042, + "step": 22235 + }, + { + "epoch": 2.4106678230702516, + "grad_norm": 0.14572401344776154, + "learning_rate": 3.943130712316844e-06, + "loss": 0.0016, + "step": 22236 + }, + { + "epoch": 2.410776235906331, + "grad_norm": 2.058230400085449, + "learning_rate": 3.942405338749457e-06, + "loss": 0.0059, + "step": 22237 + }, + { + "epoch": 2.410884648742411, + "grad_norm": 0.23293183743953705, + "learning_rate": 3.941679965182069e-06, + "loss": 0.0006, + "step": 22238 + }, + { + "epoch": 2.410993061578491, + "grad_norm": 4.939637660980225, + "learning_rate": 3.940954591614682e-06, + "loss": 0.0125, + "step": 22239 + }, + { + "epoch": 2.4111014744145707, + "grad_norm": 6.88602876663208, + "learning_rate": 3.940229218047295e-06, + "loss": 0.0827, + "step": 22240 + }, + { + "epoch": 2.4112098872506507, + "grad_norm": 1.9270830154418945, + "learning_rate": 3.939503844479908e-06, + "loss": 0.0217, + "step": 22241 + }, + { + "epoch": 2.41131830008673, + "grad_norm": 0.36783063411712646, + "learning_rate": 3.938778470912521e-06, + "loss": 0.0012, + "step": 22242 + }, + { + "epoch": 2.41142671292281, + "grad_norm": 15.185234069824219, + "learning_rate": 3.938053097345133e-06, + "loss": 0.0073, + "step": 22243 + }, + { + "epoch": 2.4115351257588897, + "grad_norm": 0.15029248595237732, + "learning_rate": 3.937327723777746e-06, + "loss": 0.0013, + "step": 22244 + }, + { + "epoch": 2.4116435385949697, + "grad_norm": 2.6526010036468506, + "learning_rate": 3.936602350210358e-06, + "loss": 0.0085, + "step": 22245 + }, + { + "epoch": 2.4117519514310493, + "grad_norm": 0.017044415697455406, + "learning_rate": 3.935876976642972e-06, + "loss": 0.0002, + "step": 22246 + }, + { + "epoch": 2.4118603642671292, + "grad_norm": 0.195074662566185, + "learning_rate": 3.935151603075584e-06, + "loss": 0.0007, + "step": 22247 + }, + { + "epoch": 2.4119687771032092, + "grad_norm": 5.838161945343018, + "learning_rate": 3.934426229508197e-06, + "loss": 0.0218, + "step": 22248 + }, + { + "epoch": 2.4120771899392888, + "grad_norm": 2.661498546600342, + "learning_rate": 3.93370085594081e-06, + "loss": 0.0045, + "step": 22249 + }, + { + "epoch": 2.4121856027753688, + "grad_norm": 1.4570313692092896, + "learning_rate": 3.932975482373423e-06, + "loss": 0.0046, + "step": 22250 + }, + { + "epoch": 2.4122940156114483, + "grad_norm": 0.2593940794467926, + "learning_rate": 3.932250108806035e-06, + "loss": 0.0009, + "step": 22251 + }, + { + "epoch": 2.4124024284475283, + "grad_norm": 0.3362039029598236, + "learning_rate": 3.931524735238648e-06, + "loss": 0.001, + "step": 22252 + }, + { + "epoch": 2.412510841283608, + "grad_norm": 2.3625497817993164, + "learning_rate": 3.930799361671261e-06, + "loss": 0.0017, + "step": 22253 + }, + { + "epoch": 2.412619254119688, + "grad_norm": 3.129382848739624, + "learning_rate": 3.930073988103874e-06, + "loss": 0.0162, + "step": 22254 + }, + { + "epoch": 2.4127276669557673, + "grad_norm": 22.674577713012695, + "learning_rate": 3.929348614536487e-06, + "loss": 0.0304, + "step": 22255 + }, + { + "epoch": 2.4128360797918473, + "grad_norm": 0.3402901589870453, + "learning_rate": 3.928623240969099e-06, + "loss": 0.0014, + "step": 22256 + }, + { + "epoch": 2.4129444926279273, + "grad_norm": 0.006179936230182648, + "learning_rate": 3.927897867401712e-06, + "loss": 0.0001, + "step": 22257 + }, + { + "epoch": 2.413052905464007, + "grad_norm": 0.007671798579394817, + "learning_rate": 3.927172493834324e-06, + "loss": 0.0001, + "step": 22258 + }, + { + "epoch": 2.413161318300087, + "grad_norm": 3.3839731216430664, + "learning_rate": 3.926447120266938e-06, + "loss": 0.0107, + "step": 22259 + }, + { + "epoch": 2.4132697311361664, + "grad_norm": 0.4306625723838806, + "learning_rate": 3.92572174669955e-06, + "loss": 0.0011, + "step": 22260 + }, + { + "epoch": 2.4133781439722464, + "grad_norm": 0.001450481591746211, + "learning_rate": 3.924996373132163e-06, + "loss": 0.0, + "step": 22261 + }, + { + "epoch": 2.4134865568083264, + "grad_norm": 3.017404556274414, + "learning_rate": 3.924270999564776e-06, + "loss": 0.0105, + "step": 22262 + }, + { + "epoch": 2.413594969644406, + "grad_norm": 11.440086364746094, + "learning_rate": 3.923545625997389e-06, + "loss": 0.0271, + "step": 22263 + }, + { + "epoch": 2.413703382480486, + "grad_norm": 7.635822296142578, + "learning_rate": 3.9228202524300015e-06, + "loss": 0.0039, + "step": 22264 + }, + { + "epoch": 2.4138117953165654, + "grad_norm": 2.5112576484680176, + "learning_rate": 3.9220948788626145e-06, + "loss": 0.016, + "step": 22265 + }, + { + "epoch": 2.4139202081526454, + "grad_norm": 0.14071042835712433, + "learning_rate": 3.9213695052952275e-06, + "loss": 0.001, + "step": 22266 + }, + { + "epoch": 2.414028620988725, + "grad_norm": 5.513850212097168, + "learning_rate": 3.9206441317278405e-06, + "loss": 0.0338, + "step": 22267 + }, + { + "epoch": 2.414137033824805, + "grad_norm": 1.5656815767288208, + "learning_rate": 3.9199187581604535e-06, + "loss": 0.0135, + "step": 22268 + }, + { + "epoch": 2.4142454466608845, + "grad_norm": 0.0536300428211689, + "learning_rate": 3.919193384593066e-06, + "loss": 0.0001, + "step": 22269 + }, + { + "epoch": 2.4143538594969645, + "grad_norm": 0.5832233428955078, + "learning_rate": 3.918468011025679e-06, + "loss": 0.0029, + "step": 22270 + }, + { + "epoch": 2.4144622723330444, + "grad_norm": 1.6615042686462402, + "learning_rate": 3.917742637458291e-06, + "loss": 0.0667, + "step": 22271 + }, + { + "epoch": 2.414570685169124, + "grad_norm": 0.075448177754879, + "learning_rate": 3.917017263890905e-06, + "loss": 0.0007, + "step": 22272 + }, + { + "epoch": 2.414679098005204, + "grad_norm": 0.06704612076282501, + "learning_rate": 3.916291890323517e-06, + "loss": 0.0009, + "step": 22273 + }, + { + "epoch": 2.4147875108412835, + "grad_norm": 4.282894611358643, + "learning_rate": 3.91556651675613e-06, + "loss": 0.0426, + "step": 22274 + }, + { + "epoch": 2.4148959236773635, + "grad_norm": 0.6508618593215942, + "learning_rate": 3.914841143188743e-06, + "loss": 0.0036, + "step": 22275 + }, + { + "epoch": 2.415004336513443, + "grad_norm": 9.510993957519531, + "learning_rate": 3.914115769621356e-06, + "loss": 0.0193, + "step": 22276 + }, + { + "epoch": 2.415112749349523, + "grad_norm": 2.4099297523498535, + "learning_rate": 3.913390396053968e-06, + "loss": 0.0024, + "step": 22277 + }, + { + "epoch": 2.4152211621856026, + "grad_norm": 0.07133053243160248, + "learning_rate": 3.912665022486581e-06, + "loss": 0.0005, + "step": 22278 + }, + { + "epoch": 2.4153295750216826, + "grad_norm": 0.35436883568763733, + "learning_rate": 3.911939648919194e-06, + "loss": 0.0012, + "step": 22279 + }, + { + "epoch": 2.4154379878577625, + "grad_norm": 4.594264030456543, + "learning_rate": 3.911214275351807e-06, + "loss": 0.0099, + "step": 22280 + }, + { + "epoch": 2.415546400693842, + "grad_norm": 2.5775060653686523, + "learning_rate": 3.91048890178442e-06, + "loss": 0.0501, + "step": 22281 + }, + { + "epoch": 2.415654813529922, + "grad_norm": 0.3535000681877136, + "learning_rate": 3.909763528217032e-06, + "loss": 0.0004, + "step": 22282 + }, + { + "epoch": 2.4157632263660016, + "grad_norm": 2.2228446006774902, + "learning_rate": 3.909038154649645e-06, + "loss": 0.0119, + "step": 22283 + }, + { + "epoch": 2.4158716392020816, + "grad_norm": 5.484980583190918, + "learning_rate": 3.908312781082257e-06, + "loss": 0.018, + "step": 22284 + }, + { + "epoch": 2.415980052038161, + "grad_norm": 0.09197817742824554, + "learning_rate": 3.907587407514871e-06, + "loss": 0.0011, + "step": 22285 + }, + { + "epoch": 2.416088464874241, + "grad_norm": 0.9274278283119202, + "learning_rate": 3.906862033947483e-06, + "loss": 0.0035, + "step": 22286 + }, + { + "epoch": 2.4161968777103207, + "grad_norm": 0.019152075052261353, + "learning_rate": 3.906136660380096e-06, + "loss": 0.0002, + "step": 22287 + }, + { + "epoch": 2.4163052905464006, + "grad_norm": 2.4322378635406494, + "learning_rate": 3.905411286812709e-06, + "loss": 0.0512, + "step": 22288 + }, + { + "epoch": 2.4164137033824806, + "grad_norm": 2.2871382236480713, + "learning_rate": 3.904685913245322e-06, + "loss": 0.0036, + "step": 22289 + }, + { + "epoch": 2.41652211621856, + "grad_norm": 10.40845775604248, + "learning_rate": 3.903960539677934e-06, + "loss": 0.0408, + "step": 22290 + }, + { + "epoch": 2.41663052905464, + "grad_norm": 2.1586368083953857, + "learning_rate": 3.903235166110547e-06, + "loss": 0.0035, + "step": 22291 + }, + { + "epoch": 2.4167389418907197, + "grad_norm": 3.2537424564361572, + "learning_rate": 3.90250979254316e-06, + "loss": 0.0282, + "step": 22292 + }, + { + "epoch": 2.4168473547267997, + "grad_norm": 5.926872730255127, + "learning_rate": 3.901784418975773e-06, + "loss": 0.0066, + "step": 22293 + }, + { + "epoch": 2.4169557675628797, + "grad_norm": 1.590756893157959, + "learning_rate": 3.901059045408386e-06, + "loss": 0.0039, + "step": 22294 + }, + { + "epoch": 2.417064180398959, + "grad_norm": 4.4187703132629395, + "learning_rate": 3.900333671840998e-06, + "loss": 0.0488, + "step": 22295 + }, + { + "epoch": 2.417172593235039, + "grad_norm": 0.03966151177883148, + "learning_rate": 3.899608298273611e-06, + "loss": 0.0002, + "step": 22296 + }, + { + "epoch": 2.4172810060711187, + "grad_norm": 2.5992627143859863, + "learning_rate": 3.898882924706223e-06, + "loss": 0.0013, + "step": 22297 + }, + { + "epoch": 2.4173894189071987, + "grad_norm": 0.5044201612472534, + "learning_rate": 3.898157551138837e-06, + "loss": 0.0023, + "step": 22298 + }, + { + "epoch": 2.4174978317432783, + "grad_norm": 0.05941058322787285, + "learning_rate": 3.897432177571449e-06, + "loss": 0.0005, + "step": 22299 + }, + { + "epoch": 2.4176062445793582, + "grad_norm": 0.5663232207298279, + "learning_rate": 3.896706804004062e-06, + "loss": 0.0019, + "step": 22300 + }, + { + "epoch": 2.417714657415438, + "grad_norm": 2.502142906188965, + "learning_rate": 3.895981430436675e-06, + "loss": 0.0135, + "step": 22301 + }, + { + "epoch": 2.4178230702515178, + "grad_norm": 0.35419946908950806, + "learning_rate": 3.895256056869288e-06, + "loss": 0.0005, + "step": 22302 + }, + { + "epoch": 2.4179314830875978, + "grad_norm": 0.16516146063804626, + "learning_rate": 3.8945306833019005e-06, + "loss": 0.0011, + "step": 22303 + }, + { + "epoch": 2.4180398959236773, + "grad_norm": 1.9849847555160522, + "learning_rate": 3.8938053097345135e-06, + "loss": 0.0053, + "step": 22304 + }, + { + "epoch": 2.4181483087597573, + "grad_norm": 2.1811397075653076, + "learning_rate": 3.8930799361671265e-06, + "loss": 0.0059, + "step": 22305 + }, + { + "epoch": 2.418256721595837, + "grad_norm": 0.06019042804837227, + "learning_rate": 3.8923545625997395e-06, + "loss": 0.0005, + "step": 22306 + }, + { + "epoch": 2.418365134431917, + "grad_norm": 0.08119890093803406, + "learning_rate": 3.8916291890323525e-06, + "loss": 0.0003, + "step": 22307 + }, + { + "epoch": 2.4184735472679963, + "grad_norm": 1.344725489616394, + "learning_rate": 3.890903815464965e-06, + "loss": 0.0051, + "step": 22308 + }, + { + "epoch": 2.4185819601040763, + "grad_norm": 5.6344828605651855, + "learning_rate": 3.890178441897578e-06, + "loss": 0.045, + "step": 22309 + }, + { + "epoch": 2.418690372940156, + "grad_norm": 0.09155447036027908, + "learning_rate": 3.88945306833019e-06, + "loss": 0.0009, + "step": 22310 + }, + { + "epoch": 2.418798785776236, + "grad_norm": 0.0385260246694088, + "learning_rate": 3.888727694762804e-06, + "loss": 0.0001, + "step": 22311 + }, + { + "epoch": 2.418907198612316, + "grad_norm": 0.34099581837654114, + "learning_rate": 3.888002321195416e-06, + "loss": 0.0005, + "step": 22312 + }, + { + "epoch": 2.4190156114483954, + "grad_norm": 1.1217167377471924, + "learning_rate": 3.887276947628029e-06, + "loss": 0.0036, + "step": 22313 + }, + { + "epoch": 2.4191240242844754, + "grad_norm": 1.5559884309768677, + "learning_rate": 3.886551574060642e-06, + "loss": 0.0249, + "step": 22314 + }, + { + "epoch": 2.419232437120555, + "grad_norm": 1.6185897588729858, + "learning_rate": 3.885826200493255e-06, + "loss": 0.0073, + "step": 22315 + }, + { + "epoch": 2.419340849956635, + "grad_norm": 2.6165611743927, + "learning_rate": 3.885100826925867e-06, + "loss": 0.0421, + "step": 22316 + }, + { + "epoch": 2.419449262792715, + "grad_norm": 0.007133503910154104, + "learning_rate": 3.88437545335848e-06, + "loss": 0.0001, + "step": 22317 + }, + { + "epoch": 2.4195576756287944, + "grad_norm": 0.14058029651641846, + "learning_rate": 3.883650079791093e-06, + "loss": 0.0011, + "step": 22318 + }, + { + "epoch": 2.4196660884648744, + "grad_norm": 3.130983591079712, + "learning_rate": 3.882924706223706e-06, + "loss": 0.0415, + "step": 22319 + }, + { + "epoch": 2.419774501300954, + "grad_norm": 1.3264752626419067, + "learning_rate": 3.882199332656319e-06, + "loss": 0.0033, + "step": 22320 + }, + { + "epoch": 2.419882914137034, + "grad_norm": 0.03038286603987217, + "learning_rate": 3.881473959088931e-06, + "loss": 0.0002, + "step": 22321 + }, + { + "epoch": 2.4199913269731135, + "grad_norm": 0.008493293076753616, + "learning_rate": 3.880748585521544e-06, + "loss": 0.0001, + "step": 22322 + }, + { + "epoch": 2.4200997398091935, + "grad_norm": 5.184611797332764, + "learning_rate": 3.880023211954156e-06, + "loss": 0.0297, + "step": 22323 + }, + { + "epoch": 2.420208152645273, + "grad_norm": 1.7130546569824219, + "learning_rate": 3.87929783838677e-06, + "loss": 0.039, + "step": 22324 + }, + { + "epoch": 2.420316565481353, + "grad_norm": 1.792041301727295, + "learning_rate": 3.878572464819382e-06, + "loss": 0.017, + "step": 22325 + }, + { + "epoch": 2.420424978317433, + "grad_norm": 12.903592109680176, + "learning_rate": 3.877847091251995e-06, + "loss": 0.0222, + "step": 22326 + }, + { + "epoch": 2.4205333911535125, + "grad_norm": 6.713070392608643, + "learning_rate": 3.877121717684608e-06, + "loss": 0.032, + "step": 22327 + }, + { + "epoch": 2.4206418039895925, + "grad_norm": 0.019299156963825226, + "learning_rate": 3.876396344117221e-06, + "loss": 0.0001, + "step": 22328 + }, + { + "epoch": 2.420750216825672, + "grad_norm": 0.6250702738761902, + "learning_rate": 3.875670970549833e-06, + "loss": 0.0008, + "step": 22329 + }, + { + "epoch": 2.420858629661752, + "grad_norm": 3.924182415008545, + "learning_rate": 3.874945596982446e-06, + "loss": 0.0027, + "step": 22330 + }, + { + "epoch": 2.4209670424978316, + "grad_norm": 0.601725161075592, + "learning_rate": 3.874220223415059e-06, + "loss": 0.0022, + "step": 22331 + }, + { + "epoch": 2.4210754553339116, + "grad_norm": 2.795243740081787, + "learning_rate": 3.873494849847672e-06, + "loss": 0.0034, + "step": 22332 + }, + { + "epoch": 2.421183868169991, + "grad_norm": 3.731321334838867, + "learning_rate": 3.872769476280285e-06, + "loss": 0.0091, + "step": 22333 + }, + { + "epoch": 2.421292281006071, + "grad_norm": 1.049645185470581, + "learning_rate": 3.872044102712897e-06, + "loss": 0.0321, + "step": 22334 + }, + { + "epoch": 2.421400693842151, + "grad_norm": 4.251276969909668, + "learning_rate": 3.87131872914551e-06, + "loss": 0.0267, + "step": 22335 + }, + { + "epoch": 2.4215091066782306, + "grad_norm": 4.413351058959961, + "learning_rate": 3.870593355578123e-06, + "loss": 0.0154, + "step": 22336 + }, + { + "epoch": 2.4216175195143106, + "grad_norm": 0.7689018845558167, + "learning_rate": 3.869867982010736e-06, + "loss": 0.002, + "step": 22337 + }, + { + "epoch": 2.42172593235039, + "grad_norm": 1.1816543340682983, + "learning_rate": 3.869142608443348e-06, + "loss": 0.0228, + "step": 22338 + }, + { + "epoch": 2.42183434518647, + "grad_norm": 0.7495330572128296, + "learning_rate": 3.868417234875961e-06, + "loss": 0.0019, + "step": 22339 + }, + { + "epoch": 2.42194275802255, + "grad_norm": 1.9366331100463867, + "learning_rate": 3.867691861308574e-06, + "loss": 0.0068, + "step": 22340 + }, + { + "epoch": 2.4220511708586296, + "grad_norm": 0.19741451740264893, + "learning_rate": 3.866966487741187e-06, + "loss": 0.001, + "step": 22341 + }, + { + "epoch": 2.4221595836947096, + "grad_norm": 7.787360191345215, + "learning_rate": 3.8662411141737995e-06, + "loss": 0.0287, + "step": 22342 + }, + { + "epoch": 2.422267996530789, + "grad_norm": 0.21366651356220245, + "learning_rate": 3.8655157406064125e-06, + "loss": 0.001, + "step": 22343 + }, + { + "epoch": 2.422376409366869, + "grad_norm": 0.061967212706804276, + "learning_rate": 3.8647903670390255e-06, + "loss": 0.0005, + "step": 22344 + }, + { + "epoch": 2.4224848222029487, + "grad_norm": 0.21989406645298004, + "learning_rate": 3.8640649934716385e-06, + "loss": 0.0008, + "step": 22345 + }, + { + "epoch": 2.4225932350390287, + "grad_norm": 0.004399886354804039, + "learning_rate": 3.8633396199042515e-06, + "loss": 0.0001, + "step": 22346 + }, + { + "epoch": 2.422701647875108, + "grad_norm": 10.296662330627441, + "learning_rate": 3.862614246336864e-06, + "loss": 0.0033, + "step": 22347 + }, + { + "epoch": 2.422810060711188, + "grad_norm": 1.6984446048736572, + "learning_rate": 3.861888872769477e-06, + "loss": 0.005, + "step": 22348 + }, + { + "epoch": 2.422918473547268, + "grad_norm": 0.08527308702468872, + "learning_rate": 3.86116349920209e-06, + "loss": 0.0004, + "step": 22349 + }, + { + "epoch": 2.4230268863833477, + "grad_norm": 3.2388908863067627, + "learning_rate": 3.860438125634703e-06, + "loss": 0.0112, + "step": 22350 + }, + { + "epoch": 2.4231352992194277, + "grad_norm": 0.16994744539260864, + "learning_rate": 3.859712752067315e-06, + "loss": 0.0005, + "step": 22351 + }, + { + "epoch": 2.4232437120555073, + "grad_norm": 0.28503164649009705, + "learning_rate": 3.858987378499928e-06, + "loss": 0.0009, + "step": 22352 + }, + { + "epoch": 2.4233521248915872, + "grad_norm": 4.352847099304199, + "learning_rate": 3.858262004932541e-06, + "loss": 0.0085, + "step": 22353 + }, + { + "epoch": 2.423460537727667, + "grad_norm": 2.028169870376587, + "learning_rate": 3.857536631365154e-06, + "loss": 0.0172, + "step": 22354 + }, + { + "epoch": 2.4235689505637468, + "grad_norm": 0.25102508068084717, + "learning_rate": 3.856811257797766e-06, + "loss": 0.0007, + "step": 22355 + }, + { + "epoch": 2.4236773633998263, + "grad_norm": 0.03929496183991432, + "learning_rate": 3.856085884230379e-06, + "loss": 0.0004, + "step": 22356 + }, + { + "epoch": 2.4237857762359063, + "grad_norm": 0.3720194399356842, + "learning_rate": 3.855360510662992e-06, + "loss": 0.0009, + "step": 22357 + }, + { + "epoch": 2.4238941890719863, + "grad_norm": 0.2782532274723053, + "learning_rate": 3.854635137095605e-06, + "loss": 0.0015, + "step": 22358 + }, + { + "epoch": 2.424002601908066, + "grad_norm": 6.975821495056152, + "learning_rate": 3.853909763528218e-06, + "loss": 0.0036, + "step": 22359 + }, + { + "epoch": 2.424111014744146, + "grad_norm": 0.939598023891449, + "learning_rate": 3.85318438996083e-06, + "loss": 0.0022, + "step": 22360 + }, + { + "epoch": 2.4242194275802254, + "grad_norm": 0.0728030800819397, + "learning_rate": 3.852459016393443e-06, + "loss": 0.0009, + "step": 22361 + }, + { + "epoch": 2.4243278404163053, + "grad_norm": 1.457308292388916, + "learning_rate": 3.851733642826056e-06, + "loss": 0.0429, + "step": 22362 + }, + { + "epoch": 2.4244362532523853, + "grad_norm": 5.582862377166748, + "learning_rate": 3.851008269258669e-06, + "loss": 0.0494, + "step": 22363 + }, + { + "epoch": 2.424544666088465, + "grad_norm": 0.007759009022265673, + "learning_rate": 3.850282895691281e-06, + "loss": 0.0001, + "step": 22364 + }, + { + "epoch": 2.424653078924545, + "grad_norm": 1.9100311994552612, + "learning_rate": 3.849557522123894e-06, + "loss": 0.0285, + "step": 22365 + }, + { + "epoch": 2.4247614917606244, + "grad_norm": 0.08198747783899307, + "learning_rate": 3.848832148556507e-06, + "loss": 0.0003, + "step": 22366 + }, + { + "epoch": 2.4248699045967044, + "grad_norm": 0.027974288910627365, + "learning_rate": 3.84810677498912e-06, + "loss": 0.0003, + "step": 22367 + }, + { + "epoch": 2.424978317432784, + "grad_norm": 4.596686840057373, + "learning_rate": 3.847381401421732e-06, + "loss": 0.0118, + "step": 22368 + }, + { + "epoch": 2.425086730268864, + "grad_norm": 0.009947129525244236, + "learning_rate": 3.846656027854345e-06, + "loss": 0.0001, + "step": 22369 + }, + { + "epoch": 2.4251951431049434, + "grad_norm": 1.7992738485336304, + "learning_rate": 3.845930654286958e-06, + "loss": 0.0091, + "step": 22370 + }, + { + "epoch": 2.4253035559410234, + "grad_norm": 1.2233169078826904, + "learning_rate": 3.845205280719571e-06, + "loss": 0.0042, + "step": 22371 + }, + { + "epoch": 2.4254119687771034, + "grad_norm": 1.0799380540847778, + "learning_rate": 3.844479907152184e-06, + "loss": 0.0026, + "step": 22372 + }, + { + "epoch": 2.425520381613183, + "grad_norm": 0.38803601264953613, + "learning_rate": 3.843754533584796e-06, + "loss": 0.0025, + "step": 22373 + }, + { + "epoch": 2.425628794449263, + "grad_norm": 2.7968554496765137, + "learning_rate": 3.843029160017409e-06, + "loss": 0.0091, + "step": 22374 + }, + { + "epoch": 2.4257372072853425, + "grad_norm": 2.6566219329833984, + "learning_rate": 3.842303786450022e-06, + "loss": 0.0078, + "step": 22375 + }, + { + "epoch": 2.4258456201214225, + "grad_norm": 0.01991698332130909, + "learning_rate": 3.841578412882635e-06, + "loss": 0.0001, + "step": 22376 + }, + { + "epoch": 2.425954032957502, + "grad_norm": 4.511301517486572, + "learning_rate": 3.840853039315247e-06, + "loss": 0.0096, + "step": 22377 + }, + { + "epoch": 2.426062445793582, + "grad_norm": 0.13102753460407257, + "learning_rate": 3.84012766574786e-06, + "loss": 0.0007, + "step": 22378 + }, + { + "epoch": 2.4261708586296615, + "grad_norm": 11.263740539550781, + "learning_rate": 3.839402292180473e-06, + "loss": 0.0147, + "step": 22379 + }, + { + "epoch": 2.4262792714657415, + "grad_norm": 0.14165407419204712, + "learning_rate": 3.838676918613086e-06, + "loss": 0.0006, + "step": 22380 + }, + { + "epoch": 2.4263876843018215, + "grad_norm": 0.012108631432056427, + "learning_rate": 3.8379515450456985e-06, + "loss": 0.0001, + "step": 22381 + }, + { + "epoch": 2.426496097137901, + "grad_norm": 0.6198515295982361, + "learning_rate": 3.8372261714783115e-06, + "loss": 0.0013, + "step": 22382 + }, + { + "epoch": 2.426604509973981, + "grad_norm": 0.17319969832897186, + "learning_rate": 3.8365007979109245e-06, + "loss": 0.0013, + "step": 22383 + }, + { + "epoch": 2.4267129228100606, + "grad_norm": 1.1182308197021484, + "learning_rate": 3.8357754243435375e-06, + "loss": 0.0054, + "step": 22384 + }, + { + "epoch": 2.4268213356461406, + "grad_norm": 0.31704285740852356, + "learning_rate": 3.8350500507761505e-06, + "loss": 0.0014, + "step": 22385 + }, + { + "epoch": 2.4269297484822205, + "grad_norm": 0.3053291141986847, + "learning_rate": 3.834324677208763e-06, + "loss": 0.0007, + "step": 22386 + }, + { + "epoch": 2.4270381613183, + "grad_norm": 1.720771074295044, + "learning_rate": 3.833599303641376e-06, + "loss": 0.0043, + "step": 22387 + }, + { + "epoch": 2.42714657415438, + "grad_norm": 0.4474078118801117, + "learning_rate": 3.832873930073989e-06, + "loss": 0.0024, + "step": 22388 + }, + { + "epoch": 2.4272549869904596, + "grad_norm": 2.430803060531616, + "learning_rate": 3.832148556506602e-06, + "loss": 0.0256, + "step": 22389 + }, + { + "epoch": 2.4273633998265396, + "grad_norm": 0.1588883399963379, + "learning_rate": 3.831423182939214e-06, + "loss": 0.0007, + "step": 22390 + }, + { + "epoch": 2.427471812662619, + "grad_norm": 22.19011878967285, + "learning_rate": 3.830697809371827e-06, + "loss": 0.0353, + "step": 22391 + }, + { + "epoch": 2.427580225498699, + "grad_norm": 0.5067956447601318, + "learning_rate": 3.82997243580444e-06, + "loss": 0.0023, + "step": 22392 + }, + { + "epoch": 2.4276886383347787, + "grad_norm": 4.323235988616943, + "learning_rate": 3.829247062237053e-06, + "loss": 0.0286, + "step": 22393 + }, + { + "epoch": 2.4277970511708586, + "grad_norm": 3.0198612213134766, + "learning_rate": 3.828521688669665e-06, + "loss": 0.0197, + "step": 22394 + }, + { + "epoch": 2.4279054640069386, + "grad_norm": 19.888816833496094, + "learning_rate": 3.827796315102278e-06, + "loss": 0.0619, + "step": 22395 + }, + { + "epoch": 2.428013876843018, + "grad_norm": 2.7949326038360596, + "learning_rate": 3.827070941534891e-06, + "loss": 0.0031, + "step": 22396 + }, + { + "epoch": 2.428122289679098, + "grad_norm": 1.6261873245239258, + "learning_rate": 3.826345567967504e-06, + "loss": 0.0373, + "step": 22397 + }, + { + "epoch": 2.4282307025151777, + "grad_norm": 0.00957221258431673, + "learning_rate": 3.825620194400117e-06, + "loss": 0.0001, + "step": 22398 + }, + { + "epoch": 2.4283391153512577, + "grad_norm": 1.560360074043274, + "learning_rate": 3.824894820832729e-06, + "loss": 0.004, + "step": 22399 + }, + { + "epoch": 2.4284475281873372, + "grad_norm": 1.8709453344345093, + "learning_rate": 3.824169447265342e-06, + "loss": 0.0026, + "step": 22400 + }, + { + "epoch": 2.428555941023417, + "grad_norm": 1.6460295915603638, + "learning_rate": 3.823444073697955e-06, + "loss": 0.0261, + "step": 22401 + }, + { + "epoch": 2.4286643538594967, + "grad_norm": 0.024617167189717293, + "learning_rate": 3.822718700130568e-06, + "loss": 0.0002, + "step": 22402 + }, + { + "epoch": 2.4287727666955767, + "grad_norm": 0.04520675167441368, + "learning_rate": 3.82199332656318e-06, + "loss": 0.0003, + "step": 22403 + }, + { + "epoch": 2.4288811795316567, + "grad_norm": 10.22967529296875, + "learning_rate": 3.821267952995793e-06, + "loss": 0.0164, + "step": 22404 + }, + { + "epoch": 2.4289895923677363, + "grad_norm": 2.4363772869110107, + "learning_rate": 3.820542579428406e-06, + "loss": 0.0158, + "step": 22405 + }, + { + "epoch": 2.4290980052038162, + "grad_norm": 0.0638505220413208, + "learning_rate": 3.819817205861019e-06, + "loss": 0.0005, + "step": 22406 + }, + { + "epoch": 2.429206418039896, + "grad_norm": 7.038509845733643, + "learning_rate": 3.819091832293631e-06, + "loss": 0.0224, + "step": 22407 + }, + { + "epoch": 2.4293148308759758, + "grad_norm": 0.01620965264737606, + "learning_rate": 3.818366458726244e-06, + "loss": 0.0001, + "step": 22408 + }, + { + "epoch": 2.4294232437120553, + "grad_norm": 0.22409719228744507, + "learning_rate": 3.817641085158857e-06, + "loss": 0.0004, + "step": 22409 + }, + { + "epoch": 2.4295316565481353, + "grad_norm": 3.825413465499878, + "learning_rate": 3.81691571159147e-06, + "loss": 0.0084, + "step": 22410 + }, + { + "epoch": 2.4296400693842153, + "grad_norm": 0.15179267525672913, + "learning_rate": 3.816190338024083e-06, + "loss": 0.0018, + "step": 22411 + }, + { + "epoch": 2.429748482220295, + "grad_norm": 1.387968897819519, + "learning_rate": 3.815464964456695e-06, + "loss": 0.0037, + "step": 22412 + }, + { + "epoch": 2.429856895056375, + "grad_norm": 0.01350694615393877, + "learning_rate": 3.814739590889308e-06, + "loss": 0.0001, + "step": 22413 + }, + { + "epoch": 2.4299653078924544, + "grad_norm": 0.040591493248939514, + "learning_rate": 3.8140142173219213e-06, + "loss": 0.0006, + "step": 22414 + }, + { + "epoch": 2.4300737207285343, + "grad_norm": 0.10019643604755402, + "learning_rate": 3.813288843754534e-06, + "loss": 0.0009, + "step": 22415 + }, + { + "epoch": 2.430182133564614, + "grad_norm": 0.8228498697280884, + "learning_rate": 3.812563470187147e-06, + "loss": 0.0034, + "step": 22416 + }, + { + "epoch": 2.430290546400694, + "grad_norm": 0.12333118915557861, + "learning_rate": 3.8118380966197594e-06, + "loss": 0.0007, + "step": 22417 + }, + { + "epoch": 2.430398959236774, + "grad_norm": 4.576864242553711, + "learning_rate": 3.811112723052372e-06, + "loss": 0.0144, + "step": 22418 + }, + { + "epoch": 2.4305073720728534, + "grad_norm": 0.3336564600467682, + "learning_rate": 3.8103873494849854e-06, + "loss": 0.0007, + "step": 22419 + }, + { + "epoch": 2.4306157849089334, + "grad_norm": 1.8065574169158936, + "learning_rate": 3.809661975917598e-06, + "loss": 0.0093, + "step": 22420 + }, + { + "epoch": 2.430724197745013, + "grad_norm": 0.44242897629737854, + "learning_rate": 3.8089366023502105e-06, + "loss": 0.0006, + "step": 22421 + }, + { + "epoch": 2.430832610581093, + "grad_norm": 0.7128519415855408, + "learning_rate": 3.808211228782823e-06, + "loss": 0.0017, + "step": 22422 + }, + { + "epoch": 2.4309410234171724, + "grad_norm": 1.7852617502212524, + "learning_rate": 3.8074858552154365e-06, + "loss": 0.0061, + "step": 22423 + }, + { + "epoch": 2.4310494362532524, + "grad_norm": 0.6441150307655334, + "learning_rate": 3.806760481648049e-06, + "loss": 0.0021, + "step": 22424 + }, + { + "epoch": 2.431157849089332, + "grad_norm": 0.5779827833175659, + "learning_rate": 3.8060351080806616e-06, + "loss": 0.0012, + "step": 22425 + }, + { + "epoch": 2.431266261925412, + "grad_norm": 0.00750353280454874, + "learning_rate": 3.8053097345132746e-06, + "loss": 0.0001, + "step": 22426 + }, + { + "epoch": 2.431374674761492, + "grad_norm": 1.6061136722564697, + "learning_rate": 3.8045843609458876e-06, + "loss": 0.0417, + "step": 22427 + }, + { + "epoch": 2.4314830875975715, + "grad_norm": 0.03641212359070778, + "learning_rate": 3.8038589873785e-06, + "loss": 0.0002, + "step": 22428 + }, + { + "epoch": 2.4315915004336515, + "grad_norm": 0.9671299457550049, + "learning_rate": 3.803133613811113e-06, + "loss": 0.0022, + "step": 22429 + }, + { + "epoch": 2.431699913269731, + "grad_norm": 0.1270403116941452, + "learning_rate": 3.8024082402437257e-06, + "loss": 0.0002, + "step": 22430 + }, + { + "epoch": 2.431808326105811, + "grad_norm": 6.049202919006348, + "learning_rate": 3.8016828666763387e-06, + "loss": 0.0194, + "step": 22431 + }, + { + "epoch": 2.4319167389418905, + "grad_norm": 0.15815377235412598, + "learning_rate": 3.8009574931089517e-06, + "loss": 0.001, + "step": 22432 + }, + { + "epoch": 2.4320251517779705, + "grad_norm": 7.0352463722229, + "learning_rate": 3.8002321195415643e-06, + "loss": 0.0259, + "step": 22433 + }, + { + "epoch": 2.43213356461405, + "grad_norm": 1.5583246946334839, + "learning_rate": 3.799506745974177e-06, + "loss": 0.0047, + "step": 22434 + }, + { + "epoch": 2.43224197745013, + "grad_norm": 2.0804357528686523, + "learning_rate": 3.7987813724067894e-06, + "loss": 0.0593, + "step": 22435 + }, + { + "epoch": 2.43235039028621, + "grad_norm": 4.8680877685546875, + "learning_rate": 3.798055998839403e-06, + "loss": 0.0445, + "step": 22436 + }, + { + "epoch": 2.4324588031222896, + "grad_norm": 2.1974432468414307, + "learning_rate": 3.7973306252720154e-06, + "loss": 0.0047, + "step": 22437 + }, + { + "epoch": 2.4325672159583696, + "grad_norm": 0.04084283486008644, + "learning_rate": 3.796605251704628e-06, + "loss": 0.0004, + "step": 22438 + }, + { + "epoch": 2.432675628794449, + "grad_norm": 4.040987491607666, + "learning_rate": 3.795879878137241e-06, + "loss": 0.0098, + "step": 22439 + }, + { + "epoch": 2.432784041630529, + "grad_norm": 0.03302513435482979, + "learning_rate": 3.795154504569854e-06, + "loss": 0.0003, + "step": 22440 + }, + { + "epoch": 2.432892454466609, + "grad_norm": 0.058844584971666336, + "learning_rate": 3.7944291310024665e-06, + "loss": 0.0006, + "step": 22441 + }, + { + "epoch": 2.4330008673026886, + "grad_norm": 3.2605388164520264, + "learning_rate": 3.7937037574350795e-06, + "loss": 0.0104, + "step": 22442 + }, + { + "epoch": 2.4331092801387686, + "grad_norm": 6.284687042236328, + "learning_rate": 3.792978383867692e-06, + "loss": 0.0166, + "step": 22443 + }, + { + "epoch": 2.433217692974848, + "grad_norm": 1.8935515880584717, + "learning_rate": 3.792253010300305e-06, + "loss": 0.0236, + "step": 22444 + }, + { + "epoch": 2.433326105810928, + "grad_norm": 0.052132558077573776, + "learning_rate": 3.791527636732918e-06, + "loss": 0.0005, + "step": 22445 + }, + { + "epoch": 2.4334345186470077, + "grad_norm": 0.7616930603981018, + "learning_rate": 3.7908022631655306e-06, + "loss": 0.003, + "step": 22446 + }, + { + "epoch": 2.4335429314830876, + "grad_norm": 25.274795532226562, + "learning_rate": 3.790076889598143e-06, + "loss": 0.0167, + "step": 22447 + }, + { + "epoch": 2.433651344319167, + "grad_norm": 0.039355117827653885, + "learning_rate": 3.7893515160307557e-06, + "loss": 0.0003, + "step": 22448 + }, + { + "epoch": 2.433759757155247, + "grad_norm": 1.5392532348632812, + "learning_rate": 3.788626142463369e-06, + "loss": 0.0037, + "step": 22449 + }, + { + "epoch": 2.433868169991327, + "grad_norm": 0.07922972738742828, + "learning_rate": 3.7879007688959817e-06, + "loss": 0.0008, + "step": 22450 + }, + { + "epoch": 2.4339765828274067, + "grad_norm": 3.26198148727417, + "learning_rate": 3.7871753953285943e-06, + "loss": 0.0111, + "step": 22451 + }, + { + "epoch": 2.4340849956634867, + "grad_norm": 3.5971157550811768, + "learning_rate": 3.7864500217612073e-06, + "loss": 0.0021, + "step": 22452 + }, + { + "epoch": 2.4341934084995662, + "grad_norm": 0.009915657341480255, + "learning_rate": 3.7857246481938203e-06, + "loss": 0.0001, + "step": 22453 + }, + { + "epoch": 2.434301821335646, + "grad_norm": 1.3072985410690308, + "learning_rate": 3.784999274626433e-06, + "loss": 0.0259, + "step": 22454 + }, + { + "epoch": 2.4344102341717258, + "grad_norm": 1.845654010772705, + "learning_rate": 3.784273901059046e-06, + "loss": 0.02, + "step": 22455 + }, + { + "epoch": 2.4345186470078057, + "grad_norm": 6.581737518310547, + "learning_rate": 3.7835485274916584e-06, + "loss": 0.079, + "step": 22456 + }, + { + "epoch": 2.4346270598438853, + "grad_norm": 0.01860242709517479, + "learning_rate": 3.7828231539242714e-06, + "loss": 0.0001, + "step": 22457 + }, + { + "epoch": 2.4347354726799653, + "grad_norm": 0.20568770170211792, + "learning_rate": 3.7820977803568844e-06, + "loss": 0.0014, + "step": 22458 + }, + { + "epoch": 2.4348438855160452, + "grad_norm": 2.4209752082824707, + "learning_rate": 3.781372406789497e-06, + "loss": 0.035, + "step": 22459 + }, + { + "epoch": 2.434952298352125, + "grad_norm": 0.9228178858757019, + "learning_rate": 3.7806470332221095e-06, + "loss": 0.0031, + "step": 22460 + }, + { + "epoch": 2.4350607111882048, + "grad_norm": 1.3491337299346924, + "learning_rate": 3.779921659654722e-06, + "loss": 0.0029, + "step": 22461 + }, + { + "epoch": 2.4351691240242843, + "grad_norm": 0.5210970640182495, + "learning_rate": 3.7791962860873355e-06, + "loss": 0.0011, + "step": 22462 + }, + { + "epoch": 2.4352775368603643, + "grad_norm": 1.809156894683838, + "learning_rate": 3.778470912519948e-06, + "loss": 0.0315, + "step": 22463 + }, + { + "epoch": 2.4353859496964443, + "grad_norm": 19.696687698364258, + "learning_rate": 3.7777455389525606e-06, + "loss": 0.0117, + "step": 22464 + }, + { + "epoch": 2.435494362532524, + "grad_norm": 0.1889200210571289, + "learning_rate": 3.7770201653851736e-06, + "loss": 0.0014, + "step": 22465 + }, + { + "epoch": 2.435602775368604, + "grad_norm": 0.22080788016319275, + "learning_rate": 3.7762947918177866e-06, + "loss": 0.0011, + "step": 22466 + }, + { + "epoch": 2.4357111882046834, + "grad_norm": 0.18898777663707733, + "learning_rate": 3.775569418250399e-06, + "loss": 0.0011, + "step": 22467 + }, + { + "epoch": 2.4358196010407633, + "grad_norm": 1.6492743492126465, + "learning_rate": 3.774844044683012e-06, + "loss": 0.014, + "step": 22468 + }, + { + "epoch": 2.435928013876843, + "grad_norm": 0.030272020027041435, + "learning_rate": 3.7741186711156247e-06, + "loss": 0.0003, + "step": 22469 + }, + { + "epoch": 2.436036426712923, + "grad_norm": 1.0868442058563232, + "learning_rate": 3.7733932975482377e-06, + "loss": 0.0032, + "step": 22470 + }, + { + "epoch": 2.4361448395490024, + "grad_norm": 10.064704895019531, + "learning_rate": 3.7726679239808507e-06, + "loss": 0.0284, + "step": 22471 + }, + { + "epoch": 2.4362532523850824, + "grad_norm": 1.6797889471054077, + "learning_rate": 3.7719425504134633e-06, + "loss": 0.0043, + "step": 22472 + }, + { + "epoch": 2.4363616652211624, + "grad_norm": 2.020601987838745, + "learning_rate": 3.771217176846076e-06, + "loss": 0.0042, + "step": 22473 + }, + { + "epoch": 2.436470078057242, + "grad_norm": 1.0563918352127075, + "learning_rate": 3.7704918032786884e-06, + "loss": 0.0038, + "step": 22474 + }, + { + "epoch": 2.436578490893322, + "grad_norm": 10.950895309448242, + "learning_rate": 3.769766429711302e-06, + "loss": 0.0226, + "step": 22475 + }, + { + "epoch": 2.4366869037294014, + "grad_norm": 3.8885509967803955, + "learning_rate": 3.7690410561439144e-06, + "loss": 0.0104, + "step": 22476 + }, + { + "epoch": 2.4367953165654814, + "grad_norm": 0.5144707560539246, + "learning_rate": 3.768315682576527e-06, + "loss": 0.0018, + "step": 22477 + }, + { + "epoch": 2.436903729401561, + "grad_norm": 1.506585955619812, + "learning_rate": 3.76759030900914e-06, + "loss": 0.03, + "step": 22478 + }, + { + "epoch": 2.437012142237641, + "grad_norm": 0.6519833207130432, + "learning_rate": 3.766864935441753e-06, + "loss": 0.0022, + "step": 22479 + }, + { + "epoch": 2.4371205550737205, + "grad_norm": 0.7509364485740662, + "learning_rate": 3.7661395618743655e-06, + "loss": 0.0033, + "step": 22480 + }, + { + "epoch": 2.4372289679098005, + "grad_norm": 13.285972595214844, + "learning_rate": 3.7654141883069785e-06, + "loss": 0.0401, + "step": 22481 + }, + { + "epoch": 2.4373373807458805, + "grad_norm": 1.571674108505249, + "learning_rate": 3.764688814739591e-06, + "loss": 0.0104, + "step": 22482 + }, + { + "epoch": 2.43744579358196, + "grad_norm": 0.0967036634683609, + "learning_rate": 3.763963441172204e-06, + "loss": 0.0006, + "step": 22483 + }, + { + "epoch": 2.43755420641804, + "grad_norm": 3.6571755409240723, + "learning_rate": 3.763238067604817e-06, + "loss": 0.0236, + "step": 22484 + }, + { + "epoch": 2.4376626192541195, + "grad_norm": 0.07938624918460846, + "learning_rate": 3.7625126940374296e-06, + "loss": 0.0007, + "step": 22485 + }, + { + "epoch": 2.4377710320901995, + "grad_norm": 0.9350726008415222, + "learning_rate": 3.761787320470042e-06, + "loss": 0.0042, + "step": 22486 + }, + { + "epoch": 2.4378794449262795, + "grad_norm": 0.2133825272321701, + "learning_rate": 3.7610619469026547e-06, + "loss": 0.0007, + "step": 22487 + }, + { + "epoch": 2.437987857762359, + "grad_norm": 1.4659253358840942, + "learning_rate": 3.760336573335268e-06, + "loss": 0.0078, + "step": 22488 + }, + { + "epoch": 2.438096270598439, + "grad_norm": 3.7571489810943604, + "learning_rate": 3.7596111997678807e-06, + "loss": 0.0032, + "step": 22489 + }, + { + "epoch": 2.4382046834345186, + "grad_norm": 2.5458474159240723, + "learning_rate": 3.7588858262004933e-06, + "loss": 0.0428, + "step": 22490 + }, + { + "epoch": 2.4383130962705986, + "grad_norm": 0.059344708919525146, + "learning_rate": 3.7581604526331063e-06, + "loss": 0.0006, + "step": 22491 + }, + { + "epoch": 2.438421509106678, + "grad_norm": 3.3494279384613037, + "learning_rate": 3.7574350790657193e-06, + "loss": 0.0083, + "step": 22492 + }, + { + "epoch": 2.438529921942758, + "grad_norm": 16.417129516601562, + "learning_rate": 3.756709705498332e-06, + "loss": 0.0435, + "step": 22493 + }, + { + "epoch": 2.4386383347788376, + "grad_norm": 8.158902168273926, + "learning_rate": 3.755984331930945e-06, + "loss": 0.0073, + "step": 22494 + }, + { + "epoch": 2.4387467476149176, + "grad_norm": 3.360258102416992, + "learning_rate": 3.7552589583635574e-06, + "loss": 0.0024, + "step": 22495 + }, + { + "epoch": 2.4388551604509976, + "grad_norm": 1.2522962093353271, + "learning_rate": 3.7545335847961704e-06, + "loss": 0.0039, + "step": 22496 + }, + { + "epoch": 2.438963573287077, + "grad_norm": 6.75607442855835, + "learning_rate": 3.7538082112287834e-06, + "loss": 0.0221, + "step": 22497 + }, + { + "epoch": 2.439071986123157, + "grad_norm": 0.14315281808376312, + "learning_rate": 3.753082837661396e-06, + "loss": 0.0008, + "step": 22498 + }, + { + "epoch": 2.4391803989592367, + "grad_norm": 1.1332584619522095, + "learning_rate": 3.7523574640940085e-06, + "loss": 0.0025, + "step": 22499 + }, + { + "epoch": 2.4392888117953166, + "grad_norm": 0.3043498992919922, + "learning_rate": 3.751632090526621e-06, + "loss": 0.0011, + "step": 22500 + }, + { + "epoch": 2.439397224631396, + "grad_norm": 1.119376301765442, + "learning_rate": 3.7509067169592345e-06, + "loss": 0.0034, + "step": 22501 + }, + { + "epoch": 2.439505637467476, + "grad_norm": 0.10118808597326279, + "learning_rate": 3.750181343391847e-06, + "loss": 0.0006, + "step": 22502 + }, + { + "epoch": 2.4396140503035557, + "grad_norm": 0.7575152516365051, + "learning_rate": 3.7494559698244596e-06, + "loss": 0.0027, + "step": 22503 + }, + { + "epoch": 2.4397224631396357, + "grad_norm": 0.03975393623113632, + "learning_rate": 3.7487305962570726e-06, + "loss": 0.0002, + "step": 22504 + }, + { + "epoch": 2.4398308759757157, + "grad_norm": 2.7391233444213867, + "learning_rate": 3.7480052226896856e-06, + "loss": 0.011, + "step": 22505 + }, + { + "epoch": 2.4399392888117952, + "grad_norm": 0.051063187420368195, + "learning_rate": 3.747279849122298e-06, + "loss": 0.0007, + "step": 22506 + }, + { + "epoch": 2.440047701647875, + "grad_norm": 5.348174095153809, + "learning_rate": 3.746554475554911e-06, + "loss": 0.0401, + "step": 22507 + }, + { + "epoch": 2.4401561144839548, + "grad_norm": 0.03293261677026749, + "learning_rate": 3.7458291019875237e-06, + "loss": 0.0003, + "step": 22508 + }, + { + "epoch": 2.4402645273200347, + "grad_norm": 0.6978655457496643, + "learning_rate": 3.7451037284201367e-06, + "loss": 0.0034, + "step": 22509 + }, + { + "epoch": 2.4403729401561147, + "grad_norm": 0.05748806893825531, + "learning_rate": 3.7443783548527497e-06, + "loss": 0.0005, + "step": 22510 + }, + { + "epoch": 2.4404813529921943, + "grad_norm": 10.145621299743652, + "learning_rate": 3.7436529812853623e-06, + "loss": 0.0406, + "step": 22511 + }, + { + "epoch": 2.4405897658282742, + "grad_norm": 0.1211642324924469, + "learning_rate": 3.742927607717975e-06, + "loss": 0.0005, + "step": 22512 + }, + { + "epoch": 2.440698178664354, + "grad_norm": 0.15652891993522644, + "learning_rate": 3.7422022341505874e-06, + "loss": 0.0008, + "step": 22513 + }, + { + "epoch": 2.4408065915004338, + "grad_norm": 0.3357810080051422, + "learning_rate": 3.741476860583201e-06, + "loss": 0.012, + "step": 22514 + }, + { + "epoch": 2.4409150043365133, + "grad_norm": 0.09027931094169617, + "learning_rate": 3.7407514870158134e-06, + "loss": 0.0011, + "step": 22515 + }, + { + "epoch": 2.4410234171725933, + "grad_norm": 0.07904716581106186, + "learning_rate": 3.740026113448426e-06, + "loss": 0.0004, + "step": 22516 + }, + { + "epoch": 2.441131830008673, + "grad_norm": 1.2064367532730103, + "learning_rate": 3.739300739881039e-06, + "loss": 0.0115, + "step": 22517 + }, + { + "epoch": 2.441240242844753, + "grad_norm": 0.040795523673295975, + "learning_rate": 3.738575366313652e-06, + "loss": 0.0004, + "step": 22518 + }, + { + "epoch": 2.441348655680833, + "grad_norm": 2.551661729812622, + "learning_rate": 3.7378499927462645e-06, + "loss": 0.0049, + "step": 22519 + }, + { + "epoch": 2.4414570685169124, + "grad_norm": 2.243518590927124, + "learning_rate": 3.7371246191788775e-06, + "loss": 0.0026, + "step": 22520 + }, + { + "epoch": 2.4415654813529923, + "grad_norm": 0.16931802034378052, + "learning_rate": 3.73639924561149e-06, + "loss": 0.001, + "step": 22521 + }, + { + "epoch": 2.441673894189072, + "grad_norm": 2.0942325592041016, + "learning_rate": 3.735673872044103e-06, + "loss": 0.0181, + "step": 22522 + }, + { + "epoch": 2.441782307025152, + "grad_norm": 0.3644603192806244, + "learning_rate": 3.734948498476716e-06, + "loss": 0.0024, + "step": 22523 + }, + { + "epoch": 2.4418907198612314, + "grad_norm": 1.1429301500320435, + "learning_rate": 3.7342231249093286e-06, + "loss": 0.0057, + "step": 22524 + }, + { + "epoch": 2.4419991326973114, + "grad_norm": 1.764967918395996, + "learning_rate": 3.733497751341941e-06, + "loss": 0.0081, + "step": 22525 + }, + { + "epoch": 2.442107545533391, + "grad_norm": 4.928824424743652, + "learning_rate": 3.7327723777745537e-06, + "loss": 0.0051, + "step": 22526 + }, + { + "epoch": 2.442215958369471, + "grad_norm": 1.9147945642471313, + "learning_rate": 3.732047004207167e-06, + "loss": 0.0105, + "step": 22527 + }, + { + "epoch": 2.442324371205551, + "grad_norm": 25.843082427978516, + "learning_rate": 3.7313216306397797e-06, + "loss": 0.0224, + "step": 22528 + }, + { + "epoch": 2.4424327840416304, + "grad_norm": 7.341709136962891, + "learning_rate": 3.7305962570723923e-06, + "loss": 0.0359, + "step": 22529 + }, + { + "epoch": 2.4425411968777104, + "grad_norm": 0.16593743860721588, + "learning_rate": 3.7298708835050053e-06, + "loss": 0.0004, + "step": 22530 + }, + { + "epoch": 2.44264960971379, + "grad_norm": 0.11335919052362442, + "learning_rate": 3.7291455099376183e-06, + "loss": 0.0003, + "step": 22531 + }, + { + "epoch": 2.44275802254987, + "grad_norm": 2.7235183715820312, + "learning_rate": 3.728420136370231e-06, + "loss": 0.0101, + "step": 22532 + }, + { + "epoch": 2.44286643538595, + "grad_norm": 0.47324854135513306, + "learning_rate": 3.727694762802844e-06, + "loss": 0.0012, + "step": 22533 + }, + { + "epoch": 2.4429748482220295, + "grad_norm": 0.056564491242170334, + "learning_rate": 3.7269693892354564e-06, + "loss": 0.0006, + "step": 22534 + }, + { + "epoch": 2.4430832610581095, + "grad_norm": 0.3139073848724365, + "learning_rate": 3.7262440156680694e-06, + "loss": 0.0008, + "step": 22535 + }, + { + "epoch": 2.443191673894189, + "grad_norm": 0.07389979064464569, + "learning_rate": 3.7255186421006824e-06, + "loss": 0.0013, + "step": 22536 + }, + { + "epoch": 2.443300086730269, + "grad_norm": 0.06519819796085358, + "learning_rate": 3.724793268533295e-06, + "loss": 0.0003, + "step": 22537 + }, + { + "epoch": 2.4434084995663485, + "grad_norm": 1.7436386346817017, + "learning_rate": 3.7240678949659075e-06, + "loss": 0.0012, + "step": 22538 + }, + { + "epoch": 2.4435169124024285, + "grad_norm": 0.14725784957408905, + "learning_rate": 3.723342521398521e-06, + "loss": 0.0009, + "step": 22539 + }, + { + "epoch": 2.443625325238508, + "grad_norm": 4.046571731567383, + "learning_rate": 3.7226171478311335e-06, + "loss": 0.0298, + "step": 22540 + }, + { + "epoch": 2.443733738074588, + "grad_norm": 15.557138442993164, + "learning_rate": 3.721891774263746e-06, + "loss": 0.0228, + "step": 22541 + }, + { + "epoch": 2.443842150910668, + "grad_norm": 0.9323905110359192, + "learning_rate": 3.7211664006963586e-06, + "loss": 0.0019, + "step": 22542 + }, + { + "epoch": 2.4439505637467476, + "grad_norm": 1.8011494874954224, + "learning_rate": 3.7204410271289716e-06, + "loss": 0.0042, + "step": 22543 + }, + { + "epoch": 2.4440589765828276, + "grad_norm": 2.5550498962402344, + "learning_rate": 3.7197156535615846e-06, + "loss": 0.0078, + "step": 22544 + }, + { + "epoch": 2.444167389418907, + "grad_norm": 3.0262680053710938, + "learning_rate": 3.718990279994197e-06, + "loss": 0.0222, + "step": 22545 + }, + { + "epoch": 2.444275802254987, + "grad_norm": 2.9438421726226807, + "learning_rate": 3.71826490642681e-06, + "loss": 0.0229, + "step": 22546 + }, + { + "epoch": 2.4443842150910666, + "grad_norm": 4.602210521697998, + "learning_rate": 3.7175395328594227e-06, + "loss": 0.0224, + "step": 22547 + }, + { + "epoch": 2.4444926279271466, + "grad_norm": 0.11469785869121552, + "learning_rate": 3.7168141592920357e-06, + "loss": 0.0007, + "step": 22548 + }, + { + "epoch": 2.444601040763226, + "grad_norm": 20.48678970336914, + "learning_rate": 3.7160887857246487e-06, + "loss": 0.0498, + "step": 22549 + }, + { + "epoch": 2.444709453599306, + "grad_norm": 0.05091095343232155, + "learning_rate": 3.7153634121572613e-06, + "loss": 0.0007, + "step": 22550 + }, + { + "epoch": 2.444817866435386, + "grad_norm": 0.3437691926956177, + "learning_rate": 3.714638038589874e-06, + "loss": 0.0011, + "step": 22551 + }, + { + "epoch": 2.4449262792714657, + "grad_norm": 5.6209821701049805, + "learning_rate": 3.7139126650224873e-06, + "loss": 0.025, + "step": 22552 + }, + { + "epoch": 2.4450346921075456, + "grad_norm": 7.299332141876221, + "learning_rate": 3.7131872914551e-06, + "loss": 0.0234, + "step": 22553 + }, + { + "epoch": 2.445143104943625, + "grad_norm": 0.048636943101882935, + "learning_rate": 3.7124619178877124e-06, + "loss": 0.0002, + "step": 22554 + }, + { + "epoch": 2.445251517779705, + "grad_norm": 0.671857476234436, + "learning_rate": 3.711736544320325e-06, + "loss": 0.0034, + "step": 22555 + }, + { + "epoch": 2.4453599306157847, + "grad_norm": 0.04625524580478668, + "learning_rate": 3.711011170752938e-06, + "loss": 0.0002, + "step": 22556 + }, + { + "epoch": 2.4454683434518647, + "grad_norm": 0.04515198990702629, + "learning_rate": 3.710285797185551e-06, + "loss": 0.0003, + "step": 22557 + }, + { + "epoch": 2.4455767562879447, + "grad_norm": 0.24741165339946747, + "learning_rate": 3.7095604236181635e-06, + "loss": 0.0021, + "step": 22558 + }, + { + "epoch": 2.4456851691240242, + "grad_norm": 0.1320907324552536, + "learning_rate": 3.7088350500507765e-06, + "loss": 0.0012, + "step": 22559 + }, + { + "epoch": 2.445793581960104, + "grad_norm": 1.2111667394638062, + "learning_rate": 3.708109676483389e-06, + "loss": 0.0027, + "step": 22560 + }, + { + "epoch": 2.4459019947961838, + "grad_norm": 1.921112298965454, + "learning_rate": 3.707384302916002e-06, + "loss": 0.0065, + "step": 22561 + }, + { + "epoch": 2.4460104076322637, + "grad_norm": 3.7088117599487305, + "learning_rate": 3.706658929348615e-06, + "loss": 0.0186, + "step": 22562 + }, + { + "epoch": 2.4461188204683433, + "grad_norm": 1.3664405345916748, + "learning_rate": 3.7059335557812276e-06, + "loss": 0.0058, + "step": 22563 + }, + { + "epoch": 2.4462272333044233, + "grad_norm": 0.02186528779566288, + "learning_rate": 3.70520818221384e-06, + "loss": 0.0002, + "step": 22564 + }, + { + "epoch": 2.4463356461405033, + "grad_norm": 3.43257212638855, + "learning_rate": 3.7044828086464536e-06, + "loss": 0.0148, + "step": 22565 + }, + { + "epoch": 2.446444058976583, + "grad_norm": 1.13480806350708, + "learning_rate": 3.703757435079066e-06, + "loss": 0.0029, + "step": 22566 + }, + { + "epoch": 2.4465524718126628, + "grad_norm": 0.24037091434001923, + "learning_rate": 3.7030320615116787e-06, + "loss": 0.0016, + "step": 22567 + }, + { + "epoch": 2.4466608846487423, + "grad_norm": 0.35498952865600586, + "learning_rate": 3.7023066879442913e-06, + "loss": 0.0003, + "step": 22568 + }, + { + "epoch": 2.4467692974848223, + "grad_norm": 0.07260527461767197, + "learning_rate": 3.7015813143769043e-06, + "loss": 0.0005, + "step": 22569 + }, + { + "epoch": 2.446877710320902, + "grad_norm": 3.720456600189209, + "learning_rate": 3.7008559408095173e-06, + "loss": 0.025, + "step": 22570 + }, + { + "epoch": 2.446986123156982, + "grad_norm": 0.5480589270591736, + "learning_rate": 3.70013056724213e-06, + "loss": 0.0021, + "step": 22571 + }, + { + "epoch": 2.4470945359930614, + "grad_norm": 0.13469627499580383, + "learning_rate": 3.699405193674743e-06, + "loss": 0.0006, + "step": 22572 + }, + { + "epoch": 2.4472029488291414, + "grad_norm": 18.803808212280273, + "learning_rate": 3.6986798201073554e-06, + "loss": 0.0336, + "step": 22573 + }, + { + "epoch": 2.4473113616652213, + "grad_norm": 0.011743483133614063, + "learning_rate": 3.6979544465399684e-06, + "loss": 0.0002, + "step": 22574 + }, + { + "epoch": 2.447419774501301, + "grad_norm": 0.17915867269039154, + "learning_rate": 3.6972290729725814e-06, + "loss": 0.0013, + "step": 22575 + }, + { + "epoch": 2.447528187337381, + "grad_norm": 0.6895989179611206, + "learning_rate": 3.696503699405194e-06, + "loss": 0.0011, + "step": 22576 + }, + { + "epoch": 2.4476366001734604, + "grad_norm": 0.10584771633148193, + "learning_rate": 3.6957783258378065e-06, + "loss": 0.0009, + "step": 22577 + }, + { + "epoch": 2.4477450130095404, + "grad_norm": 0.6140899062156677, + "learning_rate": 3.69505295227042e-06, + "loss": 0.0011, + "step": 22578 + }, + { + "epoch": 2.44785342584562, + "grad_norm": 1.6009771823883057, + "learning_rate": 3.6943275787030325e-06, + "loss": 0.0113, + "step": 22579 + }, + { + "epoch": 2.4479618386817, + "grad_norm": 0.5528267025947571, + "learning_rate": 3.693602205135645e-06, + "loss": 0.0017, + "step": 22580 + }, + { + "epoch": 2.4480702515177795, + "grad_norm": 0.1278303563594818, + "learning_rate": 3.6928768315682576e-06, + "loss": 0.0004, + "step": 22581 + }, + { + "epoch": 2.4481786643538594, + "grad_norm": 10.876830101013184, + "learning_rate": 3.6921514580008706e-06, + "loss": 0.0119, + "step": 22582 + }, + { + "epoch": 2.4482870771899394, + "grad_norm": 0.7575395107269287, + "learning_rate": 3.6914260844334836e-06, + "loss": 0.0016, + "step": 22583 + }, + { + "epoch": 2.448395490026019, + "grad_norm": 0.4867115318775177, + "learning_rate": 3.690700710866096e-06, + "loss": 0.0023, + "step": 22584 + }, + { + "epoch": 2.448503902862099, + "grad_norm": 0.2380865067243576, + "learning_rate": 3.689975337298709e-06, + "loss": 0.0005, + "step": 22585 + }, + { + "epoch": 2.4486123156981785, + "grad_norm": 0.41182413697242737, + "learning_rate": 3.6892499637313217e-06, + "loss": 0.0009, + "step": 22586 + }, + { + "epoch": 2.4487207285342585, + "grad_norm": 0.265268474817276, + "learning_rate": 3.6885245901639347e-06, + "loss": 0.0008, + "step": 22587 + }, + { + "epoch": 2.4488291413703385, + "grad_norm": 0.12190920859575272, + "learning_rate": 3.6877992165965477e-06, + "loss": 0.0006, + "step": 22588 + }, + { + "epoch": 2.448937554206418, + "grad_norm": 0.016723450273275375, + "learning_rate": 3.6870738430291603e-06, + "loss": 0.0002, + "step": 22589 + }, + { + "epoch": 2.449045967042498, + "grad_norm": 5.104631423950195, + "learning_rate": 3.686348469461773e-06, + "loss": 0.0337, + "step": 22590 + }, + { + "epoch": 2.4491543798785775, + "grad_norm": 0.057474248111248016, + "learning_rate": 3.6856230958943863e-06, + "loss": 0.0004, + "step": 22591 + }, + { + "epoch": 2.4492627927146575, + "grad_norm": 0.007906968705356121, + "learning_rate": 3.684897722326999e-06, + "loss": 0.0001, + "step": 22592 + }, + { + "epoch": 2.449371205550737, + "grad_norm": 0.381817489862442, + "learning_rate": 3.6841723487596114e-06, + "loss": 0.0018, + "step": 22593 + }, + { + "epoch": 2.449479618386817, + "grad_norm": 0.15678851306438446, + "learning_rate": 3.683446975192224e-06, + "loss": 0.0006, + "step": 22594 + }, + { + "epoch": 2.4495880312228966, + "grad_norm": 0.02754242718219757, + "learning_rate": 3.682721601624837e-06, + "loss": 0.0001, + "step": 22595 + }, + { + "epoch": 2.4496964440589766, + "grad_norm": 0.03594469279050827, + "learning_rate": 3.68199622805745e-06, + "loss": 0.0003, + "step": 22596 + }, + { + "epoch": 2.4498048568950566, + "grad_norm": 0.046080391854047775, + "learning_rate": 3.6812708544900625e-06, + "loss": 0.0006, + "step": 22597 + }, + { + "epoch": 2.449913269731136, + "grad_norm": 0.1343885213136673, + "learning_rate": 3.6805454809226755e-06, + "loss": 0.0011, + "step": 22598 + }, + { + "epoch": 2.450021682567216, + "grad_norm": 0.02099434658885002, + "learning_rate": 3.679820107355288e-06, + "loss": 0.0003, + "step": 22599 + }, + { + "epoch": 2.4501300954032956, + "grad_norm": 0.7616642713546753, + "learning_rate": 3.679094733787901e-06, + "loss": 0.0019, + "step": 22600 + }, + { + "epoch": 2.4502385082393756, + "grad_norm": 0.09767448902130127, + "learning_rate": 3.678369360220514e-06, + "loss": 0.0007, + "step": 22601 + }, + { + "epoch": 2.450346921075455, + "grad_norm": 1.3594361543655396, + "learning_rate": 3.6776439866531266e-06, + "loss": 0.0026, + "step": 22602 + }, + { + "epoch": 2.450455333911535, + "grad_norm": 3.249011278152466, + "learning_rate": 3.676918613085739e-06, + "loss": 0.0344, + "step": 22603 + }, + { + "epoch": 2.4505637467476147, + "grad_norm": 0.018639223650097847, + "learning_rate": 3.6761932395183526e-06, + "loss": 0.0001, + "step": 22604 + }, + { + "epoch": 2.4506721595836947, + "grad_norm": 4.107558727264404, + "learning_rate": 3.675467865950965e-06, + "loss": 0.0143, + "step": 22605 + }, + { + "epoch": 2.4507805724197746, + "grad_norm": 0.2854383885860443, + "learning_rate": 3.6747424923835777e-06, + "loss": 0.002, + "step": 22606 + }, + { + "epoch": 2.450888985255854, + "grad_norm": 0.5454442501068115, + "learning_rate": 3.6740171188161903e-06, + "loss": 0.0008, + "step": 22607 + }, + { + "epoch": 2.450997398091934, + "grad_norm": 0.6210414171218872, + "learning_rate": 3.6732917452488033e-06, + "loss": 0.002, + "step": 22608 + }, + { + "epoch": 2.4511058109280137, + "grad_norm": 0.11901699006557465, + "learning_rate": 3.6725663716814163e-06, + "loss": 0.0007, + "step": 22609 + }, + { + "epoch": 2.4512142237640937, + "grad_norm": 2.3029308319091797, + "learning_rate": 3.671840998114029e-06, + "loss": 0.0027, + "step": 22610 + }, + { + "epoch": 2.4513226366001737, + "grad_norm": 7.354367256164551, + "learning_rate": 3.671115624546642e-06, + "loss": 0.0883, + "step": 22611 + }, + { + "epoch": 2.4514310494362532, + "grad_norm": 6.851974964141846, + "learning_rate": 3.6703902509792544e-06, + "loss": 0.0375, + "step": 22612 + }, + { + "epoch": 2.451539462272333, + "grad_norm": 2.83927845954895, + "learning_rate": 3.669664877411868e-06, + "loss": 0.0094, + "step": 22613 + }, + { + "epoch": 2.4516478751084128, + "grad_norm": 2.2056829929351807, + "learning_rate": 3.6689395038444804e-06, + "loss": 0.0108, + "step": 22614 + }, + { + "epoch": 2.4517562879444927, + "grad_norm": 0.3022206425666809, + "learning_rate": 3.668214130277093e-06, + "loss": 0.001, + "step": 22615 + }, + { + "epoch": 2.4518647007805723, + "grad_norm": 0.06749558448791504, + "learning_rate": 3.6674887567097055e-06, + "loss": 0.0003, + "step": 22616 + }, + { + "epoch": 2.4519731136166523, + "grad_norm": 1.4060274362564087, + "learning_rate": 3.666763383142319e-06, + "loss": 0.0113, + "step": 22617 + }, + { + "epoch": 2.452081526452732, + "grad_norm": 1.4710686206817627, + "learning_rate": 3.6660380095749315e-06, + "loss": 0.0183, + "step": 22618 + }, + { + "epoch": 2.452189939288812, + "grad_norm": 0.037351831793785095, + "learning_rate": 3.665312636007544e-06, + "loss": 0.0002, + "step": 22619 + }, + { + "epoch": 2.452298352124892, + "grad_norm": 0.5983244180679321, + "learning_rate": 3.6645872624401566e-06, + "loss": 0.0018, + "step": 22620 + }, + { + "epoch": 2.4524067649609713, + "grad_norm": 0.6265637874603271, + "learning_rate": 3.6638618888727696e-06, + "loss": 0.0011, + "step": 22621 + }, + { + "epoch": 2.4525151777970513, + "grad_norm": 3.148129463195801, + "learning_rate": 3.6631365153053826e-06, + "loss": 0.0042, + "step": 22622 + }, + { + "epoch": 2.452623590633131, + "grad_norm": 0.16371139883995056, + "learning_rate": 3.662411141737995e-06, + "loss": 0.0008, + "step": 22623 + }, + { + "epoch": 2.452732003469211, + "grad_norm": 4.104379177093506, + "learning_rate": 3.661685768170608e-06, + "loss": 0.0387, + "step": 22624 + }, + { + "epoch": 2.4528404163052904, + "grad_norm": 0.0682249665260315, + "learning_rate": 3.6609603946032207e-06, + "loss": 0.0005, + "step": 22625 + }, + { + "epoch": 2.4529488291413704, + "grad_norm": 4.868601322174072, + "learning_rate": 3.660235021035834e-06, + "loss": 0.0522, + "step": 22626 + }, + { + "epoch": 2.45305724197745, + "grad_norm": 0.07158312946557999, + "learning_rate": 3.6595096474684467e-06, + "loss": 0.0005, + "step": 22627 + }, + { + "epoch": 2.45316565481353, + "grad_norm": 1.1931729316711426, + "learning_rate": 3.6587842739010593e-06, + "loss": 0.0054, + "step": 22628 + }, + { + "epoch": 2.45327406764961, + "grad_norm": 6.338059902191162, + "learning_rate": 3.658058900333672e-06, + "loss": 0.01, + "step": 22629 + }, + { + "epoch": 2.4533824804856894, + "grad_norm": 0.07966988533735275, + "learning_rate": 3.6573335267662853e-06, + "loss": 0.0004, + "step": 22630 + }, + { + "epoch": 2.4534908933217694, + "grad_norm": 16.32223129272461, + "learning_rate": 3.656608153198898e-06, + "loss": 0.0521, + "step": 22631 + }, + { + "epoch": 2.453599306157849, + "grad_norm": 0.06512479484081268, + "learning_rate": 3.6558827796315104e-06, + "loss": 0.0007, + "step": 22632 + }, + { + "epoch": 2.453707718993929, + "grad_norm": 0.22943226993083954, + "learning_rate": 3.655157406064123e-06, + "loss": 0.0006, + "step": 22633 + }, + { + "epoch": 2.453816131830009, + "grad_norm": 0.11212728172540665, + "learning_rate": 3.6544320324967364e-06, + "loss": 0.0007, + "step": 22634 + }, + { + "epoch": 2.4539245446660884, + "grad_norm": 0.02393146976828575, + "learning_rate": 3.653706658929349e-06, + "loss": 0.0001, + "step": 22635 + }, + { + "epoch": 2.4540329575021684, + "grad_norm": 0.9795918464660645, + "learning_rate": 3.6529812853619615e-06, + "loss": 0.006, + "step": 22636 + }, + { + "epoch": 2.454141370338248, + "grad_norm": 2.6905198097229004, + "learning_rate": 3.6522559117945745e-06, + "loss": 0.012, + "step": 22637 + }, + { + "epoch": 2.454249783174328, + "grad_norm": 4.122719764709473, + "learning_rate": 3.651530538227187e-06, + "loss": 0.0081, + "step": 22638 + }, + { + "epoch": 2.4543581960104075, + "grad_norm": 1.3739882707595825, + "learning_rate": 3.6508051646598005e-06, + "loss": 0.0249, + "step": 22639 + }, + { + "epoch": 2.4544666088464875, + "grad_norm": 3.5831682682037354, + "learning_rate": 3.650079791092413e-06, + "loss": 0.017, + "step": 22640 + }, + { + "epoch": 2.454575021682567, + "grad_norm": 0.25755682587623596, + "learning_rate": 3.6493544175250256e-06, + "loss": 0.0011, + "step": 22641 + }, + { + "epoch": 2.454683434518647, + "grad_norm": 0.177814781665802, + "learning_rate": 3.648629043957638e-06, + "loss": 0.0007, + "step": 22642 + }, + { + "epoch": 2.454791847354727, + "grad_norm": 2.1432716846466064, + "learning_rate": 3.6479036703902516e-06, + "loss": 0.0212, + "step": 22643 + }, + { + "epoch": 2.4549002601908065, + "grad_norm": 0.17331737279891968, + "learning_rate": 3.647178296822864e-06, + "loss": 0.0008, + "step": 22644 + }, + { + "epoch": 2.4550086730268865, + "grad_norm": 0.005954028107225895, + "learning_rate": 3.6464529232554767e-06, + "loss": 0.0001, + "step": 22645 + }, + { + "epoch": 2.455117085862966, + "grad_norm": 3.592804431915283, + "learning_rate": 3.6457275496880893e-06, + "loss": 0.0144, + "step": 22646 + }, + { + "epoch": 2.455225498699046, + "grad_norm": 0.13845129311084747, + "learning_rate": 3.6450021761207027e-06, + "loss": 0.0007, + "step": 22647 + }, + { + "epoch": 2.4553339115351256, + "grad_norm": 6.413524150848389, + "learning_rate": 3.6442768025533153e-06, + "loss": 0.0259, + "step": 22648 + }, + { + "epoch": 2.4554423243712056, + "grad_norm": 0.13316956162452698, + "learning_rate": 3.643551428985928e-06, + "loss": 0.0006, + "step": 22649 + }, + { + "epoch": 2.455550737207285, + "grad_norm": 0.03894566372036934, + "learning_rate": 3.642826055418541e-06, + "loss": 0.0003, + "step": 22650 + }, + { + "epoch": 2.455659150043365, + "grad_norm": 1.7622666358947754, + "learning_rate": 3.6421006818511534e-06, + "loss": 0.0028, + "step": 22651 + }, + { + "epoch": 2.455767562879445, + "grad_norm": 4.161493301391602, + "learning_rate": 3.641375308283767e-06, + "loss": 0.0138, + "step": 22652 + }, + { + "epoch": 2.4558759757155246, + "grad_norm": 0.023858288303017616, + "learning_rate": 3.6406499347163794e-06, + "loss": 0.0002, + "step": 22653 + }, + { + "epoch": 2.4559843885516046, + "grad_norm": 0.9216068387031555, + "learning_rate": 3.639924561148992e-06, + "loss": 0.0021, + "step": 22654 + }, + { + "epoch": 2.456092801387684, + "grad_norm": 11.025260925292969, + "learning_rate": 3.6391991875816045e-06, + "loss": 0.02, + "step": 22655 + }, + { + "epoch": 2.456201214223764, + "grad_norm": 3.0729360580444336, + "learning_rate": 3.638473814014218e-06, + "loss": 0.0695, + "step": 22656 + }, + { + "epoch": 2.456309627059844, + "grad_norm": 0.007878455333411694, + "learning_rate": 3.6377484404468305e-06, + "loss": 0.0001, + "step": 22657 + }, + { + "epoch": 2.4564180398959237, + "grad_norm": 0.06658497452735901, + "learning_rate": 3.637023066879443e-06, + "loss": 0.0007, + "step": 22658 + }, + { + "epoch": 2.4565264527320037, + "grad_norm": 0.3096611797809601, + "learning_rate": 3.6362976933120556e-06, + "loss": 0.0007, + "step": 22659 + }, + { + "epoch": 2.456634865568083, + "grad_norm": 0.009961575269699097, + "learning_rate": 3.635572319744669e-06, + "loss": 0.0001, + "step": 22660 + }, + { + "epoch": 2.456743278404163, + "grad_norm": 0.051011525094509125, + "learning_rate": 3.6348469461772816e-06, + "loss": 0.0003, + "step": 22661 + }, + { + "epoch": 2.4568516912402427, + "grad_norm": 8.296338081359863, + "learning_rate": 3.6341215726098946e-06, + "loss": 0.0063, + "step": 22662 + }, + { + "epoch": 2.4569601040763227, + "grad_norm": 4.987264633178711, + "learning_rate": 3.633396199042507e-06, + "loss": 0.0085, + "step": 22663 + }, + { + "epoch": 2.4570685169124022, + "grad_norm": 0.22334478795528412, + "learning_rate": 3.6326708254751197e-06, + "loss": 0.0005, + "step": 22664 + }, + { + "epoch": 2.4571769297484822, + "grad_norm": 0.06395742297172546, + "learning_rate": 3.631945451907733e-06, + "loss": 0.0005, + "step": 22665 + }, + { + "epoch": 2.457285342584562, + "grad_norm": 1.3544590473175049, + "learning_rate": 3.6312200783403457e-06, + "loss": 0.004, + "step": 22666 + }, + { + "epoch": 2.4573937554206418, + "grad_norm": 4.977591037750244, + "learning_rate": 3.6304947047729583e-06, + "loss": 0.0021, + "step": 22667 + }, + { + "epoch": 2.4575021682567217, + "grad_norm": 1.121760606765747, + "learning_rate": 3.629769331205571e-06, + "loss": 0.0044, + "step": 22668 + }, + { + "epoch": 2.4576105810928013, + "grad_norm": 1.1454393863677979, + "learning_rate": 3.6290439576381843e-06, + "loss": 0.0024, + "step": 22669 + }, + { + "epoch": 2.4577189939288813, + "grad_norm": 3.2630317211151123, + "learning_rate": 3.628318584070797e-06, + "loss": 0.0158, + "step": 22670 + }, + { + "epoch": 2.457827406764961, + "grad_norm": 0.9948363900184631, + "learning_rate": 3.6275932105034094e-06, + "loss": 0.0015, + "step": 22671 + }, + { + "epoch": 2.457935819601041, + "grad_norm": 0.004573095589876175, + "learning_rate": 3.626867836936022e-06, + "loss": 0.0, + "step": 22672 + }, + { + "epoch": 2.4580442324371203, + "grad_norm": 0.023131737485527992, + "learning_rate": 3.6261424633686354e-06, + "loss": 0.0001, + "step": 22673 + }, + { + "epoch": 2.4581526452732003, + "grad_norm": 0.17266441881656647, + "learning_rate": 3.625417089801248e-06, + "loss": 0.0005, + "step": 22674 + }, + { + "epoch": 2.4582610581092803, + "grad_norm": 7.7048234939575195, + "learning_rate": 3.624691716233861e-06, + "loss": 0.0623, + "step": 22675 + }, + { + "epoch": 2.45836947094536, + "grad_norm": 0.04590842127799988, + "learning_rate": 3.6239663426664735e-06, + "loss": 0.0005, + "step": 22676 + }, + { + "epoch": 2.45847788378144, + "grad_norm": 0.1506151705980301, + "learning_rate": 3.623240969099086e-06, + "loss": 0.0005, + "step": 22677 + }, + { + "epoch": 2.4585862966175194, + "grad_norm": 11.32110595703125, + "learning_rate": 3.6225155955316995e-06, + "loss": 0.0424, + "step": 22678 + }, + { + "epoch": 2.4586947094535994, + "grad_norm": 0.172790065407753, + "learning_rate": 3.621790221964312e-06, + "loss": 0.0011, + "step": 22679 + }, + { + "epoch": 2.4588031222896793, + "grad_norm": 0.0814838856458664, + "learning_rate": 3.6210648483969246e-06, + "loss": 0.0001, + "step": 22680 + }, + { + "epoch": 2.458911535125759, + "grad_norm": 0.03421064838767052, + "learning_rate": 3.620339474829537e-06, + "loss": 0.0001, + "step": 22681 + }, + { + "epoch": 2.459019947961839, + "grad_norm": 0.013697509653866291, + "learning_rate": 3.6196141012621506e-06, + "loss": 0.0001, + "step": 22682 + }, + { + "epoch": 2.4591283607979184, + "grad_norm": 0.022704431787133217, + "learning_rate": 3.618888727694763e-06, + "loss": 0.0001, + "step": 22683 + }, + { + "epoch": 2.4592367736339984, + "grad_norm": 0.1539454311132431, + "learning_rate": 3.6181633541273757e-06, + "loss": 0.0005, + "step": 22684 + }, + { + "epoch": 2.459345186470078, + "grad_norm": 0.32050612568855286, + "learning_rate": 3.6174379805599883e-06, + "loss": 0.0017, + "step": 22685 + }, + { + "epoch": 2.459453599306158, + "grad_norm": 0.724433422088623, + "learning_rate": 3.6167126069926017e-06, + "loss": 0.0028, + "step": 22686 + }, + { + "epoch": 2.4595620121422375, + "grad_norm": 11.690607070922852, + "learning_rate": 3.6159872334252143e-06, + "loss": 0.0165, + "step": 22687 + }, + { + "epoch": 2.4596704249783174, + "grad_norm": 1.4369726181030273, + "learning_rate": 3.6152618598578273e-06, + "loss": 0.0226, + "step": 22688 + }, + { + "epoch": 2.4597788378143974, + "grad_norm": 0.5063291788101196, + "learning_rate": 3.61453648629044e-06, + "loss": 0.0015, + "step": 22689 + }, + { + "epoch": 2.459887250650477, + "grad_norm": 0.12908099591732025, + "learning_rate": 3.6138111127230524e-06, + "loss": 0.0005, + "step": 22690 + }, + { + "epoch": 2.459995663486557, + "grad_norm": 3.910132646560669, + "learning_rate": 3.613085739155666e-06, + "loss": 0.0501, + "step": 22691 + }, + { + "epoch": 2.4601040763226365, + "grad_norm": 3.7981650829315186, + "learning_rate": 3.6123603655882784e-06, + "loss": 0.0019, + "step": 22692 + }, + { + "epoch": 2.4602124891587165, + "grad_norm": 0.01912383735179901, + "learning_rate": 3.611634992020891e-06, + "loss": 0.0001, + "step": 22693 + }, + { + "epoch": 2.460320901994796, + "grad_norm": 0.1629638373851776, + "learning_rate": 3.6109096184535035e-06, + "loss": 0.0009, + "step": 22694 + }, + { + "epoch": 2.460429314830876, + "grad_norm": 10.374261856079102, + "learning_rate": 3.610184244886117e-06, + "loss": 0.0177, + "step": 22695 + }, + { + "epoch": 2.4605377276669556, + "grad_norm": 0.00821156520396471, + "learning_rate": 3.6094588713187295e-06, + "loss": 0.0001, + "step": 22696 + }, + { + "epoch": 2.4606461405030355, + "grad_norm": 0.08840607851743698, + "learning_rate": 3.608733497751342e-06, + "loss": 0.0003, + "step": 22697 + }, + { + "epoch": 2.4607545533391155, + "grad_norm": 0.16839195787906647, + "learning_rate": 3.6080081241839546e-06, + "loss": 0.0003, + "step": 22698 + }, + { + "epoch": 2.460862966175195, + "grad_norm": 0.11239935457706451, + "learning_rate": 3.607282750616568e-06, + "loss": 0.0006, + "step": 22699 + }, + { + "epoch": 2.460971379011275, + "grad_norm": 0.010570777580142021, + "learning_rate": 3.6065573770491806e-06, + "loss": 0.0001, + "step": 22700 + }, + { + "epoch": 2.4610797918473546, + "grad_norm": 0.011436037719249725, + "learning_rate": 3.6058320034817936e-06, + "loss": 0.0001, + "step": 22701 + }, + { + "epoch": 2.4611882046834346, + "grad_norm": 0.15908996760845184, + "learning_rate": 3.605106629914406e-06, + "loss": 0.0002, + "step": 22702 + }, + { + "epoch": 2.461296617519514, + "grad_norm": 0.007572778034955263, + "learning_rate": 3.6043812563470187e-06, + "loss": 0.0001, + "step": 22703 + }, + { + "epoch": 2.461405030355594, + "grad_norm": 62.27206039428711, + "learning_rate": 3.603655882779632e-06, + "loss": 0.0221, + "step": 22704 + }, + { + "epoch": 2.4615134431916736, + "grad_norm": 0.6524771451950073, + "learning_rate": 3.6029305092122447e-06, + "loss": 0.0005, + "step": 22705 + }, + { + "epoch": 2.4616218560277536, + "grad_norm": 0.1574110984802246, + "learning_rate": 3.6022051356448573e-06, + "loss": 0.0006, + "step": 22706 + }, + { + "epoch": 2.4617302688638336, + "grad_norm": 1.9196535348892212, + "learning_rate": 3.60147976207747e-06, + "loss": 0.0369, + "step": 22707 + }, + { + "epoch": 2.461838681699913, + "grad_norm": 0.06581048667430878, + "learning_rate": 3.6007543885100833e-06, + "loss": 0.0004, + "step": 22708 + }, + { + "epoch": 2.461947094535993, + "grad_norm": 1.1281514167785645, + "learning_rate": 3.600029014942696e-06, + "loss": 0.0248, + "step": 22709 + }, + { + "epoch": 2.4620555073720727, + "grad_norm": 0.034454096108675, + "learning_rate": 3.5993036413753084e-06, + "loss": 0.0002, + "step": 22710 + }, + { + "epoch": 2.4621639202081527, + "grad_norm": 2.03361177444458, + "learning_rate": 3.5985782678079214e-06, + "loss": 0.0046, + "step": 22711 + }, + { + "epoch": 2.4622723330442327, + "grad_norm": 0.9032208919525146, + "learning_rate": 3.5978528942405344e-06, + "loss": 0.0011, + "step": 22712 + }, + { + "epoch": 2.462380745880312, + "grad_norm": 5.298257827758789, + "learning_rate": 3.597127520673147e-06, + "loss": 0.0014, + "step": 22713 + }, + { + "epoch": 2.462489158716392, + "grad_norm": 0.41296523809432983, + "learning_rate": 3.59640214710576e-06, + "loss": 0.0012, + "step": 22714 + }, + { + "epoch": 2.4625975715524717, + "grad_norm": 0.1034802496433258, + "learning_rate": 3.5956767735383725e-06, + "loss": 0.0008, + "step": 22715 + }, + { + "epoch": 2.4627059843885517, + "grad_norm": 1.6455742120742798, + "learning_rate": 3.594951399970985e-06, + "loss": 0.0204, + "step": 22716 + }, + { + "epoch": 2.4628143972246312, + "grad_norm": 0.011572732590138912, + "learning_rate": 3.5942260264035985e-06, + "loss": 0.0001, + "step": 22717 + }, + { + "epoch": 2.4629228100607112, + "grad_norm": 0.041274964809417725, + "learning_rate": 3.593500652836211e-06, + "loss": 0.0001, + "step": 22718 + }, + { + "epoch": 2.4630312228967908, + "grad_norm": 16.74846076965332, + "learning_rate": 3.5927752792688236e-06, + "loss": 0.0068, + "step": 22719 + }, + { + "epoch": 2.4631396357328708, + "grad_norm": 0.3770064115524292, + "learning_rate": 3.592049905701436e-06, + "loss": 0.0008, + "step": 22720 + }, + { + "epoch": 2.4632480485689507, + "grad_norm": 4.728588581085205, + "learning_rate": 3.5913245321340496e-06, + "loss": 0.0077, + "step": 22721 + }, + { + "epoch": 2.4633564614050303, + "grad_norm": 3.9489686489105225, + "learning_rate": 3.590599158566662e-06, + "loss": 0.0963, + "step": 22722 + }, + { + "epoch": 2.4634648742411103, + "grad_norm": 0.08270370960235596, + "learning_rate": 3.5898737849992747e-06, + "loss": 0.0007, + "step": 22723 + }, + { + "epoch": 2.46357328707719, + "grad_norm": 2.4367318153381348, + "learning_rate": 3.5891484114318877e-06, + "loss": 0.0121, + "step": 22724 + }, + { + "epoch": 2.46368169991327, + "grad_norm": 2.548844814300537, + "learning_rate": 3.5884230378645007e-06, + "loss": 0.023, + "step": 22725 + }, + { + "epoch": 2.4637901127493493, + "grad_norm": 0.33644625544548035, + "learning_rate": 3.5876976642971133e-06, + "loss": 0.0024, + "step": 22726 + }, + { + "epoch": 2.4638985255854293, + "grad_norm": 0.10096601396799088, + "learning_rate": 3.5869722907297263e-06, + "loss": 0.0006, + "step": 22727 + }, + { + "epoch": 2.464006938421509, + "grad_norm": 0.7509582042694092, + "learning_rate": 3.586246917162339e-06, + "loss": 0.0017, + "step": 22728 + }, + { + "epoch": 2.464115351257589, + "grad_norm": 0.1977854073047638, + "learning_rate": 3.585521543594952e-06, + "loss": 0.0008, + "step": 22729 + }, + { + "epoch": 2.464223764093669, + "grad_norm": 0.12949681282043457, + "learning_rate": 3.584796170027565e-06, + "loss": 0.0002, + "step": 22730 + }, + { + "epoch": 2.4643321769297484, + "grad_norm": 0.8085803985595703, + "learning_rate": 3.5840707964601774e-06, + "loss": 0.011, + "step": 22731 + }, + { + "epoch": 2.4644405897658284, + "grad_norm": 3.0914902687072754, + "learning_rate": 3.58334542289279e-06, + "loss": 0.0584, + "step": 22732 + }, + { + "epoch": 2.464549002601908, + "grad_norm": 5.401479721069336, + "learning_rate": 3.5826200493254025e-06, + "loss": 0.0442, + "step": 22733 + }, + { + "epoch": 2.464657415437988, + "grad_norm": 0.25177034735679626, + "learning_rate": 3.581894675758016e-06, + "loss": 0.0013, + "step": 22734 + }, + { + "epoch": 2.464765828274068, + "grad_norm": 0.06965446472167969, + "learning_rate": 3.5811693021906285e-06, + "loss": 0.0006, + "step": 22735 + }, + { + "epoch": 2.4648742411101474, + "grad_norm": 0.05330218747258186, + "learning_rate": 3.580443928623241e-06, + "loss": 0.0003, + "step": 22736 + }, + { + "epoch": 2.4649826539462274, + "grad_norm": 2.5526835918426514, + "learning_rate": 3.579718555055854e-06, + "loss": 0.015, + "step": 22737 + }, + { + "epoch": 2.465091066782307, + "grad_norm": 0.9159331917762756, + "learning_rate": 3.578993181488467e-06, + "loss": 0.0065, + "step": 22738 + }, + { + "epoch": 2.465199479618387, + "grad_norm": 1.007144808769226, + "learning_rate": 3.5782678079210796e-06, + "loss": 0.001, + "step": 22739 + }, + { + "epoch": 2.4653078924544665, + "grad_norm": 0.013971483334898949, + "learning_rate": 3.5775424343536926e-06, + "loss": 0.0001, + "step": 22740 + }, + { + "epoch": 2.4654163052905465, + "grad_norm": 0.3710242211818695, + "learning_rate": 3.576817060786305e-06, + "loss": 0.0012, + "step": 22741 + }, + { + "epoch": 2.465524718126626, + "grad_norm": 0.034311871975660324, + "learning_rate": 3.576091687218918e-06, + "loss": 0.0005, + "step": 22742 + }, + { + "epoch": 2.465633130962706, + "grad_norm": 9.726738929748535, + "learning_rate": 3.575366313651531e-06, + "loss": 0.0306, + "step": 22743 + }, + { + "epoch": 2.465741543798786, + "grad_norm": 8.593317031860352, + "learning_rate": 3.5746409400841437e-06, + "loss": 0.0465, + "step": 22744 + }, + { + "epoch": 2.4658499566348655, + "grad_norm": 1.906693696975708, + "learning_rate": 3.5739155665167563e-06, + "loss": 0.0223, + "step": 22745 + }, + { + "epoch": 2.4659583694709455, + "grad_norm": 2.1484198570251465, + "learning_rate": 3.573190192949369e-06, + "loss": 0.0013, + "step": 22746 + }, + { + "epoch": 2.466066782307025, + "grad_norm": 5.097221374511719, + "learning_rate": 3.5724648193819823e-06, + "loss": 0.0066, + "step": 22747 + }, + { + "epoch": 2.466175195143105, + "grad_norm": 0.3664957880973816, + "learning_rate": 3.571739445814595e-06, + "loss": 0.0014, + "step": 22748 + }, + { + "epoch": 2.4662836079791846, + "grad_norm": 6.208045482635498, + "learning_rate": 3.5710140722472074e-06, + "loss": 0.0158, + "step": 22749 + }, + { + "epoch": 2.4663920208152645, + "grad_norm": 0.0540798082947731, + "learning_rate": 3.5702886986798204e-06, + "loss": 0.0005, + "step": 22750 + }, + { + "epoch": 2.466500433651344, + "grad_norm": 0.013615359552204609, + "learning_rate": 3.5695633251124334e-06, + "loss": 0.0001, + "step": 22751 + }, + { + "epoch": 2.466608846487424, + "grad_norm": 6.919017314910889, + "learning_rate": 3.568837951545046e-06, + "loss": 0.0089, + "step": 22752 + }, + { + "epoch": 2.466717259323504, + "grad_norm": 0.9579170346260071, + "learning_rate": 3.568112577977659e-06, + "loss": 0.0036, + "step": 22753 + }, + { + "epoch": 2.4668256721595836, + "grad_norm": 1.0571736097335815, + "learning_rate": 3.5673872044102715e-06, + "loss": 0.005, + "step": 22754 + }, + { + "epoch": 2.4669340849956636, + "grad_norm": 0.0063821859657764435, + "learning_rate": 3.5666618308428845e-06, + "loss": 0.0, + "step": 22755 + }, + { + "epoch": 2.467042497831743, + "grad_norm": 0.5931941270828247, + "learning_rate": 3.5659364572754975e-06, + "loss": 0.0022, + "step": 22756 + }, + { + "epoch": 2.467150910667823, + "grad_norm": 0.16022813320159912, + "learning_rate": 3.56521108370811e-06, + "loss": 0.001, + "step": 22757 + }, + { + "epoch": 2.467259323503903, + "grad_norm": 1.609318733215332, + "learning_rate": 3.5644857101407226e-06, + "loss": 0.0128, + "step": 22758 + }, + { + "epoch": 2.4673677363399826, + "grad_norm": 9.170994758605957, + "learning_rate": 3.563760336573335e-06, + "loss": 0.0178, + "step": 22759 + }, + { + "epoch": 2.4674761491760626, + "grad_norm": 6.31402063369751, + "learning_rate": 3.5630349630059486e-06, + "loss": 0.0121, + "step": 22760 + }, + { + "epoch": 2.467584562012142, + "grad_norm": 0.8070083260536194, + "learning_rate": 3.562309589438561e-06, + "loss": 0.0037, + "step": 22761 + }, + { + "epoch": 2.467692974848222, + "grad_norm": 1.2841302156448364, + "learning_rate": 3.5615842158711737e-06, + "loss": 0.0011, + "step": 22762 + }, + { + "epoch": 2.4678013876843017, + "grad_norm": 1.2510381937026978, + "learning_rate": 3.5608588423037867e-06, + "loss": 0.0049, + "step": 22763 + }, + { + "epoch": 2.4679098005203817, + "grad_norm": 7.5836501121521, + "learning_rate": 3.5601334687363997e-06, + "loss": 0.0375, + "step": 22764 + }, + { + "epoch": 2.468018213356461, + "grad_norm": 0.027034301310777664, + "learning_rate": 3.5594080951690123e-06, + "loss": 0.0001, + "step": 22765 + }, + { + "epoch": 2.468126626192541, + "grad_norm": 0.0318981297314167, + "learning_rate": 3.5586827216016253e-06, + "loss": 0.0002, + "step": 22766 + }, + { + "epoch": 2.468235039028621, + "grad_norm": 2.0829076766967773, + "learning_rate": 3.557957348034238e-06, + "loss": 0.0025, + "step": 22767 + }, + { + "epoch": 2.4683434518647007, + "grad_norm": 1.849442481994629, + "learning_rate": 3.557231974466851e-06, + "loss": 0.0014, + "step": 22768 + }, + { + "epoch": 2.4684518647007807, + "grad_norm": 0.050174813717603683, + "learning_rate": 3.556506600899464e-06, + "loss": 0.0002, + "step": 22769 + }, + { + "epoch": 2.4685602775368602, + "grad_norm": 1.785459041595459, + "learning_rate": 3.5557812273320764e-06, + "loss": 0.0475, + "step": 22770 + }, + { + "epoch": 2.4686686903729402, + "grad_norm": 2.597980499267578, + "learning_rate": 3.555055853764689e-06, + "loss": 0.004, + "step": 22771 + }, + { + "epoch": 2.4687771032090198, + "grad_norm": 1.4479933977127075, + "learning_rate": 3.5543304801973015e-06, + "loss": 0.0045, + "step": 22772 + }, + { + "epoch": 2.4688855160450998, + "grad_norm": 1.5332075357437134, + "learning_rate": 3.553605106629915e-06, + "loss": 0.0031, + "step": 22773 + }, + { + "epoch": 2.4689939288811793, + "grad_norm": 0.06106260046362877, + "learning_rate": 3.5528797330625275e-06, + "loss": 0.0005, + "step": 22774 + }, + { + "epoch": 2.4691023417172593, + "grad_norm": 3.705181360244751, + "learning_rate": 3.55215435949514e-06, + "loss": 0.0604, + "step": 22775 + }, + { + "epoch": 2.4692107545533393, + "grad_norm": 0.08315224200487137, + "learning_rate": 3.551428985927753e-06, + "loss": 0.0003, + "step": 22776 + }, + { + "epoch": 2.469319167389419, + "grad_norm": 2.1120340824127197, + "learning_rate": 3.550703612360366e-06, + "loss": 0.0331, + "step": 22777 + }, + { + "epoch": 2.469427580225499, + "grad_norm": 4.946652412414551, + "learning_rate": 3.5499782387929786e-06, + "loss": 0.0086, + "step": 22778 + }, + { + "epoch": 2.4695359930615783, + "grad_norm": 2.815011501312256, + "learning_rate": 3.5492528652255916e-06, + "loss": 0.0171, + "step": 22779 + }, + { + "epoch": 2.4696444058976583, + "grad_norm": 4.360067367553711, + "learning_rate": 3.548527491658204e-06, + "loss": 0.0069, + "step": 22780 + }, + { + "epoch": 2.4697528187337383, + "grad_norm": 0.06796874105930328, + "learning_rate": 3.547802118090817e-06, + "loss": 0.0007, + "step": 22781 + }, + { + "epoch": 2.469861231569818, + "grad_norm": 6.7648844718933105, + "learning_rate": 3.54707674452343e-06, + "loss": 0.027, + "step": 22782 + }, + { + "epoch": 2.469969644405898, + "grad_norm": 0.007816973142325878, + "learning_rate": 3.5463513709560427e-06, + "loss": 0.0001, + "step": 22783 + }, + { + "epoch": 2.4700780572419774, + "grad_norm": 0.04539605230093002, + "learning_rate": 3.5456259973886553e-06, + "loss": 0.0003, + "step": 22784 + }, + { + "epoch": 2.4701864700780574, + "grad_norm": 0.3190173804759979, + "learning_rate": 3.544900623821268e-06, + "loss": 0.0016, + "step": 22785 + }, + { + "epoch": 2.470294882914137, + "grad_norm": 0.7556195259094238, + "learning_rate": 3.5441752502538813e-06, + "loss": 0.0215, + "step": 22786 + }, + { + "epoch": 2.470403295750217, + "grad_norm": 0.36798205971717834, + "learning_rate": 3.543449876686494e-06, + "loss": 0.0013, + "step": 22787 + }, + { + "epoch": 2.4705117085862964, + "grad_norm": 0.7298470139503479, + "learning_rate": 3.5427245031191064e-06, + "loss": 0.0012, + "step": 22788 + }, + { + "epoch": 2.4706201214223764, + "grad_norm": 2.2638401985168457, + "learning_rate": 3.5419991295517194e-06, + "loss": 0.0064, + "step": 22789 + }, + { + "epoch": 2.4707285342584564, + "grad_norm": 2.931885242462158, + "learning_rate": 3.5412737559843324e-06, + "loss": 0.0057, + "step": 22790 + }, + { + "epoch": 2.470836947094536, + "grad_norm": 9.30078125, + "learning_rate": 3.540548382416945e-06, + "loss": 0.0448, + "step": 22791 + }, + { + "epoch": 2.470945359930616, + "grad_norm": 0.03658786788582802, + "learning_rate": 3.539823008849558e-06, + "loss": 0.0002, + "step": 22792 + }, + { + "epoch": 2.4710537727666955, + "grad_norm": 0.04644787684082985, + "learning_rate": 3.5390976352821705e-06, + "loss": 0.0002, + "step": 22793 + }, + { + "epoch": 2.4711621856027755, + "grad_norm": 3.351245880126953, + "learning_rate": 3.5383722617147835e-06, + "loss": 0.0048, + "step": 22794 + }, + { + "epoch": 2.471270598438855, + "grad_norm": 5.747386932373047, + "learning_rate": 3.5376468881473965e-06, + "loss": 0.0023, + "step": 22795 + }, + { + "epoch": 2.471379011274935, + "grad_norm": 2.0626981258392334, + "learning_rate": 3.536921514580009e-06, + "loss": 0.0111, + "step": 22796 + }, + { + "epoch": 2.4714874241110145, + "grad_norm": 3.310215950012207, + "learning_rate": 3.5361961410126216e-06, + "loss": 0.0092, + "step": 22797 + }, + { + "epoch": 2.4715958369470945, + "grad_norm": 2.4190828800201416, + "learning_rate": 3.535470767445234e-06, + "loss": 0.003, + "step": 22798 + }, + { + "epoch": 2.4717042497831745, + "grad_norm": 0.04467387497425079, + "learning_rate": 3.5347453938778476e-06, + "loss": 0.0005, + "step": 22799 + }, + { + "epoch": 2.471812662619254, + "grad_norm": 1.694786548614502, + "learning_rate": 3.53402002031046e-06, + "loss": 0.0042, + "step": 22800 + }, + { + "epoch": 2.471921075455334, + "grad_norm": 0.46567583084106445, + "learning_rate": 3.5332946467430728e-06, + "loss": 0.0019, + "step": 22801 + }, + { + "epoch": 2.4720294882914136, + "grad_norm": 0.11522986739873886, + "learning_rate": 3.5325692731756857e-06, + "loss": 0.0009, + "step": 22802 + }, + { + "epoch": 2.4721379011274935, + "grad_norm": 0.22993843257427216, + "learning_rate": 3.5318438996082987e-06, + "loss": 0.0009, + "step": 22803 + }, + { + "epoch": 2.4722463139635735, + "grad_norm": 4.053361892700195, + "learning_rate": 3.5311185260409113e-06, + "loss": 0.0149, + "step": 22804 + }, + { + "epoch": 2.472354726799653, + "grad_norm": 1.5284672975540161, + "learning_rate": 3.5303931524735243e-06, + "loss": 0.0013, + "step": 22805 + }, + { + "epoch": 2.472463139635733, + "grad_norm": 1.9392768144607544, + "learning_rate": 3.529667778906137e-06, + "loss": 0.009, + "step": 22806 + }, + { + "epoch": 2.4725715524718126, + "grad_norm": 0.5247803926467896, + "learning_rate": 3.52894240533875e-06, + "loss": 0.0015, + "step": 22807 + }, + { + "epoch": 2.4726799653078926, + "grad_norm": 2.9981131553649902, + "learning_rate": 3.528217031771363e-06, + "loss": 0.0115, + "step": 22808 + }, + { + "epoch": 2.472788378143972, + "grad_norm": 0.0736778974533081, + "learning_rate": 3.5274916582039754e-06, + "loss": 0.0004, + "step": 22809 + }, + { + "epoch": 2.472896790980052, + "grad_norm": 0.6111676692962646, + "learning_rate": 3.526766284636588e-06, + "loss": 0.002, + "step": 22810 + }, + { + "epoch": 2.4730052038161316, + "grad_norm": 12.932849884033203, + "learning_rate": 3.5260409110692005e-06, + "loss": 0.0495, + "step": 22811 + }, + { + "epoch": 2.4731136166522116, + "grad_norm": 0.9342201352119446, + "learning_rate": 3.525315537501814e-06, + "loss": 0.0042, + "step": 22812 + }, + { + "epoch": 2.4732220294882916, + "grad_norm": 3.1257903575897217, + "learning_rate": 3.5245901639344265e-06, + "loss": 0.03, + "step": 22813 + }, + { + "epoch": 2.473330442324371, + "grad_norm": 0.8532456159591675, + "learning_rate": 3.523864790367039e-06, + "loss": 0.0019, + "step": 22814 + }, + { + "epoch": 2.473438855160451, + "grad_norm": 0.0026986575685441494, + "learning_rate": 3.523139416799652e-06, + "loss": 0.0, + "step": 22815 + }, + { + "epoch": 2.4735472679965307, + "grad_norm": 0.36899930238723755, + "learning_rate": 3.522414043232265e-06, + "loss": 0.0004, + "step": 22816 + }, + { + "epoch": 2.4736556808326107, + "grad_norm": 3.0784077644348145, + "learning_rate": 3.5216886696648776e-06, + "loss": 0.0113, + "step": 22817 + }, + { + "epoch": 2.47376409366869, + "grad_norm": 27.76267433166504, + "learning_rate": 3.5209632960974906e-06, + "loss": 0.024, + "step": 22818 + }, + { + "epoch": 2.47387250650477, + "grad_norm": 0.13386304676532745, + "learning_rate": 3.520237922530103e-06, + "loss": 0.0007, + "step": 22819 + }, + { + "epoch": 2.4739809193408497, + "grad_norm": 3.4600231647491455, + "learning_rate": 3.519512548962716e-06, + "loss": 0.0229, + "step": 22820 + }, + { + "epoch": 2.4740893321769297, + "grad_norm": 3.8755855560302734, + "learning_rate": 3.518787175395329e-06, + "loss": 0.0109, + "step": 22821 + }, + { + "epoch": 2.4741977450130097, + "grad_norm": 3.552922248840332, + "learning_rate": 3.5180618018279417e-06, + "loss": 0.0178, + "step": 22822 + }, + { + "epoch": 2.4743061578490892, + "grad_norm": 0.10339171439409256, + "learning_rate": 3.5173364282605543e-06, + "loss": 0.0007, + "step": 22823 + }, + { + "epoch": 2.4744145706851692, + "grad_norm": 8.429275512695312, + "learning_rate": 3.516611054693167e-06, + "loss": 0.0253, + "step": 22824 + }, + { + "epoch": 2.4745229835212488, + "grad_norm": 0.08906228095293045, + "learning_rate": 3.5158856811257803e-06, + "loss": 0.0003, + "step": 22825 + }, + { + "epoch": 2.4746313963573288, + "grad_norm": 0.08312173932790756, + "learning_rate": 3.515160307558393e-06, + "loss": 0.0004, + "step": 22826 + }, + { + "epoch": 2.4747398091934087, + "grad_norm": 0.23185467720031738, + "learning_rate": 3.5144349339910054e-06, + "loss": 0.0012, + "step": 22827 + }, + { + "epoch": 2.4748482220294883, + "grad_norm": 0.016809426248073578, + "learning_rate": 3.5137095604236184e-06, + "loss": 0.0001, + "step": 22828 + }, + { + "epoch": 2.4749566348655683, + "grad_norm": 5.844465255737305, + "learning_rate": 3.5129841868562314e-06, + "loss": 0.0104, + "step": 22829 + }, + { + "epoch": 2.475065047701648, + "grad_norm": 0.9784310460090637, + "learning_rate": 3.512258813288844e-06, + "loss": 0.0041, + "step": 22830 + }, + { + "epoch": 2.475173460537728, + "grad_norm": 9.40433406829834, + "learning_rate": 3.511533439721457e-06, + "loss": 0.039, + "step": 22831 + }, + { + "epoch": 2.4752818733738073, + "grad_norm": 0.14981473982334137, + "learning_rate": 3.5108080661540695e-06, + "loss": 0.0004, + "step": 22832 + }, + { + "epoch": 2.4753902862098873, + "grad_norm": 0.35991111397743225, + "learning_rate": 3.5100826925866825e-06, + "loss": 0.0015, + "step": 22833 + }, + { + "epoch": 2.475498699045967, + "grad_norm": 0.7412014603614807, + "learning_rate": 3.5093573190192955e-06, + "loss": 0.0022, + "step": 22834 + }, + { + "epoch": 2.475607111882047, + "grad_norm": 9.4749755859375, + "learning_rate": 3.508631945451908e-06, + "loss": 0.1221, + "step": 22835 + }, + { + "epoch": 2.475715524718127, + "grad_norm": 6.899288654327393, + "learning_rate": 3.5079065718845206e-06, + "loss": 0.0442, + "step": 22836 + }, + { + "epoch": 2.4758239375542064, + "grad_norm": 11.278731346130371, + "learning_rate": 3.507181198317134e-06, + "loss": 0.007, + "step": 22837 + }, + { + "epoch": 2.4759323503902864, + "grad_norm": 1.6305774450302124, + "learning_rate": 3.5064558247497466e-06, + "loss": 0.0014, + "step": 22838 + }, + { + "epoch": 2.476040763226366, + "grad_norm": 4.01875638961792, + "learning_rate": 3.505730451182359e-06, + "loss": 0.0045, + "step": 22839 + }, + { + "epoch": 2.476149176062446, + "grad_norm": 0.20699132978916168, + "learning_rate": 3.5050050776149718e-06, + "loss": 0.0012, + "step": 22840 + }, + { + "epoch": 2.4762575888985254, + "grad_norm": 0.5533220171928406, + "learning_rate": 3.5042797040475847e-06, + "loss": 0.0016, + "step": 22841 + }, + { + "epoch": 2.4763660017346054, + "grad_norm": 3.7313783168792725, + "learning_rate": 3.5035543304801977e-06, + "loss": 0.0534, + "step": 22842 + }, + { + "epoch": 2.476474414570685, + "grad_norm": 0.45158901810646057, + "learning_rate": 3.5028289569128103e-06, + "loss": 0.0018, + "step": 22843 + }, + { + "epoch": 2.476582827406765, + "grad_norm": 0.02080499939620495, + "learning_rate": 3.5021035833454233e-06, + "loss": 0.0001, + "step": 22844 + }, + { + "epoch": 2.476691240242845, + "grad_norm": 1.3728985786437988, + "learning_rate": 3.501378209778036e-06, + "loss": 0.0034, + "step": 22845 + }, + { + "epoch": 2.4767996530789245, + "grad_norm": 5.737473011016846, + "learning_rate": 3.500652836210649e-06, + "loss": 0.0176, + "step": 22846 + }, + { + "epoch": 2.4769080659150045, + "grad_norm": 0.028051983565092087, + "learning_rate": 3.499927462643262e-06, + "loss": 0.0003, + "step": 22847 + }, + { + "epoch": 2.477016478751084, + "grad_norm": 0.047460734844207764, + "learning_rate": 3.4992020890758744e-06, + "loss": 0.0004, + "step": 22848 + }, + { + "epoch": 2.477124891587164, + "grad_norm": 0.02015076018869877, + "learning_rate": 3.498476715508487e-06, + "loss": 0.0002, + "step": 22849 + }, + { + "epoch": 2.4772333044232435, + "grad_norm": 2.272613048553467, + "learning_rate": 3.4977513419411004e-06, + "loss": 0.0071, + "step": 22850 + }, + { + "epoch": 2.4773417172593235, + "grad_norm": 2.878716230392456, + "learning_rate": 3.497025968373713e-06, + "loss": 0.0042, + "step": 22851 + }, + { + "epoch": 2.477450130095403, + "grad_norm": 0.2824380695819855, + "learning_rate": 3.4963005948063255e-06, + "loss": 0.0002, + "step": 22852 + }, + { + "epoch": 2.477558542931483, + "grad_norm": 0.5149388909339905, + "learning_rate": 3.495575221238938e-06, + "loss": 0.0016, + "step": 22853 + }, + { + "epoch": 2.477666955767563, + "grad_norm": 2.000502109527588, + "learning_rate": 3.494849847671551e-06, + "loss": 0.0636, + "step": 22854 + }, + { + "epoch": 2.4777753686036426, + "grad_norm": 0.12881183624267578, + "learning_rate": 3.494124474104164e-06, + "loss": 0.0005, + "step": 22855 + }, + { + "epoch": 2.4778837814397225, + "grad_norm": 6.472160816192627, + "learning_rate": 3.4933991005367766e-06, + "loss": 0.0047, + "step": 22856 + }, + { + "epoch": 2.477992194275802, + "grad_norm": 1.7198514938354492, + "learning_rate": 3.4926737269693896e-06, + "loss": 0.0062, + "step": 22857 + }, + { + "epoch": 2.478100607111882, + "grad_norm": 0.060403622686862946, + "learning_rate": 3.491948353402002e-06, + "loss": 0.0003, + "step": 22858 + }, + { + "epoch": 2.478209019947962, + "grad_norm": 0.028894834220409393, + "learning_rate": 3.491222979834615e-06, + "loss": 0.0002, + "step": 22859 + }, + { + "epoch": 2.4783174327840416, + "grad_norm": 1.8656797409057617, + "learning_rate": 3.490497606267228e-06, + "loss": 0.0353, + "step": 22860 + }, + { + "epoch": 2.4784258456201216, + "grad_norm": 0.8227660059928894, + "learning_rate": 3.4897722326998407e-06, + "loss": 0.0017, + "step": 22861 + }, + { + "epoch": 2.478534258456201, + "grad_norm": 0.03562913462519646, + "learning_rate": 3.4890468591324533e-06, + "loss": 0.0002, + "step": 22862 + }, + { + "epoch": 2.478642671292281, + "grad_norm": 2.2750415802001953, + "learning_rate": 3.4883214855650667e-06, + "loss": 0.0066, + "step": 22863 + }, + { + "epoch": 2.4787510841283606, + "grad_norm": 2.6112706661224365, + "learning_rate": 3.4875961119976793e-06, + "loss": 0.03, + "step": 22864 + }, + { + "epoch": 2.4788594969644406, + "grad_norm": 0.05006015673279762, + "learning_rate": 3.486870738430292e-06, + "loss": 0.0004, + "step": 22865 + }, + { + "epoch": 2.47896790980052, + "grad_norm": 0.39043501019477844, + "learning_rate": 3.4861453648629044e-06, + "loss": 0.0014, + "step": 22866 + }, + { + "epoch": 2.4790763226366, + "grad_norm": 2.3243680000305176, + "learning_rate": 3.4854199912955174e-06, + "loss": 0.0046, + "step": 22867 + }, + { + "epoch": 2.47918473547268, + "grad_norm": 0.20827390253543854, + "learning_rate": 3.4846946177281304e-06, + "loss": 0.0006, + "step": 22868 + }, + { + "epoch": 2.4792931483087597, + "grad_norm": 0.15436533093452454, + "learning_rate": 3.483969244160743e-06, + "loss": 0.0011, + "step": 22869 + }, + { + "epoch": 2.4794015611448397, + "grad_norm": 2.698453426361084, + "learning_rate": 3.483243870593356e-06, + "loss": 0.0303, + "step": 22870 + }, + { + "epoch": 2.479509973980919, + "grad_norm": 4.826581954956055, + "learning_rate": 3.4825184970259685e-06, + "loss": 0.0208, + "step": 22871 + }, + { + "epoch": 2.479618386816999, + "grad_norm": 5.000153541564941, + "learning_rate": 3.4817931234585815e-06, + "loss": 0.0073, + "step": 22872 + }, + { + "epoch": 2.4797267996530787, + "grad_norm": 0.2846654951572418, + "learning_rate": 3.4810677498911945e-06, + "loss": 0.0007, + "step": 22873 + }, + { + "epoch": 2.4798352124891587, + "grad_norm": 0.1509169340133667, + "learning_rate": 3.480342376323807e-06, + "loss": 0.0002, + "step": 22874 + }, + { + "epoch": 2.4799436253252383, + "grad_norm": 0.055149950087070465, + "learning_rate": 3.4796170027564196e-06, + "loss": 0.0004, + "step": 22875 + }, + { + "epoch": 2.4800520381613183, + "grad_norm": 2.3060615062713623, + "learning_rate": 3.478891629189033e-06, + "loss": 0.0396, + "step": 22876 + }, + { + "epoch": 2.4801604509973982, + "grad_norm": 4.16880464553833, + "learning_rate": 3.4781662556216456e-06, + "loss": 0.0044, + "step": 22877 + }, + { + "epoch": 2.4802688638334778, + "grad_norm": 0.024679116904735565, + "learning_rate": 3.477440882054258e-06, + "loss": 0.0002, + "step": 22878 + }, + { + "epoch": 2.4803772766695578, + "grad_norm": 0.01913786493241787, + "learning_rate": 3.4767155084868708e-06, + "loss": 0.0002, + "step": 22879 + }, + { + "epoch": 2.4804856895056373, + "grad_norm": 0.611009418964386, + "learning_rate": 3.4759901349194837e-06, + "loss": 0.0028, + "step": 22880 + }, + { + "epoch": 2.4805941023417173, + "grad_norm": 1.90372896194458, + "learning_rate": 3.4752647613520967e-06, + "loss": 0.0354, + "step": 22881 + }, + { + "epoch": 2.4807025151777973, + "grad_norm": 0.05589364096522331, + "learning_rate": 3.4745393877847093e-06, + "loss": 0.0002, + "step": 22882 + }, + { + "epoch": 2.480810928013877, + "grad_norm": 1.3239059448242188, + "learning_rate": 3.4738140142173223e-06, + "loss": 0.0041, + "step": 22883 + }, + { + "epoch": 2.480919340849957, + "grad_norm": 0.23733212053775787, + "learning_rate": 3.473088640649935e-06, + "loss": 0.0006, + "step": 22884 + }, + { + "epoch": 2.4810277536860363, + "grad_norm": 4.257785797119141, + "learning_rate": 3.472363267082548e-06, + "loss": 0.0089, + "step": 22885 + }, + { + "epoch": 2.4811361665221163, + "grad_norm": 1.044931173324585, + "learning_rate": 3.471637893515161e-06, + "loss": 0.0061, + "step": 22886 + }, + { + "epoch": 2.481244579358196, + "grad_norm": 0.4561954438686371, + "learning_rate": 3.4709125199477734e-06, + "loss": 0.001, + "step": 22887 + }, + { + "epoch": 2.481352992194276, + "grad_norm": 15.103605270385742, + "learning_rate": 3.470187146380386e-06, + "loss": 0.0327, + "step": 22888 + }, + { + "epoch": 2.4814614050303554, + "grad_norm": 8.237529754638672, + "learning_rate": 3.4694617728129994e-06, + "loss": 0.031, + "step": 22889 + }, + { + "epoch": 2.4815698178664354, + "grad_norm": 0.33701804280281067, + "learning_rate": 3.468736399245612e-06, + "loss": 0.001, + "step": 22890 + }, + { + "epoch": 2.4816782307025154, + "grad_norm": 3.1783082485198975, + "learning_rate": 3.4680110256782245e-06, + "loss": 0.0195, + "step": 22891 + }, + { + "epoch": 2.481786643538595, + "grad_norm": 0.06289319694042206, + "learning_rate": 3.467285652110837e-06, + "loss": 0.0003, + "step": 22892 + }, + { + "epoch": 2.481895056374675, + "grad_norm": 1.3361308574676514, + "learning_rate": 3.46656027854345e-06, + "loss": 0.0044, + "step": 22893 + }, + { + "epoch": 2.4820034692107544, + "grad_norm": 0.06091023236513138, + "learning_rate": 3.465834904976063e-06, + "loss": 0.0001, + "step": 22894 + }, + { + "epoch": 2.4821118820468344, + "grad_norm": 0.01857205480337143, + "learning_rate": 3.4651095314086756e-06, + "loss": 0.0003, + "step": 22895 + }, + { + "epoch": 2.482220294882914, + "grad_norm": 1.2382138967514038, + "learning_rate": 3.4643841578412886e-06, + "loss": 0.0081, + "step": 22896 + }, + { + "epoch": 2.482328707718994, + "grad_norm": 0.31674259901046753, + "learning_rate": 3.463658784273901e-06, + "loss": 0.0012, + "step": 22897 + }, + { + "epoch": 2.4824371205550735, + "grad_norm": 0.23099324107170105, + "learning_rate": 3.462933410706514e-06, + "loss": 0.0004, + "step": 22898 + }, + { + "epoch": 2.4825455333911535, + "grad_norm": 7.179523944854736, + "learning_rate": 3.462208037139127e-06, + "loss": 0.0057, + "step": 22899 + }, + { + "epoch": 2.4826539462272335, + "grad_norm": 0.25381767749786377, + "learning_rate": 3.4614826635717397e-06, + "loss": 0.0017, + "step": 22900 + }, + { + "epoch": 2.482762359063313, + "grad_norm": 2.58506441116333, + "learning_rate": 3.4607572900043523e-06, + "loss": 0.0695, + "step": 22901 + }, + { + "epoch": 2.482870771899393, + "grad_norm": 0.2542347013950348, + "learning_rate": 3.4600319164369657e-06, + "loss": 0.0009, + "step": 22902 + }, + { + "epoch": 2.4829791847354725, + "grad_norm": 0.02722521871328354, + "learning_rate": 3.4593065428695783e-06, + "loss": 0.0003, + "step": 22903 + }, + { + "epoch": 2.4830875975715525, + "grad_norm": 0.0978928878903389, + "learning_rate": 3.458581169302191e-06, + "loss": 0.001, + "step": 22904 + }, + { + "epoch": 2.4831960104076325, + "grad_norm": 4.4842143058776855, + "learning_rate": 3.4578557957348034e-06, + "loss": 0.0469, + "step": 22905 + }, + { + "epoch": 2.483304423243712, + "grad_norm": 4.527277946472168, + "learning_rate": 3.4571304221674164e-06, + "loss": 0.0295, + "step": 22906 + }, + { + "epoch": 2.483412836079792, + "grad_norm": 0.06359902024269104, + "learning_rate": 3.4564050486000294e-06, + "loss": 0.0006, + "step": 22907 + }, + { + "epoch": 2.4835212489158716, + "grad_norm": 2.3636858463287354, + "learning_rate": 3.455679675032642e-06, + "loss": 0.0083, + "step": 22908 + }, + { + "epoch": 2.4836296617519515, + "grad_norm": 8.74274730682373, + "learning_rate": 3.454954301465255e-06, + "loss": 0.0284, + "step": 22909 + }, + { + "epoch": 2.483738074588031, + "grad_norm": 3.6581695079803467, + "learning_rate": 3.4542289278978675e-06, + "loss": 0.0039, + "step": 22910 + }, + { + "epoch": 2.483846487424111, + "grad_norm": 1.3452273607254028, + "learning_rate": 3.4535035543304805e-06, + "loss": 0.0025, + "step": 22911 + }, + { + "epoch": 2.4839549002601906, + "grad_norm": 0.0865858718752861, + "learning_rate": 3.4527781807630935e-06, + "loss": 0.0004, + "step": 22912 + }, + { + "epoch": 2.4840633130962706, + "grad_norm": 0.5842334628105164, + "learning_rate": 3.452052807195706e-06, + "loss": 0.0017, + "step": 22913 + }, + { + "epoch": 2.4841717259323506, + "grad_norm": 3.102238893508911, + "learning_rate": 3.4513274336283186e-06, + "loss": 0.0473, + "step": 22914 + }, + { + "epoch": 2.48428013876843, + "grad_norm": 0.7130181193351746, + "learning_rate": 3.450602060060932e-06, + "loss": 0.0016, + "step": 22915 + }, + { + "epoch": 2.48438855160451, + "grad_norm": 0.23352384567260742, + "learning_rate": 3.4498766864935446e-06, + "loss": 0.001, + "step": 22916 + }, + { + "epoch": 2.4844969644405897, + "grad_norm": 0.060385461896657944, + "learning_rate": 3.449151312926157e-06, + "loss": 0.0002, + "step": 22917 + }, + { + "epoch": 2.4846053772766696, + "grad_norm": 0.2768060266971588, + "learning_rate": 3.4484259393587698e-06, + "loss": 0.0012, + "step": 22918 + }, + { + "epoch": 2.484713790112749, + "grad_norm": 3.2049126625061035, + "learning_rate": 3.4477005657913828e-06, + "loss": 0.0061, + "step": 22919 + }, + { + "epoch": 2.484822202948829, + "grad_norm": 0.3435153663158417, + "learning_rate": 3.4469751922239957e-06, + "loss": 0.0015, + "step": 22920 + }, + { + "epoch": 2.4849306157849087, + "grad_norm": 1.873500943183899, + "learning_rate": 3.4462498186566083e-06, + "loss": 0.0035, + "step": 22921 + }, + { + "epoch": 2.4850390286209887, + "grad_norm": 0.7002717852592468, + "learning_rate": 3.4455244450892213e-06, + "loss": 0.003, + "step": 22922 + }, + { + "epoch": 2.4851474414570687, + "grad_norm": 3.613816499710083, + "learning_rate": 3.444799071521834e-06, + "loss": 0.0131, + "step": 22923 + }, + { + "epoch": 2.485255854293148, + "grad_norm": 24.859905242919922, + "learning_rate": 3.444073697954447e-06, + "loss": 0.0112, + "step": 22924 + }, + { + "epoch": 2.485364267129228, + "grad_norm": 0.04268530383706093, + "learning_rate": 3.44334832438706e-06, + "loss": 0.0006, + "step": 22925 + }, + { + "epoch": 2.4854726799653077, + "grad_norm": 1.9237840175628662, + "learning_rate": 3.4426229508196724e-06, + "loss": 0.0335, + "step": 22926 + }, + { + "epoch": 2.4855810928013877, + "grad_norm": 0.029365185648202896, + "learning_rate": 3.441897577252285e-06, + "loss": 0.0001, + "step": 22927 + }, + { + "epoch": 2.4856895056374677, + "grad_norm": 0.24216002225875854, + "learning_rate": 3.4411722036848984e-06, + "loss": 0.0011, + "step": 22928 + }, + { + "epoch": 2.4857979184735473, + "grad_norm": 0.018204450607299805, + "learning_rate": 3.440446830117511e-06, + "loss": 0.0002, + "step": 22929 + }, + { + "epoch": 2.4859063313096272, + "grad_norm": 0.09465119242668152, + "learning_rate": 3.4397214565501235e-06, + "loss": 0.001, + "step": 22930 + }, + { + "epoch": 2.486014744145707, + "grad_norm": 1.3062210083007812, + "learning_rate": 3.438996082982736e-06, + "loss": 0.0312, + "step": 22931 + }, + { + "epoch": 2.4861231569817868, + "grad_norm": 0.9683482646942139, + "learning_rate": 3.4382707094153495e-06, + "loss": 0.0021, + "step": 22932 + }, + { + "epoch": 2.4862315698178663, + "grad_norm": 6.110036849975586, + "learning_rate": 3.437545335847962e-06, + "loss": 0.0154, + "step": 22933 + }, + { + "epoch": 2.4863399826539463, + "grad_norm": 0.05574308708310127, + "learning_rate": 3.4368199622805746e-06, + "loss": 0.0004, + "step": 22934 + }, + { + "epoch": 2.486448395490026, + "grad_norm": 4.902334213256836, + "learning_rate": 3.4360945887131876e-06, + "loss": 0.0388, + "step": 22935 + }, + { + "epoch": 2.486556808326106, + "grad_norm": 0.024015674367547035, + "learning_rate": 3.4353692151458e-06, + "loss": 0.0002, + "step": 22936 + }, + { + "epoch": 2.486665221162186, + "grad_norm": 0.04854520037770271, + "learning_rate": 3.434643841578413e-06, + "loss": 0.0003, + "step": 22937 + }, + { + "epoch": 2.4867736339982653, + "grad_norm": 0.8066280484199524, + "learning_rate": 3.433918468011026e-06, + "loss": 0.0007, + "step": 22938 + }, + { + "epoch": 2.4868820468343453, + "grad_norm": 0.005483965389430523, + "learning_rate": 3.4331930944436387e-06, + "loss": 0.0001, + "step": 22939 + }, + { + "epoch": 2.486990459670425, + "grad_norm": 2.8637535572052, + "learning_rate": 3.4324677208762513e-06, + "loss": 0.0115, + "step": 22940 + }, + { + "epoch": 2.487098872506505, + "grad_norm": 1.189000129699707, + "learning_rate": 3.4317423473088647e-06, + "loss": 0.004, + "step": 22941 + }, + { + "epoch": 2.4872072853425844, + "grad_norm": 0.100205197930336, + "learning_rate": 3.4310169737414773e-06, + "loss": 0.0009, + "step": 22942 + }, + { + "epoch": 2.4873156981786644, + "grad_norm": 0.9595112800598145, + "learning_rate": 3.43029160017409e-06, + "loss": 0.0045, + "step": 22943 + }, + { + "epoch": 2.487424111014744, + "grad_norm": 3.710970640182495, + "learning_rate": 3.4295662266067024e-06, + "loss": 0.0167, + "step": 22944 + }, + { + "epoch": 2.487532523850824, + "grad_norm": 0.4018854796886444, + "learning_rate": 3.428840853039316e-06, + "loss": 0.0015, + "step": 22945 + }, + { + "epoch": 2.487640936686904, + "grad_norm": 0.0707402378320694, + "learning_rate": 3.4281154794719284e-06, + "loss": 0.0005, + "step": 22946 + }, + { + "epoch": 2.4877493495229834, + "grad_norm": 0.13588008284568787, + "learning_rate": 3.427390105904541e-06, + "loss": 0.0006, + "step": 22947 + }, + { + "epoch": 2.4878577623590634, + "grad_norm": 0.3270334303379059, + "learning_rate": 3.426664732337154e-06, + "loss": 0.0016, + "step": 22948 + }, + { + "epoch": 2.487966175195143, + "grad_norm": 8.103228569030762, + "learning_rate": 3.4259393587697665e-06, + "loss": 0.0472, + "step": 22949 + }, + { + "epoch": 2.488074588031223, + "grad_norm": 2.3987808227539062, + "learning_rate": 3.4252139852023795e-06, + "loss": 0.0062, + "step": 22950 + }, + { + "epoch": 2.488183000867303, + "grad_norm": 0.03702884167432785, + "learning_rate": 3.4244886116349925e-06, + "loss": 0.0001, + "step": 22951 + }, + { + "epoch": 2.4882914137033825, + "grad_norm": 0.05978510156273842, + "learning_rate": 3.423763238067605e-06, + "loss": 0.0006, + "step": 22952 + }, + { + "epoch": 2.4883998265394625, + "grad_norm": 0.038534555584192276, + "learning_rate": 3.4230378645002176e-06, + "loss": 0.0002, + "step": 22953 + }, + { + "epoch": 2.488508239375542, + "grad_norm": 0.04278263449668884, + "learning_rate": 3.422312490932831e-06, + "loss": 0.0003, + "step": 22954 + }, + { + "epoch": 2.488616652211622, + "grad_norm": 2.3404035568237305, + "learning_rate": 3.4215871173654436e-06, + "loss": 0.0298, + "step": 22955 + }, + { + "epoch": 2.4887250650477015, + "grad_norm": 0.3442111909389496, + "learning_rate": 3.420861743798056e-06, + "loss": 0.0016, + "step": 22956 + }, + { + "epoch": 2.4888334778837815, + "grad_norm": 3.0075464248657227, + "learning_rate": 3.4201363702306688e-06, + "loss": 0.0126, + "step": 22957 + }, + { + "epoch": 2.488941890719861, + "grad_norm": 0.49087315797805786, + "learning_rate": 3.419410996663282e-06, + "loss": 0.0006, + "step": 22958 + }, + { + "epoch": 2.489050303555941, + "grad_norm": 13.369634628295898, + "learning_rate": 3.4186856230958947e-06, + "loss": 0.0197, + "step": 22959 + }, + { + "epoch": 2.489158716392021, + "grad_norm": 2.2839088439941406, + "learning_rate": 3.4179602495285073e-06, + "loss": 0.0043, + "step": 22960 + }, + { + "epoch": 2.4892671292281006, + "grad_norm": 0.03330478072166443, + "learning_rate": 3.4172348759611203e-06, + "loss": 0.0003, + "step": 22961 + }, + { + "epoch": 2.4893755420641805, + "grad_norm": 0.035550303757190704, + "learning_rate": 3.416509502393733e-06, + "loss": 0.0004, + "step": 22962 + }, + { + "epoch": 2.48948395490026, + "grad_norm": 0.01060801837593317, + "learning_rate": 3.415784128826346e-06, + "loss": 0.0001, + "step": 22963 + }, + { + "epoch": 2.48959236773634, + "grad_norm": 3.709089994430542, + "learning_rate": 3.415058755258959e-06, + "loss": 0.0456, + "step": 22964 + }, + { + "epoch": 2.4897007805724196, + "grad_norm": 0.02274378389120102, + "learning_rate": 3.4143333816915714e-06, + "loss": 0.0001, + "step": 22965 + }, + { + "epoch": 2.4898091934084996, + "grad_norm": 2.491024971008301, + "learning_rate": 3.413608008124184e-06, + "loss": 0.0048, + "step": 22966 + }, + { + "epoch": 2.489917606244579, + "grad_norm": 10.738842964172363, + "learning_rate": 3.4128826345567974e-06, + "loss": 0.0081, + "step": 22967 + }, + { + "epoch": 2.490026019080659, + "grad_norm": 1.9880499839782715, + "learning_rate": 3.41215726098941e-06, + "loss": 0.0025, + "step": 22968 + }, + { + "epoch": 2.490134431916739, + "grad_norm": 4.568389892578125, + "learning_rate": 3.4114318874220225e-06, + "loss": 0.0147, + "step": 22969 + }, + { + "epoch": 2.4902428447528187, + "grad_norm": 15.830391883850098, + "learning_rate": 3.410706513854635e-06, + "loss": 0.0328, + "step": 22970 + }, + { + "epoch": 2.4903512575888986, + "grad_norm": 0.7142783999443054, + "learning_rate": 3.4099811402872485e-06, + "loss": 0.0016, + "step": 22971 + }, + { + "epoch": 2.490459670424978, + "grad_norm": 0.028856920078396797, + "learning_rate": 3.409255766719861e-06, + "loss": 0.0003, + "step": 22972 + }, + { + "epoch": 2.490568083261058, + "grad_norm": 0.7934007048606873, + "learning_rate": 3.4085303931524736e-06, + "loss": 0.0028, + "step": 22973 + }, + { + "epoch": 2.490676496097138, + "grad_norm": 0.1584736555814743, + "learning_rate": 3.4078050195850866e-06, + "loss": 0.0008, + "step": 22974 + }, + { + "epoch": 2.4907849089332177, + "grad_norm": 0.1899283528327942, + "learning_rate": 3.407079646017699e-06, + "loss": 0.0012, + "step": 22975 + }, + { + "epoch": 2.4908933217692977, + "grad_norm": 1.8390110731124878, + "learning_rate": 3.406354272450312e-06, + "loss": 0.005, + "step": 22976 + }, + { + "epoch": 2.491001734605377, + "grad_norm": 0.13974244892597198, + "learning_rate": 3.405628898882925e-06, + "loss": 0.0007, + "step": 22977 + }, + { + "epoch": 2.491110147441457, + "grad_norm": 1.893828272819519, + "learning_rate": 3.4049035253155378e-06, + "loss": 0.0024, + "step": 22978 + }, + { + "epoch": 2.4912185602775367, + "grad_norm": 3.2191991806030273, + "learning_rate": 3.4041781517481503e-06, + "loss": 0.048, + "step": 22979 + }, + { + "epoch": 2.4913269731136167, + "grad_norm": 2.965564727783203, + "learning_rate": 3.4034527781807637e-06, + "loss": 0.0435, + "step": 22980 + }, + { + "epoch": 2.4914353859496963, + "grad_norm": 3.389590263366699, + "learning_rate": 3.4027274046133763e-06, + "loss": 0.0181, + "step": 22981 + }, + { + "epoch": 2.4915437987857763, + "grad_norm": 0.24464832246303558, + "learning_rate": 3.402002031045989e-06, + "loss": 0.001, + "step": 22982 + }, + { + "epoch": 2.4916522116218562, + "grad_norm": 1.9846466779708862, + "learning_rate": 3.4012766574786014e-06, + "loss": 0.0119, + "step": 22983 + }, + { + "epoch": 2.491760624457936, + "grad_norm": 6.194106101989746, + "learning_rate": 3.400551283911215e-06, + "loss": 0.03, + "step": 22984 + }, + { + "epoch": 2.4918690372940158, + "grad_norm": 0.38033798336982727, + "learning_rate": 3.3998259103438274e-06, + "loss": 0.0014, + "step": 22985 + }, + { + "epoch": 2.4919774501300953, + "grad_norm": 1.2134299278259277, + "learning_rate": 3.39910053677644e-06, + "loss": 0.002, + "step": 22986 + }, + { + "epoch": 2.4920858629661753, + "grad_norm": 0.14503757655620575, + "learning_rate": 3.398375163209053e-06, + "loss": 0.0016, + "step": 22987 + }, + { + "epoch": 2.492194275802255, + "grad_norm": 0.05658329650759697, + "learning_rate": 3.3976497896416655e-06, + "loss": 0.0006, + "step": 22988 + }, + { + "epoch": 2.492302688638335, + "grad_norm": 2.010242462158203, + "learning_rate": 3.3969244160742785e-06, + "loss": 0.0031, + "step": 22989 + }, + { + "epoch": 2.4924111014744144, + "grad_norm": 0.9347105622291565, + "learning_rate": 3.3961990425068915e-06, + "loss": 0.0034, + "step": 22990 + }, + { + "epoch": 2.4925195143104943, + "grad_norm": 3.0961201190948486, + "learning_rate": 3.395473668939504e-06, + "loss": 0.0279, + "step": 22991 + }, + { + "epoch": 2.4926279271465743, + "grad_norm": 1.064948320388794, + "learning_rate": 3.3947482953721167e-06, + "loss": 0.0033, + "step": 22992 + }, + { + "epoch": 2.492736339982654, + "grad_norm": 0.017122574150562286, + "learning_rate": 3.39402292180473e-06, + "loss": 0.0002, + "step": 22993 + }, + { + "epoch": 2.492844752818734, + "grad_norm": 0.06742701679468155, + "learning_rate": 3.3932975482373426e-06, + "loss": 0.0006, + "step": 22994 + }, + { + "epoch": 2.4929531656548134, + "grad_norm": 0.6920676827430725, + "learning_rate": 3.392572174669955e-06, + "loss": 0.0013, + "step": 22995 + }, + { + "epoch": 2.4930615784908934, + "grad_norm": 0.8016438484191895, + "learning_rate": 3.3918468011025678e-06, + "loss": 0.0023, + "step": 22996 + }, + { + "epoch": 2.493169991326973, + "grad_norm": 4.005145072937012, + "learning_rate": 3.391121427535181e-06, + "loss": 0.0183, + "step": 22997 + }, + { + "epoch": 2.493278404163053, + "grad_norm": 10.304577827453613, + "learning_rate": 3.3903960539677937e-06, + "loss": 0.0142, + "step": 22998 + }, + { + "epoch": 2.4933868169991324, + "grad_norm": 0.6569831371307373, + "learning_rate": 3.3896706804004063e-06, + "loss": 0.0012, + "step": 22999 + }, + { + "epoch": 2.4934952298352124, + "grad_norm": 2.3588716983795166, + "learning_rate": 3.3889453068330193e-06, + "loss": 0.0135, + "step": 23000 + }, + { + "epoch": 2.4936036426712924, + "grad_norm": 0.06463166326284409, + "learning_rate": 3.388219933265632e-06, + "loss": 0.0006, + "step": 23001 + }, + { + "epoch": 2.493712055507372, + "grad_norm": 7.612314224243164, + "learning_rate": 3.387494559698245e-06, + "loss": 0.0527, + "step": 23002 + }, + { + "epoch": 2.493820468343452, + "grad_norm": 0.22595803439617157, + "learning_rate": 3.386769186130858e-06, + "loss": 0.0015, + "step": 23003 + }, + { + "epoch": 2.4939288811795315, + "grad_norm": 0.19318650662899017, + "learning_rate": 3.3860438125634704e-06, + "loss": 0.0013, + "step": 23004 + }, + { + "epoch": 2.4940372940156115, + "grad_norm": 0.013607192784547806, + "learning_rate": 3.385318438996083e-06, + "loss": 0.0001, + "step": 23005 + }, + { + "epoch": 2.4941457068516915, + "grad_norm": 0.1181551143527031, + "learning_rate": 3.3845930654286964e-06, + "loss": 0.0015, + "step": 23006 + }, + { + "epoch": 2.494254119687771, + "grad_norm": 0.28578752279281616, + "learning_rate": 3.383867691861309e-06, + "loss": 0.0008, + "step": 23007 + }, + { + "epoch": 2.494362532523851, + "grad_norm": 0.5179052352905273, + "learning_rate": 3.3831423182939215e-06, + "loss": 0.0023, + "step": 23008 + }, + { + "epoch": 2.4944709453599305, + "grad_norm": 0.07988764345645905, + "learning_rate": 3.382416944726534e-06, + "loss": 0.0003, + "step": 23009 + }, + { + "epoch": 2.4945793581960105, + "grad_norm": 3.369380474090576, + "learning_rate": 3.3816915711591475e-06, + "loss": 0.005, + "step": 23010 + }, + { + "epoch": 2.49468777103209, + "grad_norm": 5.167849540710449, + "learning_rate": 3.38096619759176e-06, + "loss": 0.0154, + "step": 23011 + }, + { + "epoch": 2.49479618386817, + "grad_norm": 0.05787253379821777, + "learning_rate": 3.3802408240243726e-06, + "loss": 0.0007, + "step": 23012 + }, + { + "epoch": 2.4949045967042496, + "grad_norm": 0.23587666451931, + "learning_rate": 3.3795154504569856e-06, + "loss": 0.0011, + "step": 23013 + }, + { + "epoch": 2.4950130095403296, + "grad_norm": 0.27785390615463257, + "learning_rate": 3.378790076889598e-06, + "loss": 0.0013, + "step": 23014 + }, + { + "epoch": 2.4951214223764095, + "grad_norm": 2.7756075859069824, + "learning_rate": 3.378064703322211e-06, + "loss": 0.0123, + "step": 23015 + }, + { + "epoch": 2.495229835212489, + "grad_norm": 2.5054256916046143, + "learning_rate": 3.377339329754824e-06, + "loss": 0.018, + "step": 23016 + }, + { + "epoch": 2.495338248048569, + "grad_norm": 0.13528086245059967, + "learning_rate": 3.3766139561874368e-06, + "loss": 0.0004, + "step": 23017 + }, + { + "epoch": 2.4954466608846486, + "grad_norm": 4.193513870239258, + "learning_rate": 3.3758885826200493e-06, + "loss": 0.0196, + "step": 23018 + }, + { + "epoch": 2.4955550737207286, + "grad_norm": 0.1114054024219513, + "learning_rate": 3.3751632090526627e-06, + "loss": 0.0009, + "step": 23019 + }, + { + "epoch": 2.495663486556808, + "grad_norm": 7.362367630004883, + "learning_rate": 3.3744378354852753e-06, + "loss": 0.0135, + "step": 23020 + }, + { + "epoch": 2.495771899392888, + "grad_norm": 0.029277903959155083, + "learning_rate": 3.373712461917888e-06, + "loss": 0.0003, + "step": 23021 + }, + { + "epoch": 2.4958803122289677, + "grad_norm": 0.04570770636200905, + "learning_rate": 3.3729870883505004e-06, + "loss": 0.0003, + "step": 23022 + }, + { + "epoch": 2.4959887250650477, + "grad_norm": 0.04980771243572235, + "learning_rate": 3.372261714783114e-06, + "loss": 0.0004, + "step": 23023 + }, + { + "epoch": 2.4960971379011276, + "grad_norm": 0.14868488907814026, + "learning_rate": 3.3715363412157264e-06, + "loss": 0.0004, + "step": 23024 + }, + { + "epoch": 2.496205550737207, + "grad_norm": 2.0798535346984863, + "learning_rate": 3.370810967648339e-06, + "loss": 0.0037, + "step": 23025 + }, + { + "epoch": 2.496313963573287, + "grad_norm": 0.0465661957859993, + "learning_rate": 3.370085594080952e-06, + "loss": 0.0003, + "step": 23026 + }, + { + "epoch": 2.4964223764093667, + "grad_norm": 1.008539080619812, + "learning_rate": 3.369360220513565e-06, + "loss": 0.0018, + "step": 23027 + }, + { + "epoch": 2.4965307892454467, + "grad_norm": 15.631096839904785, + "learning_rate": 3.3686348469461775e-06, + "loss": 0.0425, + "step": 23028 + }, + { + "epoch": 2.4966392020815267, + "grad_norm": 0.11460760235786438, + "learning_rate": 3.3679094733787905e-06, + "loss": 0.0009, + "step": 23029 + }, + { + "epoch": 2.496747614917606, + "grad_norm": 30.70831871032715, + "learning_rate": 3.367184099811403e-06, + "loss": 0.0131, + "step": 23030 + }, + { + "epoch": 2.496856027753686, + "grad_norm": 1.865046739578247, + "learning_rate": 3.3664587262440157e-06, + "loss": 0.0167, + "step": 23031 + }, + { + "epoch": 2.4969644405897657, + "grad_norm": 0.17631085216999054, + "learning_rate": 3.365733352676629e-06, + "loss": 0.0005, + "step": 23032 + }, + { + "epoch": 2.4970728534258457, + "grad_norm": 2.135159969329834, + "learning_rate": 3.3650079791092416e-06, + "loss": 0.0205, + "step": 23033 + }, + { + "epoch": 2.4971812662619253, + "grad_norm": 2.9237794876098633, + "learning_rate": 3.364282605541854e-06, + "loss": 0.0342, + "step": 23034 + }, + { + "epoch": 2.4972896790980053, + "grad_norm": 0.9685220718383789, + "learning_rate": 3.3635572319744668e-06, + "loss": 0.0031, + "step": 23035 + }, + { + "epoch": 2.497398091934085, + "grad_norm": 10.215263366699219, + "learning_rate": 3.36283185840708e-06, + "loss": 0.0149, + "step": 23036 + }, + { + "epoch": 2.497506504770165, + "grad_norm": 0.0228715967386961, + "learning_rate": 3.3621064848396928e-06, + "loss": 0.0002, + "step": 23037 + }, + { + "epoch": 2.4976149176062448, + "grad_norm": 0.05072406306862831, + "learning_rate": 3.3613811112723053e-06, + "loss": 0.0001, + "step": 23038 + }, + { + "epoch": 2.4977233304423243, + "grad_norm": 0.10466495901346207, + "learning_rate": 3.3606557377049183e-06, + "loss": 0.0001, + "step": 23039 + }, + { + "epoch": 2.4978317432784043, + "grad_norm": 0.07557450234889984, + "learning_rate": 3.3599303641375313e-06, + "loss": 0.0004, + "step": 23040 + }, + { + "epoch": 2.497940156114484, + "grad_norm": 0.06405511498451233, + "learning_rate": 3.359204990570144e-06, + "loss": 0.0005, + "step": 23041 + }, + { + "epoch": 2.498048568950564, + "grad_norm": 7.219640254974365, + "learning_rate": 3.358479617002757e-06, + "loss": 0.0471, + "step": 23042 + }, + { + "epoch": 2.4981569817866434, + "grad_norm": 0.011558802798390388, + "learning_rate": 3.3577542434353694e-06, + "loss": 0.0001, + "step": 23043 + }, + { + "epoch": 2.4982653946227233, + "grad_norm": 0.030001932755112648, + "learning_rate": 3.357028869867982e-06, + "loss": 0.0003, + "step": 23044 + }, + { + "epoch": 2.498373807458803, + "grad_norm": 0.07570812106132507, + "learning_rate": 3.3563034963005954e-06, + "loss": 0.0004, + "step": 23045 + }, + { + "epoch": 2.498482220294883, + "grad_norm": 7.0669450759887695, + "learning_rate": 3.355578122733208e-06, + "loss": 0.0039, + "step": 23046 + }, + { + "epoch": 2.498590633130963, + "grad_norm": 0.007962443865835667, + "learning_rate": 3.3548527491658205e-06, + "loss": 0.0001, + "step": 23047 + }, + { + "epoch": 2.4986990459670424, + "grad_norm": 0.25812652707099915, + "learning_rate": 3.354127375598433e-06, + "loss": 0.0004, + "step": 23048 + }, + { + "epoch": 2.4988074588031224, + "grad_norm": 0.020397411659359932, + "learning_rate": 3.3534020020310465e-06, + "loss": 0.0001, + "step": 23049 + }, + { + "epoch": 2.498915871639202, + "grad_norm": 1.418874740600586, + "learning_rate": 3.352676628463659e-06, + "loss": 0.0139, + "step": 23050 + }, + { + "epoch": 2.499024284475282, + "grad_norm": 6.750314712524414, + "learning_rate": 3.3519512548962717e-06, + "loss": 0.0069, + "step": 23051 + }, + { + "epoch": 2.499132697311362, + "grad_norm": 1.5785155296325684, + "learning_rate": 3.3512258813288846e-06, + "loss": 0.0455, + "step": 23052 + }, + { + "epoch": 2.4992411101474414, + "grad_norm": 0.42342230677604675, + "learning_rate": 3.3505005077614976e-06, + "loss": 0.002, + "step": 23053 + }, + { + "epoch": 2.4993495229835214, + "grad_norm": 3.3356356620788574, + "learning_rate": 3.34977513419411e-06, + "loss": 0.0237, + "step": 23054 + }, + { + "epoch": 2.499457935819601, + "grad_norm": 4.077708721160889, + "learning_rate": 3.349049760626723e-06, + "loss": 0.0274, + "step": 23055 + }, + { + "epoch": 2.499566348655681, + "grad_norm": 0.6935098767280579, + "learning_rate": 3.3483243870593358e-06, + "loss": 0.0023, + "step": 23056 + }, + { + "epoch": 2.4996747614917605, + "grad_norm": 0.08043117821216583, + "learning_rate": 3.3475990134919483e-06, + "loss": 0.0005, + "step": 23057 + }, + { + "epoch": 2.4997831743278405, + "grad_norm": 0.0905199944972992, + "learning_rate": 3.3468736399245617e-06, + "loss": 0.0003, + "step": 23058 + }, + { + "epoch": 2.49989158716392, + "grad_norm": 2.9060187339782715, + "learning_rate": 3.3461482663571743e-06, + "loss": 0.0116, + "step": 23059 + }, + { + "epoch": 2.5, + "grad_norm": 1.0470942258834839, + "learning_rate": 3.345422892789787e-06, + "loss": 0.0013, + "step": 23060 + }, + { + "epoch": 2.50010841283608, + "grad_norm": 2.506145477294922, + "learning_rate": 3.3446975192223994e-06, + "loss": 0.0115, + "step": 23061 + }, + { + "epoch": 2.5002168256721595, + "grad_norm": 1.5411491394042969, + "learning_rate": 3.343972145655013e-06, + "loss": 0.0239, + "step": 23062 + }, + { + "epoch": 2.5003252385082395, + "grad_norm": 0.053076740354299545, + "learning_rate": 3.3432467720876254e-06, + "loss": 0.0003, + "step": 23063 + }, + { + "epoch": 2.500433651344319, + "grad_norm": 2.638312339782715, + "learning_rate": 3.342521398520238e-06, + "loss": 0.0295, + "step": 23064 + }, + { + "epoch": 2.500542064180399, + "grad_norm": 0.03782406449317932, + "learning_rate": 3.341796024952851e-06, + "loss": 0.0001, + "step": 23065 + }, + { + "epoch": 2.500650477016479, + "grad_norm": 2.0757267475128174, + "learning_rate": 3.341070651385464e-06, + "loss": 0.0073, + "step": 23066 + }, + { + "epoch": 2.5007588898525586, + "grad_norm": 0.26002824306488037, + "learning_rate": 3.3403452778180765e-06, + "loss": 0.0015, + "step": 23067 + }, + { + "epoch": 2.500867302688638, + "grad_norm": 5.367881774902344, + "learning_rate": 3.3396199042506895e-06, + "loss": 0.0336, + "step": 23068 + }, + { + "epoch": 2.500975715524718, + "grad_norm": 0.33987686038017273, + "learning_rate": 3.338894530683302e-06, + "loss": 0.0014, + "step": 23069 + }, + { + "epoch": 2.501084128360798, + "grad_norm": 6.316983699798584, + "learning_rate": 3.3381691571159147e-06, + "loss": 0.0086, + "step": 23070 + }, + { + "epoch": 2.5011925411968776, + "grad_norm": 0.10460709780454636, + "learning_rate": 3.337443783548528e-06, + "loss": 0.0004, + "step": 23071 + }, + { + "epoch": 2.5013009540329576, + "grad_norm": 0.5232975482940674, + "learning_rate": 3.3367184099811406e-06, + "loss": 0.0012, + "step": 23072 + }, + { + "epoch": 2.501409366869037, + "grad_norm": 0.4653708040714264, + "learning_rate": 3.335993036413753e-06, + "loss": 0.0016, + "step": 23073 + }, + { + "epoch": 2.501517779705117, + "grad_norm": 0.12586835026741028, + "learning_rate": 3.3352676628463658e-06, + "loss": 0.001, + "step": 23074 + }, + { + "epoch": 2.501626192541197, + "grad_norm": 0.033306241035461426, + "learning_rate": 3.334542289278979e-06, + "loss": 0.0003, + "step": 23075 + }, + { + "epoch": 2.5017346053772767, + "grad_norm": 0.1444377601146698, + "learning_rate": 3.3338169157115918e-06, + "loss": 0.0009, + "step": 23076 + }, + { + "epoch": 2.501843018213356, + "grad_norm": 2.1656010150909424, + "learning_rate": 3.3330915421442043e-06, + "loss": 0.0303, + "step": 23077 + }, + { + "epoch": 2.501951431049436, + "grad_norm": 3.677572727203369, + "learning_rate": 3.3323661685768173e-06, + "loss": 0.0049, + "step": 23078 + }, + { + "epoch": 2.502059843885516, + "grad_norm": 0.029248327016830444, + "learning_rate": 3.3316407950094303e-06, + "loss": 0.0004, + "step": 23079 + }, + { + "epoch": 2.5021682567215957, + "grad_norm": 0.7452442646026611, + "learning_rate": 3.330915421442043e-06, + "loss": 0.0008, + "step": 23080 + }, + { + "epoch": 2.5022766695576757, + "grad_norm": 0.04902763292193413, + "learning_rate": 3.330190047874656e-06, + "loss": 0.0004, + "step": 23081 + }, + { + "epoch": 2.5023850823937552, + "grad_norm": 21.575529098510742, + "learning_rate": 3.3294646743072684e-06, + "loss": 0.0073, + "step": 23082 + }, + { + "epoch": 2.502493495229835, + "grad_norm": 0.35017305612564087, + "learning_rate": 3.328739300739881e-06, + "loss": 0.0014, + "step": 23083 + }, + { + "epoch": 2.502601908065915, + "grad_norm": 3.3464455604553223, + "learning_rate": 3.3280139271724944e-06, + "loss": 0.0017, + "step": 23084 + }, + { + "epoch": 2.5027103209019947, + "grad_norm": 0.08145283907651901, + "learning_rate": 3.327288553605107e-06, + "loss": 0.0004, + "step": 23085 + }, + { + "epoch": 2.5028187337380747, + "grad_norm": 6.722464084625244, + "learning_rate": 3.3265631800377195e-06, + "loss": 0.0363, + "step": 23086 + }, + { + "epoch": 2.5029271465741543, + "grad_norm": 0.04095333442091942, + "learning_rate": 3.325837806470332e-06, + "loss": 0.0003, + "step": 23087 + }, + { + "epoch": 2.5030355594102343, + "grad_norm": 4.607501029968262, + "learning_rate": 3.3251124329029455e-06, + "loss": 0.0044, + "step": 23088 + }, + { + "epoch": 2.503143972246314, + "grad_norm": 4.198335647583008, + "learning_rate": 3.324387059335558e-06, + "loss": 0.0316, + "step": 23089 + }, + { + "epoch": 2.503252385082394, + "grad_norm": 6.795025825500488, + "learning_rate": 3.3236616857681707e-06, + "loss": 0.0382, + "step": 23090 + }, + { + "epoch": 2.5033607979184733, + "grad_norm": 0.028351299464702606, + "learning_rate": 3.3229363122007836e-06, + "loss": 0.0002, + "step": 23091 + }, + { + "epoch": 2.5034692107545533, + "grad_norm": 0.6774212121963501, + "learning_rate": 3.3222109386333966e-06, + "loss": 0.0035, + "step": 23092 + }, + { + "epoch": 2.5035776235906333, + "grad_norm": 0.6621464490890503, + "learning_rate": 3.321485565066009e-06, + "loss": 0.0017, + "step": 23093 + }, + { + "epoch": 2.503686036426713, + "grad_norm": 0.046035535633563995, + "learning_rate": 3.320760191498622e-06, + "loss": 0.0002, + "step": 23094 + }, + { + "epoch": 2.503794449262793, + "grad_norm": 0.12113010138273239, + "learning_rate": 3.3200348179312348e-06, + "loss": 0.001, + "step": 23095 + }, + { + "epoch": 2.5039028620988724, + "grad_norm": 0.10424575209617615, + "learning_rate": 3.3193094443638473e-06, + "loss": 0.0004, + "step": 23096 + }, + { + "epoch": 2.5040112749349523, + "grad_norm": 0.5036395788192749, + "learning_rate": 3.3185840707964607e-06, + "loss": 0.002, + "step": 23097 + }, + { + "epoch": 2.5041196877710323, + "grad_norm": 0.23220710456371307, + "learning_rate": 3.3178586972290733e-06, + "loss": 0.0005, + "step": 23098 + }, + { + "epoch": 2.504228100607112, + "grad_norm": 6.3824381828308105, + "learning_rate": 3.317133323661686e-06, + "loss": 0.0123, + "step": 23099 + }, + { + "epoch": 2.5043365134431914, + "grad_norm": 1.4693673849105835, + "learning_rate": 3.3164079500942984e-06, + "loss": 0.0044, + "step": 23100 + }, + { + "epoch": 2.5044449262792714, + "grad_norm": 4.043229103088379, + "learning_rate": 3.315682576526912e-06, + "loss": 0.0478, + "step": 23101 + }, + { + "epoch": 2.5045533391153514, + "grad_norm": 6.876430034637451, + "learning_rate": 3.3149572029595244e-06, + "loss": 0.0062, + "step": 23102 + }, + { + "epoch": 2.504661751951431, + "grad_norm": 0.2867426574230194, + "learning_rate": 3.314231829392137e-06, + "loss": 0.0009, + "step": 23103 + }, + { + "epoch": 2.504770164787511, + "grad_norm": 3.1702330112457275, + "learning_rate": 3.31350645582475e-06, + "loss": 0.0112, + "step": 23104 + }, + { + "epoch": 2.5048785776235905, + "grad_norm": 1.6146199703216553, + "learning_rate": 3.312781082257363e-06, + "loss": 0.0036, + "step": 23105 + }, + { + "epoch": 2.5049869904596704, + "grad_norm": 5.220978736877441, + "learning_rate": 3.3120557086899755e-06, + "loss": 0.0053, + "step": 23106 + }, + { + "epoch": 2.5050954032957504, + "grad_norm": 1.1239898204803467, + "learning_rate": 3.3113303351225885e-06, + "loss": 0.0021, + "step": 23107 + }, + { + "epoch": 2.50520381613183, + "grad_norm": 2.4617972373962402, + "learning_rate": 3.310604961555201e-06, + "loss": 0.004, + "step": 23108 + }, + { + "epoch": 2.50531222896791, + "grad_norm": 4.825112342834473, + "learning_rate": 3.3098795879878137e-06, + "loss": 0.0172, + "step": 23109 + }, + { + "epoch": 2.5054206418039895, + "grad_norm": 23.472469329833984, + "learning_rate": 3.309154214420427e-06, + "loss": 0.0502, + "step": 23110 + }, + { + "epoch": 2.5055290546400695, + "grad_norm": 0.08499959111213684, + "learning_rate": 3.3084288408530396e-06, + "loss": 0.0005, + "step": 23111 + }, + { + "epoch": 2.505637467476149, + "grad_norm": 0.04074171558022499, + "learning_rate": 3.307703467285652e-06, + "loss": 0.0003, + "step": 23112 + }, + { + "epoch": 2.505745880312229, + "grad_norm": 0.19896431267261505, + "learning_rate": 3.3069780937182648e-06, + "loss": 0.0008, + "step": 23113 + }, + { + "epoch": 2.5058542931483085, + "grad_norm": 9.651290893554688, + "learning_rate": 3.306252720150878e-06, + "loss": 0.0274, + "step": 23114 + }, + { + "epoch": 2.5059627059843885, + "grad_norm": 1.8264073133468628, + "learning_rate": 3.3055273465834908e-06, + "loss": 0.0048, + "step": 23115 + }, + { + "epoch": 2.5060711188204685, + "grad_norm": 0.026261821389198303, + "learning_rate": 3.3048019730161033e-06, + "loss": 0.0003, + "step": 23116 + }, + { + "epoch": 2.506179531656548, + "grad_norm": 0.026311099529266357, + "learning_rate": 3.3040765994487163e-06, + "loss": 0.0001, + "step": 23117 + }, + { + "epoch": 2.506287944492628, + "grad_norm": 1.9243885278701782, + "learning_rate": 3.3033512258813293e-06, + "loss": 0.0025, + "step": 23118 + }, + { + "epoch": 2.5063963573287076, + "grad_norm": 0.04399709403514862, + "learning_rate": 3.302625852313942e-06, + "loss": 0.0004, + "step": 23119 + }, + { + "epoch": 2.5065047701647876, + "grad_norm": 0.09400345385074615, + "learning_rate": 3.301900478746555e-06, + "loss": 0.0006, + "step": 23120 + }, + { + "epoch": 2.5066131830008676, + "grad_norm": 0.02858584187924862, + "learning_rate": 3.3011751051791674e-06, + "loss": 0.0003, + "step": 23121 + }, + { + "epoch": 2.506721595836947, + "grad_norm": 0.03055950626730919, + "learning_rate": 3.30044973161178e-06, + "loss": 0.0001, + "step": 23122 + }, + { + "epoch": 2.5068300086730266, + "grad_norm": 6.692137718200684, + "learning_rate": 3.2997243580443934e-06, + "loss": 0.0084, + "step": 23123 + }, + { + "epoch": 2.5069384215091066, + "grad_norm": 2.03277850151062, + "learning_rate": 3.298998984477006e-06, + "loss": 0.0092, + "step": 23124 + }, + { + "epoch": 2.5070468343451866, + "grad_norm": 2.992905855178833, + "learning_rate": 3.2982736109096185e-06, + "loss": 0.0133, + "step": 23125 + }, + { + "epoch": 2.507155247181266, + "grad_norm": 5.2147932052612305, + "learning_rate": 3.297548237342231e-06, + "loss": 0.0166, + "step": 23126 + }, + { + "epoch": 2.507263660017346, + "grad_norm": 0.02266807109117508, + "learning_rate": 3.2968228637748445e-06, + "loss": 0.0002, + "step": 23127 + }, + { + "epoch": 2.5073720728534257, + "grad_norm": 9.457221984863281, + "learning_rate": 3.296097490207457e-06, + "loss": 0.0366, + "step": 23128 + }, + { + "epoch": 2.5074804856895057, + "grad_norm": 1.196782112121582, + "learning_rate": 3.2953721166400697e-06, + "loss": 0.0028, + "step": 23129 + }, + { + "epoch": 2.5075888985255856, + "grad_norm": 7.475391864776611, + "learning_rate": 3.2946467430726826e-06, + "loss": 0.0375, + "step": 23130 + }, + { + "epoch": 2.507697311361665, + "grad_norm": 0.6946357488632202, + "learning_rate": 3.2939213695052956e-06, + "loss": 0.0021, + "step": 23131 + }, + { + "epoch": 2.507805724197745, + "grad_norm": 7.152445316314697, + "learning_rate": 3.293195995937908e-06, + "loss": 0.0195, + "step": 23132 + }, + { + "epoch": 2.5079141370338247, + "grad_norm": 0.1789320558309555, + "learning_rate": 3.292470622370521e-06, + "loss": 0.001, + "step": 23133 + }, + { + "epoch": 2.5080225498699047, + "grad_norm": 2.5606603622436523, + "learning_rate": 3.2917452488031338e-06, + "loss": 0.0175, + "step": 23134 + }, + { + "epoch": 2.5081309627059842, + "grad_norm": 2.309295654296875, + "learning_rate": 3.2910198752357468e-06, + "loss": 0.003, + "step": 23135 + }, + { + "epoch": 2.508239375542064, + "grad_norm": 6.265074729919434, + "learning_rate": 3.2902945016683597e-06, + "loss": 0.0179, + "step": 23136 + }, + { + "epoch": 2.5083477883781438, + "grad_norm": 0.6895466446876526, + "learning_rate": 3.2895691281009723e-06, + "loss": 0.0009, + "step": 23137 + }, + { + "epoch": 2.5084562012142237, + "grad_norm": 5.786100387573242, + "learning_rate": 3.288843754533585e-06, + "loss": 0.0154, + "step": 23138 + }, + { + "epoch": 2.5085646140503037, + "grad_norm": 1.236444354057312, + "learning_rate": 3.2881183809661974e-06, + "loss": 0.0064, + "step": 23139 + }, + { + "epoch": 2.5086730268863833, + "grad_norm": 0.04924030974507332, + "learning_rate": 3.287393007398811e-06, + "loss": 0.0003, + "step": 23140 + }, + { + "epoch": 2.5087814397224633, + "grad_norm": 2.796950101852417, + "learning_rate": 3.2866676338314234e-06, + "loss": 0.0185, + "step": 23141 + }, + { + "epoch": 2.508889852558543, + "grad_norm": 2.484281539916992, + "learning_rate": 3.285942260264036e-06, + "loss": 0.0421, + "step": 23142 + }, + { + "epoch": 2.508998265394623, + "grad_norm": 0.06839334219694138, + "learning_rate": 3.285216886696649e-06, + "loss": 0.0003, + "step": 23143 + }, + { + "epoch": 2.5091066782307028, + "grad_norm": 0.4745764136314392, + "learning_rate": 3.284491513129262e-06, + "loss": 0.0016, + "step": 23144 + }, + { + "epoch": 2.5092150910667823, + "grad_norm": 2.2666702270507812, + "learning_rate": 3.2837661395618745e-06, + "loss": 0.0243, + "step": 23145 + }, + { + "epoch": 2.509323503902862, + "grad_norm": 2.2489888668060303, + "learning_rate": 3.2830407659944875e-06, + "loss": 0.0091, + "step": 23146 + }, + { + "epoch": 2.509431916738942, + "grad_norm": 0.33710935711860657, + "learning_rate": 3.2823153924271e-06, + "loss": 0.0009, + "step": 23147 + }, + { + "epoch": 2.509540329575022, + "grad_norm": 11.229442596435547, + "learning_rate": 3.281590018859713e-06, + "loss": 0.0067, + "step": 23148 + }, + { + "epoch": 2.5096487424111014, + "grad_norm": 1.1938755512237549, + "learning_rate": 3.280864645292326e-06, + "loss": 0.0043, + "step": 23149 + }, + { + "epoch": 2.5097571552471813, + "grad_norm": 0.8591384291648865, + "learning_rate": 3.2801392717249386e-06, + "loss": 0.0012, + "step": 23150 + }, + { + "epoch": 2.509865568083261, + "grad_norm": 2.6978471279144287, + "learning_rate": 3.2794138981575512e-06, + "loss": 0.024, + "step": 23151 + }, + { + "epoch": 2.509973980919341, + "grad_norm": 6.148781776428223, + "learning_rate": 3.2786885245901638e-06, + "loss": 0.0028, + "step": 23152 + }, + { + "epoch": 2.510082393755421, + "grad_norm": 0.026058441027998924, + "learning_rate": 3.277963151022777e-06, + "loss": 0.0001, + "step": 23153 + }, + { + "epoch": 2.5101908065915004, + "grad_norm": 1.001194953918457, + "learning_rate": 3.2772377774553898e-06, + "loss": 0.0336, + "step": 23154 + }, + { + "epoch": 2.5102992194275804, + "grad_norm": 0.9282289147377014, + "learning_rate": 3.2765124038880023e-06, + "loss": 0.0007, + "step": 23155 + }, + { + "epoch": 2.51040763226366, + "grad_norm": 0.12489286810159683, + "learning_rate": 3.2757870303206153e-06, + "loss": 0.0006, + "step": 23156 + }, + { + "epoch": 2.51051604509974, + "grad_norm": 1.4245245456695557, + "learning_rate": 3.2750616567532283e-06, + "loss": 0.0006, + "step": 23157 + }, + { + "epoch": 2.5106244579358195, + "grad_norm": 0.06409212201833725, + "learning_rate": 3.274336283185841e-06, + "loss": 0.0003, + "step": 23158 + }, + { + "epoch": 2.5107328707718994, + "grad_norm": 2.8858189582824707, + "learning_rate": 3.273610909618454e-06, + "loss": 0.0216, + "step": 23159 + }, + { + "epoch": 2.510841283607979, + "grad_norm": 2.8883211612701416, + "learning_rate": 3.2728855360510664e-06, + "loss": 0.0086, + "step": 23160 + }, + { + "epoch": 2.510949696444059, + "grad_norm": 0.09193634986877441, + "learning_rate": 3.2721601624836794e-06, + "loss": 0.0006, + "step": 23161 + }, + { + "epoch": 2.511058109280139, + "grad_norm": 2.7986104488372803, + "learning_rate": 3.2714347889162924e-06, + "loss": 0.0062, + "step": 23162 + }, + { + "epoch": 2.5111665221162185, + "grad_norm": 2.3570895195007324, + "learning_rate": 3.270709415348905e-06, + "loss": 0.0074, + "step": 23163 + }, + { + "epoch": 2.5112749349522985, + "grad_norm": 0.1168927326798439, + "learning_rate": 3.2699840417815175e-06, + "loss": 0.0007, + "step": 23164 + }, + { + "epoch": 2.511383347788378, + "grad_norm": 0.08105102926492691, + "learning_rate": 3.26925866821413e-06, + "loss": 0.0003, + "step": 23165 + }, + { + "epoch": 2.511491760624458, + "grad_norm": 0.01511822547763586, + "learning_rate": 3.2685332946467435e-06, + "loss": 0.0002, + "step": 23166 + }, + { + "epoch": 2.511600173460538, + "grad_norm": 0.16764052212238312, + "learning_rate": 3.267807921079356e-06, + "loss": 0.0011, + "step": 23167 + }, + { + "epoch": 2.5117085862966175, + "grad_norm": 2.2811765670776367, + "learning_rate": 3.2670825475119687e-06, + "loss": 0.024, + "step": 23168 + }, + { + "epoch": 2.511816999132697, + "grad_norm": 0.3577325642108917, + "learning_rate": 3.2663571739445817e-06, + "loss": 0.0007, + "step": 23169 + }, + { + "epoch": 2.511925411968777, + "grad_norm": 0.07568301260471344, + "learning_rate": 3.2656318003771946e-06, + "loss": 0.0005, + "step": 23170 + }, + { + "epoch": 2.512033824804857, + "grad_norm": 0.14495956897735596, + "learning_rate": 3.264906426809807e-06, + "loss": 0.0006, + "step": 23171 + }, + { + "epoch": 2.5121422376409366, + "grad_norm": 2.428513526916504, + "learning_rate": 3.26418105324242e-06, + "loss": 0.0123, + "step": 23172 + }, + { + "epoch": 2.5122506504770166, + "grad_norm": 6.019289493560791, + "learning_rate": 3.2634556796750328e-06, + "loss": 0.0244, + "step": 23173 + }, + { + "epoch": 2.512359063313096, + "grad_norm": 2.914562702178955, + "learning_rate": 3.2627303061076458e-06, + "loss": 0.0033, + "step": 23174 + }, + { + "epoch": 2.512467476149176, + "grad_norm": 1.7217406034469604, + "learning_rate": 3.2620049325402587e-06, + "loss": 0.0016, + "step": 23175 + }, + { + "epoch": 2.512575888985256, + "grad_norm": 0.2664574086666107, + "learning_rate": 3.2612795589728713e-06, + "loss": 0.0007, + "step": 23176 + }, + { + "epoch": 2.5126843018213356, + "grad_norm": 0.20785044133663177, + "learning_rate": 3.260554185405484e-06, + "loss": 0.0006, + "step": 23177 + }, + { + "epoch": 2.512792714657415, + "grad_norm": 0.41287827491760254, + "learning_rate": 3.2598288118380964e-06, + "loss": 0.0008, + "step": 23178 + }, + { + "epoch": 2.512901127493495, + "grad_norm": 0.2787158191204071, + "learning_rate": 3.25910343827071e-06, + "loss": 0.0007, + "step": 23179 + }, + { + "epoch": 2.513009540329575, + "grad_norm": 2.0240375995635986, + "learning_rate": 3.2583780647033224e-06, + "loss": 0.0067, + "step": 23180 + }, + { + "epoch": 2.5131179531656547, + "grad_norm": 6.579709053039551, + "learning_rate": 3.257652691135935e-06, + "loss": 0.0072, + "step": 23181 + }, + { + "epoch": 2.5132263660017347, + "grad_norm": 0.7276666164398193, + "learning_rate": 3.256927317568548e-06, + "loss": 0.0028, + "step": 23182 + }, + { + "epoch": 2.513334778837814, + "grad_norm": 2.9195656776428223, + "learning_rate": 3.256201944001161e-06, + "loss": 0.0264, + "step": 23183 + }, + { + "epoch": 2.513443191673894, + "grad_norm": 0.4013950228691101, + "learning_rate": 3.2554765704337735e-06, + "loss": 0.0023, + "step": 23184 + }, + { + "epoch": 2.513551604509974, + "grad_norm": 0.951185405254364, + "learning_rate": 3.2547511968663865e-06, + "loss": 0.0015, + "step": 23185 + }, + { + "epoch": 2.5136600173460537, + "grad_norm": 0.9712311625480652, + "learning_rate": 3.254025823298999e-06, + "loss": 0.0038, + "step": 23186 + }, + { + "epoch": 2.5137684301821337, + "grad_norm": 1.0103825330734253, + "learning_rate": 3.253300449731612e-06, + "loss": 0.0034, + "step": 23187 + }, + { + "epoch": 2.5138768430182132, + "grad_norm": 0.2695821225643158, + "learning_rate": 3.252575076164225e-06, + "loss": 0.0006, + "step": 23188 + }, + { + "epoch": 2.513985255854293, + "grad_norm": 1.3675062656402588, + "learning_rate": 3.2518497025968376e-06, + "loss": 0.0029, + "step": 23189 + }, + { + "epoch": 2.514093668690373, + "grad_norm": 7.613471508026123, + "learning_rate": 3.2511243290294502e-06, + "loss": 0.0158, + "step": 23190 + }, + { + "epoch": 2.5142020815264527, + "grad_norm": 0.04212918505072594, + "learning_rate": 3.2503989554620628e-06, + "loss": 0.0001, + "step": 23191 + }, + { + "epoch": 2.5143104943625323, + "grad_norm": 0.252713143825531, + "learning_rate": 3.249673581894676e-06, + "loss": 0.0017, + "step": 23192 + }, + { + "epoch": 2.5144189071986123, + "grad_norm": 4.487370014190674, + "learning_rate": 3.2489482083272888e-06, + "loss": 0.0126, + "step": 23193 + }, + { + "epoch": 2.5145273200346923, + "grad_norm": 0.15066660940647125, + "learning_rate": 3.2482228347599013e-06, + "loss": 0.001, + "step": 23194 + }, + { + "epoch": 2.514635732870772, + "grad_norm": 0.05116598308086395, + "learning_rate": 3.2474974611925143e-06, + "loss": 0.0001, + "step": 23195 + }, + { + "epoch": 2.514744145706852, + "grad_norm": 3.4629428386688232, + "learning_rate": 3.2467720876251273e-06, + "loss": 0.0144, + "step": 23196 + }, + { + "epoch": 2.5148525585429313, + "grad_norm": 1.1148183345794678, + "learning_rate": 3.24604671405774e-06, + "loss": 0.0043, + "step": 23197 + }, + { + "epoch": 2.5149609713790113, + "grad_norm": 11.816411018371582, + "learning_rate": 3.245321340490353e-06, + "loss": 0.0115, + "step": 23198 + }, + { + "epoch": 2.5150693842150913, + "grad_norm": 1.8296071290969849, + "learning_rate": 3.2445959669229654e-06, + "loss": 0.0045, + "step": 23199 + }, + { + "epoch": 2.515177797051171, + "grad_norm": 0.07795760035514832, + "learning_rate": 3.2438705933555784e-06, + "loss": 0.0002, + "step": 23200 + }, + { + "epoch": 2.5152862098872504, + "grad_norm": 3.6816673278808594, + "learning_rate": 3.2431452197881914e-06, + "loss": 0.0084, + "step": 23201 + }, + { + "epoch": 2.5153946227233304, + "grad_norm": 0.035351600497961044, + "learning_rate": 3.242419846220804e-06, + "loss": 0.0002, + "step": 23202 + }, + { + "epoch": 2.5155030355594103, + "grad_norm": 0.7359643578529358, + "learning_rate": 3.2416944726534165e-06, + "loss": 0.0026, + "step": 23203 + }, + { + "epoch": 2.51561144839549, + "grad_norm": 3.433589220046997, + "learning_rate": 3.240969099086029e-06, + "loss": 0.0231, + "step": 23204 + }, + { + "epoch": 2.51571986123157, + "grad_norm": 1.3070979118347168, + "learning_rate": 3.2402437255186425e-06, + "loss": 0.0024, + "step": 23205 + }, + { + "epoch": 2.5158282740676494, + "grad_norm": 0.5481331944465637, + "learning_rate": 3.239518351951255e-06, + "loss": 0.0009, + "step": 23206 + }, + { + "epoch": 2.5159366869037294, + "grad_norm": 5.082079887390137, + "learning_rate": 3.2387929783838677e-06, + "loss": 0.0189, + "step": 23207 + }, + { + "epoch": 2.5160450997398094, + "grad_norm": 0.5088738799095154, + "learning_rate": 3.2380676048164807e-06, + "loss": 0.0012, + "step": 23208 + }, + { + "epoch": 2.516153512575889, + "grad_norm": 4.583163261413574, + "learning_rate": 3.2373422312490936e-06, + "loss": 0.0255, + "step": 23209 + }, + { + "epoch": 2.516261925411969, + "grad_norm": 2.255098581314087, + "learning_rate": 3.2366168576817062e-06, + "loss": 0.0152, + "step": 23210 + }, + { + "epoch": 2.5163703382480485, + "grad_norm": 0.19356219470500946, + "learning_rate": 3.235891484114319e-06, + "loss": 0.0005, + "step": 23211 + }, + { + "epoch": 2.5164787510841284, + "grad_norm": 1.3219215869903564, + "learning_rate": 3.2351661105469318e-06, + "loss": 0.0075, + "step": 23212 + }, + { + "epoch": 2.5165871639202084, + "grad_norm": 2.0696465969085693, + "learning_rate": 3.2344407369795448e-06, + "loss": 0.0411, + "step": 23213 + }, + { + "epoch": 2.516695576756288, + "grad_norm": 0.17596769332885742, + "learning_rate": 3.2337153634121578e-06, + "loss": 0.0006, + "step": 23214 + }, + { + "epoch": 2.5168039895923675, + "grad_norm": 2.0126867294311523, + "learning_rate": 3.2329899898447703e-06, + "loss": 0.0129, + "step": 23215 + }, + { + "epoch": 2.5169124024284475, + "grad_norm": 1.314692497253418, + "learning_rate": 3.232264616277383e-06, + "loss": 0.0216, + "step": 23216 + }, + { + "epoch": 2.5170208152645275, + "grad_norm": 5.432213306427002, + "learning_rate": 3.2315392427099955e-06, + "loss": 0.02, + "step": 23217 + }, + { + "epoch": 2.517129228100607, + "grad_norm": 4.017956733703613, + "learning_rate": 3.230813869142609e-06, + "loss": 0.0072, + "step": 23218 + }, + { + "epoch": 2.517237640936687, + "grad_norm": 0.04861486330628395, + "learning_rate": 3.2300884955752214e-06, + "loss": 0.0003, + "step": 23219 + }, + { + "epoch": 2.5173460537727665, + "grad_norm": 0.06905590742826462, + "learning_rate": 3.229363122007834e-06, + "loss": 0.0005, + "step": 23220 + }, + { + "epoch": 2.5174544666088465, + "grad_norm": 0.01651264727115631, + "learning_rate": 3.228637748440447e-06, + "loss": 0.0001, + "step": 23221 + }, + { + "epoch": 2.5175628794449265, + "grad_norm": 1.495392084121704, + "learning_rate": 3.22791237487306e-06, + "loss": 0.011, + "step": 23222 + }, + { + "epoch": 2.517671292281006, + "grad_norm": 0.06252479553222656, + "learning_rate": 3.2271870013056725e-06, + "loss": 0.0003, + "step": 23223 + }, + { + "epoch": 2.5177797051170856, + "grad_norm": 7.0618367195129395, + "learning_rate": 3.2264616277382855e-06, + "loss": 0.0136, + "step": 23224 + }, + { + "epoch": 2.5178881179531656, + "grad_norm": 1.7862728834152222, + "learning_rate": 3.225736254170898e-06, + "loss": 0.0532, + "step": 23225 + }, + { + "epoch": 2.5179965307892456, + "grad_norm": 0.13394030928611755, + "learning_rate": 3.2250108806035115e-06, + "loss": 0.0004, + "step": 23226 + }, + { + "epoch": 2.518104943625325, + "grad_norm": 0.04536798596382141, + "learning_rate": 3.224285507036124e-06, + "loss": 0.0006, + "step": 23227 + }, + { + "epoch": 2.518213356461405, + "grad_norm": 21.029422760009766, + "learning_rate": 3.2235601334687367e-06, + "loss": 0.0196, + "step": 23228 + }, + { + "epoch": 2.5183217692974846, + "grad_norm": 0.09302005171775818, + "learning_rate": 3.2228347599013492e-06, + "loss": 0.0005, + "step": 23229 + }, + { + "epoch": 2.5184301821335646, + "grad_norm": 2.172055244445801, + "learning_rate": 3.2221093863339626e-06, + "loss": 0.0084, + "step": 23230 + }, + { + "epoch": 2.5185385949696446, + "grad_norm": 9.366679191589355, + "learning_rate": 3.221384012766575e-06, + "loss": 0.0097, + "step": 23231 + }, + { + "epoch": 2.518647007805724, + "grad_norm": 1.4975568056106567, + "learning_rate": 3.2206586391991878e-06, + "loss": 0.0037, + "step": 23232 + }, + { + "epoch": 2.518755420641804, + "grad_norm": 1.5568666458129883, + "learning_rate": 3.2199332656318003e-06, + "loss": 0.0147, + "step": 23233 + }, + { + "epoch": 2.5188638334778837, + "grad_norm": 0.23326686024665833, + "learning_rate": 3.2192078920644133e-06, + "loss": 0.0004, + "step": 23234 + }, + { + "epoch": 2.5189722463139637, + "grad_norm": 0.009077114053070545, + "learning_rate": 3.2184825184970263e-06, + "loss": 0.0001, + "step": 23235 + }, + { + "epoch": 2.519080659150043, + "grad_norm": 4.051299095153809, + "learning_rate": 3.217757144929639e-06, + "loss": 0.0035, + "step": 23236 + }, + { + "epoch": 2.519189071986123, + "grad_norm": 0.030237596482038498, + "learning_rate": 3.217031771362252e-06, + "loss": 0.0004, + "step": 23237 + }, + { + "epoch": 2.5192974848222027, + "grad_norm": 5.727712631225586, + "learning_rate": 3.2163063977948644e-06, + "loss": 0.033, + "step": 23238 + }, + { + "epoch": 2.5194058976582827, + "grad_norm": 10.481780052185059, + "learning_rate": 3.215581024227478e-06, + "loss": 0.0184, + "step": 23239 + }, + { + "epoch": 2.5195143104943627, + "grad_norm": 18.235795974731445, + "learning_rate": 3.2148556506600904e-06, + "loss": 0.0163, + "step": 23240 + }, + { + "epoch": 2.5196227233304422, + "grad_norm": 2.9504482746124268, + "learning_rate": 3.214130277092703e-06, + "loss": 0.0129, + "step": 23241 + }, + { + "epoch": 2.5197311361665222, + "grad_norm": 1.7244579792022705, + "learning_rate": 3.2134049035253156e-06, + "loss": 0.0034, + "step": 23242 + }, + { + "epoch": 2.5198395490026018, + "grad_norm": 0.015171739272773266, + "learning_rate": 3.212679529957929e-06, + "loss": 0.0001, + "step": 23243 + }, + { + "epoch": 2.5199479618386817, + "grad_norm": 0.04379216581583023, + "learning_rate": 3.2119541563905415e-06, + "loss": 0.0002, + "step": 23244 + }, + { + "epoch": 2.5200563746747617, + "grad_norm": 0.010111062787473202, + "learning_rate": 3.211228782823154e-06, + "loss": 0.0001, + "step": 23245 + }, + { + "epoch": 2.5201647875108413, + "grad_norm": 0.07886205613613129, + "learning_rate": 3.2105034092557667e-06, + "loss": 0.0004, + "step": 23246 + }, + { + "epoch": 2.520273200346921, + "grad_norm": 21.164976119995117, + "learning_rate": 3.2097780356883797e-06, + "loss": 0.0469, + "step": 23247 + }, + { + "epoch": 2.520381613183001, + "grad_norm": 1.2636078596115112, + "learning_rate": 3.2090526621209926e-06, + "loss": 0.0027, + "step": 23248 + }, + { + "epoch": 2.520490026019081, + "grad_norm": 11.685040473937988, + "learning_rate": 3.2083272885536052e-06, + "loss": 0.0411, + "step": 23249 + }, + { + "epoch": 2.5205984388551603, + "grad_norm": 3.943932294845581, + "learning_rate": 3.207601914986218e-06, + "loss": 0.0558, + "step": 23250 + }, + { + "epoch": 2.5207068516912403, + "grad_norm": 0.04893547669053078, + "learning_rate": 3.2068765414188308e-06, + "loss": 0.0003, + "step": 23251 + }, + { + "epoch": 2.52081526452732, + "grad_norm": 0.7750381827354431, + "learning_rate": 3.206151167851444e-06, + "loss": 0.0035, + "step": 23252 + }, + { + "epoch": 2.5209236773634, + "grad_norm": 0.3857215344905853, + "learning_rate": 3.2054257942840568e-06, + "loss": 0.001, + "step": 23253 + }, + { + "epoch": 2.52103209019948, + "grad_norm": 0.0213617030531168, + "learning_rate": 3.2047004207166693e-06, + "loss": 0.0001, + "step": 23254 + }, + { + "epoch": 2.5211405030355594, + "grad_norm": 1.7165919542312622, + "learning_rate": 3.203975047149282e-06, + "loss": 0.0125, + "step": 23255 + }, + { + "epoch": 2.5212489158716394, + "grad_norm": 3.513678789138794, + "learning_rate": 3.2032496735818953e-06, + "loss": 0.02, + "step": 23256 + }, + { + "epoch": 2.521357328707719, + "grad_norm": 4.850953102111816, + "learning_rate": 3.202524300014508e-06, + "loss": 0.0035, + "step": 23257 + }, + { + "epoch": 2.521465741543799, + "grad_norm": 0.052215661853551865, + "learning_rate": 3.2017989264471204e-06, + "loss": 0.0005, + "step": 23258 + }, + { + "epoch": 2.5215741543798784, + "grad_norm": 0.42541396617889404, + "learning_rate": 3.201073552879733e-06, + "loss": 0.0013, + "step": 23259 + }, + { + "epoch": 2.5216825672159584, + "grad_norm": 0.4449712634086609, + "learning_rate": 3.200348179312346e-06, + "loss": 0.0011, + "step": 23260 + }, + { + "epoch": 2.521790980052038, + "grad_norm": 5.84028434753418, + "learning_rate": 3.199622805744959e-06, + "loss": 0.0335, + "step": 23261 + }, + { + "epoch": 2.521899392888118, + "grad_norm": 0.49105900526046753, + "learning_rate": 3.1988974321775715e-06, + "loss": 0.002, + "step": 23262 + }, + { + "epoch": 2.522007805724198, + "grad_norm": 1.677687168121338, + "learning_rate": 3.1981720586101845e-06, + "loss": 0.0121, + "step": 23263 + }, + { + "epoch": 2.5221162185602775, + "grad_norm": 0.06916840374469757, + "learning_rate": 3.197446685042797e-06, + "loss": 0.0006, + "step": 23264 + }, + { + "epoch": 2.5222246313963574, + "grad_norm": 2.9817302227020264, + "learning_rate": 3.1967213114754105e-06, + "loss": 0.0022, + "step": 23265 + }, + { + "epoch": 2.522333044232437, + "grad_norm": 3.5362987518310547, + "learning_rate": 3.195995937908023e-06, + "loss": 0.0563, + "step": 23266 + }, + { + "epoch": 2.522441457068517, + "grad_norm": 0.23020416498184204, + "learning_rate": 3.1952705643406357e-06, + "loss": 0.0006, + "step": 23267 + }, + { + "epoch": 2.522549869904597, + "grad_norm": 1.8380601406097412, + "learning_rate": 3.1945451907732482e-06, + "loss": 0.0062, + "step": 23268 + }, + { + "epoch": 2.5226582827406765, + "grad_norm": 0.1878986805677414, + "learning_rate": 3.1938198172058616e-06, + "loss": 0.0007, + "step": 23269 + }, + { + "epoch": 2.522766695576756, + "grad_norm": 0.1431504487991333, + "learning_rate": 3.193094443638474e-06, + "loss": 0.0004, + "step": 23270 + }, + { + "epoch": 2.522875108412836, + "grad_norm": 0.40370967984199524, + "learning_rate": 3.1923690700710868e-06, + "loss": 0.0019, + "step": 23271 + }, + { + "epoch": 2.522983521248916, + "grad_norm": 0.0201315525919199, + "learning_rate": 3.1916436965036993e-06, + "loss": 0.0002, + "step": 23272 + }, + { + "epoch": 2.5230919340849955, + "grad_norm": 0.019989870488643646, + "learning_rate": 3.1909183229363123e-06, + "loss": 0.0002, + "step": 23273 + }, + { + "epoch": 2.5232003469210755, + "grad_norm": 13.21264362335205, + "learning_rate": 3.1901929493689253e-06, + "loss": 0.055, + "step": 23274 + }, + { + "epoch": 2.523308759757155, + "grad_norm": 0.12521201372146606, + "learning_rate": 3.1894675758015383e-06, + "loss": 0.0011, + "step": 23275 + }, + { + "epoch": 2.523417172593235, + "grad_norm": 4.318758964538574, + "learning_rate": 3.188742202234151e-06, + "loss": 0.0082, + "step": 23276 + }, + { + "epoch": 2.523525585429315, + "grad_norm": 0.020910009741783142, + "learning_rate": 3.1880168286667634e-06, + "loss": 0.0002, + "step": 23277 + }, + { + "epoch": 2.5236339982653946, + "grad_norm": 2.7127315998077393, + "learning_rate": 3.187291455099377e-06, + "loss": 0.0133, + "step": 23278 + }, + { + "epoch": 2.5237424111014746, + "grad_norm": 0.8153867125511169, + "learning_rate": 3.1865660815319894e-06, + "loss": 0.0026, + "step": 23279 + }, + { + "epoch": 2.523850823937554, + "grad_norm": 5.157268047332764, + "learning_rate": 3.185840707964602e-06, + "loss": 0.0122, + "step": 23280 + }, + { + "epoch": 2.523959236773634, + "grad_norm": 0.12503397464752197, + "learning_rate": 3.1851153343972146e-06, + "loss": 0.0011, + "step": 23281 + }, + { + "epoch": 2.5240676496097136, + "grad_norm": 1.4336438179016113, + "learning_rate": 3.184389960829828e-06, + "loss": 0.0036, + "step": 23282 + }, + { + "epoch": 2.5241760624457936, + "grad_norm": 0.09692084789276123, + "learning_rate": 3.1836645872624405e-06, + "loss": 0.0007, + "step": 23283 + }, + { + "epoch": 2.524284475281873, + "grad_norm": 0.01813042350113392, + "learning_rate": 3.182939213695053e-06, + "loss": 0.0001, + "step": 23284 + }, + { + "epoch": 2.524392888117953, + "grad_norm": 0.06775465607643127, + "learning_rate": 3.1822138401276657e-06, + "loss": 0.0004, + "step": 23285 + }, + { + "epoch": 2.524501300954033, + "grad_norm": 2.0126824378967285, + "learning_rate": 3.1814884665602787e-06, + "loss": 0.0198, + "step": 23286 + }, + { + "epoch": 2.5246097137901127, + "grad_norm": 9.20689868927002, + "learning_rate": 3.1807630929928917e-06, + "loss": 0.006, + "step": 23287 + }, + { + "epoch": 2.5247181266261927, + "grad_norm": 0.06003108248114586, + "learning_rate": 3.1800377194255046e-06, + "loss": 0.0005, + "step": 23288 + }, + { + "epoch": 2.524826539462272, + "grad_norm": 0.8467705249786377, + "learning_rate": 3.179312345858117e-06, + "loss": 0.0022, + "step": 23289 + }, + { + "epoch": 2.524934952298352, + "grad_norm": 2.8971662521362305, + "learning_rate": 3.1785869722907298e-06, + "loss": 0.0299, + "step": 23290 + }, + { + "epoch": 2.525043365134432, + "grad_norm": 1.8992536067962646, + "learning_rate": 3.177861598723343e-06, + "loss": 0.0102, + "step": 23291 + }, + { + "epoch": 2.5251517779705117, + "grad_norm": 10.133697509765625, + "learning_rate": 3.1771362251559558e-06, + "loss": 0.0312, + "step": 23292 + }, + { + "epoch": 2.5252601908065913, + "grad_norm": 0.03521937131881714, + "learning_rate": 3.1764108515885683e-06, + "loss": 0.0005, + "step": 23293 + }, + { + "epoch": 2.5253686036426712, + "grad_norm": 0.07726652175188065, + "learning_rate": 3.175685478021181e-06, + "loss": 0.0001, + "step": 23294 + }, + { + "epoch": 2.5254770164787512, + "grad_norm": 1.3194817304611206, + "learning_rate": 3.1749601044537943e-06, + "loss": 0.0287, + "step": 23295 + }, + { + "epoch": 2.5255854293148308, + "grad_norm": 1.434572696685791, + "learning_rate": 3.174234730886407e-06, + "loss": 0.0041, + "step": 23296 + }, + { + "epoch": 2.5256938421509108, + "grad_norm": 4.189521312713623, + "learning_rate": 3.1735093573190194e-06, + "loss": 0.0086, + "step": 23297 + }, + { + "epoch": 2.5258022549869903, + "grad_norm": 1.6493209600448608, + "learning_rate": 3.172783983751632e-06, + "loss": 0.0051, + "step": 23298 + }, + { + "epoch": 2.5259106678230703, + "grad_norm": 2.6427013874053955, + "learning_rate": 3.172058610184245e-06, + "loss": 0.069, + "step": 23299 + }, + { + "epoch": 2.5260190806591503, + "grad_norm": 1.217434287071228, + "learning_rate": 3.171333236616858e-06, + "loss": 0.0041, + "step": 23300 + }, + { + "epoch": 2.52612749349523, + "grad_norm": 4.262243270874023, + "learning_rate": 3.170607863049471e-06, + "loss": 0.0347, + "step": 23301 + }, + { + "epoch": 2.52623590633131, + "grad_norm": 0.04017924889922142, + "learning_rate": 3.1698824894820835e-06, + "loss": 0.0002, + "step": 23302 + }, + { + "epoch": 2.5263443191673893, + "grad_norm": 0.08313768357038498, + "learning_rate": 3.169157115914696e-06, + "loss": 0.0004, + "step": 23303 + }, + { + "epoch": 2.5264527320034693, + "grad_norm": 1.0892975330352783, + "learning_rate": 3.1684317423473095e-06, + "loss": 0.0018, + "step": 23304 + }, + { + "epoch": 2.526561144839549, + "grad_norm": 3.594688653945923, + "learning_rate": 3.167706368779922e-06, + "loss": 0.0234, + "step": 23305 + }, + { + "epoch": 2.526669557675629, + "grad_norm": 0.15315140783786774, + "learning_rate": 3.1669809952125347e-06, + "loss": 0.0012, + "step": 23306 + }, + { + "epoch": 2.5267779705117084, + "grad_norm": 13.547383308410645, + "learning_rate": 3.1662556216451472e-06, + "loss": 0.0207, + "step": 23307 + }, + { + "epoch": 2.5268863833477884, + "grad_norm": 8.740274429321289, + "learning_rate": 3.1655302480777606e-06, + "loss": 0.0083, + "step": 23308 + }, + { + "epoch": 2.5269947961838684, + "grad_norm": 0.07580766081809998, + "learning_rate": 3.164804874510373e-06, + "loss": 0.0004, + "step": 23309 + }, + { + "epoch": 2.527103209019948, + "grad_norm": 0.42106547951698303, + "learning_rate": 3.1640795009429858e-06, + "loss": 0.001, + "step": 23310 + }, + { + "epoch": 2.527211621856028, + "grad_norm": 3.8190743923187256, + "learning_rate": 3.1633541273755983e-06, + "loss": 0.0468, + "step": 23311 + }, + { + "epoch": 2.5273200346921074, + "grad_norm": 0.13251923024654388, + "learning_rate": 3.1626287538082113e-06, + "loss": 0.0008, + "step": 23312 + }, + { + "epoch": 2.5274284475281874, + "grad_norm": 2.0655105113983154, + "learning_rate": 3.1619033802408243e-06, + "loss": 0.0256, + "step": 23313 + }, + { + "epoch": 2.5275368603642674, + "grad_norm": 6.123299598693848, + "learning_rate": 3.1611780066734373e-06, + "loss": 0.0248, + "step": 23314 + }, + { + "epoch": 2.527645273200347, + "grad_norm": 0.01427322905510664, + "learning_rate": 3.16045263310605e-06, + "loss": 0.0001, + "step": 23315 + }, + { + "epoch": 2.5277536860364265, + "grad_norm": 0.12114159017801285, + "learning_rate": 3.1597272595386624e-06, + "loss": 0.0005, + "step": 23316 + }, + { + "epoch": 2.5278620988725065, + "grad_norm": 1.9135048389434814, + "learning_rate": 3.159001885971276e-06, + "loss": 0.0016, + "step": 23317 + }, + { + "epoch": 2.5279705117085864, + "grad_norm": 0.037850409746170044, + "learning_rate": 3.1582765124038884e-06, + "loss": 0.0002, + "step": 23318 + }, + { + "epoch": 2.528078924544666, + "grad_norm": 0.7940524220466614, + "learning_rate": 3.157551138836501e-06, + "loss": 0.0008, + "step": 23319 + }, + { + "epoch": 2.528187337380746, + "grad_norm": 3.1101155281066895, + "learning_rate": 3.1568257652691136e-06, + "loss": 0.004, + "step": 23320 + }, + { + "epoch": 2.5282957502168255, + "grad_norm": 5.415589809417725, + "learning_rate": 3.156100391701727e-06, + "loss": 0.0367, + "step": 23321 + }, + { + "epoch": 2.5284041630529055, + "grad_norm": 11.277595520019531, + "learning_rate": 3.1553750181343395e-06, + "loss": 0.0292, + "step": 23322 + }, + { + "epoch": 2.5285125758889855, + "grad_norm": 0.291687548160553, + "learning_rate": 3.154649644566952e-06, + "loss": 0.0081, + "step": 23323 + }, + { + "epoch": 2.528620988725065, + "grad_norm": 0.11629121750593185, + "learning_rate": 3.1539242709995647e-06, + "loss": 0.0004, + "step": 23324 + }, + { + "epoch": 2.5287294015611446, + "grad_norm": 10.538558959960938, + "learning_rate": 3.1531988974321777e-06, + "loss": 0.005, + "step": 23325 + }, + { + "epoch": 2.5288378143972245, + "grad_norm": 0.22392378747463226, + "learning_rate": 3.1524735238647907e-06, + "loss": 0.0007, + "step": 23326 + }, + { + "epoch": 2.5289462272333045, + "grad_norm": 0.05530986562371254, + "learning_rate": 3.1517481502974036e-06, + "loss": 0.0002, + "step": 23327 + }, + { + "epoch": 2.529054640069384, + "grad_norm": 0.2890476882457733, + "learning_rate": 3.1510227767300162e-06, + "loss": 0.0014, + "step": 23328 + }, + { + "epoch": 2.529163052905464, + "grad_norm": 4.110182762145996, + "learning_rate": 3.1502974031626288e-06, + "loss": 0.0103, + "step": 23329 + }, + { + "epoch": 2.5292714657415436, + "grad_norm": 17.760011672973633, + "learning_rate": 3.149572029595242e-06, + "loss": 0.0985, + "step": 23330 + }, + { + "epoch": 2.5293798785776236, + "grad_norm": 9.082544326782227, + "learning_rate": 3.1488466560278548e-06, + "loss": 0.0045, + "step": 23331 + }, + { + "epoch": 2.5294882914137036, + "grad_norm": 0.07898245006799698, + "learning_rate": 3.1481212824604673e-06, + "loss": 0.0007, + "step": 23332 + }, + { + "epoch": 2.529596704249783, + "grad_norm": 0.25625672936439514, + "learning_rate": 3.14739590889308e-06, + "loss": 0.001, + "step": 23333 + }, + { + "epoch": 2.529705117085863, + "grad_norm": 6.161498546600342, + "learning_rate": 3.1466705353256933e-06, + "loss": 0.0312, + "step": 23334 + }, + { + "epoch": 2.5298135299219426, + "grad_norm": 4.602473258972168, + "learning_rate": 3.145945161758306e-06, + "loss": 0.0112, + "step": 23335 + }, + { + "epoch": 2.5299219427580226, + "grad_norm": 1.6100022792816162, + "learning_rate": 3.1452197881909184e-06, + "loss": 0.0082, + "step": 23336 + }, + { + "epoch": 2.5300303555941026, + "grad_norm": 1.6217230558395386, + "learning_rate": 3.1444944146235314e-06, + "loss": 0.0025, + "step": 23337 + }, + { + "epoch": 2.530138768430182, + "grad_norm": 0.07467596977949142, + "learning_rate": 3.1437690410561444e-06, + "loss": 0.0006, + "step": 23338 + }, + { + "epoch": 2.5302471812662617, + "grad_norm": 0.18124042451381683, + "learning_rate": 3.143043667488757e-06, + "loss": 0.0008, + "step": 23339 + }, + { + "epoch": 2.5303555941023417, + "grad_norm": 0.09049277007579803, + "learning_rate": 3.14231829392137e-06, + "loss": 0.0006, + "step": 23340 + }, + { + "epoch": 2.5304640069384217, + "grad_norm": 0.2690546214580536, + "learning_rate": 3.1415929203539825e-06, + "loss": 0.0018, + "step": 23341 + }, + { + "epoch": 2.530572419774501, + "grad_norm": 2.2129719257354736, + "learning_rate": 3.140867546786595e-06, + "loss": 0.029, + "step": 23342 + }, + { + "epoch": 2.530680832610581, + "grad_norm": 0.037910670042037964, + "learning_rate": 3.1401421732192085e-06, + "loss": 0.0003, + "step": 23343 + }, + { + "epoch": 2.5307892454466607, + "grad_norm": 1.5263259410858154, + "learning_rate": 3.139416799651821e-06, + "loss": 0.0154, + "step": 23344 + }, + { + "epoch": 2.5308976582827407, + "grad_norm": 5.3781561851501465, + "learning_rate": 3.1386914260844337e-06, + "loss": 0.0072, + "step": 23345 + }, + { + "epoch": 2.5310060711188207, + "grad_norm": 1.8441755771636963, + "learning_rate": 3.1379660525170462e-06, + "loss": 0.0058, + "step": 23346 + }, + { + "epoch": 2.5311144839549002, + "grad_norm": 0.0981832817196846, + "learning_rate": 3.1372406789496596e-06, + "loss": 0.0004, + "step": 23347 + }, + { + "epoch": 2.53122289679098, + "grad_norm": 0.4916994273662567, + "learning_rate": 3.136515305382272e-06, + "loss": 0.0007, + "step": 23348 + }, + { + "epoch": 2.5313313096270598, + "grad_norm": 7.502678871154785, + "learning_rate": 3.1357899318148848e-06, + "loss": 0.0233, + "step": 23349 + }, + { + "epoch": 2.5314397224631398, + "grad_norm": 0.00796958152204752, + "learning_rate": 3.1350645582474978e-06, + "loss": 0.0001, + "step": 23350 + }, + { + "epoch": 2.5315481352992193, + "grad_norm": 0.12218590825796127, + "learning_rate": 3.1343391846801108e-06, + "loss": 0.0006, + "step": 23351 + }, + { + "epoch": 2.5316565481352993, + "grad_norm": 3.0390970706939697, + "learning_rate": 3.1336138111127233e-06, + "loss": 0.0683, + "step": 23352 + }, + { + "epoch": 2.531764960971379, + "grad_norm": 3.7974634170532227, + "learning_rate": 3.1328884375453363e-06, + "loss": 0.0335, + "step": 23353 + }, + { + "epoch": 2.531873373807459, + "grad_norm": 0.9888657331466675, + "learning_rate": 3.132163063977949e-06, + "loss": 0.0032, + "step": 23354 + }, + { + "epoch": 2.531981786643539, + "grad_norm": 0.07908987253904343, + "learning_rate": 3.1314376904105614e-06, + "loss": 0.0008, + "step": 23355 + }, + { + "epoch": 2.5320901994796183, + "grad_norm": 6.545213222503662, + "learning_rate": 3.130712316843175e-06, + "loss": 0.0285, + "step": 23356 + }, + { + "epoch": 2.5321986123156983, + "grad_norm": 0.01384714338928461, + "learning_rate": 3.1299869432757874e-06, + "loss": 0.0001, + "step": 23357 + }, + { + "epoch": 2.532307025151778, + "grad_norm": 0.13511766493320465, + "learning_rate": 3.1292615697084e-06, + "loss": 0.0005, + "step": 23358 + }, + { + "epoch": 2.532415437987858, + "grad_norm": 0.632635772228241, + "learning_rate": 3.1285361961410126e-06, + "loss": 0.0023, + "step": 23359 + }, + { + "epoch": 2.532523850823938, + "grad_norm": 0.19954900443553925, + "learning_rate": 3.127810822573626e-06, + "loss": 0.0008, + "step": 23360 + }, + { + "epoch": 2.5326322636600174, + "grad_norm": 0.2876003086566925, + "learning_rate": 3.1270854490062385e-06, + "loss": 0.0017, + "step": 23361 + }, + { + "epoch": 2.532740676496097, + "grad_norm": 0.12507547438144684, + "learning_rate": 3.126360075438851e-06, + "loss": 0.0006, + "step": 23362 + }, + { + "epoch": 2.532849089332177, + "grad_norm": 2.8153533935546875, + "learning_rate": 3.125634701871464e-06, + "loss": 0.0136, + "step": 23363 + }, + { + "epoch": 2.532957502168257, + "grad_norm": 3.636373519897461, + "learning_rate": 3.124909328304077e-06, + "loss": 0.0173, + "step": 23364 + }, + { + "epoch": 2.5330659150043364, + "grad_norm": 0.32352373003959656, + "learning_rate": 3.1241839547366897e-06, + "loss": 0.0016, + "step": 23365 + }, + { + "epoch": 2.5331743278404164, + "grad_norm": 0.09991651773452759, + "learning_rate": 3.1234585811693026e-06, + "loss": 0.0006, + "step": 23366 + }, + { + "epoch": 2.533282740676496, + "grad_norm": 1.4334473609924316, + "learning_rate": 3.1227332076019152e-06, + "loss": 0.0029, + "step": 23367 + }, + { + "epoch": 2.533391153512576, + "grad_norm": 0.05422719568014145, + "learning_rate": 3.1220078340345278e-06, + "loss": 0.0003, + "step": 23368 + }, + { + "epoch": 2.533499566348656, + "grad_norm": 2.1871516704559326, + "learning_rate": 3.121282460467141e-06, + "loss": 0.0015, + "step": 23369 + }, + { + "epoch": 2.5336079791847355, + "grad_norm": 5.728876113891602, + "learning_rate": 3.1205570868997538e-06, + "loss": 0.006, + "step": 23370 + }, + { + "epoch": 2.533716392020815, + "grad_norm": 1.5882779359817505, + "learning_rate": 3.1198317133323663e-06, + "loss": 0.0044, + "step": 23371 + }, + { + "epoch": 2.533824804856895, + "grad_norm": 7.442884922027588, + "learning_rate": 3.119106339764979e-06, + "loss": 0.0464, + "step": 23372 + }, + { + "epoch": 2.533933217692975, + "grad_norm": 0.062025245279073715, + "learning_rate": 3.1183809661975923e-06, + "loss": 0.0005, + "step": 23373 + }, + { + "epoch": 2.5340416305290545, + "grad_norm": 0.5220057964324951, + "learning_rate": 3.117655592630205e-06, + "loss": 0.0022, + "step": 23374 + }, + { + "epoch": 2.5341500433651345, + "grad_norm": 0.15257930755615234, + "learning_rate": 3.1169302190628174e-06, + "loss": 0.0007, + "step": 23375 + }, + { + "epoch": 2.534258456201214, + "grad_norm": 3.1134886741638184, + "learning_rate": 3.1162048454954304e-06, + "loss": 0.0145, + "step": 23376 + }, + { + "epoch": 2.534366869037294, + "grad_norm": 0.08945423364639282, + "learning_rate": 3.1154794719280434e-06, + "loss": 0.0006, + "step": 23377 + }, + { + "epoch": 2.534475281873374, + "grad_norm": 1.364946722984314, + "learning_rate": 3.114754098360656e-06, + "loss": 0.0048, + "step": 23378 + }, + { + "epoch": 2.5345836947094535, + "grad_norm": 0.04853127896785736, + "learning_rate": 3.114028724793269e-06, + "loss": 0.0002, + "step": 23379 + }, + { + "epoch": 2.5346921075455335, + "grad_norm": 4.6798481941223145, + "learning_rate": 3.1133033512258815e-06, + "loss": 0.0195, + "step": 23380 + }, + { + "epoch": 2.534800520381613, + "grad_norm": 14.027507781982422, + "learning_rate": 3.112577977658494e-06, + "loss": 0.0355, + "step": 23381 + }, + { + "epoch": 2.534908933217693, + "grad_norm": 1.2644622325897217, + "learning_rate": 3.1118526040911075e-06, + "loss": 0.0275, + "step": 23382 + }, + { + "epoch": 2.5350173460537726, + "grad_norm": 0.044883158057928085, + "learning_rate": 3.11112723052372e-06, + "loss": 0.0001, + "step": 23383 + }, + { + "epoch": 2.5351257588898526, + "grad_norm": 0.06780777126550674, + "learning_rate": 3.1104018569563327e-06, + "loss": 0.0003, + "step": 23384 + }, + { + "epoch": 2.535234171725932, + "grad_norm": 0.06008235365152359, + "learning_rate": 3.1096764833889452e-06, + "loss": 0.0005, + "step": 23385 + }, + { + "epoch": 2.535342584562012, + "grad_norm": 0.009465291164815426, + "learning_rate": 3.1089511098215586e-06, + "loss": 0.0001, + "step": 23386 + }, + { + "epoch": 2.535450997398092, + "grad_norm": 0.061862122267484665, + "learning_rate": 3.1082257362541712e-06, + "loss": 0.0003, + "step": 23387 + }, + { + "epoch": 2.5355594102341716, + "grad_norm": 0.03568060323596001, + "learning_rate": 3.1075003626867838e-06, + "loss": 0.0003, + "step": 23388 + }, + { + "epoch": 2.5356678230702516, + "grad_norm": 0.017712825909256935, + "learning_rate": 3.1067749891193968e-06, + "loss": 0.0001, + "step": 23389 + }, + { + "epoch": 2.535776235906331, + "grad_norm": 2.9403939247131348, + "learning_rate": 3.1060496155520098e-06, + "loss": 0.0048, + "step": 23390 + }, + { + "epoch": 2.535884648742411, + "grad_norm": 0.232812762260437, + "learning_rate": 3.1053242419846223e-06, + "loss": 0.0006, + "step": 23391 + }, + { + "epoch": 2.535993061578491, + "grad_norm": 6.491919994354248, + "learning_rate": 3.1045988684172353e-06, + "loss": 0.0055, + "step": 23392 + }, + { + "epoch": 2.5361014744145707, + "grad_norm": 0.0033020921982824802, + "learning_rate": 3.103873494849848e-06, + "loss": 0.0, + "step": 23393 + }, + { + "epoch": 2.53620988725065, + "grad_norm": 2.7608139514923096, + "learning_rate": 3.1031481212824605e-06, + "loss": 0.0386, + "step": 23394 + }, + { + "epoch": 2.53631830008673, + "grad_norm": 1.2865257263183594, + "learning_rate": 3.102422747715074e-06, + "loss": 0.0049, + "step": 23395 + }, + { + "epoch": 2.53642671292281, + "grad_norm": 0.054167672991752625, + "learning_rate": 3.1016973741476864e-06, + "loss": 0.0003, + "step": 23396 + }, + { + "epoch": 2.5365351257588897, + "grad_norm": 0.7583296298980713, + "learning_rate": 3.100972000580299e-06, + "loss": 0.0006, + "step": 23397 + }, + { + "epoch": 2.5366435385949697, + "grad_norm": 2.394688844680786, + "learning_rate": 3.1002466270129116e-06, + "loss": 0.0065, + "step": 23398 + }, + { + "epoch": 2.5367519514310493, + "grad_norm": 0.10376225411891937, + "learning_rate": 3.099521253445525e-06, + "loss": 0.0003, + "step": 23399 + }, + { + "epoch": 2.5368603642671292, + "grad_norm": 0.09723907709121704, + "learning_rate": 3.0987958798781375e-06, + "loss": 0.0006, + "step": 23400 + }, + { + "epoch": 2.5369687771032092, + "grad_norm": 4.740965843200684, + "learning_rate": 3.09807050631075e-06, + "loss": 0.0199, + "step": 23401 + }, + { + "epoch": 2.5370771899392888, + "grad_norm": 0.1691821962594986, + "learning_rate": 3.097345132743363e-06, + "loss": 0.0003, + "step": 23402 + }, + { + "epoch": 2.5371856027753688, + "grad_norm": 0.17584675550460815, + "learning_rate": 3.096619759175976e-06, + "loss": 0.0012, + "step": 23403 + }, + { + "epoch": 2.5372940156114483, + "grad_norm": 7.168611526489258, + "learning_rate": 3.0958943856085887e-06, + "loss": 0.0414, + "step": 23404 + }, + { + "epoch": 2.5374024284475283, + "grad_norm": 6.455456256866455, + "learning_rate": 3.0951690120412017e-06, + "loss": 0.0319, + "step": 23405 + }, + { + "epoch": 2.537510841283608, + "grad_norm": 0.3898620307445526, + "learning_rate": 3.0944436384738142e-06, + "loss": 0.0009, + "step": 23406 + }, + { + "epoch": 2.537619254119688, + "grad_norm": 0.24310162663459778, + "learning_rate": 3.0937182649064268e-06, + "loss": 0.0007, + "step": 23407 + }, + { + "epoch": 2.5377276669557673, + "grad_norm": 1.8228000402450562, + "learning_rate": 3.09299289133904e-06, + "loss": 0.0019, + "step": 23408 + }, + { + "epoch": 2.5378360797918473, + "grad_norm": 0.13406187295913696, + "learning_rate": 3.0922675177716528e-06, + "loss": 0.0007, + "step": 23409 + }, + { + "epoch": 2.5379444926279273, + "grad_norm": 0.14984625577926636, + "learning_rate": 3.0915421442042653e-06, + "loss": 0.0005, + "step": 23410 + }, + { + "epoch": 2.538052905464007, + "grad_norm": 0.07668543606996536, + "learning_rate": 3.090816770636878e-06, + "loss": 0.0002, + "step": 23411 + }, + { + "epoch": 2.538161318300087, + "grad_norm": 0.03462367504835129, + "learning_rate": 3.0900913970694913e-06, + "loss": 0.0002, + "step": 23412 + }, + { + "epoch": 2.5382697311361664, + "grad_norm": 0.17750293016433716, + "learning_rate": 3.089366023502104e-06, + "loss": 0.0008, + "step": 23413 + }, + { + "epoch": 2.5383781439722464, + "grad_norm": 3.3845677375793457, + "learning_rate": 3.0886406499347164e-06, + "loss": 0.0451, + "step": 23414 + }, + { + "epoch": 2.5384865568083264, + "grad_norm": 0.0694703459739685, + "learning_rate": 3.0879152763673294e-06, + "loss": 0.0004, + "step": 23415 + }, + { + "epoch": 2.538594969644406, + "grad_norm": 0.08063601702451706, + "learning_rate": 3.0871899027999424e-06, + "loss": 0.0005, + "step": 23416 + }, + { + "epoch": 2.5387033824804854, + "grad_norm": 0.1013510674238205, + "learning_rate": 3.086464529232555e-06, + "loss": 0.0004, + "step": 23417 + }, + { + "epoch": 2.5388117953165654, + "grad_norm": 2.0415680408477783, + "learning_rate": 3.085739155665168e-06, + "loss": 0.0078, + "step": 23418 + }, + { + "epoch": 2.5389202081526454, + "grad_norm": 0.18486875295639038, + "learning_rate": 3.0850137820977806e-06, + "loss": 0.0011, + "step": 23419 + }, + { + "epoch": 2.539028620988725, + "grad_norm": 0.0011939476244151592, + "learning_rate": 3.084288408530393e-06, + "loss": 0.0, + "step": 23420 + }, + { + "epoch": 2.539137033824805, + "grad_norm": 0.38996008038520813, + "learning_rate": 3.0835630349630065e-06, + "loss": 0.0005, + "step": 23421 + }, + { + "epoch": 2.5392454466608845, + "grad_norm": 1.082331657409668, + "learning_rate": 3.082837661395619e-06, + "loss": 0.0031, + "step": 23422 + }, + { + "epoch": 2.5393538594969645, + "grad_norm": 0.6720934510231018, + "learning_rate": 3.0821122878282317e-06, + "loss": 0.0015, + "step": 23423 + }, + { + "epoch": 2.5394622723330444, + "grad_norm": 0.03750425949692726, + "learning_rate": 3.0813869142608442e-06, + "loss": 0.0003, + "step": 23424 + }, + { + "epoch": 2.539570685169124, + "grad_norm": 1.4893293380737305, + "learning_rate": 3.0806615406934576e-06, + "loss": 0.003, + "step": 23425 + }, + { + "epoch": 2.539679098005204, + "grad_norm": 0.055554646998643875, + "learning_rate": 3.0799361671260702e-06, + "loss": 0.0006, + "step": 23426 + }, + { + "epoch": 2.5397875108412835, + "grad_norm": 4.929266929626465, + "learning_rate": 3.0792107935586828e-06, + "loss": 0.0208, + "step": 23427 + }, + { + "epoch": 2.5398959236773635, + "grad_norm": 0.06808780878782272, + "learning_rate": 3.0784854199912958e-06, + "loss": 0.0001, + "step": 23428 + }, + { + "epoch": 2.540004336513443, + "grad_norm": 6.2967400550842285, + "learning_rate": 3.0777600464239088e-06, + "loss": 0.0048, + "step": 23429 + }, + { + "epoch": 2.540112749349523, + "grad_norm": 0.04401346668601036, + "learning_rate": 3.0770346728565213e-06, + "loss": 0.0003, + "step": 23430 + }, + { + "epoch": 2.5402211621856026, + "grad_norm": 3.1608009338378906, + "learning_rate": 3.0763092992891343e-06, + "loss": 0.0216, + "step": 23431 + }, + { + "epoch": 2.5403295750216826, + "grad_norm": 3.7251954078674316, + "learning_rate": 3.075583925721747e-06, + "loss": 0.0172, + "step": 23432 + }, + { + "epoch": 2.5404379878577625, + "grad_norm": 0.023358460515737534, + "learning_rate": 3.07485855215436e-06, + "loss": 0.0001, + "step": 23433 + }, + { + "epoch": 2.540546400693842, + "grad_norm": 0.028482981026172638, + "learning_rate": 3.074133178586973e-06, + "loss": 0.0003, + "step": 23434 + }, + { + "epoch": 2.540654813529922, + "grad_norm": 1.685150146484375, + "learning_rate": 3.0734078050195854e-06, + "loss": 0.0275, + "step": 23435 + }, + { + "epoch": 2.5407632263660016, + "grad_norm": 6.165113925933838, + "learning_rate": 3.072682431452198e-06, + "loss": 0.0269, + "step": 23436 + }, + { + "epoch": 2.5408716392020816, + "grad_norm": 0.008874970488250256, + "learning_rate": 3.0719570578848106e-06, + "loss": 0.0, + "step": 23437 + }, + { + "epoch": 2.5409800520381616, + "grad_norm": 0.0390455387532711, + "learning_rate": 3.071231684317424e-06, + "loss": 0.0002, + "step": 23438 + }, + { + "epoch": 2.541088464874241, + "grad_norm": 0.02074950933456421, + "learning_rate": 3.0705063107500365e-06, + "loss": 0.0001, + "step": 23439 + }, + { + "epoch": 2.5411968777103207, + "grad_norm": 13.915626525878906, + "learning_rate": 3.069780937182649e-06, + "loss": 0.006, + "step": 23440 + }, + { + "epoch": 2.5413052905464006, + "grad_norm": 0.05395475775003433, + "learning_rate": 3.069055563615262e-06, + "loss": 0.0004, + "step": 23441 + }, + { + "epoch": 2.5414137033824806, + "grad_norm": 0.03935147076845169, + "learning_rate": 3.068330190047875e-06, + "loss": 0.0002, + "step": 23442 + }, + { + "epoch": 2.54152211621856, + "grad_norm": 0.06761067360639572, + "learning_rate": 3.0676048164804877e-06, + "loss": 0.0002, + "step": 23443 + }, + { + "epoch": 2.54163052905464, + "grad_norm": 2.9683902263641357, + "learning_rate": 3.0668794429131007e-06, + "loss": 0.0038, + "step": 23444 + }, + { + "epoch": 2.5417389418907197, + "grad_norm": 0.3614649474620819, + "learning_rate": 3.0661540693457132e-06, + "loss": 0.0006, + "step": 23445 + }, + { + "epoch": 2.5418473547267997, + "grad_norm": 0.4150138795375824, + "learning_rate": 3.0654286957783262e-06, + "loss": 0.0007, + "step": 23446 + }, + { + "epoch": 2.5419557675628797, + "grad_norm": 0.00963500514626503, + "learning_rate": 3.064703322210939e-06, + "loss": 0.0001, + "step": 23447 + }, + { + "epoch": 2.542064180398959, + "grad_norm": 1.045740008354187, + "learning_rate": 3.0639779486435518e-06, + "loss": 0.0018, + "step": 23448 + }, + { + "epoch": 2.542172593235039, + "grad_norm": 0.023647600784897804, + "learning_rate": 3.0632525750761643e-06, + "loss": 0.0001, + "step": 23449 + }, + { + "epoch": 2.5422810060711187, + "grad_norm": 0.018536696210503578, + "learning_rate": 3.062527201508777e-06, + "loss": 0.0001, + "step": 23450 + }, + { + "epoch": 2.5423894189071987, + "grad_norm": 3.1847777366638184, + "learning_rate": 3.0618018279413903e-06, + "loss": 0.0074, + "step": 23451 + }, + { + "epoch": 2.5424978317432783, + "grad_norm": 0.013581466861069202, + "learning_rate": 3.061076454374003e-06, + "loss": 0.0002, + "step": 23452 + }, + { + "epoch": 2.5426062445793582, + "grad_norm": 0.4551873207092285, + "learning_rate": 3.0603510808066155e-06, + "loss": 0.0022, + "step": 23453 + }, + { + "epoch": 2.542714657415438, + "grad_norm": 13.873645782470703, + "learning_rate": 3.0596257072392284e-06, + "loss": 0.0532, + "step": 23454 + }, + { + "epoch": 2.5428230702515178, + "grad_norm": 8.863815307617188, + "learning_rate": 3.0589003336718414e-06, + "loss": 0.0049, + "step": 23455 + }, + { + "epoch": 2.5429314830875978, + "grad_norm": 0.10513697564601898, + "learning_rate": 3.058174960104454e-06, + "loss": 0.0012, + "step": 23456 + }, + { + "epoch": 2.5430398959236773, + "grad_norm": 3.1674022674560547, + "learning_rate": 3.057449586537067e-06, + "loss": 0.0239, + "step": 23457 + }, + { + "epoch": 2.5431483087597573, + "grad_norm": 0.6448012590408325, + "learning_rate": 3.0567242129696796e-06, + "loss": 0.0023, + "step": 23458 + }, + { + "epoch": 2.543256721595837, + "grad_norm": 0.2053733915090561, + "learning_rate": 3.0559988394022925e-06, + "loss": 0.0004, + "step": 23459 + }, + { + "epoch": 2.543365134431917, + "grad_norm": 2.72501802444458, + "learning_rate": 3.0552734658349055e-06, + "loss": 0.0322, + "step": 23460 + }, + { + "epoch": 2.543473547267997, + "grad_norm": 4.937190055847168, + "learning_rate": 3.054548092267518e-06, + "loss": 0.0196, + "step": 23461 + }, + { + "epoch": 2.5435819601040763, + "grad_norm": 4.116786956787109, + "learning_rate": 3.0538227187001307e-06, + "loss": 0.0266, + "step": 23462 + }, + { + "epoch": 2.543690372940156, + "grad_norm": 1.9511429071426392, + "learning_rate": 3.0530973451327432e-06, + "loss": 0.0028, + "step": 23463 + }, + { + "epoch": 2.543798785776236, + "grad_norm": 0.037365008145570755, + "learning_rate": 3.0523719715653567e-06, + "loss": 0.0005, + "step": 23464 + }, + { + "epoch": 2.543907198612316, + "grad_norm": 2.935035228729248, + "learning_rate": 3.0516465979979692e-06, + "loss": 0.0082, + "step": 23465 + }, + { + "epoch": 2.5440156114483954, + "grad_norm": 0.07571356743574142, + "learning_rate": 3.0509212244305818e-06, + "loss": 0.0005, + "step": 23466 + }, + { + "epoch": 2.5441240242844754, + "grad_norm": 0.8030273914337158, + "learning_rate": 3.0501958508631948e-06, + "loss": 0.0033, + "step": 23467 + }, + { + "epoch": 2.544232437120555, + "grad_norm": 0.7501683235168457, + "learning_rate": 3.0494704772958078e-06, + "loss": 0.0022, + "step": 23468 + }, + { + "epoch": 2.544340849956635, + "grad_norm": 2.377293825149536, + "learning_rate": 3.0487451037284203e-06, + "loss": 0.015, + "step": 23469 + }, + { + "epoch": 2.544449262792715, + "grad_norm": 0.06758018583059311, + "learning_rate": 3.0480197301610333e-06, + "loss": 0.0004, + "step": 23470 + }, + { + "epoch": 2.5445576756287944, + "grad_norm": 3.2012813091278076, + "learning_rate": 3.047294356593646e-06, + "loss": 0.0368, + "step": 23471 + }, + { + "epoch": 2.544666088464874, + "grad_norm": 2.711674690246582, + "learning_rate": 3.046568983026259e-06, + "loss": 0.0465, + "step": 23472 + }, + { + "epoch": 2.544774501300954, + "grad_norm": 0.2832036018371582, + "learning_rate": 3.045843609458872e-06, + "loss": 0.0026, + "step": 23473 + }, + { + "epoch": 2.544882914137034, + "grad_norm": 5.568324565887451, + "learning_rate": 3.0451182358914844e-06, + "loss": 0.0478, + "step": 23474 + }, + { + "epoch": 2.5449913269731135, + "grad_norm": 0.08030623197555542, + "learning_rate": 3.044392862324097e-06, + "loss": 0.0007, + "step": 23475 + }, + { + "epoch": 2.5450997398091935, + "grad_norm": 2.4129908084869385, + "learning_rate": 3.0436674887567096e-06, + "loss": 0.0061, + "step": 23476 + }, + { + "epoch": 2.545208152645273, + "grad_norm": 0.5892312526702881, + "learning_rate": 3.042942115189323e-06, + "loss": 0.0026, + "step": 23477 + }, + { + "epoch": 2.545316565481353, + "grad_norm": 2.7657482624053955, + "learning_rate": 3.0422167416219356e-06, + "loss": 0.0209, + "step": 23478 + }, + { + "epoch": 2.545424978317433, + "grad_norm": 0.2978633940219879, + "learning_rate": 3.041491368054548e-06, + "loss": 0.0014, + "step": 23479 + }, + { + "epoch": 2.5455333911535125, + "grad_norm": 4.888904094696045, + "learning_rate": 3.040765994487161e-06, + "loss": 0.0476, + "step": 23480 + }, + { + "epoch": 2.5456418039895925, + "grad_norm": 9.602312088012695, + "learning_rate": 3.040040620919774e-06, + "loss": 0.0111, + "step": 23481 + }, + { + "epoch": 2.545750216825672, + "grad_norm": 0.364872545003891, + "learning_rate": 3.0393152473523867e-06, + "loss": 0.0008, + "step": 23482 + }, + { + "epoch": 2.545858629661752, + "grad_norm": 0.038417745381593704, + "learning_rate": 3.0385898737849997e-06, + "loss": 0.0002, + "step": 23483 + }, + { + "epoch": 2.545967042497832, + "grad_norm": 0.03234262764453888, + "learning_rate": 3.0378645002176122e-06, + "loss": 0.0003, + "step": 23484 + }, + { + "epoch": 2.5460754553339116, + "grad_norm": 0.025793997570872307, + "learning_rate": 3.0371391266502252e-06, + "loss": 0.0004, + "step": 23485 + }, + { + "epoch": 2.546183868169991, + "grad_norm": 0.2038676142692566, + "learning_rate": 3.036413753082838e-06, + "loss": 0.0005, + "step": 23486 + }, + { + "epoch": 2.546292281006071, + "grad_norm": 1.454999327659607, + "learning_rate": 3.0356883795154508e-06, + "loss": 0.0249, + "step": 23487 + }, + { + "epoch": 2.546400693842151, + "grad_norm": 0.06641917675733566, + "learning_rate": 3.0349630059480633e-06, + "loss": 0.0003, + "step": 23488 + }, + { + "epoch": 2.5465091066782306, + "grad_norm": 2.19897723197937, + "learning_rate": 3.034237632380676e-06, + "loss": 0.002, + "step": 23489 + }, + { + "epoch": 2.5466175195143106, + "grad_norm": 0.12729120254516602, + "learning_rate": 3.0335122588132893e-06, + "loss": 0.0005, + "step": 23490 + }, + { + "epoch": 2.54672593235039, + "grad_norm": 1.7476435899734497, + "learning_rate": 3.032786885245902e-06, + "loss": 0.0597, + "step": 23491 + }, + { + "epoch": 2.54683434518647, + "grad_norm": 1.6948649883270264, + "learning_rate": 3.0320615116785145e-06, + "loss": 0.0265, + "step": 23492 + }, + { + "epoch": 2.54694275802255, + "grad_norm": 0.25694265961647034, + "learning_rate": 3.0313361381111274e-06, + "loss": 0.0016, + "step": 23493 + }, + { + "epoch": 2.5470511708586296, + "grad_norm": 2.648648738861084, + "learning_rate": 3.0306107645437404e-06, + "loss": 0.0072, + "step": 23494 + }, + { + "epoch": 2.547159583694709, + "grad_norm": 36.52552032470703, + "learning_rate": 3.029885390976353e-06, + "loss": 0.0444, + "step": 23495 + }, + { + "epoch": 2.547267996530789, + "grad_norm": 0.02976774238049984, + "learning_rate": 3.029160017408966e-06, + "loss": 0.0003, + "step": 23496 + }, + { + "epoch": 2.547376409366869, + "grad_norm": 0.28230416774749756, + "learning_rate": 3.0284346438415786e-06, + "loss": 0.0012, + "step": 23497 + }, + { + "epoch": 2.5474848222029487, + "grad_norm": 2.455521583557129, + "learning_rate": 3.0277092702741915e-06, + "loss": 0.0058, + "step": 23498 + }, + { + "epoch": 2.5475932350390287, + "grad_norm": 2.453577756881714, + "learning_rate": 3.0269838967068045e-06, + "loss": 0.01, + "step": 23499 + }, + { + "epoch": 2.547701647875108, + "grad_norm": 0.9563501477241516, + "learning_rate": 3.026258523139417e-06, + "loss": 0.045, + "step": 23500 + }, + { + "epoch": 2.547810060711188, + "grad_norm": 0.9104248881340027, + "learning_rate": 3.0255331495720297e-06, + "loss": 0.0015, + "step": 23501 + }, + { + "epoch": 2.547918473547268, + "grad_norm": 2.145657539367676, + "learning_rate": 3.0248077760046422e-06, + "loss": 0.0396, + "step": 23502 + }, + { + "epoch": 2.5480268863833477, + "grad_norm": 9.224148750305176, + "learning_rate": 3.0240824024372557e-06, + "loss": 0.0374, + "step": 23503 + }, + { + "epoch": 2.5481352992194277, + "grad_norm": 0.5655502676963806, + "learning_rate": 3.0233570288698682e-06, + "loss": 0.0015, + "step": 23504 + }, + { + "epoch": 2.5482437120555073, + "grad_norm": 0.006646964233368635, + "learning_rate": 3.022631655302481e-06, + "loss": 0.0001, + "step": 23505 + }, + { + "epoch": 2.5483521248915872, + "grad_norm": 0.053511325269937515, + "learning_rate": 3.0219062817350938e-06, + "loss": 0.0004, + "step": 23506 + }, + { + "epoch": 2.5484605377276672, + "grad_norm": 0.14186617732048035, + "learning_rate": 3.0211809081677068e-06, + "loss": 0.0008, + "step": 23507 + }, + { + "epoch": 2.5485689505637468, + "grad_norm": 0.06991133838891983, + "learning_rate": 3.0204555346003193e-06, + "loss": 0.0004, + "step": 23508 + }, + { + "epoch": 2.5486773633998263, + "grad_norm": 0.12093944847583771, + "learning_rate": 3.0197301610329323e-06, + "loss": 0.0008, + "step": 23509 + }, + { + "epoch": 2.5487857762359063, + "grad_norm": 0.0932457223534584, + "learning_rate": 3.019004787465545e-06, + "loss": 0.0004, + "step": 23510 + }, + { + "epoch": 2.5488941890719863, + "grad_norm": 3.529062032699585, + "learning_rate": 3.018279413898158e-06, + "loss": 0.0105, + "step": 23511 + }, + { + "epoch": 2.549002601908066, + "grad_norm": 0.3061758875846863, + "learning_rate": 3.017554040330771e-06, + "loss": 0.0009, + "step": 23512 + }, + { + "epoch": 2.549111014744146, + "grad_norm": 0.1586528718471527, + "learning_rate": 3.0168286667633834e-06, + "loss": 0.0005, + "step": 23513 + }, + { + "epoch": 2.5492194275802254, + "grad_norm": 0.03503969684243202, + "learning_rate": 3.016103293195996e-06, + "loss": 0.0002, + "step": 23514 + }, + { + "epoch": 2.5493278404163053, + "grad_norm": 0.06061824411153793, + "learning_rate": 3.0153779196286086e-06, + "loss": 0.0004, + "step": 23515 + }, + { + "epoch": 2.5494362532523853, + "grad_norm": 1.3532469272613525, + "learning_rate": 3.014652546061222e-06, + "loss": 0.0016, + "step": 23516 + }, + { + "epoch": 2.549544666088465, + "grad_norm": 1.136035680770874, + "learning_rate": 3.0139271724938346e-06, + "loss": 0.0023, + "step": 23517 + }, + { + "epoch": 2.5496530789245444, + "grad_norm": 0.04599950090050697, + "learning_rate": 3.013201798926447e-06, + "loss": 0.0003, + "step": 23518 + }, + { + "epoch": 2.5497614917606244, + "grad_norm": 2.029892683029175, + "learning_rate": 3.01247642535906e-06, + "loss": 0.0234, + "step": 23519 + }, + { + "epoch": 2.5498699045967044, + "grad_norm": 0.27538129687309265, + "learning_rate": 3.011751051791673e-06, + "loss": 0.0009, + "step": 23520 + }, + { + "epoch": 2.549978317432784, + "grad_norm": 0.01391767431050539, + "learning_rate": 3.0110256782242857e-06, + "loss": 0.0002, + "step": 23521 + }, + { + "epoch": 2.550086730268864, + "grad_norm": 4.287509441375732, + "learning_rate": 3.0103003046568987e-06, + "loss": 0.0141, + "step": 23522 + }, + { + "epoch": 2.5501951431049434, + "grad_norm": 0.11157096177339554, + "learning_rate": 3.0095749310895112e-06, + "loss": 0.0004, + "step": 23523 + }, + { + "epoch": 2.5503035559410234, + "grad_norm": 0.039445795118808746, + "learning_rate": 3.0088495575221242e-06, + "loss": 0.0002, + "step": 23524 + }, + { + "epoch": 2.5504119687771034, + "grad_norm": 3.570814847946167, + "learning_rate": 3.008124183954737e-06, + "loss": 0.0036, + "step": 23525 + }, + { + "epoch": 2.550520381613183, + "grad_norm": 1.7376132011413574, + "learning_rate": 3.0073988103873498e-06, + "loss": 0.0374, + "step": 23526 + }, + { + "epoch": 2.550628794449263, + "grad_norm": 4.8491621017456055, + "learning_rate": 3.0066734368199623e-06, + "loss": 0.0135, + "step": 23527 + }, + { + "epoch": 2.5507372072853425, + "grad_norm": 1.4257128238677979, + "learning_rate": 3.0059480632525758e-06, + "loss": 0.0034, + "step": 23528 + }, + { + "epoch": 2.5508456201214225, + "grad_norm": 0.10665471851825714, + "learning_rate": 3.0052226896851883e-06, + "loss": 0.0006, + "step": 23529 + }, + { + "epoch": 2.550954032957502, + "grad_norm": 0.11517055332660675, + "learning_rate": 3.004497316117801e-06, + "loss": 0.0006, + "step": 23530 + }, + { + "epoch": 2.551062445793582, + "grad_norm": 4.160928249359131, + "learning_rate": 3.0037719425504135e-06, + "loss": 0.0199, + "step": 23531 + }, + { + "epoch": 2.5511708586296615, + "grad_norm": 8.62647533416748, + "learning_rate": 3.0030465689830264e-06, + "loss": 0.0126, + "step": 23532 + }, + { + "epoch": 2.5512792714657415, + "grad_norm": 0.01828037202358246, + "learning_rate": 3.0023211954156394e-06, + "loss": 0.0001, + "step": 23533 + }, + { + "epoch": 2.5513876843018215, + "grad_norm": 0.12052911520004272, + "learning_rate": 3.001595821848252e-06, + "loss": 0.0007, + "step": 23534 + }, + { + "epoch": 2.551496097137901, + "grad_norm": 1.0771220922470093, + "learning_rate": 3.000870448280865e-06, + "loss": 0.0059, + "step": 23535 + }, + { + "epoch": 2.551604509973981, + "grad_norm": 1.4820157289505005, + "learning_rate": 3.0001450747134776e-06, + "loss": 0.046, + "step": 23536 + }, + { + "epoch": 2.5517129228100606, + "grad_norm": 0.16204692423343658, + "learning_rate": 2.9994197011460906e-06, + "loss": 0.0007, + "step": 23537 + }, + { + "epoch": 2.5518213356461406, + "grad_norm": 0.009647411294281483, + "learning_rate": 2.9986943275787035e-06, + "loss": 0.0001, + "step": 23538 + }, + { + "epoch": 2.5519297484822205, + "grad_norm": 0.2666403353214264, + "learning_rate": 2.997968954011316e-06, + "loss": 0.0012, + "step": 23539 + }, + { + "epoch": 2.5520381613183, + "grad_norm": 2.8866076469421387, + "learning_rate": 2.9972435804439287e-06, + "loss": 0.0161, + "step": 23540 + }, + { + "epoch": 2.5521465741543796, + "grad_norm": 1.7067452669143677, + "learning_rate": 2.996518206876542e-06, + "loss": 0.0298, + "step": 23541 + }, + { + "epoch": 2.5522549869904596, + "grad_norm": 3.6577563285827637, + "learning_rate": 2.9957928333091547e-06, + "loss": 0.0443, + "step": 23542 + }, + { + "epoch": 2.5523633998265396, + "grad_norm": 0.33298352360725403, + "learning_rate": 2.9950674597417672e-06, + "loss": 0.0005, + "step": 23543 + }, + { + "epoch": 2.552471812662619, + "grad_norm": 0.030660338699817657, + "learning_rate": 2.99434208617438e-06, + "loss": 0.0003, + "step": 23544 + }, + { + "epoch": 2.552580225498699, + "grad_norm": 0.07733698934316635, + "learning_rate": 2.9936167126069928e-06, + "loss": 0.0004, + "step": 23545 + }, + { + "epoch": 2.5526886383347787, + "grad_norm": 0.17866940796375275, + "learning_rate": 2.9928913390396058e-06, + "loss": 0.0007, + "step": 23546 + }, + { + "epoch": 2.5527970511708586, + "grad_norm": 0.2548574209213257, + "learning_rate": 2.9921659654722183e-06, + "loss": 0.0006, + "step": 23547 + }, + { + "epoch": 2.5529054640069386, + "grad_norm": 1.3802385330200195, + "learning_rate": 2.9914405919048313e-06, + "loss": 0.0083, + "step": 23548 + }, + { + "epoch": 2.553013876843018, + "grad_norm": 5.279461860656738, + "learning_rate": 2.990715218337444e-06, + "loss": 0.0413, + "step": 23549 + }, + { + "epoch": 2.553122289679098, + "grad_norm": 0.5406032204627991, + "learning_rate": 2.989989844770057e-06, + "loss": 0.0012, + "step": 23550 + }, + { + "epoch": 2.5532307025151777, + "grad_norm": 0.3767838478088379, + "learning_rate": 2.98926447120267e-06, + "loss": 0.0016, + "step": 23551 + }, + { + "epoch": 2.5533391153512577, + "grad_norm": 0.12104971706867218, + "learning_rate": 2.9885390976352824e-06, + "loss": 0.0006, + "step": 23552 + }, + { + "epoch": 2.5534475281873372, + "grad_norm": 7.3425726890563965, + "learning_rate": 2.987813724067895e-06, + "loss": 0.0235, + "step": 23553 + }, + { + "epoch": 2.553555941023417, + "grad_norm": 3.5646800994873047, + "learning_rate": 2.9870883505005084e-06, + "loss": 0.0645, + "step": 23554 + }, + { + "epoch": 2.5536643538594967, + "grad_norm": 0.9822933077812195, + "learning_rate": 2.986362976933121e-06, + "loss": 0.0019, + "step": 23555 + }, + { + "epoch": 2.5537727666955767, + "grad_norm": 0.05057888850569725, + "learning_rate": 2.9856376033657336e-06, + "loss": 0.0004, + "step": 23556 + }, + { + "epoch": 2.5538811795316567, + "grad_norm": 0.029774975031614304, + "learning_rate": 2.984912229798346e-06, + "loss": 0.0002, + "step": 23557 + }, + { + "epoch": 2.5539895923677363, + "grad_norm": 0.23975178599357605, + "learning_rate": 2.984186856230959e-06, + "loss": 0.0011, + "step": 23558 + }, + { + "epoch": 2.5540980052038162, + "grad_norm": 3.0924293994903564, + "learning_rate": 2.983461482663572e-06, + "loss": 0.0408, + "step": 23559 + }, + { + "epoch": 2.554206418039896, + "grad_norm": 6.316823482513428, + "learning_rate": 2.9827361090961847e-06, + "loss": 0.0256, + "step": 23560 + }, + { + "epoch": 2.5543148308759758, + "grad_norm": 10.473893165588379, + "learning_rate": 2.9820107355287977e-06, + "loss": 0.0289, + "step": 23561 + }, + { + "epoch": 2.5544232437120558, + "grad_norm": 0.1446746438741684, + "learning_rate": 2.9812853619614102e-06, + "loss": 0.0008, + "step": 23562 + }, + { + "epoch": 2.5545316565481353, + "grad_norm": 3.0308239459991455, + "learning_rate": 2.9805599883940232e-06, + "loss": 0.013, + "step": 23563 + }, + { + "epoch": 2.554640069384215, + "grad_norm": 0.07496921718120575, + "learning_rate": 2.9798346148266362e-06, + "loss": 0.0006, + "step": 23564 + }, + { + "epoch": 2.554748482220295, + "grad_norm": 1.6789625883102417, + "learning_rate": 2.9791092412592488e-06, + "loss": 0.006, + "step": 23565 + }, + { + "epoch": 2.554856895056375, + "grad_norm": 2.0336523056030273, + "learning_rate": 2.9783838676918613e-06, + "loss": 0.0021, + "step": 23566 + }, + { + "epoch": 2.5549653078924544, + "grad_norm": 12.35665225982666, + "learning_rate": 2.9776584941244748e-06, + "loss": 0.0376, + "step": 23567 + }, + { + "epoch": 2.5550737207285343, + "grad_norm": 0.19410274922847748, + "learning_rate": 2.9769331205570873e-06, + "loss": 0.0013, + "step": 23568 + }, + { + "epoch": 2.555182133564614, + "grad_norm": 3.0441734790802, + "learning_rate": 2.9762077469897e-06, + "loss": 0.0584, + "step": 23569 + }, + { + "epoch": 2.555290546400694, + "grad_norm": 0.9626722931861877, + "learning_rate": 2.9754823734223125e-06, + "loss": 0.0022, + "step": 23570 + }, + { + "epoch": 2.555398959236774, + "grad_norm": 0.8878354430198669, + "learning_rate": 2.9747569998549255e-06, + "loss": 0.0022, + "step": 23571 + }, + { + "epoch": 2.5555073720728534, + "grad_norm": 7.126128196716309, + "learning_rate": 2.9740316262875384e-06, + "loss": 0.0169, + "step": 23572 + }, + { + "epoch": 2.5556157849089334, + "grad_norm": 0.6728407740592957, + "learning_rate": 2.973306252720151e-06, + "loss": 0.0035, + "step": 23573 + }, + { + "epoch": 2.555724197745013, + "grad_norm": 0.5374983549118042, + "learning_rate": 2.972580879152764e-06, + "loss": 0.0027, + "step": 23574 + }, + { + "epoch": 2.555832610581093, + "grad_norm": 0.02223791927099228, + "learning_rate": 2.9718555055853766e-06, + "loss": 0.0002, + "step": 23575 + }, + { + "epoch": 2.5559410234171724, + "grad_norm": 0.049870721995830536, + "learning_rate": 2.9711301320179896e-06, + "loss": 0.0005, + "step": 23576 + }, + { + "epoch": 2.5560494362532524, + "grad_norm": 0.028732987120747566, + "learning_rate": 2.9704047584506025e-06, + "loss": 0.0002, + "step": 23577 + }, + { + "epoch": 2.556157849089332, + "grad_norm": 0.7764158844947815, + "learning_rate": 2.969679384883215e-06, + "loss": 0.0032, + "step": 23578 + }, + { + "epoch": 2.556266261925412, + "grad_norm": 5.6649017333984375, + "learning_rate": 2.9689540113158277e-06, + "loss": 0.0127, + "step": 23579 + }, + { + "epoch": 2.556374674761492, + "grad_norm": 0.12351462990045547, + "learning_rate": 2.968228637748441e-06, + "loss": 0.0009, + "step": 23580 + }, + { + "epoch": 2.5564830875975715, + "grad_norm": 0.12901510298252106, + "learning_rate": 2.9675032641810537e-06, + "loss": 0.0007, + "step": 23581 + }, + { + "epoch": 2.5565915004336515, + "grad_norm": 0.49903690814971924, + "learning_rate": 2.9667778906136662e-06, + "loss": 0.0008, + "step": 23582 + }, + { + "epoch": 2.556699913269731, + "grad_norm": 0.18406789004802704, + "learning_rate": 2.966052517046279e-06, + "loss": 0.0017, + "step": 23583 + }, + { + "epoch": 2.556808326105811, + "grad_norm": 1.6765226125717163, + "learning_rate": 2.9653271434788918e-06, + "loss": 0.0073, + "step": 23584 + }, + { + "epoch": 2.556916738941891, + "grad_norm": 1.8271570205688477, + "learning_rate": 2.9646017699115048e-06, + "loss": 0.0186, + "step": 23585 + }, + { + "epoch": 2.5570251517779705, + "grad_norm": 0.031815700232982635, + "learning_rate": 2.9638763963441173e-06, + "loss": 0.0002, + "step": 23586 + }, + { + "epoch": 2.55713356461405, + "grad_norm": 2.5507709980010986, + "learning_rate": 2.9631510227767303e-06, + "loss": 0.0215, + "step": 23587 + }, + { + "epoch": 2.55724197745013, + "grad_norm": 0.08630301058292389, + "learning_rate": 2.962425649209343e-06, + "loss": 0.0011, + "step": 23588 + }, + { + "epoch": 2.55735039028621, + "grad_norm": 2.2458200454711914, + "learning_rate": 2.961700275641956e-06, + "loss": 0.0082, + "step": 23589 + }, + { + "epoch": 2.5574588031222896, + "grad_norm": 0.5311598777770996, + "learning_rate": 2.960974902074569e-06, + "loss": 0.0015, + "step": 23590 + }, + { + "epoch": 2.5575672159583696, + "grad_norm": 1.7870947122573853, + "learning_rate": 2.9602495285071814e-06, + "loss": 0.0054, + "step": 23591 + }, + { + "epoch": 2.557675628794449, + "grad_norm": 3.9918453693389893, + "learning_rate": 2.959524154939794e-06, + "loss": 0.0076, + "step": 23592 + }, + { + "epoch": 2.557784041630529, + "grad_norm": 0.11620545387268066, + "learning_rate": 2.9587987813724074e-06, + "loss": 0.0009, + "step": 23593 + }, + { + "epoch": 2.557892454466609, + "grad_norm": 0.11243082582950592, + "learning_rate": 2.95807340780502e-06, + "loss": 0.001, + "step": 23594 + }, + { + "epoch": 2.5580008673026886, + "grad_norm": 0.7181580662727356, + "learning_rate": 2.9573480342376326e-06, + "loss": 0.0027, + "step": 23595 + }, + { + "epoch": 2.5581092801387686, + "grad_norm": 0.11358712613582611, + "learning_rate": 2.956622660670245e-06, + "loss": 0.0006, + "step": 23596 + }, + { + "epoch": 2.558217692974848, + "grad_norm": 10.047438621520996, + "learning_rate": 2.955897287102858e-06, + "loss": 0.0073, + "step": 23597 + }, + { + "epoch": 2.558326105810928, + "grad_norm": 0.10462488979101181, + "learning_rate": 2.955171913535471e-06, + "loss": 0.0007, + "step": 23598 + }, + { + "epoch": 2.5584345186470077, + "grad_norm": 2.7909629344940186, + "learning_rate": 2.9544465399680837e-06, + "loss": 0.0101, + "step": 23599 + }, + { + "epoch": 2.5585429314830876, + "grad_norm": 0.9596793055534363, + "learning_rate": 2.9537211664006967e-06, + "loss": 0.0021, + "step": 23600 + }, + { + "epoch": 2.558651344319167, + "grad_norm": 0.8208740949630737, + "learning_rate": 2.9529957928333092e-06, + "loss": 0.0007, + "step": 23601 + }, + { + "epoch": 2.558759757155247, + "grad_norm": 15.411975860595703, + "learning_rate": 2.9522704192659222e-06, + "loss": 0.0103, + "step": 23602 + }, + { + "epoch": 2.558868169991327, + "grad_norm": 0.54674232006073, + "learning_rate": 2.9515450456985352e-06, + "loss": 0.0015, + "step": 23603 + }, + { + "epoch": 2.5589765828274067, + "grad_norm": 0.2757827043533325, + "learning_rate": 2.9508196721311478e-06, + "loss": 0.0014, + "step": 23604 + }, + { + "epoch": 2.5590849956634867, + "grad_norm": 0.0750565230846405, + "learning_rate": 2.9500942985637603e-06, + "loss": 0.0004, + "step": 23605 + }, + { + "epoch": 2.5591934084995662, + "grad_norm": 0.024519171565771103, + "learning_rate": 2.9493689249963738e-06, + "loss": 0.0001, + "step": 23606 + }, + { + "epoch": 2.559301821335646, + "grad_norm": 0.14034445583820343, + "learning_rate": 2.9486435514289863e-06, + "loss": 0.0009, + "step": 23607 + }, + { + "epoch": 2.559410234171726, + "grad_norm": 0.03202557936310768, + "learning_rate": 2.947918177861599e-06, + "loss": 0.0003, + "step": 23608 + }, + { + "epoch": 2.5595186470078057, + "grad_norm": 0.20134301483631134, + "learning_rate": 2.9471928042942115e-06, + "loss": 0.0011, + "step": 23609 + }, + { + "epoch": 2.5596270598438853, + "grad_norm": 0.048636406660079956, + "learning_rate": 2.9464674307268245e-06, + "loss": 0.0005, + "step": 23610 + }, + { + "epoch": 2.5597354726799653, + "grad_norm": 0.5551431179046631, + "learning_rate": 2.9457420571594374e-06, + "loss": 0.0023, + "step": 23611 + }, + { + "epoch": 2.5598438855160452, + "grad_norm": 2.920341968536377, + "learning_rate": 2.94501668359205e-06, + "loss": 0.0055, + "step": 23612 + }, + { + "epoch": 2.559952298352125, + "grad_norm": 0.03500868007540703, + "learning_rate": 2.944291310024663e-06, + "loss": 0.0003, + "step": 23613 + }, + { + "epoch": 2.5600607111882048, + "grad_norm": 1.7953755855560303, + "learning_rate": 2.9435659364572756e-06, + "loss": 0.0065, + "step": 23614 + }, + { + "epoch": 2.5601691240242843, + "grad_norm": 3.5634031295776367, + "learning_rate": 2.9428405628898886e-06, + "loss": 0.0101, + "step": 23615 + }, + { + "epoch": 2.5602775368603643, + "grad_norm": 0.07891176640987396, + "learning_rate": 2.9421151893225015e-06, + "loss": 0.0009, + "step": 23616 + }, + { + "epoch": 2.5603859496964443, + "grad_norm": 0.026776202023029327, + "learning_rate": 2.941389815755114e-06, + "loss": 0.0003, + "step": 23617 + }, + { + "epoch": 2.560494362532524, + "grad_norm": 1.0549967288970947, + "learning_rate": 2.9406644421877267e-06, + "loss": 0.0033, + "step": 23618 + }, + { + "epoch": 2.5606027753686034, + "grad_norm": 0.0073793507181108, + "learning_rate": 2.93993906862034e-06, + "loss": 0.0001, + "step": 23619 + }, + { + "epoch": 2.5607111882046834, + "grad_norm": 4.097942352294922, + "learning_rate": 2.9392136950529527e-06, + "loss": 0.0262, + "step": 23620 + }, + { + "epoch": 2.5608196010407633, + "grad_norm": 0.06132596358656883, + "learning_rate": 2.9384883214855652e-06, + "loss": 0.0004, + "step": 23621 + }, + { + "epoch": 2.560928013876843, + "grad_norm": 2.4326303005218506, + "learning_rate": 2.937762947918178e-06, + "loss": 0.0141, + "step": 23622 + }, + { + "epoch": 2.561036426712923, + "grad_norm": 0.11943872272968292, + "learning_rate": 2.937037574350791e-06, + "loss": 0.0006, + "step": 23623 + }, + { + "epoch": 2.5611448395490024, + "grad_norm": 1.5026997327804565, + "learning_rate": 2.9363122007834038e-06, + "loss": 0.0057, + "step": 23624 + }, + { + "epoch": 2.5612532523850824, + "grad_norm": 0.04679204151034355, + "learning_rate": 2.9355868272160163e-06, + "loss": 0.0003, + "step": 23625 + }, + { + "epoch": 2.5613616652211624, + "grad_norm": 2.7487130165100098, + "learning_rate": 2.9348614536486293e-06, + "loss": 0.0387, + "step": 23626 + }, + { + "epoch": 2.561470078057242, + "grad_norm": 0.1074925884604454, + "learning_rate": 2.934136080081242e-06, + "loss": 0.0007, + "step": 23627 + }, + { + "epoch": 2.561578490893322, + "grad_norm": 2.1879475116729736, + "learning_rate": 2.933410706513855e-06, + "loss": 0.0298, + "step": 23628 + }, + { + "epoch": 2.5616869037294014, + "grad_norm": 2.7562832832336426, + "learning_rate": 2.932685332946468e-06, + "loss": 0.0231, + "step": 23629 + }, + { + "epoch": 2.5617953165654814, + "grad_norm": 0.006077961064875126, + "learning_rate": 2.9319599593790805e-06, + "loss": 0.0001, + "step": 23630 + }, + { + "epoch": 2.5619037294015614, + "grad_norm": 0.01633935421705246, + "learning_rate": 2.931234585811693e-06, + "loss": 0.0001, + "step": 23631 + }, + { + "epoch": 2.562012142237641, + "grad_norm": 0.42062392830848694, + "learning_rate": 2.9305092122443064e-06, + "loss": 0.002, + "step": 23632 + }, + { + "epoch": 2.5621205550737205, + "grad_norm": 0.08020927757024765, + "learning_rate": 2.929783838676919e-06, + "loss": 0.0004, + "step": 23633 + }, + { + "epoch": 2.5622289679098005, + "grad_norm": 0.14075663685798645, + "learning_rate": 2.9290584651095316e-06, + "loss": 0.0006, + "step": 23634 + }, + { + "epoch": 2.5623373807458805, + "grad_norm": 0.1855076253414154, + "learning_rate": 2.928333091542144e-06, + "loss": 0.001, + "step": 23635 + }, + { + "epoch": 2.56244579358196, + "grad_norm": 0.04833170026540756, + "learning_rate": 2.9276077179747575e-06, + "loss": 0.0004, + "step": 23636 + }, + { + "epoch": 2.56255420641804, + "grad_norm": 0.04960288479924202, + "learning_rate": 2.92688234440737e-06, + "loss": 0.0002, + "step": 23637 + }, + { + "epoch": 2.5626626192541195, + "grad_norm": 7.35042667388916, + "learning_rate": 2.9261569708399827e-06, + "loss": 0.0198, + "step": 23638 + }, + { + "epoch": 2.5627710320901995, + "grad_norm": 0.024694429710507393, + "learning_rate": 2.9254315972725957e-06, + "loss": 0.0003, + "step": 23639 + }, + { + "epoch": 2.5628794449262795, + "grad_norm": 0.05596467852592468, + "learning_rate": 2.9247062237052082e-06, + "loss": 0.0004, + "step": 23640 + }, + { + "epoch": 2.562987857762359, + "grad_norm": 49.65784454345703, + "learning_rate": 2.9239808501378212e-06, + "loss": 0.0155, + "step": 23641 + }, + { + "epoch": 2.5630962705984386, + "grad_norm": 0.13137187063694, + "learning_rate": 2.9232554765704342e-06, + "loss": 0.0007, + "step": 23642 + }, + { + "epoch": 2.5632046834345186, + "grad_norm": 0.0880216509103775, + "learning_rate": 2.9225301030030468e-06, + "loss": 0.0006, + "step": 23643 + }, + { + "epoch": 2.5633130962705986, + "grad_norm": 1.775878667831421, + "learning_rate": 2.9218047294356594e-06, + "loss": 0.0059, + "step": 23644 + }, + { + "epoch": 2.563421509106678, + "grad_norm": 1.2177908420562744, + "learning_rate": 2.9210793558682728e-06, + "loss": 0.0068, + "step": 23645 + }, + { + "epoch": 2.563529921942758, + "grad_norm": 0.44038841128349304, + "learning_rate": 2.9203539823008853e-06, + "loss": 0.0009, + "step": 23646 + }, + { + "epoch": 2.5636383347788376, + "grad_norm": 3.758509397506714, + "learning_rate": 2.919628608733498e-06, + "loss": 0.0076, + "step": 23647 + }, + { + "epoch": 2.5637467476149176, + "grad_norm": 1.1463488340377808, + "learning_rate": 2.9189032351661105e-06, + "loss": 0.0032, + "step": 23648 + }, + { + "epoch": 2.5638551604509976, + "grad_norm": 0.01671917736530304, + "learning_rate": 2.918177861598724e-06, + "loss": 0.0001, + "step": 23649 + }, + { + "epoch": 2.563963573287077, + "grad_norm": 0.010425742715597153, + "learning_rate": 2.9174524880313364e-06, + "loss": 0.0001, + "step": 23650 + }, + { + "epoch": 2.564071986123157, + "grad_norm": 0.0256031546741724, + "learning_rate": 2.916727114463949e-06, + "loss": 0.0004, + "step": 23651 + }, + { + "epoch": 2.5641803989592367, + "grad_norm": 8.891314506530762, + "learning_rate": 2.916001740896562e-06, + "loss": 0.0177, + "step": 23652 + }, + { + "epoch": 2.5642888117953166, + "grad_norm": 2.2113118171691895, + "learning_rate": 2.9152763673291746e-06, + "loss": 0.0081, + "step": 23653 + }, + { + "epoch": 2.5643972246313966, + "grad_norm": 0.09560073167085648, + "learning_rate": 2.9145509937617876e-06, + "loss": 0.0003, + "step": 23654 + }, + { + "epoch": 2.564505637467476, + "grad_norm": 2.2217984199523926, + "learning_rate": 2.9138256201944006e-06, + "loss": 0.0016, + "step": 23655 + }, + { + "epoch": 2.5646140503035557, + "grad_norm": 0.06696271896362305, + "learning_rate": 2.913100246627013e-06, + "loss": 0.0005, + "step": 23656 + }, + { + "epoch": 2.5647224631396357, + "grad_norm": 0.04375347122550011, + "learning_rate": 2.9123748730596257e-06, + "loss": 0.0006, + "step": 23657 + }, + { + "epoch": 2.5648308759757157, + "grad_norm": 0.08104395121335983, + "learning_rate": 2.911649499492239e-06, + "loss": 0.0007, + "step": 23658 + }, + { + "epoch": 2.5649392888117952, + "grad_norm": 0.4080607295036316, + "learning_rate": 2.9109241259248517e-06, + "loss": 0.0019, + "step": 23659 + }, + { + "epoch": 2.565047701647875, + "grad_norm": 0.1853242814540863, + "learning_rate": 2.9101987523574642e-06, + "loss": 0.0005, + "step": 23660 + }, + { + "epoch": 2.5651561144839548, + "grad_norm": 0.2846909463405609, + "learning_rate": 2.909473378790077e-06, + "loss": 0.0016, + "step": 23661 + }, + { + "epoch": 2.5652645273200347, + "grad_norm": 2.585055351257324, + "learning_rate": 2.9087480052226902e-06, + "loss": 0.0108, + "step": 23662 + }, + { + "epoch": 2.5653729401561147, + "grad_norm": 0.6035091280937195, + "learning_rate": 2.9080226316553028e-06, + "loss": 0.0024, + "step": 23663 + }, + { + "epoch": 2.5654813529921943, + "grad_norm": 0.39002329111099243, + "learning_rate": 2.9072972580879153e-06, + "loss": 0.0007, + "step": 23664 + }, + { + "epoch": 2.565589765828274, + "grad_norm": 0.12031260132789612, + "learning_rate": 2.9065718845205283e-06, + "loss": 0.0008, + "step": 23665 + }, + { + "epoch": 2.565698178664354, + "grad_norm": 7.7206315994262695, + "learning_rate": 2.905846510953141e-06, + "loss": 0.0933, + "step": 23666 + }, + { + "epoch": 2.5658065915004338, + "grad_norm": 0.453102171421051, + "learning_rate": 2.905121137385754e-06, + "loss": 0.0005, + "step": 23667 + }, + { + "epoch": 2.5659150043365133, + "grad_norm": 0.9698959589004517, + "learning_rate": 2.904395763818367e-06, + "loss": 0.0017, + "step": 23668 + }, + { + "epoch": 2.5660234171725933, + "grad_norm": 4.062713623046875, + "learning_rate": 2.9036703902509795e-06, + "loss": 0.0161, + "step": 23669 + }, + { + "epoch": 2.566131830008673, + "grad_norm": 0.21829363703727722, + "learning_rate": 2.902945016683592e-06, + "loss": 0.0005, + "step": 23670 + }, + { + "epoch": 2.566240242844753, + "grad_norm": 0.2741905748844147, + "learning_rate": 2.9022196431162054e-06, + "loss": 0.0007, + "step": 23671 + }, + { + "epoch": 2.566348655680833, + "grad_norm": 0.8809848427772522, + "learning_rate": 2.901494269548818e-06, + "loss": 0.0037, + "step": 23672 + }, + { + "epoch": 2.5664570685169124, + "grad_norm": 2.7237977981567383, + "learning_rate": 2.9007688959814306e-06, + "loss": 0.0025, + "step": 23673 + }, + { + "epoch": 2.5665654813529923, + "grad_norm": 0.01738269440829754, + "learning_rate": 2.900043522414043e-06, + "loss": 0.0001, + "step": 23674 + }, + { + "epoch": 2.566673894189072, + "grad_norm": 0.8596656918525696, + "learning_rate": 2.8993181488466565e-06, + "loss": 0.0017, + "step": 23675 + }, + { + "epoch": 2.566782307025152, + "grad_norm": 4.269963264465332, + "learning_rate": 2.898592775279269e-06, + "loss": 0.0079, + "step": 23676 + }, + { + "epoch": 2.5668907198612314, + "grad_norm": 19.35723114013672, + "learning_rate": 2.8978674017118817e-06, + "loss": 0.0556, + "step": 23677 + }, + { + "epoch": 2.5669991326973114, + "grad_norm": 5.0761942863464355, + "learning_rate": 2.8971420281444947e-06, + "loss": 0.05, + "step": 23678 + }, + { + "epoch": 2.567107545533391, + "grad_norm": 0.008276750333607197, + "learning_rate": 2.8964166545771072e-06, + "loss": 0.0001, + "step": 23679 + }, + { + "epoch": 2.567215958369471, + "grad_norm": 9.70030689239502, + "learning_rate": 2.8956912810097202e-06, + "loss": 0.0237, + "step": 23680 + }, + { + "epoch": 2.567324371205551, + "grad_norm": 2.5608158111572266, + "learning_rate": 2.8949659074423332e-06, + "loss": 0.0086, + "step": 23681 + }, + { + "epoch": 2.5674327840416304, + "grad_norm": 3.672759532928467, + "learning_rate": 2.894240533874946e-06, + "loss": 0.0198, + "step": 23682 + }, + { + "epoch": 2.5675411968777104, + "grad_norm": 0.01184969674795866, + "learning_rate": 2.8935151603075584e-06, + "loss": 0.0001, + "step": 23683 + }, + { + "epoch": 2.56764960971379, + "grad_norm": 0.17534543573856354, + "learning_rate": 2.8927897867401718e-06, + "loss": 0.0004, + "step": 23684 + }, + { + "epoch": 2.56775802254987, + "grad_norm": 2.500976085662842, + "learning_rate": 2.8920644131727843e-06, + "loss": 0.0163, + "step": 23685 + }, + { + "epoch": 2.56786643538595, + "grad_norm": 0.029407355934381485, + "learning_rate": 2.891339039605397e-06, + "loss": 0.0002, + "step": 23686 + }, + { + "epoch": 2.5679748482220295, + "grad_norm": 0.0289187952876091, + "learning_rate": 2.8906136660380095e-06, + "loss": 0.0002, + "step": 23687 + }, + { + "epoch": 2.568083261058109, + "grad_norm": 1.070989966392517, + "learning_rate": 2.889888292470623e-06, + "loss": 0.0054, + "step": 23688 + }, + { + "epoch": 2.568191673894189, + "grad_norm": 0.013753487728536129, + "learning_rate": 2.8891629189032355e-06, + "loss": 0.0001, + "step": 23689 + }, + { + "epoch": 2.568300086730269, + "grad_norm": 0.28833436965942383, + "learning_rate": 2.888437545335848e-06, + "loss": 0.0013, + "step": 23690 + }, + { + "epoch": 2.5684084995663485, + "grad_norm": 0.4785486161708832, + "learning_rate": 2.887712171768461e-06, + "loss": 0.0008, + "step": 23691 + }, + { + "epoch": 2.5685169124024285, + "grad_norm": 1.8177447319030762, + "learning_rate": 2.8869867982010736e-06, + "loss": 0.0553, + "step": 23692 + }, + { + "epoch": 2.568625325238508, + "grad_norm": 0.33297526836395264, + "learning_rate": 2.8862614246336866e-06, + "loss": 0.002, + "step": 23693 + }, + { + "epoch": 2.568733738074588, + "grad_norm": 4.895390033721924, + "learning_rate": 2.8855360510662996e-06, + "loss": 0.0289, + "step": 23694 + }, + { + "epoch": 2.568842150910668, + "grad_norm": 3.871514081954956, + "learning_rate": 2.884810677498912e-06, + "loss": 0.0296, + "step": 23695 + }, + { + "epoch": 2.5689505637467476, + "grad_norm": 0.055997833609580994, + "learning_rate": 2.8840853039315247e-06, + "loss": 0.0002, + "step": 23696 + }, + { + "epoch": 2.5690589765828276, + "grad_norm": 0.129867821931839, + "learning_rate": 2.883359930364138e-06, + "loss": 0.0005, + "step": 23697 + }, + { + "epoch": 2.569167389418907, + "grad_norm": 0.8837271332740784, + "learning_rate": 2.8826345567967507e-06, + "loss": 0.0022, + "step": 23698 + }, + { + "epoch": 2.569275802254987, + "grad_norm": 0.1769833266735077, + "learning_rate": 2.8819091832293632e-06, + "loss": 0.0007, + "step": 23699 + }, + { + "epoch": 2.5693842150910666, + "grad_norm": 1.3383535146713257, + "learning_rate": 2.881183809661976e-06, + "loss": 0.0378, + "step": 23700 + }, + { + "epoch": 2.5694926279271466, + "grad_norm": 0.023308267816901207, + "learning_rate": 2.8804584360945892e-06, + "loss": 0.0002, + "step": 23701 + }, + { + "epoch": 2.569601040763226, + "grad_norm": 0.4286562502384186, + "learning_rate": 2.8797330625272018e-06, + "loss": 0.0005, + "step": 23702 + }, + { + "epoch": 2.569709453599306, + "grad_norm": 0.013385608792304993, + "learning_rate": 2.8790076889598144e-06, + "loss": 0.0001, + "step": 23703 + }, + { + "epoch": 2.569817866435386, + "grad_norm": 0.10913419723510742, + "learning_rate": 2.8782823153924273e-06, + "loss": 0.0004, + "step": 23704 + }, + { + "epoch": 2.5699262792714657, + "grad_norm": 6.4225616455078125, + "learning_rate": 2.87755694182504e-06, + "loss": 0.0173, + "step": 23705 + }, + { + "epoch": 2.5700346921075456, + "grad_norm": 0.003665217896923423, + "learning_rate": 2.876831568257653e-06, + "loss": 0.0, + "step": 23706 + }, + { + "epoch": 2.570143104943625, + "grad_norm": 1.2539703845977783, + "learning_rate": 2.876106194690266e-06, + "loss": 0.0026, + "step": 23707 + }, + { + "epoch": 2.570251517779705, + "grad_norm": 22.164596557617188, + "learning_rate": 2.8753808211228785e-06, + "loss": 0.0086, + "step": 23708 + }, + { + "epoch": 2.570359930615785, + "grad_norm": 0.01857280358672142, + "learning_rate": 2.874655447555491e-06, + "loss": 0.0002, + "step": 23709 + }, + { + "epoch": 2.5704683434518647, + "grad_norm": 0.0434427447617054, + "learning_rate": 2.8739300739881044e-06, + "loss": 0.0002, + "step": 23710 + }, + { + "epoch": 2.5705767562879442, + "grad_norm": 0.09223911911249161, + "learning_rate": 2.873204700420717e-06, + "loss": 0.0003, + "step": 23711 + }, + { + "epoch": 2.5706851691240242, + "grad_norm": 0.1408899873495102, + "learning_rate": 2.8724793268533296e-06, + "loss": 0.0008, + "step": 23712 + }, + { + "epoch": 2.570793581960104, + "grad_norm": 2.8646905422210693, + "learning_rate": 2.871753953285942e-06, + "loss": 0.0403, + "step": 23713 + }, + { + "epoch": 2.5709019947961838, + "grad_norm": 9.845159530639648, + "learning_rate": 2.8710285797185556e-06, + "loss": 0.004, + "step": 23714 + }, + { + "epoch": 2.5710104076322637, + "grad_norm": 0.2902294397354126, + "learning_rate": 2.870303206151168e-06, + "loss": 0.0008, + "step": 23715 + }, + { + "epoch": 2.5711188204683433, + "grad_norm": 0.2919647693634033, + "learning_rate": 2.8695778325837807e-06, + "loss": 0.001, + "step": 23716 + }, + { + "epoch": 2.5712272333044233, + "grad_norm": 0.07486744970083237, + "learning_rate": 2.8688524590163937e-06, + "loss": 0.0005, + "step": 23717 + }, + { + "epoch": 2.5713356461405033, + "grad_norm": 0.018127305433154106, + "learning_rate": 2.8681270854490062e-06, + "loss": 0.0002, + "step": 23718 + }, + { + "epoch": 2.571444058976583, + "grad_norm": 0.038774725049734116, + "learning_rate": 2.8674017118816192e-06, + "loss": 0.0004, + "step": 23719 + }, + { + "epoch": 2.5715524718126628, + "grad_norm": 0.05470125004649162, + "learning_rate": 2.8666763383142322e-06, + "loss": 0.0003, + "step": 23720 + }, + { + "epoch": 2.5716608846487423, + "grad_norm": 0.028598841279745102, + "learning_rate": 2.865950964746845e-06, + "loss": 0.0002, + "step": 23721 + }, + { + "epoch": 2.5717692974848223, + "grad_norm": 7.336206436157227, + "learning_rate": 2.8652255911794574e-06, + "loss": 0.0461, + "step": 23722 + }, + { + "epoch": 2.571877710320902, + "grad_norm": 0.04151669517159462, + "learning_rate": 2.8645002176120708e-06, + "loss": 0.0002, + "step": 23723 + }, + { + "epoch": 2.571986123156982, + "grad_norm": 0.27282285690307617, + "learning_rate": 2.8637748440446833e-06, + "loss": 0.0007, + "step": 23724 + }, + { + "epoch": 2.5720945359930614, + "grad_norm": 0.01346192043274641, + "learning_rate": 2.863049470477296e-06, + "loss": 0.0001, + "step": 23725 + }, + { + "epoch": 2.5722029488291414, + "grad_norm": 3.5295021533966064, + "learning_rate": 2.8623240969099085e-06, + "loss": 0.0625, + "step": 23726 + }, + { + "epoch": 2.5723113616652213, + "grad_norm": 0.016902640461921692, + "learning_rate": 2.861598723342522e-06, + "loss": 0.0002, + "step": 23727 + }, + { + "epoch": 2.572419774501301, + "grad_norm": 0.8923588395118713, + "learning_rate": 2.8608733497751345e-06, + "loss": 0.0028, + "step": 23728 + }, + { + "epoch": 2.572528187337381, + "grad_norm": 5.230202674865723, + "learning_rate": 2.860147976207747e-06, + "loss": 0.0212, + "step": 23729 + }, + { + "epoch": 2.5726366001734604, + "grad_norm": 4.217369556427002, + "learning_rate": 2.85942260264036e-06, + "loss": 0.0208, + "step": 23730 + }, + { + "epoch": 2.5727450130095404, + "grad_norm": 0.009319179691374302, + "learning_rate": 2.858697229072973e-06, + "loss": 0.0001, + "step": 23731 + }, + { + "epoch": 2.5728534258456204, + "grad_norm": 1.4758760929107666, + "learning_rate": 2.8579718555055856e-06, + "loss": 0.0084, + "step": 23732 + }, + { + "epoch": 2.5729618386817, + "grad_norm": 4.772251605987549, + "learning_rate": 2.8572464819381986e-06, + "loss": 0.075, + "step": 23733 + }, + { + "epoch": 2.5730702515177795, + "grad_norm": 1.2782697677612305, + "learning_rate": 2.856521108370811e-06, + "loss": 0.031, + "step": 23734 + }, + { + "epoch": 2.5731786643538594, + "grad_norm": 0.08795052766799927, + "learning_rate": 2.8557957348034237e-06, + "loss": 0.0005, + "step": 23735 + }, + { + "epoch": 2.5732870771899394, + "grad_norm": 2.5402677059173584, + "learning_rate": 2.855070361236037e-06, + "loss": 0.0084, + "step": 23736 + }, + { + "epoch": 2.573395490026019, + "grad_norm": 10.173088073730469, + "learning_rate": 2.8543449876686497e-06, + "loss": 0.0119, + "step": 23737 + }, + { + "epoch": 2.573503902862099, + "grad_norm": 1.0693966150283813, + "learning_rate": 2.8536196141012622e-06, + "loss": 0.0014, + "step": 23738 + }, + { + "epoch": 2.5736123156981785, + "grad_norm": 0.16069738566875458, + "learning_rate": 2.852894240533875e-06, + "loss": 0.0015, + "step": 23739 + }, + { + "epoch": 2.5737207285342585, + "grad_norm": 0.27185988426208496, + "learning_rate": 2.8521688669664882e-06, + "loss": 0.0011, + "step": 23740 + }, + { + "epoch": 2.5738291413703385, + "grad_norm": 1.7444727420806885, + "learning_rate": 2.851443493399101e-06, + "loss": 0.0151, + "step": 23741 + }, + { + "epoch": 2.573937554206418, + "grad_norm": 0.12421026080846786, + "learning_rate": 2.8507181198317134e-06, + "loss": 0.0005, + "step": 23742 + }, + { + "epoch": 2.574045967042498, + "grad_norm": 2.741037607192993, + "learning_rate": 2.8499927462643263e-06, + "loss": 0.0287, + "step": 23743 + }, + { + "epoch": 2.5741543798785775, + "grad_norm": 5.582421779632568, + "learning_rate": 2.8492673726969393e-06, + "loss": 0.029, + "step": 23744 + }, + { + "epoch": 2.5742627927146575, + "grad_norm": 0.3879126012325287, + "learning_rate": 2.848541999129552e-06, + "loss": 0.0025, + "step": 23745 + }, + { + "epoch": 2.574371205550737, + "grad_norm": 4.654974460601807, + "learning_rate": 2.847816625562165e-06, + "loss": 0.0206, + "step": 23746 + }, + { + "epoch": 2.574479618386817, + "grad_norm": 1.2435266971588135, + "learning_rate": 2.8470912519947775e-06, + "loss": 0.0032, + "step": 23747 + }, + { + "epoch": 2.5745880312228966, + "grad_norm": 0.07347574084997177, + "learning_rate": 2.84636587842739e-06, + "loss": 0.0005, + "step": 23748 + }, + { + "epoch": 2.5746964440589766, + "grad_norm": 2.2898993492126465, + "learning_rate": 2.8456405048600034e-06, + "loss": 0.0321, + "step": 23749 + }, + { + "epoch": 2.5748048568950566, + "grad_norm": 2.477106809616089, + "learning_rate": 2.844915131292616e-06, + "loss": 0.0048, + "step": 23750 + }, + { + "epoch": 2.574913269731136, + "grad_norm": 0.44571229815483093, + "learning_rate": 2.8441897577252286e-06, + "loss": 0.0012, + "step": 23751 + }, + { + "epoch": 2.575021682567216, + "grad_norm": 3.7847042083740234, + "learning_rate": 2.843464384157841e-06, + "loss": 0.0277, + "step": 23752 + }, + { + "epoch": 2.5751300954032956, + "grad_norm": 3.6512839794158936, + "learning_rate": 2.8427390105904546e-06, + "loss": 0.0343, + "step": 23753 + }, + { + "epoch": 2.5752385082393756, + "grad_norm": 0.03292321041226387, + "learning_rate": 2.842013637023067e-06, + "loss": 0.0003, + "step": 23754 + }, + { + "epoch": 2.5753469210754556, + "grad_norm": 0.09400726854801178, + "learning_rate": 2.8412882634556797e-06, + "loss": 0.0005, + "step": 23755 + }, + { + "epoch": 2.575455333911535, + "grad_norm": 10.161081314086914, + "learning_rate": 2.8405628898882927e-06, + "loss": 0.042, + "step": 23756 + }, + { + "epoch": 2.5755637467476147, + "grad_norm": 6.1193528175354, + "learning_rate": 2.8398375163209057e-06, + "loss": 0.0176, + "step": 23757 + }, + { + "epoch": 2.5756721595836947, + "grad_norm": 0.953641414642334, + "learning_rate": 2.8391121427535182e-06, + "loss": 0.0041, + "step": 23758 + }, + { + "epoch": 2.5757805724197746, + "grad_norm": 4.333050727844238, + "learning_rate": 2.8383867691861312e-06, + "loss": 0.0021, + "step": 23759 + }, + { + "epoch": 2.575888985255854, + "grad_norm": 0.6635342836380005, + "learning_rate": 2.837661395618744e-06, + "loss": 0.0026, + "step": 23760 + }, + { + "epoch": 2.575997398091934, + "grad_norm": 1.5146733522415161, + "learning_rate": 2.8369360220513564e-06, + "loss": 0.0061, + "step": 23761 + }, + { + "epoch": 2.5761058109280137, + "grad_norm": 0.020865580067038536, + "learning_rate": 2.8362106484839698e-06, + "loss": 0.0002, + "step": 23762 + }, + { + "epoch": 2.5762142237640937, + "grad_norm": 0.027127623558044434, + "learning_rate": 2.8354852749165823e-06, + "loss": 0.0003, + "step": 23763 + }, + { + "epoch": 2.5763226366001737, + "grad_norm": 1.8784887790679932, + "learning_rate": 2.834759901349195e-06, + "loss": 0.0252, + "step": 23764 + }, + { + "epoch": 2.5764310494362532, + "grad_norm": 0.21516183018684387, + "learning_rate": 2.8340345277818075e-06, + "loss": 0.0006, + "step": 23765 + }, + { + "epoch": 2.5765394622723328, + "grad_norm": 2.69681978225708, + "learning_rate": 2.833309154214421e-06, + "loss": 0.0466, + "step": 23766 + }, + { + "epoch": 2.5766478751084128, + "grad_norm": 0.8378734588623047, + "learning_rate": 2.8325837806470335e-06, + "loss": 0.0039, + "step": 23767 + }, + { + "epoch": 2.5767562879444927, + "grad_norm": 0.05837245658040047, + "learning_rate": 2.831858407079646e-06, + "loss": 0.0005, + "step": 23768 + }, + { + "epoch": 2.5768647007805723, + "grad_norm": 0.18698802590370178, + "learning_rate": 2.831133033512259e-06, + "loss": 0.0007, + "step": 23769 + }, + { + "epoch": 2.5769731136166523, + "grad_norm": 0.6525990962982178, + "learning_rate": 2.830407659944872e-06, + "loss": 0.0031, + "step": 23770 + }, + { + "epoch": 2.577081526452732, + "grad_norm": 0.08185398578643799, + "learning_rate": 2.8296822863774846e-06, + "loss": 0.0007, + "step": 23771 + }, + { + "epoch": 2.577189939288812, + "grad_norm": 1.5384007692337036, + "learning_rate": 2.8289569128100976e-06, + "loss": 0.0647, + "step": 23772 + }, + { + "epoch": 2.577298352124892, + "grad_norm": 0.38330957293510437, + "learning_rate": 2.82823153924271e-06, + "loss": 0.0018, + "step": 23773 + }, + { + "epoch": 2.5774067649609713, + "grad_norm": 12.305401802062988, + "learning_rate": 2.8275061656753227e-06, + "loss": 0.045, + "step": 23774 + }, + { + "epoch": 2.5775151777970513, + "grad_norm": 8.065171241760254, + "learning_rate": 2.826780792107936e-06, + "loss": 0.0141, + "step": 23775 + }, + { + "epoch": 2.577623590633131, + "grad_norm": 0.46755585074424744, + "learning_rate": 2.8260554185405487e-06, + "loss": 0.0019, + "step": 23776 + }, + { + "epoch": 2.577732003469211, + "grad_norm": 0.560376763343811, + "learning_rate": 2.8253300449731612e-06, + "loss": 0.0017, + "step": 23777 + }, + { + "epoch": 2.577840416305291, + "grad_norm": 0.10968054085969925, + "learning_rate": 2.824604671405774e-06, + "loss": 0.0007, + "step": 23778 + }, + { + "epoch": 2.5779488291413704, + "grad_norm": 0.11422789096832275, + "learning_rate": 2.8238792978383872e-06, + "loss": 0.0005, + "step": 23779 + }, + { + "epoch": 2.57805724197745, + "grad_norm": 1.8527861833572388, + "learning_rate": 2.823153924271e-06, + "loss": 0.0014, + "step": 23780 + }, + { + "epoch": 2.57816565481353, + "grad_norm": 1.6509370803833008, + "learning_rate": 2.8224285507036124e-06, + "loss": 0.0021, + "step": 23781 + }, + { + "epoch": 2.57827406764961, + "grad_norm": 0.6108525395393372, + "learning_rate": 2.8217031771362253e-06, + "loss": 0.0007, + "step": 23782 + }, + { + "epoch": 2.5783824804856894, + "grad_norm": 4.530231475830078, + "learning_rate": 2.8209778035688383e-06, + "loss": 0.0696, + "step": 23783 + }, + { + "epoch": 2.5784908933217694, + "grad_norm": 0.1025109589099884, + "learning_rate": 2.820252430001451e-06, + "loss": 0.0005, + "step": 23784 + }, + { + "epoch": 2.578599306157849, + "grad_norm": 9.38994026184082, + "learning_rate": 2.819527056434064e-06, + "loss": 0.0124, + "step": 23785 + }, + { + "epoch": 2.578707718993929, + "grad_norm": 2.2257394790649414, + "learning_rate": 2.8188016828666765e-06, + "loss": 0.0327, + "step": 23786 + }, + { + "epoch": 2.578816131830009, + "grad_norm": 1.005028247833252, + "learning_rate": 2.818076309299289e-06, + "loss": 0.0025, + "step": 23787 + }, + { + "epoch": 2.5789245446660884, + "grad_norm": 8.796512603759766, + "learning_rate": 2.8173509357319024e-06, + "loss": 0.0097, + "step": 23788 + }, + { + "epoch": 2.579032957502168, + "grad_norm": 0.02437015064060688, + "learning_rate": 2.816625562164515e-06, + "loss": 0.0002, + "step": 23789 + }, + { + "epoch": 2.579141370338248, + "grad_norm": 0.04330054670572281, + "learning_rate": 2.8159001885971276e-06, + "loss": 0.0004, + "step": 23790 + }, + { + "epoch": 2.579249783174328, + "grad_norm": 4.738725662231445, + "learning_rate": 2.81517481502974e-06, + "loss": 0.0473, + "step": 23791 + }, + { + "epoch": 2.5793581960104075, + "grad_norm": 0.8728928565979004, + "learning_rate": 2.8144494414623536e-06, + "loss": 0.0013, + "step": 23792 + }, + { + "epoch": 2.5794666088464875, + "grad_norm": 0.6750698089599609, + "learning_rate": 2.813724067894966e-06, + "loss": 0.0017, + "step": 23793 + }, + { + "epoch": 2.579575021682567, + "grad_norm": 3.3146543502807617, + "learning_rate": 2.8129986943275787e-06, + "loss": 0.0143, + "step": 23794 + }, + { + "epoch": 2.579683434518647, + "grad_norm": 14.209171295166016, + "learning_rate": 2.8122733207601917e-06, + "loss": 0.0565, + "step": 23795 + }, + { + "epoch": 2.579791847354727, + "grad_norm": 0.02002963423728943, + "learning_rate": 2.8115479471928047e-06, + "loss": 0.0002, + "step": 23796 + }, + { + "epoch": 2.5799002601908065, + "grad_norm": 0.17427565157413483, + "learning_rate": 2.8108225736254172e-06, + "loss": 0.0009, + "step": 23797 + }, + { + "epoch": 2.5800086730268865, + "grad_norm": 0.09016314893960953, + "learning_rate": 2.8100972000580302e-06, + "loss": 0.0005, + "step": 23798 + }, + { + "epoch": 2.580117085862966, + "grad_norm": 0.04228604584932327, + "learning_rate": 2.809371826490643e-06, + "loss": 0.0003, + "step": 23799 + }, + { + "epoch": 2.580225498699046, + "grad_norm": 3.020726203918457, + "learning_rate": 2.8086464529232554e-06, + "loss": 0.0137, + "step": 23800 + }, + { + "epoch": 2.580333911535126, + "grad_norm": 0.02847825549542904, + "learning_rate": 2.8079210793558688e-06, + "loss": 0.0002, + "step": 23801 + }, + { + "epoch": 2.5804423243712056, + "grad_norm": 3.5915825366973877, + "learning_rate": 2.8071957057884813e-06, + "loss": 0.033, + "step": 23802 + }, + { + "epoch": 2.580550737207285, + "grad_norm": 1.6700323820114136, + "learning_rate": 2.806470332221094e-06, + "loss": 0.0228, + "step": 23803 + }, + { + "epoch": 2.580659150043365, + "grad_norm": 0.012778690084815025, + "learning_rate": 2.8057449586537065e-06, + "loss": 0.0001, + "step": 23804 + }, + { + "epoch": 2.580767562879445, + "grad_norm": 0.15684844553470612, + "learning_rate": 2.80501958508632e-06, + "loss": 0.0005, + "step": 23805 + }, + { + "epoch": 2.5808759757155246, + "grad_norm": 0.057762786746025085, + "learning_rate": 2.8042942115189325e-06, + "loss": 0.0004, + "step": 23806 + }, + { + "epoch": 2.5809843885516046, + "grad_norm": 1.4604042768478394, + "learning_rate": 2.803568837951545e-06, + "loss": 0.0359, + "step": 23807 + }, + { + "epoch": 2.581092801387684, + "grad_norm": 0.01276471745222807, + "learning_rate": 2.802843464384158e-06, + "loss": 0.0001, + "step": 23808 + }, + { + "epoch": 2.581201214223764, + "grad_norm": 0.10309474170207977, + "learning_rate": 2.802118090816771e-06, + "loss": 0.0007, + "step": 23809 + }, + { + "epoch": 2.581309627059844, + "grad_norm": 0.0746648907661438, + "learning_rate": 2.8013927172493836e-06, + "loss": 0.0006, + "step": 23810 + }, + { + "epoch": 2.5814180398959237, + "grad_norm": 0.04219725728034973, + "learning_rate": 2.8006673436819966e-06, + "loss": 0.0003, + "step": 23811 + }, + { + "epoch": 2.581526452732003, + "grad_norm": 0.4307193160057068, + "learning_rate": 2.799941970114609e-06, + "loss": 0.0016, + "step": 23812 + }, + { + "epoch": 2.581634865568083, + "grad_norm": 3.217313528060913, + "learning_rate": 2.7992165965472217e-06, + "loss": 0.0225, + "step": 23813 + }, + { + "epoch": 2.581743278404163, + "grad_norm": 3.550842761993408, + "learning_rate": 2.798491222979835e-06, + "loss": 0.013, + "step": 23814 + }, + { + "epoch": 2.5818516912402427, + "grad_norm": 1.6132869720458984, + "learning_rate": 2.7977658494124477e-06, + "loss": 0.0219, + "step": 23815 + }, + { + "epoch": 2.5819601040763227, + "grad_norm": 0.057395342737436295, + "learning_rate": 2.7970404758450602e-06, + "loss": 0.0004, + "step": 23816 + }, + { + "epoch": 2.5820685169124022, + "grad_norm": 0.031531527638435364, + "learning_rate": 2.796315102277673e-06, + "loss": 0.0003, + "step": 23817 + }, + { + "epoch": 2.5821769297484822, + "grad_norm": 0.9605210423469543, + "learning_rate": 2.7955897287102862e-06, + "loss": 0.0034, + "step": 23818 + }, + { + "epoch": 2.582285342584562, + "grad_norm": 0.07201379537582397, + "learning_rate": 2.794864355142899e-06, + "loss": 0.001, + "step": 23819 + }, + { + "epoch": 2.5823937554206418, + "grad_norm": 0.05065497010946274, + "learning_rate": 2.7941389815755114e-06, + "loss": 0.0007, + "step": 23820 + }, + { + "epoch": 2.5825021682567217, + "grad_norm": 1.5748964548110962, + "learning_rate": 2.7934136080081244e-06, + "loss": 0.0012, + "step": 23821 + }, + { + "epoch": 2.5826105810928013, + "grad_norm": 1.8788524866104126, + "learning_rate": 2.7926882344407373e-06, + "loss": 0.0068, + "step": 23822 + }, + { + "epoch": 2.5827189939288813, + "grad_norm": 0.034481193870306015, + "learning_rate": 2.79196286087335e-06, + "loss": 0.0004, + "step": 23823 + }, + { + "epoch": 2.582827406764961, + "grad_norm": 1.849879503250122, + "learning_rate": 2.791237487305963e-06, + "loss": 0.0116, + "step": 23824 + }, + { + "epoch": 2.582935819601041, + "grad_norm": 6.649807929992676, + "learning_rate": 2.7905121137385755e-06, + "loss": 0.0188, + "step": 23825 + }, + { + "epoch": 2.5830442324371203, + "grad_norm": 0.7788732051849365, + "learning_rate": 2.7897867401711885e-06, + "loss": 0.0519, + "step": 23826 + }, + { + "epoch": 2.5831526452732003, + "grad_norm": 5.356814861297607, + "learning_rate": 2.7890613666038014e-06, + "loss": 0.0651, + "step": 23827 + }, + { + "epoch": 2.5832610581092803, + "grad_norm": 2.341240882873535, + "learning_rate": 2.788335993036414e-06, + "loss": 0.0047, + "step": 23828 + }, + { + "epoch": 2.58336947094536, + "grad_norm": 4.204939365386963, + "learning_rate": 2.7876106194690266e-06, + "loss": 0.0054, + "step": 23829 + }, + { + "epoch": 2.58347788378144, + "grad_norm": 3.956526041030884, + "learning_rate": 2.786885245901639e-06, + "loss": 0.0265, + "step": 23830 + }, + { + "epoch": 2.5835862966175194, + "grad_norm": 1.5557029247283936, + "learning_rate": 2.7861598723342526e-06, + "loss": 0.0018, + "step": 23831 + }, + { + "epoch": 2.5836947094535994, + "grad_norm": 0.5450884103775024, + "learning_rate": 2.785434498766865e-06, + "loss": 0.0041, + "step": 23832 + }, + { + "epoch": 2.5838031222896793, + "grad_norm": 1.2906420230865479, + "learning_rate": 2.7847091251994777e-06, + "loss": 0.0065, + "step": 23833 + }, + { + "epoch": 2.583911535125759, + "grad_norm": 0.9246016144752502, + "learning_rate": 2.7839837516320907e-06, + "loss": 0.0008, + "step": 23834 + }, + { + "epoch": 2.5840199479618384, + "grad_norm": 0.019006792455911636, + "learning_rate": 2.7832583780647037e-06, + "loss": 0.0002, + "step": 23835 + }, + { + "epoch": 2.5841283607979184, + "grad_norm": 0.024159163236618042, + "learning_rate": 2.7825330044973162e-06, + "loss": 0.0002, + "step": 23836 + }, + { + "epoch": 2.5842367736339984, + "grad_norm": 0.05391081050038338, + "learning_rate": 2.7818076309299292e-06, + "loss": 0.0002, + "step": 23837 + }, + { + "epoch": 2.584345186470078, + "grad_norm": 19.03917694091797, + "learning_rate": 2.781082257362542e-06, + "loss": 0.0182, + "step": 23838 + }, + { + "epoch": 2.584453599306158, + "grad_norm": 3.8572471141815186, + "learning_rate": 2.7803568837951552e-06, + "loss": 0.0123, + "step": 23839 + }, + { + "epoch": 2.5845620121422375, + "grad_norm": 1.3036335706710815, + "learning_rate": 2.7796315102277678e-06, + "loss": 0.0041, + "step": 23840 + }, + { + "epoch": 2.5846704249783174, + "grad_norm": 0.10814281553030014, + "learning_rate": 2.7789061366603803e-06, + "loss": 0.0011, + "step": 23841 + }, + { + "epoch": 2.5847788378143974, + "grad_norm": 0.9253572821617126, + "learning_rate": 2.778180763092993e-06, + "loss": 0.0011, + "step": 23842 + }, + { + "epoch": 2.584887250650477, + "grad_norm": 2.48758602142334, + "learning_rate": 2.7774553895256055e-06, + "loss": 0.01, + "step": 23843 + }, + { + "epoch": 2.584995663486557, + "grad_norm": 2.588076591491699, + "learning_rate": 2.776730015958219e-06, + "loss": 0.0068, + "step": 23844 + }, + { + "epoch": 2.5851040763226365, + "grad_norm": 0.15099658071994781, + "learning_rate": 2.7760046423908315e-06, + "loss": 0.0009, + "step": 23845 + }, + { + "epoch": 2.5852124891587165, + "grad_norm": 0.1817522794008255, + "learning_rate": 2.775279268823444e-06, + "loss": 0.0008, + "step": 23846 + }, + { + "epoch": 2.585320901994796, + "grad_norm": 1.2006336450576782, + "learning_rate": 2.774553895256057e-06, + "loss": 0.0055, + "step": 23847 + }, + { + "epoch": 2.585429314830876, + "grad_norm": 4.436214923858643, + "learning_rate": 2.77382852168867e-06, + "loss": 0.0085, + "step": 23848 + }, + { + "epoch": 2.5855377276669556, + "grad_norm": 0.05887038633227348, + "learning_rate": 2.7731031481212826e-06, + "loss": 0.0005, + "step": 23849 + }, + { + "epoch": 2.5856461405030355, + "grad_norm": 0.054508306086063385, + "learning_rate": 2.7723777745538956e-06, + "loss": 0.0004, + "step": 23850 + }, + { + "epoch": 2.5857545533391155, + "grad_norm": 0.09022326022386551, + "learning_rate": 2.771652400986508e-06, + "loss": 0.0006, + "step": 23851 + }, + { + "epoch": 2.585862966175195, + "grad_norm": 0.05254964530467987, + "learning_rate": 2.7709270274191215e-06, + "loss": 0.0004, + "step": 23852 + }, + { + "epoch": 2.585971379011275, + "grad_norm": 0.16800594329833984, + "learning_rate": 2.770201653851734e-06, + "loss": 0.0012, + "step": 23853 + }, + { + "epoch": 2.5860797918473546, + "grad_norm": 0.07116346806287766, + "learning_rate": 2.7694762802843467e-06, + "loss": 0.0004, + "step": 23854 + }, + { + "epoch": 2.5861882046834346, + "grad_norm": 0.6219594478607178, + "learning_rate": 2.7687509067169593e-06, + "loss": 0.001, + "step": 23855 + }, + { + "epoch": 2.5862966175195146, + "grad_norm": 3.6912522315979004, + "learning_rate": 2.768025533149572e-06, + "loss": 0.0042, + "step": 23856 + }, + { + "epoch": 2.586405030355594, + "grad_norm": 0.29122498631477356, + "learning_rate": 2.7673001595821852e-06, + "loss": 0.0014, + "step": 23857 + }, + { + "epoch": 2.5865134431916736, + "grad_norm": 3.9264628887176514, + "learning_rate": 2.766574786014798e-06, + "loss": 0.0188, + "step": 23858 + }, + { + "epoch": 2.5866218560277536, + "grad_norm": 3.1317477226257324, + "learning_rate": 2.7658494124474104e-06, + "loss": 0.0063, + "step": 23859 + }, + { + "epoch": 2.5867302688638336, + "grad_norm": 0.03232845291495323, + "learning_rate": 2.7651240388800234e-06, + "loss": 0.0004, + "step": 23860 + }, + { + "epoch": 2.586838681699913, + "grad_norm": 0.17100763320922852, + "learning_rate": 2.7643986653126363e-06, + "loss": 0.0007, + "step": 23861 + }, + { + "epoch": 2.586947094535993, + "grad_norm": 0.31789618730545044, + "learning_rate": 2.763673291745249e-06, + "loss": 0.0025, + "step": 23862 + }, + { + "epoch": 2.5870555073720727, + "grad_norm": 5.089783191680908, + "learning_rate": 2.762947918177862e-06, + "loss": 0.0354, + "step": 23863 + }, + { + "epoch": 2.5871639202081527, + "grad_norm": 0.892017662525177, + "learning_rate": 2.7622225446104745e-06, + "loss": 0.0014, + "step": 23864 + }, + { + "epoch": 2.5872723330442327, + "grad_norm": 0.6840763688087463, + "learning_rate": 2.761497171043088e-06, + "loss": 0.0024, + "step": 23865 + }, + { + "epoch": 2.587380745880312, + "grad_norm": 0.04733268916606903, + "learning_rate": 2.7607717974757005e-06, + "loss": 0.0004, + "step": 23866 + }, + { + "epoch": 2.587489158716392, + "grad_norm": 5.043186664581299, + "learning_rate": 2.760046423908313e-06, + "loss": 0.0336, + "step": 23867 + }, + { + "epoch": 2.5875975715524717, + "grad_norm": 0.030734596773982048, + "learning_rate": 2.7593210503409256e-06, + "loss": 0.0003, + "step": 23868 + }, + { + "epoch": 2.5877059843885517, + "grad_norm": 0.571975827217102, + "learning_rate": 2.758595676773538e-06, + "loss": 0.0011, + "step": 23869 + }, + { + "epoch": 2.5878143972246312, + "grad_norm": 0.45844584703445435, + "learning_rate": 2.7578703032061516e-06, + "loss": 0.0015, + "step": 23870 + }, + { + "epoch": 2.5879228100607112, + "grad_norm": 5.436036586761475, + "learning_rate": 2.757144929638764e-06, + "loss": 0.0087, + "step": 23871 + }, + { + "epoch": 2.5880312228967908, + "grad_norm": 0.08987163007259369, + "learning_rate": 2.7564195560713767e-06, + "loss": 0.0004, + "step": 23872 + }, + { + "epoch": 2.5881396357328708, + "grad_norm": 1.8030773401260376, + "learning_rate": 2.7556941825039897e-06, + "loss": 0.0018, + "step": 23873 + }, + { + "epoch": 2.5882480485689507, + "grad_norm": 0.47284525632858276, + "learning_rate": 2.7549688089366027e-06, + "loss": 0.0021, + "step": 23874 + }, + { + "epoch": 2.5883564614050303, + "grad_norm": 1.6451389789581299, + "learning_rate": 2.7542434353692152e-06, + "loss": 0.0237, + "step": 23875 + }, + { + "epoch": 2.5884648742411103, + "grad_norm": 1.371523141860962, + "learning_rate": 2.7535180618018282e-06, + "loss": 0.0287, + "step": 23876 + }, + { + "epoch": 2.58857328707719, + "grad_norm": 0.126088485121727, + "learning_rate": 2.752792688234441e-06, + "loss": 0.0002, + "step": 23877 + }, + { + "epoch": 2.58868169991327, + "grad_norm": 1.5494080781936646, + "learning_rate": 2.7520673146670542e-06, + "loss": 0.0477, + "step": 23878 + }, + { + "epoch": 2.58879011274935, + "grad_norm": 0.12197650969028473, + "learning_rate": 2.7513419410996668e-06, + "loss": 0.0009, + "step": 23879 + }, + { + "epoch": 2.5888985255854293, + "grad_norm": 0.29400351643562317, + "learning_rate": 2.7506165675322794e-06, + "loss": 0.0011, + "step": 23880 + }, + { + "epoch": 2.589006938421509, + "grad_norm": 0.44762203097343445, + "learning_rate": 2.749891193964892e-06, + "loss": 0.0015, + "step": 23881 + }, + { + "epoch": 2.589115351257589, + "grad_norm": 0.04354926943778992, + "learning_rate": 2.7491658203975045e-06, + "loss": 0.0003, + "step": 23882 + }, + { + "epoch": 2.589223764093669, + "grad_norm": 3.334141731262207, + "learning_rate": 2.748440446830118e-06, + "loss": 0.0043, + "step": 23883 + }, + { + "epoch": 2.5893321769297484, + "grad_norm": 5.139706134796143, + "learning_rate": 2.7477150732627305e-06, + "loss": 0.0205, + "step": 23884 + }, + { + "epoch": 2.5894405897658284, + "grad_norm": 1.0730654001235962, + "learning_rate": 2.746989699695343e-06, + "loss": 0.0031, + "step": 23885 + }, + { + "epoch": 2.589549002601908, + "grad_norm": 0.1136806383728981, + "learning_rate": 2.746264326127956e-06, + "loss": 0.0006, + "step": 23886 + }, + { + "epoch": 2.589657415437988, + "grad_norm": 0.16335919499397278, + "learning_rate": 2.745538952560569e-06, + "loss": 0.0009, + "step": 23887 + }, + { + "epoch": 2.589765828274068, + "grad_norm": 0.22467927634716034, + "learning_rate": 2.7448135789931816e-06, + "loss": 0.0009, + "step": 23888 + }, + { + "epoch": 2.5898742411101474, + "grad_norm": 0.012087891809642315, + "learning_rate": 2.7440882054257946e-06, + "loss": 0.0001, + "step": 23889 + }, + { + "epoch": 2.5899826539462274, + "grad_norm": 0.3557167947292328, + "learning_rate": 2.743362831858407e-06, + "loss": 0.0014, + "step": 23890 + }, + { + "epoch": 2.590091066782307, + "grad_norm": 0.017560157924890518, + "learning_rate": 2.7426374582910206e-06, + "loss": 0.0002, + "step": 23891 + }, + { + "epoch": 2.590199479618387, + "grad_norm": 12.023716926574707, + "learning_rate": 2.741912084723633e-06, + "loss": 0.0134, + "step": 23892 + }, + { + "epoch": 2.5903078924544665, + "grad_norm": 3.3235576152801514, + "learning_rate": 2.7411867111562457e-06, + "loss": 0.0205, + "step": 23893 + }, + { + "epoch": 2.5904163052905465, + "grad_norm": 2.4164092540740967, + "learning_rate": 2.7404613375888583e-06, + "loss": 0.0317, + "step": 23894 + }, + { + "epoch": 2.590524718126626, + "grad_norm": 9.914508819580078, + "learning_rate": 2.739735964021471e-06, + "loss": 0.0585, + "step": 23895 + }, + { + "epoch": 2.590633130962706, + "grad_norm": 0.22825807332992554, + "learning_rate": 2.7390105904540842e-06, + "loss": 0.0014, + "step": 23896 + }, + { + "epoch": 2.590741543798786, + "grad_norm": 4.607588291168213, + "learning_rate": 2.738285216886697e-06, + "loss": 0.0087, + "step": 23897 + }, + { + "epoch": 2.5908499566348655, + "grad_norm": 0.07453028112649918, + "learning_rate": 2.7375598433193094e-06, + "loss": 0.0006, + "step": 23898 + }, + { + "epoch": 2.5909583694709455, + "grad_norm": 0.8672386407852173, + "learning_rate": 2.7368344697519224e-06, + "loss": 0.0022, + "step": 23899 + }, + { + "epoch": 2.591066782307025, + "grad_norm": 9.563602447509766, + "learning_rate": 2.7361090961845353e-06, + "loss": 0.0112, + "step": 23900 + }, + { + "epoch": 2.591175195143105, + "grad_norm": 1.0471504926681519, + "learning_rate": 2.7353837226171483e-06, + "loss": 0.0029, + "step": 23901 + }, + { + "epoch": 2.591283607979185, + "grad_norm": 0.033372096717357635, + "learning_rate": 2.734658349049761e-06, + "loss": 0.0004, + "step": 23902 + }, + { + "epoch": 2.5913920208152645, + "grad_norm": 10.282169342041016, + "learning_rate": 2.7339329754823735e-06, + "loss": 0.0189, + "step": 23903 + }, + { + "epoch": 2.591500433651344, + "grad_norm": 2.889404296875, + "learning_rate": 2.733207601914987e-06, + "loss": 0.0328, + "step": 23904 + }, + { + "epoch": 2.591608846487424, + "grad_norm": 0.15155962109565735, + "learning_rate": 2.7324822283475995e-06, + "loss": 0.0011, + "step": 23905 + }, + { + "epoch": 2.591717259323504, + "grad_norm": 0.09774511307477951, + "learning_rate": 2.731756854780212e-06, + "loss": 0.0005, + "step": 23906 + }, + { + "epoch": 2.5918256721595836, + "grad_norm": 0.7147024869918823, + "learning_rate": 2.7310314812128246e-06, + "loss": 0.001, + "step": 23907 + }, + { + "epoch": 2.5919340849956636, + "grad_norm": 2.6990649700164795, + "learning_rate": 2.730306107645437e-06, + "loss": 0.0357, + "step": 23908 + }, + { + "epoch": 2.592042497831743, + "grad_norm": 3.1245357990264893, + "learning_rate": 2.7295807340780506e-06, + "loss": 0.0177, + "step": 23909 + }, + { + "epoch": 2.592150910667823, + "grad_norm": 7.096277713775635, + "learning_rate": 2.728855360510663e-06, + "loss": 0.0314, + "step": 23910 + }, + { + "epoch": 2.592259323503903, + "grad_norm": 0.038900550454854965, + "learning_rate": 2.7281299869432757e-06, + "loss": 0.0005, + "step": 23911 + }, + { + "epoch": 2.5923677363399826, + "grad_norm": 0.07902458310127258, + "learning_rate": 2.7274046133758887e-06, + "loss": 0.0008, + "step": 23912 + }, + { + "epoch": 2.592476149176062, + "grad_norm": 2.8172576427459717, + "learning_rate": 2.7266792398085017e-06, + "loss": 0.0077, + "step": 23913 + }, + { + "epoch": 2.592584562012142, + "grad_norm": 0.9367786645889282, + "learning_rate": 2.7259538662411147e-06, + "loss": 0.0028, + "step": 23914 + }, + { + "epoch": 2.592692974848222, + "grad_norm": 0.7311958074569702, + "learning_rate": 2.7252284926737272e-06, + "loss": 0.0026, + "step": 23915 + }, + { + "epoch": 2.5928013876843017, + "grad_norm": 0.08842200040817261, + "learning_rate": 2.72450311910634e-06, + "loss": 0.0005, + "step": 23916 + }, + { + "epoch": 2.5929098005203817, + "grad_norm": 0.10756010562181473, + "learning_rate": 2.7237777455389532e-06, + "loss": 0.0006, + "step": 23917 + }, + { + "epoch": 2.593018213356461, + "grad_norm": 1.7620635032653809, + "learning_rate": 2.723052371971566e-06, + "loss": 0.0048, + "step": 23918 + }, + { + "epoch": 2.593126626192541, + "grad_norm": 1.2730073928833008, + "learning_rate": 2.7223269984041784e-06, + "loss": 0.0305, + "step": 23919 + }, + { + "epoch": 2.593235039028621, + "grad_norm": 0.3601134717464447, + "learning_rate": 2.721601624836791e-06, + "loss": 0.0017, + "step": 23920 + }, + { + "epoch": 2.5933434518647007, + "grad_norm": 0.12690484523773193, + "learning_rate": 2.7208762512694035e-06, + "loss": 0.001, + "step": 23921 + }, + { + "epoch": 2.5934518647007807, + "grad_norm": 0.18997906148433685, + "learning_rate": 2.720150877702017e-06, + "loss": 0.0014, + "step": 23922 + }, + { + "epoch": 2.5935602775368602, + "grad_norm": 4.075896263122559, + "learning_rate": 2.7194255041346295e-06, + "loss": 0.0125, + "step": 23923 + }, + { + "epoch": 2.5936686903729402, + "grad_norm": 0.08288193494081497, + "learning_rate": 2.718700130567242e-06, + "loss": 0.0002, + "step": 23924 + }, + { + "epoch": 2.59377710320902, + "grad_norm": 2.120915651321411, + "learning_rate": 2.717974756999855e-06, + "loss": 0.0045, + "step": 23925 + }, + { + "epoch": 2.5938855160450998, + "grad_norm": 2.516986608505249, + "learning_rate": 2.717249383432468e-06, + "loss": 0.0085, + "step": 23926 + }, + { + "epoch": 2.5939939288811793, + "grad_norm": 0.013398692943155766, + "learning_rate": 2.716524009865081e-06, + "loss": 0.0001, + "step": 23927 + }, + { + "epoch": 2.5941023417172593, + "grad_norm": 0.9740664958953857, + "learning_rate": 2.7157986362976936e-06, + "loss": 0.0029, + "step": 23928 + }, + { + "epoch": 2.5942107545533393, + "grad_norm": 5.449305057525635, + "learning_rate": 2.715073262730306e-06, + "loss": 0.0309, + "step": 23929 + }, + { + "epoch": 2.594319167389419, + "grad_norm": 0.13318684697151184, + "learning_rate": 2.7143478891629196e-06, + "loss": 0.0007, + "step": 23930 + }, + { + "epoch": 2.594427580225499, + "grad_norm": 1.9682618379592896, + "learning_rate": 2.713622515595532e-06, + "loss": 0.0063, + "step": 23931 + }, + { + "epoch": 2.5945359930615783, + "grad_norm": 0.12928421795368195, + "learning_rate": 2.7128971420281447e-06, + "loss": 0.0006, + "step": 23932 + }, + { + "epoch": 2.5946444058976583, + "grad_norm": 0.0219286922365427, + "learning_rate": 2.7121717684607573e-06, + "loss": 0.0003, + "step": 23933 + }, + { + "epoch": 2.5947528187337383, + "grad_norm": 2.1797873973846436, + "learning_rate": 2.7114463948933707e-06, + "loss": 0.0201, + "step": 23934 + }, + { + "epoch": 2.594861231569818, + "grad_norm": 0.03885975480079651, + "learning_rate": 2.7107210213259832e-06, + "loss": 0.0003, + "step": 23935 + }, + { + "epoch": 2.5949696444058974, + "grad_norm": 0.13420888781547546, + "learning_rate": 2.709995647758596e-06, + "loss": 0.0006, + "step": 23936 + }, + { + "epoch": 2.5950780572419774, + "grad_norm": 1.6352622509002686, + "learning_rate": 2.7092702741912084e-06, + "loss": 0.0377, + "step": 23937 + }, + { + "epoch": 2.5951864700780574, + "grad_norm": 0.04779983311891556, + "learning_rate": 2.7085449006238214e-06, + "loss": 0.0004, + "step": 23938 + }, + { + "epoch": 2.595294882914137, + "grad_norm": 0.5907377600669861, + "learning_rate": 2.7078195270564344e-06, + "loss": 0.0019, + "step": 23939 + }, + { + "epoch": 2.595403295750217, + "grad_norm": 0.7467541098594666, + "learning_rate": 2.7070941534890473e-06, + "loss": 0.0046, + "step": 23940 + }, + { + "epoch": 2.5955117085862964, + "grad_norm": 0.8583745360374451, + "learning_rate": 2.70636877992166e-06, + "loss": 0.0026, + "step": 23941 + }, + { + "epoch": 2.5956201214223764, + "grad_norm": 0.09358161687850952, + "learning_rate": 2.7056434063542725e-06, + "loss": 0.0004, + "step": 23942 + }, + { + "epoch": 2.5957285342584564, + "grad_norm": 0.14885081350803375, + "learning_rate": 2.704918032786886e-06, + "loss": 0.0009, + "step": 23943 + }, + { + "epoch": 2.595836947094536, + "grad_norm": 0.6080542206764221, + "learning_rate": 2.7041926592194985e-06, + "loss": 0.0022, + "step": 23944 + }, + { + "epoch": 2.595945359930616, + "grad_norm": 1.1621938943862915, + "learning_rate": 2.703467285652111e-06, + "loss": 0.0056, + "step": 23945 + }, + { + "epoch": 2.5960537727666955, + "grad_norm": 0.11819062381982803, + "learning_rate": 2.7027419120847236e-06, + "loss": 0.0006, + "step": 23946 + }, + { + "epoch": 2.5961621856027755, + "grad_norm": 0.6728739142417908, + "learning_rate": 2.702016538517337e-06, + "loss": 0.0036, + "step": 23947 + }, + { + "epoch": 2.5962705984388554, + "grad_norm": 0.09962882101535797, + "learning_rate": 2.7012911649499496e-06, + "loss": 0.0008, + "step": 23948 + }, + { + "epoch": 2.596379011274935, + "grad_norm": 4.0096659660339355, + "learning_rate": 2.700565791382562e-06, + "loss": 0.0058, + "step": 23949 + }, + { + "epoch": 2.5964874241110145, + "grad_norm": 0.16518200933933258, + "learning_rate": 2.699840417815175e-06, + "loss": 0.0006, + "step": 23950 + }, + { + "epoch": 2.5965958369470945, + "grad_norm": 0.30354025959968567, + "learning_rate": 2.6991150442477877e-06, + "loss": 0.0011, + "step": 23951 + }, + { + "epoch": 2.5967042497831745, + "grad_norm": 4.522220134735107, + "learning_rate": 2.6983896706804007e-06, + "loss": 0.0073, + "step": 23952 + }, + { + "epoch": 2.596812662619254, + "grad_norm": 8.169692993164062, + "learning_rate": 2.6976642971130137e-06, + "loss": 0.0386, + "step": 23953 + }, + { + "epoch": 2.596921075455334, + "grad_norm": 2.5226078033447266, + "learning_rate": 2.6969389235456262e-06, + "loss": 0.0108, + "step": 23954 + }, + { + "epoch": 2.5970294882914136, + "grad_norm": 2.7423384189605713, + "learning_rate": 2.696213549978239e-06, + "loss": 0.0101, + "step": 23955 + }, + { + "epoch": 2.5971379011274935, + "grad_norm": 0.0763651430606842, + "learning_rate": 2.6954881764108522e-06, + "loss": 0.0007, + "step": 23956 + }, + { + "epoch": 2.5972463139635735, + "grad_norm": 1.701261281967163, + "learning_rate": 2.694762802843465e-06, + "loss": 0.0074, + "step": 23957 + }, + { + "epoch": 2.597354726799653, + "grad_norm": 2.2802183628082275, + "learning_rate": 2.6940374292760774e-06, + "loss": 0.0072, + "step": 23958 + }, + { + "epoch": 2.5974631396357326, + "grad_norm": 0.25496676564216614, + "learning_rate": 2.69331205570869e-06, + "loss": 0.0012, + "step": 23959 + }, + { + "epoch": 2.5975715524718126, + "grad_norm": 1.9567712545394897, + "learning_rate": 2.6925866821413033e-06, + "loss": 0.0053, + "step": 23960 + }, + { + "epoch": 2.5976799653078926, + "grad_norm": 0.1791684627532959, + "learning_rate": 2.691861308573916e-06, + "loss": 0.0015, + "step": 23961 + }, + { + "epoch": 2.597788378143972, + "grad_norm": 0.09133285284042358, + "learning_rate": 2.6911359350065285e-06, + "loss": 0.0007, + "step": 23962 + }, + { + "epoch": 2.597896790980052, + "grad_norm": 0.012852574698626995, + "learning_rate": 2.6904105614391415e-06, + "loss": 0.0001, + "step": 23963 + }, + { + "epoch": 2.5980052038161316, + "grad_norm": 0.25547218322753906, + "learning_rate": 2.689685187871754e-06, + "loss": 0.0009, + "step": 23964 + }, + { + "epoch": 2.5981136166522116, + "grad_norm": 0.9310212731361389, + "learning_rate": 2.688959814304367e-06, + "loss": 0.0461, + "step": 23965 + }, + { + "epoch": 2.5982220294882916, + "grad_norm": 1.886917233467102, + "learning_rate": 2.68823444073698e-06, + "loss": 0.0239, + "step": 23966 + }, + { + "epoch": 2.598330442324371, + "grad_norm": 1.9510436058044434, + "learning_rate": 2.6875090671695926e-06, + "loss": 0.0133, + "step": 23967 + }, + { + "epoch": 2.598438855160451, + "grad_norm": 2.2773172855377197, + "learning_rate": 2.686783693602205e-06, + "loss": 0.009, + "step": 23968 + }, + { + "epoch": 2.5985472679965307, + "grad_norm": 1.0711323022842407, + "learning_rate": 2.6860583200348186e-06, + "loss": 0.0033, + "step": 23969 + }, + { + "epoch": 2.5986556808326107, + "grad_norm": 4.605634689331055, + "learning_rate": 2.685332946467431e-06, + "loss": 0.0175, + "step": 23970 + }, + { + "epoch": 2.59876409366869, + "grad_norm": 1.8059237003326416, + "learning_rate": 2.6846075729000437e-06, + "loss": 0.0069, + "step": 23971 + }, + { + "epoch": 2.59887250650477, + "grad_norm": 0.31929898262023926, + "learning_rate": 2.6838821993326563e-06, + "loss": 0.001, + "step": 23972 + }, + { + "epoch": 2.5989809193408497, + "grad_norm": 0.023868028074502945, + "learning_rate": 2.6831568257652697e-06, + "loss": 0.0002, + "step": 23973 + }, + { + "epoch": 2.5990893321769297, + "grad_norm": 0.10634051263332367, + "learning_rate": 2.6824314521978822e-06, + "loss": 0.0007, + "step": 23974 + }, + { + "epoch": 2.5991977450130097, + "grad_norm": 0.5746660828590393, + "learning_rate": 2.681706078630495e-06, + "loss": 0.0021, + "step": 23975 + }, + { + "epoch": 2.5993061578490892, + "grad_norm": 0.1742907464504242, + "learning_rate": 2.680980705063108e-06, + "loss": 0.001, + "step": 23976 + }, + { + "epoch": 2.5994145706851692, + "grad_norm": 0.2728632390499115, + "learning_rate": 2.6802553314957204e-06, + "loss": 0.0005, + "step": 23977 + }, + { + "epoch": 2.5995229835212488, + "grad_norm": 0.9443195462226868, + "learning_rate": 2.6795299579283334e-06, + "loss": 0.0039, + "step": 23978 + }, + { + "epoch": 2.5996313963573288, + "grad_norm": 3.5904042720794678, + "learning_rate": 2.6788045843609463e-06, + "loss": 0.0091, + "step": 23979 + }, + { + "epoch": 2.5997398091934087, + "grad_norm": 0.653799831867218, + "learning_rate": 2.678079210793559e-06, + "loss": 0.0048, + "step": 23980 + }, + { + "epoch": 2.5998482220294883, + "grad_norm": 3.1735188961029053, + "learning_rate": 2.6773538372261715e-06, + "loss": 0.0447, + "step": 23981 + }, + { + "epoch": 2.599956634865568, + "grad_norm": 1.4088188409805298, + "learning_rate": 2.676628463658785e-06, + "loss": 0.0174, + "step": 23982 + }, + { + "epoch": 2.600065047701648, + "grad_norm": 3.230412006378174, + "learning_rate": 2.6759030900913975e-06, + "loss": 0.0066, + "step": 23983 + }, + { + "epoch": 2.600173460537728, + "grad_norm": 0.008086715824902058, + "learning_rate": 2.67517771652401e-06, + "loss": 0.0001, + "step": 23984 + }, + { + "epoch": 2.6002818733738073, + "grad_norm": 4.067490100860596, + "learning_rate": 2.6744523429566226e-06, + "loss": 0.0172, + "step": 23985 + }, + { + "epoch": 2.6003902862098873, + "grad_norm": 2.8709475994110107, + "learning_rate": 2.673726969389236e-06, + "loss": 0.0122, + "step": 23986 + }, + { + "epoch": 2.600498699045967, + "grad_norm": 0.20154862105846405, + "learning_rate": 2.6730015958218486e-06, + "loss": 0.0007, + "step": 23987 + }, + { + "epoch": 2.600607111882047, + "grad_norm": 3.5753767490386963, + "learning_rate": 2.672276222254461e-06, + "loss": 0.0118, + "step": 23988 + }, + { + "epoch": 2.600715524718127, + "grad_norm": 0.10164760053157806, + "learning_rate": 2.671550848687074e-06, + "loss": 0.0006, + "step": 23989 + }, + { + "epoch": 2.6008239375542064, + "grad_norm": 0.12079226970672607, + "learning_rate": 2.6708254751196867e-06, + "loss": 0.0007, + "step": 23990 + }, + { + "epoch": 2.6009323503902864, + "grad_norm": 5.4709553718566895, + "learning_rate": 2.6701001015522997e-06, + "loss": 0.025, + "step": 23991 + }, + { + "epoch": 2.601040763226366, + "grad_norm": 0.15843336284160614, + "learning_rate": 2.6693747279849127e-06, + "loss": 0.0003, + "step": 23992 + }, + { + "epoch": 2.601149176062446, + "grad_norm": 0.8547499775886536, + "learning_rate": 2.6686493544175252e-06, + "loss": 0.0039, + "step": 23993 + }, + { + "epoch": 2.6012575888985254, + "grad_norm": 3.462451934814453, + "learning_rate": 2.667923980850138e-06, + "loss": 0.0412, + "step": 23994 + }, + { + "epoch": 2.6013660017346054, + "grad_norm": 8.15174388885498, + "learning_rate": 2.6671986072827512e-06, + "loss": 0.0036, + "step": 23995 + }, + { + "epoch": 2.601474414570685, + "grad_norm": 5.42504358291626, + "learning_rate": 2.666473233715364e-06, + "loss": 0.0684, + "step": 23996 + }, + { + "epoch": 2.601582827406765, + "grad_norm": 0.029368532821536064, + "learning_rate": 2.6657478601479764e-06, + "loss": 0.0003, + "step": 23997 + }, + { + "epoch": 2.601691240242845, + "grad_norm": 0.5648863315582275, + "learning_rate": 2.665022486580589e-06, + "loss": 0.0014, + "step": 23998 + }, + { + "epoch": 2.6017996530789245, + "grad_norm": 3.100011110305786, + "learning_rate": 2.6642971130132023e-06, + "loss": 0.0133, + "step": 23999 + }, + { + "epoch": 2.6019080659150045, + "grad_norm": 0.24640367925167084, + "learning_rate": 2.663571739445815e-06, + "loss": 0.0005, + "step": 24000 + }, + { + "epoch": 2.602016478751084, + "grad_norm": 0.2306446135044098, + "learning_rate": 2.6628463658784275e-06, + "loss": 0.0011, + "step": 24001 + }, + { + "epoch": 2.602124891587164, + "grad_norm": 3.5175445079803467, + "learning_rate": 2.6621209923110405e-06, + "loss": 0.003, + "step": 24002 + }, + { + "epoch": 2.602233304423244, + "grad_norm": 0.06245701014995575, + "learning_rate": 2.661395618743653e-06, + "loss": 0.0004, + "step": 24003 + }, + { + "epoch": 2.6023417172593235, + "grad_norm": 0.05021801218390465, + "learning_rate": 2.660670245176266e-06, + "loss": 0.0002, + "step": 24004 + }, + { + "epoch": 2.602450130095403, + "grad_norm": 0.0470590814948082, + "learning_rate": 2.659944871608879e-06, + "loss": 0.0003, + "step": 24005 + }, + { + "epoch": 2.602558542931483, + "grad_norm": 0.13587084412574768, + "learning_rate": 2.6592194980414916e-06, + "loss": 0.0014, + "step": 24006 + }, + { + "epoch": 2.602666955767563, + "grad_norm": 0.18819788098335266, + "learning_rate": 2.658494124474104e-06, + "loss": 0.0011, + "step": 24007 + }, + { + "epoch": 2.6027753686036426, + "grad_norm": 0.29691290855407715, + "learning_rate": 2.6577687509067176e-06, + "loss": 0.0008, + "step": 24008 + }, + { + "epoch": 2.6028837814397225, + "grad_norm": 2.6009902954101562, + "learning_rate": 2.65704337733933e-06, + "loss": 0.0071, + "step": 24009 + }, + { + "epoch": 2.602992194275802, + "grad_norm": 0.6967536211013794, + "learning_rate": 2.6563180037719427e-06, + "loss": 0.002, + "step": 24010 + }, + { + "epoch": 2.603100607111882, + "grad_norm": 0.22795841097831726, + "learning_rate": 2.6555926302045553e-06, + "loss": 0.001, + "step": 24011 + }, + { + "epoch": 2.603209019947962, + "grad_norm": 1.1580069065093994, + "learning_rate": 2.6548672566371687e-06, + "loss": 0.0058, + "step": 24012 + }, + { + "epoch": 2.6033174327840416, + "grad_norm": 0.022788256406784058, + "learning_rate": 2.6541418830697812e-06, + "loss": 0.0003, + "step": 24013 + }, + { + "epoch": 2.6034258456201216, + "grad_norm": 0.03998308628797531, + "learning_rate": 2.653416509502394e-06, + "loss": 0.0004, + "step": 24014 + }, + { + "epoch": 2.603534258456201, + "grad_norm": 1.9002680778503418, + "learning_rate": 2.652691135935007e-06, + "loss": 0.0197, + "step": 24015 + }, + { + "epoch": 2.603642671292281, + "grad_norm": 0.050335902720689774, + "learning_rate": 2.6519657623676194e-06, + "loss": 0.0005, + "step": 24016 + }, + { + "epoch": 2.6037510841283606, + "grad_norm": 2.761840581893921, + "learning_rate": 2.6512403888002324e-06, + "loss": 0.0082, + "step": 24017 + }, + { + "epoch": 2.6038594969644406, + "grad_norm": 10.21642017364502, + "learning_rate": 2.6505150152328453e-06, + "loss": 0.0095, + "step": 24018 + }, + { + "epoch": 2.60396790980052, + "grad_norm": 0.045312732458114624, + "learning_rate": 2.649789641665458e-06, + "loss": 0.0004, + "step": 24019 + }, + { + "epoch": 2.6040763226366, + "grad_norm": 1.3055322170257568, + "learning_rate": 2.6490642680980705e-06, + "loss": 0.0048, + "step": 24020 + }, + { + "epoch": 2.60418473547268, + "grad_norm": 2.832524061203003, + "learning_rate": 2.648338894530684e-06, + "loss": 0.0037, + "step": 24021 + }, + { + "epoch": 2.6042931483087597, + "grad_norm": 5.034911155700684, + "learning_rate": 2.6476135209632965e-06, + "loss": 0.0201, + "step": 24022 + }, + { + "epoch": 2.6044015611448397, + "grad_norm": 0.04033897444605827, + "learning_rate": 2.646888147395909e-06, + "loss": 0.0004, + "step": 24023 + }, + { + "epoch": 2.604509973980919, + "grad_norm": 0.016239771619439125, + "learning_rate": 2.6461627738285216e-06, + "loss": 0.0002, + "step": 24024 + }, + { + "epoch": 2.604618386816999, + "grad_norm": 2.015606164932251, + "learning_rate": 2.645437400261135e-06, + "loss": 0.0036, + "step": 24025 + }, + { + "epoch": 2.604726799653079, + "grad_norm": 3.4899353981018066, + "learning_rate": 2.6447120266937476e-06, + "loss": 0.012, + "step": 24026 + }, + { + "epoch": 2.6048352124891587, + "grad_norm": 1.4332952499389648, + "learning_rate": 2.64398665312636e-06, + "loss": 0.0025, + "step": 24027 + }, + { + "epoch": 2.6049436253252383, + "grad_norm": 0.07579173147678375, + "learning_rate": 2.643261279558973e-06, + "loss": 0.0003, + "step": 24028 + }, + { + "epoch": 2.6050520381613183, + "grad_norm": 2.229739189147949, + "learning_rate": 2.642535905991586e-06, + "loss": 0.0212, + "step": 24029 + }, + { + "epoch": 2.6051604509973982, + "grad_norm": 2.9155213832855225, + "learning_rate": 2.6418105324241987e-06, + "loss": 0.0074, + "step": 24030 + }, + { + "epoch": 2.6052688638334778, + "grad_norm": 4.27819299697876, + "learning_rate": 2.6410851588568117e-06, + "loss": 0.0287, + "step": 24031 + }, + { + "epoch": 2.6053772766695578, + "grad_norm": 0.11667560040950775, + "learning_rate": 2.6403597852894243e-06, + "loss": 0.0006, + "step": 24032 + }, + { + "epoch": 2.6054856895056373, + "grad_norm": 0.3684682548046112, + "learning_rate": 2.639634411722037e-06, + "loss": 0.0014, + "step": 24033 + }, + { + "epoch": 2.6055941023417173, + "grad_norm": 3.7282612323760986, + "learning_rate": 2.6389090381546502e-06, + "loss": 0.0575, + "step": 24034 + }, + { + "epoch": 2.6057025151777973, + "grad_norm": 5.087857723236084, + "learning_rate": 2.638183664587263e-06, + "loss": 0.0279, + "step": 24035 + }, + { + "epoch": 2.605810928013877, + "grad_norm": 0.23633478581905365, + "learning_rate": 2.6374582910198754e-06, + "loss": 0.0006, + "step": 24036 + }, + { + "epoch": 2.605919340849957, + "grad_norm": 2.1325347423553467, + "learning_rate": 2.636732917452488e-06, + "loss": 0.0327, + "step": 24037 + }, + { + "epoch": 2.6060277536860363, + "grad_norm": 2.5613772869110107, + "learning_rate": 2.6360075438851013e-06, + "loss": 0.0061, + "step": 24038 + }, + { + "epoch": 2.6061361665221163, + "grad_norm": 0.05099768936634064, + "learning_rate": 2.635282170317714e-06, + "loss": 0.0006, + "step": 24039 + }, + { + "epoch": 2.606244579358196, + "grad_norm": 0.09637753665447235, + "learning_rate": 2.6345567967503265e-06, + "loss": 0.0007, + "step": 24040 + }, + { + "epoch": 2.606352992194276, + "grad_norm": 5.633955478668213, + "learning_rate": 2.6338314231829395e-06, + "loss": 0.0379, + "step": 24041 + }, + { + "epoch": 2.6064614050303554, + "grad_norm": 3.397038221359253, + "learning_rate": 2.6331060496155525e-06, + "loss": 0.0764, + "step": 24042 + }, + { + "epoch": 2.6065698178664354, + "grad_norm": 0.5066317915916443, + "learning_rate": 2.632380676048165e-06, + "loss": 0.0009, + "step": 24043 + }, + { + "epoch": 2.6066782307025154, + "grad_norm": 6.334993839263916, + "learning_rate": 2.631655302480778e-06, + "loss": 0.0101, + "step": 24044 + }, + { + "epoch": 2.606786643538595, + "grad_norm": 0.16497397422790527, + "learning_rate": 2.6309299289133906e-06, + "loss": 0.0005, + "step": 24045 + }, + { + "epoch": 2.606895056374675, + "grad_norm": 2.5264251232147217, + "learning_rate": 2.630204555346003e-06, + "loss": 0.0068, + "step": 24046 + }, + { + "epoch": 2.6070034692107544, + "grad_norm": 0.14382849633693695, + "learning_rate": 2.6294791817786166e-06, + "loss": 0.0004, + "step": 24047 + }, + { + "epoch": 2.6071118820468344, + "grad_norm": 0.5685508251190186, + "learning_rate": 2.628753808211229e-06, + "loss": 0.0021, + "step": 24048 + }, + { + "epoch": 2.6072202948829144, + "grad_norm": 25.046438217163086, + "learning_rate": 2.6280284346438417e-06, + "loss": 0.0112, + "step": 24049 + }, + { + "epoch": 2.607328707718994, + "grad_norm": 2.5949418544769287, + "learning_rate": 2.6273030610764543e-06, + "loss": 0.007, + "step": 24050 + }, + { + "epoch": 2.6074371205550735, + "grad_norm": 24.114225387573242, + "learning_rate": 2.6265776875090677e-06, + "loss": 0.0239, + "step": 24051 + }, + { + "epoch": 2.6075455333911535, + "grad_norm": 0.038538847118616104, + "learning_rate": 2.6258523139416802e-06, + "loss": 0.0004, + "step": 24052 + }, + { + "epoch": 2.6076539462272335, + "grad_norm": 0.0327209047973156, + "learning_rate": 2.625126940374293e-06, + "loss": 0.0003, + "step": 24053 + }, + { + "epoch": 2.607762359063313, + "grad_norm": 2.554835796356201, + "learning_rate": 2.624401566806906e-06, + "loss": 0.0103, + "step": 24054 + }, + { + "epoch": 2.607870771899393, + "grad_norm": 13.233697891235352, + "learning_rate": 2.623676193239519e-06, + "loss": 0.0268, + "step": 24055 + }, + { + "epoch": 2.6079791847354725, + "grad_norm": 0.9738522171974182, + "learning_rate": 2.6229508196721314e-06, + "loss": 0.0046, + "step": 24056 + }, + { + "epoch": 2.6080875975715525, + "grad_norm": 19.54837417602539, + "learning_rate": 2.6222254461047444e-06, + "loss": 0.0103, + "step": 24057 + }, + { + "epoch": 2.6081960104076325, + "grad_norm": 0.0761069506406784, + "learning_rate": 2.621500072537357e-06, + "loss": 0.0005, + "step": 24058 + }, + { + "epoch": 2.608304423243712, + "grad_norm": 0.05077681690454483, + "learning_rate": 2.6207746989699695e-06, + "loss": 0.0002, + "step": 24059 + }, + { + "epoch": 2.6084128360797916, + "grad_norm": 0.43857109546661377, + "learning_rate": 2.620049325402583e-06, + "loss": 0.0014, + "step": 24060 + }, + { + "epoch": 2.6085212489158716, + "grad_norm": 0.020451752468943596, + "learning_rate": 2.6193239518351955e-06, + "loss": 0.0002, + "step": 24061 + }, + { + "epoch": 2.6086296617519515, + "grad_norm": 7.650907039642334, + "learning_rate": 2.618598578267808e-06, + "loss": 0.0073, + "step": 24062 + }, + { + "epoch": 2.608738074588031, + "grad_norm": 0.06022392958402634, + "learning_rate": 2.6178732047004206e-06, + "loss": 0.0003, + "step": 24063 + }, + { + "epoch": 2.608846487424111, + "grad_norm": 0.1962122768163681, + "learning_rate": 2.617147831133034e-06, + "loss": 0.0009, + "step": 24064 + }, + { + "epoch": 2.6089549002601906, + "grad_norm": 13.185137748718262, + "learning_rate": 2.6164224575656466e-06, + "loss": 0.0077, + "step": 24065 + }, + { + "epoch": 2.6090633130962706, + "grad_norm": 0.2113422304391861, + "learning_rate": 2.615697083998259e-06, + "loss": 0.0009, + "step": 24066 + }, + { + "epoch": 2.6091717259323506, + "grad_norm": 0.04969540983438492, + "learning_rate": 2.614971710430872e-06, + "loss": 0.0005, + "step": 24067 + }, + { + "epoch": 2.60928013876843, + "grad_norm": 0.09457767009735107, + "learning_rate": 2.614246336863485e-06, + "loss": 0.0003, + "step": 24068 + }, + { + "epoch": 2.60938855160451, + "grad_norm": 6.518744945526123, + "learning_rate": 2.6135209632960977e-06, + "loss": 0.0206, + "step": 24069 + }, + { + "epoch": 2.6094969644405897, + "grad_norm": 0.03403053432703018, + "learning_rate": 2.6127955897287107e-06, + "loss": 0.0004, + "step": 24070 + }, + { + "epoch": 2.6096053772766696, + "grad_norm": 0.3840258717536926, + "learning_rate": 2.6120702161613233e-06, + "loss": 0.0014, + "step": 24071 + }, + { + "epoch": 2.6097137901127496, + "grad_norm": 0.032202690839767456, + "learning_rate": 2.611344842593936e-06, + "loss": 0.0004, + "step": 24072 + }, + { + "epoch": 2.609822202948829, + "grad_norm": 1.6152796745300293, + "learning_rate": 2.6106194690265492e-06, + "loss": 0.0047, + "step": 24073 + }, + { + "epoch": 2.6099306157849087, + "grad_norm": 4.574814796447754, + "learning_rate": 2.609894095459162e-06, + "loss": 0.0297, + "step": 24074 + }, + { + "epoch": 2.6100390286209887, + "grad_norm": 2.2084221839904785, + "learning_rate": 2.6091687218917744e-06, + "loss": 0.0022, + "step": 24075 + }, + { + "epoch": 2.6101474414570687, + "grad_norm": 0.11588495969772339, + "learning_rate": 2.608443348324387e-06, + "loss": 0.0006, + "step": 24076 + }, + { + "epoch": 2.610255854293148, + "grad_norm": 3.8189382553100586, + "learning_rate": 2.6077179747570003e-06, + "loss": 0.0188, + "step": 24077 + }, + { + "epoch": 2.610364267129228, + "grad_norm": 0.29889973998069763, + "learning_rate": 2.606992601189613e-06, + "loss": 0.0015, + "step": 24078 + }, + { + "epoch": 2.6104726799653077, + "grad_norm": 0.6298391819000244, + "learning_rate": 2.6062672276222255e-06, + "loss": 0.002, + "step": 24079 + }, + { + "epoch": 2.6105810928013877, + "grad_norm": 0.12764288485050201, + "learning_rate": 2.6055418540548385e-06, + "loss": 0.0002, + "step": 24080 + }, + { + "epoch": 2.6106895056374677, + "grad_norm": 2.2624690532684326, + "learning_rate": 2.6048164804874515e-06, + "loss": 0.0236, + "step": 24081 + }, + { + "epoch": 2.6107979184735473, + "grad_norm": 0.05161842703819275, + "learning_rate": 2.604091106920064e-06, + "loss": 0.0002, + "step": 24082 + }, + { + "epoch": 2.610906331309627, + "grad_norm": 0.6829041242599487, + "learning_rate": 2.603365733352677e-06, + "loss": 0.001, + "step": 24083 + }, + { + "epoch": 2.611014744145707, + "grad_norm": 3.121504068374634, + "learning_rate": 2.6026403597852896e-06, + "loss": 0.0063, + "step": 24084 + }, + { + "epoch": 2.6111231569817868, + "grad_norm": 2.328770637512207, + "learning_rate": 2.601914986217902e-06, + "loss": 0.015, + "step": 24085 + }, + { + "epoch": 2.6112315698178663, + "grad_norm": 0.05872402340173721, + "learning_rate": 2.6011896126505156e-06, + "loss": 0.0003, + "step": 24086 + }, + { + "epoch": 2.6113399826539463, + "grad_norm": 2.7435975074768066, + "learning_rate": 2.600464239083128e-06, + "loss": 0.0011, + "step": 24087 + }, + { + "epoch": 2.611448395490026, + "grad_norm": 1.9128096103668213, + "learning_rate": 2.5997388655157407e-06, + "loss": 0.0328, + "step": 24088 + }, + { + "epoch": 2.611556808326106, + "grad_norm": 2.181542158126831, + "learning_rate": 2.5990134919483533e-06, + "loss": 0.0035, + "step": 24089 + }, + { + "epoch": 2.611665221162186, + "grad_norm": 0.13495926558971405, + "learning_rate": 2.5982881183809667e-06, + "loss": 0.0006, + "step": 24090 + }, + { + "epoch": 2.6117736339982653, + "grad_norm": 5.304355144500732, + "learning_rate": 2.5975627448135793e-06, + "loss": 0.0299, + "step": 24091 + }, + { + "epoch": 2.6118820468343453, + "grad_norm": 4.446225643157959, + "learning_rate": 2.596837371246192e-06, + "loss": 0.0317, + "step": 24092 + }, + { + "epoch": 2.611990459670425, + "grad_norm": 0.015893954783678055, + "learning_rate": 2.596111997678805e-06, + "loss": 0.0002, + "step": 24093 + }, + { + "epoch": 2.612098872506505, + "grad_norm": 3.693844795227051, + "learning_rate": 2.595386624111418e-06, + "loss": 0.0029, + "step": 24094 + }, + { + "epoch": 2.612207285342585, + "grad_norm": 0.22474664449691772, + "learning_rate": 2.5946612505440304e-06, + "loss": 0.0009, + "step": 24095 + }, + { + "epoch": 2.6123156981786644, + "grad_norm": 0.14860641956329346, + "learning_rate": 2.5939358769766434e-06, + "loss": 0.0004, + "step": 24096 + }, + { + "epoch": 2.612424111014744, + "grad_norm": 0.05776388198137283, + "learning_rate": 2.593210503409256e-06, + "loss": 0.0004, + "step": 24097 + }, + { + "epoch": 2.612532523850824, + "grad_norm": 2.613105297088623, + "learning_rate": 2.5924851298418685e-06, + "loss": 0.0043, + "step": 24098 + }, + { + "epoch": 2.612640936686904, + "grad_norm": 0.16514788568019867, + "learning_rate": 2.591759756274482e-06, + "loss": 0.0009, + "step": 24099 + }, + { + "epoch": 2.6127493495229834, + "grad_norm": 1.3704838752746582, + "learning_rate": 2.5910343827070945e-06, + "loss": 0.0074, + "step": 24100 + }, + { + "epoch": 2.6128577623590634, + "grad_norm": 2.271699905395508, + "learning_rate": 2.590309009139707e-06, + "loss": 0.0049, + "step": 24101 + }, + { + "epoch": 2.612966175195143, + "grad_norm": 0.01544521376490593, + "learning_rate": 2.5895836355723196e-06, + "loss": 0.0001, + "step": 24102 + }, + { + "epoch": 2.613074588031223, + "grad_norm": 4.650588512420654, + "learning_rate": 2.588858262004933e-06, + "loss": 0.0152, + "step": 24103 + }, + { + "epoch": 2.613183000867303, + "grad_norm": 0.18296201527118683, + "learning_rate": 2.5881328884375456e-06, + "loss": 0.0005, + "step": 24104 + }, + { + "epoch": 2.6132914137033825, + "grad_norm": 1.5226776599884033, + "learning_rate": 2.587407514870158e-06, + "loss": 0.0297, + "step": 24105 + }, + { + "epoch": 2.613399826539462, + "grad_norm": 0.5358450412750244, + "learning_rate": 2.586682141302771e-06, + "loss": 0.0013, + "step": 24106 + }, + { + "epoch": 2.613508239375542, + "grad_norm": 7.769946575164795, + "learning_rate": 2.585956767735384e-06, + "loss": 0.0473, + "step": 24107 + }, + { + "epoch": 2.613616652211622, + "grad_norm": 0.169852614402771, + "learning_rate": 2.5852313941679967e-06, + "loss": 0.0007, + "step": 24108 + }, + { + "epoch": 2.6137250650477015, + "grad_norm": 1.9745310544967651, + "learning_rate": 2.5845060206006097e-06, + "loss": 0.0022, + "step": 24109 + }, + { + "epoch": 2.6138334778837815, + "grad_norm": 1.3059194087982178, + "learning_rate": 2.5837806470332223e-06, + "loss": 0.0051, + "step": 24110 + }, + { + "epoch": 2.613941890719861, + "grad_norm": 0.03680489584803581, + "learning_rate": 2.583055273465835e-06, + "loss": 0.0002, + "step": 24111 + }, + { + "epoch": 2.614050303555941, + "grad_norm": 3.644890308380127, + "learning_rate": 2.5823298998984482e-06, + "loss": 0.0268, + "step": 24112 + }, + { + "epoch": 2.614158716392021, + "grad_norm": 0.022169429808855057, + "learning_rate": 2.581604526331061e-06, + "loss": 0.0002, + "step": 24113 + }, + { + "epoch": 2.6142671292281006, + "grad_norm": 0.09524784237146378, + "learning_rate": 2.5808791527636734e-06, + "loss": 0.0007, + "step": 24114 + }, + { + "epoch": 2.6143755420641805, + "grad_norm": 0.015066939406096935, + "learning_rate": 2.580153779196286e-06, + "loss": 0.0001, + "step": 24115 + }, + { + "epoch": 2.61448395490026, + "grad_norm": 3.0759449005126953, + "learning_rate": 2.5794284056288994e-06, + "loss": 0.0289, + "step": 24116 + }, + { + "epoch": 2.61459236773634, + "grad_norm": 4.874696254730225, + "learning_rate": 2.578703032061512e-06, + "loss": 0.0053, + "step": 24117 + }, + { + "epoch": 2.6147007805724196, + "grad_norm": 0.0036457933019846678, + "learning_rate": 2.5779776584941245e-06, + "loss": 0.0, + "step": 24118 + }, + { + "epoch": 2.6148091934084996, + "grad_norm": 4.48215389251709, + "learning_rate": 2.5772522849267375e-06, + "loss": 0.0229, + "step": 24119 + }, + { + "epoch": 2.614917606244579, + "grad_norm": 1.0379153490066528, + "learning_rate": 2.5765269113593505e-06, + "loss": 0.0319, + "step": 24120 + }, + { + "epoch": 2.615026019080659, + "grad_norm": 3.5557472705841064, + "learning_rate": 2.575801537791963e-06, + "loss": 0.0253, + "step": 24121 + }, + { + "epoch": 2.615134431916739, + "grad_norm": 17.730499267578125, + "learning_rate": 2.575076164224576e-06, + "loss": 0.0542, + "step": 24122 + }, + { + "epoch": 2.6152428447528187, + "grad_norm": 0.20625732839107513, + "learning_rate": 2.5743507906571886e-06, + "loss": 0.0007, + "step": 24123 + }, + { + "epoch": 2.6153512575888986, + "grad_norm": 0.5740917325019836, + "learning_rate": 2.573625417089801e-06, + "loss": 0.0012, + "step": 24124 + }, + { + "epoch": 2.615459670424978, + "grad_norm": 3.357308864593506, + "learning_rate": 2.5729000435224146e-06, + "loss": 0.006, + "step": 24125 + }, + { + "epoch": 2.615568083261058, + "grad_norm": 6.400149345397949, + "learning_rate": 2.572174669955027e-06, + "loss": 0.0111, + "step": 24126 + }, + { + "epoch": 2.615676496097138, + "grad_norm": 8.764751434326172, + "learning_rate": 2.5714492963876397e-06, + "loss": 0.0058, + "step": 24127 + }, + { + "epoch": 2.6157849089332177, + "grad_norm": 2.303215742111206, + "learning_rate": 2.5707239228202523e-06, + "loss": 0.0262, + "step": 24128 + }, + { + "epoch": 2.6158933217692972, + "grad_norm": 0.1595579981803894, + "learning_rate": 2.5699985492528657e-06, + "loss": 0.0009, + "step": 24129 + }, + { + "epoch": 2.616001734605377, + "grad_norm": 0.02449055388569832, + "learning_rate": 2.5692731756854783e-06, + "loss": 0.0002, + "step": 24130 + }, + { + "epoch": 2.616110147441457, + "grad_norm": 0.0713699609041214, + "learning_rate": 2.568547802118091e-06, + "loss": 0.0005, + "step": 24131 + }, + { + "epoch": 2.6162185602775367, + "grad_norm": 0.4897448420524597, + "learning_rate": 2.567822428550704e-06, + "loss": 0.0015, + "step": 24132 + }, + { + "epoch": 2.6163269731136167, + "grad_norm": 0.0983903631567955, + "learning_rate": 2.567097054983317e-06, + "loss": 0.0004, + "step": 24133 + }, + { + "epoch": 2.6164353859496963, + "grad_norm": 0.028501154854893684, + "learning_rate": 2.5663716814159294e-06, + "loss": 0.0003, + "step": 24134 + }, + { + "epoch": 2.6165437987857763, + "grad_norm": 2.5206093788146973, + "learning_rate": 2.5656463078485424e-06, + "loss": 0.0029, + "step": 24135 + }, + { + "epoch": 2.6166522116218562, + "grad_norm": 0.7049729824066162, + "learning_rate": 2.564920934281155e-06, + "loss": 0.001, + "step": 24136 + }, + { + "epoch": 2.616760624457936, + "grad_norm": 1.393684983253479, + "learning_rate": 2.564195560713768e-06, + "loss": 0.0011, + "step": 24137 + }, + { + "epoch": 2.6168690372940158, + "grad_norm": 0.036161672323942184, + "learning_rate": 2.563470187146381e-06, + "loss": 0.0004, + "step": 24138 + }, + { + "epoch": 2.6169774501300953, + "grad_norm": 0.03773040324449539, + "learning_rate": 2.5627448135789935e-06, + "loss": 0.0002, + "step": 24139 + }, + { + "epoch": 2.6170858629661753, + "grad_norm": 0.004672633949667215, + "learning_rate": 2.562019440011606e-06, + "loss": 0.0001, + "step": 24140 + }, + { + "epoch": 2.617194275802255, + "grad_norm": 0.15903432667255402, + "learning_rate": 2.5612940664442186e-06, + "loss": 0.0008, + "step": 24141 + }, + { + "epoch": 2.617302688638335, + "grad_norm": 2.4446158409118652, + "learning_rate": 2.560568692876832e-06, + "loss": 0.0128, + "step": 24142 + }, + { + "epoch": 2.6174111014744144, + "grad_norm": 0.031104136258363724, + "learning_rate": 2.5598433193094446e-06, + "loss": 0.0003, + "step": 24143 + }, + { + "epoch": 2.6175195143104943, + "grad_norm": 0.48272424936294556, + "learning_rate": 2.559117945742057e-06, + "loss": 0.002, + "step": 24144 + }, + { + "epoch": 2.6176279271465743, + "grad_norm": 0.26953983306884766, + "learning_rate": 2.55839257217467e-06, + "loss": 0.0019, + "step": 24145 + }, + { + "epoch": 2.617736339982654, + "grad_norm": 0.08128009736537933, + "learning_rate": 2.557667198607283e-06, + "loss": 0.0007, + "step": 24146 + }, + { + "epoch": 2.617844752818734, + "grad_norm": 0.26447927951812744, + "learning_rate": 2.5569418250398957e-06, + "loss": 0.0009, + "step": 24147 + }, + { + "epoch": 2.6179531656548134, + "grad_norm": 0.013520834036171436, + "learning_rate": 2.5562164514725087e-06, + "loss": 0.0001, + "step": 24148 + }, + { + "epoch": 2.6180615784908934, + "grad_norm": 0.08837388455867767, + "learning_rate": 2.5554910779051213e-06, + "loss": 0.0005, + "step": 24149 + }, + { + "epoch": 2.6181699913269734, + "grad_norm": 5.521915435791016, + "learning_rate": 2.5547657043377343e-06, + "loss": 0.0153, + "step": 24150 + }, + { + "epoch": 2.618278404163053, + "grad_norm": 6.752490997314453, + "learning_rate": 2.5540403307703472e-06, + "loss": 0.0131, + "step": 24151 + }, + { + "epoch": 2.6183868169991324, + "grad_norm": 3.652937889099121, + "learning_rate": 2.55331495720296e-06, + "loss": 0.0301, + "step": 24152 + }, + { + "epoch": 2.6184952298352124, + "grad_norm": 0.0203399695456028, + "learning_rate": 2.5525895836355724e-06, + "loss": 0.0001, + "step": 24153 + }, + { + "epoch": 2.6186036426712924, + "grad_norm": 0.041938889771699905, + "learning_rate": 2.551864210068185e-06, + "loss": 0.0003, + "step": 24154 + }, + { + "epoch": 2.618712055507372, + "grad_norm": 2.3788509368896484, + "learning_rate": 2.5511388365007984e-06, + "loss": 0.0209, + "step": 24155 + }, + { + "epoch": 2.618820468343452, + "grad_norm": 0.09202288836240768, + "learning_rate": 2.550413462933411e-06, + "loss": 0.0005, + "step": 24156 + }, + { + "epoch": 2.6189288811795315, + "grad_norm": 1.7882479429244995, + "learning_rate": 2.5496880893660235e-06, + "loss": 0.0156, + "step": 24157 + }, + { + "epoch": 2.6190372940156115, + "grad_norm": 11.856088638305664, + "learning_rate": 2.5489627157986365e-06, + "loss": 0.01, + "step": 24158 + }, + { + "epoch": 2.6191457068516915, + "grad_norm": 0.34059348702430725, + "learning_rate": 2.5482373422312495e-06, + "loss": 0.0015, + "step": 24159 + }, + { + "epoch": 2.619254119687771, + "grad_norm": 0.11127150058746338, + "learning_rate": 2.547511968663862e-06, + "loss": 0.0007, + "step": 24160 + }, + { + "epoch": 2.619362532523851, + "grad_norm": 0.1480190008878708, + "learning_rate": 2.546786595096475e-06, + "loss": 0.0007, + "step": 24161 + }, + { + "epoch": 2.6194709453599305, + "grad_norm": 1.6645567417144775, + "learning_rate": 2.5460612215290876e-06, + "loss": 0.003, + "step": 24162 + }, + { + "epoch": 2.6195793581960105, + "grad_norm": 0.6132275462150574, + "learning_rate": 2.5453358479617006e-06, + "loss": 0.0015, + "step": 24163 + }, + { + "epoch": 2.61968777103209, + "grad_norm": 1.432173252105713, + "learning_rate": 2.5446104743943136e-06, + "loss": 0.0282, + "step": 24164 + }, + { + "epoch": 2.61979618386817, + "grad_norm": 0.3661094009876251, + "learning_rate": 2.543885100826926e-06, + "loss": 0.0017, + "step": 24165 + }, + { + "epoch": 2.6199045967042496, + "grad_norm": 2.13407301902771, + "learning_rate": 2.5431597272595387e-06, + "loss": 0.0039, + "step": 24166 + }, + { + "epoch": 2.6200130095403296, + "grad_norm": 8.738712310791016, + "learning_rate": 2.5424343536921513e-06, + "loss": 0.0219, + "step": 24167 + }, + { + "epoch": 2.6201214223764095, + "grad_norm": 3.736617088317871, + "learning_rate": 2.5417089801247647e-06, + "loss": 0.0187, + "step": 24168 + }, + { + "epoch": 2.620229835212489, + "grad_norm": 3.4679994583129883, + "learning_rate": 2.5409836065573773e-06, + "loss": 0.0307, + "step": 24169 + }, + { + "epoch": 2.620338248048569, + "grad_norm": 7.378194808959961, + "learning_rate": 2.54025823298999e-06, + "loss": 0.0187, + "step": 24170 + }, + { + "epoch": 2.6204466608846486, + "grad_norm": 0.43634089827537537, + "learning_rate": 2.539532859422603e-06, + "loss": 0.0018, + "step": 24171 + }, + { + "epoch": 2.6205550737207286, + "grad_norm": 0.26309114694595337, + "learning_rate": 2.538807485855216e-06, + "loss": 0.0011, + "step": 24172 + }, + { + "epoch": 2.6206634865568086, + "grad_norm": 0.847161054611206, + "learning_rate": 2.5380821122878284e-06, + "loss": 0.0039, + "step": 24173 + }, + { + "epoch": 2.620771899392888, + "grad_norm": 0.07285935431718826, + "learning_rate": 2.5373567387204414e-06, + "loss": 0.0004, + "step": 24174 + }, + { + "epoch": 2.6208803122289677, + "grad_norm": 2.23974871635437, + "learning_rate": 2.536631365153054e-06, + "loss": 0.0024, + "step": 24175 + }, + { + "epoch": 2.6209887250650477, + "grad_norm": 3.5243256092071533, + "learning_rate": 2.535905991585667e-06, + "loss": 0.024, + "step": 24176 + }, + { + "epoch": 2.6210971379011276, + "grad_norm": 0.058772336691617966, + "learning_rate": 2.53518061801828e-06, + "loss": 0.0004, + "step": 24177 + }, + { + "epoch": 2.621205550737207, + "grad_norm": 5.920992851257324, + "learning_rate": 2.5344552444508925e-06, + "loss": 0.0187, + "step": 24178 + }, + { + "epoch": 2.621313963573287, + "grad_norm": 0.06881868839263916, + "learning_rate": 2.533729870883505e-06, + "loss": 0.0004, + "step": 24179 + }, + { + "epoch": 2.6214223764093667, + "grad_norm": 2.879114866256714, + "learning_rate": 2.5330044973161176e-06, + "loss": 0.0156, + "step": 24180 + }, + { + "epoch": 2.6215307892454467, + "grad_norm": 1.5258890390396118, + "learning_rate": 2.532279123748731e-06, + "loss": 0.0065, + "step": 24181 + }, + { + "epoch": 2.6216392020815267, + "grad_norm": 0.03353152424097061, + "learning_rate": 2.5315537501813436e-06, + "loss": 0.0004, + "step": 24182 + }, + { + "epoch": 2.621747614917606, + "grad_norm": 0.019572164863348007, + "learning_rate": 2.530828376613956e-06, + "loss": 0.0001, + "step": 24183 + }, + { + "epoch": 2.621856027753686, + "grad_norm": 0.13691839575767517, + "learning_rate": 2.530103003046569e-06, + "loss": 0.0008, + "step": 24184 + }, + { + "epoch": 2.6219644405897657, + "grad_norm": 0.016464240849018097, + "learning_rate": 2.529377629479182e-06, + "loss": 0.0002, + "step": 24185 + }, + { + "epoch": 2.6220728534258457, + "grad_norm": 0.07466953247785568, + "learning_rate": 2.5286522559117947e-06, + "loss": 0.0003, + "step": 24186 + }, + { + "epoch": 2.6221812662619253, + "grad_norm": 0.027915582060813904, + "learning_rate": 2.5279268823444077e-06, + "loss": 0.0002, + "step": 24187 + }, + { + "epoch": 2.6222896790980053, + "grad_norm": 0.04668563976883888, + "learning_rate": 2.5272015087770203e-06, + "loss": 0.0005, + "step": 24188 + }, + { + "epoch": 2.622398091934085, + "grad_norm": 5.524749279022217, + "learning_rate": 2.5264761352096333e-06, + "loss": 0.0178, + "step": 24189 + }, + { + "epoch": 2.622506504770165, + "grad_norm": 11.292316436767578, + "learning_rate": 2.5257507616422462e-06, + "loss": 0.0124, + "step": 24190 + }, + { + "epoch": 2.6226149176062448, + "grad_norm": 2.749122381210327, + "learning_rate": 2.525025388074859e-06, + "loss": 0.0205, + "step": 24191 + }, + { + "epoch": 2.6227233304423243, + "grad_norm": 2.5674757957458496, + "learning_rate": 2.5243000145074714e-06, + "loss": 0.0073, + "step": 24192 + }, + { + "epoch": 2.6228317432784043, + "grad_norm": 0.012436711229383945, + "learning_rate": 2.523574640940084e-06, + "loss": 0.0001, + "step": 24193 + }, + { + "epoch": 2.622940156114484, + "grad_norm": 0.0029255577828735113, + "learning_rate": 2.5228492673726974e-06, + "loss": 0.0, + "step": 24194 + }, + { + "epoch": 2.623048568950564, + "grad_norm": 0.9746977686882019, + "learning_rate": 2.52212389380531e-06, + "loss": 0.0025, + "step": 24195 + }, + { + "epoch": 2.623156981786644, + "grad_norm": 2.0914862155914307, + "learning_rate": 2.5213985202379225e-06, + "loss": 0.0249, + "step": 24196 + }, + { + "epoch": 2.6232653946227233, + "grad_norm": 0.03623955696821213, + "learning_rate": 2.5206731466705355e-06, + "loss": 0.0003, + "step": 24197 + }, + { + "epoch": 2.623373807458803, + "grad_norm": 0.008307970128953457, + "learning_rate": 2.5199477731031485e-06, + "loss": 0.0001, + "step": 24198 + }, + { + "epoch": 2.623482220294883, + "grad_norm": 0.03897521272301674, + "learning_rate": 2.519222399535761e-06, + "loss": 0.0003, + "step": 24199 + }, + { + "epoch": 2.623590633130963, + "grad_norm": 0.526931881904602, + "learning_rate": 2.518497025968374e-06, + "loss": 0.0009, + "step": 24200 + }, + { + "epoch": 2.6236990459670424, + "grad_norm": 2.9163239002227783, + "learning_rate": 2.5177716524009866e-06, + "loss": 0.007, + "step": 24201 + }, + { + "epoch": 2.6238074588031224, + "grad_norm": 5.242954254150391, + "learning_rate": 2.5170462788335996e-06, + "loss": 0.0373, + "step": 24202 + }, + { + "epoch": 2.623915871639202, + "grad_norm": 0.011351263150572777, + "learning_rate": 2.5163209052662126e-06, + "loss": 0.0001, + "step": 24203 + }, + { + "epoch": 2.624024284475282, + "grad_norm": 3.793882369995117, + "learning_rate": 2.515595531698825e-06, + "loss": 0.0275, + "step": 24204 + }, + { + "epoch": 2.624132697311362, + "grad_norm": 0.07688208669424057, + "learning_rate": 2.5148701581314377e-06, + "loss": 0.0004, + "step": 24205 + }, + { + "epoch": 2.6242411101474414, + "grad_norm": 3.1793999671936035, + "learning_rate": 2.5141447845640503e-06, + "loss": 0.0105, + "step": 24206 + }, + { + "epoch": 2.624349522983521, + "grad_norm": 4.411421298980713, + "learning_rate": 2.5134194109966637e-06, + "loss": 0.0191, + "step": 24207 + }, + { + "epoch": 2.624457935819601, + "grad_norm": 0.31157541275024414, + "learning_rate": 2.5126940374292763e-06, + "loss": 0.0011, + "step": 24208 + }, + { + "epoch": 2.624566348655681, + "grad_norm": 0.03746241703629494, + "learning_rate": 2.511968663861889e-06, + "loss": 0.0004, + "step": 24209 + }, + { + "epoch": 2.6246747614917605, + "grad_norm": 0.05843653529882431, + "learning_rate": 2.511243290294502e-06, + "loss": 0.0004, + "step": 24210 + }, + { + "epoch": 2.6247831743278405, + "grad_norm": 0.9203205704689026, + "learning_rate": 2.510517916727115e-06, + "loss": 0.0032, + "step": 24211 + }, + { + "epoch": 2.62489158716392, + "grad_norm": 13.080164909362793, + "learning_rate": 2.5097925431597274e-06, + "loss": 0.0223, + "step": 24212 + }, + { + "epoch": 2.625, + "grad_norm": 0.22080491483211517, + "learning_rate": 2.5090671695923404e-06, + "loss": 0.0013, + "step": 24213 + }, + { + "epoch": 2.62510841283608, + "grad_norm": 3.0024333000183105, + "learning_rate": 2.508341796024953e-06, + "loss": 0.0205, + "step": 24214 + }, + { + "epoch": 2.6252168256721595, + "grad_norm": 13.704266548156738, + "learning_rate": 2.507616422457566e-06, + "loss": 0.0154, + "step": 24215 + }, + { + "epoch": 2.6253252385082395, + "grad_norm": 4.389416694641113, + "learning_rate": 2.506891048890179e-06, + "loss": 0.0102, + "step": 24216 + }, + { + "epoch": 2.625433651344319, + "grad_norm": 1.6809461116790771, + "learning_rate": 2.5061656753227915e-06, + "loss": 0.051, + "step": 24217 + }, + { + "epoch": 2.625542064180399, + "grad_norm": 3.952893018722534, + "learning_rate": 2.505440301755404e-06, + "loss": 0.0033, + "step": 24218 + }, + { + "epoch": 2.625650477016479, + "grad_norm": 1.0154516696929932, + "learning_rate": 2.5047149281880166e-06, + "loss": 0.0138, + "step": 24219 + }, + { + "epoch": 2.6257588898525586, + "grad_norm": 0.02148997224867344, + "learning_rate": 2.50398955462063e-06, + "loss": 0.0002, + "step": 24220 + }, + { + "epoch": 2.625867302688638, + "grad_norm": 0.696972131729126, + "learning_rate": 2.5032641810532426e-06, + "loss": 0.0024, + "step": 24221 + }, + { + "epoch": 2.625975715524718, + "grad_norm": 4.806148529052734, + "learning_rate": 2.502538807485855e-06, + "loss": 0.0368, + "step": 24222 + }, + { + "epoch": 2.626084128360798, + "grad_norm": 0.013258855789899826, + "learning_rate": 2.501813433918468e-06, + "loss": 0.0001, + "step": 24223 + }, + { + "epoch": 2.6261925411968776, + "grad_norm": 3.06978702545166, + "learning_rate": 2.501088060351081e-06, + "loss": 0.0045, + "step": 24224 + }, + { + "epoch": 2.6263009540329576, + "grad_norm": 0.06390015780925751, + "learning_rate": 2.5003626867836937e-06, + "loss": 0.0002, + "step": 24225 + }, + { + "epoch": 2.626409366869037, + "grad_norm": 8.563924789428711, + "learning_rate": 2.4996373132163067e-06, + "loss": 0.0294, + "step": 24226 + }, + { + "epoch": 2.626517779705117, + "grad_norm": 0.16724525392055511, + "learning_rate": 2.4989119396489197e-06, + "loss": 0.0007, + "step": 24227 + }, + { + "epoch": 2.626626192541197, + "grad_norm": 1.9611577987670898, + "learning_rate": 2.4981865660815323e-06, + "loss": 0.0111, + "step": 24228 + }, + { + "epoch": 2.6267346053772767, + "grad_norm": 0.9361093640327454, + "learning_rate": 2.4974611925141452e-06, + "loss": 0.039, + "step": 24229 + }, + { + "epoch": 2.626843018213356, + "grad_norm": 5.893643379211426, + "learning_rate": 2.496735818946758e-06, + "loss": 0.0134, + "step": 24230 + }, + { + "epoch": 2.626951431049436, + "grad_norm": 0.10299261659383774, + "learning_rate": 2.496010445379371e-06, + "loss": 0.0006, + "step": 24231 + }, + { + "epoch": 2.627059843885516, + "grad_norm": 0.04319422319531441, + "learning_rate": 2.4952850718119834e-06, + "loss": 0.0004, + "step": 24232 + }, + { + "epoch": 2.6271682567215957, + "grad_norm": 0.00668247789144516, + "learning_rate": 2.494559698244596e-06, + "loss": 0.0001, + "step": 24233 + }, + { + "epoch": 2.6272766695576757, + "grad_norm": 0.9957588315010071, + "learning_rate": 2.493834324677209e-06, + "loss": 0.007, + "step": 24234 + }, + { + "epoch": 2.6273850823937552, + "grad_norm": 1.6840485334396362, + "learning_rate": 2.4931089511098215e-06, + "loss": 0.004, + "step": 24235 + }, + { + "epoch": 2.627493495229835, + "grad_norm": 0.22055861353874207, + "learning_rate": 2.4923835775424345e-06, + "loss": 0.0013, + "step": 24236 + }, + { + "epoch": 2.627601908065915, + "grad_norm": 0.07613809406757355, + "learning_rate": 2.491658203975047e-06, + "loss": 0.0006, + "step": 24237 + }, + { + "epoch": 2.6277103209019947, + "grad_norm": 0.08333646506071091, + "learning_rate": 2.49093283040766e-06, + "loss": 0.0008, + "step": 24238 + }, + { + "epoch": 2.6278187337380747, + "grad_norm": 0.017903417348861694, + "learning_rate": 2.490207456840273e-06, + "loss": 0.0001, + "step": 24239 + }, + { + "epoch": 2.6279271465741543, + "grad_norm": 0.540445864200592, + "learning_rate": 2.489482083272886e-06, + "loss": 0.0033, + "step": 24240 + }, + { + "epoch": 2.6280355594102343, + "grad_norm": 7.337190628051758, + "learning_rate": 2.4887567097054986e-06, + "loss": 0.0173, + "step": 24241 + }, + { + "epoch": 2.628143972246314, + "grad_norm": 0.08216787129640579, + "learning_rate": 2.4880313361381116e-06, + "loss": 0.0004, + "step": 24242 + }, + { + "epoch": 2.628252385082394, + "grad_norm": 0.034215111285448074, + "learning_rate": 2.487305962570724e-06, + "loss": 0.0002, + "step": 24243 + }, + { + "epoch": 2.6283607979184733, + "grad_norm": 0.04510440677404404, + "learning_rate": 2.486580589003337e-06, + "loss": 0.0003, + "step": 24244 + }, + { + "epoch": 2.6284692107545533, + "grad_norm": 0.008929413743317127, + "learning_rate": 2.4858552154359497e-06, + "loss": 0.0001, + "step": 24245 + }, + { + "epoch": 2.6285776235906333, + "grad_norm": 2.801920175552368, + "learning_rate": 2.4851298418685623e-06, + "loss": 0.0213, + "step": 24246 + }, + { + "epoch": 2.628686036426713, + "grad_norm": 2.959505081176758, + "learning_rate": 2.4844044683011753e-06, + "loss": 0.0081, + "step": 24247 + }, + { + "epoch": 2.628794449262793, + "grad_norm": 0.07038618624210358, + "learning_rate": 2.483679094733788e-06, + "loss": 0.0006, + "step": 24248 + }, + { + "epoch": 2.6289028620988724, + "grad_norm": 5.159634113311768, + "learning_rate": 2.482953721166401e-06, + "loss": 0.0109, + "step": 24249 + }, + { + "epoch": 2.6290112749349523, + "grad_norm": 0.08440511673688889, + "learning_rate": 2.4822283475990134e-06, + "loss": 0.0006, + "step": 24250 + }, + { + "epoch": 2.6291196877710323, + "grad_norm": 0.08213754743337631, + "learning_rate": 2.4815029740316264e-06, + "loss": 0.0005, + "step": 24251 + }, + { + "epoch": 2.629228100607112, + "grad_norm": 0.3820410668849945, + "learning_rate": 2.4807776004642394e-06, + "loss": 0.0016, + "step": 24252 + }, + { + "epoch": 2.6293365134431914, + "grad_norm": 2.1864302158355713, + "learning_rate": 2.4800522268968524e-06, + "loss": 0.0066, + "step": 24253 + }, + { + "epoch": 2.6294449262792714, + "grad_norm": 1.6968002319335938, + "learning_rate": 2.479326853329465e-06, + "loss": 0.0013, + "step": 24254 + }, + { + "epoch": 2.6295533391153514, + "grad_norm": 2.2018589973449707, + "learning_rate": 2.478601479762078e-06, + "loss": 0.0119, + "step": 24255 + }, + { + "epoch": 2.629661751951431, + "grad_norm": 2.1238090991973877, + "learning_rate": 2.4778761061946905e-06, + "loss": 0.0148, + "step": 24256 + }, + { + "epoch": 2.629770164787511, + "grad_norm": 0.009525920264422894, + "learning_rate": 2.4771507326273035e-06, + "loss": 0.0001, + "step": 24257 + }, + { + "epoch": 2.6298785776235905, + "grad_norm": 0.15929095447063446, + "learning_rate": 2.476425359059916e-06, + "loss": 0.0011, + "step": 24258 + }, + { + "epoch": 2.6299869904596704, + "grad_norm": 0.05248512327671051, + "learning_rate": 2.475699985492529e-06, + "loss": 0.0004, + "step": 24259 + }, + { + "epoch": 2.6300954032957504, + "grad_norm": 0.1660023331642151, + "learning_rate": 2.4749746119251416e-06, + "loss": 0.0008, + "step": 24260 + }, + { + "epoch": 2.63020381613183, + "grad_norm": 0.23073460161685944, + "learning_rate": 2.474249238357754e-06, + "loss": 0.001, + "step": 24261 + }, + { + "epoch": 2.63031222896791, + "grad_norm": 13.510080337524414, + "learning_rate": 2.473523864790367e-06, + "loss": 0.0645, + "step": 24262 + }, + { + "epoch": 2.6304206418039895, + "grad_norm": 0.07572022080421448, + "learning_rate": 2.47279849122298e-06, + "loss": 0.0002, + "step": 24263 + }, + { + "epoch": 2.6305290546400695, + "grad_norm": 11.913260459899902, + "learning_rate": 2.4720731176555927e-06, + "loss": 0.0515, + "step": 24264 + }, + { + "epoch": 2.630637467476149, + "grad_norm": 9.902149200439453, + "learning_rate": 2.4713477440882057e-06, + "loss": 0.0209, + "step": 24265 + }, + { + "epoch": 2.630745880312229, + "grad_norm": 0.04199698939919472, + "learning_rate": 2.4706223705208187e-06, + "loss": 0.0004, + "step": 24266 + }, + { + "epoch": 2.6308542931483085, + "grad_norm": 0.943612813949585, + "learning_rate": 2.4698969969534313e-06, + "loss": 0.0055, + "step": 24267 + }, + { + "epoch": 2.6309627059843885, + "grad_norm": 2.110135078430176, + "learning_rate": 2.4691716233860443e-06, + "loss": 0.0068, + "step": 24268 + }, + { + "epoch": 2.6310711188204685, + "grad_norm": 2.9160308837890625, + "learning_rate": 2.468446249818657e-06, + "loss": 0.0063, + "step": 24269 + }, + { + "epoch": 2.631179531656548, + "grad_norm": 0.30595317482948303, + "learning_rate": 2.46772087625127e-06, + "loss": 0.0006, + "step": 24270 + }, + { + "epoch": 2.631287944492628, + "grad_norm": 0.6000205278396606, + "learning_rate": 2.4669955026838824e-06, + "loss": 0.0034, + "step": 24271 + }, + { + "epoch": 2.6313963573287076, + "grad_norm": 0.030397554859519005, + "learning_rate": 2.4662701291164954e-06, + "loss": 0.0002, + "step": 24272 + }, + { + "epoch": 2.6315047701647876, + "grad_norm": 2.4114415645599365, + "learning_rate": 2.465544755549108e-06, + "loss": 0.0011, + "step": 24273 + }, + { + "epoch": 2.6316131830008676, + "grad_norm": 0.09479644149541855, + "learning_rate": 2.4648193819817205e-06, + "loss": 0.0008, + "step": 24274 + }, + { + "epoch": 2.631721595836947, + "grad_norm": 0.03428554907441139, + "learning_rate": 2.4640940084143335e-06, + "loss": 0.0003, + "step": 24275 + }, + { + "epoch": 2.6318300086730266, + "grad_norm": 3.517749547958374, + "learning_rate": 2.4633686348469465e-06, + "loss": 0.0119, + "step": 24276 + }, + { + "epoch": 2.6319384215091066, + "grad_norm": 0.3790040910243988, + "learning_rate": 2.462643261279559e-06, + "loss": 0.0008, + "step": 24277 + }, + { + "epoch": 2.6320468343451866, + "grad_norm": 2.0837175846099854, + "learning_rate": 2.461917887712172e-06, + "loss": 0.0279, + "step": 24278 + }, + { + "epoch": 2.632155247181266, + "grad_norm": 2.5903313159942627, + "learning_rate": 2.461192514144785e-06, + "loss": 0.0031, + "step": 24279 + }, + { + "epoch": 2.632263660017346, + "grad_norm": 2.1242024898529053, + "learning_rate": 2.4604671405773976e-06, + "loss": 0.0344, + "step": 24280 + }, + { + "epoch": 2.6323720728534257, + "grad_norm": 0.009405243210494518, + "learning_rate": 2.4597417670100106e-06, + "loss": 0.0001, + "step": 24281 + }, + { + "epoch": 2.6324804856895057, + "grad_norm": 0.10711059719324112, + "learning_rate": 2.459016393442623e-06, + "loss": 0.0003, + "step": 24282 + }, + { + "epoch": 2.6325888985255856, + "grad_norm": 2.1663269996643066, + "learning_rate": 2.458291019875236e-06, + "loss": 0.0101, + "step": 24283 + }, + { + "epoch": 2.632697311361665, + "grad_norm": 2.552690029144287, + "learning_rate": 2.4575656463078487e-06, + "loss": 0.0089, + "step": 24284 + }, + { + "epoch": 2.632805724197745, + "grad_norm": 0.161740243434906, + "learning_rate": 2.4568402727404617e-06, + "loss": 0.0014, + "step": 24285 + }, + { + "epoch": 2.6329141370338247, + "grad_norm": 0.23858655989170074, + "learning_rate": 2.4561148991730743e-06, + "loss": 0.0009, + "step": 24286 + }, + { + "epoch": 2.6330225498699047, + "grad_norm": 0.07568191736936569, + "learning_rate": 2.455389525605687e-06, + "loss": 0.0002, + "step": 24287 + }, + { + "epoch": 2.6331309627059842, + "grad_norm": 4.616508960723877, + "learning_rate": 2.4546641520383e-06, + "loss": 0.0077, + "step": 24288 + }, + { + "epoch": 2.633239375542064, + "grad_norm": 0.1500527262687683, + "learning_rate": 2.453938778470913e-06, + "loss": 0.0007, + "step": 24289 + }, + { + "epoch": 2.6333477883781438, + "grad_norm": 5.372107982635498, + "learning_rate": 2.4532134049035254e-06, + "loss": 0.0281, + "step": 24290 + }, + { + "epoch": 2.6334562012142237, + "grad_norm": 0.3060569167137146, + "learning_rate": 2.4524880313361384e-06, + "loss": 0.0012, + "step": 24291 + }, + { + "epoch": 2.6335646140503037, + "grad_norm": 1.7180652618408203, + "learning_rate": 2.4517626577687514e-06, + "loss": 0.0071, + "step": 24292 + }, + { + "epoch": 2.6336730268863833, + "grad_norm": 2.603381633758545, + "learning_rate": 2.451037284201364e-06, + "loss": 0.0055, + "step": 24293 + }, + { + "epoch": 2.6337814397224633, + "grad_norm": 0.06750226765871048, + "learning_rate": 2.450311910633977e-06, + "loss": 0.0008, + "step": 24294 + }, + { + "epoch": 2.633889852558543, + "grad_norm": 0.36716338992118835, + "learning_rate": 2.4495865370665895e-06, + "loss": 0.0011, + "step": 24295 + }, + { + "epoch": 2.633998265394623, + "grad_norm": 0.06103228032588959, + "learning_rate": 2.4488611634992025e-06, + "loss": 0.0002, + "step": 24296 + }, + { + "epoch": 2.6341066782307028, + "grad_norm": 2.0068275928497314, + "learning_rate": 2.448135789931815e-06, + "loss": 0.0069, + "step": 24297 + }, + { + "epoch": 2.6342150910667823, + "grad_norm": 5.718835353851318, + "learning_rate": 2.447410416364428e-06, + "loss": 0.0149, + "step": 24298 + }, + { + "epoch": 2.634323503902862, + "grad_norm": 0.05413884297013283, + "learning_rate": 2.4466850427970406e-06, + "loss": 0.0003, + "step": 24299 + }, + { + "epoch": 2.634431916738942, + "grad_norm": 0.061503101140260696, + "learning_rate": 2.445959669229653e-06, + "loss": 0.0004, + "step": 24300 + }, + { + "epoch": 2.634540329575022, + "grad_norm": 0.4464994966983795, + "learning_rate": 2.445234295662266e-06, + "loss": 0.0019, + "step": 24301 + }, + { + "epoch": 2.6346487424111014, + "grad_norm": 0.04616552218794823, + "learning_rate": 2.444508922094879e-06, + "loss": 0.0003, + "step": 24302 + }, + { + "epoch": 2.6347571552471813, + "grad_norm": 0.08805619925260544, + "learning_rate": 2.4437835485274917e-06, + "loss": 0.0003, + "step": 24303 + }, + { + "epoch": 2.634865568083261, + "grad_norm": 3.8771419525146484, + "learning_rate": 2.4430581749601047e-06, + "loss": 0.0789, + "step": 24304 + }, + { + "epoch": 2.634973980919341, + "grad_norm": 3.4683756828308105, + "learning_rate": 2.4423328013927177e-06, + "loss": 0.0253, + "step": 24305 + }, + { + "epoch": 2.635082393755421, + "grad_norm": 2.0299580097198486, + "learning_rate": 2.4416074278253303e-06, + "loss": 0.011, + "step": 24306 + }, + { + "epoch": 2.6351908065915004, + "grad_norm": 1.073228359222412, + "learning_rate": 2.4408820542579433e-06, + "loss": 0.0032, + "step": 24307 + }, + { + "epoch": 2.6352992194275804, + "grad_norm": 0.04307666793465614, + "learning_rate": 2.440156680690556e-06, + "loss": 0.0003, + "step": 24308 + }, + { + "epoch": 2.63540763226366, + "grad_norm": 0.00362875172868371, + "learning_rate": 2.439431307123169e-06, + "loss": 0.0, + "step": 24309 + }, + { + "epoch": 2.63551604509974, + "grad_norm": 3.2703871726989746, + "learning_rate": 2.4387059335557814e-06, + "loss": 0.0084, + "step": 24310 + }, + { + "epoch": 2.6356244579358195, + "grad_norm": 0.5422548055648804, + "learning_rate": 2.4379805599883944e-06, + "loss": 0.0025, + "step": 24311 + }, + { + "epoch": 2.6357328707718994, + "grad_norm": 0.18357326090335846, + "learning_rate": 2.437255186421007e-06, + "loss": 0.0008, + "step": 24312 + }, + { + "epoch": 2.635841283607979, + "grad_norm": 0.46099910140037537, + "learning_rate": 2.43652981285362e-06, + "loss": 0.0017, + "step": 24313 + }, + { + "epoch": 2.635949696444059, + "grad_norm": 5.161045551300049, + "learning_rate": 2.4358044392862325e-06, + "loss": 0.0272, + "step": 24314 + }, + { + "epoch": 2.636058109280139, + "grad_norm": 2.686640977859497, + "learning_rate": 2.4350790657188455e-06, + "loss": 0.0179, + "step": 24315 + }, + { + "epoch": 2.6361665221162185, + "grad_norm": 1.6337398290634155, + "learning_rate": 2.434353692151458e-06, + "loss": 0.0101, + "step": 24316 + }, + { + "epoch": 2.6362749349522985, + "grad_norm": 11.618966102600098, + "learning_rate": 2.433628318584071e-06, + "loss": 0.0199, + "step": 24317 + }, + { + "epoch": 2.636383347788378, + "grad_norm": 1.2464605569839478, + "learning_rate": 2.432902945016684e-06, + "loss": 0.0022, + "step": 24318 + }, + { + "epoch": 2.636491760624458, + "grad_norm": 0.032756507396698, + "learning_rate": 2.4321775714492966e-06, + "loss": 0.0002, + "step": 24319 + }, + { + "epoch": 2.636600173460538, + "grad_norm": 0.0762033611536026, + "learning_rate": 2.4314521978819096e-06, + "loss": 0.0004, + "step": 24320 + }, + { + "epoch": 2.6367085862966175, + "grad_norm": 0.01712660677731037, + "learning_rate": 2.430726824314522e-06, + "loss": 0.0001, + "step": 24321 + }, + { + "epoch": 2.636816999132697, + "grad_norm": 0.11098957806825638, + "learning_rate": 2.430001450747135e-06, + "loss": 0.0009, + "step": 24322 + }, + { + "epoch": 2.636925411968777, + "grad_norm": 0.700242817401886, + "learning_rate": 2.4292760771797477e-06, + "loss": 0.0021, + "step": 24323 + }, + { + "epoch": 2.637033824804857, + "grad_norm": 9.347847938537598, + "learning_rate": 2.4285507036123607e-06, + "loss": 0.021, + "step": 24324 + }, + { + "epoch": 2.6371422376409366, + "grad_norm": 0.3344951868057251, + "learning_rate": 2.4278253300449733e-06, + "loss": 0.0016, + "step": 24325 + }, + { + "epoch": 2.6372506504770166, + "grad_norm": 0.46405866742134094, + "learning_rate": 2.4270999564775863e-06, + "loss": 0.0013, + "step": 24326 + }, + { + "epoch": 2.637359063313096, + "grad_norm": 0.02556079626083374, + "learning_rate": 2.426374582910199e-06, + "loss": 0.0001, + "step": 24327 + }, + { + "epoch": 2.637467476149176, + "grad_norm": 0.3140481114387512, + "learning_rate": 2.425649209342812e-06, + "loss": 0.0015, + "step": 24328 + }, + { + "epoch": 2.637575888985256, + "grad_norm": 0.043681394308805466, + "learning_rate": 2.4249238357754244e-06, + "loss": 0.0003, + "step": 24329 + }, + { + "epoch": 2.6376843018213356, + "grad_norm": 0.035922806710004807, + "learning_rate": 2.4241984622080374e-06, + "loss": 0.0003, + "step": 24330 + }, + { + "epoch": 2.637792714657415, + "grad_norm": 0.9949558973312378, + "learning_rate": 2.4234730886406504e-06, + "loss": 0.003, + "step": 24331 + }, + { + "epoch": 2.637901127493495, + "grad_norm": 0.5795319676399231, + "learning_rate": 2.422747715073263e-06, + "loss": 0.0016, + "step": 24332 + }, + { + "epoch": 2.638009540329575, + "grad_norm": 0.5695132613182068, + "learning_rate": 2.422022341505876e-06, + "loss": 0.0042, + "step": 24333 + }, + { + "epoch": 2.6381179531656547, + "grad_norm": 0.03175148367881775, + "learning_rate": 2.4212969679384885e-06, + "loss": 0.0003, + "step": 24334 + }, + { + "epoch": 2.6382263660017347, + "grad_norm": 0.03615665063261986, + "learning_rate": 2.4205715943711015e-06, + "loss": 0.0005, + "step": 24335 + }, + { + "epoch": 2.638334778837814, + "grad_norm": 3.020718812942505, + "learning_rate": 2.419846220803714e-06, + "loss": 0.0073, + "step": 24336 + }, + { + "epoch": 2.638443191673894, + "grad_norm": 0.08797173202037811, + "learning_rate": 2.419120847236327e-06, + "loss": 0.0004, + "step": 24337 + }, + { + "epoch": 2.638551604509974, + "grad_norm": 1.3191341161727905, + "learning_rate": 2.4183954736689396e-06, + "loss": 0.0047, + "step": 24338 + }, + { + "epoch": 2.6386600173460537, + "grad_norm": 2.5559189319610596, + "learning_rate": 2.4176701001015526e-06, + "loss": 0.0105, + "step": 24339 + }, + { + "epoch": 2.6387684301821337, + "grad_norm": 0.6214781999588013, + "learning_rate": 2.416944726534165e-06, + "loss": 0.0015, + "step": 24340 + }, + { + "epoch": 2.6388768430182132, + "grad_norm": 0.23003491759300232, + "learning_rate": 2.416219352966778e-06, + "loss": 0.0011, + "step": 24341 + }, + { + "epoch": 2.638985255854293, + "grad_norm": 0.09938966482877731, + "learning_rate": 2.4154939793993907e-06, + "loss": 0.0004, + "step": 24342 + }, + { + "epoch": 2.639093668690373, + "grad_norm": 0.017001178115606308, + "learning_rate": 2.4147686058320037e-06, + "loss": 0.0001, + "step": 24343 + }, + { + "epoch": 2.6392020815264527, + "grad_norm": 0.13795489072799683, + "learning_rate": 2.4140432322646167e-06, + "loss": 0.001, + "step": 24344 + }, + { + "epoch": 2.6393104943625323, + "grad_norm": 4.359557628631592, + "learning_rate": 2.4133178586972293e-06, + "loss": 0.0061, + "step": 24345 + }, + { + "epoch": 2.6394189071986123, + "grad_norm": 0.9016037583351135, + "learning_rate": 2.4125924851298423e-06, + "loss": 0.0013, + "step": 24346 + }, + { + "epoch": 2.6395273200346923, + "grad_norm": 0.7578262686729431, + "learning_rate": 2.411867111562455e-06, + "loss": 0.0023, + "step": 24347 + }, + { + "epoch": 2.639635732870772, + "grad_norm": 0.3458806574344635, + "learning_rate": 2.411141737995068e-06, + "loss": 0.001, + "step": 24348 + }, + { + "epoch": 2.639744145706852, + "grad_norm": 0.06416682153940201, + "learning_rate": 2.4104163644276804e-06, + "loss": 0.0004, + "step": 24349 + }, + { + "epoch": 2.6398525585429313, + "grad_norm": 2.974200487136841, + "learning_rate": 2.4096909908602934e-06, + "loss": 0.0116, + "step": 24350 + }, + { + "epoch": 2.6399609713790113, + "grad_norm": 0.12792886793613434, + "learning_rate": 2.408965617292906e-06, + "loss": 0.0002, + "step": 24351 + }, + { + "epoch": 2.6400693842150913, + "grad_norm": 0.6593775153160095, + "learning_rate": 2.408240243725519e-06, + "loss": 0.0021, + "step": 24352 + }, + { + "epoch": 2.640177797051171, + "grad_norm": 4.345442295074463, + "learning_rate": 2.4075148701581315e-06, + "loss": 0.0161, + "step": 24353 + }, + { + "epoch": 2.6402862098872504, + "grad_norm": 2.3365676403045654, + "learning_rate": 2.4067894965907445e-06, + "loss": 0.0055, + "step": 24354 + }, + { + "epoch": 2.6403946227233304, + "grad_norm": 0.022721294313669205, + "learning_rate": 2.406064123023357e-06, + "loss": 0.0003, + "step": 24355 + }, + { + "epoch": 2.6405030355594103, + "grad_norm": 0.18518099188804626, + "learning_rate": 2.40533874945597e-06, + "loss": 0.0009, + "step": 24356 + }, + { + "epoch": 2.64061144839549, + "grad_norm": 0.47325193881988525, + "learning_rate": 2.404613375888583e-06, + "loss": 0.0018, + "step": 24357 + }, + { + "epoch": 2.64071986123157, + "grad_norm": 0.005271479487419128, + "learning_rate": 2.4038880023211956e-06, + "loss": 0.0001, + "step": 24358 + }, + { + "epoch": 2.6408282740676494, + "grad_norm": 5.484884262084961, + "learning_rate": 2.4031626287538086e-06, + "loss": 0.0225, + "step": 24359 + }, + { + "epoch": 2.6409366869037294, + "grad_norm": 2.7715952396392822, + "learning_rate": 2.402437255186421e-06, + "loss": 0.0432, + "step": 24360 + }, + { + "epoch": 2.6410450997398094, + "grad_norm": 0.6327748894691467, + "learning_rate": 2.401711881619034e-06, + "loss": 0.0049, + "step": 24361 + }, + { + "epoch": 2.641153512575889, + "grad_norm": 8.792044639587402, + "learning_rate": 2.4009865080516467e-06, + "loss": 0.0101, + "step": 24362 + }, + { + "epoch": 2.641261925411969, + "grad_norm": 0.996187686920166, + "learning_rate": 2.4002611344842597e-06, + "loss": 0.0489, + "step": 24363 + }, + { + "epoch": 2.6413703382480485, + "grad_norm": 0.23035715520381927, + "learning_rate": 2.3995357609168723e-06, + "loss": 0.0014, + "step": 24364 + }, + { + "epoch": 2.6414787510841284, + "grad_norm": 0.010670274496078491, + "learning_rate": 2.3988103873494853e-06, + "loss": 0.0001, + "step": 24365 + }, + { + "epoch": 2.6415871639202084, + "grad_norm": 0.046774350106716156, + "learning_rate": 2.398085013782098e-06, + "loss": 0.0001, + "step": 24366 + }, + { + "epoch": 2.641695576756288, + "grad_norm": 0.010409320704638958, + "learning_rate": 2.397359640214711e-06, + "loss": 0.0001, + "step": 24367 + }, + { + "epoch": 2.6418039895923675, + "grad_norm": 0.7093136310577393, + "learning_rate": 2.3966342666473234e-06, + "loss": 0.0016, + "step": 24368 + }, + { + "epoch": 2.6419124024284475, + "grad_norm": 0.06579176336526871, + "learning_rate": 2.3959088930799364e-06, + "loss": 0.0003, + "step": 24369 + }, + { + "epoch": 2.6420208152645275, + "grad_norm": 0.09974999725818634, + "learning_rate": 2.3951835195125494e-06, + "loss": 0.0006, + "step": 24370 + }, + { + "epoch": 2.642129228100607, + "grad_norm": 0.1468673199415207, + "learning_rate": 2.394458145945162e-06, + "loss": 0.0006, + "step": 24371 + }, + { + "epoch": 2.642237640936687, + "grad_norm": 0.004827218595892191, + "learning_rate": 2.393732772377775e-06, + "loss": 0.0, + "step": 24372 + }, + { + "epoch": 2.6423460537727665, + "grad_norm": 0.06199664995074272, + "learning_rate": 2.3930073988103875e-06, + "loss": 0.0005, + "step": 24373 + }, + { + "epoch": 2.6424544666088465, + "grad_norm": 0.13621871173381805, + "learning_rate": 2.3922820252430005e-06, + "loss": 0.001, + "step": 24374 + }, + { + "epoch": 2.6425628794449265, + "grad_norm": 12.19931411743164, + "learning_rate": 2.391556651675613e-06, + "loss": 0.0162, + "step": 24375 + }, + { + "epoch": 2.642671292281006, + "grad_norm": 1.7784491777420044, + "learning_rate": 2.390831278108226e-06, + "loss": 0.0021, + "step": 24376 + }, + { + "epoch": 2.6427797051170856, + "grad_norm": 13.54544734954834, + "learning_rate": 2.3901059045408386e-06, + "loss": 0.0498, + "step": 24377 + }, + { + "epoch": 2.6428881179531656, + "grad_norm": 1.701491117477417, + "learning_rate": 2.3893805309734516e-06, + "loss": 0.0047, + "step": 24378 + }, + { + "epoch": 2.6429965307892456, + "grad_norm": 0.07495464384555817, + "learning_rate": 2.388655157406064e-06, + "loss": 0.0005, + "step": 24379 + }, + { + "epoch": 2.643104943625325, + "grad_norm": 0.03571658581495285, + "learning_rate": 2.387929783838677e-06, + "loss": 0.0002, + "step": 24380 + }, + { + "epoch": 2.643213356461405, + "grad_norm": 5.930858612060547, + "learning_rate": 2.3872044102712897e-06, + "loss": 0.0389, + "step": 24381 + }, + { + "epoch": 2.6433217692974846, + "grad_norm": 3.1350553035736084, + "learning_rate": 2.3864790367039027e-06, + "loss": 0.007, + "step": 24382 + }, + { + "epoch": 2.6434301821335646, + "grad_norm": 4.568159580230713, + "learning_rate": 2.3857536631365157e-06, + "loss": 0.005, + "step": 24383 + }, + { + "epoch": 2.6435385949696446, + "grad_norm": 0.03194253146648407, + "learning_rate": 2.3850282895691283e-06, + "loss": 0.0003, + "step": 24384 + }, + { + "epoch": 2.643647007805724, + "grad_norm": 1.4520468711853027, + "learning_rate": 2.3843029160017413e-06, + "loss": 0.0029, + "step": 24385 + }, + { + "epoch": 2.643755420641804, + "grad_norm": 1.6503618955612183, + "learning_rate": 2.383577542434354e-06, + "loss": 0.0379, + "step": 24386 + }, + { + "epoch": 2.6438638334778837, + "grad_norm": 0.018629588186740875, + "learning_rate": 2.382852168866967e-06, + "loss": 0.0001, + "step": 24387 + }, + { + "epoch": 2.6439722463139637, + "grad_norm": 4.058461666107178, + "learning_rate": 2.3821267952995794e-06, + "loss": 0.0399, + "step": 24388 + }, + { + "epoch": 2.644080659150043, + "grad_norm": 3.022076368331909, + "learning_rate": 2.3814014217321924e-06, + "loss": 0.0089, + "step": 24389 + }, + { + "epoch": 2.644189071986123, + "grad_norm": 0.12537391483783722, + "learning_rate": 2.380676048164805e-06, + "loss": 0.0009, + "step": 24390 + }, + { + "epoch": 2.6442974848222027, + "grad_norm": 0.06880855560302734, + "learning_rate": 2.379950674597418e-06, + "loss": 0.0003, + "step": 24391 + }, + { + "epoch": 2.6444058976582827, + "grad_norm": 9.288342475891113, + "learning_rate": 2.3792253010300305e-06, + "loss": 0.0468, + "step": 24392 + }, + { + "epoch": 2.6445143104943627, + "grad_norm": 0.007043763063848019, + "learning_rate": 2.3784999274626435e-06, + "loss": 0.0, + "step": 24393 + }, + { + "epoch": 2.6446227233304422, + "grad_norm": 6.557852268218994, + "learning_rate": 2.377774553895256e-06, + "loss": 0.0247, + "step": 24394 + }, + { + "epoch": 2.6447311361665222, + "grad_norm": 1.3647078275680542, + "learning_rate": 2.377049180327869e-06, + "loss": 0.0023, + "step": 24395 + }, + { + "epoch": 2.6448395490026018, + "grad_norm": 13.085151672363281, + "learning_rate": 2.376323806760482e-06, + "loss": 0.0329, + "step": 24396 + }, + { + "epoch": 2.6449479618386817, + "grad_norm": 0.26484382152557373, + "learning_rate": 2.3755984331930946e-06, + "loss": 0.0011, + "step": 24397 + }, + { + "epoch": 2.6450563746747617, + "grad_norm": 2.95786714553833, + "learning_rate": 2.3748730596257076e-06, + "loss": 0.0341, + "step": 24398 + }, + { + "epoch": 2.6451647875108413, + "grad_norm": 0.02049904130399227, + "learning_rate": 2.37414768605832e-06, + "loss": 0.0001, + "step": 24399 + }, + { + "epoch": 2.645273200346921, + "grad_norm": 4.6568169593811035, + "learning_rate": 2.373422312490933e-06, + "loss": 0.0129, + "step": 24400 + }, + { + "epoch": 2.645381613183001, + "grad_norm": 0.029781542718410492, + "learning_rate": 2.3726969389235457e-06, + "loss": 0.0002, + "step": 24401 + }, + { + "epoch": 2.645490026019081, + "grad_norm": 0.2957752048969269, + "learning_rate": 2.3719715653561587e-06, + "loss": 0.0004, + "step": 24402 + }, + { + "epoch": 2.6455984388551603, + "grad_norm": 4.14267110824585, + "learning_rate": 2.3712461917887713e-06, + "loss": 0.0486, + "step": 24403 + }, + { + "epoch": 2.6457068516912403, + "grad_norm": 26.136873245239258, + "learning_rate": 2.3705208182213843e-06, + "loss": 0.0395, + "step": 24404 + }, + { + "epoch": 2.64581526452732, + "grad_norm": 0.029805978760123253, + "learning_rate": 2.369795444653997e-06, + "loss": 0.0003, + "step": 24405 + }, + { + "epoch": 2.6459236773634, + "grad_norm": 0.33972862362861633, + "learning_rate": 2.36907007108661e-06, + "loss": 0.0006, + "step": 24406 + }, + { + "epoch": 2.64603209019948, + "grad_norm": 1.5812175273895264, + "learning_rate": 2.3683446975192224e-06, + "loss": 0.0076, + "step": 24407 + }, + { + "epoch": 2.6461405030355594, + "grad_norm": 0.1837885081768036, + "learning_rate": 2.3676193239518354e-06, + "loss": 0.0007, + "step": 24408 + }, + { + "epoch": 2.6462489158716394, + "grad_norm": 0.029210736975073814, + "learning_rate": 2.3668939503844484e-06, + "loss": 0.0002, + "step": 24409 + }, + { + "epoch": 2.646357328707719, + "grad_norm": 1.0835932493209839, + "learning_rate": 2.366168576817061e-06, + "loss": 0.0037, + "step": 24410 + }, + { + "epoch": 2.646465741543799, + "grad_norm": 0.4177204668521881, + "learning_rate": 2.365443203249674e-06, + "loss": 0.002, + "step": 24411 + }, + { + "epoch": 2.6465741543798784, + "grad_norm": 0.9478458166122437, + "learning_rate": 2.3647178296822865e-06, + "loss": 0.0024, + "step": 24412 + }, + { + "epoch": 2.6466825672159584, + "grad_norm": 0.9434697031974792, + "learning_rate": 2.3639924561148995e-06, + "loss": 0.0034, + "step": 24413 + }, + { + "epoch": 2.646790980052038, + "grad_norm": 0.023500969633460045, + "learning_rate": 2.363267082547512e-06, + "loss": 0.0001, + "step": 24414 + }, + { + "epoch": 2.646899392888118, + "grad_norm": 3.2970328330993652, + "learning_rate": 2.362541708980125e-06, + "loss": 0.0153, + "step": 24415 + }, + { + "epoch": 2.647007805724198, + "grad_norm": 5.635837078094482, + "learning_rate": 2.3618163354127376e-06, + "loss": 0.0503, + "step": 24416 + }, + { + "epoch": 2.6471162185602775, + "grad_norm": 0.041307926177978516, + "learning_rate": 2.3610909618453506e-06, + "loss": 0.0003, + "step": 24417 + }, + { + "epoch": 2.6472246313963574, + "grad_norm": 4.085740566253662, + "learning_rate": 2.360365588277963e-06, + "loss": 0.0148, + "step": 24418 + }, + { + "epoch": 2.647333044232437, + "grad_norm": 0.14561328291893005, + "learning_rate": 2.359640214710576e-06, + "loss": 0.0012, + "step": 24419 + }, + { + "epoch": 2.647441457068517, + "grad_norm": 9.006285667419434, + "learning_rate": 2.3589148411431887e-06, + "loss": 0.0071, + "step": 24420 + }, + { + "epoch": 2.647549869904597, + "grad_norm": 0.21404185891151428, + "learning_rate": 2.3581894675758017e-06, + "loss": 0.0007, + "step": 24421 + }, + { + "epoch": 2.6476582827406765, + "grad_norm": 2.2625575065612793, + "learning_rate": 2.3574640940084147e-06, + "loss": 0.006, + "step": 24422 + }, + { + "epoch": 2.647766695576756, + "grad_norm": 0.11748562008142471, + "learning_rate": 2.3567387204410273e-06, + "loss": 0.0007, + "step": 24423 + }, + { + "epoch": 2.647875108412836, + "grad_norm": 0.024928269907832146, + "learning_rate": 2.3560133468736403e-06, + "loss": 0.0002, + "step": 24424 + }, + { + "epoch": 2.647983521248916, + "grad_norm": 15.51931095123291, + "learning_rate": 2.355287973306253e-06, + "loss": 0.0087, + "step": 24425 + }, + { + "epoch": 2.6480919340849955, + "grad_norm": 4.399986743927002, + "learning_rate": 2.354562599738866e-06, + "loss": 0.0184, + "step": 24426 + }, + { + "epoch": 2.6482003469210755, + "grad_norm": 1.6859347820281982, + "learning_rate": 2.3538372261714784e-06, + "loss": 0.0043, + "step": 24427 + }, + { + "epoch": 2.648308759757155, + "grad_norm": 3.2625081539154053, + "learning_rate": 2.3531118526040914e-06, + "loss": 0.0114, + "step": 24428 + }, + { + "epoch": 2.648417172593235, + "grad_norm": 0.012359794229269028, + "learning_rate": 2.352386479036704e-06, + "loss": 0.0001, + "step": 24429 + }, + { + "epoch": 2.648525585429315, + "grad_norm": 0.2947982847690582, + "learning_rate": 2.351661105469317e-06, + "loss": 0.0008, + "step": 24430 + }, + { + "epoch": 2.6486339982653946, + "grad_norm": 0.035536736249923706, + "learning_rate": 2.3509357319019295e-06, + "loss": 0.0004, + "step": 24431 + }, + { + "epoch": 2.6487424111014746, + "grad_norm": 0.7018653154373169, + "learning_rate": 2.3502103583345425e-06, + "loss": 0.0015, + "step": 24432 + }, + { + "epoch": 2.648850823937554, + "grad_norm": 0.1648055762052536, + "learning_rate": 2.349484984767155e-06, + "loss": 0.0006, + "step": 24433 + }, + { + "epoch": 2.648959236773634, + "grad_norm": 0.012256196700036526, + "learning_rate": 2.348759611199768e-06, + "loss": 0.0001, + "step": 24434 + }, + { + "epoch": 2.6490676496097136, + "grad_norm": 0.3309594988822937, + "learning_rate": 2.348034237632381e-06, + "loss": 0.0009, + "step": 24435 + }, + { + "epoch": 2.6491760624457936, + "grad_norm": 3.0525269508361816, + "learning_rate": 2.3473088640649936e-06, + "loss": 0.0039, + "step": 24436 + }, + { + "epoch": 2.649284475281873, + "grad_norm": 2.642603635787964, + "learning_rate": 2.3465834904976066e-06, + "loss": 0.0066, + "step": 24437 + }, + { + "epoch": 2.649392888117953, + "grad_norm": 4.809952259063721, + "learning_rate": 2.345858116930219e-06, + "loss": 0.0182, + "step": 24438 + }, + { + "epoch": 2.649501300954033, + "grad_norm": 4.125296592712402, + "learning_rate": 2.345132743362832e-06, + "loss": 0.008, + "step": 24439 + }, + { + "epoch": 2.6496097137901127, + "grad_norm": 21.824371337890625, + "learning_rate": 2.3444073697954447e-06, + "loss": 0.0176, + "step": 24440 + }, + { + "epoch": 2.6497181266261927, + "grad_norm": 0.16198739409446716, + "learning_rate": 2.3436819962280577e-06, + "loss": 0.0009, + "step": 24441 + }, + { + "epoch": 2.649826539462272, + "grad_norm": 1.2550276517868042, + "learning_rate": 2.3429566226606703e-06, + "loss": 0.0031, + "step": 24442 + }, + { + "epoch": 2.649934952298352, + "grad_norm": 0.26248034834861755, + "learning_rate": 2.3422312490932833e-06, + "loss": 0.001, + "step": 24443 + }, + { + "epoch": 2.650043365134432, + "grad_norm": 4.871161937713623, + "learning_rate": 2.341505875525896e-06, + "loss": 0.0379, + "step": 24444 + }, + { + "epoch": 2.6501517779705117, + "grad_norm": 0.1377662867307663, + "learning_rate": 2.340780501958509e-06, + "loss": 0.0003, + "step": 24445 + }, + { + "epoch": 2.6502601908065913, + "grad_norm": 0.0998632162809372, + "learning_rate": 2.3400551283911214e-06, + "loss": 0.0004, + "step": 24446 + }, + { + "epoch": 2.6503686036426712, + "grad_norm": 13.417989730834961, + "learning_rate": 2.3393297548237344e-06, + "loss": 0.0176, + "step": 24447 + }, + { + "epoch": 2.6504770164787512, + "grad_norm": 1.9871444702148438, + "learning_rate": 2.3386043812563474e-06, + "loss": 0.0061, + "step": 24448 + }, + { + "epoch": 2.6505854293148308, + "grad_norm": 4.331192970275879, + "learning_rate": 2.33787900768896e-06, + "loss": 0.0273, + "step": 24449 + }, + { + "epoch": 2.6506938421509108, + "grad_norm": 0.3636082112789154, + "learning_rate": 2.337153634121573e-06, + "loss": 0.001, + "step": 24450 + }, + { + "epoch": 2.6508022549869903, + "grad_norm": 9.236614227294922, + "learning_rate": 2.3364282605541855e-06, + "loss": 0.0479, + "step": 24451 + }, + { + "epoch": 2.6509106678230703, + "grad_norm": 3.8634135723114014, + "learning_rate": 2.3357028869867985e-06, + "loss": 0.0187, + "step": 24452 + }, + { + "epoch": 2.6510190806591503, + "grad_norm": 0.6699310541152954, + "learning_rate": 2.334977513419411e-06, + "loss": 0.001, + "step": 24453 + }, + { + "epoch": 2.65112749349523, + "grad_norm": 0.6599105000495911, + "learning_rate": 2.334252139852024e-06, + "loss": 0.0012, + "step": 24454 + }, + { + "epoch": 2.65123590633131, + "grad_norm": 0.38160693645477295, + "learning_rate": 2.3335267662846366e-06, + "loss": 0.001, + "step": 24455 + }, + { + "epoch": 2.6513443191673893, + "grad_norm": 0.14376428723335266, + "learning_rate": 2.3328013927172496e-06, + "loss": 0.0005, + "step": 24456 + }, + { + "epoch": 2.6514527320034693, + "grad_norm": 0.027883091941475868, + "learning_rate": 2.332076019149862e-06, + "loss": 0.0002, + "step": 24457 + }, + { + "epoch": 2.651561144839549, + "grad_norm": 5.544048309326172, + "learning_rate": 2.331350645582475e-06, + "loss": 0.0059, + "step": 24458 + }, + { + "epoch": 2.651669557675629, + "grad_norm": 5.341982364654541, + "learning_rate": 2.3306252720150877e-06, + "loss": 0.0247, + "step": 24459 + }, + { + "epoch": 2.6517779705117084, + "grad_norm": 0.31211113929748535, + "learning_rate": 2.3298998984477007e-06, + "loss": 0.0015, + "step": 24460 + }, + { + "epoch": 2.6518863833477884, + "grad_norm": 0.0843963772058487, + "learning_rate": 2.3291745248803137e-06, + "loss": 0.0007, + "step": 24461 + }, + { + "epoch": 2.6519947961838684, + "grad_norm": 0.014873486012220383, + "learning_rate": 2.3284491513129263e-06, + "loss": 0.0002, + "step": 24462 + }, + { + "epoch": 2.652103209019948, + "grad_norm": 0.003991160076111555, + "learning_rate": 2.3277237777455393e-06, + "loss": 0.0, + "step": 24463 + }, + { + "epoch": 2.652211621856028, + "grad_norm": 0.18330280482769012, + "learning_rate": 2.326998404178152e-06, + "loss": 0.0008, + "step": 24464 + }, + { + "epoch": 2.6523200346921074, + "grad_norm": 2.228663206100464, + "learning_rate": 2.326273030610765e-06, + "loss": 0.0455, + "step": 24465 + }, + { + "epoch": 2.6524284475281874, + "grad_norm": 0.09819180518388748, + "learning_rate": 2.3255476570433774e-06, + "loss": 0.0004, + "step": 24466 + }, + { + "epoch": 2.6525368603642674, + "grad_norm": 0.18282850086688995, + "learning_rate": 2.3248222834759904e-06, + "loss": 0.0006, + "step": 24467 + }, + { + "epoch": 2.652645273200347, + "grad_norm": 0.03538743034005165, + "learning_rate": 2.324096909908603e-06, + "loss": 0.0003, + "step": 24468 + }, + { + "epoch": 2.6527536860364265, + "grad_norm": 0.2742311358451843, + "learning_rate": 2.323371536341216e-06, + "loss": 0.0017, + "step": 24469 + }, + { + "epoch": 2.6528620988725065, + "grad_norm": 8.562432289123535, + "learning_rate": 2.3226461627738285e-06, + "loss": 0.0471, + "step": 24470 + }, + { + "epoch": 2.6529705117085864, + "grad_norm": 0.2444608062505722, + "learning_rate": 2.3219207892064415e-06, + "loss": 0.0009, + "step": 24471 + }, + { + "epoch": 2.653078924544666, + "grad_norm": 0.2848562002182007, + "learning_rate": 2.321195415639054e-06, + "loss": 0.0004, + "step": 24472 + }, + { + "epoch": 2.653187337380746, + "grad_norm": 17.66016960144043, + "learning_rate": 2.320470042071667e-06, + "loss": 0.0424, + "step": 24473 + }, + { + "epoch": 2.6532957502168255, + "grad_norm": 0.15618716180324554, + "learning_rate": 2.31974466850428e-06, + "loss": 0.0004, + "step": 24474 + }, + { + "epoch": 2.6534041630529055, + "grad_norm": 0.04292403534054756, + "learning_rate": 2.3190192949368926e-06, + "loss": 0.0002, + "step": 24475 + }, + { + "epoch": 2.6535125758889855, + "grad_norm": 0.1789282262325287, + "learning_rate": 2.3182939213695056e-06, + "loss": 0.0011, + "step": 24476 + }, + { + "epoch": 2.653620988725065, + "grad_norm": 0.2508993446826935, + "learning_rate": 2.317568547802118e-06, + "loss": 0.0013, + "step": 24477 + }, + { + "epoch": 2.6537294015611446, + "grad_norm": 0.266589492559433, + "learning_rate": 2.316843174234731e-06, + "loss": 0.001, + "step": 24478 + }, + { + "epoch": 2.6538378143972245, + "grad_norm": 2.723876476287842, + "learning_rate": 2.3161178006673437e-06, + "loss": 0.0012, + "step": 24479 + }, + { + "epoch": 2.6539462272333045, + "grad_norm": 0.006503662094473839, + "learning_rate": 2.3153924270999567e-06, + "loss": 0.0, + "step": 24480 + }, + { + "epoch": 2.654054640069384, + "grad_norm": 3.310838460922241, + "learning_rate": 2.3146670535325693e-06, + "loss": 0.0242, + "step": 24481 + }, + { + "epoch": 2.654163052905464, + "grad_norm": 0.6688703894615173, + "learning_rate": 2.3139416799651823e-06, + "loss": 0.0017, + "step": 24482 + }, + { + "epoch": 2.6542714657415436, + "grad_norm": 2.0011284351348877, + "learning_rate": 2.313216306397795e-06, + "loss": 0.0148, + "step": 24483 + }, + { + "epoch": 2.6543798785776236, + "grad_norm": 1.2345463037490845, + "learning_rate": 2.312490932830408e-06, + "loss": 0.0305, + "step": 24484 + }, + { + "epoch": 2.6544882914137036, + "grad_norm": 0.13218821585178375, + "learning_rate": 2.3117655592630204e-06, + "loss": 0.0008, + "step": 24485 + }, + { + "epoch": 2.654596704249783, + "grad_norm": 0.5822924375534058, + "learning_rate": 2.3110401856956334e-06, + "loss": 0.0028, + "step": 24486 + }, + { + "epoch": 2.654705117085863, + "grad_norm": 0.057263895869255066, + "learning_rate": 2.3103148121282464e-06, + "loss": 0.0003, + "step": 24487 + }, + { + "epoch": 2.6548135299219426, + "grad_norm": 1.9499107599258423, + "learning_rate": 2.309589438560859e-06, + "loss": 0.0121, + "step": 24488 + }, + { + "epoch": 2.6549219427580226, + "grad_norm": 0.8122075200080872, + "learning_rate": 2.308864064993472e-06, + "loss": 0.001, + "step": 24489 + }, + { + "epoch": 2.6550303555941026, + "grad_norm": 0.010198586620390415, + "learning_rate": 2.3081386914260845e-06, + "loss": 0.0001, + "step": 24490 + }, + { + "epoch": 2.655138768430182, + "grad_norm": 3.187174081802368, + "learning_rate": 2.3074133178586975e-06, + "loss": 0.0066, + "step": 24491 + }, + { + "epoch": 2.6552471812662617, + "grad_norm": 0.6749705672264099, + "learning_rate": 2.30668794429131e-06, + "loss": 0.0016, + "step": 24492 + }, + { + "epoch": 2.6553555941023417, + "grad_norm": 0.25346648693084717, + "learning_rate": 2.305962570723923e-06, + "loss": 0.0012, + "step": 24493 + }, + { + "epoch": 2.6554640069384217, + "grad_norm": 1.4360321760177612, + "learning_rate": 2.3052371971565356e-06, + "loss": 0.0022, + "step": 24494 + }, + { + "epoch": 2.655572419774501, + "grad_norm": 0.15685923397541046, + "learning_rate": 2.3045118235891486e-06, + "loss": 0.0004, + "step": 24495 + }, + { + "epoch": 2.655680832610581, + "grad_norm": 0.28694191575050354, + "learning_rate": 2.303786450021761e-06, + "loss": 0.0003, + "step": 24496 + }, + { + "epoch": 2.6557892454466607, + "grad_norm": 4.264604568481445, + "learning_rate": 2.303061076454374e-06, + "loss": 0.0289, + "step": 24497 + }, + { + "epoch": 2.6558976582827407, + "grad_norm": 1.264490008354187, + "learning_rate": 2.3023357028869867e-06, + "loss": 0.0031, + "step": 24498 + }, + { + "epoch": 2.6560060711188207, + "grad_norm": 0.8186969757080078, + "learning_rate": 2.3016103293195997e-06, + "loss": 0.0007, + "step": 24499 + }, + { + "epoch": 2.6561144839549002, + "grad_norm": 7.525094509124756, + "learning_rate": 2.3008849557522127e-06, + "loss": 0.0216, + "step": 24500 + }, + { + "epoch": 2.65622289679098, + "grad_norm": 2.055575370788574, + "learning_rate": 2.3001595821848253e-06, + "loss": 0.0037, + "step": 24501 + }, + { + "epoch": 2.6563313096270598, + "grad_norm": 0.022951707243919373, + "learning_rate": 2.2994342086174383e-06, + "loss": 0.0002, + "step": 24502 + }, + { + "epoch": 2.6564397224631398, + "grad_norm": 3.0059850215911865, + "learning_rate": 2.2987088350500513e-06, + "loss": 0.0029, + "step": 24503 + }, + { + "epoch": 2.6565481352992193, + "grad_norm": 0.16316643357276917, + "learning_rate": 2.297983461482664e-06, + "loss": 0.0006, + "step": 24504 + }, + { + "epoch": 2.6566565481352993, + "grad_norm": 3.15647029876709, + "learning_rate": 2.2972580879152764e-06, + "loss": 0.024, + "step": 24505 + }, + { + "epoch": 2.656764960971379, + "grad_norm": 2.501408576965332, + "learning_rate": 2.2965327143478894e-06, + "loss": 0.0608, + "step": 24506 + }, + { + "epoch": 2.656873373807459, + "grad_norm": 0.06353028118610382, + "learning_rate": 2.295807340780502e-06, + "loss": 0.0006, + "step": 24507 + }, + { + "epoch": 2.656981786643539, + "grad_norm": 2.1725800037384033, + "learning_rate": 2.295081967213115e-06, + "loss": 0.034, + "step": 24508 + }, + { + "epoch": 2.6570901994796183, + "grad_norm": 6.80984354019165, + "learning_rate": 2.2943565936457275e-06, + "loss": 0.012, + "step": 24509 + }, + { + "epoch": 2.6571986123156983, + "grad_norm": 1.8189904689788818, + "learning_rate": 2.2936312200783405e-06, + "loss": 0.0153, + "step": 24510 + }, + { + "epoch": 2.657307025151778, + "grad_norm": 0.004707351326942444, + "learning_rate": 2.292905846510953e-06, + "loss": 0.0001, + "step": 24511 + }, + { + "epoch": 2.657415437987858, + "grad_norm": 0.3885191082954407, + "learning_rate": 2.292180472943566e-06, + "loss": 0.0006, + "step": 24512 + }, + { + "epoch": 2.657523850823938, + "grad_norm": 3.132861375808716, + "learning_rate": 2.291455099376179e-06, + "loss": 0.041, + "step": 24513 + }, + { + "epoch": 2.6576322636600174, + "grad_norm": 1.0558608770370483, + "learning_rate": 2.290729725808792e-06, + "loss": 0.001, + "step": 24514 + }, + { + "epoch": 2.657740676496097, + "grad_norm": 0.1407509297132492, + "learning_rate": 2.2900043522414046e-06, + "loss": 0.001, + "step": 24515 + }, + { + "epoch": 2.657849089332177, + "grad_norm": 14.423991203308105, + "learning_rate": 2.2892789786740176e-06, + "loss": 0.0178, + "step": 24516 + }, + { + "epoch": 2.657957502168257, + "grad_norm": 2.042301893234253, + "learning_rate": 2.28855360510663e-06, + "loss": 0.0011, + "step": 24517 + }, + { + "epoch": 2.6580659150043364, + "grad_norm": 0.14501367509365082, + "learning_rate": 2.2878282315392427e-06, + "loss": 0.0007, + "step": 24518 + }, + { + "epoch": 2.6581743278404164, + "grad_norm": 4.45037841796875, + "learning_rate": 2.2871028579718557e-06, + "loss": 0.0201, + "step": 24519 + }, + { + "epoch": 2.658282740676496, + "grad_norm": 1.7913256883621216, + "learning_rate": 2.2863774844044683e-06, + "loss": 0.0543, + "step": 24520 + }, + { + "epoch": 2.658391153512576, + "grad_norm": 4.338158130645752, + "learning_rate": 2.2856521108370813e-06, + "loss": 0.0262, + "step": 24521 + }, + { + "epoch": 2.658499566348656, + "grad_norm": 12.59195613861084, + "learning_rate": 2.284926737269694e-06, + "loss": 0.0061, + "step": 24522 + }, + { + "epoch": 2.6586079791847355, + "grad_norm": 0.3154245615005493, + "learning_rate": 2.284201363702307e-06, + "loss": 0.0008, + "step": 24523 + }, + { + "epoch": 2.658716392020815, + "grad_norm": 0.29393118619918823, + "learning_rate": 2.2834759901349194e-06, + "loss": 0.0011, + "step": 24524 + }, + { + "epoch": 2.658824804856895, + "grad_norm": 0.02312697470188141, + "learning_rate": 2.2827506165675324e-06, + "loss": 0.0002, + "step": 24525 + }, + { + "epoch": 2.658933217692975, + "grad_norm": 0.5501700639724731, + "learning_rate": 2.2820252430001454e-06, + "loss": 0.002, + "step": 24526 + }, + { + "epoch": 2.6590416305290545, + "grad_norm": 0.07174219936132431, + "learning_rate": 2.2812998694327584e-06, + "loss": 0.0004, + "step": 24527 + }, + { + "epoch": 2.6591500433651345, + "grad_norm": 0.8614675998687744, + "learning_rate": 2.280574495865371e-06, + "loss": 0.0039, + "step": 24528 + }, + { + "epoch": 2.659258456201214, + "grad_norm": 0.389494925737381, + "learning_rate": 2.279849122297984e-06, + "loss": 0.0018, + "step": 24529 + }, + { + "epoch": 2.659366869037294, + "grad_norm": 0.08492225408554077, + "learning_rate": 2.2791237487305965e-06, + "loss": 0.0005, + "step": 24530 + }, + { + "epoch": 2.659475281873374, + "grad_norm": 2.0026323795318604, + "learning_rate": 2.278398375163209e-06, + "loss": 0.0045, + "step": 24531 + }, + { + "epoch": 2.6595836947094535, + "grad_norm": 0.3803599178791046, + "learning_rate": 2.277673001595822e-06, + "loss": 0.0007, + "step": 24532 + }, + { + "epoch": 2.6596921075455335, + "grad_norm": 0.36321648955345154, + "learning_rate": 2.2769476280284346e-06, + "loss": 0.0017, + "step": 24533 + }, + { + "epoch": 2.659800520381613, + "grad_norm": 0.017064537853002548, + "learning_rate": 2.2762222544610476e-06, + "loss": 0.0001, + "step": 24534 + }, + { + "epoch": 2.659908933217693, + "grad_norm": 1.3499943017959595, + "learning_rate": 2.27549688089366e-06, + "loss": 0.0035, + "step": 24535 + }, + { + "epoch": 2.6600173460537726, + "grad_norm": 3.2656970024108887, + "learning_rate": 2.274771507326273e-06, + "loss": 0.005, + "step": 24536 + }, + { + "epoch": 2.6601257588898526, + "grad_norm": 6.286787033081055, + "learning_rate": 2.2740461337588857e-06, + "loss": 0.0113, + "step": 24537 + }, + { + "epoch": 2.660234171725932, + "grad_norm": 10.467555046081543, + "learning_rate": 2.2733207601914987e-06, + "loss": 0.0084, + "step": 24538 + }, + { + "epoch": 2.660342584562012, + "grad_norm": 1.0157514810562134, + "learning_rate": 2.2725953866241117e-06, + "loss": 0.0017, + "step": 24539 + }, + { + "epoch": 2.660450997398092, + "grad_norm": 0.3472788333892822, + "learning_rate": 2.2718700130567247e-06, + "loss": 0.0008, + "step": 24540 + }, + { + "epoch": 2.6605594102341716, + "grad_norm": 0.03245626389980316, + "learning_rate": 2.2711446394893373e-06, + "loss": 0.0004, + "step": 24541 + }, + { + "epoch": 2.6606678230702516, + "grad_norm": 0.6687833070755005, + "learning_rate": 2.2704192659219503e-06, + "loss": 0.0016, + "step": 24542 + }, + { + "epoch": 2.660776235906331, + "grad_norm": 0.07192642986774445, + "learning_rate": 2.269693892354563e-06, + "loss": 0.0002, + "step": 24543 + }, + { + "epoch": 2.660884648742411, + "grad_norm": 0.165268212556839, + "learning_rate": 2.2689685187871754e-06, + "loss": 0.0006, + "step": 24544 + }, + { + "epoch": 2.660993061578491, + "grad_norm": 0.06402410566806793, + "learning_rate": 2.2682431452197884e-06, + "loss": 0.0002, + "step": 24545 + }, + { + "epoch": 2.6611014744145707, + "grad_norm": 4.3556647300720215, + "learning_rate": 2.267517771652401e-06, + "loss": 0.0018, + "step": 24546 + }, + { + "epoch": 2.66120988725065, + "grad_norm": 1.1823699474334717, + "learning_rate": 2.266792398085014e-06, + "loss": 0.0021, + "step": 24547 + }, + { + "epoch": 2.66131830008673, + "grad_norm": 2.4756033420562744, + "learning_rate": 2.2660670245176265e-06, + "loss": 0.0129, + "step": 24548 + }, + { + "epoch": 2.66142671292281, + "grad_norm": 0.6667988896369934, + "learning_rate": 2.2653416509502395e-06, + "loss": 0.0015, + "step": 24549 + }, + { + "epoch": 2.6615351257588897, + "grad_norm": 1.9160254001617432, + "learning_rate": 2.264616277382852e-06, + "loss": 0.0035, + "step": 24550 + }, + { + "epoch": 2.6616435385949697, + "grad_norm": 4.955742835998535, + "learning_rate": 2.263890903815465e-06, + "loss": 0.0165, + "step": 24551 + }, + { + "epoch": 2.6617519514310493, + "grad_norm": 0.015188895165920258, + "learning_rate": 2.263165530248078e-06, + "loss": 0.0001, + "step": 24552 + }, + { + "epoch": 2.6618603642671292, + "grad_norm": 0.03087012469768524, + "learning_rate": 2.262440156680691e-06, + "loss": 0.0002, + "step": 24553 + }, + { + "epoch": 2.6619687771032092, + "grad_norm": 0.11036337167024612, + "learning_rate": 2.2617147831133036e-06, + "loss": 0.0011, + "step": 24554 + }, + { + "epoch": 2.6620771899392888, + "grad_norm": 0.21415570378303528, + "learning_rate": 2.2609894095459166e-06, + "loss": 0.0011, + "step": 24555 + }, + { + "epoch": 2.6621856027753688, + "grad_norm": 0.6116686463356018, + "learning_rate": 2.260264035978529e-06, + "loss": 0.0007, + "step": 24556 + }, + { + "epoch": 2.6622940156114483, + "grad_norm": 0.5793582797050476, + "learning_rate": 2.259538662411142e-06, + "loss": 0.0024, + "step": 24557 + }, + { + "epoch": 2.6624024284475283, + "grad_norm": 1.0365185737609863, + "learning_rate": 2.2588132888437547e-06, + "loss": 0.0035, + "step": 24558 + }, + { + "epoch": 2.662510841283608, + "grad_norm": 0.10352573543787003, + "learning_rate": 2.2580879152763673e-06, + "loss": 0.0008, + "step": 24559 + }, + { + "epoch": 2.662619254119688, + "grad_norm": 9.99291706085205, + "learning_rate": 2.2573625417089803e-06, + "loss": 0.0056, + "step": 24560 + }, + { + "epoch": 2.6627276669557673, + "grad_norm": 1.9965075254440308, + "learning_rate": 2.256637168141593e-06, + "loss": 0.0077, + "step": 24561 + }, + { + "epoch": 2.6628360797918473, + "grad_norm": 0.025758517906069756, + "learning_rate": 2.255911794574206e-06, + "loss": 0.0001, + "step": 24562 + }, + { + "epoch": 2.6629444926279273, + "grad_norm": 1.8976430892944336, + "learning_rate": 2.255186421006819e-06, + "loss": 0.0456, + "step": 24563 + }, + { + "epoch": 2.663052905464007, + "grad_norm": 8.577408790588379, + "learning_rate": 2.2544610474394314e-06, + "loss": 0.0127, + "step": 24564 + }, + { + "epoch": 2.663161318300087, + "grad_norm": 0.028559992089867592, + "learning_rate": 2.2537356738720444e-06, + "loss": 0.0002, + "step": 24565 + }, + { + "epoch": 2.6632697311361664, + "grad_norm": 0.08763763308525085, + "learning_rate": 2.2530103003046574e-06, + "loss": 0.0008, + "step": 24566 + }, + { + "epoch": 2.6633781439722464, + "grad_norm": 1.463056206703186, + "learning_rate": 2.25228492673727e-06, + "loss": 0.0048, + "step": 24567 + }, + { + "epoch": 2.6634865568083264, + "grad_norm": 0.03968790918588638, + "learning_rate": 2.251559553169883e-06, + "loss": 0.0002, + "step": 24568 + }, + { + "epoch": 2.663594969644406, + "grad_norm": 7.8829026222229, + "learning_rate": 2.2508341796024955e-06, + "loss": 0.0309, + "step": 24569 + }, + { + "epoch": 2.6637033824804854, + "grad_norm": 0.04187208041548729, + "learning_rate": 2.2501088060351085e-06, + "loss": 0.0003, + "step": 24570 + }, + { + "epoch": 2.6638117953165654, + "grad_norm": 0.04681907966732979, + "learning_rate": 2.249383432467721e-06, + "loss": 0.0004, + "step": 24571 + }, + { + "epoch": 2.6639202081526454, + "grad_norm": 0.037183940410614014, + "learning_rate": 2.2486580589003336e-06, + "loss": 0.0001, + "step": 24572 + }, + { + "epoch": 2.664028620988725, + "grad_norm": 7.300706386566162, + "learning_rate": 2.2479326853329466e-06, + "loss": 0.0046, + "step": 24573 + }, + { + "epoch": 2.664137033824805, + "grad_norm": 13.744421005249023, + "learning_rate": 2.247207311765559e-06, + "loss": 0.0245, + "step": 24574 + }, + { + "epoch": 2.6642454466608845, + "grad_norm": 2.2933812141418457, + "learning_rate": 2.246481938198172e-06, + "loss": 0.0174, + "step": 24575 + }, + { + "epoch": 2.6643538594969645, + "grad_norm": 3.283031702041626, + "learning_rate": 2.245756564630785e-06, + "loss": 0.0086, + "step": 24576 + }, + { + "epoch": 2.6644622723330444, + "grad_norm": 10.565449714660645, + "learning_rate": 2.2450311910633977e-06, + "loss": 0.0523, + "step": 24577 + }, + { + "epoch": 2.664570685169124, + "grad_norm": 4.777799129486084, + "learning_rate": 2.2443058174960107e-06, + "loss": 0.0077, + "step": 24578 + }, + { + "epoch": 2.664679098005204, + "grad_norm": 6.467898845672607, + "learning_rate": 2.2435804439286237e-06, + "loss": 0.0192, + "step": 24579 + }, + { + "epoch": 2.6647875108412835, + "grad_norm": 0.02131008356809616, + "learning_rate": 2.2428550703612363e-06, + "loss": 0.0002, + "step": 24580 + }, + { + "epoch": 2.6648959236773635, + "grad_norm": 0.1450444608926773, + "learning_rate": 2.2421296967938493e-06, + "loss": 0.0005, + "step": 24581 + }, + { + "epoch": 2.665004336513443, + "grad_norm": 0.06062959134578705, + "learning_rate": 2.241404323226462e-06, + "loss": 0.0006, + "step": 24582 + }, + { + "epoch": 2.665112749349523, + "grad_norm": 3.622959852218628, + "learning_rate": 2.240678949659075e-06, + "loss": 0.0309, + "step": 24583 + }, + { + "epoch": 2.6652211621856026, + "grad_norm": 0.07609258592128754, + "learning_rate": 2.2399535760916874e-06, + "loss": 0.0003, + "step": 24584 + }, + { + "epoch": 2.6653295750216826, + "grad_norm": 1.0067386627197266, + "learning_rate": 2.2392282025243e-06, + "loss": 0.0023, + "step": 24585 + }, + { + "epoch": 2.6654379878577625, + "grad_norm": 10.14189338684082, + "learning_rate": 2.238502828956913e-06, + "loss": 0.0371, + "step": 24586 + }, + { + "epoch": 2.665546400693842, + "grad_norm": 1.3031560182571411, + "learning_rate": 2.2377774553895255e-06, + "loss": 0.0053, + "step": 24587 + }, + { + "epoch": 2.665654813529922, + "grad_norm": 0.12961620092391968, + "learning_rate": 2.2370520818221385e-06, + "loss": 0.0008, + "step": 24588 + }, + { + "epoch": 2.6657632263660016, + "grad_norm": 2.2803242206573486, + "learning_rate": 2.2363267082547515e-06, + "loss": 0.0128, + "step": 24589 + }, + { + "epoch": 2.6658716392020816, + "grad_norm": 2.95224928855896, + "learning_rate": 2.235601334687364e-06, + "loss": 0.0084, + "step": 24590 + }, + { + "epoch": 2.6659800520381616, + "grad_norm": 0.021597178652882576, + "learning_rate": 2.234875961119977e-06, + "loss": 0.0001, + "step": 24591 + }, + { + "epoch": 2.666088464874241, + "grad_norm": 0.39972683787345886, + "learning_rate": 2.23415058755259e-06, + "loss": 0.0026, + "step": 24592 + }, + { + "epoch": 2.6661968777103207, + "grad_norm": 1.404660701751709, + "learning_rate": 2.2334252139852026e-06, + "loss": 0.0022, + "step": 24593 + }, + { + "epoch": 2.6663052905464006, + "grad_norm": 7.275085926055908, + "learning_rate": 2.2326998404178156e-06, + "loss": 0.0374, + "step": 24594 + }, + { + "epoch": 2.6664137033824806, + "grad_norm": 0.5798134803771973, + "learning_rate": 2.231974466850428e-06, + "loss": 0.001, + "step": 24595 + }, + { + "epoch": 2.66652211621856, + "grad_norm": 1.7586009502410889, + "learning_rate": 2.231249093283041e-06, + "loss": 0.0035, + "step": 24596 + }, + { + "epoch": 2.66663052905464, + "grad_norm": 1.177790641784668, + "learning_rate": 2.2305237197156537e-06, + "loss": 0.0042, + "step": 24597 + }, + { + "epoch": 2.6667389418907197, + "grad_norm": 0.29781845211982727, + "learning_rate": 2.2297983461482663e-06, + "loss": 0.0007, + "step": 24598 + }, + { + "epoch": 2.6668473547267997, + "grad_norm": 0.21236754953861237, + "learning_rate": 2.2290729725808793e-06, + "loss": 0.0009, + "step": 24599 + }, + { + "epoch": 2.6669557675628797, + "grad_norm": 1.1237852573394775, + "learning_rate": 2.228347599013492e-06, + "loss": 0.0021, + "step": 24600 + }, + { + "epoch": 2.667064180398959, + "grad_norm": 0.028349557891488075, + "learning_rate": 2.227622225446105e-06, + "loss": 0.0002, + "step": 24601 + }, + { + "epoch": 2.667172593235039, + "grad_norm": 0.03057469241321087, + "learning_rate": 2.226896851878718e-06, + "loss": 0.0002, + "step": 24602 + }, + { + "epoch": 2.6672810060711187, + "grad_norm": 2.9531421661376953, + "learning_rate": 2.2261714783113304e-06, + "loss": 0.0289, + "step": 24603 + }, + { + "epoch": 2.6673894189071987, + "grad_norm": 7.070066928863525, + "learning_rate": 2.2254461047439434e-06, + "loss": 0.0355, + "step": 24604 + }, + { + "epoch": 2.6674978317432783, + "grad_norm": 7.0981903076171875, + "learning_rate": 2.2247207311765564e-06, + "loss": 0.0084, + "step": 24605 + }, + { + "epoch": 2.6676062445793582, + "grad_norm": 3.742652416229248, + "learning_rate": 2.223995357609169e-06, + "loss": 0.0111, + "step": 24606 + }, + { + "epoch": 2.667714657415438, + "grad_norm": 7.0333170890808105, + "learning_rate": 2.223269984041782e-06, + "loss": 0.0067, + "step": 24607 + }, + { + "epoch": 2.6678230702515178, + "grad_norm": 2.045543670654297, + "learning_rate": 2.2225446104743945e-06, + "loss": 0.0219, + "step": 24608 + }, + { + "epoch": 2.6679314830875978, + "grad_norm": 0.02136567048728466, + "learning_rate": 2.2218192369070075e-06, + "loss": 0.0001, + "step": 24609 + }, + { + "epoch": 2.6680398959236773, + "grad_norm": 3.225044012069702, + "learning_rate": 2.22109386333962e-06, + "loss": 0.0238, + "step": 24610 + }, + { + "epoch": 2.6681483087597573, + "grad_norm": 0.5142294764518738, + "learning_rate": 2.220368489772233e-06, + "loss": 0.0016, + "step": 24611 + }, + { + "epoch": 2.668256721595837, + "grad_norm": 0.2741697430610657, + "learning_rate": 2.2196431162048456e-06, + "loss": 0.0007, + "step": 24612 + }, + { + "epoch": 2.668365134431917, + "grad_norm": 0.3831062316894531, + "learning_rate": 2.218917742637458e-06, + "loss": 0.0014, + "step": 24613 + }, + { + "epoch": 2.668473547267997, + "grad_norm": 3.2852845191955566, + "learning_rate": 2.218192369070071e-06, + "loss": 0.0255, + "step": 24614 + }, + { + "epoch": 2.6685819601040763, + "grad_norm": 0.11739760637283325, + "learning_rate": 2.217466995502684e-06, + "loss": 0.0004, + "step": 24615 + }, + { + "epoch": 2.668690372940156, + "grad_norm": 0.029055774211883545, + "learning_rate": 2.2167416219352967e-06, + "loss": 0.0001, + "step": 24616 + }, + { + "epoch": 2.668798785776236, + "grad_norm": 0.5633889436721802, + "learning_rate": 2.2160162483679097e-06, + "loss": 0.0015, + "step": 24617 + }, + { + "epoch": 2.668907198612316, + "grad_norm": 0.5053237080574036, + "learning_rate": 2.2152908748005227e-06, + "loss": 0.0018, + "step": 24618 + }, + { + "epoch": 2.6690156114483954, + "grad_norm": 0.40897467732429504, + "learning_rate": 2.2145655012331353e-06, + "loss": 0.0018, + "step": 24619 + }, + { + "epoch": 2.6691240242844754, + "grad_norm": 6.698975563049316, + "learning_rate": 2.2138401276657483e-06, + "loss": 0.014, + "step": 24620 + }, + { + "epoch": 2.669232437120555, + "grad_norm": 0.06428024917840958, + "learning_rate": 2.213114754098361e-06, + "loss": 0.0005, + "step": 24621 + }, + { + "epoch": 2.669340849956635, + "grad_norm": 1.101189136505127, + "learning_rate": 2.212389380530974e-06, + "loss": 0.0021, + "step": 24622 + }, + { + "epoch": 2.669449262792715, + "grad_norm": 3.590590715408325, + "learning_rate": 2.2116640069635864e-06, + "loss": 0.0117, + "step": 24623 + }, + { + "epoch": 2.6695576756287944, + "grad_norm": 0.06170498579740524, + "learning_rate": 2.2109386333961994e-06, + "loss": 0.0002, + "step": 24624 + }, + { + "epoch": 2.669666088464874, + "grad_norm": 1.7908987998962402, + "learning_rate": 2.210213259828812e-06, + "loss": 0.0194, + "step": 24625 + }, + { + "epoch": 2.669774501300954, + "grad_norm": 8.630441665649414, + "learning_rate": 2.2094878862614245e-06, + "loss": 0.0143, + "step": 24626 + }, + { + "epoch": 2.669882914137034, + "grad_norm": 2.4132940769195557, + "learning_rate": 2.2087625126940375e-06, + "loss": 0.0116, + "step": 24627 + }, + { + "epoch": 2.6699913269731135, + "grad_norm": 1.165867805480957, + "learning_rate": 2.2080371391266505e-06, + "loss": 0.0034, + "step": 24628 + }, + { + "epoch": 2.6700997398091935, + "grad_norm": 0.047856248915195465, + "learning_rate": 2.207311765559263e-06, + "loss": 0.0004, + "step": 24629 + }, + { + "epoch": 2.670208152645273, + "grad_norm": 0.006782818585634232, + "learning_rate": 2.206586391991876e-06, + "loss": 0.0, + "step": 24630 + }, + { + "epoch": 2.670316565481353, + "grad_norm": 0.15779581665992737, + "learning_rate": 2.205861018424489e-06, + "loss": 0.0004, + "step": 24631 + }, + { + "epoch": 2.670424978317433, + "grad_norm": 1.316948652267456, + "learning_rate": 2.2051356448571016e-06, + "loss": 0.0073, + "step": 24632 + }, + { + "epoch": 2.6705333911535125, + "grad_norm": 0.34306371212005615, + "learning_rate": 2.2044102712897146e-06, + "loss": 0.0004, + "step": 24633 + }, + { + "epoch": 2.6706418039895925, + "grad_norm": 2.4139437675476074, + "learning_rate": 2.203684897722327e-06, + "loss": 0.0042, + "step": 24634 + }, + { + "epoch": 2.670750216825672, + "grad_norm": 1.842239260673523, + "learning_rate": 2.20295952415494e-06, + "loss": 0.0392, + "step": 24635 + }, + { + "epoch": 2.670858629661752, + "grad_norm": 0.174677312374115, + "learning_rate": 2.2022341505875527e-06, + "loss": 0.0006, + "step": 24636 + }, + { + "epoch": 2.670967042497832, + "grad_norm": 0.13158604502677917, + "learning_rate": 2.2015087770201657e-06, + "loss": 0.0003, + "step": 24637 + }, + { + "epoch": 2.6710754553339116, + "grad_norm": 1.8656927347183228, + "learning_rate": 2.2007834034527783e-06, + "loss": 0.0331, + "step": 24638 + }, + { + "epoch": 2.671183868169991, + "grad_norm": 0.8723618984222412, + "learning_rate": 2.200058029885391e-06, + "loss": 0.0021, + "step": 24639 + }, + { + "epoch": 2.671292281006071, + "grad_norm": 4.442853927612305, + "learning_rate": 2.199332656318004e-06, + "loss": 0.0242, + "step": 24640 + }, + { + "epoch": 2.671400693842151, + "grad_norm": 0.07226825505495071, + "learning_rate": 2.198607282750617e-06, + "loss": 0.0006, + "step": 24641 + }, + { + "epoch": 2.6715091066782306, + "grad_norm": 0.9099981188774109, + "learning_rate": 2.1978819091832294e-06, + "loss": 0.0021, + "step": 24642 + }, + { + "epoch": 2.6716175195143106, + "grad_norm": 0.36672595143318176, + "learning_rate": 2.1971565356158424e-06, + "loss": 0.0004, + "step": 24643 + }, + { + "epoch": 2.67172593235039, + "grad_norm": 1.2169904708862305, + "learning_rate": 2.1964311620484554e-06, + "loss": 0.0387, + "step": 24644 + }, + { + "epoch": 2.67183434518647, + "grad_norm": 1.9093762636184692, + "learning_rate": 2.195705788481068e-06, + "loss": 0.029, + "step": 24645 + }, + { + "epoch": 2.67194275802255, + "grad_norm": 5.69031286239624, + "learning_rate": 2.194980414913681e-06, + "loss": 0.0255, + "step": 24646 + }, + { + "epoch": 2.6720511708586296, + "grad_norm": 22.382705688476562, + "learning_rate": 2.1942550413462935e-06, + "loss": 0.0277, + "step": 24647 + }, + { + "epoch": 2.672159583694709, + "grad_norm": 0.836571455001831, + "learning_rate": 2.1935296677789065e-06, + "loss": 0.003, + "step": 24648 + }, + { + "epoch": 2.672267996530789, + "grad_norm": 3.0763418674468994, + "learning_rate": 2.192804294211519e-06, + "loss": 0.0341, + "step": 24649 + }, + { + "epoch": 2.672376409366869, + "grad_norm": 0.24685855209827423, + "learning_rate": 2.192078920644132e-06, + "loss": 0.0004, + "step": 24650 + }, + { + "epoch": 2.6724848222029487, + "grad_norm": 0.5460277199745178, + "learning_rate": 2.1913535470767446e-06, + "loss": 0.0013, + "step": 24651 + }, + { + "epoch": 2.6725932350390287, + "grad_norm": 1.1201280355453491, + "learning_rate": 2.1906281735093576e-06, + "loss": 0.0042, + "step": 24652 + }, + { + "epoch": 2.672701647875108, + "grad_norm": 2.232980251312256, + "learning_rate": 2.18990279994197e-06, + "loss": 0.0083, + "step": 24653 + }, + { + "epoch": 2.672810060711188, + "grad_norm": 1.2236430644989014, + "learning_rate": 2.189177426374583e-06, + "loss": 0.011, + "step": 24654 + }, + { + "epoch": 2.672918473547268, + "grad_norm": 3.1310408115386963, + "learning_rate": 2.1884520528071957e-06, + "loss": 0.0497, + "step": 24655 + }, + { + "epoch": 2.6730268863833477, + "grad_norm": 8.043804168701172, + "learning_rate": 2.1877266792398087e-06, + "loss": 0.0236, + "step": 24656 + }, + { + "epoch": 2.6731352992194277, + "grad_norm": 0.5993661284446716, + "learning_rate": 2.1870013056724217e-06, + "loss": 0.0023, + "step": 24657 + }, + { + "epoch": 2.6732437120555073, + "grad_norm": 2.366401433944702, + "learning_rate": 2.1862759321050343e-06, + "loss": 0.0089, + "step": 24658 + }, + { + "epoch": 2.6733521248915872, + "grad_norm": 2.8031907081604004, + "learning_rate": 2.1855505585376473e-06, + "loss": 0.0028, + "step": 24659 + }, + { + "epoch": 2.6734605377276672, + "grad_norm": 0.007366080768406391, + "learning_rate": 2.18482518497026e-06, + "loss": 0.0, + "step": 24660 + }, + { + "epoch": 2.6735689505637468, + "grad_norm": 0.03491806611418724, + "learning_rate": 2.184099811402873e-06, + "loss": 0.0002, + "step": 24661 + }, + { + "epoch": 2.6736773633998263, + "grad_norm": 0.10056544840335846, + "learning_rate": 2.1833744378354854e-06, + "loss": 0.0007, + "step": 24662 + }, + { + "epoch": 2.6737857762359063, + "grad_norm": 0.018498113378882408, + "learning_rate": 2.1826490642680984e-06, + "loss": 0.0001, + "step": 24663 + }, + { + "epoch": 2.6738941890719863, + "grad_norm": 0.464790016412735, + "learning_rate": 2.181923690700711e-06, + "loss": 0.0006, + "step": 24664 + }, + { + "epoch": 2.674002601908066, + "grad_norm": 0.09670479595661163, + "learning_rate": 2.181198317133324e-06, + "loss": 0.0011, + "step": 24665 + }, + { + "epoch": 2.674111014744146, + "grad_norm": 0.14289745688438416, + "learning_rate": 2.1804729435659365e-06, + "loss": 0.0007, + "step": 24666 + }, + { + "epoch": 2.6742194275802254, + "grad_norm": 2.980508804321289, + "learning_rate": 2.1797475699985495e-06, + "loss": 0.0037, + "step": 24667 + }, + { + "epoch": 2.6743278404163053, + "grad_norm": 6.654005527496338, + "learning_rate": 2.179022196431162e-06, + "loss": 0.0051, + "step": 24668 + }, + { + "epoch": 2.6744362532523853, + "grad_norm": 2.7978293895721436, + "learning_rate": 2.178296822863775e-06, + "loss": 0.0242, + "step": 24669 + }, + { + "epoch": 2.674544666088465, + "grad_norm": 0.1179128885269165, + "learning_rate": 2.177571449296388e-06, + "loss": 0.0007, + "step": 24670 + }, + { + "epoch": 2.6746530789245444, + "grad_norm": 0.054399047046899796, + "learning_rate": 2.1768460757290006e-06, + "loss": 0.0004, + "step": 24671 + }, + { + "epoch": 2.6747614917606244, + "grad_norm": 2.0808112621307373, + "learning_rate": 2.1761207021616136e-06, + "loss": 0.0039, + "step": 24672 + }, + { + "epoch": 2.6748699045967044, + "grad_norm": 1.8456392288208008, + "learning_rate": 2.175395328594226e-06, + "loss": 0.0217, + "step": 24673 + }, + { + "epoch": 2.674978317432784, + "grad_norm": 0.038572702556848526, + "learning_rate": 2.174669955026839e-06, + "loss": 0.0003, + "step": 24674 + }, + { + "epoch": 2.675086730268864, + "grad_norm": 0.0059914062730968, + "learning_rate": 2.1739445814594517e-06, + "loss": 0.0001, + "step": 24675 + }, + { + "epoch": 2.6751951431049434, + "grad_norm": 0.09531830996274948, + "learning_rate": 2.1732192078920647e-06, + "loss": 0.0007, + "step": 24676 + }, + { + "epoch": 2.6753035559410234, + "grad_norm": 3.714862823486328, + "learning_rate": 2.1724938343246773e-06, + "loss": 0.0032, + "step": 24677 + }, + { + "epoch": 2.6754119687771034, + "grad_norm": 14.825431823730469, + "learning_rate": 2.1717684607572903e-06, + "loss": 0.0796, + "step": 24678 + }, + { + "epoch": 2.675520381613183, + "grad_norm": 3.3747241497039795, + "learning_rate": 2.171043087189903e-06, + "loss": 0.0296, + "step": 24679 + }, + { + "epoch": 2.675628794449263, + "grad_norm": 1.0597097873687744, + "learning_rate": 2.170317713622516e-06, + "loss": 0.0462, + "step": 24680 + }, + { + "epoch": 2.6757372072853425, + "grad_norm": 0.18706758320331573, + "learning_rate": 2.1695923400551284e-06, + "loss": 0.0001, + "step": 24681 + }, + { + "epoch": 2.6758456201214225, + "grad_norm": 1.1265194416046143, + "learning_rate": 2.1688669664877414e-06, + "loss": 0.0117, + "step": 24682 + }, + { + "epoch": 2.675954032957502, + "grad_norm": 0.13535575568675995, + "learning_rate": 2.1681415929203544e-06, + "loss": 0.0006, + "step": 24683 + }, + { + "epoch": 2.676062445793582, + "grad_norm": 0.6646108627319336, + "learning_rate": 2.167416219352967e-06, + "loss": 0.0024, + "step": 24684 + }, + { + "epoch": 2.6761708586296615, + "grad_norm": 0.04209219291806221, + "learning_rate": 2.16669084578558e-06, + "loss": 0.0004, + "step": 24685 + }, + { + "epoch": 2.6762792714657415, + "grad_norm": 2.6298155784606934, + "learning_rate": 2.1659654722181925e-06, + "loss": 0.0052, + "step": 24686 + }, + { + "epoch": 2.6763876843018215, + "grad_norm": 0.02218487486243248, + "learning_rate": 2.1652400986508055e-06, + "loss": 0.0002, + "step": 24687 + }, + { + "epoch": 2.676496097137901, + "grad_norm": 0.16941972076892853, + "learning_rate": 2.164514725083418e-06, + "loss": 0.0007, + "step": 24688 + }, + { + "epoch": 2.676604509973981, + "grad_norm": 4.3530144691467285, + "learning_rate": 2.163789351516031e-06, + "loss": 0.0489, + "step": 24689 + }, + { + "epoch": 2.6767129228100606, + "grad_norm": 0.7262647151947021, + "learning_rate": 2.1630639779486436e-06, + "loss": 0.0016, + "step": 24690 + }, + { + "epoch": 2.6768213356461406, + "grad_norm": 0.18569473922252655, + "learning_rate": 2.1623386043812566e-06, + "loss": 0.0008, + "step": 24691 + }, + { + "epoch": 2.6769297484822205, + "grad_norm": 3.45415997505188, + "learning_rate": 2.161613230813869e-06, + "loss": 0.009, + "step": 24692 + }, + { + "epoch": 2.6770381613183, + "grad_norm": 0.012034005485475063, + "learning_rate": 2.160887857246482e-06, + "loss": 0.0001, + "step": 24693 + }, + { + "epoch": 2.6771465741543796, + "grad_norm": 0.1239970251917839, + "learning_rate": 2.1601624836790947e-06, + "loss": 0.0005, + "step": 24694 + }, + { + "epoch": 2.6772549869904596, + "grad_norm": 2.5924081802368164, + "learning_rate": 2.1594371101117077e-06, + "loss": 0.0165, + "step": 24695 + }, + { + "epoch": 2.6773633998265396, + "grad_norm": 3.6629889011383057, + "learning_rate": 2.1587117365443207e-06, + "loss": 0.0195, + "step": 24696 + }, + { + "epoch": 2.677471812662619, + "grad_norm": 4.5921311378479, + "learning_rate": 2.1579863629769333e-06, + "loss": 0.0071, + "step": 24697 + }, + { + "epoch": 2.677580225498699, + "grad_norm": 0.11957083642482758, + "learning_rate": 2.1572609894095463e-06, + "loss": 0.0008, + "step": 24698 + }, + { + "epoch": 2.6776886383347787, + "grad_norm": 0.1604892611503601, + "learning_rate": 2.156535615842159e-06, + "loss": 0.0007, + "step": 24699 + }, + { + "epoch": 2.6777970511708586, + "grad_norm": 1.5084089040756226, + "learning_rate": 2.155810242274772e-06, + "loss": 0.0041, + "step": 24700 + }, + { + "epoch": 2.6779054640069386, + "grad_norm": 11.45145034790039, + "learning_rate": 2.1550848687073844e-06, + "loss": 0.0473, + "step": 24701 + }, + { + "epoch": 2.678013876843018, + "grad_norm": 0.3273624777793884, + "learning_rate": 2.1543594951399974e-06, + "loss": 0.001, + "step": 24702 + }, + { + "epoch": 2.678122289679098, + "grad_norm": 2.3068058490753174, + "learning_rate": 2.15363412157261e-06, + "loss": 0.0175, + "step": 24703 + }, + { + "epoch": 2.6782307025151777, + "grad_norm": 7.207469463348389, + "learning_rate": 2.152908748005223e-06, + "loss": 0.0068, + "step": 24704 + }, + { + "epoch": 2.6783391153512577, + "grad_norm": 6.522165298461914, + "learning_rate": 2.1521833744378355e-06, + "loss": 0.0158, + "step": 24705 + }, + { + "epoch": 2.6784475281873372, + "grad_norm": 4.657563209533691, + "learning_rate": 2.1514580008704485e-06, + "loss": 0.0106, + "step": 24706 + }, + { + "epoch": 2.678555941023417, + "grad_norm": 0.4004858434200287, + "learning_rate": 2.150732627303061e-06, + "loss": 0.003, + "step": 24707 + }, + { + "epoch": 2.6786643538594967, + "grad_norm": 0.014117243699729443, + "learning_rate": 2.150007253735674e-06, + "loss": 0.0001, + "step": 24708 + }, + { + "epoch": 2.6787727666955767, + "grad_norm": 0.17558600008487701, + "learning_rate": 2.149281880168287e-06, + "loss": 0.0007, + "step": 24709 + }, + { + "epoch": 2.6788811795316567, + "grad_norm": 2.104668378829956, + "learning_rate": 2.1485565066008996e-06, + "loss": 0.0289, + "step": 24710 + }, + { + "epoch": 2.6789895923677363, + "grad_norm": 0.13242337107658386, + "learning_rate": 2.1478311330335126e-06, + "loss": 0.0002, + "step": 24711 + }, + { + "epoch": 2.6790980052038162, + "grad_norm": 8.083820343017578, + "learning_rate": 2.147105759466125e-06, + "loss": 0.0066, + "step": 24712 + }, + { + "epoch": 2.679206418039896, + "grad_norm": 3.2544384002685547, + "learning_rate": 2.146380385898738e-06, + "loss": 0.0146, + "step": 24713 + }, + { + "epoch": 2.6793148308759758, + "grad_norm": 1.8912277221679688, + "learning_rate": 2.1456550123313507e-06, + "loss": 0.0028, + "step": 24714 + }, + { + "epoch": 2.6794232437120558, + "grad_norm": 0.050833385437726974, + "learning_rate": 2.1449296387639637e-06, + "loss": 0.0003, + "step": 24715 + }, + { + "epoch": 2.6795316565481353, + "grad_norm": 1.2117999792099, + "learning_rate": 2.1442042651965763e-06, + "loss": 0.0056, + "step": 24716 + }, + { + "epoch": 2.679640069384215, + "grad_norm": 2.182460069656372, + "learning_rate": 2.1434788916291893e-06, + "loss": 0.019, + "step": 24717 + }, + { + "epoch": 2.679748482220295, + "grad_norm": 0.11022651940584183, + "learning_rate": 2.142753518061802e-06, + "loss": 0.0008, + "step": 24718 + }, + { + "epoch": 2.679856895056375, + "grad_norm": 0.10308574885129929, + "learning_rate": 2.142028144494415e-06, + "loss": 0.0006, + "step": 24719 + }, + { + "epoch": 2.6799653078924544, + "grad_norm": 2.037996292114258, + "learning_rate": 2.1413027709270274e-06, + "loss": 0.0143, + "step": 24720 + }, + { + "epoch": 2.6800737207285343, + "grad_norm": 0.01317291334271431, + "learning_rate": 2.1405773973596404e-06, + "loss": 0.0002, + "step": 24721 + }, + { + "epoch": 2.680182133564614, + "grad_norm": 0.03169567137956619, + "learning_rate": 2.1398520237922534e-06, + "loss": 0.0003, + "step": 24722 + }, + { + "epoch": 2.680290546400694, + "grad_norm": 7.304134845733643, + "learning_rate": 2.139126650224866e-06, + "loss": 0.0161, + "step": 24723 + }, + { + "epoch": 2.680398959236774, + "grad_norm": 0.2959238886833191, + "learning_rate": 2.138401276657479e-06, + "loss": 0.0008, + "step": 24724 + }, + { + "epoch": 2.6805073720728534, + "grad_norm": 0.08420561999082565, + "learning_rate": 2.1376759030900915e-06, + "loss": 0.0009, + "step": 24725 + }, + { + "epoch": 2.6806157849089334, + "grad_norm": 1.7165589332580566, + "learning_rate": 2.1369505295227045e-06, + "loss": 0.0268, + "step": 24726 + }, + { + "epoch": 2.680724197745013, + "grad_norm": 0.024888327345252037, + "learning_rate": 2.136225155955317e-06, + "loss": 0.0002, + "step": 24727 + }, + { + "epoch": 2.680832610581093, + "grad_norm": 1.2431979179382324, + "learning_rate": 2.13549978238793e-06, + "loss": 0.0035, + "step": 24728 + }, + { + "epoch": 2.6809410234171724, + "grad_norm": 0.15341365337371826, + "learning_rate": 2.1347744088205426e-06, + "loss": 0.0009, + "step": 24729 + }, + { + "epoch": 2.6810494362532524, + "grad_norm": 0.20683586597442627, + "learning_rate": 2.1340490352531556e-06, + "loss": 0.0008, + "step": 24730 + }, + { + "epoch": 2.681157849089332, + "grad_norm": 0.01324024610221386, + "learning_rate": 2.133323661685768e-06, + "loss": 0.0001, + "step": 24731 + }, + { + "epoch": 2.681266261925412, + "grad_norm": 5.499157905578613, + "learning_rate": 2.132598288118381e-06, + "loss": 0.0608, + "step": 24732 + }, + { + "epoch": 2.681374674761492, + "grad_norm": 3.5891330242156982, + "learning_rate": 2.1318729145509937e-06, + "loss": 0.0453, + "step": 24733 + }, + { + "epoch": 2.6814830875975715, + "grad_norm": 0.7699668407440186, + "learning_rate": 2.1311475409836067e-06, + "loss": 0.0036, + "step": 24734 + }, + { + "epoch": 2.6815915004336515, + "grad_norm": 3.685642719268799, + "learning_rate": 2.1304221674162197e-06, + "loss": 0.0277, + "step": 24735 + }, + { + "epoch": 2.681699913269731, + "grad_norm": 5.889967441558838, + "learning_rate": 2.1296967938488323e-06, + "loss": 0.0093, + "step": 24736 + }, + { + "epoch": 2.681808326105811, + "grad_norm": 2.6535356044769287, + "learning_rate": 2.1289714202814453e-06, + "loss": 0.0054, + "step": 24737 + }, + { + "epoch": 2.681916738941891, + "grad_norm": 0.44634655117988586, + "learning_rate": 2.128246046714058e-06, + "loss": 0.0007, + "step": 24738 + }, + { + "epoch": 2.6820251517779705, + "grad_norm": 0.1450607031583786, + "learning_rate": 2.127520673146671e-06, + "loss": 0.0009, + "step": 24739 + }, + { + "epoch": 2.68213356461405, + "grad_norm": 0.009172733873128891, + "learning_rate": 2.1267952995792834e-06, + "loss": 0.0001, + "step": 24740 + }, + { + "epoch": 2.68224197745013, + "grad_norm": 0.049053169786930084, + "learning_rate": 2.1260699260118964e-06, + "loss": 0.0002, + "step": 24741 + }, + { + "epoch": 2.68235039028621, + "grad_norm": 0.27725446224212646, + "learning_rate": 2.125344552444509e-06, + "loss": 0.0008, + "step": 24742 + }, + { + "epoch": 2.6824588031222896, + "grad_norm": 0.8095142245292664, + "learning_rate": 2.124619178877122e-06, + "loss": 0.002, + "step": 24743 + }, + { + "epoch": 2.6825672159583696, + "grad_norm": 3.0899949073791504, + "learning_rate": 2.1238938053097345e-06, + "loss": 0.0334, + "step": 24744 + }, + { + "epoch": 2.682675628794449, + "grad_norm": 0.9174472689628601, + "learning_rate": 2.1231684317423475e-06, + "loss": 0.0018, + "step": 24745 + }, + { + "epoch": 2.682784041630529, + "grad_norm": 0.6486953496932983, + "learning_rate": 2.12244305817496e-06, + "loss": 0.0014, + "step": 24746 + }, + { + "epoch": 2.682892454466609, + "grad_norm": 2.18426775932312, + "learning_rate": 2.121717684607573e-06, + "loss": 0.0047, + "step": 24747 + }, + { + "epoch": 2.6830008673026886, + "grad_norm": 3.6421875953674316, + "learning_rate": 2.120992311040186e-06, + "loss": 0.0384, + "step": 24748 + }, + { + "epoch": 2.6831092801387686, + "grad_norm": 4.106327533721924, + "learning_rate": 2.1202669374727986e-06, + "loss": 0.0105, + "step": 24749 + }, + { + "epoch": 2.683217692974848, + "grad_norm": 0.31574684381484985, + "learning_rate": 2.1195415639054116e-06, + "loss": 0.0008, + "step": 24750 + }, + { + "epoch": 2.683326105810928, + "grad_norm": 0.8241278529167175, + "learning_rate": 2.118816190338024e-06, + "loss": 0.0025, + "step": 24751 + }, + { + "epoch": 2.6834345186470077, + "grad_norm": 3.1416687965393066, + "learning_rate": 2.118090816770637e-06, + "loss": 0.0273, + "step": 24752 + }, + { + "epoch": 2.6835429314830876, + "grad_norm": 0.3013346791267395, + "learning_rate": 2.1173654432032497e-06, + "loss": 0.0015, + "step": 24753 + }, + { + "epoch": 2.683651344319167, + "grad_norm": 1.2937238216400146, + "learning_rate": 2.1166400696358627e-06, + "loss": 0.0037, + "step": 24754 + }, + { + "epoch": 2.683759757155247, + "grad_norm": 0.09850909560918808, + "learning_rate": 2.1159146960684753e-06, + "loss": 0.0002, + "step": 24755 + }, + { + "epoch": 2.683868169991327, + "grad_norm": 0.16447311639785767, + "learning_rate": 2.1151893225010883e-06, + "loss": 0.001, + "step": 24756 + }, + { + "epoch": 2.6839765828274067, + "grad_norm": 0.046773165464401245, + "learning_rate": 2.114463948933701e-06, + "loss": 0.0002, + "step": 24757 + }, + { + "epoch": 2.6840849956634867, + "grad_norm": 0.49846234917640686, + "learning_rate": 2.113738575366314e-06, + "loss": 0.0006, + "step": 24758 + }, + { + "epoch": 2.6841934084995662, + "grad_norm": 1.717913269996643, + "learning_rate": 2.1130132017989264e-06, + "loss": 0.0268, + "step": 24759 + }, + { + "epoch": 2.684301821335646, + "grad_norm": 6.061157703399658, + "learning_rate": 2.1122878282315394e-06, + "loss": 0.0582, + "step": 24760 + }, + { + "epoch": 2.684410234171726, + "grad_norm": 1.2742878198623657, + "learning_rate": 2.1115624546641524e-06, + "loss": 0.0009, + "step": 24761 + }, + { + "epoch": 2.6845186470078057, + "grad_norm": 0.22205378115177155, + "learning_rate": 2.110837081096765e-06, + "loss": 0.0008, + "step": 24762 + }, + { + "epoch": 2.6846270598438853, + "grad_norm": 2.7171378135681152, + "learning_rate": 2.110111707529378e-06, + "loss": 0.0274, + "step": 24763 + }, + { + "epoch": 2.6847354726799653, + "grad_norm": 7.430253982543945, + "learning_rate": 2.1093863339619905e-06, + "loss": 0.0275, + "step": 24764 + }, + { + "epoch": 2.6848438855160452, + "grad_norm": 0.12981948256492615, + "learning_rate": 2.1086609603946035e-06, + "loss": 0.0005, + "step": 24765 + }, + { + "epoch": 2.684952298352125, + "grad_norm": 2.2007622718811035, + "learning_rate": 2.107935586827216e-06, + "loss": 0.0049, + "step": 24766 + }, + { + "epoch": 2.6850607111882048, + "grad_norm": 0.054038189351558685, + "learning_rate": 2.107210213259829e-06, + "loss": 0.0002, + "step": 24767 + }, + { + "epoch": 2.6851691240242843, + "grad_norm": 0.05527050793170929, + "learning_rate": 2.1064848396924416e-06, + "loss": 0.0004, + "step": 24768 + }, + { + "epoch": 2.6852775368603643, + "grad_norm": 4.405053615570068, + "learning_rate": 2.1057594661250546e-06, + "loss": 0.0088, + "step": 24769 + }, + { + "epoch": 2.6853859496964443, + "grad_norm": 0.28806859254837036, + "learning_rate": 2.105034092557667e-06, + "loss": 0.0006, + "step": 24770 + }, + { + "epoch": 2.685494362532524, + "grad_norm": 0.17437519133090973, + "learning_rate": 2.10430871899028e-06, + "loss": 0.0008, + "step": 24771 + }, + { + "epoch": 2.6856027753686034, + "grad_norm": 0.04205983504652977, + "learning_rate": 2.1035833454228927e-06, + "loss": 0.0004, + "step": 24772 + }, + { + "epoch": 2.6857111882046834, + "grad_norm": 0.5203530788421631, + "learning_rate": 2.1028579718555057e-06, + "loss": 0.0011, + "step": 24773 + }, + { + "epoch": 2.6858196010407633, + "grad_norm": 0.7730057239532471, + "learning_rate": 2.1021325982881187e-06, + "loss": 0.0017, + "step": 24774 + }, + { + "epoch": 2.685928013876843, + "grad_norm": 4.054991722106934, + "learning_rate": 2.1014072247207313e-06, + "loss": 0.0251, + "step": 24775 + }, + { + "epoch": 2.686036426712923, + "grad_norm": 0.49351340532302856, + "learning_rate": 2.1006818511533443e-06, + "loss": 0.0015, + "step": 24776 + }, + { + "epoch": 2.6861448395490024, + "grad_norm": 0.04430963471531868, + "learning_rate": 2.099956477585957e-06, + "loss": 0.0003, + "step": 24777 + }, + { + "epoch": 2.6862532523850824, + "grad_norm": 2.290275812149048, + "learning_rate": 2.09923110401857e-06, + "loss": 0.0133, + "step": 24778 + }, + { + "epoch": 2.6863616652211624, + "grad_norm": 1.5240036249160767, + "learning_rate": 2.0985057304511824e-06, + "loss": 0.0333, + "step": 24779 + }, + { + "epoch": 2.686470078057242, + "grad_norm": 0.05546494200825691, + "learning_rate": 2.0977803568837954e-06, + "loss": 0.0002, + "step": 24780 + }, + { + "epoch": 2.686578490893322, + "grad_norm": 0.03745713457465172, + "learning_rate": 2.097054983316408e-06, + "loss": 0.0003, + "step": 24781 + }, + { + "epoch": 2.6866869037294014, + "grad_norm": 4.333979606628418, + "learning_rate": 2.096329609749021e-06, + "loss": 0.0331, + "step": 24782 + }, + { + "epoch": 2.6867953165654814, + "grad_norm": 2.361988067626953, + "learning_rate": 2.0956042361816335e-06, + "loss": 0.0226, + "step": 24783 + }, + { + "epoch": 2.6869037294015614, + "grad_norm": 0.04192462936043739, + "learning_rate": 2.0948788626142465e-06, + "loss": 0.0004, + "step": 24784 + }, + { + "epoch": 2.687012142237641, + "grad_norm": 3.289900541305542, + "learning_rate": 2.094153489046859e-06, + "loss": 0.0027, + "step": 24785 + }, + { + "epoch": 2.6871205550737205, + "grad_norm": 8.154207229614258, + "learning_rate": 2.093428115479472e-06, + "loss": 0.067, + "step": 24786 + }, + { + "epoch": 2.6872289679098005, + "grad_norm": 5.707393169403076, + "learning_rate": 2.092702741912085e-06, + "loss": 0.0356, + "step": 24787 + }, + { + "epoch": 2.6873373807458805, + "grad_norm": 0.19298186898231506, + "learning_rate": 2.0919773683446976e-06, + "loss": 0.0009, + "step": 24788 + }, + { + "epoch": 2.68744579358196, + "grad_norm": 0.1530885398387909, + "learning_rate": 2.0912519947773106e-06, + "loss": 0.001, + "step": 24789 + }, + { + "epoch": 2.68755420641804, + "grad_norm": 3.0183823108673096, + "learning_rate": 2.090526621209923e-06, + "loss": 0.0443, + "step": 24790 + }, + { + "epoch": 2.6876626192541195, + "grad_norm": 5.096798419952393, + "learning_rate": 2.089801247642536e-06, + "loss": 0.0301, + "step": 24791 + }, + { + "epoch": 2.6877710320901995, + "grad_norm": 0.06812180578708649, + "learning_rate": 2.0890758740751487e-06, + "loss": 0.0004, + "step": 24792 + }, + { + "epoch": 2.6878794449262795, + "grad_norm": 0.11972064524888992, + "learning_rate": 2.0883505005077617e-06, + "loss": 0.001, + "step": 24793 + }, + { + "epoch": 2.687987857762359, + "grad_norm": 0.09687559306621552, + "learning_rate": 2.0876251269403743e-06, + "loss": 0.0006, + "step": 24794 + }, + { + "epoch": 2.6880962705984386, + "grad_norm": 5.7185282707214355, + "learning_rate": 2.0868997533729873e-06, + "loss": 0.0039, + "step": 24795 + }, + { + "epoch": 2.6882046834345186, + "grad_norm": 1.9775563478469849, + "learning_rate": 2.0861743798056e-06, + "loss": 0.002, + "step": 24796 + }, + { + "epoch": 2.6883130962705986, + "grad_norm": 0.12136320024728775, + "learning_rate": 2.085449006238213e-06, + "loss": 0.0009, + "step": 24797 + }, + { + "epoch": 2.688421509106678, + "grad_norm": 0.05260765552520752, + "learning_rate": 2.0847236326708254e-06, + "loss": 0.0004, + "step": 24798 + }, + { + "epoch": 2.688529921942758, + "grad_norm": 6.327090263366699, + "learning_rate": 2.0839982591034384e-06, + "loss": 0.0033, + "step": 24799 + }, + { + "epoch": 2.6886383347788376, + "grad_norm": 17.685897827148438, + "learning_rate": 2.0832728855360514e-06, + "loss": 0.0047, + "step": 24800 + }, + { + "epoch": 2.6887467476149176, + "grad_norm": 0.6834630966186523, + "learning_rate": 2.082547511968664e-06, + "loss": 0.0025, + "step": 24801 + }, + { + "epoch": 2.6888551604509976, + "grad_norm": 0.0666714683175087, + "learning_rate": 2.081822138401277e-06, + "loss": 0.0004, + "step": 24802 + }, + { + "epoch": 2.688963573287077, + "grad_norm": 2.8840537071228027, + "learning_rate": 2.0810967648338895e-06, + "loss": 0.0211, + "step": 24803 + }, + { + "epoch": 2.689071986123157, + "grad_norm": 0.046789780259132385, + "learning_rate": 2.0803713912665025e-06, + "loss": 0.0002, + "step": 24804 + }, + { + "epoch": 2.6891803989592367, + "grad_norm": 0.013630896806716919, + "learning_rate": 2.079646017699115e-06, + "loss": 0.0001, + "step": 24805 + }, + { + "epoch": 2.6892888117953166, + "grad_norm": 0.35220351815223694, + "learning_rate": 2.078920644131728e-06, + "loss": 0.0011, + "step": 24806 + }, + { + "epoch": 2.6893972246313966, + "grad_norm": 0.08078151941299438, + "learning_rate": 2.0781952705643406e-06, + "loss": 0.0003, + "step": 24807 + }, + { + "epoch": 2.689505637467476, + "grad_norm": 1.1296733617782593, + "learning_rate": 2.0774698969969536e-06, + "loss": 0.0008, + "step": 24808 + }, + { + "epoch": 2.6896140503035557, + "grad_norm": 7.513167381286621, + "learning_rate": 2.076744523429566e-06, + "loss": 0.011, + "step": 24809 + }, + { + "epoch": 2.6897224631396357, + "grad_norm": 0.06282518804073334, + "learning_rate": 2.076019149862179e-06, + "loss": 0.0004, + "step": 24810 + }, + { + "epoch": 2.6898308759757157, + "grad_norm": 2.275679588317871, + "learning_rate": 2.0752937762947917e-06, + "loss": 0.0295, + "step": 24811 + }, + { + "epoch": 2.6899392888117952, + "grad_norm": 2.8560540676116943, + "learning_rate": 2.0745684027274047e-06, + "loss": 0.0305, + "step": 24812 + }, + { + "epoch": 2.690047701647875, + "grad_norm": 10.060979843139648, + "learning_rate": 2.0738430291600177e-06, + "loss": 0.0119, + "step": 24813 + }, + { + "epoch": 2.6901561144839548, + "grad_norm": 5.828256130218506, + "learning_rate": 2.0731176555926303e-06, + "loss": 0.035, + "step": 24814 + }, + { + "epoch": 2.6902645273200347, + "grad_norm": 3.298031806945801, + "learning_rate": 2.0723922820252433e-06, + "loss": 0.0035, + "step": 24815 + }, + { + "epoch": 2.6903729401561147, + "grad_norm": 0.5726021528244019, + "learning_rate": 2.071666908457856e-06, + "loss": 0.0019, + "step": 24816 + }, + { + "epoch": 2.6904813529921943, + "grad_norm": 8.514517784118652, + "learning_rate": 2.070941534890469e-06, + "loss": 0.0212, + "step": 24817 + }, + { + "epoch": 2.690589765828274, + "grad_norm": 1.0069262981414795, + "learning_rate": 2.0702161613230814e-06, + "loss": 0.0024, + "step": 24818 + }, + { + "epoch": 2.690698178664354, + "grad_norm": 0.8679126501083374, + "learning_rate": 2.0694907877556944e-06, + "loss": 0.0026, + "step": 24819 + }, + { + "epoch": 2.6908065915004338, + "grad_norm": 0.02056766487658024, + "learning_rate": 2.068765414188307e-06, + "loss": 0.0002, + "step": 24820 + }, + { + "epoch": 2.6909150043365133, + "grad_norm": 0.05200397968292236, + "learning_rate": 2.06804004062092e-06, + "loss": 0.0003, + "step": 24821 + }, + { + "epoch": 2.6910234171725933, + "grad_norm": 0.10836753994226456, + "learning_rate": 2.0673146670535325e-06, + "loss": 0.0007, + "step": 24822 + }, + { + "epoch": 2.691131830008673, + "grad_norm": 0.9288218021392822, + "learning_rate": 2.0665892934861455e-06, + "loss": 0.003, + "step": 24823 + }, + { + "epoch": 2.691240242844753, + "grad_norm": 0.27080684900283813, + "learning_rate": 2.065863919918758e-06, + "loss": 0.0005, + "step": 24824 + }, + { + "epoch": 2.691348655680833, + "grad_norm": 5.714078903198242, + "learning_rate": 2.065138546351371e-06, + "loss": 0.0052, + "step": 24825 + }, + { + "epoch": 2.6914570685169124, + "grad_norm": 0.09216438978910446, + "learning_rate": 2.064413172783984e-06, + "loss": 0.0002, + "step": 24826 + }, + { + "epoch": 2.6915654813529923, + "grad_norm": 0.03794896602630615, + "learning_rate": 2.063687799216597e-06, + "loss": 0.0003, + "step": 24827 + }, + { + "epoch": 2.691673894189072, + "grad_norm": 4.917944431304932, + "learning_rate": 2.0629624256492096e-06, + "loss": 0.0506, + "step": 24828 + }, + { + "epoch": 2.691782307025152, + "grad_norm": 0.03736380487680435, + "learning_rate": 2.062237052081822e-06, + "loss": 0.0003, + "step": 24829 + }, + { + "epoch": 2.6918907198612314, + "grad_norm": 2.1774637699127197, + "learning_rate": 2.061511678514435e-06, + "loss": 0.0057, + "step": 24830 + }, + { + "epoch": 2.6919991326973114, + "grad_norm": 0.03075243905186653, + "learning_rate": 2.0607863049470477e-06, + "loss": 0.0003, + "step": 24831 + }, + { + "epoch": 2.692107545533391, + "grad_norm": 0.01920873112976551, + "learning_rate": 2.0600609313796607e-06, + "loss": 0.0001, + "step": 24832 + }, + { + "epoch": 2.692215958369471, + "grad_norm": 2.479318141937256, + "learning_rate": 2.0593355578122733e-06, + "loss": 0.0199, + "step": 24833 + }, + { + "epoch": 2.692324371205551, + "grad_norm": 0.04834697023034096, + "learning_rate": 2.0586101842448863e-06, + "loss": 0.0003, + "step": 24834 + }, + { + "epoch": 2.6924327840416304, + "grad_norm": 0.07418975979089737, + "learning_rate": 2.057884810677499e-06, + "loss": 0.0004, + "step": 24835 + }, + { + "epoch": 2.6925411968777104, + "grad_norm": 2.152219295501709, + "learning_rate": 2.057159437110112e-06, + "loss": 0.0021, + "step": 24836 + }, + { + "epoch": 2.69264960971379, + "grad_norm": 0.6903537511825562, + "learning_rate": 2.0564340635427244e-06, + "loss": 0.0021, + "step": 24837 + }, + { + "epoch": 2.69275802254987, + "grad_norm": 3.233893871307373, + "learning_rate": 2.0557086899753374e-06, + "loss": 0.0072, + "step": 24838 + }, + { + "epoch": 2.69286643538595, + "grad_norm": 2.270695686340332, + "learning_rate": 2.0549833164079504e-06, + "loss": 0.0062, + "step": 24839 + }, + { + "epoch": 2.6929748482220295, + "grad_norm": 0.07088292390108109, + "learning_rate": 2.0542579428405634e-06, + "loss": 0.0003, + "step": 24840 + }, + { + "epoch": 2.693083261058109, + "grad_norm": 3.4077742099761963, + "learning_rate": 2.053532569273176e-06, + "loss": 0.0402, + "step": 24841 + }, + { + "epoch": 2.693191673894189, + "grad_norm": 0.14715950191020966, + "learning_rate": 2.0528071957057885e-06, + "loss": 0.0004, + "step": 24842 + }, + { + "epoch": 2.693300086730269, + "grad_norm": 0.04645087569952011, + "learning_rate": 2.0520818221384015e-06, + "loss": 0.0003, + "step": 24843 + }, + { + "epoch": 2.6934084995663485, + "grad_norm": 4.13987922668457, + "learning_rate": 2.051356448571014e-06, + "loss": 0.0319, + "step": 24844 + }, + { + "epoch": 2.6935169124024285, + "grad_norm": 5.948329448699951, + "learning_rate": 2.050631075003627e-06, + "loss": 0.0554, + "step": 24845 + }, + { + "epoch": 2.693625325238508, + "grad_norm": 4.407208442687988, + "learning_rate": 2.0499057014362396e-06, + "loss": 0.0084, + "step": 24846 + }, + { + "epoch": 2.693733738074588, + "grad_norm": 2.590949058532715, + "learning_rate": 2.0491803278688526e-06, + "loss": 0.0039, + "step": 24847 + }, + { + "epoch": 2.693842150910668, + "grad_norm": 0.1476660817861557, + "learning_rate": 2.048454954301465e-06, + "loss": 0.0009, + "step": 24848 + }, + { + "epoch": 2.6939505637467476, + "grad_norm": 2.334124803543091, + "learning_rate": 2.047729580734078e-06, + "loss": 0.0113, + "step": 24849 + }, + { + "epoch": 2.6940589765828276, + "grad_norm": 2.9710350036621094, + "learning_rate": 2.0470042071666908e-06, + "loss": 0.0186, + "step": 24850 + }, + { + "epoch": 2.694167389418907, + "grad_norm": 10.668034553527832, + "learning_rate": 2.0462788335993037e-06, + "loss": 0.0049, + "step": 24851 + }, + { + "epoch": 2.694275802254987, + "grad_norm": 3.167888641357422, + "learning_rate": 2.0455534600319167e-06, + "loss": 0.0328, + "step": 24852 + }, + { + "epoch": 2.6943842150910666, + "grad_norm": 10.027278900146484, + "learning_rate": 2.0448280864645297e-06, + "loss": 0.0264, + "step": 24853 + }, + { + "epoch": 2.6944926279271466, + "grad_norm": 0.47648558020591736, + "learning_rate": 2.0441027128971423e-06, + "loss": 0.0013, + "step": 24854 + }, + { + "epoch": 2.694601040763226, + "grad_norm": 4.349597930908203, + "learning_rate": 2.0433773393297553e-06, + "loss": 0.0102, + "step": 24855 + }, + { + "epoch": 2.694709453599306, + "grad_norm": 0.16222909092903137, + "learning_rate": 2.042651965762368e-06, + "loss": 0.0007, + "step": 24856 + }, + { + "epoch": 2.694817866435386, + "grad_norm": 2.355353355407715, + "learning_rate": 2.0419265921949804e-06, + "loss": 0.002, + "step": 24857 + }, + { + "epoch": 2.6949262792714657, + "grad_norm": 0.1101708859205246, + "learning_rate": 2.0412012186275934e-06, + "loss": 0.0006, + "step": 24858 + }, + { + "epoch": 2.6950346921075456, + "grad_norm": 4.2117815017700195, + "learning_rate": 2.040475845060206e-06, + "loss": 0.0362, + "step": 24859 + }, + { + "epoch": 2.695143104943625, + "grad_norm": 0.07540721446275711, + "learning_rate": 2.039750471492819e-06, + "loss": 0.0005, + "step": 24860 + }, + { + "epoch": 2.695251517779705, + "grad_norm": 1.884120225906372, + "learning_rate": 2.0390250979254315e-06, + "loss": 0.0054, + "step": 24861 + }, + { + "epoch": 2.695359930615785, + "grad_norm": 4.77868127822876, + "learning_rate": 2.0382997243580445e-06, + "loss": 0.025, + "step": 24862 + }, + { + "epoch": 2.6954683434518647, + "grad_norm": 2.8431413173675537, + "learning_rate": 2.037574350790657e-06, + "loss": 0.036, + "step": 24863 + }, + { + "epoch": 2.6955767562879442, + "grad_norm": 0.38303396105766296, + "learning_rate": 2.03684897722327e-06, + "loss": 0.0004, + "step": 24864 + }, + { + "epoch": 2.6956851691240242, + "grad_norm": 0.7793504595756531, + "learning_rate": 2.036123603655883e-06, + "loss": 0.0016, + "step": 24865 + }, + { + "epoch": 2.695793581960104, + "grad_norm": 0.05742126703262329, + "learning_rate": 2.035398230088496e-06, + "loss": 0.0005, + "step": 24866 + }, + { + "epoch": 2.6959019947961838, + "grad_norm": 3.5473291873931885, + "learning_rate": 2.0346728565211086e-06, + "loss": 0.0471, + "step": 24867 + }, + { + "epoch": 2.6960104076322637, + "grad_norm": 6.676717281341553, + "learning_rate": 2.0339474829537216e-06, + "loss": 0.0484, + "step": 24868 + }, + { + "epoch": 2.6961188204683433, + "grad_norm": 0.4352557361125946, + "learning_rate": 2.033222109386334e-06, + "loss": 0.0016, + "step": 24869 + }, + { + "epoch": 2.6962272333044233, + "grad_norm": 0.4794057309627533, + "learning_rate": 2.0324967358189467e-06, + "loss": 0.0018, + "step": 24870 + }, + { + "epoch": 2.6963356461405033, + "grad_norm": 4.652782440185547, + "learning_rate": 2.0317713622515597e-06, + "loss": 0.0178, + "step": 24871 + }, + { + "epoch": 2.696444058976583, + "grad_norm": 0.17877568304538727, + "learning_rate": 2.0310459886841723e-06, + "loss": 0.0006, + "step": 24872 + }, + { + "epoch": 2.6965524718126628, + "grad_norm": 0.16585680842399597, + "learning_rate": 2.0303206151167853e-06, + "loss": 0.0003, + "step": 24873 + }, + { + "epoch": 2.6966608846487423, + "grad_norm": 0.13519853353500366, + "learning_rate": 2.029595241549398e-06, + "loss": 0.001, + "step": 24874 + }, + { + "epoch": 2.6967692974848223, + "grad_norm": 1.703169345855713, + "learning_rate": 2.028869867982011e-06, + "loss": 0.047, + "step": 24875 + }, + { + "epoch": 2.696877710320902, + "grad_norm": 0.13834764063358307, + "learning_rate": 2.028144494414624e-06, + "loss": 0.0012, + "step": 24876 + }, + { + "epoch": 2.696986123156982, + "grad_norm": 0.07536491751670837, + "learning_rate": 2.0274191208472364e-06, + "loss": 0.0004, + "step": 24877 + }, + { + "epoch": 2.6970945359930614, + "grad_norm": 0.07789649069309235, + "learning_rate": 2.0266937472798494e-06, + "loss": 0.0007, + "step": 24878 + }, + { + "epoch": 2.6972029488291414, + "grad_norm": 0.46585071086883545, + "learning_rate": 2.0259683737124624e-06, + "loss": 0.0016, + "step": 24879 + }, + { + "epoch": 2.6973113616652213, + "grad_norm": 6.282087326049805, + "learning_rate": 2.025243000145075e-06, + "loss": 0.0219, + "step": 24880 + }, + { + "epoch": 2.697419774501301, + "grad_norm": 0.07240636646747589, + "learning_rate": 2.024517626577688e-06, + "loss": 0.0004, + "step": 24881 + }, + { + "epoch": 2.697528187337381, + "grad_norm": 15.103816986083984, + "learning_rate": 2.0237922530103005e-06, + "loss": 0.0425, + "step": 24882 + }, + { + "epoch": 2.6976366001734604, + "grad_norm": 0.03285818547010422, + "learning_rate": 2.023066879442913e-06, + "loss": 0.0003, + "step": 24883 + }, + { + "epoch": 2.6977450130095404, + "grad_norm": 1.826608419418335, + "learning_rate": 2.022341505875526e-06, + "loss": 0.0033, + "step": 24884 + }, + { + "epoch": 2.6978534258456204, + "grad_norm": 0.9103164672851562, + "learning_rate": 2.0216161323081386e-06, + "loss": 0.0021, + "step": 24885 + }, + { + "epoch": 2.6979618386817, + "grad_norm": 2.6725146770477295, + "learning_rate": 2.0208907587407516e-06, + "loss": 0.0423, + "step": 24886 + }, + { + "epoch": 2.6980702515177795, + "grad_norm": 4.693857192993164, + "learning_rate": 2.020165385173364e-06, + "loss": 0.008, + "step": 24887 + }, + { + "epoch": 2.6981786643538594, + "grad_norm": 11.20413589477539, + "learning_rate": 2.019440011605977e-06, + "loss": 0.0105, + "step": 24888 + }, + { + "epoch": 2.6982870771899394, + "grad_norm": 0.13551123440265656, + "learning_rate": 2.01871463803859e-06, + "loss": 0.0005, + "step": 24889 + }, + { + "epoch": 2.698395490026019, + "grad_norm": 1.0707696676254272, + "learning_rate": 2.0179892644712027e-06, + "loss": 0.0012, + "step": 24890 + }, + { + "epoch": 2.698503902862099, + "grad_norm": 0.6623835563659668, + "learning_rate": 2.0172638909038157e-06, + "loss": 0.0008, + "step": 24891 + }, + { + "epoch": 2.6986123156981785, + "grad_norm": 4.590192794799805, + "learning_rate": 2.0165385173364287e-06, + "loss": 0.0119, + "step": 24892 + }, + { + "epoch": 2.6987207285342585, + "grad_norm": 3.898834466934204, + "learning_rate": 2.0158131437690413e-06, + "loss": 0.0076, + "step": 24893 + }, + { + "epoch": 2.6988291413703385, + "grad_norm": 2.393981456756592, + "learning_rate": 2.0150877702016543e-06, + "loss": 0.0035, + "step": 24894 + }, + { + "epoch": 2.698937554206418, + "grad_norm": 0.13843251764774323, + "learning_rate": 2.014362396634267e-06, + "loss": 0.0006, + "step": 24895 + }, + { + "epoch": 2.699045967042498, + "grad_norm": 0.3039526641368866, + "learning_rate": 2.0136370230668794e-06, + "loss": 0.0019, + "step": 24896 + }, + { + "epoch": 2.6991543798785775, + "grad_norm": 0.32423344254493713, + "learning_rate": 2.0129116494994924e-06, + "loss": 0.0006, + "step": 24897 + }, + { + "epoch": 2.6992627927146575, + "grad_norm": 0.10749619454145432, + "learning_rate": 2.012186275932105e-06, + "loss": 0.0006, + "step": 24898 + }, + { + "epoch": 2.699371205550737, + "grad_norm": 0.5262140035629272, + "learning_rate": 2.011460902364718e-06, + "loss": 0.0017, + "step": 24899 + }, + { + "epoch": 2.699479618386817, + "grad_norm": 0.030045565217733383, + "learning_rate": 2.0107355287973305e-06, + "loss": 0.0003, + "step": 24900 + }, + { + "epoch": 2.6995880312228966, + "grad_norm": 0.03809398040175438, + "learning_rate": 2.0100101552299435e-06, + "loss": 0.0004, + "step": 24901 + }, + { + "epoch": 2.6996964440589766, + "grad_norm": 0.16850732266902924, + "learning_rate": 2.0092847816625565e-06, + "loss": 0.0005, + "step": 24902 + }, + { + "epoch": 2.6998048568950566, + "grad_norm": 0.6116867065429688, + "learning_rate": 2.008559408095169e-06, + "loss": 0.0021, + "step": 24903 + }, + { + "epoch": 2.699913269731136, + "grad_norm": 1.1323920488357544, + "learning_rate": 2.007834034527782e-06, + "loss": 0.0032, + "step": 24904 + }, + { + "epoch": 2.700021682567216, + "grad_norm": 1.3385885953903198, + "learning_rate": 2.007108660960395e-06, + "loss": 0.0053, + "step": 24905 + }, + { + "epoch": 2.7001300954032956, + "grad_norm": 0.1451994925737381, + "learning_rate": 2.0063832873930076e-06, + "loss": 0.0007, + "step": 24906 + }, + { + "epoch": 2.7002385082393756, + "grad_norm": 0.04220118746161461, + "learning_rate": 2.0056579138256206e-06, + "loss": 0.0005, + "step": 24907 + }, + { + "epoch": 2.7003469210754556, + "grad_norm": 0.36042463779449463, + "learning_rate": 2.004932540258233e-06, + "loss": 0.0012, + "step": 24908 + }, + { + "epoch": 2.700455333911535, + "grad_norm": 2.383510112762451, + "learning_rate": 2.004207166690846e-06, + "loss": 0.0025, + "step": 24909 + }, + { + "epoch": 2.7005637467476147, + "grad_norm": 1.0794260501861572, + "learning_rate": 2.0034817931234587e-06, + "loss": 0.0008, + "step": 24910 + }, + { + "epoch": 2.7006721595836947, + "grad_norm": 1.254827857017517, + "learning_rate": 2.0027564195560713e-06, + "loss": 0.0031, + "step": 24911 + }, + { + "epoch": 2.7007805724197746, + "grad_norm": 0.12432502955198288, + "learning_rate": 2.0020310459886843e-06, + "loss": 0.0002, + "step": 24912 + }, + { + "epoch": 2.700888985255854, + "grad_norm": 4.026916027069092, + "learning_rate": 2.001305672421297e-06, + "loss": 0.0619, + "step": 24913 + }, + { + "epoch": 2.700997398091934, + "grad_norm": 2.835146188735962, + "learning_rate": 2.00058029885391e-06, + "loss": 0.007, + "step": 24914 + }, + { + "epoch": 2.7011058109280137, + "grad_norm": 0.09857823699712753, + "learning_rate": 1.999854925286523e-06, + "loss": 0.0003, + "step": 24915 + }, + { + "epoch": 2.7012142237640937, + "grad_norm": 3.4661142826080322, + "learning_rate": 1.9991295517191354e-06, + "loss": 0.0226, + "step": 24916 + }, + { + "epoch": 2.7013226366001737, + "grad_norm": 1.1487009525299072, + "learning_rate": 1.9984041781517484e-06, + "loss": 0.0016, + "step": 24917 + }, + { + "epoch": 2.7014310494362532, + "grad_norm": 0.08071038126945496, + "learning_rate": 1.9976788045843614e-06, + "loss": 0.0005, + "step": 24918 + }, + { + "epoch": 2.7015394622723328, + "grad_norm": 0.040964070707559586, + "learning_rate": 1.996953431016974e-06, + "loss": 0.0003, + "step": 24919 + }, + { + "epoch": 2.7016478751084128, + "grad_norm": 0.01993127539753914, + "learning_rate": 1.996228057449587e-06, + "loss": 0.0001, + "step": 24920 + }, + { + "epoch": 2.7017562879444927, + "grad_norm": 3.0385055541992188, + "learning_rate": 1.9955026838821995e-06, + "loss": 0.0113, + "step": 24921 + }, + { + "epoch": 2.7018647007805723, + "grad_norm": 0.14826391637325287, + "learning_rate": 1.9947773103148125e-06, + "loss": 0.0008, + "step": 24922 + }, + { + "epoch": 2.7019731136166523, + "grad_norm": 1.4404278993606567, + "learning_rate": 1.994051936747425e-06, + "loss": 0.0505, + "step": 24923 + }, + { + "epoch": 2.702081526452732, + "grad_norm": 3.186241388320923, + "learning_rate": 1.9933265631800376e-06, + "loss": 0.0075, + "step": 24924 + }, + { + "epoch": 2.702189939288812, + "grad_norm": 0.016735175624489784, + "learning_rate": 1.9926011896126506e-06, + "loss": 0.0002, + "step": 24925 + }, + { + "epoch": 2.702298352124892, + "grad_norm": 4.091002941131592, + "learning_rate": 1.991875816045263e-06, + "loss": 0.0034, + "step": 24926 + }, + { + "epoch": 2.7024067649609713, + "grad_norm": 2.0033581256866455, + "learning_rate": 1.991150442477876e-06, + "loss": 0.0403, + "step": 24927 + }, + { + "epoch": 2.7025151777970513, + "grad_norm": 2.4555625915527344, + "learning_rate": 1.990425068910489e-06, + "loss": 0.0089, + "step": 24928 + }, + { + "epoch": 2.702623590633131, + "grad_norm": 0.3742988407611847, + "learning_rate": 1.9896996953431017e-06, + "loss": 0.0005, + "step": 24929 + }, + { + "epoch": 2.702732003469211, + "grad_norm": 0.06495766341686249, + "learning_rate": 1.9889743217757147e-06, + "loss": 0.0003, + "step": 24930 + }, + { + "epoch": 2.702840416305291, + "grad_norm": 3.7519824504852295, + "learning_rate": 1.9882489482083277e-06, + "loss": 0.0547, + "step": 24931 + }, + { + "epoch": 2.7029488291413704, + "grad_norm": 2.306691884994507, + "learning_rate": 1.9875235746409403e-06, + "loss": 0.0058, + "step": 24932 + }, + { + "epoch": 2.70305724197745, + "grad_norm": 1.9141658544540405, + "learning_rate": 1.9867982010735533e-06, + "loss": 0.0074, + "step": 24933 + }, + { + "epoch": 2.70316565481353, + "grad_norm": 2.341460943222046, + "learning_rate": 1.986072827506166e-06, + "loss": 0.005, + "step": 24934 + }, + { + "epoch": 2.70327406764961, + "grad_norm": 1.040223479270935, + "learning_rate": 1.985347453938779e-06, + "loss": 0.0034, + "step": 24935 + }, + { + "epoch": 2.7033824804856894, + "grad_norm": 0.17464040219783783, + "learning_rate": 1.9846220803713914e-06, + "loss": 0.0007, + "step": 24936 + }, + { + "epoch": 2.7034908933217694, + "grad_norm": 4.116101264953613, + "learning_rate": 1.983896706804004e-06, + "loss": 0.036, + "step": 24937 + }, + { + "epoch": 2.703599306157849, + "grad_norm": 2.2242257595062256, + "learning_rate": 1.983171333236617e-06, + "loss": 0.0381, + "step": 24938 + }, + { + "epoch": 2.703707718993929, + "grad_norm": 0.021856021136045456, + "learning_rate": 1.9824459596692295e-06, + "loss": 0.0002, + "step": 24939 + }, + { + "epoch": 2.703816131830009, + "grad_norm": 0.8938284516334534, + "learning_rate": 1.9817205861018425e-06, + "loss": 0.0047, + "step": 24940 + }, + { + "epoch": 2.7039245446660884, + "grad_norm": 1.9955008029937744, + "learning_rate": 1.9809952125344555e-06, + "loss": 0.0198, + "step": 24941 + }, + { + "epoch": 2.704032957502168, + "grad_norm": 5.612492561340332, + "learning_rate": 1.980269838967068e-06, + "loss": 0.0578, + "step": 24942 + }, + { + "epoch": 2.704141370338248, + "grad_norm": 0.11025036126375198, + "learning_rate": 1.979544465399681e-06, + "loss": 0.0009, + "step": 24943 + }, + { + "epoch": 2.704249783174328, + "grad_norm": 0.033137399703264236, + "learning_rate": 1.978819091832294e-06, + "loss": 0.0002, + "step": 24944 + }, + { + "epoch": 2.7043581960104075, + "grad_norm": 0.5605559349060059, + "learning_rate": 1.9780937182649066e-06, + "loss": 0.0022, + "step": 24945 + }, + { + "epoch": 2.7044666088464875, + "grad_norm": 0.0074172778986394405, + "learning_rate": 1.9773683446975196e-06, + "loss": 0.0001, + "step": 24946 + }, + { + "epoch": 2.704575021682567, + "grad_norm": 5.653763771057129, + "learning_rate": 1.976642971130132e-06, + "loss": 0.0441, + "step": 24947 + }, + { + "epoch": 2.704683434518647, + "grad_norm": 0.37022703886032104, + "learning_rate": 1.975917597562745e-06, + "loss": 0.0009, + "step": 24948 + }, + { + "epoch": 2.704791847354727, + "grad_norm": 1.8587205410003662, + "learning_rate": 1.9751922239953577e-06, + "loss": 0.0483, + "step": 24949 + }, + { + "epoch": 2.7049002601908065, + "grad_norm": 0.24669067561626434, + "learning_rate": 1.9744668504279707e-06, + "loss": 0.0013, + "step": 24950 + }, + { + "epoch": 2.7050086730268865, + "grad_norm": 0.007108059246093035, + "learning_rate": 1.9737414768605833e-06, + "loss": 0.0001, + "step": 24951 + }, + { + "epoch": 2.705117085862966, + "grad_norm": 1.805344581604004, + "learning_rate": 1.973016103293196e-06, + "loss": 0.0252, + "step": 24952 + }, + { + "epoch": 2.705225498699046, + "grad_norm": 0.6358561515808105, + "learning_rate": 1.972290729725809e-06, + "loss": 0.0026, + "step": 24953 + }, + { + "epoch": 2.705333911535126, + "grad_norm": 0.13526713848114014, + "learning_rate": 1.971565356158422e-06, + "loss": 0.0005, + "step": 24954 + }, + { + "epoch": 2.7054423243712056, + "grad_norm": 1.8736259937286377, + "learning_rate": 1.9708399825910344e-06, + "loss": 0.0119, + "step": 24955 + }, + { + "epoch": 2.705550737207285, + "grad_norm": 0.032282836735248566, + "learning_rate": 1.9701146090236474e-06, + "loss": 0.0002, + "step": 24956 + }, + { + "epoch": 2.705659150043365, + "grad_norm": 0.038855697959661484, + "learning_rate": 1.9693892354562604e-06, + "loss": 0.0003, + "step": 24957 + }, + { + "epoch": 2.705767562879445, + "grad_norm": 3.298774242401123, + "learning_rate": 1.968663861888873e-06, + "loss": 0.0096, + "step": 24958 + }, + { + "epoch": 2.7058759757155246, + "grad_norm": 0.03665568679571152, + "learning_rate": 1.967938488321486e-06, + "loss": 0.0002, + "step": 24959 + }, + { + "epoch": 2.7059843885516046, + "grad_norm": 0.0185418538749218, + "learning_rate": 1.9672131147540985e-06, + "loss": 0.0002, + "step": 24960 + }, + { + "epoch": 2.706092801387684, + "grad_norm": 1.6492342948913574, + "learning_rate": 1.9664877411867115e-06, + "loss": 0.0393, + "step": 24961 + }, + { + "epoch": 2.706201214223764, + "grad_norm": 0.02498948574066162, + "learning_rate": 1.965762367619324e-06, + "loss": 0.0002, + "step": 24962 + }, + { + "epoch": 2.706309627059844, + "grad_norm": 0.038406964391469955, + "learning_rate": 1.965036994051937e-06, + "loss": 0.0004, + "step": 24963 + }, + { + "epoch": 2.7064180398959237, + "grad_norm": 1.4411132335662842, + "learning_rate": 1.9643116204845496e-06, + "loss": 0.0038, + "step": 24964 + }, + { + "epoch": 2.706526452732003, + "grad_norm": 0.18070702254772186, + "learning_rate": 1.963586246917162e-06, + "loss": 0.0012, + "step": 24965 + }, + { + "epoch": 2.706634865568083, + "grad_norm": 0.11509387940168381, + "learning_rate": 1.962860873349775e-06, + "loss": 0.0008, + "step": 24966 + }, + { + "epoch": 2.706743278404163, + "grad_norm": 0.12010890990495682, + "learning_rate": 1.962135499782388e-06, + "loss": 0.0012, + "step": 24967 + }, + { + "epoch": 2.7068516912402427, + "grad_norm": 0.24340026080608368, + "learning_rate": 1.9614101262150008e-06, + "loss": 0.0003, + "step": 24968 + }, + { + "epoch": 2.7069601040763227, + "grad_norm": 0.15734094381332397, + "learning_rate": 1.9606847526476137e-06, + "loss": 0.0003, + "step": 24969 + }, + { + "epoch": 2.7070685169124022, + "grad_norm": 1.3526332378387451, + "learning_rate": 1.9599593790802267e-06, + "loss": 0.0033, + "step": 24970 + }, + { + "epoch": 2.7071769297484822, + "grad_norm": 7.5469841957092285, + "learning_rate": 1.9592340055128393e-06, + "loss": 0.0224, + "step": 24971 + }, + { + "epoch": 2.707285342584562, + "grad_norm": 1.2510720491409302, + "learning_rate": 1.9585086319454523e-06, + "loss": 0.0024, + "step": 24972 + }, + { + "epoch": 2.7073937554206418, + "grad_norm": 0.3326786160469055, + "learning_rate": 1.957783258378065e-06, + "loss": 0.0012, + "step": 24973 + }, + { + "epoch": 2.7075021682567217, + "grad_norm": 3.530795097351074, + "learning_rate": 1.957057884810678e-06, + "loss": 0.0173, + "step": 24974 + }, + { + "epoch": 2.7076105810928013, + "grad_norm": 0.057083651423454285, + "learning_rate": 1.9563325112432904e-06, + "loss": 0.0005, + "step": 24975 + }, + { + "epoch": 2.7077189939288813, + "grad_norm": 0.0165041983127594, + "learning_rate": 1.9556071376759034e-06, + "loss": 0.0002, + "step": 24976 + }, + { + "epoch": 2.707827406764961, + "grad_norm": 0.3422052562236786, + "learning_rate": 1.954881764108516e-06, + "loss": 0.0014, + "step": 24977 + }, + { + "epoch": 2.707935819601041, + "grad_norm": 0.848265528678894, + "learning_rate": 1.9541563905411285e-06, + "loss": 0.0038, + "step": 24978 + }, + { + "epoch": 2.7080442324371203, + "grad_norm": 0.05814285948872566, + "learning_rate": 1.9534310169737415e-06, + "loss": 0.0006, + "step": 24979 + }, + { + "epoch": 2.7081526452732003, + "grad_norm": 2.7018492221832275, + "learning_rate": 1.9527056434063545e-06, + "loss": 0.0269, + "step": 24980 + }, + { + "epoch": 2.7082610581092803, + "grad_norm": 14.555512428283691, + "learning_rate": 1.951980269838967e-06, + "loss": 0.0446, + "step": 24981 + }, + { + "epoch": 2.70836947094536, + "grad_norm": 0.019551675766706467, + "learning_rate": 1.95125489627158e-06, + "loss": 0.0001, + "step": 24982 + }, + { + "epoch": 2.70847788378144, + "grad_norm": 0.029088657349348068, + "learning_rate": 1.950529522704193e-06, + "loss": 0.0003, + "step": 24983 + }, + { + "epoch": 2.7085862966175194, + "grad_norm": 0.2202809900045395, + "learning_rate": 1.9498041491368056e-06, + "loss": 0.001, + "step": 24984 + }, + { + "epoch": 2.7086947094535994, + "grad_norm": 3.712090253829956, + "learning_rate": 1.9490787755694186e-06, + "loss": 0.0095, + "step": 24985 + }, + { + "epoch": 2.7088031222896793, + "grad_norm": 0.5225943326950073, + "learning_rate": 1.948353402002031e-06, + "loss": 0.0015, + "step": 24986 + }, + { + "epoch": 2.708911535125759, + "grad_norm": 0.15819236636161804, + "learning_rate": 1.947628028434644e-06, + "loss": 0.0005, + "step": 24987 + }, + { + "epoch": 2.7090199479618384, + "grad_norm": 0.0953870490193367, + "learning_rate": 1.9469026548672567e-06, + "loss": 0.0005, + "step": 24988 + }, + { + "epoch": 2.7091283607979184, + "grad_norm": 3.3667140007019043, + "learning_rate": 1.9461772812998697e-06, + "loss": 0.0048, + "step": 24989 + }, + { + "epoch": 2.7092367736339984, + "grad_norm": 2.7315492630004883, + "learning_rate": 1.9454519077324823e-06, + "loss": 0.0145, + "step": 24990 + }, + { + "epoch": 2.709345186470078, + "grad_norm": 0.033626724034547806, + "learning_rate": 1.944726534165095e-06, + "loss": 0.0002, + "step": 24991 + }, + { + "epoch": 2.709453599306158, + "grad_norm": 0.11647552996873856, + "learning_rate": 1.944001160597708e-06, + "loss": 0.0006, + "step": 24992 + }, + { + "epoch": 2.7095620121422375, + "grad_norm": 12.506294250488281, + "learning_rate": 1.943275787030321e-06, + "loss": 0.0252, + "step": 24993 + }, + { + "epoch": 2.7096704249783174, + "grad_norm": 0.07694334536790848, + "learning_rate": 1.9425504134629334e-06, + "loss": 0.0004, + "step": 24994 + }, + { + "epoch": 2.7097788378143974, + "grad_norm": 0.2874806523323059, + "learning_rate": 1.9418250398955464e-06, + "loss": 0.0019, + "step": 24995 + }, + { + "epoch": 2.709887250650477, + "grad_norm": 2.065749406814575, + "learning_rate": 1.9410996663281594e-06, + "loss": 0.0506, + "step": 24996 + }, + { + "epoch": 2.709995663486557, + "grad_norm": 1.0220673084259033, + "learning_rate": 1.940374292760772e-06, + "loss": 0.0047, + "step": 24997 + }, + { + "epoch": 2.7101040763226365, + "grad_norm": 0.02574232593178749, + "learning_rate": 1.939648919193385e-06, + "loss": 0.0002, + "step": 24998 + }, + { + "epoch": 2.7102124891587165, + "grad_norm": 0.0753568634390831, + "learning_rate": 1.9389235456259975e-06, + "loss": 0.0005, + "step": 24999 + }, + { + "epoch": 2.710320901994796, + "grad_norm": 9.491629600524902, + "learning_rate": 1.9381981720586105e-06, + "loss": 0.0105, + "step": 25000 + }, + { + "epoch": 2.710429314830876, + "grad_norm": 0.29576221108436584, + "learning_rate": 1.937472798491223e-06, + "loss": 0.0013, + "step": 25001 + }, + { + "epoch": 2.7105377276669556, + "grad_norm": 0.005893752444535494, + "learning_rate": 1.936747424923836e-06, + "loss": 0.0001, + "step": 25002 + }, + { + "epoch": 2.7106461405030355, + "grad_norm": 9.626896858215332, + "learning_rate": 1.9360220513564486e-06, + "loss": 0.0747, + "step": 25003 + }, + { + "epoch": 2.7107545533391155, + "grad_norm": 0.17998969554901123, + "learning_rate": 1.9352966777890616e-06, + "loss": 0.0008, + "step": 25004 + }, + { + "epoch": 2.710862966175195, + "grad_norm": 0.028231266885995865, + "learning_rate": 1.934571304221674e-06, + "loss": 0.0003, + "step": 25005 + }, + { + "epoch": 2.710971379011275, + "grad_norm": 0.012158910743892193, + "learning_rate": 1.933845930654287e-06, + "loss": 0.0001, + "step": 25006 + }, + { + "epoch": 2.7110797918473546, + "grad_norm": 0.17366176843643188, + "learning_rate": 1.9331205570868998e-06, + "loss": 0.0008, + "step": 25007 + }, + { + "epoch": 2.7111882046834346, + "grad_norm": 0.3796105980873108, + "learning_rate": 1.9323951835195127e-06, + "loss": 0.0015, + "step": 25008 + }, + { + "epoch": 2.7112966175195146, + "grad_norm": 0.7078077793121338, + "learning_rate": 1.9316698099521257e-06, + "loss": 0.0008, + "step": 25009 + }, + { + "epoch": 2.711405030355594, + "grad_norm": 4.4724884033203125, + "learning_rate": 1.9309444363847383e-06, + "loss": 0.0262, + "step": 25010 + }, + { + "epoch": 2.7115134431916736, + "grad_norm": 0.061602115631103516, + "learning_rate": 1.9302190628173513e-06, + "loss": 0.0005, + "step": 25011 + }, + { + "epoch": 2.7116218560277536, + "grad_norm": 0.007356954738497734, + "learning_rate": 1.929493689249964e-06, + "loss": 0.0001, + "step": 25012 + }, + { + "epoch": 2.7117302688638336, + "grad_norm": 7.993110656738281, + "learning_rate": 1.928768315682577e-06, + "loss": 0.0386, + "step": 25013 + }, + { + "epoch": 2.711838681699913, + "grad_norm": 0.2556190490722656, + "learning_rate": 1.9280429421151894e-06, + "loss": 0.0015, + "step": 25014 + }, + { + "epoch": 2.711947094535993, + "grad_norm": 0.8945001363754272, + "learning_rate": 1.9273175685478024e-06, + "loss": 0.0022, + "step": 25015 + }, + { + "epoch": 2.7120555073720727, + "grad_norm": 0.373258501291275, + "learning_rate": 1.926592194980415e-06, + "loss": 0.0015, + "step": 25016 + }, + { + "epoch": 2.7121639202081527, + "grad_norm": 1.3425229787826538, + "learning_rate": 1.925866821413028e-06, + "loss": 0.0034, + "step": 25017 + }, + { + "epoch": 2.7122723330442327, + "grad_norm": 3.0403287410736084, + "learning_rate": 1.9251414478456405e-06, + "loss": 0.0016, + "step": 25018 + }, + { + "epoch": 2.712380745880312, + "grad_norm": 3.7528347969055176, + "learning_rate": 1.9244160742782535e-06, + "loss": 0.0119, + "step": 25019 + }, + { + "epoch": 2.712489158716392, + "grad_norm": 3.427281141281128, + "learning_rate": 1.923690700710866e-06, + "loss": 0.0181, + "step": 25020 + }, + { + "epoch": 2.7125975715524717, + "grad_norm": 2.821622371673584, + "learning_rate": 1.922965327143479e-06, + "loss": 0.0354, + "step": 25021 + }, + { + "epoch": 2.7127059843885517, + "grad_norm": 0.2462134063243866, + "learning_rate": 1.922239953576092e-06, + "loss": 0.0016, + "step": 25022 + }, + { + "epoch": 2.7128143972246312, + "grad_norm": 0.0479305274784565, + "learning_rate": 1.9215145800087046e-06, + "loss": 0.0003, + "step": 25023 + }, + { + "epoch": 2.7129228100607112, + "grad_norm": 0.03350391611456871, + "learning_rate": 1.9207892064413176e-06, + "loss": 0.0002, + "step": 25024 + }, + { + "epoch": 2.7130312228967908, + "grad_norm": 0.2824784815311432, + "learning_rate": 1.92006383287393e-06, + "loss": 0.0016, + "step": 25025 + }, + { + "epoch": 2.7131396357328708, + "grad_norm": 5.213517189025879, + "learning_rate": 1.919338459306543e-06, + "loss": 0.0047, + "step": 25026 + }, + { + "epoch": 2.7132480485689507, + "grad_norm": 8.325776100158691, + "learning_rate": 1.9186130857391558e-06, + "loss": 0.0043, + "step": 25027 + }, + { + "epoch": 2.7133564614050303, + "grad_norm": 0.03717264160513878, + "learning_rate": 1.9178877121717687e-06, + "loss": 0.0003, + "step": 25028 + }, + { + "epoch": 2.7134648742411103, + "grad_norm": 0.02416323870420456, + "learning_rate": 1.9171623386043813e-06, + "loss": 0.0001, + "step": 25029 + }, + { + "epoch": 2.71357328707719, + "grad_norm": 0.3770367503166199, + "learning_rate": 1.9164369650369943e-06, + "loss": 0.0014, + "step": 25030 + }, + { + "epoch": 2.71368169991327, + "grad_norm": 8.700261116027832, + "learning_rate": 1.915711591469607e-06, + "loss": 0.0049, + "step": 25031 + }, + { + "epoch": 2.71379011274935, + "grad_norm": 0.06895903497934341, + "learning_rate": 1.91498621790222e-06, + "loss": 0.0005, + "step": 25032 + }, + { + "epoch": 2.7138985255854293, + "grad_norm": 0.21914610266685486, + "learning_rate": 1.9142608443348324e-06, + "loss": 0.0011, + "step": 25033 + }, + { + "epoch": 2.714006938421509, + "grad_norm": 2.4297215938568115, + "learning_rate": 1.9135354707674454e-06, + "loss": 0.0234, + "step": 25034 + }, + { + "epoch": 2.714115351257589, + "grad_norm": 0.5666629076004028, + "learning_rate": 1.9128100972000584e-06, + "loss": 0.001, + "step": 25035 + }, + { + "epoch": 2.714223764093669, + "grad_norm": 0.9394813179969788, + "learning_rate": 1.912084723632671e-06, + "loss": 0.0032, + "step": 25036 + }, + { + "epoch": 2.7143321769297484, + "grad_norm": 0.8514184355735779, + "learning_rate": 1.911359350065284e-06, + "loss": 0.0026, + "step": 25037 + }, + { + "epoch": 2.7144405897658284, + "grad_norm": 0.024716323241591454, + "learning_rate": 1.9106339764978965e-06, + "loss": 0.0001, + "step": 25038 + }, + { + "epoch": 2.714549002601908, + "grad_norm": 0.0687485784292221, + "learning_rate": 1.9099086029305095e-06, + "loss": 0.0006, + "step": 25039 + }, + { + "epoch": 2.714657415437988, + "grad_norm": 8.59521484375, + "learning_rate": 1.909183229363122e-06, + "loss": 0.0393, + "step": 25040 + }, + { + "epoch": 2.714765828274068, + "grad_norm": 0.03127097338438034, + "learning_rate": 1.908457855795735e-06, + "loss": 0.0003, + "step": 25041 + }, + { + "epoch": 2.7148742411101474, + "grad_norm": 3.6975960731506348, + "learning_rate": 1.9077324822283476e-06, + "loss": 0.0147, + "step": 25042 + }, + { + "epoch": 2.7149826539462274, + "grad_norm": 1.1711946725845337, + "learning_rate": 1.9070071086609606e-06, + "loss": 0.0367, + "step": 25043 + }, + { + "epoch": 2.715091066782307, + "grad_norm": 0.05924195051193237, + "learning_rate": 1.9062817350935734e-06, + "loss": 0.0006, + "step": 25044 + }, + { + "epoch": 2.715199479618387, + "grad_norm": 0.026948735117912292, + "learning_rate": 1.905556361526186e-06, + "loss": 0.0003, + "step": 25045 + }, + { + "epoch": 2.7153078924544665, + "grad_norm": 0.11452794820070267, + "learning_rate": 1.904830987958799e-06, + "loss": 0.0005, + "step": 25046 + }, + { + "epoch": 2.7154163052905465, + "grad_norm": 0.028846658766269684, + "learning_rate": 1.9041056143914115e-06, + "loss": 0.0002, + "step": 25047 + }, + { + "epoch": 2.715524718126626, + "grad_norm": 0.09248897433280945, + "learning_rate": 1.9033802408240245e-06, + "loss": 0.0005, + "step": 25048 + }, + { + "epoch": 2.715633130962706, + "grad_norm": 1.2977585792541504, + "learning_rate": 1.9026548672566373e-06, + "loss": 0.0032, + "step": 25049 + }, + { + "epoch": 2.715741543798786, + "grad_norm": 1.0103750228881836, + "learning_rate": 1.90192949368925e-06, + "loss": 0.0022, + "step": 25050 + }, + { + "epoch": 2.7158499566348655, + "grad_norm": 3.075021743774414, + "learning_rate": 1.9012041201218629e-06, + "loss": 0.0273, + "step": 25051 + }, + { + "epoch": 2.7159583694709455, + "grad_norm": 0.05513829365372658, + "learning_rate": 1.9004787465544759e-06, + "loss": 0.0002, + "step": 25052 + }, + { + "epoch": 2.716066782307025, + "grad_norm": 0.09180736541748047, + "learning_rate": 1.8997533729870884e-06, + "loss": 0.0005, + "step": 25053 + }, + { + "epoch": 2.716175195143105, + "grad_norm": 3.4388439655303955, + "learning_rate": 1.8990279994197014e-06, + "loss": 0.0027, + "step": 25054 + }, + { + "epoch": 2.716283607979185, + "grad_norm": 4.4014668464660645, + "learning_rate": 1.898302625852314e-06, + "loss": 0.0257, + "step": 25055 + }, + { + "epoch": 2.7163920208152645, + "grad_norm": 0.01360412035137415, + "learning_rate": 1.897577252284927e-06, + "loss": 0.0001, + "step": 25056 + }, + { + "epoch": 2.716500433651344, + "grad_norm": 0.7877592444419861, + "learning_rate": 1.8968518787175397e-06, + "loss": 0.0016, + "step": 25057 + }, + { + "epoch": 2.716608846487424, + "grad_norm": 6.028838157653809, + "learning_rate": 1.8961265051501525e-06, + "loss": 0.0326, + "step": 25058 + }, + { + "epoch": 2.716717259323504, + "grad_norm": 6.030372142791748, + "learning_rate": 1.8954011315827653e-06, + "loss": 0.0124, + "step": 25059 + }, + { + "epoch": 2.7168256721595836, + "grad_norm": 3.950737476348877, + "learning_rate": 1.8946757580153779e-06, + "loss": 0.0047, + "step": 25060 + }, + { + "epoch": 2.7169340849956636, + "grad_norm": 0.20195722579956055, + "learning_rate": 1.8939503844479909e-06, + "loss": 0.0008, + "step": 25061 + }, + { + "epoch": 2.717042497831743, + "grad_norm": 8.11735725402832, + "learning_rate": 1.8932250108806036e-06, + "loss": 0.018, + "step": 25062 + }, + { + "epoch": 2.717150910667823, + "grad_norm": 2.9509952068328857, + "learning_rate": 1.8924996373132164e-06, + "loss": 0.0021, + "step": 25063 + }, + { + "epoch": 2.717259323503903, + "grad_norm": 0.05847335606813431, + "learning_rate": 1.8917742637458292e-06, + "loss": 0.0007, + "step": 25064 + }, + { + "epoch": 2.7173677363399826, + "grad_norm": 0.12186657637357712, + "learning_rate": 1.8910488901784422e-06, + "loss": 0.0007, + "step": 25065 + }, + { + "epoch": 2.717476149176062, + "grad_norm": 0.1761654168367386, + "learning_rate": 1.8903235166110548e-06, + "loss": 0.0004, + "step": 25066 + }, + { + "epoch": 2.717584562012142, + "grad_norm": 12.659040451049805, + "learning_rate": 1.8895981430436677e-06, + "loss": 0.0152, + "step": 25067 + }, + { + "epoch": 2.717692974848222, + "grad_norm": 0.04230492562055588, + "learning_rate": 1.8888727694762803e-06, + "loss": 0.0004, + "step": 25068 + }, + { + "epoch": 2.7178013876843017, + "grad_norm": 3.733593225479126, + "learning_rate": 1.8881473959088933e-06, + "loss": 0.0133, + "step": 25069 + }, + { + "epoch": 2.7179098005203817, + "grad_norm": 0.055695272982120514, + "learning_rate": 1.887422022341506e-06, + "loss": 0.0004, + "step": 25070 + }, + { + "epoch": 2.718018213356461, + "grad_norm": 0.017882904037833214, + "learning_rate": 1.8866966487741189e-06, + "loss": 0.0002, + "step": 25071 + }, + { + "epoch": 2.718126626192541, + "grad_norm": 2.845463275909424, + "learning_rate": 1.8859712752067316e-06, + "loss": 0.0344, + "step": 25072 + }, + { + "epoch": 2.718235039028621, + "grad_norm": 3.68357515335083, + "learning_rate": 1.8852459016393442e-06, + "loss": 0.0482, + "step": 25073 + }, + { + "epoch": 2.7183434518647007, + "grad_norm": 0.33750903606414795, + "learning_rate": 1.8845205280719572e-06, + "loss": 0.0011, + "step": 25074 + }, + { + "epoch": 2.7184518647007807, + "grad_norm": 0.2513013482093811, + "learning_rate": 1.88379515450457e-06, + "loss": 0.0007, + "step": 25075 + }, + { + "epoch": 2.7185602775368602, + "grad_norm": 10.513773918151855, + "learning_rate": 1.8830697809371828e-06, + "loss": 0.0223, + "step": 25076 + }, + { + "epoch": 2.7186686903729402, + "grad_norm": 1.1199672222137451, + "learning_rate": 1.8823444073697955e-06, + "loss": 0.0023, + "step": 25077 + }, + { + "epoch": 2.71877710320902, + "grad_norm": 5.241086959838867, + "learning_rate": 1.8816190338024085e-06, + "loss": 0.0152, + "step": 25078 + }, + { + "epoch": 2.7188855160450998, + "grad_norm": 0.9237756133079529, + "learning_rate": 1.880893660235021e-06, + "loss": 0.002, + "step": 25079 + }, + { + "epoch": 2.7189939288811793, + "grad_norm": 0.05532742664217949, + "learning_rate": 1.880168286667634e-06, + "loss": 0.0003, + "step": 25080 + }, + { + "epoch": 2.7191023417172593, + "grad_norm": 2.0316483974456787, + "learning_rate": 1.8794429131002466e-06, + "loss": 0.0127, + "step": 25081 + }, + { + "epoch": 2.7192107545533393, + "grad_norm": 0.5286709666252136, + "learning_rate": 1.8787175395328596e-06, + "loss": 0.0012, + "step": 25082 + }, + { + "epoch": 2.719319167389419, + "grad_norm": 0.06848252564668655, + "learning_rate": 1.8779921659654724e-06, + "loss": 0.0002, + "step": 25083 + }, + { + "epoch": 2.719427580225499, + "grad_norm": 0.7425723075866699, + "learning_rate": 1.8772667923980852e-06, + "loss": 0.0024, + "step": 25084 + }, + { + "epoch": 2.7195359930615783, + "grad_norm": 0.1928863525390625, + "learning_rate": 1.876541418830698e-06, + "loss": 0.0009, + "step": 25085 + }, + { + "epoch": 2.7196444058976583, + "grad_norm": 0.14497867226600647, + "learning_rate": 1.8758160452633105e-06, + "loss": 0.0005, + "step": 25086 + }, + { + "epoch": 2.7197528187337383, + "grad_norm": 0.7959333062171936, + "learning_rate": 1.8750906716959235e-06, + "loss": 0.0111, + "step": 25087 + }, + { + "epoch": 2.719861231569818, + "grad_norm": 0.18232671916484833, + "learning_rate": 1.8743652981285363e-06, + "loss": 0.0012, + "step": 25088 + }, + { + "epoch": 2.7199696444058974, + "grad_norm": 3.016474962234497, + "learning_rate": 1.873639924561149e-06, + "loss": 0.0103, + "step": 25089 + }, + { + "epoch": 2.7200780572419774, + "grad_norm": 0.029239796102046967, + "learning_rate": 1.8729145509937619e-06, + "loss": 0.0003, + "step": 25090 + }, + { + "epoch": 2.7201864700780574, + "grad_norm": 0.09720788151025772, + "learning_rate": 1.8721891774263749e-06, + "loss": 0.001, + "step": 25091 + }, + { + "epoch": 2.720294882914137, + "grad_norm": 0.0648091733455658, + "learning_rate": 1.8714638038589874e-06, + "loss": 0.0003, + "step": 25092 + }, + { + "epoch": 2.720403295750217, + "grad_norm": 8.158123970031738, + "learning_rate": 1.8707384302916004e-06, + "loss": 0.0078, + "step": 25093 + }, + { + "epoch": 2.7205117085862964, + "grad_norm": 1.3548504114151, + "learning_rate": 1.870013056724213e-06, + "loss": 0.0044, + "step": 25094 + }, + { + "epoch": 2.7206201214223764, + "grad_norm": 0.6764635443687439, + "learning_rate": 1.869287683156826e-06, + "loss": 0.0012, + "step": 25095 + }, + { + "epoch": 2.7207285342584564, + "grad_norm": 0.06500931829214096, + "learning_rate": 1.8685623095894387e-06, + "loss": 0.0005, + "step": 25096 + }, + { + "epoch": 2.720836947094536, + "grad_norm": 9.619181632995605, + "learning_rate": 1.8678369360220515e-06, + "loss": 0.015, + "step": 25097 + }, + { + "epoch": 2.720945359930616, + "grad_norm": 0.06095263734459877, + "learning_rate": 1.8671115624546643e-06, + "loss": 0.0005, + "step": 25098 + }, + { + "epoch": 2.7210537727666955, + "grad_norm": 0.21056383848190308, + "learning_rate": 1.8663861888872769e-06, + "loss": 0.0005, + "step": 25099 + }, + { + "epoch": 2.7211621856027755, + "grad_norm": 0.04566028341650963, + "learning_rate": 1.8656608153198899e-06, + "loss": 0.0004, + "step": 25100 + }, + { + "epoch": 2.7212705984388554, + "grad_norm": 0.017814844846725464, + "learning_rate": 1.8649354417525026e-06, + "loss": 0.0002, + "step": 25101 + }, + { + "epoch": 2.721379011274935, + "grad_norm": 0.648160457611084, + "learning_rate": 1.8642100681851154e-06, + "loss": 0.0026, + "step": 25102 + }, + { + "epoch": 2.7214874241110145, + "grad_norm": 3.1635029315948486, + "learning_rate": 1.8634846946177282e-06, + "loss": 0.0174, + "step": 25103 + }, + { + "epoch": 2.7215958369470945, + "grad_norm": 0.08687081933021545, + "learning_rate": 1.8627593210503412e-06, + "loss": 0.0005, + "step": 25104 + }, + { + "epoch": 2.7217042497831745, + "grad_norm": 0.06448229402303696, + "learning_rate": 1.8620339474829538e-06, + "loss": 0.0003, + "step": 25105 + }, + { + "epoch": 2.721812662619254, + "grad_norm": 0.6143262982368469, + "learning_rate": 1.8613085739155667e-06, + "loss": 0.0016, + "step": 25106 + }, + { + "epoch": 2.721921075455334, + "grad_norm": 0.11541827768087387, + "learning_rate": 1.8605832003481793e-06, + "loss": 0.0004, + "step": 25107 + }, + { + "epoch": 2.7220294882914136, + "grad_norm": 0.03461156785488129, + "learning_rate": 1.8598578267807923e-06, + "loss": 0.0001, + "step": 25108 + }, + { + "epoch": 2.7221379011274935, + "grad_norm": 0.15505775809288025, + "learning_rate": 1.859132453213405e-06, + "loss": 0.0009, + "step": 25109 + }, + { + "epoch": 2.7222463139635735, + "grad_norm": 1.7439249753952026, + "learning_rate": 1.8584070796460179e-06, + "loss": 0.007, + "step": 25110 + }, + { + "epoch": 2.722354726799653, + "grad_norm": 0.6432111859321594, + "learning_rate": 1.8576817060786306e-06, + "loss": 0.0039, + "step": 25111 + }, + { + "epoch": 2.7224631396357326, + "grad_norm": 2.2895843982696533, + "learning_rate": 1.8569563325112436e-06, + "loss": 0.0496, + "step": 25112 + }, + { + "epoch": 2.7225715524718126, + "grad_norm": 5.461183547973633, + "learning_rate": 1.8562309589438562e-06, + "loss": 0.0472, + "step": 25113 + }, + { + "epoch": 2.7226799653078926, + "grad_norm": 1.2283397912979126, + "learning_rate": 1.855505585376469e-06, + "loss": 0.0025, + "step": 25114 + }, + { + "epoch": 2.722788378143972, + "grad_norm": 1.7389700412750244, + "learning_rate": 1.8547802118090818e-06, + "loss": 0.0015, + "step": 25115 + }, + { + "epoch": 2.722896790980052, + "grad_norm": 5.7815752029418945, + "learning_rate": 1.8540548382416945e-06, + "loss": 0.0354, + "step": 25116 + }, + { + "epoch": 2.7230052038161316, + "grad_norm": 0.04188861325383186, + "learning_rate": 1.8533294646743075e-06, + "loss": 0.0002, + "step": 25117 + }, + { + "epoch": 2.7231136166522116, + "grad_norm": 0.13364483416080475, + "learning_rate": 1.85260409110692e-06, + "loss": 0.0008, + "step": 25118 + }, + { + "epoch": 2.7232220294882916, + "grad_norm": 3.013617515563965, + "learning_rate": 1.851878717539533e-06, + "loss": 0.0036, + "step": 25119 + }, + { + "epoch": 2.723330442324371, + "grad_norm": 0.6186890006065369, + "learning_rate": 1.8511533439721456e-06, + "loss": 0.0031, + "step": 25120 + }, + { + "epoch": 2.723438855160451, + "grad_norm": 3.6604056358337402, + "learning_rate": 1.8504279704047586e-06, + "loss": 0.0097, + "step": 25121 + }, + { + "epoch": 2.7235472679965307, + "grad_norm": 0.3544164299964905, + "learning_rate": 1.8497025968373714e-06, + "loss": 0.0008, + "step": 25122 + }, + { + "epoch": 2.7236556808326107, + "grad_norm": 0.44197022914886475, + "learning_rate": 1.8489772232699842e-06, + "loss": 0.0009, + "step": 25123 + }, + { + "epoch": 2.72376409366869, + "grad_norm": 1.6152015924453735, + "learning_rate": 1.848251849702597e-06, + "loss": 0.0143, + "step": 25124 + }, + { + "epoch": 2.72387250650477, + "grad_norm": 0.06569994240999222, + "learning_rate": 1.84752647613521e-06, + "loss": 0.0004, + "step": 25125 + }, + { + "epoch": 2.7239809193408497, + "grad_norm": 0.04722592979669571, + "learning_rate": 1.8468011025678225e-06, + "loss": 0.0004, + "step": 25126 + }, + { + "epoch": 2.7240893321769297, + "grad_norm": 2.164290189743042, + "learning_rate": 1.8460757290004353e-06, + "loss": 0.0151, + "step": 25127 + }, + { + "epoch": 2.7241977450130097, + "grad_norm": 0.8539131283760071, + "learning_rate": 1.845350355433048e-06, + "loss": 0.0016, + "step": 25128 + }, + { + "epoch": 2.7243061578490892, + "grad_norm": 0.17449067533016205, + "learning_rate": 1.8446249818656609e-06, + "loss": 0.0007, + "step": 25129 + }, + { + "epoch": 2.7244145706851692, + "grad_norm": 0.004255042877048254, + "learning_rate": 1.8438996082982739e-06, + "loss": 0.0, + "step": 25130 + }, + { + "epoch": 2.7245229835212488, + "grad_norm": 2.8283417224884033, + "learning_rate": 1.8431742347308864e-06, + "loss": 0.011, + "step": 25131 + }, + { + "epoch": 2.7246313963573288, + "grad_norm": 2.1230602264404297, + "learning_rate": 1.8424488611634994e-06, + "loss": 0.0396, + "step": 25132 + }, + { + "epoch": 2.7247398091934087, + "grad_norm": 0.061354827135801315, + "learning_rate": 1.841723487596112e-06, + "loss": 0.0003, + "step": 25133 + }, + { + "epoch": 2.7248482220294883, + "grad_norm": 0.1793445199728012, + "learning_rate": 1.840998114028725e-06, + "loss": 0.001, + "step": 25134 + }, + { + "epoch": 2.724956634865568, + "grad_norm": 2.460669755935669, + "learning_rate": 1.8402727404613378e-06, + "loss": 0.0189, + "step": 25135 + }, + { + "epoch": 2.725065047701648, + "grad_norm": 0.08626207709312439, + "learning_rate": 1.8395473668939505e-06, + "loss": 0.0006, + "step": 25136 + }, + { + "epoch": 2.725173460537728, + "grad_norm": 2.1749532222747803, + "learning_rate": 1.8388219933265633e-06, + "loss": 0.0035, + "step": 25137 + }, + { + "epoch": 2.7252818733738073, + "grad_norm": 0.02918139658868313, + "learning_rate": 1.8380966197591763e-06, + "loss": 0.0004, + "step": 25138 + }, + { + "epoch": 2.7253902862098873, + "grad_norm": 0.1545974165201187, + "learning_rate": 1.8373712461917889e-06, + "loss": 0.001, + "step": 25139 + }, + { + "epoch": 2.725498699045967, + "grad_norm": 0.7191433906555176, + "learning_rate": 1.8366458726244016e-06, + "loss": 0.0011, + "step": 25140 + }, + { + "epoch": 2.725607111882047, + "grad_norm": 0.026021629571914673, + "learning_rate": 1.8359204990570144e-06, + "loss": 0.0003, + "step": 25141 + }, + { + "epoch": 2.725715524718127, + "grad_norm": 10.705796241760254, + "learning_rate": 1.8351951254896272e-06, + "loss": 0.02, + "step": 25142 + }, + { + "epoch": 2.7258239375542064, + "grad_norm": 2.9233791828155518, + "learning_rate": 1.8344697519222402e-06, + "loss": 0.0241, + "step": 25143 + }, + { + "epoch": 2.7259323503902864, + "grad_norm": 0.007355198729783297, + "learning_rate": 1.8337443783548528e-06, + "loss": 0.0001, + "step": 25144 + }, + { + "epoch": 2.726040763226366, + "grad_norm": 0.13529092073440552, + "learning_rate": 1.8330190047874658e-06, + "loss": 0.0004, + "step": 25145 + }, + { + "epoch": 2.726149176062446, + "grad_norm": 0.5163116455078125, + "learning_rate": 1.8322936312200783e-06, + "loss": 0.0009, + "step": 25146 + }, + { + "epoch": 2.7262575888985254, + "grad_norm": 0.9543068408966064, + "learning_rate": 1.8315682576526913e-06, + "loss": 0.0028, + "step": 25147 + }, + { + "epoch": 2.7263660017346054, + "grad_norm": 0.10622506588697433, + "learning_rate": 1.830842884085304e-06, + "loss": 0.0004, + "step": 25148 + }, + { + "epoch": 2.726474414570685, + "grad_norm": 0.04220174252986908, + "learning_rate": 1.830117510517917e-06, + "loss": 0.0004, + "step": 25149 + }, + { + "epoch": 2.726582827406765, + "grad_norm": 0.16409306228160858, + "learning_rate": 1.8293921369505296e-06, + "loss": 0.0004, + "step": 25150 + }, + { + "epoch": 2.726691240242845, + "grad_norm": 3.277048110961914, + "learning_rate": 1.8286667633831426e-06, + "loss": 0.022, + "step": 25151 + }, + { + "epoch": 2.7267996530789245, + "grad_norm": 0.3905029594898224, + "learning_rate": 1.8279413898157552e-06, + "loss": 0.0012, + "step": 25152 + }, + { + "epoch": 2.7269080659150045, + "grad_norm": 0.07239624112844467, + "learning_rate": 1.8272160162483682e-06, + "loss": 0.0003, + "step": 25153 + }, + { + "epoch": 2.727016478751084, + "grad_norm": 0.030895939096808434, + "learning_rate": 1.8264906426809808e-06, + "loss": 0.0002, + "step": 25154 + }, + { + "epoch": 2.727124891587164, + "grad_norm": 0.5920107364654541, + "learning_rate": 1.8257652691135935e-06, + "loss": 0.0033, + "step": 25155 + }, + { + "epoch": 2.727233304423244, + "grad_norm": 0.33390024304389954, + "learning_rate": 1.8250398955462065e-06, + "loss": 0.0017, + "step": 25156 + }, + { + "epoch": 2.7273417172593235, + "grad_norm": 2.075761556625366, + "learning_rate": 1.824314521978819e-06, + "loss": 0.0053, + "step": 25157 + }, + { + "epoch": 2.727450130095403, + "grad_norm": 4.6824445724487305, + "learning_rate": 1.823589148411432e-06, + "loss": 0.014, + "step": 25158 + }, + { + "epoch": 2.727558542931483, + "grad_norm": 0.7870026230812073, + "learning_rate": 1.8228637748440447e-06, + "loss": 0.0017, + "step": 25159 + }, + { + "epoch": 2.727666955767563, + "grad_norm": 1.8298674821853638, + "learning_rate": 1.8221384012766576e-06, + "loss": 0.0263, + "step": 25160 + }, + { + "epoch": 2.7277753686036426, + "grad_norm": 0.4898289144039154, + "learning_rate": 1.8214130277092704e-06, + "loss": 0.0009, + "step": 25161 + }, + { + "epoch": 2.7278837814397225, + "grad_norm": 2.7374789714813232, + "learning_rate": 1.8206876541418834e-06, + "loss": 0.0081, + "step": 25162 + }, + { + "epoch": 2.727992194275802, + "grad_norm": 3.3680903911590576, + "learning_rate": 1.819962280574496e-06, + "loss": 0.0077, + "step": 25163 + }, + { + "epoch": 2.728100607111882, + "grad_norm": 3.588611602783203, + "learning_rate": 1.819236907007109e-06, + "loss": 0.0525, + "step": 25164 + }, + { + "epoch": 2.728209019947962, + "grad_norm": 0.39088380336761475, + "learning_rate": 1.8185115334397215e-06, + "loss": 0.0009, + "step": 25165 + }, + { + "epoch": 2.7283174327840416, + "grad_norm": 0.14027927815914154, + "learning_rate": 1.8177861598723345e-06, + "loss": 0.001, + "step": 25166 + }, + { + "epoch": 2.7284258456201216, + "grad_norm": 0.09569951146841049, + "learning_rate": 1.8170607863049473e-06, + "loss": 0.0006, + "step": 25167 + }, + { + "epoch": 2.728534258456201, + "grad_norm": 1.7559394836425781, + "learning_rate": 1.8163354127375599e-06, + "loss": 0.0217, + "step": 25168 + }, + { + "epoch": 2.728642671292281, + "grad_norm": 0.5576099753379822, + "learning_rate": 1.8156100391701729e-06, + "loss": 0.0009, + "step": 25169 + }, + { + "epoch": 2.7287510841283606, + "grad_norm": 0.005144859664142132, + "learning_rate": 1.8148846656027854e-06, + "loss": 0.0001, + "step": 25170 + }, + { + "epoch": 2.7288594969644406, + "grad_norm": 6.03516149520874, + "learning_rate": 1.8141592920353984e-06, + "loss": 0.0415, + "step": 25171 + }, + { + "epoch": 2.72896790980052, + "grad_norm": 1.9225939512252808, + "learning_rate": 1.813433918468011e-06, + "loss": 0.0285, + "step": 25172 + }, + { + "epoch": 2.7290763226366, + "grad_norm": 2.6894569396972656, + "learning_rate": 1.812708544900624e-06, + "loss": 0.0359, + "step": 25173 + }, + { + "epoch": 2.72918473547268, + "grad_norm": 0.36233165860176086, + "learning_rate": 1.8119831713332368e-06, + "loss": 0.0011, + "step": 25174 + }, + { + "epoch": 2.7292931483087597, + "grad_norm": 1.529349684715271, + "learning_rate": 1.8112577977658497e-06, + "loss": 0.0056, + "step": 25175 + }, + { + "epoch": 2.7294015611448397, + "grad_norm": 0.06312384456396103, + "learning_rate": 1.8105324241984623e-06, + "loss": 0.0004, + "step": 25176 + }, + { + "epoch": 2.729509973980919, + "grad_norm": 0.6504772901535034, + "learning_rate": 1.8098070506310753e-06, + "loss": 0.0009, + "step": 25177 + }, + { + "epoch": 2.729618386816999, + "grad_norm": 0.4629475176334381, + "learning_rate": 1.8090816770636879e-06, + "loss": 0.0028, + "step": 25178 + }, + { + "epoch": 2.729726799653079, + "grad_norm": 1.7129664421081543, + "learning_rate": 1.8083563034963009e-06, + "loss": 0.0299, + "step": 25179 + }, + { + "epoch": 2.7298352124891587, + "grad_norm": 3.787034511566162, + "learning_rate": 1.8076309299289136e-06, + "loss": 0.0549, + "step": 25180 + }, + { + "epoch": 2.7299436253252383, + "grad_norm": 5.0271806716918945, + "learning_rate": 1.8069055563615262e-06, + "loss": 0.0463, + "step": 25181 + }, + { + "epoch": 2.7300520381613183, + "grad_norm": 9.630471229553223, + "learning_rate": 1.8061801827941392e-06, + "loss": 0.0334, + "step": 25182 + }, + { + "epoch": 2.7301604509973982, + "grad_norm": 1.5382401943206787, + "learning_rate": 1.8054548092267518e-06, + "loss": 0.0061, + "step": 25183 + }, + { + "epoch": 2.7302688638334778, + "grad_norm": 0.34222039580345154, + "learning_rate": 1.8047294356593648e-06, + "loss": 0.0021, + "step": 25184 + }, + { + "epoch": 2.7303772766695578, + "grad_norm": 0.155926913022995, + "learning_rate": 1.8040040620919773e-06, + "loss": 0.0005, + "step": 25185 + }, + { + "epoch": 2.7304856895056373, + "grad_norm": 2.0394628047943115, + "learning_rate": 1.8032786885245903e-06, + "loss": 0.0211, + "step": 25186 + }, + { + "epoch": 2.7305941023417173, + "grad_norm": 0.04702012613415718, + "learning_rate": 1.802553314957203e-06, + "loss": 0.0003, + "step": 25187 + }, + { + "epoch": 2.7307025151777973, + "grad_norm": 2.1232047080993652, + "learning_rate": 1.801827941389816e-06, + "loss": 0.0287, + "step": 25188 + }, + { + "epoch": 2.730810928013877, + "grad_norm": 3.233480930328369, + "learning_rate": 1.8011025678224286e-06, + "loss": 0.0178, + "step": 25189 + }, + { + "epoch": 2.730919340849957, + "grad_norm": 0.07505161315202713, + "learning_rate": 1.8003771942550416e-06, + "loss": 0.0005, + "step": 25190 + }, + { + "epoch": 2.7310277536860363, + "grad_norm": 2.761676073074341, + "learning_rate": 1.7996518206876542e-06, + "loss": 0.0387, + "step": 25191 + }, + { + "epoch": 2.7311361665221163, + "grad_norm": 7.826925277709961, + "learning_rate": 1.7989264471202672e-06, + "loss": 0.0423, + "step": 25192 + }, + { + "epoch": 2.731244579358196, + "grad_norm": 0.07908377051353455, + "learning_rate": 1.79820107355288e-06, + "loss": 0.0005, + "step": 25193 + }, + { + "epoch": 2.731352992194276, + "grad_norm": 0.13683021068572998, + "learning_rate": 1.7974756999854925e-06, + "loss": 0.001, + "step": 25194 + }, + { + "epoch": 2.7314614050303554, + "grad_norm": 5.148063659667969, + "learning_rate": 1.7967503264181055e-06, + "loss": 0.0367, + "step": 25195 + }, + { + "epoch": 2.7315698178664354, + "grad_norm": 0.21477720141410828, + "learning_rate": 1.796024952850718e-06, + "loss": 0.0007, + "step": 25196 + }, + { + "epoch": 2.7316782307025154, + "grad_norm": 0.6560062170028687, + "learning_rate": 1.795299579283331e-06, + "loss": 0.0035, + "step": 25197 + }, + { + "epoch": 2.731786643538595, + "grad_norm": 0.08023848384618759, + "learning_rate": 1.7945742057159439e-06, + "loss": 0.0002, + "step": 25198 + }, + { + "epoch": 2.731895056374675, + "grad_norm": 0.6769822239875793, + "learning_rate": 1.7938488321485566e-06, + "loss": 0.0034, + "step": 25199 + }, + { + "epoch": 2.7320034692107544, + "grad_norm": 2.354048490524292, + "learning_rate": 1.7931234585811694e-06, + "loss": 0.0042, + "step": 25200 + }, + { + "epoch": 2.7321118820468344, + "grad_norm": 4.352388858795166, + "learning_rate": 1.7923980850137824e-06, + "loss": 0.0036, + "step": 25201 + }, + { + "epoch": 2.7322202948829144, + "grad_norm": 0.06106116622686386, + "learning_rate": 1.791672711446395e-06, + "loss": 0.0002, + "step": 25202 + }, + { + "epoch": 2.732328707718994, + "grad_norm": 1.2652617692947388, + "learning_rate": 1.790947337879008e-06, + "loss": 0.031, + "step": 25203 + }, + { + "epoch": 2.7324371205550735, + "grad_norm": 1.3832366466522217, + "learning_rate": 1.7902219643116205e-06, + "loss": 0.0053, + "step": 25204 + }, + { + "epoch": 2.7325455333911535, + "grad_norm": 0.17719008028507233, + "learning_rate": 1.7894965907442335e-06, + "loss": 0.0013, + "step": 25205 + }, + { + "epoch": 2.7326539462272335, + "grad_norm": 1.5024620294570923, + "learning_rate": 1.7887712171768463e-06, + "loss": 0.0025, + "step": 25206 + }, + { + "epoch": 2.732762359063313, + "grad_norm": 3.021012544631958, + "learning_rate": 1.788045843609459e-06, + "loss": 0.0052, + "step": 25207 + }, + { + "epoch": 2.732870771899393, + "grad_norm": 0.09848565608263016, + "learning_rate": 1.7873204700420719e-06, + "loss": 0.0003, + "step": 25208 + }, + { + "epoch": 2.7329791847354725, + "grad_norm": 0.7117263078689575, + "learning_rate": 1.7865950964746844e-06, + "loss": 0.0015, + "step": 25209 + }, + { + "epoch": 2.7330875975715525, + "grad_norm": 7.618786811828613, + "learning_rate": 1.7858697229072974e-06, + "loss": 0.0171, + "step": 25210 + }, + { + "epoch": 2.7331960104076325, + "grad_norm": 0.6548209190368652, + "learning_rate": 1.7851443493399102e-06, + "loss": 0.0025, + "step": 25211 + }, + { + "epoch": 2.733304423243712, + "grad_norm": 5.96690034866333, + "learning_rate": 1.784418975772523e-06, + "loss": 0.0087, + "step": 25212 + }, + { + "epoch": 2.7334128360797916, + "grad_norm": 0.03607505187392235, + "learning_rate": 1.7836936022051358e-06, + "loss": 0.0003, + "step": 25213 + }, + { + "epoch": 2.7335212489158716, + "grad_norm": 5.504568099975586, + "learning_rate": 1.7829682286377487e-06, + "loss": 0.0079, + "step": 25214 + }, + { + "epoch": 2.7336296617519515, + "grad_norm": 12.590047836303711, + "learning_rate": 1.7822428550703613e-06, + "loss": 0.0181, + "step": 25215 + }, + { + "epoch": 2.733738074588031, + "grad_norm": 1.7461133003234863, + "learning_rate": 1.7815174815029743e-06, + "loss": 0.0088, + "step": 25216 + }, + { + "epoch": 2.733846487424111, + "grad_norm": 19.209375381469727, + "learning_rate": 1.7807921079355869e-06, + "loss": 0.0075, + "step": 25217 + }, + { + "epoch": 2.7339549002601906, + "grad_norm": 0.08719327300786972, + "learning_rate": 1.7800667343681999e-06, + "loss": 0.0007, + "step": 25218 + }, + { + "epoch": 2.7340633130962706, + "grad_norm": 0.04901750013232231, + "learning_rate": 1.7793413608008126e-06, + "loss": 0.0003, + "step": 25219 + }, + { + "epoch": 2.7341717259323506, + "grad_norm": 0.20804840326309204, + "learning_rate": 1.7786159872334254e-06, + "loss": 0.001, + "step": 25220 + }, + { + "epoch": 2.73428013876843, + "grad_norm": 0.18916183710098267, + "learning_rate": 1.7778906136660382e-06, + "loss": 0.0008, + "step": 25221 + }, + { + "epoch": 2.73438855160451, + "grad_norm": 0.12349212914705276, + "learning_rate": 1.7771652400986508e-06, + "loss": 0.0005, + "step": 25222 + }, + { + "epoch": 2.7344969644405897, + "grad_norm": 3.918959140777588, + "learning_rate": 1.7764398665312638e-06, + "loss": 0.0121, + "step": 25223 + }, + { + "epoch": 2.7346053772766696, + "grad_norm": 0.11407598853111267, + "learning_rate": 1.7757144929638765e-06, + "loss": 0.0004, + "step": 25224 + }, + { + "epoch": 2.7347137901127496, + "grad_norm": 0.0846547931432724, + "learning_rate": 1.7749891193964893e-06, + "loss": 0.0009, + "step": 25225 + }, + { + "epoch": 2.734822202948829, + "grad_norm": 1.5029858350753784, + "learning_rate": 1.774263745829102e-06, + "loss": 0.0061, + "step": 25226 + }, + { + "epoch": 2.7349306157849087, + "grad_norm": 11.236976623535156, + "learning_rate": 1.773538372261715e-06, + "loss": 0.0373, + "step": 25227 + }, + { + "epoch": 2.7350390286209887, + "grad_norm": 0.18120218813419342, + "learning_rate": 1.7728129986943277e-06, + "loss": 0.0009, + "step": 25228 + }, + { + "epoch": 2.7351474414570687, + "grad_norm": 0.06410646438598633, + "learning_rate": 1.7720876251269406e-06, + "loss": 0.0007, + "step": 25229 + }, + { + "epoch": 2.735255854293148, + "grad_norm": 0.036944929510354996, + "learning_rate": 1.7713622515595532e-06, + "loss": 0.0003, + "step": 25230 + }, + { + "epoch": 2.735364267129228, + "grad_norm": 2.866929531097412, + "learning_rate": 1.7706368779921662e-06, + "loss": 0.0226, + "step": 25231 + }, + { + "epoch": 2.7354726799653077, + "grad_norm": 1.7784630060195923, + "learning_rate": 1.769911504424779e-06, + "loss": 0.0038, + "step": 25232 + }, + { + "epoch": 2.7355810928013877, + "grad_norm": 4.30573034286499, + "learning_rate": 1.7691861308573918e-06, + "loss": 0.0098, + "step": 25233 + }, + { + "epoch": 2.7356895056374677, + "grad_norm": 0.04342320188879967, + "learning_rate": 1.7684607572900045e-06, + "loss": 0.0004, + "step": 25234 + }, + { + "epoch": 2.7357979184735473, + "grad_norm": 0.6353260278701782, + "learning_rate": 1.767735383722617e-06, + "loss": 0.0027, + "step": 25235 + }, + { + "epoch": 2.735906331309627, + "grad_norm": 0.09124000370502472, + "learning_rate": 1.76701001015523e-06, + "loss": 0.0006, + "step": 25236 + }, + { + "epoch": 2.736014744145707, + "grad_norm": 9.783740997314453, + "learning_rate": 1.7662846365878429e-06, + "loss": 0.0536, + "step": 25237 + }, + { + "epoch": 2.7361231569817868, + "grad_norm": 0.6031156182289124, + "learning_rate": 1.7655592630204556e-06, + "loss": 0.0021, + "step": 25238 + }, + { + "epoch": 2.7362315698178663, + "grad_norm": 2.2711586952209473, + "learning_rate": 1.7648338894530684e-06, + "loss": 0.0081, + "step": 25239 + }, + { + "epoch": 2.7363399826539463, + "grad_norm": 0.16250073909759521, + "learning_rate": 1.7641085158856814e-06, + "loss": 0.0007, + "step": 25240 + }, + { + "epoch": 2.736448395490026, + "grad_norm": 0.43725869059562683, + "learning_rate": 1.763383142318294e-06, + "loss": 0.0011, + "step": 25241 + }, + { + "epoch": 2.736556808326106, + "grad_norm": 0.10240843147039413, + "learning_rate": 1.762657768750907e-06, + "loss": 0.0009, + "step": 25242 + }, + { + "epoch": 2.736665221162186, + "grad_norm": 0.19361472129821777, + "learning_rate": 1.7619323951835195e-06, + "loss": 0.0005, + "step": 25243 + }, + { + "epoch": 2.7367736339982653, + "grad_norm": 0.0267446618527174, + "learning_rate": 1.7612070216161325e-06, + "loss": 0.0003, + "step": 25244 + }, + { + "epoch": 2.7368820468343453, + "grad_norm": 0.8012142181396484, + "learning_rate": 1.7604816480487453e-06, + "loss": 0.0036, + "step": 25245 + }, + { + "epoch": 2.736990459670425, + "grad_norm": 1.2535042762756348, + "learning_rate": 1.759756274481358e-06, + "loss": 0.0119, + "step": 25246 + }, + { + "epoch": 2.737098872506505, + "grad_norm": 0.06245221197605133, + "learning_rate": 1.7590309009139709e-06, + "loss": 0.0002, + "step": 25247 + }, + { + "epoch": 2.737207285342585, + "grad_norm": 1.3855724334716797, + "learning_rate": 1.7583055273465834e-06, + "loss": 0.0051, + "step": 25248 + }, + { + "epoch": 2.7373156981786644, + "grad_norm": 0.8626423478126526, + "learning_rate": 1.7575801537791964e-06, + "loss": 0.003, + "step": 25249 + }, + { + "epoch": 2.737424111014744, + "grad_norm": 0.09754469990730286, + "learning_rate": 1.7568547802118092e-06, + "loss": 0.001, + "step": 25250 + }, + { + "epoch": 2.737532523850824, + "grad_norm": 0.08672648668289185, + "learning_rate": 1.756129406644422e-06, + "loss": 0.0007, + "step": 25251 + }, + { + "epoch": 2.737640936686904, + "grad_norm": 0.08093667030334473, + "learning_rate": 1.7554040330770348e-06, + "loss": 0.0006, + "step": 25252 + }, + { + "epoch": 2.7377493495229834, + "grad_norm": 2.008042573928833, + "learning_rate": 1.7546786595096478e-06, + "loss": 0.0061, + "step": 25253 + }, + { + "epoch": 2.7378577623590634, + "grad_norm": 1.0392512083053589, + "learning_rate": 1.7539532859422603e-06, + "loss": 0.0049, + "step": 25254 + }, + { + "epoch": 2.737966175195143, + "grad_norm": 1.35308039188385, + "learning_rate": 1.7532279123748733e-06, + "loss": 0.0022, + "step": 25255 + }, + { + "epoch": 2.738074588031223, + "grad_norm": 4.899337291717529, + "learning_rate": 1.7525025388074859e-06, + "loss": 0.015, + "step": 25256 + }, + { + "epoch": 2.738183000867303, + "grad_norm": 0.1446191966533661, + "learning_rate": 1.7517771652400989e-06, + "loss": 0.0005, + "step": 25257 + }, + { + "epoch": 2.7382914137033825, + "grad_norm": 1.5031932592391968, + "learning_rate": 1.7510517916727116e-06, + "loss": 0.0062, + "step": 25258 + }, + { + "epoch": 2.738399826539462, + "grad_norm": 14.731014251708984, + "learning_rate": 1.7503264181053244e-06, + "loss": 0.0096, + "step": 25259 + }, + { + "epoch": 2.738508239375542, + "grad_norm": 7.294739723205566, + "learning_rate": 1.7496010445379372e-06, + "loss": 0.0525, + "step": 25260 + }, + { + "epoch": 2.738616652211622, + "grad_norm": 1.8864651918411255, + "learning_rate": 1.7488756709705502e-06, + "loss": 0.0336, + "step": 25261 + }, + { + "epoch": 2.7387250650477015, + "grad_norm": 4.690243244171143, + "learning_rate": 1.7481502974031628e-06, + "loss": 0.0163, + "step": 25262 + }, + { + "epoch": 2.7388334778837815, + "grad_norm": 4.2508931159973145, + "learning_rate": 1.7474249238357755e-06, + "loss": 0.0145, + "step": 25263 + }, + { + "epoch": 2.738941890719861, + "grad_norm": 0.12456002831459045, + "learning_rate": 1.7466995502683883e-06, + "loss": 0.0004, + "step": 25264 + }, + { + "epoch": 2.739050303555941, + "grad_norm": 0.06461792439222336, + "learning_rate": 1.745974176701001e-06, + "loss": 0.0005, + "step": 25265 + }, + { + "epoch": 2.739158716392021, + "grad_norm": 3.3056514263153076, + "learning_rate": 1.745248803133614e-06, + "loss": 0.0389, + "step": 25266 + }, + { + "epoch": 2.7392671292281006, + "grad_norm": 0.1278197169303894, + "learning_rate": 1.7445234295662267e-06, + "loss": 0.0005, + "step": 25267 + }, + { + "epoch": 2.7393755420641805, + "grad_norm": 0.03042527288198471, + "learning_rate": 1.7437980559988396e-06, + "loss": 0.0002, + "step": 25268 + }, + { + "epoch": 2.73948395490026, + "grad_norm": 1.7353678941726685, + "learning_rate": 1.7430726824314522e-06, + "loss": 0.0045, + "step": 25269 + }, + { + "epoch": 2.73959236773634, + "grad_norm": 0.10401829332113266, + "learning_rate": 1.7423473088640652e-06, + "loss": 0.0007, + "step": 25270 + }, + { + "epoch": 2.7397007805724196, + "grad_norm": 2.320589303970337, + "learning_rate": 1.741621935296678e-06, + "loss": 0.0424, + "step": 25271 + }, + { + "epoch": 2.7398091934084996, + "grad_norm": 0.05680947005748749, + "learning_rate": 1.7408965617292908e-06, + "loss": 0.0004, + "step": 25272 + }, + { + "epoch": 2.739917606244579, + "grad_norm": 3.0409228801727295, + "learning_rate": 1.7401711881619035e-06, + "loss": 0.0297, + "step": 25273 + }, + { + "epoch": 2.740026019080659, + "grad_norm": 0.7466889023780823, + "learning_rate": 1.7394458145945165e-06, + "loss": 0.0028, + "step": 25274 + }, + { + "epoch": 2.740134431916739, + "grad_norm": 5.087950706481934, + "learning_rate": 1.738720441027129e-06, + "loss": 0.0232, + "step": 25275 + }, + { + "epoch": 2.7402428447528187, + "grad_norm": 1.8988069295883179, + "learning_rate": 1.7379950674597419e-06, + "loss": 0.0265, + "step": 25276 + }, + { + "epoch": 2.7403512575888986, + "grad_norm": 4.166316509246826, + "learning_rate": 1.7372696938923547e-06, + "loss": 0.0152, + "step": 25277 + }, + { + "epoch": 2.740459670424978, + "grad_norm": 0.17088329792022705, + "learning_rate": 1.7365443203249674e-06, + "loss": 0.0007, + "step": 25278 + }, + { + "epoch": 2.740568083261058, + "grad_norm": 0.011762499809265137, + "learning_rate": 1.7358189467575804e-06, + "loss": 0.0001, + "step": 25279 + }, + { + "epoch": 2.740676496097138, + "grad_norm": 2.7828924655914307, + "learning_rate": 1.735093573190193e-06, + "loss": 0.0356, + "step": 25280 + }, + { + "epoch": 2.7407849089332177, + "grad_norm": 8.892274856567383, + "learning_rate": 1.734368199622806e-06, + "loss": 0.014, + "step": 25281 + }, + { + "epoch": 2.7408933217692972, + "grad_norm": 0.07270143926143646, + "learning_rate": 1.7336428260554185e-06, + "loss": 0.0002, + "step": 25282 + }, + { + "epoch": 2.741001734605377, + "grad_norm": 0.08673550188541412, + "learning_rate": 1.7329174524880315e-06, + "loss": 0.0004, + "step": 25283 + }, + { + "epoch": 2.741110147441457, + "grad_norm": 4.1806416511535645, + "learning_rate": 1.7321920789206443e-06, + "loss": 0.0365, + "step": 25284 + }, + { + "epoch": 2.7412185602775367, + "grad_norm": 1.5617623329162598, + "learning_rate": 1.731466705353257e-06, + "loss": 0.002, + "step": 25285 + }, + { + "epoch": 2.7413269731136167, + "grad_norm": 1.5191411972045898, + "learning_rate": 1.7307413317858699e-06, + "loss": 0.0055, + "step": 25286 + }, + { + "epoch": 2.7414353859496963, + "grad_norm": 0.3229946196079254, + "learning_rate": 1.7300159582184829e-06, + "loss": 0.0007, + "step": 25287 + }, + { + "epoch": 2.7415437987857763, + "grad_norm": 0.10329650342464447, + "learning_rate": 1.7292905846510954e-06, + "loss": 0.0007, + "step": 25288 + }, + { + "epoch": 2.7416522116218562, + "grad_norm": 5.514350414276123, + "learning_rate": 1.7285652110837082e-06, + "loss": 0.0666, + "step": 25289 + }, + { + "epoch": 2.741760624457936, + "grad_norm": 0.6227088570594788, + "learning_rate": 1.727839837516321e-06, + "loss": 0.0062, + "step": 25290 + }, + { + "epoch": 2.7418690372940158, + "grad_norm": 0.411499559879303, + "learning_rate": 1.7271144639489338e-06, + "loss": 0.0014, + "step": 25291 + }, + { + "epoch": 2.7419774501300953, + "grad_norm": 3.4554548263549805, + "learning_rate": 1.7263890903815468e-06, + "loss": 0.0171, + "step": 25292 + }, + { + "epoch": 2.7420858629661753, + "grad_norm": 0.18772897124290466, + "learning_rate": 1.7256637168141593e-06, + "loss": 0.0009, + "step": 25293 + }, + { + "epoch": 2.742194275802255, + "grad_norm": 0.8072370290756226, + "learning_rate": 1.7249383432467723e-06, + "loss": 0.0026, + "step": 25294 + }, + { + "epoch": 2.742302688638335, + "grad_norm": 0.07460878044366837, + "learning_rate": 1.7242129696793849e-06, + "loss": 0.0003, + "step": 25295 + }, + { + "epoch": 2.7424111014744144, + "grad_norm": 0.04663817957043648, + "learning_rate": 1.7234875961119979e-06, + "loss": 0.0004, + "step": 25296 + }, + { + "epoch": 2.7425195143104943, + "grad_norm": 0.020590748637914658, + "learning_rate": 1.7227622225446106e-06, + "loss": 0.0002, + "step": 25297 + }, + { + "epoch": 2.7426279271465743, + "grad_norm": 0.14794211089611053, + "learning_rate": 1.7220368489772234e-06, + "loss": 0.0011, + "step": 25298 + }, + { + "epoch": 2.742736339982654, + "grad_norm": 0.29507461190223694, + "learning_rate": 1.7213114754098362e-06, + "loss": 0.0007, + "step": 25299 + }, + { + "epoch": 2.742844752818734, + "grad_norm": 0.21298043429851532, + "learning_rate": 1.7205861018424492e-06, + "loss": 0.0012, + "step": 25300 + }, + { + "epoch": 2.7429531656548134, + "grad_norm": 0.555971086025238, + "learning_rate": 1.7198607282750618e-06, + "loss": 0.0013, + "step": 25301 + }, + { + "epoch": 2.7430615784908934, + "grad_norm": 8.368040084838867, + "learning_rate": 1.7191353547076748e-06, + "loss": 0.0325, + "step": 25302 + }, + { + "epoch": 2.7431699913269734, + "grad_norm": 2.326268434524536, + "learning_rate": 1.7184099811402873e-06, + "loss": 0.0136, + "step": 25303 + }, + { + "epoch": 2.743278404163053, + "grad_norm": 0.04485670104622841, + "learning_rate": 1.7176846075729e-06, + "loss": 0.0005, + "step": 25304 + }, + { + "epoch": 2.7433868169991324, + "grad_norm": 3.7730391025543213, + "learning_rate": 1.716959234005513e-06, + "loss": 0.0533, + "step": 25305 + }, + { + "epoch": 2.7434952298352124, + "grad_norm": 0.2601701617240906, + "learning_rate": 1.7162338604381257e-06, + "loss": 0.0014, + "step": 25306 + }, + { + "epoch": 2.7436036426712924, + "grad_norm": 6.130241870880127, + "learning_rate": 1.7155084868707386e-06, + "loss": 0.0769, + "step": 25307 + }, + { + "epoch": 2.743712055507372, + "grad_norm": 0.9432198405265808, + "learning_rate": 1.7147831133033512e-06, + "loss": 0.0044, + "step": 25308 + }, + { + "epoch": 2.743820468343452, + "grad_norm": 0.6070910096168518, + "learning_rate": 1.7140577397359642e-06, + "loss": 0.001, + "step": 25309 + }, + { + "epoch": 2.7439288811795315, + "grad_norm": 3.4043076038360596, + "learning_rate": 1.713332366168577e-06, + "loss": 0.0574, + "step": 25310 + }, + { + "epoch": 2.7440372940156115, + "grad_norm": 2.8150012493133545, + "learning_rate": 1.7126069926011898e-06, + "loss": 0.0076, + "step": 25311 + }, + { + "epoch": 2.7441457068516915, + "grad_norm": 0.22300194203853607, + "learning_rate": 1.7118816190338025e-06, + "loss": 0.0008, + "step": 25312 + }, + { + "epoch": 2.744254119687771, + "grad_norm": 3.3417298793792725, + "learning_rate": 1.7111562454664155e-06, + "loss": 0.0405, + "step": 25313 + }, + { + "epoch": 2.744362532523851, + "grad_norm": 0.07790175080299377, + "learning_rate": 1.710430871899028e-06, + "loss": 0.0003, + "step": 25314 + }, + { + "epoch": 2.7444709453599305, + "grad_norm": 0.20725195109844208, + "learning_rate": 1.709705498331641e-06, + "loss": 0.0012, + "step": 25315 + }, + { + "epoch": 2.7445793581960105, + "grad_norm": 0.8398565053939819, + "learning_rate": 1.7089801247642537e-06, + "loss": 0.0445, + "step": 25316 + }, + { + "epoch": 2.74468777103209, + "grad_norm": 8.141420364379883, + "learning_rate": 1.7082547511968664e-06, + "loss": 0.0145, + "step": 25317 + }, + { + "epoch": 2.74479618386817, + "grad_norm": 0.9439080357551575, + "learning_rate": 1.7075293776294794e-06, + "loss": 0.0023, + "step": 25318 + }, + { + "epoch": 2.7449045967042496, + "grad_norm": 0.006572357844561338, + "learning_rate": 1.706804004062092e-06, + "loss": 0.0001, + "step": 25319 + }, + { + "epoch": 2.7450130095403296, + "grad_norm": 0.2423451691865921, + "learning_rate": 1.706078630494705e-06, + "loss": 0.001, + "step": 25320 + }, + { + "epoch": 2.7451214223764095, + "grad_norm": 4.141980171203613, + "learning_rate": 1.7053532569273175e-06, + "loss": 0.0367, + "step": 25321 + }, + { + "epoch": 2.745229835212489, + "grad_norm": 0.31469982862472534, + "learning_rate": 1.7046278833599305e-06, + "loss": 0.0004, + "step": 25322 + }, + { + "epoch": 2.745338248048569, + "grad_norm": 1.2029343843460083, + "learning_rate": 1.7039025097925433e-06, + "loss": 0.0032, + "step": 25323 + }, + { + "epoch": 2.7454466608846486, + "grad_norm": 3.147352695465088, + "learning_rate": 1.703177136225156e-06, + "loss": 0.0193, + "step": 25324 + }, + { + "epoch": 2.7455550737207286, + "grad_norm": 1.7953683137893677, + "learning_rate": 1.7024517626577689e-06, + "loss": 0.0037, + "step": 25325 + }, + { + "epoch": 2.7456634865568086, + "grad_norm": 0.3979185223579407, + "learning_rate": 1.7017263890903819e-06, + "loss": 0.001, + "step": 25326 + }, + { + "epoch": 2.745771899392888, + "grad_norm": 0.07791487872600555, + "learning_rate": 1.7010010155229944e-06, + "loss": 0.0005, + "step": 25327 + }, + { + "epoch": 2.7458803122289677, + "grad_norm": 1.9521763324737549, + "learning_rate": 1.7002756419556074e-06, + "loss": 0.0097, + "step": 25328 + }, + { + "epoch": 2.7459887250650477, + "grad_norm": 0.049905624240636826, + "learning_rate": 1.69955026838822e-06, + "loss": 0.0004, + "step": 25329 + }, + { + "epoch": 2.7460971379011276, + "grad_norm": 0.44575339555740356, + "learning_rate": 1.6988248948208328e-06, + "loss": 0.0031, + "step": 25330 + }, + { + "epoch": 2.746205550737207, + "grad_norm": 0.07043907791376114, + "learning_rate": 1.6980995212534458e-06, + "loss": 0.0005, + "step": 25331 + }, + { + "epoch": 2.746313963573287, + "grad_norm": 4.244042873382568, + "learning_rate": 1.6973741476860583e-06, + "loss": 0.026, + "step": 25332 + }, + { + "epoch": 2.7464223764093667, + "grad_norm": 0.027150088921189308, + "learning_rate": 1.6966487741186713e-06, + "loss": 0.0002, + "step": 25333 + }, + { + "epoch": 2.7465307892454467, + "grad_norm": 2.0989863872528076, + "learning_rate": 1.6959234005512839e-06, + "loss": 0.0045, + "step": 25334 + }, + { + "epoch": 2.7466392020815267, + "grad_norm": 2.0649869441986084, + "learning_rate": 1.6951980269838969e-06, + "loss": 0.0267, + "step": 25335 + }, + { + "epoch": 2.746747614917606, + "grad_norm": 2.3220579624176025, + "learning_rate": 1.6944726534165097e-06, + "loss": 0.0167, + "step": 25336 + }, + { + "epoch": 2.746856027753686, + "grad_norm": 5.579751968383789, + "learning_rate": 1.6937472798491224e-06, + "loss": 0.0255, + "step": 25337 + }, + { + "epoch": 2.7469644405897657, + "grad_norm": 0.3368285298347473, + "learning_rate": 1.6930219062817352e-06, + "loss": 0.0011, + "step": 25338 + }, + { + "epoch": 2.7470728534258457, + "grad_norm": 2.988190174102783, + "learning_rate": 1.6922965327143482e-06, + "loss": 0.0044, + "step": 25339 + }, + { + "epoch": 2.7471812662619253, + "grad_norm": 0.04096880555152893, + "learning_rate": 1.6915711591469608e-06, + "loss": 0.0003, + "step": 25340 + }, + { + "epoch": 2.7472896790980053, + "grad_norm": 0.05973992496728897, + "learning_rate": 1.6908457855795738e-06, + "loss": 0.0005, + "step": 25341 + }, + { + "epoch": 2.747398091934085, + "grad_norm": 0.62014240026474, + "learning_rate": 1.6901204120121863e-06, + "loss": 0.0014, + "step": 25342 + }, + { + "epoch": 2.747506504770165, + "grad_norm": 1.7078531980514526, + "learning_rate": 1.689395038444799e-06, + "loss": 0.0053, + "step": 25343 + }, + { + "epoch": 2.7476149176062448, + "grad_norm": 0.01935761421918869, + "learning_rate": 1.688669664877412e-06, + "loss": 0.0002, + "step": 25344 + }, + { + "epoch": 2.7477233304423243, + "grad_norm": 2.7449798583984375, + "learning_rate": 1.6879442913100247e-06, + "loss": 0.0464, + "step": 25345 + }, + { + "epoch": 2.7478317432784043, + "grad_norm": 0.05840207636356354, + "learning_rate": 1.6872189177426377e-06, + "loss": 0.0003, + "step": 25346 + }, + { + "epoch": 2.747940156114484, + "grad_norm": 8.356955528259277, + "learning_rate": 1.6864935441752502e-06, + "loss": 0.0418, + "step": 25347 + }, + { + "epoch": 2.748048568950564, + "grad_norm": 0.9419794082641602, + "learning_rate": 1.6857681706078632e-06, + "loss": 0.002, + "step": 25348 + }, + { + "epoch": 2.748156981786644, + "grad_norm": 3.1172618865966797, + "learning_rate": 1.685042797040476e-06, + "loss": 0.0109, + "step": 25349 + }, + { + "epoch": 2.7482653946227233, + "grad_norm": 0.0684276893734932, + "learning_rate": 1.6843174234730888e-06, + "loss": 0.0003, + "step": 25350 + }, + { + "epoch": 2.748373807458803, + "grad_norm": 4.245641231536865, + "learning_rate": 1.6835920499057015e-06, + "loss": 0.0242, + "step": 25351 + }, + { + "epoch": 2.748482220294883, + "grad_norm": 6.75007963180542, + "learning_rate": 1.6828666763383145e-06, + "loss": 0.018, + "step": 25352 + }, + { + "epoch": 2.748590633130963, + "grad_norm": 8.40573501586914, + "learning_rate": 1.682141302770927e-06, + "loss": 0.0357, + "step": 25353 + }, + { + "epoch": 2.7486990459670424, + "grad_norm": 0.05890091881155968, + "learning_rate": 1.68141592920354e-06, + "loss": 0.0007, + "step": 25354 + }, + { + "epoch": 2.7488074588031224, + "grad_norm": 2.1792919635772705, + "learning_rate": 1.6806905556361527e-06, + "loss": 0.0042, + "step": 25355 + }, + { + "epoch": 2.748915871639202, + "grad_norm": 23.605514526367188, + "learning_rate": 1.6799651820687656e-06, + "loss": 0.0478, + "step": 25356 + }, + { + "epoch": 2.749024284475282, + "grad_norm": 0.497852087020874, + "learning_rate": 1.6792398085013784e-06, + "loss": 0.002, + "step": 25357 + }, + { + "epoch": 2.749132697311362, + "grad_norm": 0.03277408331632614, + "learning_rate": 1.678514434933991e-06, + "loss": 0.0003, + "step": 25358 + }, + { + "epoch": 2.7492411101474414, + "grad_norm": 23.38892936706543, + "learning_rate": 1.677789061366604e-06, + "loss": 0.0202, + "step": 25359 + }, + { + "epoch": 2.749349522983521, + "grad_norm": 0.03611040860414505, + "learning_rate": 1.6770636877992166e-06, + "loss": 0.0003, + "step": 25360 + }, + { + "epoch": 2.749457935819601, + "grad_norm": 0.18998666107654572, + "learning_rate": 1.6763383142318295e-06, + "loss": 0.0009, + "step": 25361 + }, + { + "epoch": 2.749566348655681, + "grad_norm": 7.602951526641846, + "learning_rate": 1.6756129406644423e-06, + "loss": 0.0554, + "step": 25362 + }, + { + "epoch": 2.7496747614917605, + "grad_norm": 0.21783596277236938, + "learning_rate": 1.674887567097055e-06, + "loss": 0.0011, + "step": 25363 + }, + { + "epoch": 2.7497831743278405, + "grad_norm": 3.0044667720794678, + "learning_rate": 1.6741621935296679e-06, + "loss": 0.0047, + "step": 25364 + }, + { + "epoch": 2.74989158716392, + "grad_norm": 3.5358753204345703, + "learning_rate": 1.6734368199622809e-06, + "loss": 0.0024, + "step": 25365 + }, + { + "epoch": 2.75, + "grad_norm": 1.5416125059127808, + "learning_rate": 1.6727114463948934e-06, + "loss": 0.039, + "step": 25366 + }, + { + "epoch": 2.75010841283608, + "grad_norm": 4.168554306030273, + "learning_rate": 1.6719860728275064e-06, + "loss": 0.094, + "step": 25367 + }, + { + "epoch": 2.7502168256721595, + "grad_norm": 4.826910495758057, + "learning_rate": 1.671260699260119e-06, + "loss": 0.0249, + "step": 25368 + }, + { + "epoch": 2.7503252385082395, + "grad_norm": 1.5806219577789307, + "learning_rate": 1.670535325692732e-06, + "loss": 0.0012, + "step": 25369 + }, + { + "epoch": 2.750433651344319, + "grad_norm": 2.491309881210327, + "learning_rate": 1.6698099521253448e-06, + "loss": 0.0042, + "step": 25370 + }, + { + "epoch": 2.750542064180399, + "grad_norm": 11.90081787109375, + "learning_rate": 1.6690845785579573e-06, + "loss": 0.01, + "step": 25371 + }, + { + "epoch": 2.750650477016479, + "grad_norm": 0.24609051644802094, + "learning_rate": 1.6683592049905703e-06, + "loss": 0.0009, + "step": 25372 + }, + { + "epoch": 2.7507588898525586, + "grad_norm": 0.6640579700469971, + "learning_rate": 1.6676338314231829e-06, + "loss": 0.0017, + "step": 25373 + }, + { + "epoch": 2.750867302688638, + "grad_norm": 0.018482573330402374, + "learning_rate": 1.6669084578557959e-06, + "loss": 0.0002, + "step": 25374 + }, + { + "epoch": 2.750975715524718, + "grad_norm": 4.233335018157959, + "learning_rate": 1.6661830842884087e-06, + "loss": 0.0085, + "step": 25375 + }, + { + "epoch": 2.751084128360798, + "grad_norm": 0.30897727608680725, + "learning_rate": 1.6654577107210214e-06, + "loss": 0.0011, + "step": 25376 + }, + { + "epoch": 2.7511925411968776, + "grad_norm": 4.35746955871582, + "learning_rate": 1.6647323371536342e-06, + "loss": 0.0599, + "step": 25377 + }, + { + "epoch": 2.7513009540329576, + "grad_norm": 3.7186200618743896, + "learning_rate": 1.6640069635862472e-06, + "loss": 0.0292, + "step": 25378 + }, + { + "epoch": 2.751409366869037, + "grad_norm": 2.672189712524414, + "learning_rate": 1.6632815900188598e-06, + "loss": 0.0357, + "step": 25379 + }, + { + "epoch": 2.751517779705117, + "grad_norm": 2.862581968307495, + "learning_rate": 1.6625562164514728e-06, + "loss": 0.003, + "step": 25380 + }, + { + "epoch": 2.751626192541197, + "grad_norm": 0.4917086958885193, + "learning_rate": 1.6618308428840853e-06, + "loss": 0.0012, + "step": 25381 + }, + { + "epoch": 2.7517346053772767, + "grad_norm": 0.2273269146680832, + "learning_rate": 1.6611054693166983e-06, + "loss": 0.0009, + "step": 25382 + }, + { + "epoch": 2.751843018213356, + "grad_norm": 0.03698209673166275, + "learning_rate": 1.660380095749311e-06, + "loss": 0.0004, + "step": 25383 + }, + { + "epoch": 2.751951431049436, + "grad_norm": 0.2693656086921692, + "learning_rate": 1.6596547221819237e-06, + "loss": 0.0021, + "step": 25384 + }, + { + "epoch": 2.752059843885516, + "grad_norm": 0.05418986454606056, + "learning_rate": 1.6589293486145367e-06, + "loss": 0.0003, + "step": 25385 + }, + { + "epoch": 2.7521682567215957, + "grad_norm": 0.3644929826259613, + "learning_rate": 1.6582039750471492e-06, + "loss": 0.0019, + "step": 25386 + }, + { + "epoch": 2.7522766695576757, + "grad_norm": 0.0379604808986187, + "learning_rate": 1.6574786014797622e-06, + "loss": 0.0003, + "step": 25387 + }, + { + "epoch": 2.7523850823937552, + "grad_norm": 0.020995013415813446, + "learning_rate": 1.656753227912375e-06, + "loss": 0.0002, + "step": 25388 + }, + { + "epoch": 2.752493495229835, + "grad_norm": 0.9336962699890137, + "learning_rate": 1.6560278543449878e-06, + "loss": 0.0052, + "step": 25389 + }, + { + "epoch": 2.752601908065915, + "grad_norm": 0.172201007604599, + "learning_rate": 1.6553024807776005e-06, + "loss": 0.0012, + "step": 25390 + }, + { + "epoch": 2.7527103209019947, + "grad_norm": 0.028096182271838188, + "learning_rate": 1.6545771072102135e-06, + "loss": 0.0004, + "step": 25391 + }, + { + "epoch": 2.7528187337380747, + "grad_norm": 0.22459810972213745, + "learning_rate": 1.653851733642826e-06, + "loss": 0.0015, + "step": 25392 + }, + { + "epoch": 2.7529271465741543, + "grad_norm": 0.022253645583987236, + "learning_rate": 1.653126360075439e-06, + "loss": 0.0001, + "step": 25393 + }, + { + "epoch": 2.7530355594102343, + "grad_norm": 0.0661705881357193, + "learning_rate": 1.6524009865080517e-06, + "loss": 0.0004, + "step": 25394 + }, + { + "epoch": 2.753143972246314, + "grad_norm": 2.969851016998291, + "learning_rate": 1.6516756129406647e-06, + "loss": 0.0051, + "step": 25395 + }, + { + "epoch": 2.753252385082394, + "grad_norm": 0.05069350451231003, + "learning_rate": 1.6509502393732774e-06, + "loss": 0.0003, + "step": 25396 + }, + { + "epoch": 2.7533607979184733, + "grad_norm": 13.226797103881836, + "learning_rate": 1.65022486580589e-06, + "loss": 0.0207, + "step": 25397 + }, + { + "epoch": 2.7534692107545533, + "grad_norm": 0.8545718193054199, + "learning_rate": 1.649499492238503e-06, + "loss": 0.0037, + "step": 25398 + }, + { + "epoch": 2.7535776235906333, + "grad_norm": 0.1883523017168045, + "learning_rate": 1.6487741186711156e-06, + "loss": 0.0011, + "step": 25399 + }, + { + "epoch": 2.753686036426713, + "grad_norm": 0.21115848422050476, + "learning_rate": 1.6480487451037285e-06, + "loss": 0.0012, + "step": 25400 + }, + { + "epoch": 2.753794449262793, + "grad_norm": 11.44749641418457, + "learning_rate": 1.6473233715363413e-06, + "loss": 0.0179, + "step": 25401 + }, + { + "epoch": 2.7539028620988724, + "grad_norm": 10.813238143920898, + "learning_rate": 1.646597997968954e-06, + "loss": 0.0338, + "step": 25402 + }, + { + "epoch": 2.7540112749349523, + "grad_norm": 0.4534319341182709, + "learning_rate": 1.6458726244015669e-06, + "loss": 0.0018, + "step": 25403 + }, + { + "epoch": 2.7541196877710323, + "grad_norm": 0.01798819750547409, + "learning_rate": 1.6451472508341799e-06, + "loss": 0.0002, + "step": 25404 + }, + { + "epoch": 2.754228100607112, + "grad_norm": 0.7267655730247498, + "learning_rate": 1.6444218772667924e-06, + "loss": 0.002, + "step": 25405 + }, + { + "epoch": 2.7543365134431914, + "grad_norm": 0.9661199450492859, + "learning_rate": 1.6436965036994054e-06, + "loss": 0.0039, + "step": 25406 + }, + { + "epoch": 2.7544449262792714, + "grad_norm": 0.2138792872428894, + "learning_rate": 1.642971130132018e-06, + "loss": 0.0018, + "step": 25407 + }, + { + "epoch": 2.7545533391153514, + "grad_norm": 0.3036765456199646, + "learning_rate": 1.642245756564631e-06, + "loss": 0.0014, + "step": 25408 + }, + { + "epoch": 2.754661751951431, + "grad_norm": 0.014215840958058834, + "learning_rate": 1.6415203829972438e-06, + "loss": 0.0001, + "step": 25409 + }, + { + "epoch": 2.754770164787511, + "grad_norm": 1.319041132926941, + "learning_rate": 1.6407950094298565e-06, + "loss": 0.0432, + "step": 25410 + }, + { + "epoch": 2.7548785776235905, + "grad_norm": 2.5209054946899414, + "learning_rate": 1.6400696358624693e-06, + "loss": 0.0072, + "step": 25411 + }, + { + "epoch": 2.7549869904596704, + "grad_norm": 2.9009225368499756, + "learning_rate": 1.6393442622950819e-06, + "loss": 0.0042, + "step": 25412 + }, + { + "epoch": 2.7550954032957504, + "grad_norm": 0.01203666627407074, + "learning_rate": 1.6386188887276949e-06, + "loss": 0.0001, + "step": 25413 + }, + { + "epoch": 2.75520381613183, + "grad_norm": 2.582245349884033, + "learning_rate": 1.6378935151603077e-06, + "loss": 0.0752, + "step": 25414 + }, + { + "epoch": 2.75531222896791, + "grad_norm": 0.40203937888145447, + "learning_rate": 1.6371681415929204e-06, + "loss": 0.0016, + "step": 25415 + }, + { + "epoch": 2.7554206418039895, + "grad_norm": 0.09494637697935104, + "learning_rate": 1.6364427680255332e-06, + "loss": 0.0008, + "step": 25416 + }, + { + "epoch": 2.7555290546400695, + "grad_norm": 4.025813579559326, + "learning_rate": 1.6357173944581462e-06, + "loss": 0.0281, + "step": 25417 + }, + { + "epoch": 2.755637467476149, + "grad_norm": 0.02015460841357708, + "learning_rate": 1.6349920208907588e-06, + "loss": 0.0001, + "step": 25418 + }, + { + "epoch": 2.755745880312229, + "grad_norm": 0.06057926267385483, + "learning_rate": 1.6342666473233718e-06, + "loss": 0.0002, + "step": 25419 + }, + { + "epoch": 2.7558542931483085, + "grad_norm": 0.317180871963501, + "learning_rate": 1.6335412737559843e-06, + "loss": 0.001, + "step": 25420 + }, + { + "epoch": 2.7559627059843885, + "grad_norm": 5.165653705596924, + "learning_rate": 1.6328159001885973e-06, + "loss": 0.0142, + "step": 25421 + }, + { + "epoch": 2.7560711188204685, + "grad_norm": 3.6922335624694824, + "learning_rate": 1.63209052662121e-06, + "loss": 0.0415, + "step": 25422 + }, + { + "epoch": 2.756179531656548, + "grad_norm": 0.12009845674037933, + "learning_rate": 1.6313651530538229e-06, + "loss": 0.0005, + "step": 25423 + }, + { + "epoch": 2.756287944492628, + "grad_norm": 11.722025871276855, + "learning_rate": 1.6306397794864357e-06, + "loss": 0.0273, + "step": 25424 + }, + { + "epoch": 2.7563963573287076, + "grad_norm": 1.222543716430664, + "learning_rate": 1.6299144059190482e-06, + "loss": 0.013, + "step": 25425 + }, + { + "epoch": 2.7565047701647876, + "grad_norm": 7.306241035461426, + "learning_rate": 1.6291890323516612e-06, + "loss": 0.0218, + "step": 25426 + }, + { + "epoch": 2.7566131830008676, + "grad_norm": 4.446935176849365, + "learning_rate": 1.628463658784274e-06, + "loss": 0.0054, + "step": 25427 + }, + { + "epoch": 2.756721595836947, + "grad_norm": 4.190691947937012, + "learning_rate": 1.6277382852168868e-06, + "loss": 0.016, + "step": 25428 + }, + { + "epoch": 2.7568300086730266, + "grad_norm": 0.2145371437072754, + "learning_rate": 1.6270129116494996e-06, + "loss": 0.0019, + "step": 25429 + }, + { + "epoch": 2.7569384215091066, + "grad_norm": 0.03071781061589718, + "learning_rate": 1.6262875380821125e-06, + "loss": 0.0002, + "step": 25430 + }, + { + "epoch": 2.7570468343451866, + "grad_norm": 0.46838656067848206, + "learning_rate": 1.6255621645147251e-06, + "loss": 0.001, + "step": 25431 + }, + { + "epoch": 2.757155247181266, + "grad_norm": 0.5996551513671875, + "learning_rate": 1.624836790947338e-06, + "loss": 0.0038, + "step": 25432 + }, + { + "epoch": 2.757263660017346, + "grad_norm": 4.560791015625, + "learning_rate": 1.6241114173799507e-06, + "loss": 0.0076, + "step": 25433 + }, + { + "epoch": 2.7573720728534257, + "grad_norm": 0.2414545714855194, + "learning_rate": 1.6233860438125637e-06, + "loss": 0.0014, + "step": 25434 + }, + { + "epoch": 2.7574804856895057, + "grad_norm": 0.7326303124427795, + "learning_rate": 1.6226606702451764e-06, + "loss": 0.0023, + "step": 25435 + }, + { + "epoch": 2.7575888985255856, + "grad_norm": 7.188728332519531, + "learning_rate": 1.6219352966777892e-06, + "loss": 0.0104, + "step": 25436 + }, + { + "epoch": 2.757697311361665, + "grad_norm": 2.3554446697235107, + "learning_rate": 1.621209923110402e-06, + "loss": 0.009, + "step": 25437 + }, + { + "epoch": 2.757805724197745, + "grad_norm": 2.0445799827575684, + "learning_rate": 1.6204845495430146e-06, + "loss": 0.0038, + "step": 25438 + }, + { + "epoch": 2.7579141370338247, + "grad_norm": 8.11579704284668, + "learning_rate": 1.6197591759756275e-06, + "loss": 0.0337, + "step": 25439 + }, + { + "epoch": 2.7580225498699047, + "grad_norm": 0.1360407918691635, + "learning_rate": 1.6190338024082403e-06, + "loss": 0.0003, + "step": 25440 + }, + { + "epoch": 2.7581309627059842, + "grad_norm": 3.249338388442993, + "learning_rate": 1.6183084288408531e-06, + "loss": 0.0071, + "step": 25441 + }, + { + "epoch": 2.758239375542064, + "grad_norm": 3.4644317626953125, + "learning_rate": 1.6175830552734659e-06, + "loss": 0.046, + "step": 25442 + }, + { + "epoch": 2.7583477883781438, + "grad_norm": 0.2580900192260742, + "learning_rate": 1.6168576817060789e-06, + "loss": 0.0008, + "step": 25443 + }, + { + "epoch": 2.7584562012142237, + "grad_norm": 3.552654266357422, + "learning_rate": 1.6161323081386914e-06, + "loss": 0.0168, + "step": 25444 + }, + { + "epoch": 2.7585646140503037, + "grad_norm": 2.5544497966766357, + "learning_rate": 1.6154069345713044e-06, + "loss": 0.003, + "step": 25445 + }, + { + "epoch": 2.7586730268863833, + "grad_norm": 0.15289220213890076, + "learning_rate": 1.614681561003917e-06, + "loss": 0.001, + "step": 25446 + }, + { + "epoch": 2.7587814397224633, + "grad_norm": 0.48373642563819885, + "learning_rate": 1.61395618743653e-06, + "loss": 0.0012, + "step": 25447 + }, + { + "epoch": 2.758889852558543, + "grad_norm": 0.11995875090360641, + "learning_rate": 1.6132308138691428e-06, + "loss": 0.0004, + "step": 25448 + }, + { + "epoch": 2.758998265394623, + "grad_norm": 1.9975348711013794, + "learning_rate": 1.6125054403017558e-06, + "loss": 0.0038, + "step": 25449 + }, + { + "epoch": 2.7591066782307028, + "grad_norm": 1.4071052074432373, + "learning_rate": 1.6117800667343683e-06, + "loss": 0.0028, + "step": 25450 + }, + { + "epoch": 2.7592150910667823, + "grad_norm": 0.044328778982162476, + "learning_rate": 1.6110546931669813e-06, + "loss": 0.0003, + "step": 25451 + }, + { + "epoch": 2.759323503902862, + "grad_norm": 1.6635891199111938, + "learning_rate": 1.6103293195995939e-06, + "loss": 0.0379, + "step": 25452 + }, + { + "epoch": 2.759431916738942, + "grad_norm": 5.766568183898926, + "learning_rate": 1.6096039460322067e-06, + "loss": 0.0032, + "step": 25453 + }, + { + "epoch": 2.759540329575022, + "grad_norm": 2.3052217960357666, + "learning_rate": 1.6088785724648194e-06, + "loss": 0.0034, + "step": 25454 + }, + { + "epoch": 2.7596487424111014, + "grad_norm": 0.390340656042099, + "learning_rate": 1.6081531988974322e-06, + "loss": 0.0024, + "step": 25455 + }, + { + "epoch": 2.7597571552471813, + "grad_norm": 1.3840769529342651, + "learning_rate": 1.6074278253300452e-06, + "loss": 0.0009, + "step": 25456 + }, + { + "epoch": 2.759865568083261, + "grad_norm": 0.03900754824280739, + "learning_rate": 1.6067024517626578e-06, + "loss": 0.0002, + "step": 25457 + }, + { + "epoch": 2.759973980919341, + "grad_norm": 2.458038091659546, + "learning_rate": 1.6059770781952708e-06, + "loss": 0.0039, + "step": 25458 + }, + { + "epoch": 2.760082393755421, + "grad_norm": 5.776155948638916, + "learning_rate": 1.6052517046278833e-06, + "loss": 0.0055, + "step": 25459 + }, + { + "epoch": 2.7601908065915004, + "grad_norm": 2.629720687866211, + "learning_rate": 1.6045263310604963e-06, + "loss": 0.007, + "step": 25460 + }, + { + "epoch": 2.7602992194275804, + "grad_norm": 0.10514616966247559, + "learning_rate": 1.603800957493109e-06, + "loss": 0.0007, + "step": 25461 + }, + { + "epoch": 2.76040763226366, + "grad_norm": 1.602982997894287, + "learning_rate": 1.603075583925722e-06, + "loss": 0.0378, + "step": 25462 + }, + { + "epoch": 2.76051604509974, + "grad_norm": 2.650937557220459, + "learning_rate": 1.6023502103583347e-06, + "loss": 0.0084, + "step": 25463 + }, + { + "epoch": 2.7606244579358195, + "grad_norm": 0.8633095026016235, + "learning_rate": 1.6016248367909477e-06, + "loss": 0.0018, + "step": 25464 + }, + { + "epoch": 2.7607328707718994, + "grad_norm": 1.875847339630127, + "learning_rate": 1.6008994632235602e-06, + "loss": 0.0331, + "step": 25465 + }, + { + "epoch": 2.760841283607979, + "grad_norm": 0.7583966851234436, + "learning_rate": 1.600174089656173e-06, + "loss": 0.0034, + "step": 25466 + }, + { + "epoch": 2.760949696444059, + "grad_norm": 0.16511985659599304, + "learning_rate": 1.5994487160887858e-06, + "loss": 0.0009, + "step": 25467 + }, + { + "epoch": 2.761058109280139, + "grad_norm": 0.2041768580675125, + "learning_rate": 1.5987233425213986e-06, + "loss": 0.0005, + "step": 25468 + }, + { + "epoch": 2.7611665221162185, + "grad_norm": 0.07347875088453293, + "learning_rate": 1.5979979689540115e-06, + "loss": 0.0004, + "step": 25469 + }, + { + "epoch": 2.7612749349522985, + "grad_norm": 0.1806480586528778, + "learning_rate": 1.5972725953866241e-06, + "loss": 0.0011, + "step": 25470 + }, + { + "epoch": 2.761383347788378, + "grad_norm": 4.248746395111084, + "learning_rate": 1.596547221819237e-06, + "loss": 0.0102, + "step": 25471 + }, + { + "epoch": 2.761491760624458, + "grad_norm": 1.5966053009033203, + "learning_rate": 1.5958218482518497e-06, + "loss": 0.0022, + "step": 25472 + }, + { + "epoch": 2.761600173460538, + "grad_norm": 0.9506717920303345, + "learning_rate": 1.5950964746844627e-06, + "loss": 0.0344, + "step": 25473 + }, + { + "epoch": 2.7617085862966175, + "grad_norm": 0.6296620965003967, + "learning_rate": 1.5943711011170754e-06, + "loss": 0.0018, + "step": 25474 + }, + { + "epoch": 2.761816999132697, + "grad_norm": 0.3223021626472473, + "learning_rate": 1.5936457275496884e-06, + "loss": 0.0013, + "step": 25475 + }, + { + "epoch": 2.761925411968777, + "grad_norm": 0.6929975152015686, + "learning_rate": 1.592920353982301e-06, + "loss": 0.0027, + "step": 25476 + }, + { + "epoch": 2.762033824804857, + "grad_norm": 0.22419574856758118, + "learning_rate": 1.592194980414914e-06, + "loss": 0.001, + "step": 25477 + }, + { + "epoch": 2.7621422376409366, + "grad_norm": 1.08229398727417, + "learning_rate": 1.5914696068475266e-06, + "loss": 0.0049, + "step": 25478 + }, + { + "epoch": 2.7622506504770166, + "grad_norm": 2.031362771987915, + "learning_rate": 1.5907442332801393e-06, + "loss": 0.0101, + "step": 25479 + }, + { + "epoch": 2.762359063313096, + "grad_norm": 1.57134211063385, + "learning_rate": 1.5900188597127523e-06, + "loss": 0.0036, + "step": 25480 + }, + { + "epoch": 2.762467476149176, + "grad_norm": 0.1506151705980301, + "learning_rate": 1.5892934861453649e-06, + "loss": 0.0012, + "step": 25481 + }, + { + "epoch": 2.762575888985256, + "grad_norm": 0.2307075709104538, + "learning_rate": 1.5885681125779779e-06, + "loss": 0.0011, + "step": 25482 + }, + { + "epoch": 2.7626843018213356, + "grad_norm": 4.521736145019531, + "learning_rate": 1.5878427390105904e-06, + "loss": 0.023, + "step": 25483 + }, + { + "epoch": 2.762792714657415, + "grad_norm": 0.17508913576602936, + "learning_rate": 1.5871173654432034e-06, + "loss": 0.0007, + "step": 25484 + }, + { + "epoch": 2.762901127493495, + "grad_norm": 0.43903636932373047, + "learning_rate": 1.586391991875816e-06, + "loss": 0.0012, + "step": 25485 + }, + { + "epoch": 2.763009540329575, + "grad_norm": 0.5364540815353394, + "learning_rate": 1.585666618308429e-06, + "loss": 0.0016, + "step": 25486 + }, + { + "epoch": 2.7631179531656547, + "grad_norm": 0.03713678941130638, + "learning_rate": 1.5849412447410418e-06, + "loss": 0.0003, + "step": 25487 + }, + { + "epoch": 2.7632263660017347, + "grad_norm": 0.06494444608688354, + "learning_rate": 1.5842158711736548e-06, + "loss": 0.0004, + "step": 25488 + }, + { + "epoch": 2.763334778837814, + "grad_norm": 0.0987180843949318, + "learning_rate": 1.5834904976062673e-06, + "loss": 0.0006, + "step": 25489 + }, + { + "epoch": 2.763443191673894, + "grad_norm": 0.09477730840444565, + "learning_rate": 1.5827651240388803e-06, + "loss": 0.0004, + "step": 25490 + }, + { + "epoch": 2.763551604509974, + "grad_norm": 0.7180166840553284, + "learning_rate": 1.5820397504714929e-06, + "loss": 0.0031, + "step": 25491 + }, + { + "epoch": 2.7636600173460537, + "grad_norm": 0.7326316833496094, + "learning_rate": 1.5813143769041057e-06, + "loss": 0.0022, + "step": 25492 + }, + { + "epoch": 2.7637684301821337, + "grad_norm": 0.02180751971900463, + "learning_rate": 1.5805890033367187e-06, + "loss": 0.0002, + "step": 25493 + }, + { + "epoch": 2.7638768430182132, + "grad_norm": 0.093333400785923, + "learning_rate": 1.5798636297693312e-06, + "loss": 0.0004, + "step": 25494 + }, + { + "epoch": 2.763985255854293, + "grad_norm": 0.026477819308638573, + "learning_rate": 1.5791382562019442e-06, + "loss": 0.0003, + "step": 25495 + }, + { + "epoch": 2.764093668690373, + "grad_norm": 11.438267707824707, + "learning_rate": 1.5784128826345568e-06, + "loss": 0.0316, + "step": 25496 + }, + { + "epoch": 2.7642020815264527, + "grad_norm": 0.06869212538003922, + "learning_rate": 1.5776875090671698e-06, + "loss": 0.0003, + "step": 25497 + }, + { + "epoch": 2.7643104943625323, + "grad_norm": 0.6292932629585266, + "learning_rate": 1.5769621354997823e-06, + "loss": 0.0027, + "step": 25498 + }, + { + "epoch": 2.7644189071986123, + "grad_norm": 0.28047940135002136, + "learning_rate": 1.5762367619323953e-06, + "loss": 0.0011, + "step": 25499 + }, + { + "epoch": 2.7645273200346923, + "grad_norm": 0.9730250835418701, + "learning_rate": 1.5755113883650081e-06, + "loss": 0.0024, + "step": 25500 + }, + { + "epoch": 2.764635732870772, + "grad_norm": 0.1555224508047104, + "learning_rate": 1.574786014797621e-06, + "loss": 0.0011, + "step": 25501 + }, + { + "epoch": 2.764744145706852, + "grad_norm": 0.40687552094459534, + "learning_rate": 1.5740606412302337e-06, + "loss": 0.0014, + "step": 25502 + }, + { + "epoch": 2.7648525585429313, + "grad_norm": 7.966742992401123, + "learning_rate": 1.5733352676628467e-06, + "loss": 0.0267, + "step": 25503 + }, + { + "epoch": 2.7649609713790113, + "grad_norm": 0.021060973405838013, + "learning_rate": 1.5726098940954592e-06, + "loss": 0.0002, + "step": 25504 + }, + { + "epoch": 2.7650693842150913, + "grad_norm": 0.019318655133247375, + "learning_rate": 1.5718845205280722e-06, + "loss": 0.0002, + "step": 25505 + }, + { + "epoch": 2.765177797051171, + "grad_norm": 0.0466323085129261, + "learning_rate": 1.571159146960685e-06, + "loss": 0.0003, + "step": 25506 + }, + { + "epoch": 2.7652862098872504, + "grad_norm": 0.022627735510468483, + "learning_rate": 1.5704337733932976e-06, + "loss": 0.0003, + "step": 25507 + }, + { + "epoch": 2.7653946227233304, + "grad_norm": 2.537118673324585, + "learning_rate": 1.5697083998259105e-06, + "loss": 0.0031, + "step": 25508 + }, + { + "epoch": 2.7655030355594103, + "grad_norm": 0.16575263440608978, + "learning_rate": 1.5689830262585231e-06, + "loss": 0.0006, + "step": 25509 + }, + { + "epoch": 2.76561144839549, + "grad_norm": 2.431966781616211, + "learning_rate": 1.568257652691136e-06, + "loss": 0.0031, + "step": 25510 + }, + { + "epoch": 2.76571986123157, + "grad_norm": 0.2117435336112976, + "learning_rate": 1.5675322791237489e-06, + "loss": 0.0006, + "step": 25511 + }, + { + "epoch": 2.7658282740676494, + "grad_norm": 5.857620716094971, + "learning_rate": 1.5668069055563617e-06, + "loss": 0.0127, + "step": 25512 + }, + { + "epoch": 2.7659366869037294, + "grad_norm": 0.2298690527677536, + "learning_rate": 1.5660815319889744e-06, + "loss": 0.0005, + "step": 25513 + }, + { + "epoch": 2.7660450997398094, + "grad_norm": 0.11877159774303436, + "learning_rate": 1.5653561584215874e-06, + "loss": 0.0007, + "step": 25514 + }, + { + "epoch": 2.766153512575889, + "grad_norm": 0.08042991161346436, + "learning_rate": 1.5646307848542e-06, + "loss": 0.0005, + "step": 25515 + }, + { + "epoch": 2.766261925411969, + "grad_norm": 7.024108409881592, + "learning_rate": 1.563905411286813e-06, + "loss": 0.006, + "step": 25516 + }, + { + "epoch": 2.7663703382480485, + "grad_norm": 0.024191875010728836, + "learning_rate": 1.5631800377194256e-06, + "loss": 0.0003, + "step": 25517 + }, + { + "epoch": 2.7664787510841284, + "grad_norm": 0.7168506979942322, + "learning_rate": 1.5624546641520385e-06, + "loss": 0.0006, + "step": 25518 + }, + { + "epoch": 2.7665871639202084, + "grad_norm": 3.1175811290740967, + "learning_rate": 1.5617292905846513e-06, + "loss": 0.0019, + "step": 25519 + }, + { + "epoch": 2.766695576756288, + "grad_norm": 0.10656717419624329, + "learning_rate": 1.5610039170172639e-06, + "loss": 0.0008, + "step": 25520 + }, + { + "epoch": 2.7668039895923675, + "grad_norm": 1.4489080905914307, + "learning_rate": 1.5602785434498769e-06, + "loss": 0.0018, + "step": 25521 + }, + { + "epoch": 2.7669124024284475, + "grad_norm": 0.026241322979331017, + "learning_rate": 1.5595531698824894e-06, + "loss": 0.0003, + "step": 25522 + }, + { + "epoch": 2.7670208152645275, + "grad_norm": 0.03377898409962654, + "learning_rate": 1.5588277963151024e-06, + "loss": 0.0003, + "step": 25523 + }, + { + "epoch": 2.767129228100607, + "grad_norm": 0.02943534590303898, + "learning_rate": 1.5581024227477152e-06, + "loss": 0.0003, + "step": 25524 + }, + { + "epoch": 2.767237640936687, + "grad_norm": 1.6365954875946045, + "learning_rate": 1.557377049180328e-06, + "loss": 0.006, + "step": 25525 + }, + { + "epoch": 2.7673460537727665, + "grad_norm": 0.17199663817882538, + "learning_rate": 1.5566516756129408e-06, + "loss": 0.0009, + "step": 25526 + }, + { + "epoch": 2.7674544666088465, + "grad_norm": 0.15174050629138947, + "learning_rate": 1.5559263020455538e-06, + "loss": 0.0007, + "step": 25527 + }, + { + "epoch": 2.7675628794449265, + "grad_norm": 0.33126968145370483, + "learning_rate": 1.5552009284781663e-06, + "loss": 0.0015, + "step": 25528 + }, + { + "epoch": 2.767671292281006, + "grad_norm": 0.9438703060150146, + "learning_rate": 1.5544755549107793e-06, + "loss": 0.0021, + "step": 25529 + }, + { + "epoch": 2.7677797051170856, + "grad_norm": 0.02559368871152401, + "learning_rate": 1.5537501813433919e-06, + "loss": 0.0003, + "step": 25530 + }, + { + "epoch": 2.7678881179531656, + "grad_norm": 0.0679592415690422, + "learning_rate": 1.5530248077760049e-06, + "loss": 0.0002, + "step": 25531 + }, + { + "epoch": 2.7679965307892456, + "grad_norm": 0.3747839629650116, + "learning_rate": 1.5522994342086177e-06, + "loss": 0.0013, + "step": 25532 + }, + { + "epoch": 2.768104943625325, + "grad_norm": 1.6374962329864502, + "learning_rate": 1.5515740606412302e-06, + "loss": 0.0282, + "step": 25533 + }, + { + "epoch": 2.768213356461405, + "grad_norm": 0.03008667193353176, + "learning_rate": 1.5508486870738432e-06, + "loss": 0.0002, + "step": 25534 + }, + { + "epoch": 2.7683217692974846, + "grad_norm": 1.4315838813781738, + "learning_rate": 1.5501233135064558e-06, + "loss": 0.0078, + "step": 25535 + }, + { + "epoch": 2.7684301821335646, + "grad_norm": 0.14978466928005219, + "learning_rate": 1.5493979399390688e-06, + "loss": 0.0007, + "step": 25536 + }, + { + "epoch": 2.7685385949696446, + "grad_norm": 0.6313934326171875, + "learning_rate": 1.5486725663716816e-06, + "loss": 0.0023, + "step": 25537 + }, + { + "epoch": 2.768647007805724, + "grad_norm": 0.028461506590247154, + "learning_rate": 1.5479471928042943e-06, + "loss": 0.0002, + "step": 25538 + }, + { + "epoch": 2.768755420641804, + "grad_norm": 2.109570026397705, + "learning_rate": 1.5472218192369071e-06, + "loss": 0.0247, + "step": 25539 + }, + { + "epoch": 2.7688638334778837, + "grad_norm": 0.0555730015039444, + "learning_rate": 1.54649644566952e-06, + "loss": 0.0003, + "step": 25540 + }, + { + "epoch": 2.7689722463139637, + "grad_norm": 0.012637940235435963, + "learning_rate": 1.5457710721021327e-06, + "loss": 0.0001, + "step": 25541 + }, + { + "epoch": 2.769080659150043, + "grad_norm": 9.634576797485352, + "learning_rate": 1.5450456985347457e-06, + "loss": 0.0154, + "step": 25542 + }, + { + "epoch": 2.769189071986123, + "grad_norm": 3.713188648223877, + "learning_rate": 1.5443203249673582e-06, + "loss": 0.0799, + "step": 25543 + }, + { + "epoch": 2.7692974848222027, + "grad_norm": 0.45087912678718567, + "learning_rate": 1.5435949513999712e-06, + "loss": 0.0014, + "step": 25544 + }, + { + "epoch": 2.7694058976582827, + "grad_norm": 0.8124468326568604, + "learning_rate": 1.542869577832584e-06, + "loss": 0.0033, + "step": 25545 + }, + { + "epoch": 2.7695143104943627, + "grad_norm": 1.9493210315704346, + "learning_rate": 1.5421442042651966e-06, + "loss": 0.0284, + "step": 25546 + }, + { + "epoch": 2.7696227233304422, + "grad_norm": 0.04865284264087677, + "learning_rate": 1.5414188306978096e-06, + "loss": 0.0003, + "step": 25547 + }, + { + "epoch": 2.7697311361665222, + "grad_norm": 18.464757919311523, + "learning_rate": 1.5406934571304221e-06, + "loss": 0.0374, + "step": 25548 + }, + { + "epoch": 2.7698395490026018, + "grad_norm": 0.2715914845466614, + "learning_rate": 1.5399680835630351e-06, + "loss": 0.0015, + "step": 25549 + }, + { + "epoch": 2.7699479618386817, + "grad_norm": 0.060292765498161316, + "learning_rate": 1.5392427099956479e-06, + "loss": 0.0004, + "step": 25550 + }, + { + "epoch": 2.7700563746747617, + "grad_norm": 11.522427558898926, + "learning_rate": 1.5385173364282607e-06, + "loss": 0.0354, + "step": 25551 + }, + { + "epoch": 2.7701647875108413, + "grad_norm": 1.874566674232483, + "learning_rate": 1.5377919628608734e-06, + "loss": 0.099, + "step": 25552 + }, + { + "epoch": 2.770273200346921, + "grad_norm": 1.9446709156036377, + "learning_rate": 1.5370665892934864e-06, + "loss": 0.0126, + "step": 25553 + }, + { + "epoch": 2.770381613183001, + "grad_norm": 0.17079396545886993, + "learning_rate": 1.536341215726099e-06, + "loss": 0.0003, + "step": 25554 + }, + { + "epoch": 2.770490026019081, + "grad_norm": 7.186304569244385, + "learning_rate": 1.535615842158712e-06, + "loss": 0.0335, + "step": 25555 + }, + { + "epoch": 2.7705984388551603, + "grad_norm": 0.5851389169692993, + "learning_rate": 1.5348904685913246e-06, + "loss": 0.0011, + "step": 25556 + }, + { + "epoch": 2.7707068516912403, + "grad_norm": 15.79233455657959, + "learning_rate": 1.5341650950239375e-06, + "loss": 0.0093, + "step": 25557 + }, + { + "epoch": 2.77081526452732, + "grad_norm": 0.41950905323028564, + "learning_rate": 1.5334397214565503e-06, + "loss": 0.001, + "step": 25558 + }, + { + "epoch": 2.7709236773634, + "grad_norm": 0.8218328356742859, + "learning_rate": 1.5327143478891631e-06, + "loss": 0.0021, + "step": 25559 + }, + { + "epoch": 2.77103209019948, + "grad_norm": 0.7151196002960205, + "learning_rate": 1.5319889743217759e-06, + "loss": 0.0009, + "step": 25560 + }, + { + "epoch": 2.7711405030355594, + "grad_norm": 0.06603129953145981, + "learning_rate": 1.5312636007543885e-06, + "loss": 0.0004, + "step": 25561 + }, + { + "epoch": 2.7712489158716394, + "grad_norm": 0.8584426045417786, + "learning_rate": 1.5305382271870014e-06, + "loss": 0.0012, + "step": 25562 + }, + { + "epoch": 2.771357328707719, + "grad_norm": 1.8604384660720825, + "learning_rate": 1.5298128536196142e-06, + "loss": 0.0447, + "step": 25563 + }, + { + "epoch": 2.771465741543799, + "grad_norm": 2.235133647918701, + "learning_rate": 1.529087480052227e-06, + "loss": 0.0252, + "step": 25564 + }, + { + "epoch": 2.7715741543798784, + "grad_norm": 1.7005988359451294, + "learning_rate": 1.5283621064848398e-06, + "loss": 0.0043, + "step": 25565 + }, + { + "epoch": 2.7716825672159584, + "grad_norm": 3.112257957458496, + "learning_rate": 1.5276367329174528e-06, + "loss": 0.0125, + "step": 25566 + }, + { + "epoch": 2.771790980052038, + "grad_norm": 1.274471640586853, + "learning_rate": 1.5269113593500653e-06, + "loss": 0.002, + "step": 25567 + }, + { + "epoch": 2.771899392888118, + "grad_norm": 0.015466113574802876, + "learning_rate": 1.5261859857826783e-06, + "loss": 0.0002, + "step": 25568 + }, + { + "epoch": 2.772007805724198, + "grad_norm": 0.0650678426027298, + "learning_rate": 1.5254606122152909e-06, + "loss": 0.0004, + "step": 25569 + }, + { + "epoch": 2.7721162185602775, + "grad_norm": 0.8019073605537415, + "learning_rate": 1.5247352386479039e-06, + "loss": 0.0035, + "step": 25570 + }, + { + "epoch": 2.7722246313963574, + "grad_norm": 0.04058261215686798, + "learning_rate": 1.5240098650805167e-06, + "loss": 0.0002, + "step": 25571 + }, + { + "epoch": 2.772333044232437, + "grad_norm": 0.050363387912511826, + "learning_rate": 1.5232844915131294e-06, + "loss": 0.0003, + "step": 25572 + }, + { + "epoch": 2.772441457068517, + "grad_norm": 0.022014686837792397, + "learning_rate": 1.5225591179457422e-06, + "loss": 0.0003, + "step": 25573 + }, + { + "epoch": 2.772549869904597, + "grad_norm": 0.021743513643741608, + "learning_rate": 1.5218337443783548e-06, + "loss": 0.0003, + "step": 25574 + }, + { + "epoch": 2.7726582827406765, + "grad_norm": 7.13040018081665, + "learning_rate": 1.5211083708109678e-06, + "loss": 0.0409, + "step": 25575 + }, + { + "epoch": 2.772766695576756, + "grad_norm": 5.98463249206543, + "learning_rate": 1.5203829972435806e-06, + "loss": 0.1316, + "step": 25576 + }, + { + "epoch": 2.772875108412836, + "grad_norm": 0.13016536831855774, + "learning_rate": 1.5196576236761933e-06, + "loss": 0.0007, + "step": 25577 + }, + { + "epoch": 2.772983521248916, + "grad_norm": 1.8650059700012207, + "learning_rate": 1.5189322501088061e-06, + "loss": 0.0029, + "step": 25578 + }, + { + "epoch": 2.7730919340849955, + "grad_norm": 3.7466681003570557, + "learning_rate": 1.518206876541419e-06, + "loss": 0.0156, + "step": 25579 + }, + { + "epoch": 2.7732003469210755, + "grad_norm": 0.2987041473388672, + "learning_rate": 1.5174815029740317e-06, + "loss": 0.0007, + "step": 25580 + }, + { + "epoch": 2.773308759757155, + "grad_norm": 2.6685690879821777, + "learning_rate": 1.5167561294066447e-06, + "loss": 0.007, + "step": 25581 + }, + { + "epoch": 2.773417172593235, + "grad_norm": 0.018222233280539513, + "learning_rate": 1.5160307558392572e-06, + "loss": 0.0002, + "step": 25582 + }, + { + "epoch": 2.773525585429315, + "grad_norm": 0.06719662994146347, + "learning_rate": 1.5153053822718702e-06, + "loss": 0.0001, + "step": 25583 + }, + { + "epoch": 2.7736339982653946, + "grad_norm": 0.11295773088932037, + "learning_rate": 1.514580008704483e-06, + "loss": 0.0006, + "step": 25584 + }, + { + "epoch": 2.7737424111014746, + "grad_norm": 0.14332075417041779, + "learning_rate": 1.5138546351370958e-06, + "loss": 0.0012, + "step": 25585 + }, + { + "epoch": 2.773850823937554, + "grad_norm": 3.764181137084961, + "learning_rate": 1.5131292615697086e-06, + "loss": 0.0021, + "step": 25586 + }, + { + "epoch": 2.773959236773634, + "grad_norm": 0.03494086489081383, + "learning_rate": 1.5124038880023211e-06, + "loss": 0.0002, + "step": 25587 + }, + { + "epoch": 2.7740676496097136, + "grad_norm": 0.5098382830619812, + "learning_rate": 1.5116785144349341e-06, + "loss": 0.0017, + "step": 25588 + }, + { + "epoch": 2.7741760624457936, + "grad_norm": 1.8213990926742554, + "learning_rate": 1.5109531408675469e-06, + "loss": 0.0022, + "step": 25589 + }, + { + "epoch": 2.774284475281873, + "grad_norm": 3.1136467456817627, + "learning_rate": 1.5102277673001597e-06, + "loss": 0.0243, + "step": 25590 + }, + { + "epoch": 2.774392888117953, + "grad_norm": 6.108033180236816, + "learning_rate": 1.5095023937327724e-06, + "loss": 0.0406, + "step": 25591 + }, + { + "epoch": 2.774501300954033, + "grad_norm": 0.8012953996658325, + "learning_rate": 1.5087770201653854e-06, + "loss": 0.0034, + "step": 25592 + }, + { + "epoch": 2.7746097137901127, + "grad_norm": 4.394175052642822, + "learning_rate": 1.508051646597998e-06, + "loss": 0.0473, + "step": 25593 + }, + { + "epoch": 2.7747181266261927, + "grad_norm": 0.516069769859314, + "learning_rate": 1.507326273030611e-06, + "loss": 0.0014, + "step": 25594 + }, + { + "epoch": 2.774826539462272, + "grad_norm": 0.08964160829782486, + "learning_rate": 1.5066008994632236e-06, + "loss": 0.0007, + "step": 25595 + }, + { + "epoch": 2.774934952298352, + "grad_norm": 0.1301235556602478, + "learning_rate": 1.5058755258958366e-06, + "loss": 0.0008, + "step": 25596 + }, + { + "epoch": 2.775043365134432, + "grad_norm": 6.586206912994385, + "learning_rate": 1.5051501523284493e-06, + "loss": 0.0064, + "step": 25597 + }, + { + "epoch": 2.7751517779705117, + "grad_norm": 0.03579812869429588, + "learning_rate": 1.5044247787610621e-06, + "loss": 0.0002, + "step": 25598 + }, + { + "epoch": 2.7752601908065913, + "grad_norm": 0.08108942955732346, + "learning_rate": 1.5036994051936749e-06, + "loss": 0.0005, + "step": 25599 + }, + { + "epoch": 2.7753686036426712, + "grad_norm": 0.028371622785925865, + "learning_rate": 1.5029740316262879e-06, + "loss": 0.0002, + "step": 25600 + }, + { + "epoch": 2.7754770164787512, + "grad_norm": 4.334846019744873, + "learning_rate": 1.5022486580589004e-06, + "loss": 0.0141, + "step": 25601 + }, + { + "epoch": 2.7755854293148308, + "grad_norm": 8.021038055419922, + "learning_rate": 1.5015232844915132e-06, + "loss": 0.017, + "step": 25602 + }, + { + "epoch": 2.7756938421509108, + "grad_norm": 0.586708128452301, + "learning_rate": 1.500797910924126e-06, + "loss": 0.0015, + "step": 25603 + }, + { + "epoch": 2.7758022549869903, + "grad_norm": 4.804261684417725, + "learning_rate": 1.5000725373567388e-06, + "loss": 0.0145, + "step": 25604 + }, + { + "epoch": 2.7759106678230703, + "grad_norm": 5.0723161697387695, + "learning_rate": 1.4993471637893518e-06, + "loss": 0.0078, + "step": 25605 + }, + { + "epoch": 2.7760190806591503, + "grad_norm": 0.5848421454429626, + "learning_rate": 1.4986217902219643e-06, + "loss": 0.0015, + "step": 25606 + }, + { + "epoch": 2.77612749349523, + "grad_norm": 0.018758973106741905, + "learning_rate": 1.4978964166545773e-06, + "loss": 0.0002, + "step": 25607 + }, + { + "epoch": 2.77623590633131, + "grad_norm": 0.8173112273216248, + "learning_rate": 1.49717104308719e-06, + "loss": 0.0007, + "step": 25608 + }, + { + "epoch": 2.7763443191673893, + "grad_norm": 0.041660457849502563, + "learning_rate": 1.4964456695198029e-06, + "loss": 0.0002, + "step": 25609 + }, + { + "epoch": 2.7764527320034693, + "grad_norm": 0.14563827216625214, + "learning_rate": 1.4957202959524157e-06, + "loss": 0.0008, + "step": 25610 + }, + { + "epoch": 2.776561144839549, + "grad_norm": 12.856391906738281, + "learning_rate": 1.4949949223850284e-06, + "loss": 0.0281, + "step": 25611 + }, + { + "epoch": 2.776669557675629, + "grad_norm": 0.029783517122268677, + "learning_rate": 1.4942695488176412e-06, + "loss": 0.0003, + "step": 25612 + }, + { + "epoch": 2.7767779705117084, + "grad_norm": 10.948652267456055, + "learning_rate": 1.4935441752502542e-06, + "loss": 0.0102, + "step": 25613 + }, + { + "epoch": 2.7768863833477884, + "grad_norm": 0.5367522239685059, + "learning_rate": 1.4928188016828668e-06, + "loss": 0.0012, + "step": 25614 + }, + { + "epoch": 2.7769947961838684, + "grad_norm": 0.02576325088739395, + "learning_rate": 1.4920934281154796e-06, + "loss": 0.0003, + "step": 25615 + }, + { + "epoch": 2.777103209019948, + "grad_norm": 0.28750643134117126, + "learning_rate": 1.4913680545480923e-06, + "loss": 0.0012, + "step": 25616 + }, + { + "epoch": 2.777211621856028, + "grad_norm": 0.12885764241218567, + "learning_rate": 1.4906426809807051e-06, + "loss": 0.0006, + "step": 25617 + }, + { + "epoch": 2.7773200346921074, + "grad_norm": 0.1857021450996399, + "learning_rate": 1.4899173074133181e-06, + "loss": 0.0009, + "step": 25618 + }, + { + "epoch": 2.7774284475281874, + "grad_norm": 0.12044818699359894, + "learning_rate": 1.4891919338459307e-06, + "loss": 0.0005, + "step": 25619 + }, + { + "epoch": 2.7775368603642674, + "grad_norm": 0.04884742200374603, + "learning_rate": 1.4884665602785437e-06, + "loss": 0.0003, + "step": 25620 + }, + { + "epoch": 2.777645273200347, + "grad_norm": 1.2616685628890991, + "learning_rate": 1.4877411867111562e-06, + "loss": 0.0016, + "step": 25621 + }, + { + "epoch": 2.7777536860364265, + "grad_norm": 1.4406501054763794, + "learning_rate": 1.4870158131437692e-06, + "loss": 0.0196, + "step": 25622 + }, + { + "epoch": 2.7778620988725065, + "grad_norm": 0.7065409421920776, + "learning_rate": 1.486290439576382e-06, + "loss": 0.0019, + "step": 25623 + }, + { + "epoch": 2.7779705117085864, + "grad_norm": 0.14997464418411255, + "learning_rate": 1.4855650660089948e-06, + "loss": 0.0007, + "step": 25624 + }, + { + "epoch": 2.778078924544666, + "grad_norm": 1.9625972509384155, + "learning_rate": 1.4848396924416076e-06, + "loss": 0.0324, + "step": 25625 + }, + { + "epoch": 2.778187337380746, + "grad_norm": 2.4916090965270996, + "learning_rate": 1.4841143188742205e-06, + "loss": 0.0279, + "step": 25626 + }, + { + "epoch": 2.7782957502168255, + "grad_norm": 0.14517919719219208, + "learning_rate": 1.4833889453068331e-06, + "loss": 0.0007, + "step": 25627 + }, + { + "epoch": 2.7784041630529055, + "grad_norm": 0.12642696499824524, + "learning_rate": 1.4826635717394459e-06, + "loss": 0.0007, + "step": 25628 + }, + { + "epoch": 2.7785125758889855, + "grad_norm": 3.5019261837005615, + "learning_rate": 1.4819381981720587e-06, + "loss": 0.0154, + "step": 25629 + }, + { + "epoch": 2.778620988725065, + "grad_norm": 0.08810213208198547, + "learning_rate": 1.4812128246046715e-06, + "loss": 0.0006, + "step": 25630 + }, + { + "epoch": 2.7787294015611446, + "grad_norm": 7.275830268859863, + "learning_rate": 1.4804874510372844e-06, + "loss": 0.0435, + "step": 25631 + }, + { + "epoch": 2.7788378143972245, + "grad_norm": 0.03694671764969826, + "learning_rate": 1.479762077469897e-06, + "loss": 0.0002, + "step": 25632 + }, + { + "epoch": 2.7789462272333045, + "grad_norm": 0.028274275362491608, + "learning_rate": 1.47903670390251e-06, + "loss": 0.0004, + "step": 25633 + }, + { + "epoch": 2.779054640069384, + "grad_norm": 0.017778804525732994, + "learning_rate": 1.4783113303351226e-06, + "loss": 0.0002, + "step": 25634 + }, + { + "epoch": 2.779163052905464, + "grad_norm": 0.0039062032010406256, + "learning_rate": 1.4775859567677356e-06, + "loss": 0.0001, + "step": 25635 + }, + { + "epoch": 2.7792714657415436, + "grad_norm": 2.4874186515808105, + "learning_rate": 1.4768605832003483e-06, + "loss": 0.0557, + "step": 25636 + }, + { + "epoch": 2.7793798785776236, + "grad_norm": 0.5628606677055359, + "learning_rate": 1.4761352096329611e-06, + "loss": 0.0009, + "step": 25637 + }, + { + "epoch": 2.7794882914137036, + "grad_norm": 2.7121329307556152, + "learning_rate": 1.4754098360655739e-06, + "loss": 0.0069, + "step": 25638 + }, + { + "epoch": 2.779596704249783, + "grad_norm": 0.04561837762594223, + "learning_rate": 1.4746844624981869e-06, + "loss": 0.0005, + "step": 25639 + }, + { + "epoch": 2.779705117085863, + "grad_norm": 0.5699895620346069, + "learning_rate": 1.4739590889307994e-06, + "loss": 0.0009, + "step": 25640 + }, + { + "epoch": 2.7798135299219426, + "grad_norm": 0.5458474159240723, + "learning_rate": 1.4732337153634122e-06, + "loss": 0.0013, + "step": 25641 + }, + { + "epoch": 2.7799219427580226, + "grad_norm": 0.08236925303936005, + "learning_rate": 1.472508341796025e-06, + "loss": 0.0005, + "step": 25642 + }, + { + "epoch": 2.7800303555941026, + "grad_norm": 12.627121925354004, + "learning_rate": 1.4717829682286378e-06, + "loss": 0.0553, + "step": 25643 + }, + { + "epoch": 2.780138768430182, + "grad_norm": 10.03546142578125, + "learning_rate": 1.4710575946612508e-06, + "loss": 0.0221, + "step": 25644 + }, + { + "epoch": 2.7802471812662617, + "grad_norm": 0.0460822768509388, + "learning_rate": 1.4703322210938633e-06, + "loss": 0.0004, + "step": 25645 + }, + { + "epoch": 2.7803555941023417, + "grad_norm": 6.761567115783691, + "learning_rate": 1.4696068475264763e-06, + "loss": 0.0136, + "step": 25646 + }, + { + "epoch": 2.7804640069384217, + "grad_norm": 0.5295208096504211, + "learning_rate": 1.468881473959089e-06, + "loss": 0.0013, + "step": 25647 + }, + { + "epoch": 2.780572419774501, + "grad_norm": 0.28151974081993103, + "learning_rate": 1.4681561003917019e-06, + "loss": 0.0004, + "step": 25648 + }, + { + "epoch": 2.780680832610581, + "grad_norm": 0.43039482831954956, + "learning_rate": 1.4674307268243147e-06, + "loss": 0.0015, + "step": 25649 + }, + { + "epoch": 2.7807892454466607, + "grad_norm": 0.8088445663452148, + "learning_rate": 1.4667053532569274e-06, + "loss": 0.0493, + "step": 25650 + }, + { + "epoch": 2.7808976582827407, + "grad_norm": 7.94480037689209, + "learning_rate": 1.4659799796895402e-06, + "loss": 0.0203, + "step": 25651 + }, + { + "epoch": 2.7810060711188207, + "grad_norm": 6.537252426147461, + "learning_rate": 1.4652546061221532e-06, + "loss": 0.0218, + "step": 25652 + }, + { + "epoch": 2.7811144839549002, + "grad_norm": 0.2542820870876312, + "learning_rate": 1.4645292325547658e-06, + "loss": 0.0007, + "step": 25653 + }, + { + "epoch": 2.78122289679098, + "grad_norm": 4.035383701324463, + "learning_rate": 1.4638038589873788e-06, + "loss": 0.0064, + "step": 25654 + }, + { + "epoch": 2.7813313096270598, + "grad_norm": 0.48119455575942993, + "learning_rate": 1.4630784854199913e-06, + "loss": 0.0014, + "step": 25655 + }, + { + "epoch": 2.7814397224631398, + "grad_norm": 0.09981264173984528, + "learning_rate": 1.4623531118526041e-06, + "loss": 0.0004, + "step": 25656 + }, + { + "epoch": 2.7815481352992193, + "grad_norm": 0.029027599841356277, + "learning_rate": 1.4616277382852171e-06, + "loss": 0.0002, + "step": 25657 + }, + { + "epoch": 2.7816565481352993, + "grad_norm": 2.614661931991577, + "learning_rate": 1.4609023647178297e-06, + "loss": 0.0163, + "step": 25658 + }, + { + "epoch": 2.781764960971379, + "grad_norm": 0.17854559421539307, + "learning_rate": 1.4601769911504427e-06, + "loss": 0.0008, + "step": 25659 + }, + { + "epoch": 2.781873373807459, + "grad_norm": 0.02331884205341339, + "learning_rate": 1.4594516175830552e-06, + "loss": 0.0002, + "step": 25660 + }, + { + "epoch": 2.781981786643539, + "grad_norm": 9.94495677947998, + "learning_rate": 1.4587262440156682e-06, + "loss": 0.0091, + "step": 25661 + }, + { + "epoch": 2.7820901994796183, + "grad_norm": 1.9925074577331543, + "learning_rate": 1.458000870448281e-06, + "loss": 0.0785, + "step": 25662 + }, + { + "epoch": 2.7821986123156983, + "grad_norm": 0.04373433440923691, + "learning_rate": 1.4572754968808938e-06, + "loss": 0.0003, + "step": 25663 + }, + { + "epoch": 2.782307025151778, + "grad_norm": 1.9564241170883179, + "learning_rate": 1.4565501233135066e-06, + "loss": 0.0383, + "step": 25664 + }, + { + "epoch": 2.782415437987858, + "grad_norm": 0.17317500710487366, + "learning_rate": 1.4558247497461196e-06, + "loss": 0.0006, + "step": 25665 + }, + { + "epoch": 2.782523850823938, + "grad_norm": 4.1405863761901855, + "learning_rate": 1.4550993761787321e-06, + "loss": 0.0219, + "step": 25666 + }, + { + "epoch": 2.7826322636600174, + "grad_norm": 0.15488730370998383, + "learning_rate": 1.4543740026113451e-06, + "loss": 0.0013, + "step": 25667 + }, + { + "epoch": 2.782740676496097, + "grad_norm": 0.7244945168495178, + "learning_rate": 1.4536486290439577e-06, + "loss": 0.0011, + "step": 25668 + }, + { + "epoch": 2.782849089332177, + "grad_norm": 3.292914867401123, + "learning_rate": 1.4529232554765705e-06, + "loss": 0.0128, + "step": 25669 + }, + { + "epoch": 2.782957502168257, + "grad_norm": 0.03879079222679138, + "learning_rate": 1.4521978819091834e-06, + "loss": 0.0002, + "step": 25670 + }, + { + "epoch": 2.7830659150043364, + "grad_norm": 0.011042464524507523, + "learning_rate": 1.451472508341796e-06, + "loss": 0.0001, + "step": 25671 + }, + { + "epoch": 2.7831743278404164, + "grad_norm": 5.378154754638672, + "learning_rate": 1.450747134774409e-06, + "loss": 0.0097, + "step": 25672 + }, + { + "epoch": 2.783282740676496, + "grad_norm": 0.5243215560913086, + "learning_rate": 1.4500217612070216e-06, + "loss": 0.0016, + "step": 25673 + }, + { + "epoch": 2.783391153512576, + "grad_norm": 0.31354063749313354, + "learning_rate": 1.4492963876396346e-06, + "loss": 0.0007, + "step": 25674 + }, + { + "epoch": 2.783499566348656, + "grad_norm": 0.6589662432670593, + "learning_rate": 1.4485710140722473e-06, + "loss": 0.0037, + "step": 25675 + }, + { + "epoch": 2.7836079791847355, + "grad_norm": 0.4049249589443207, + "learning_rate": 1.4478456405048601e-06, + "loss": 0.0008, + "step": 25676 + }, + { + "epoch": 2.783716392020815, + "grad_norm": 1.6442515850067139, + "learning_rate": 1.447120266937473e-06, + "loss": 0.0018, + "step": 25677 + }, + { + "epoch": 2.783824804856895, + "grad_norm": 9.68319320678711, + "learning_rate": 1.4463948933700859e-06, + "loss": 0.0041, + "step": 25678 + }, + { + "epoch": 2.783933217692975, + "grad_norm": 1.7211003303527832, + "learning_rate": 1.4456695198026985e-06, + "loss": 0.023, + "step": 25679 + }, + { + "epoch": 2.7840416305290545, + "grad_norm": 2.6702871322631836, + "learning_rate": 1.4449441462353114e-06, + "loss": 0.0207, + "step": 25680 + }, + { + "epoch": 2.7841500433651345, + "grad_norm": 0.16589024662971497, + "learning_rate": 1.444218772667924e-06, + "loss": 0.0013, + "step": 25681 + }, + { + "epoch": 2.784258456201214, + "grad_norm": 2.304241895675659, + "learning_rate": 1.4434933991005368e-06, + "loss": 0.02, + "step": 25682 + }, + { + "epoch": 2.784366869037294, + "grad_norm": 0.04483109340071678, + "learning_rate": 1.4427680255331498e-06, + "loss": 0.0004, + "step": 25683 + }, + { + "epoch": 2.784475281873374, + "grad_norm": 2.519871950149536, + "learning_rate": 1.4420426519657623e-06, + "loss": 0.0068, + "step": 25684 + }, + { + "epoch": 2.7845836947094535, + "grad_norm": 1.0260697603225708, + "learning_rate": 1.4413172783983753e-06, + "loss": 0.0023, + "step": 25685 + }, + { + "epoch": 2.7846921075455335, + "grad_norm": 0.07846149057149887, + "learning_rate": 1.440591904830988e-06, + "loss": 0.0004, + "step": 25686 + }, + { + "epoch": 2.784800520381613, + "grad_norm": 0.4240860641002655, + "learning_rate": 1.4398665312636009e-06, + "loss": 0.0015, + "step": 25687 + }, + { + "epoch": 2.784908933217693, + "grad_norm": 0.33462586998939514, + "learning_rate": 1.4391411576962137e-06, + "loss": 0.0012, + "step": 25688 + }, + { + "epoch": 2.7850173460537726, + "grad_norm": 1.8734488487243652, + "learning_rate": 1.4384157841288265e-06, + "loss": 0.0216, + "step": 25689 + }, + { + "epoch": 2.7851257588898526, + "grad_norm": 1.5536268949508667, + "learning_rate": 1.4376904105614392e-06, + "loss": 0.0069, + "step": 25690 + }, + { + "epoch": 2.785234171725932, + "grad_norm": 2.8207578659057617, + "learning_rate": 1.4369650369940522e-06, + "loss": 0.0512, + "step": 25691 + }, + { + "epoch": 2.785342584562012, + "grad_norm": 0.03411433845758438, + "learning_rate": 1.4362396634266648e-06, + "loss": 0.0002, + "step": 25692 + }, + { + "epoch": 2.785450997398092, + "grad_norm": 0.03552717715501785, + "learning_rate": 1.4355142898592778e-06, + "loss": 0.0004, + "step": 25693 + }, + { + "epoch": 2.7855594102341716, + "grad_norm": 0.3437926769256592, + "learning_rate": 1.4347889162918903e-06, + "loss": 0.0008, + "step": 25694 + }, + { + "epoch": 2.7856678230702516, + "grad_norm": 0.10979323834180832, + "learning_rate": 1.4340635427245031e-06, + "loss": 0.0007, + "step": 25695 + }, + { + "epoch": 2.785776235906331, + "grad_norm": 5.135624885559082, + "learning_rate": 1.4333381691571161e-06, + "loss": 0.0045, + "step": 25696 + }, + { + "epoch": 2.785884648742411, + "grad_norm": 0.14055867493152618, + "learning_rate": 1.4326127955897287e-06, + "loss": 0.001, + "step": 25697 + }, + { + "epoch": 2.785993061578491, + "grad_norm": 0.03661951422691345, + "learning_rate": 1.4318874220223417e-06, + "loss": 0.0005, + "step": 25698 + }, + { + "epoch": 2.7861014744145707, + "grad_norm": 2.354647397994995, + "learning_rate": 1.4311620484549542e-06, + "loss": 0.0037, + "step": 25699 + }, + { + "epoch": 2.78620988725065, + "grad_norm": 4.869307041168213, + "learning_rate": 1.4304366748875672e-06, + "loss": 0.0251, + "step": 25700 + }, + { + "epoch": 2.78631830008673, + "grad_norm": 1.0308213233947754, + "learning_rate": 1.42971130132018e-06, + "loss": 0.0052, + "step": 25701 + }, + { + "epoch": 2.78642671292281, + "grad_norm": 0.15914270281791687, + "learning_rate": 1.4289859277527928e-06, + "loss": 0.0003, + "step": 25702 + }, + { + "epoch": 2.7865351257588897, + "grad_norm": 0.12009922415018082, + "learning_rate": 1.4282605541854056e-06, + "loss": 0.0005, + "step": 25703 + }, + { + "epoch": 2.7866435385949697, + "grad_norm": 0.02516266331076622, + "learning_rate": 1.4275351806180186e-06, + "loss": 0.0002, + "step": 25704 + }, + { + "epoch": 2.7867519514310493, + "grad_norm": 0.8247891068458557, + "learning_rate": 1.4268098070506311e-06, + "loss": 0.0014, + "step": 25705 + }, + { + "epoch": 2.7868603642671292, + "grad_norm": 0.12117139250040054, + "learning_rate": 1.4260844334832441e-06, + "loss": 0.0007, + "step": 25706 + }, + { + "epoch": 2.7869687771032092, + "grad_norm": 0.0344744510948658, + "learning_rate": 1.4253590599158567e-06, + "loss": 0.0002, + "step": 25707 + }, + { + "epoch": 2.7870771899392888, + "grad_norm": 0.6703178882598877, + "learning_rate": 1.4246336863484697e-06, + "loss": 0.0025, + "step": 25708 + }, + { + "epoch": 2.7871856027753688, + "grad_norm": 0.1383034884929657, + "learning_rate": 1.4239083127810824e-06, + "loss": 0.0008, + "step": 25709 + }, + { + "epoch": 2.7872940156114483, + "grad_norm": 0.05591337010264397, + "learning_rate": 1.423182939213695e-06, + "loss": 0.0005, + "step": 25710 + }, + { + "epoch": 2.7874024284475283, + "grad_norm": 0.07433341443538666, + "learning_rate": 1.422457565646308e-06, + "loss": 0.0006, + "step": 25711 + }, + { + "epoch": 2.787510841283608, + "grad_norm": 0.0271264910697937, + "learning_rate": 1.4217321920789206e-06, + "loss": 0.0002, + "step": 25712 + }, + { + "epoch": 2.787619254119688, + "grad_norm": 0.5410543084144592, + "learning_rate": 1.4210068185115336e-06, + "loss": 0.0021, + "step": 25713 + }, + { + "epoch": 2.7877276669557673, + "grad_norm": 6.659892559051514, + "learning_rate": 1.4202814449441463e-06, + "loss": 0.0091, + "step": 25714 + }, + { + "epoch": 2.7878360797918473, + "grad_norm": 0.5626164674758911, + "learning_rate": 1.4195560713767591e-06, + "loss": 0.0027, + "step": 25715 + }, + { + "epoch": 2.7879444926279273, + "grad_norm": 0.024492742493748665, + "learning_rate": 1.418830697809372e-06, + "loss": 0.0003, + "step": 25716 + }, + { + "epoch": 2.788052905464007, + "grad_norm": 1.3164570331573486, + "learning_rate": 1.4181053242419849e-06, + "loss": 0.0038, + "step": 25717 + }, + { + "epoch": 2.788161318300087, + "grad_norm": 0.014045409858226776, + "learning_rate": 1.4173799506745975e-06, + "loss": 0.0001, + "step": 25718 + }, + { + "epoch": 2.7882697311361664, + "grad_norm": 4.036916732788086, + "learning_rate": 1.4166545771072104e-06, + "loss": 0.0128, + "step": 25719 + }, + { + "epoch": 2.7883781439722464, + "grad_norm": 0.9161553978919983, + "learning_rate": 1.415929203539823e-06, + "loss": 0.0271, + "step": 25720 + }, + { + "epoch": 2.7884865568083264, + "grad_norm": 0.15212082862854004, + "learning_rate": 1.415203829972436e-06, + "loss": 0.0008, + "step": 25721 + }, + { + "epoch": 2.788594969644406, + "grad_norm": 0.10571721196174622, + "learning_rate": 1.4144784564050488e-06, + "loss": 0.0007, + "step": 25722 + }, + { + "epoch": 2.7887033824804854, + "grad_norm": 0.7256295680999756, + "learning_rate": 1.4137530828376613e-06, + "loss": 0.0015, + "step": 25723 + }, + { + "epoch": 2.7888117953165654, + "grad_norm": 0.10749535262584686, + "learning_rate": 1.4130277092702743e-06, + "loss": 0.0008, + "step": 25724 + }, + { + "epoch": 2.7889202081526454, + "grad_norm": 4.469177722930908, + "learning_rate": 1.412302335702887e-06, + "loss": 0.0301, + "step": 25725 + }, + { + "epoch": 2.789028620988725, + "grad_norm": 0.06860534846782684, + "learning_rate": 1.4115769621355e-06, + "loss": 0.0005, + "step": 25726 + }, + { + "epoch": 2.789137033824805, + "grad_norm": 0.13583919405937195, + "learning_rate": 1.4108515885681127e-06, + "loss": 0.0004, + "step": 25727 + }, + { + "epoch": 2.7892454466608845, + "grad_norm": 1.5171524286270142, + "learning_rate": 1.4101262150007255e-06, + "loss": 0.0073, + "step": 25728 + }, + { + "epoch": 2.7893538594969645, + "grad_norm": 0.11046597361564636, + "learning_rate": 1.4094008414333382e-06, + "loss": 0.0004, + "step": 25729 + }, + { + "epoch": 2.7894622723330444, + "grad_norm": 0.8302558064460754, + "learning_rate": 1.4086754678659512e-06, + "loss": 0.0026, + "step": 25730 + }, + { + "epoch": 2.789570685169124, + "grad_norm": 0.2673746645450592, + "learning_rate": 1.4079500942985638e-06, + "loss": 0.0009, + "step": 25731 + }, + { + "epoch": 2.789679098005204, + "grad_norm": 4.164857387542725, + "learning_rate": 1.4072247207311768e-06, + "loss": 0.0014, + "step": 25732 + }, + { + "epoch": 2.7897875108412835, + "grad_norm": 0.04235069826245308, + "learning_rate": 1.4064993471637893e-06, + "loss": 0.0003, + "step": 25733 + }, + { + "epoch": 2.7898959236773635, + "grad_norm": 0.14721918106079102, + "learning_rate": 1.4057739735964023e-06, + "loss": 0.0008, + "step": 25734 + }, + { + "epoch": 2.790004336513443, + "grad_norm": 0.057790402323007584, + "learning_rate": 1.4050486000290151e-06, + "loss": 0.0003, + "step": 25735 + }, + { + "epoch": 2.790112749349523, + "grad_norm": 0.02900470606982708, + "learning_rate": 1.4043232264616277e-06, + "loss": 0.0002, + "step": 25736 + }, + { + "epoch": 2.7902211621856026, + "grad_norm": 4.072452068328857, + "learning_rate": 1.4035978528942407e-06, + "loss": 0.0218, + "step": 25737 + }, + { + "epoch": 2.7903295750216826, + "grad_norm": 0.2167074829339981, + "learning_rate": 1.4028724793268532e-06, + "loss": 0.0009, + "step": 25738 + }, + { + "epoch": 2.7904379878577625, + "grad_norm": 3.142956256866455, + "learning_rate": 1.4021471057594662e-06, + "loss": 0.049, + "step": 25739 + }, + { + "epoch": 2.790546400693842, + "grad_norm": 2.1178178787231445, + "learning_rate": 1.401421732192079e-06, + "loss": 0.0195, + "step": 25740 + }, + { + "epoch": 2.790654813529922, + "grad_norm": 1.0299065113067627, + "learning_rate": 1.4006963586246918e-06, + "loss": 0.0031, + "step": 25741 + }, + { + "epoch": 2.7907632263660016, + "grad_norm": 4.91571569442749, + "learning_rate": 1.3999709850573046e-06, + "loss": 0.0305, + "step": 25742 + }, + { + "epoch": 2.7908716392020816, + "grad_norm": 0.08878809958696365, + "learning_rate": 1.3992456114899176e-06, + "loss": 0.0004, + "step": 25743 + }, + { + "epoch": 2.7909800520381616, + "grad_norm": 2.4212865829467773, + "learning_rate": 1.3985202379225301e-06, + "loss": 0.0278, + "step": 25744 + }, + { + "epoch": 2.791088464874241, + "grad_norm": 7.998253345489502, + "learning_rate": 1.3977948643551431e-06, + "loss": 0.0096, + "step": 25745 + }, + { + "epoch": 2.7911968777103207, + "grad_norm": 0.06631167978048325, + "learning_rate": 1.3970694907877557e-06, + "loss": 0.0006, + "step": 25746 + }, + { + "epoch": 2.7913052905464006, + "grad_norm": 0.6321985125541687, + "learning_rate": 1.3963441172203687e-06, + "loss": 0.0015, + "step": 25747 + }, + { + "epoch": 2.7914137033824806, + "grad_norm": 0.023809785023331642, + "learning_rate": 1.3956187436529815e-06, + "loss": 0.0001, + "step": 25748 + }, + { + "epoch": 2.79152211621856, + "grad_norm": 0.15873156487941742, + "learning_rate": 1.3948933700855942e-06, + "loss": 0.0004, + "step": 25749 + }, + { + "epoch": 2.79163052905464, + "grad_norm": 9.991229057312012, + "learning_rate": 1.394167996518207e-06, + "loss": 0.0278, + "step": 25750 + }, + { + "epoch": 2.7917389418907197, + "grad_norm": 1.551347017288208, + "learning_rate": 1.3934426229508196e-06, + "loss": 0.0237, + "step": 25751 + }, + { + "epoch": 2.7918473547267997, + "grad_norm": 2.6713693141937256, + "learning_rate": 1.3927172493834326e-06, + "loss": 0.0379, + "step": 25752 + }, + { + "epoch": 2.7919557675628797, + "grad_norm": 0.779398500919342, + "learning_rate": 1.3919918758160453e-06, + "loss": 0.0136, + "step": 25753 + }, + { + "epoch": 2.792064180398959, + "grad_norm": 0.38770392537117004, + "learning_rate": 1.3912665022486581e-06, + "loss": 0.0017, + "step": 25754 + }, + { + "epoch": 2.792172593235039, + "grad_norm": 2.734056234359741, + "learning_rate": 1.390541128681271e-06, + "loss": 0.0231, + "step": 25755 + }, + { + "epoch": 2.7922810060711187, + "grad_norm": 0.2982708215713501, + "learning_rate": 1.3898157551138839e-06, + "loss": 0.001, + "step": 25756 + }, + { + "epoch": 2.7923894189071987, + "grad_norm": 0.023916488513350487, + "learning_rate": 1.3890903815464965e-06, + "loss": 0.0002, + "step": 25757 + }, + { + "epoch": 2.7924978317432783, + "grad_norm": 1.2526326179504395, + "learning_rate": 1.3883650079791094e-06, + "loss": 0.0056, + "step": 25758 + }, + { + "epoch": 2.7926062445793582, + "grad_norm": 3.942491054534912, + "learning_rate": 1.387639634411722e-06, + "loss": 0.0043, + "step": 25759 + }, + { + "epoch": 2.792714657415438, + "grad_norm": 0.44790276885032654, + "learning_rate": 1.386914260844335e-06, + "loss": 0.0013, + "step": 25760 + }, + { + "epoch": 2.7928230702515178, + "grad_norm": 0.11237802356481552, + "learning_rate": 1.3861888872769478e-06, + "loss": 0.0007, + "step": 25761 + }, + { + "epoch": 2.7929314830875978, + "grad_norm": 0.061753302812576294, + "learning_rate": 1.3854635137095608e-06, + "loss": 0.0006, + "step": 25762 + }, + { + "epoch": 2.7930398959236773, + "grad_norm": 1.8881621360778809, + "learning_rate": 1.3847381401421733e-06, + "loss": 0.0364, + "step": 25763 + }, + { + "epoch": 2.7931483087597573, + "grad_norm": 0.15875773131847382, + "learning_rate": 1.384012766574786e-06, + "loss": 0.0011, + "step": 25764 + }, + { + "epoch": 2.793256721595837, + "grad_norm": 7.2390947341918945, + "learning_rate": 1.383287393007399e-06, + "loss": 0.0252, + "step": 25765 + }, + { + "epoch": 2.793365134431917, + "grad_norm": 0.06529966741800308, + "learning_rate": 1.3825620194400117e-06, + "loss": 0.0004, + "step": 25766 + }, + { + "epoch": 2.793473547267997, + "grad_norm": 1.0159087181091309, + "learning_rate": 1.3818366458726245e-06, + "loss": 0.0037, + "step": 25767 + }, + { + "epoch": 2.7935819601040763, + "grad_norm": 0.07102145999670029, + "learning_rate": 1.3811112723052372e-06, + "loss": 0.0003, + "step": 25768 + }, + { + "epoch": 2.793690372940156, + "grad_norm": 0.5278835296630859, + "learning_rate": 1.3803858987378502e-06, + "loss": 0.0013, + "step": 25769 + }, + { + "epoch": 2.793798785776236, + "grad_norm": 6.748746395111084, + "learning_rate": 1.3796605251704628e-06, + "loss": 0.0345, + "step": 25770 + }, + { + "epoch": 2.793907198612316, + "grad_norm": 7.687887668609619, + "learning_rate": 1.3789351516030758e-06, + "loss": 0.0198, + "step": 25771 + }, + { + "epoch": 2.7940156114483954, + "grad_norm": 4.966721534729004, + "learning_rate": 1.3782097780356883e-06, + "loss": 0.0423, + "step": 25772 + }, + { + "epoch": 2.7941240242844754, + "grad_norm": 2.1070380210876465, + "learning_rate": 1.3774844044683013e-06, + "loss": 0.0122, + "step": 25773 + }, + { + "epoch": 2.794232437120555, + "grad_norm": 0.0777822732925415, + "learning_rate": 1.3767590309009141e-06, + "loss": 0.0005, + "step": 25774 + }, + { + "epoch": 2.794340849956635, + "grad_norm": 0.10581532120704651, + "learning_rate": 1.3760336573335271e-06, + "loss": 0.0002, + "step": 25775 + }, + { + "epoch": 2.794449262792715, + "grad_norm": 1.0290266275405884, + "learning_rate": 1.3753082837661397e-06, + "loss": 0.0029, + "step": 25776 + }, + { + "epoch": 2.7945576756287944, + "grad_norm": 1.941733956336975, + "learning_rate": 1.3745829101987522e-06, + "loss": 0.0168, + "step": 25777 + }, + { + "epoch": 2.794666088464874, + "grad_norm": 0.036072421818971634, + "learning_rate": 1.3738575366313652e-06, + "loss": 0.0004, + "step": 25778 + }, + { + "epoch": 2.794774501300954, + "grad_norm": 0.7304647564888, + "learning_rate": 1.373132163063978e-06, + "loss": 0.0044, + "step": 25779 + }, + { + "epoch": 2.794882914137034, + "grad_norm": 0.8812364935874939, + "learning_rate": 1.3724067894965908e-06, + "loss": 0.004, + "step": 25780 + }, + { + "epoch": 2.7949913269731135, + "grad_norm": 3.3829522132873535, + "learning_rate": 1.3716814159292036e-06, + "loss": 0.0044, + "step": 25781 + }, + { + "epoch": 2.7950997398091935, + "grad_norm": 10.304856300354004, + "learning_rate": 1.3709560423618166e-06, + "loss": 0.0109, + "step": 25782 + }, + { + "epoch": 2.795208152645273, + "grad_norm": 1.9086350202560425, + "learning_rate": 1.3702306687944291e-06, + "loss": 0.0032, + "step": 25783 + }, + { + "epoch": 2.795316565481353, + "grad_norm": 18.006938934326172, + "learning_rate": 1.3695052952270421e-06, + "loss": 0.0358, + "step": 25784 + }, + { + "epoch": 2.795424978317433, + "grad_norm": 4.866331100463867, + "learning_rate": 1.3687799216596547e-06, + "loss": 0.0381, + "step": 25785 + }, + { + "epoch": 2.7955333911535125, + "grad_norm": 0.04520338773727417, + "learning_rate": 1.3680545480922677e-06, + "loss": 0.0004, + "step": 25786 + }, + { + "epoch": 2.7956418039895925, + "grad_norm": 0.023120442405343056, + "learning_rate": 1.3673291745248805e-06, + "loss": 0.0001, + "step": 25787 + }, + { + "epoch": 2.795750216825672, + "grad_norm": 2.6118035316467285, + "learning_rate": 1.3666038009574934e-06, + "loss": 0.0072, + "step": 25788 + }, + { + "epoch": 2.795858629661752, + "grad_norm": 0.6820158958435059, + "learning_rate": 1.365878427390106e-06, + "loss": 0.0019, + "step": 25789 + }, + { + "epoch": 2.795967042497832, + "grad_norm": 0.08466995507478714, + "learning_rate": 1.3651530538227186e-06, + "loss": 0.0003, + "step": 25790 + }, + { + "epoch": 2.7960754553339116, + "grad_norm": 3.5876100063323975, + "learning_rate": 1.3644276802553316e-06, + "loss": 0.0058, + "step": 25791 + }, + { + "epoch": 2.796183868169991, + "grad_norm": 3.4333581924438477, + "learning_rate": 1.3637023066879443e-06, + "loss": 0.0153, + "step": 25792 + }, + { + "epoch": 2.796292281006071, + "grad_norm": 0.11066628992557526, + "learning_rate": 1.3629769331205573e-06, + "loss": 0.0012, + "step": 25793 + }, + { + "epoch": 2.796400693842151, + "grad_norm": 0.07035814970731735, + "learning_rate": 1.36225155955317e-06, + "loss": 0.0006, + "step": 25794 + }, + { + "epoch": 2.7965091066782306, + "grad_norm": 0.0799659788608551, + "learning_rate": 1.361526185985783e-06, + "loss": 0.0006, + "step": 25795 + }, + { + "epoch": 2.7966175195143106, + "grad_norm": 1.2672221660614014, + "learning_rate": 1.3608008124183955e-06, + "loss": 0.004, + "step": 25796 + }, + { + "epoch": 2.79672593235039, + "grad_norm": 0.052661649882793427, + "learning_rate": 1.3600754388510085e-06, + "loss": 0.0005, + "step": 25797 + }, + { + "epoch": 2.79683434518647, + "grad_norm": 5.746230602264404, + "learning_rate": 1.359350065283621e-06, + "loss": 0.0634, + "step": 25798 + }, + { + "epoch": 2.79694275802255, + "grad_norm": 0.8690258264541626, + "learning_rate": 1.358624691716234e-06, + "loss": 0.0034, + "step": 25799 + }, + { + "epoch": 2.7970511708586296, + "grad_norm": 0.0650436133146286, + "learning_rate": 1.3578993181488468e-06, + "loss": 0.0004, + "step": 25800 + }, + { + "epoch": 2.797159583694709, + "grad_norm": 2.4840428829193115, + "learning_rate": 1.3571739445814598e-06, + "loss": 0.0422, + "step": 25801 + }, + { + "epoch": 2.797267996530789, + "grad_norm": 0.03406301885843277, + "learning_rate": 1.3564485710140723e-06, + "loss": 0.0002, + "step": 25802 + }, + { + "epoch": 2.797376409366869, + "grad_norm": 0.305980384349823, + "learning_rate": 1.3557231974466853e-06, + "loss": 0.0013, + "step": 25803 + }, + { + "epoch": 2.7974848222029487, + "grad_norm": 0.019485019147396088, + "learning_rate": 1.354997823879298e-06, + "loss": 0.0003, + "step": 25804 + }, + { + "epoch": 2.7975932350390287, + "grad_norm": 0.8413969278335571, + "learning_rate": 1.3542724503119107e-06, + "loss": 0.0434, + "step": 25805 + }, + { + "epoch": 2.797701647875108, + "grad_norm": 2.1749815940856934, + "learning_rate": 1.3535470767445237e-06, + "loss": 0.0051, + "step": 25806 + }, + { + "epoch": 2.797810060711188, + "grad_norm": 5.701659202575684, + "learning_rate": 1.3528217031771362e-06, + "loss": 0.0138, + "step": 25807 + }, + { + "epoch": 2.797918473547268, + "grad_norm": 0.18489940464496613, + "learning_rate": 1.3520963296097492e-06, + "loss": 0.0018, + "step": 25808 + }, + { + "epoch": 2.7980268863833477, + "grad_norm": 0.03314264863729477, + "learning_rate": 1.3513709560423618e-06, + "loss": 0.0003, + "step": 25809 + }, + { + "epoch": 2.7981352992194277, + "grad_norm": 0.0547795444726944, + "learning_rate": 1.3506455824749748e-06, + "loss": 0.0004, + "step": 25810 + }, + { + "epoch": 2.7982437120555073, + "grad_norm": 0.5888808965682983, + "learning_rate": 1.3499202089075876e-06, + "loss": 0.0025, + "step": 25811 + }, + { + "epoch": 2.7983521248915872, + "grad_norm": 0.0877666249871254, + "learning_rate": 1.3491948353402003e-06, + "loss": 0.0006, + "step": 25812 + }, + { + "epoch": 2.7984605377276672, + "grad_norm": 0.0259051863104105, + "learning_rate": 1.3484694617728131e-06, + "loss": 0.0003, + "step": 25813 + }, + { + "epoch": 2.7985689505637468, + "grad_norm": 4.371073246002197, + "learning_rate": 1.3477440882054261e-06, + "loss": 0.0071, + "step": 25814 + }, + { + "epoch": 2.7986773633998263, + "grad_norm": 1.7683360576629639, + "learning_rate": 1.3470187146380387e-06, + "loss": 0.0202, + "step": 25815 + }, + { + "epoch": 2.7987857762359063, + "grad_norm": 3.866987705230713, + "learning_rate": 1.3462933410706517e-06, + "loss": 0.0302, + "step": 25816 + }, + { + "epoch": 2.7988941890719863, + "grad_norm": 0.016386963427066803, + "learning_rate": 1.3455679675032642e-06, + "loss": 0.0001, + "step": 25817 + }, + { + "epoch": 2.799002601908066, + "grad_norm": 2.2184512615203857, + "learning_rate": 1.344842593935877e-06, + "loss": 0.0216, + "step": 25818 + }, + { + "epoch": 2.799111014744146, + "grad_norm": 3.2848589420318604, + "learning_rate": 1.34411722036849e-06, + "loss": 0.0264, + "step": 25819 + }, + { + "epoch": 2.7992194275802254, + "grad_norm": 1.7875608205795288, + "learning_rate": 1.3433918468011026e-06, + "loss": 0.0038, + "step": 25820 + }, + { + "epoch": 2.7993278404163053, + "grad_norm": 0.15378037095069885, + "learning_rate": 1.3426664732337156e-06, + "loss": 0.0006, + "step": 25821 + }, + { + "epoch": 2.7994362532523853, + "grad_norm": 1.6846376657485962, + "learning_rate": 1.3419410996663281e-06, + "loss": 0.0429, + "step": 25822 + }, + { + "epoch": 2.799544666088465, + "grad_norm": 3.2040154933929443, + "learning_rate": 1.3412157260989411e-06, + "loss": 0.0175, + "step": 25823 + }, + { + "epoch": 2.7996530789245444, + "grad_norm": 5.881544589996338, + "learning_rate": 1.340490352531554e-06, + "loss": 0.0203, + "step": 25824 + }, + { + "epoch": 2.7997614917606244, + "grad_norm": 0.06970010697841644, + "learning_rate": 1.3397649789641667e-06, + "loss": 0.0006, + "step": 25825 + }, + { + "epoch": 2.7998699045967044, + "grad_norm": 2.7385029792785645, + "learning_rate": 1.3390396053967795e-06, + "loss": 0.0115, + "step": 25826 + }, + { + "epoch": 2.799978317432784, + "grad_norm": 1.1513532400131226, + "learning_rate": 1.3383142318293924e-06, + "loss": 0.0416, + "step": 25827 + }, + { + "epoch": 2.800086730268864, + "grad_norm": 0.040736112743616104, + "learning_rate": 1.337588858262005e-06, + "loss": 0.0003, + "step": 25828 + }, + { + "epoch": 2.8001951431049434, + "grad_norm": 0.02090640179812908, + "learning_rate": 1.336863484694618e-06, + "loss": 0.0001, + "step": 25829 + }, + { + "epoch": 2.8003035559410234, + "grad_norm": 0.09319453686475754, + "learning_rate": 1.3361381111272306e-06, + "loss": 0.0006, + "step": 25830 + }, + { + "epoch": 2.8004119687771034, + "grad_norm": 0.1193733960390091, + "learning_rate": 1.3354127375598433e-06, + "loss": 0.0008, + "step": 25831 + }, + { + "epoch": 2.800520381613183, + "grad_norm": 9.545111656188965, + "learning_rate": 1.3346873639924563e-06, + "loss": 0.0505, + "step": 25832 + }, + { + "epoch": 2.800628794449263, + "grad_norm": 0.015414855442941189, + "learning_rate": 1.333961990425069e-06, + "loss": 0.0002, + "step": 25833 + }, + { + "epoch": 2.8007372072853425, + "grad_norm": 0.05138927698135376, + "learning_rate": 1.333236616857682e-06, + "loss": 0.0004, + "step": 25834 + }, + { + "epoch": 2.8008456201214225, + "grad_norm": 0.3327862322330475, + "learning_rate": 1.3325112432902945e-06, + "loss": 0.0016, + "step": 25835 + }, + { + "epoch": 2.800954032957502, + "grad_norm": 0.1067948266863823, + "learning_rate": 1.3317858697229075e-06, + "loss": 0.0003, + "step": 25836 + }, + { + "epoch": 2.801062445793582, + "grad_norm": 0.9184653162956238, + "learning_rate": 1.3310604961555202e-06, + "loss": 0.0026, + "step": 25837 + }, + { + "epoch": 2.8011708586296615, + "grad_norm": 5.257067680358887, + "learning_rate": 1.330335122588133e-06, + "loss": 0.0054, + "step": 25838 + }, + { + "epoch": 2.8012792714657415, + "grad_norm": 0.1607881784439087, + "learning_rate": 1.3296097490207458e-06, + "loss": 0.0008, + "step": 25839 + }, + { + "epoch": 2.8013876843018215, + "grad_norm": 0.2671785056591034, + "learning_rate": 1.3288843754533588e-06, + "loss": 0.0013, + "step": 25840 + }, + { + "epoch": 2.801496097137901, + "grad_norm": 0.25344234704971313, + "learning_rate": 1.3281590018859713e-06, + "loss": 0.0006, + "step": 25841 + }, + { + "epoch": 2.801604509973981, + "grad_norm": 1.3068633079528809, + "learning_rate": 1.3274336283185843e-06, + "loss": 0.0037, + "step": 25842 + }, + { + "epoch": 2.8017129228100606, + "grad_norm": 0.03538435325026512, + "learning_rate": 1.326708254751197e-06, + "loss": 0.0003, + "step": 25843 + }, + { + "epoch": 2.8018213356461406, + "grad_norm": 4.55037784576416, + "learning_rate": 1.3259828811838097e-06, + "loss": 0.0207, + "step": 25844 + }, + { + "epoch": 2.8019297484822205, + "grad_norm": 3.606003999710083, + "learning_rate": 1.3252575076164227e-06, + "loss": 0.0082, + "step": 25845 + }, + { + "epoch": 2.8020381613183, + "grad_norm": 2.184396505355835, + "learning_rate": 1.3245321340490352e-06, + "loss": 0.0051, + "step": 25846 + }, + { + "epoch": 2.8021465741543796, + "grad_norm": 0.15689803659915924, + "learning_rate": 1.3238067604816482e-06, + "loss": 0.0009, + "step": 25847 + }, + { + "epoch": 2.8022549869904596, + "grad_norm": 1.555847406387329, + "learning_rate": 1.3230813869142608e-06, + "loss": 0.0017, + "step": 25848 + }, + { + "epoch": 2.8023633998265396, + "grad_norm": 1.140657663345337, + "learning_rate": 1.3223560133468738e-06, + "loss": 0.0011, + "step": 25849 + }, + { + "epoch": 2.802471812662619, + "grad_norm": 0.08514352142810822, + "learning_rate": 1.3216306397794866e-06, + "loss": 0.0008, + "step": 25850 + }, + { + "epoch": 2.802580225498699, + "grad_norm": 0.015607262961566448, + "learning_rate": 1.3209052662120993e-06, + "loss": 0.0002, + "step": 25851 + }, + { + "epoch": 2.8026886383347787, + "grad_norm": 0.019702782854437828, + "learning_rate": 1.3201798926447121e-06, + "loss": 0.0002, + "step": 25852 + }, + { + "epoch": 2.8027970511708586, + "grad_norm": 2.955967426300049, + "learning_rate": 1.3194545190773251e-06, + "loss": 0.0104, + "step": 25853 + }, + { + "epoch": 2.8029054640069386, + "grad_norm": 2.7565438747406006, + "learning_rate": 1.3187291455099377e-06, + "loss": 0.0242, + "step": 25854 + }, + { + "epoch": 2.803013876843018, + "grad_norm": 1.120044231414795, + "learning_rate": 1.3180037719425507e-06, + "loss": 0.0039, + "step": 25855 + }, + { + "epoch": 2.803122289679098, + "grad_norm": 4.748197078704834, + "learning_rate": 1.3172783983751632e-06, + "loss": 0.0048, + "step": 25856 + }, + { + "epoch": 2.8032307025151777, + "grad_norm": 0.2987425923347473, + "learning_rate": 1.3165530248077762e-06, + "loss": 0.0015, + "step": 25857 + }, + { + "epoch": 2.8033391153512577, + "grad_norm": 0.13008445501327515, + "learning_rate": 1.315827651240389e-06, + "loss": 0.0006, + "step": 25858 + }, + { + "epoch": 2.8034475281873372, + "grad_norm": 0.045703887939453125, + "learning_rate": 1.3151022776730016e-06, + "loss": 0.0002, + "step": 25859 + }, + { + "epoch": 2.803555941023417, + "grad_norm": 0.26216500997543335, + "learning_rate": 1.3143769041056146e-06, + "loss": 0.0013, + "step": 25860 + }, + { + "epoch": 2.8036643538594967, + "grad_norm": 0.09021522849798203, + "learning_rate": 1.3136515305382271e-06, + "loss": 0.0007, + "step": 25861 + }, + { + "epoch": 2.8037727666955767, + "grad_norm": 4.637558937072754, + "learning_rate": 1.3129261569708401e-06, + "loss": 0.0082, + "step": 25862 + }, + { + "epoch": 2.8038811795316567, + "grad_norm": 3.800583839416504, + "learning_rate": 1.312200783403453e-06, + "loss": 0.0499, + "step": 25863 + }, + { + "epoch": 2.8039895923677363, + "grad_norm": 9.820136070251465, + "learning_rate": 1.3114754098360657e-06, + "loss": 0.0252, + "step": 25864 + }, + { + "epoch": 2.8040980052038162, + "grad_norm": 0.4379783868789673, + "learning_rate": 1.3107500362686785e-06, + "loss": 0.0006, + "step": 25865 + }, + { + "epoch": 2.804206418039896, + "grad_norm": 0.21889734268188477, + "learning_rate": 1.3100246627012915e-06, + "loss": 0.001, + "step": 25866 + }, + { + "epoch": 2.8043148308759758, + "grad_norm": 0.03785479813814163, + "learning_rate": 1.309299289133904e-06, + "loss": 0.0003, + "step": 25867 + }, + { + "epoch": 2.8044232437120558, + "grad_norm": 0.9831738471984863, + "learning_rate": 1.308573915566517e-06, + "loss": 0.0048, + "step": 25868 + }, + { + "epoch": 2.8045316565481353, + "grad_norm": 0.0417504757642746, + "learning_rate": 1.3078485419991296e-06, + "loss": 0.0004, + "step": 25869 + }, + { + "epoch": 2.804640069384215, + "grad_norm": 0.16175411641597748, + "learning_rate": 1.3071231684317426e-06, + "loss": 0.0005, + "step": 25870 + }, + { + "epoch": 2.804748482220295, + "grad_norm": 0.0644327700138092, + "learning_rate": 1.3063977948643553e-06, + "loss": 0.0004, + "step": 25871 + }, + { + "epoch": 2.804856895056375, + "grad_norm": 0.016589848324656487, + "learning_rate": 1.305672421296968e-06, + "loss": 0.0002, + "step": 25872 + }, + { + "epoch": 2.8049653078924544, + "grad_norm": 9.995378494262695, + "learning_rate": 1.304947047729581e-06, + "loss": 0.0213, + "step": 25873 + }, + { + "epoch": 2.8050737207285343, + "grad_norm": 0.22132983803749084, + "learning_rate": 1.3042216741621935e-06, + "loss": 0.0007, + "step": 25874 + }, + { + "epoch": 2.805182133564614, + "grad_norm": 2.2430531978607178, + "learning_rate": 1.3034963005948065e-06, + "loss": 0.0283, + "step": 25875 + }, + { + "epoch": 2.805290546400694, + "grad_norm": 0.1397007256746292, + "learning_rate": 1.3027709270274192e-06, + "loss": 0.0008, + "step": 25876 + }, + { + "epoch": 2.805398959236774, + "grad_norm": 0.048023611307144165, + "learning_rate": 1.302045553460032e-06, + "loss": 0.0002, + "step": 25877 + }, + { + "epoch": 2.8055073720728534, + "grad_norm": 0.08836964517831802, + "learning_rate": 1.3013201798926448e-06, + "loss": 0.0003, + "step": 25878 + }, + { + "epoch": 2.8056157849089334, + "grad_norm": 6.050665855407715, + "learning_rate": 1.3005948063252578e-06, + "loss": 0.0275, + "step": 25879 + }, + { + "epoch": 2.805724197745013, + "grad_norm": 1.9251490831375122, + "learning_rate": 1.2998694327578704e-06, + "loss": 0.0167, + "step": 25880 + }, + { + "epoch": 2.805832610581093, + "grad_norm": 1.9257822036743164, + "learning_rate": 1.2991440591904833e-06, + "loss": 0.0046, + "step": 25881 + }, + { + "epoch": 2.8059410234171724, + "grad_norm": 0.06649111956357956, + "learning_rate": 1.298418685623096e-06, + "loss": 0.0004, + "step": 25882 + }, + { + "epoch": 2.8060494362532524, + "grad_norm": 0.1289932280778885, + "learning_rate": 1.297693312055709e-06, + "loss": 0.0008, + "step": 25883 + }, + { + "epoch": 2.806157849089332, + "grad_norm": 0.022154979407787323, + "learning_rate": 1.2969679384883217e-06, + "loss": 0.0002, + "step": 25884 + }, + { + "epoch": 2.806266261925412, + "grad_norm": 0.016162529587745667, + "learning_rate": 1.2962425649209342e-06, + "loss": 0.0002, + "step": 25885 + }, + { + "epoch": 2.806374674761492, + "grad_norm": 2.965200662612915, + "learning_rate": 1.2955171913535472e-06, + "loss": 0.0155, + "step": 25886 + }, + { + "epoch": 2.8064830875975715, + "grad_norm": 0.10272977501153946, + "learning_rate": 1.2947918177861598e-06, + "loss": 0.0004, + "step": 25887 + }, + { + "epoch": 2.8065915004336515, + "grad_norm": 5.91910982131958, + "learning_rate": 1.2940664442187728e-06, + "loss": 0.0195, + "step": 25888 + }, + { + "epoch": 2.806699913269731, + "grad_norm": 0.16772396862506866, + "learning_rate": 1.2933410706513856e-06, + "loss": 0.0011, + "step": 25889 + }, + { + "epoch": 2.806808326105811, + "grad_norm": 0.1321963220834732, + "learning_rate": 1.2926156970839983e-06, + "loss": 0.0007, + "step": 25890 + }, + { + "epoch": 2.806916738941891, + "grad_norm": 3.8785560131073, + "learning_rate": 1.2918903235166111e-06, + "loss": 0.0072, + "step": 25891 + }, + { + "epoch": 2.8070251517779705, + "grad_norm": 5.914047718048096, + "learning_rate": 1.2911649499492241e-06, + "loss": 0.0329, + "step": 25892 + }, + { + "epoch": 2.80713356461405, + "grad_norm": 0.09116619825363159, + "learning_rate": 1.2904395763818367e-06, + "loss": 0.0005, + "step": 25893 + }, + { + "epoch": 2.80724197745013, + "grad_norm": 1.4851828813552856, + "learning_rate": 1.2897142028144497e-06, + "loss": 0.0198, + "step": 25894 + }, + { + "epoch": 2.80735039028621, + "grad_norm": 0.28755027055740356, + "learning_rate": 1.2889888292470622e-06, + "loss": 0.002, + "step": 25895 + }, + { + "epoch": 2.8074588031222896, + "grad_norm": 2.0615248680114746, + "learning_rate": 1.2882634556796752e-06, + "loss": 0.0436, + "step": 25896 + }, + { + "epoch": 2.8075672159583696, + "grad_norm": 0.614291787147522, + "learning_rate": 1.287538082112288e-06, + "loss": 0.0012, + "step": 25897 + }, + { + "epoch": 2.807675628794449, + "grad_norm": 4.872426986694336, + "learning_rate": 1.2868127085449006e-06, + "loss": 0.0233, + "step": 25898 + }, + { + "epoch": 2.807784041630529, + "grad_norm": 0.5549275875091553, + "learning_rate": 1.2860873349775136e-06, + "loss": 0.0012, + "step": 25899 + }, + { + "epoch": 2.807892454466609, + "grad_norm": 0.11595601588487625, + "learning_rate": 1.2853619614101261e-06, + "loss": 0.0007, + "step": 25900 + }, + { + "epoch": 2.8080008673026886, + "grad_norm": 2.343005895614624, + "learning_rate": 1.2846365878427391e-06, + "loss": 0.0062, + "step": 25901 + }, + { + "epoch": 2.8081092801387686, + "grad_norm": 0.3010680675506592, + "learning_rate": 1.283911214275352e-06, + "loss": 0.0016, + "step": 25902 + }, + { + "epoch": 2.808217692974848, + "grad_norm": 0.17877686023712158, + "learning_rate": 1.2831858407079647e-06, + "loss": 0.0007, + "step": 25903 + }, + { + "epoch": 2.808326105810928, + "grad_norm": 0.06378515064716339, + "learning_rate": 1.2824604671405775e-06, + "loss": 0.0003, + "step": 25904 + }, + { + "epoch": 2.8084345186470077, + "grad_norm": 0.21989765763282776, + "learning_rate": 1.2817350935731905e-06, + "loss": 0.0008, + "step": 25905 + }, + { + "epoch": 2.8085429314830876, + "grad_norm": 0.3119744062423706, + "learning_rate": 1.281009720005803e-06, + "loss": 0.0013, + "step": 25906 + }, + { + "epoch": 2.808651344319167, + "grad_norm": 0.4534643590450287, + "learning_rate": 1.280284346438416e-06, + "loss": 0.0014, + "step": 25907 + }, + { + "epoch": 2.808759757155247, + "grad_norm": 0.647839367389679, + "learning_rate": 1.2795589728710286e-06, + "loss": 0.002, + "step": 25908 + }, + { + "epoch": 2.808868169991327, + "grad_norm": 0.04749377816915512, + "learning_rate": 1.2788335993036416e-06, + "loss": 0.0004, + "step": 25909 + }, + { + "epoch": 2.8089765828274067, + "grad_norm": 1.768282175064087, + "learning_rate": 1.2781082257362543e-06, + "loss": 0.023, + "step": 25910 + }, + { + "epoch": 2.8090849956634867, + "grad_norm": 0.0221868809312582, + "learning_rate": 1.2773828521688671e-06, + "loss": 0.0002, + "step": 25911 + }, + { + "epoch": 2.8091934084995662, + "grad_norm": 1.7451691627502441, + "learning_rate": 1.27665747860148e-06, + "loss": 0.0331, + "step": 25912 + }, + { + "epoch": 2.809301821335646, + "grad_norm": 0.07099729031324387, + "learning_rate": 1.2759321050340925e-06, + "loss": 0.0002, + "step": 25913 + }, + { + "epoch": 2.809410234171726, + "grad_norm": 0.16181977093219757, + "learning_rate": 1.2752067314667055e-06, + "loss": 0.001, + "step": 25914 + }, + { + "epoch": 2.8095186470078057, + "grad_norm": 1.049548864364624, + "learning_rate": 1.2744813578993182e-06, + "loss": 0.0022, + "step": 25915 + }, + { + "epoch": 2.8096270598438853, + "grad_norm": 0.4482952654361725, + "learning_rate": 1.273755984331931e-06, + "loss": 0.0019, + "step": 25916 + }, + { + "epoch": 2.8097354726799653, + "grad_norm": 1.5199344158172607, + "learning_rate": 1.2730306107645438e-06, + "loss": 0.0341, + "step": 25917 + }, + { + "epoch": 2.8098438855160452, + "grad_norm": 1.0759252309799194, + "learning_rate": 1.2723052371971568e-06, + "loss": 0.0029, + "step": 25918 + }, + { + "epoch": 2.809952298352125, + "grad_norm": 0.020372610539197922, + "learning_rate": 1.2715798636297694e-06, + "loss": 0.0003, + "step": 25919 + }, + { + "epoch": 2.8100607111882048, + "grad_norm": 2.4121885299682617, + "learning_rate": 1.2708544900623823e-06, + "loss": 0.0322, + "step": 25920 + }, + { + "epoch": 2.8101691240242843, + "grad_norm": 0.17540331184864044, + "learning_rate": 1.270129116494995e-06, + "loss": 0.0011, + "step": 25921 + }, + { + "epoch": 2.8102775368603643, + "grad_norm": 0.13453221321105957, + "learning_rate": 1.269403742927608e-06, + "loss": 0.001, + "step": 25922 + }, + { + "epoch": 2.8103859496964443, + "grad_norm": 0.09113068878650665, + "learning_rate": 1.2686783693602207e-06, + "loss": 0.0007, + "step": 25923 + }, + { + "epoch": 2.810494362532524, + "grad_norm": 0.9302343726158142, + "learning_rate": 1.2679529957928335e-06, + "loss": 0.0014, + "step": 25924 + }, + { + "epoch": 2.8106027753686034, + "grad_norm": 0.015345037914812565, + "learning_rate": 1.2672276222254462e-06, + "loss": 0.0001, + "step": 25925 + }, + { + "epoch": 2.8107111882046834, + "grad_norm": 3.0341079235076904, + "learning_rate": 1.2665022486580588e-06, + "loss": 0.0161, + "step": 25926 + }, + { + "epoch": 2.8108196010407633, + "grad_norm": 0.04283234477043152, + "learning_rate": 1.2657768750906718e-06, + "loss": 0.0003, + "step": 25927 + }, + { + "epoch": 2.810928013876843, + "grad_norm": 0.2422824501991272, + "learning_rate": 1.2650515015232846e-06, + "loss": 0.0011, + "step": 25928 + }, + { + "epoch": 2.811036426712923, + "grad_norm": 5.090002059936523, + "learning_rate": 1.2643261279558974e-06, + "loss": 0.0186, + "step": 25929 + }, + { + "epoch": 2.8111448395490024, + "grad_norm": 0.048259906470775604, + "learning_rate": 1.2636007543885101e-06, + "loss": 0.0002, + "step": 25930 + }, + { + "epoch": 2.8112532523850824, + "grad_norm": 0.12028884887695312, + "learning_rate": 1.2628753808211231e-06, + "loss": 0.001, + "step": 25931 + }, + { + "epoch": 2.8113616652211624, + "grad_norm": 0.07306738942861557, + "learning_rate": 1.2621500072537357e-06, + "loss": 0.0004, + "step": 25932 + }, + { + "epoch": 2.811470078057242, + "grad_norm": 0.9402682185173035, + "learning_rate": 1.2614246336863487e-06, + "loss": 0.003, + "step": 25933 + }, + { + "epoch": 2.811578490893322, + "grad_norm": 2.0656628608703613, + "learning_rate": 1.2606992601189612e-06, + "loss": 0.0027, + "step": 25934 + }, + { + "epoch": 2.8116869037294014, + "grad_norm": 0.03847223520278931, + "learning_rate": 1.2599738865515742e-06, + "loss": 0.0003, + "step": 25935 + }, + { + "epoch": 2.8117953165654814, + "grad_norm": 10.58192253112793, + "learning_rate": 1.259248512984187e-06, + "loss": 0.0066, + "step": 25936 + }, + { + "epoch": 2.8119037294015614, + "grad_norm": 3.4817774295806885, + "learning_rate": 1.2585231394167998e-06, + "loss": 0.0245, + "step": 25937 + }, + { + "epoch": 2.812012142237641, + "grad_norm": 2.3687667846679688, + "learning_rate": 1.2577977658494126e-06, + "loss": 0.0013, + "step": 25938 + }, + { + "epoch": 2.8121205550737205, + "grad_norm": 2.2195420265197754, + "learning_rate": 1.2570723922820251e-06, + "loss": 0.008, + "step": 25939 + }, + { + "epoch": 2.8122289679098005, + "grad_norm": 1.3471739292144775, + "learning_rate": 1.2563470187146381e-06, + "loss": 0.0027, + "step": 25940 + }, + { + "epoch": 2.8123373807458805, + "grad_norm": 1.5715770721435547, + "learning_rate": 1.255621645147251e-06, + "loss": 0.0304, + "step": 25941 + }, + { + "epoch": 2.81244579358196, + "grad_norm": 0.8403339982032776, + "learning_rate": 1.2548962715798637e-06, + "loss": 0.0015, + "step": 25942 + }, + { + "epoch": 2.81255420641804, + "grad_norm": 3.0155434608459473, + "learning_rate": 1.2541708980124765e-06, + "loss": 0.0393, + "step": 25943 + }, + { + "epoch": 2.8126626192541195, + "grad_norm": 3.087664842605591, + "learning_rate": 1.2534455244450895e-06, + "loss": 0.0051, + "step": 25944 + }, + { + "epoch": 2.8127710320901995, + "grad_norm": 0.1178269237279892, + "learning_rate": 1.252720150877702e-06, + "loss": 0.0006, + "step": 25945 + }, + { + "epoch": 2.8128794449262795, + "grad_norm": 3.545858383178711, + "learning_rate": 1.251994777310315e-06, + "loss": 0.015, + "step": 25946 + }, + { + "epoch": 2.812987857762359, + "grad_norm": 0.08911146223545074, + "learning_rate": 1.2512694037429276e-06, + "loss": 0.0004, + "step": 25947 + }, + { + "epoch": 2.8130962705984386, + "grad_norm": 7.6080546379089355, + "learning_rate": 1.2505440301755406e-06, + "loss": 0.0399, + "step": 25948 + }, + { + "epoch": 2.8132046834345186, + "grad_norm": 2.930363893508911, + "learning_rate": 1.2498186566081533e-06, + "loss": 0.0061, + "step": 25949 + }, + { + "epoch": 2.8133130962705986, + "grad_norm": 5.153585433959961, + "learning_rate": 1.2490932830407661e-06, + "loss": 0.0156, + "step": 25950 + }, + { + "epoch": 2.813421509106678, + "grad_norm": 0.0361747108399868, + "learning_rate": 1.248367909473379e-06, + "loss": 0.0003, + "step": 25951 + }, + { + "epoch": 2.813529921942758, + "grad_norm": 0.14263081550598145, + "learning_rate": 1.2476425359059917e-06, + "loss": 0.0006, + "step": 25952 + }, + { + "epoch": 2.8136383347788376, + "grad_norm": 0.1706509292125702, + "learning_rate": 1.2469171623386045e-06, + "loss": 0.001, + "step": 25953 + }, + { + "epoch": 2.8137467476149176, + "grad_norm": 5.800680160522461, + "learning_rate": 1.2461917887712172e-06, + "loss": 0.026, + "step": 25954 + }, + { + "epoch": 2.8138551604509976, + "grad_norm": 0.03881622850894928, + "learning_rate": 1.24546641520383e-06, + "loss": 0.0001, + "step": 25955 + }, + { + "epoch": 2.813963573287077, + "grad_norm": 0.03176730126142502, + "learning_rate": 1.244741041636443e-06, + "loss": 0.0003, + "step": 25956 + }, + { + "epoch": 2.814071986123157, + "grad_norm": 3.169858932495117, + "learning_rate": 1.2440156680690558e-06, + "loss": 0.0255, + "step": 25957 + }, + { + "epoch": 2.8141803989592367, + "grad_norm": 3.089730978012085, + "learning_rate": 1.2432902945016686e-06, + "loss": 0.0439, + "step": 25958 + }, + { + "epoch": 2.8142888117953166, + "grad_norm": 3.3695809841156006, + "learning_rate": 1.2425649209342811e-06, + "loss": 0.0117, + "step": 25959 + }, + { + "epoch": 2.8143972246313966, + "grad_norm": 0.1016862764954567, + "learning_rate": 1.241839547366894e-06, + "loss": 0.0009, + "step": 25960 + }, + { + "epoch": 2.814505637467476, + "grad_norm": 0.40160176157951355, + "learning_rate": 1.2411141737995067e-06, + "loss": 0.0019, + "step": 25961 + }, + { + "epoch": 2.8146140503035557, + "grad_norm": 6.490359306335449, + "learning_rate": 1.2403888002321197e-06, + "loss": 0.0173, + "step": 25962 + }, + { + "epoch": 2.8147224631396357, + "grad_norm": 0.019279273226857185, + "learning_rate": 1.2396634266647325e-06, + "loss": 0.0001, + "step": 25963 + }, + { + "epoch": 2.8148308759757157, + "grad_norm": 1.455541968345642, + "learning_rate": 1.2389380530973452e-06, + "loss": 0.049, + "step": 25964 + }, + { + "epoch": 2.8149392888117952, + "grad_norm": 0.5249350666999817, + "learning_rate": 1.238212679529958e-06, + "loss": 0.007, + "step": 25965 + }, + { + "epoch": 2.815047701647875, + "grad_norm": 0.26029422879219055, + "learning_rate": 1.2374873059625708e-06, + "loss": 0.0016, + "step": 25966 + }, + { + "epoch": 2.8151561144839548, + "grad_norm": 3.9678287506103516, + "learning_rate": 1.2367619323951836e-06, + "loss": 0.0079, + "step": 25967 + }, + { + "epoch": 2.8152645273200347, + "grad_norm": 1.3480720520019531, + "learning_rate": 1.2360365588277964e-06, + "loss": 0.0047, + "step": 25968 + }, + { + "epoch": 2.8153729401561147, + "grad_norm": 0.10545438528060913, + "learning_rate": 1.2353111852604093e-06, + "loss": 0.0009, + "step": 25969 + }, + { + "epoch": 2.8154813529921943, + "grad_norm": 1.5975631475448608, + "learning_rate": 1.2345858116930221e-06, + "loss": 0.0257, + "step": 25970 + }, + { + "epoch": 2.815589765828274, + "grad_norm": 0.04583188518881798, + "learning_rate": 1.233860438125635e-06, + "loss": 0.0003, + "step": 25971 + }, + { + "epoch": 2.815698178664354, + "grad_norm": 3.0540692806243896, + "learning_rate": 1.2331350645582477e-06, + "loss": 0.003, + "step": 25972 + }, + { + "epoch": 2.8158065915004338, + "grad_norm": 0.013688808307051659, + "learning_rate": 1.2324096909908602e-06, + "loss": 0.0002, + "step": 25973 + }, + { + "epoch": 2.8159150043365133, + "grad_norm": 2.362051248550415, + "learning_rate": 1.2316843174234732e-06, + "loss": 0.0163, + "step": 25974 + }, + { + "epoch": 2.8160234171725933, + "grad_norm": 0.025279982015490532, + "learning_rate": 1.230958943856086e-06, + "loss": 0.0002, + "step": 25975 + }, + { + "epoch": 2.816131830008673, + "grad_norm": 6.462640285491943, + "learning_rate": 1.2302335702886988e-06, + "loss": 0.006, + "step": 25976 + }, + { + "epoch": 2.816240242844753, + "grad_norm": 0.17553868889808655, + "learning_rate": 1.2295081967213116e-06, + "loss": 0.0013, + "step": 25977 + }, + { + "epoch": 2.816348655680833, + "grad_norm": 0.7354332804679871, + "learning_rate": 1.2287828231539244e-06, + "loss": 0.0044, + "step": 25978 + }, + { + "epoch": 2.8164570685169124, + "grad_norm": 0.06769530475139618, + "learning_rate": 1.2280574495865371e-06, + "loss": 0.0004, + "step": 25979 + }, + { + "epoch": 2.8165654813529923, + "grad_norm": 0.10503804683685303, + "learning_rate": 1.22733207601915e-06, + "loss": 0.0008, + "step": 25980 + }, + { + "epoch": 2.816673894189072, + "grad_norm": 7.278690338134766, + "learning_rate": 1.2266067024517627e-06, + "loss": 0.0672, + "step": 25981 + }, + { + "epoch": 2.816782307025152, + "grad_norm": 0.1423773616552353, + "learning_rate": 1.2258813288843757e-06, + "loss": 0.001, + "step": 25982 + }, + { + "epoch": 2.8168907198612314, + "grad_norm": 3.3821873664855957, + "learning_rate": 1.2251559553169885e-06, + "loss": 0.0259, + "step": 25983 + }, + { + "epoch": 2.8169991326973114, + "grad_norm": 0.07603494822978973, + "learning_rate": 1.2244305817496012e-06, + "loss": 0.0005, + "step": 25984 + }, + { + "epoch": 2.817107545533391, + "grad_norm": 2.641819715499878, + "learning_rate": 1.223705208182214e-06, + "loss": 0.0093, + "step": 25985 + }, + { + "epoch": 2.817215958369471, + "grad_norm": 1.3093992471694946, + "learning_rate": 1.2229798346148266e-06, + "loss": 0.0477, + "step": 25986 + }, + { + "epoch": 2.817324371205551, + "grad_norm": 11.105440139770508, + "learning_rate": 1.2222544610474396e-06, + "loss": 0.0058, + "step": 25987 + }, + { + "epoch": 2.8174327840416304, + "grad_norm": 1.841160535812378, + "learning_rate": 1.2215290874800524e-06, + "loss": 0.0058, + "step": 25988 + }, + { + "epoch": 2.8175411968777104, + "grad_norm": 4.023859024047852, + "learning_rate": 1.2208037139126651e-06, + "loss": 0.0158, + "step": 25989 + }, + { + "epoch": 2.81764960971379, + "grad_norm": 0.06371266394853592, + "learning_rate": 1.220078340345278e-06, + "loss": 0.0004, + "step": 25990 + }, + { + "epoch": 2.81775802254987, + "grad_norm": 0.05764627829194069, + "learning_rate": 1.2193529667778907e-06, + "loss": 0.0004, + "step": 25991 + }, + { + "epoch": 2.81786643538595, + "grad_norm": 0.4640924632549286, + "learning_rate": 1.2186275932105035e-06, + "loss": 0.001, + "step": 25992 + }, + { + "epoch": 2.8179748482220295, + "grad_norm": 3.8057026863098145, + "learning_rate": 1.2179022196431162e-06, + "loss": 0.0309, + "step": 25993 + }, + { + "epoch": 2.818083261058109, + "grad_norm": 0.3802582919597626, + "learning_rate": 1.217176846075729e-06, + "loss": 0.0014, + "step": 25994 + }, + { + "epoch": 2.818191673894189, + "grad_norm": 11.161175727844238, + "learning_rate": 1.216451472508342e-06, + "loss": 0.0316, + "step": 25995 + }, + { + "epoch": 2.818300086730269, + "grad_norm": 1.9556009769439697, + "learning_rate": 1.2157260989409548e-06, + "loss": 0.0511, + "step": 25996 + }, + { + "epoch": 2.8184084995663485, + "grad_norm": 0.05419471859931946, + "learning_rate": 1.2150007253735676e-06, + "loss": 0.0007, + "step": 25997 + }, + { + "epoch": 2.8185169124024285, + "grad_norm": 0.02596196159720421, + "learning_rate": 1.2142753518061804e-06, + "loss": 0.0002, + "step": 25998 + }, + { + "epoch": 2.818625325238508, + "grad_norm": 0.13585638999938965, + "learning_rate": 1.2135499782387931e-06, + "loss": 0.0009, + "step": 25999 + }, + { + "epoch": 2.818733738074588, + "grad_norm": 0.0699445828795433, + "learning_rate": 1.212824604671406e-06, + "loss": 0.0004, + "step": 26000 + }, + { + "epoch": 2.818842150910668, + "grad_norm": 0.5825780034065247, + "learning_rate": 1.2120992311040187e-06, + "loss": 0.0027, + "step": 26001 + }, + { + "epoch": 2.8189505637467476, + "grad_norm": 8.045068740844727, + "learning_rate": 1.2113738575366315e-06, + "loss": 0.0137, + "step": 26002 + }, + { + "epoch": 2.8190589765828276, + "grad_norm": 0.06355854868888855, + "learning_rate": 1.2106484839692442e-06, + "loss": 0.0004, + "step": 26003 + }, + { + "epoch": 2.819167389418907, + "grad_norm": 4.09496545791626, + "learning_rate": 1.209923110401857e-06, + "loss": 0.0361, + "step": 26004 + }, + { + "epoch": 2.819275802254987, + "grad_norm": 0.021830949932336807, + "learning_rate": 1.2091977368344698e-06, + "loss": 0.0003, + "step": 26005 + }, + { + "epoch": 2.8193842150910666, + "grad_norm": 0.12481319904327393, + "learning_rate": 1.2084723632670826e-06, + "loss": 0.0005, + "step": 26006 + }, + { + "epoch": 2.8194926279271466, + "grad_norm": 0.09794748574495316, + "learning_rate": 1.2077469896996954e-06, + "loss": 0.0009, + "step": 26007 + }, + { + "epoch": 2.819601040763226, + "grad_norm": 0.11501260846853256, + "learning_rate": 1.2070216161323083e-06, + "loss": 0.0005, + "step": 26008 + }, + { + "epoch": 2.819709453599306, + "grad_norm": 0.18659788370132446, + "learning_rate": 1.2062962425649211e-06, + "loss": 0.0017, + "step": 26009 + }, + { + "epoch": 2.819817866435386, + "grad_norm": 0.041439589112997055, + "learning_rate": 1.205570868997534e-06, + "loss": 0.0003, + "step": 26010 + }, + { + "epoch": 2.8199262792714657, + "grad_norm": 0.5394419431686401, + "learning_rate": 1.2048454954301467e-06, + "loss": 0.0024, + "step": 26011 + }, + { + "epoch": 2.8200346921075456, + "grad_norm": 2.093954086303711, + "learning_rate": 1.2041201218627595e-06, + "loss": 0.0333, + "step": 26012 + }, + { + "epoch": 2.820143104943625, + "grad_norm": 3.637690782546997, + "learning_rate": 1.2033947482953722e-06, + "loss": 0.0126, + "step": 26013 + }, + { + "epoch": 2.820251517779705, + "grad_norm": 0.5624164342880249, + "learning_rate": 1.202669374727985e-06, + "loss": 0.0013, + "step": 26014 + }, + { + "epoch": 2.820359930615785, + "grad_norm": 0.08765976876020432, + "learning_rate": 1.2019440011605978e-06, + "loss": 0.0005, + "step": 26015 + }, + { + "epoch": 2.8204683434518647, + "grad_norm": 0.5112531185150146, + "learning_rate": 1.2012186275932106e-06, + "loss": 0.0014, + "step": 26016 + }, + { + "epoch": 2.8205767562879442, + "grad_norm": 1.3309649229049683, + "learning_rate": 1.2004932540258234e-06, + "loss": 0.0041, + "step": 26017 + }, + { + "epoch": 2.8206851691240242, + "grad_norm": 0.8145954608917236, + "learning_rate": 1.1997678804584361e-06, + "loss": 0.0017, + "step": 26018 + }, + { + "epoch": 2.820793581960104, + "grad_norm": 0.005830876063555479, + "learning_rate": 1.199042506891049e-06, + "loss": 0.0001, + "step": 26019 + }, + { + "epoch": 2.8209019947961838, + "grad_norm": 0.14058493077754974, + "learning_rate": 1.1983171333236617e-06, + "loss": 0.0011, + "step": 26020 + }, + { + "epoch": 2.8210104076322637, + "grad_norm": 1.1242480278015137, + "learning_rate": 1.1975917597562747e-06, + "loss": 0.0046, + "step": 26021 + }, + { + "epoch": 2.8211188204683433, + "grad_norm": 0.0787397250533104, + "learning_rate": 1.1968663861888875e-06, + "loss": 0.0006, + "step": 26022 + }, + { + "epoch": 2.8212272333044233, + "grad_norm": 0.05789044871926308, + "learning_rate": 1.1961410126215002e-06, + "loss": 0.0005, + "step": 26023 + }, + { + "epoch": 2.8213356461405033, + "grad_norm": 0.024509213864803314, + "learning_rate": 1.195415639054113e-06, + "loss": 0.0002, + "step": 26024 + }, + { + "epoch": 2.821444058976583, + "grad_norm": 0.3344206213951111, + "learning_rate": 1.1946902654867258e-06, + "loss": 0.0014, + "step": 26025 + }, + { + "epoch": 2.8215524718126628, + "grad_norm": 1.3770400285720825, + "learning_rate": 1.1939648919193386e-06, + "loss": 0.0037, + "step": 26026 + }, + { + "epoch": 2.8216608846487423, + "grad_norm": 0.1893649399280548, + "learning_rate": 1.1932395183519514e-06, + "loss": 0.0006, + "step": 26027 + }, + { + "epoch": 2.8217692974848223, + "grad_norm": 0.10571222752332687, + "learning_rate": 1.1925141447845641e-06, + "loss": 0.0005, + "step": 26028 + }, + { + "epoch": 2.821877710320902, + "grad_norm": 1.5923186540603638, + "learning_rate": 1.191788771217177e-06, + "loss": 0.0022, + "step": 26029 + }, + { + "epoch": 2.821986123156982, + "grad_norm": 1.1171607971191406, + "learning_rate": 1.1910633976497897e-06, + "loss": 0.0285, + "step": 26030 + }, + { + "epoch": 2.8220945359930614, + "grad_norm": 0.16999046504497528, + "learning_rate": 1.1903380240824025e-06, + "loss": 0.0012, + "step": 26031 + }, + { + "epoch": 2.8222029488291414, + "grad_norm": 0.7861710786819458, + "learning_rate": 1.1896126505150152e-06, + "loss": 0.0034, + "step": 26032 + }, + { + "epoch": 2.8223113616652213, + "grad_norm": 0.12930767238140106, + "learning_rate": 1.188887276947628e-06, + "loss": 0.001, + "step": 26033 + }, + { + "epoch": 2.822419774501301, + "grad_norm": 0.09025196731090546, + "learning_rate": 1.188161903380241e-06, + "loss": 0.0006, + "step": 26034 + }, + { + "epoch": 2.822528187337381, + "grad_norm": 2.1661245822906494, + "learning_rate": 1.1874365298128538e-06, + "loss": 0.0126, + "step": 26035 + }, + { + "epoch": 2.8226366001734604, + "grad_norm": 0.14087124168872833, + "learning_rate": 1.1867111562454666e-06, + "loss": 0.001, + "step": 26036 + }, + { + "epoch": 2.8227450130095404, + "grad_norm": 0.013060824014246464, + "learning_rate": 1.1859857826780794e-06, + "loss": 0.0001, + "step": 26037 + }, + { + "epoch": 2.8228534258456204, + "grad_norm": 0.03921951726078987, + "learning_rate": 1.1852604091106921e-06, + "loss": 0.0003, + "step": 26038 + }, + { + "epoch": 2.8229618386817, + "grad_norm": 0.00900336354970932, + "learning_rate": 1.184535035543305e-06, + "loss": 0.0001, + "step": 26039 + }, + { + "epoch": 2.8230702515177795, + "grad_norm": 0.17528598010540009, + "learning_rate": 1.1838096619759177e-06, + "loss": 0.0006, + "step": 26040 + }, + { + "epoch": 2.8231786643538594, + "grad_norm": 0.1860947161912918, + "learning_rate": 1.1830842884085305e-06, + "loss": 0.0007, + "step": 26041 + }, + { + "epoch": 2.8232870771899394, + "grad_norm": 0.03344360738992691, + "learning_rate": 1.1823589148411432e-06, + "loss": 0.0004, + "step": 26042 + }, + { + "epoch": 2.823395490026019, + "grad_norm": 6.277355670928955, + "learning_rate": 1.181633541273756e-06, + "loss": 0.0196, + "step": 26043 + }, + { + "epoch": 2.823503902862099, + "grad_norm": 13.786133766174316, + "learning_rate": 1.1809081677063688e-06, + "loss": 0.0455, + "step": 26044 + }, + { + "epoch": 2.8236123156981785, + "grad_norm": 0.7169220447540283, + "learning_rate": 1.1801827941389816e-06, + "loss": 0.0018, + "step": 26045 + }, + { + "epoch": 2.8237207285342585, + "grad_norm": 0.6057900190353394, + "learning_rate": 1.1794574205715944e-06, + "loss": 0.0019, + "step": 26046 + }, + { + "epoch": 2.8238291413703385, + "grad_norm": 1.9581692218780518, + "learning_rate": 1.1787320470042074e-06, + "loss": 0.0287, + "step": 26047 + }, + { + "epoch": 2.823937554206418, + "grad_norm": 0.036666009575128555, + "learning_rate": 1.1780066734368201e-06, + "loss": 0.0004, + "step": 26048 + }, + { + "epoch": 2.824045967042498, + "grad_norm": 0.025899676606059074, + "learning_rate": 1.177281299869433e-06, + "loss": 0.0002, + "step": 26049 + }, + { + "epoch": 2.8241543798785775, + "grad_norm": 4.2142205238342285, + "learning_rate": 1.1765559263020457e-06, + "loss": 0.024, + "step": 26050 + }, + { + "epoch": 2.8242627927146575, + "grad_norm": 9.4452543258667, + "learning_rate": 1.1758305527346585e-06, + "loss": 0.0418, + "step": 26051 + }, + { + "epoch": 2.824371205550737, + "grad_norm": 0.6694247126579285, + "learning_rate": 1.1751051791672712e-06, + "loss": 0.0032, + "step": 26052 + }, + { + "epoch": 2.824479618386817, + "grad_norm": 4.006790637969971, + "learning_rate": 1.174379805599884e-06, + "loss": 0.0117, + "step": 26053 + }, + { + "epoch": 2.8245880312228966, + "grad_norm": 0.014777830801904202, + "learning_rate": 1.1736544320324968e-06, + "loss": 0.0002, + "step": 26054 + }, + { + "epoch": 2.8246964440589766, + "grad_norm": 0.4307473301887512, + "learning_rate": 1.1729290584651096e-06, + "loss": 0.0018, + "step": 26055 + }, + { + "epoch": 2.8248048568950566, + "grad_norm": 2.2426493167877197, + "learning_rate": 1.1722036848977224e-06, + "loss": 0.0196, + "step": 26056 + }, + { + "epoch": 2.824913269731136, + "grad_norm": 0.2729799747467041, + "learning_rate": 1.1714783113303351e-06, + "loss": 0.0015, + "step": 26057 + }, + { + "epoch": 2.825021682567216, + "grad_norm": 0.3229120671749115, + "learning_rate": 1.170752937762948e-06, + "loss": 0.0018, + "step": 26058 + }, + { + "epoch": 2.8251300954032956, + "grad_norm": 0.5741593241691589, + "learning_rate": 1.1700275641955607e-06, + "loss": 0.001, + "step": 26059 + }, + { + "epoch": 2.8252385082393756, + "grad_norm": 0.11645493656396866, + "learning_rate": 1.1693021906281737e-06, + "loss": 0.0004, + "step": 26060 + }, + { + "epoch": 2.8253469210754556, + "grad_norm": 0.24542509019374847, + "learning_rate": 1.1685768170607865e-06, + "loss": 0.0004, + "step": 26061 + }, + { + "epoch": 2.825455333911535, + "grad_norm": 0.00872285757213831, + "learning_rate": 1.1678514434933992e-06, + "loss": 0.0001, + "step": 26062 + }, + { + "epoch": 2.8255637467476147, + "grad_norm": 0.08442534506320953, + "learning_rate": 1.167126069926012e-06, + "loss": 0.0005, + "step": 26063 + }, + { + "epoch": 2.8256721595836947, + "grad_norm": 4.015103816986084, + "learning_rate": 1.1664006963586248e-06, + "loss": 0.0317, + "step": 26064 + }, + { + "epoch": 2.8257805724197746, + "grad_norm": 9.876860618591309, + "learning_rate": 1.1656753227912376e-06, + "loss": 0.0249, + "step": 26065 + }, + { + "epoch": 2.825888985255854, + "grad_norm": 4.10395622253418, + "learning_rate": 1.1649499492238504e-06, + "loss": 0.0337, + "step": 26066 + }, + { + "epoch": 2.825997398091934, + "grad_norm": 0.042908746749162674, + "learning_rate": 1.1642245756564631e-06, + "loss": 0.0005, + "step": 26067 + }, + { + "epoch": 2.8261058109280137, + "grad_norm": 0.2972700297832489, + "learning_rate": 1.163499202089076e-06, + "loss": 0.001, + "step": 26068 + }, + { + "epoch": 2.8262142237640937, + "grad_norm": 0.06566017866134644, + "learning_rate": 1.1627738285216887e-06, + "loss": 0.0007, + "step": 26069 + }, + { + "epoch": 2.8263226366001737, + "grad_norm": 19.99687957763672, + "learning_rate": 1.1620484549543015e-06, + "loss": 0.0403, + "step": 26070 + }, + { + "epoch": 2.8264310494362532, + "grad_norm": 2.709738254547119, + "learning_rate": 1.1613230813869143e-06, + "loss": 0.0293, + "step": 26071 + }, + { + "epoch": 2.8265394622723328, + "grad_norm": 0.04457589238882065, + "learning_rate": 1.160597707819527e-06, + "loss": 0.0004, + "step": 26072 + }, + { + "epoch": 2.8266478751084128, + "grad_norm": 1.3284449577331543, + "learning_rate": 1.15987233425214e-06, + "loss": 0.0025, + "step": 26073 + }, + { + "epoch": 2.8267562879444927, + "grad_norm": 1.4490811824798584, + "learning_rate": 1.1591469606847528e-06, + "loss": 0.0056, + "step": 26074 + }, + { + "epoch": 2.8268647007805723, + "grad_norm": 0.05058828741312027, + "learning_rate": 1.1584215871173656e-06, + "loss": 0.0003, + "step": 26075 + }, + { + "epoch": 2.8269731136166523, + "grad_norm": 9.054451942443848, + "learning_rate": 1.1576962135499784e-06, + "loss": 0.0334, + "step": 26076 + }, + { + "epoch": 2.827081526452732, + "grad_norm": 0.029274117201566696, + "learning_rate": 1.1569708399825911e-06, + "loss": 0.0002, + "step": 26077 + }, + { + "epoch": 2.827189939288812, + "grad_norm": 0.03325812146067619, + "learning_rate": 1.156245466415204e-06, + "loss": 0.0003, + "step": 26078 + }, + { + "epoch": 2.827298352124892, + "grad_norm": 3.1590476036071777, + "learning_rate": 1.1555200928478167e-06, + "loss": 0.0296, + "step": 26079 + }, + { + "epoch": 2.8274067649609713, + "grad_norm": 0.06393340975046158, + "learning_rate": 1.1547947192804295e-06, + "loss": 0.0006, + "step": 26080 + }, + { + "epoch": 2.8275151777970513, + "grad_norm": 9.229759216308594, + "learning_rate": 1.1540693457130423e-06, + "loss": 0.0338, + "step": 26081 + }, + { + "epoch": 2.827623590633131, + "grad_norm": 5.047642707824707, + "learning_rate": 1.153343972145655e-06, + "loss": 0.0395, + "step": 26082 + }, + { + "epoch": 2.827732003469211, + "grad_norm": 0.11221469938755035, + "learning_rate": 1.1526185985782678e-06, + "loss": 0.0005, + "step": 26083 + }, + { + "epoch": 2.827840416305291, + "grad_norm": 0.2712711691856384, + "learning_rate": 1.1518932250108806e-06, + "loss": 0.0014, + "step": 26084 + }, + { + "epoch": 2.8279488291413704, + "grad_norm": 3.0452451705932617, + "learning_rate": 1.1511678514434934e-06, + "loss": 0.0039, + "step": 26085 + }, + { + "epoch": 2.82805724197745, + "grad_norm": 0.5905617475509644, + "learning_rate": 1.1504424778761064e-06, + "loss": 0.0013, + "step": 26086 + }, + { + "epoch": 2.82816565481353, + "grad_norm": 3.880997657775879, + "learning_rate": 1.1497171043087191e-06, + "loss": 0.0334, + "step": 26087 + }, + { + "epoch": 2.82827406764961, + "grad_norm": 1.9738606214523315, + "learning_rate": 1.148991730741332e-06, + "loss": 0.0091, + "step": 26088 + }, + { + "epoch": 2.8283824804856894, + "grad_norm": 2.542772054672241, + "learning_rate": 1.1482663571739447e-06, + "loss": 0.0294, + "step": 26089 + }, + { + "epoch": 2.8284908933217694, + "grad_norm": 0.37978503108024597, + "learning_rate": 1.1475409836065575e-06, + "loss": 0.0015, + "step": 26090 + }, + { + "epoch": 2.828599306157849, + "grad_norm": 0.08644428104162216, + "learning_rate": 1.1468156100391702e-06, + "loss": 0.0004, + "step": 26091 + }, + { + "epoch": 2.828707718993929, + "grad_norm": 0.04305649176239967, + "learning_rate": 1.146090236471783e-06, + "loss": 0.0004, + "step": 26092 + }, + { + "epoch": 2.828816131830009, + "grad_norm": 1.812085509300232, + "learning_rate": 1.145364862904396e-06, + "loss": 0.0384, + "step": 26093 + }, + { + "epoch": 2.8289245446660884, + "grad_norm": 0.07863198965787888, + "learning_rate": 1.1446394893370088e-06, + "loss": 0.0005, + "step": 26094 + }, + { + "epoch": 2.829032957502168, + "grad_norm": 2.63442325592041, + "learning_rate": 1.1439141157696214e-06, + "loss": 0.0037, + "step": 26095 + }, + { + "epoch": 2.829141370338248, + "grad_norm": 0.969534695148468, + "learning_rate": 1.1431887422022341e-06, + "loss": 0.0047, + "step": 26096 + }, + { + "epoch": 2.829249783174328, + "grad_norm": 0.09912921488285065, + "learning_rate": 1.142463368634847e-06, + "loss": 0.0006, + "step": 26097 + }, + { + "epoch": 2.8293581960104075, + "grad_norm": 0.06385865062475204, + "learning_rate": 1.1417379950674597e-06, + "loss": 0.0003, + "step": 26098 + }, + { + "epoch": 2.8294666088464875, + "grad_norm": 0.03471203148365021, + "learning_rate": 1.1410126215000727e-06, + "loss": 0.0004, + "step": 26099 + }, + { + "epoch": 2.829575021682567, + "grad_norm": 1.9963538646697998, + "learning_rate": 1.1402872479326855e-06, + "loss": 0.0283, + "step": 26100 + }, + { + "epoch": 2.829683434518647, + "grad_norm": 0.0649707093834877, + "learning_rate": 1.1395618743652982e-06, + "loss": 0.0004, + "step": 26101 + }, + { + "epoch": 2.829791847354727, + "grad_norm": 0.5717220306396484, + "learning_rate": 1.138836500797911e-06, + "loss": 0.0011, + "step": 26102 + }, + { + "epoch": 2.8299002601908065, + "grad_norm": 5.4927849769592285, + "learning_rate": 1.1381111272305238e-06, + "loss": 0.0322, + "step": 26103 + }, + { + "epoch": 2.8300086730268865, + "grad_norm": 0.22443634271621704, + "learning_rate": 1.1373857536631366e-06, + "loss": 0.0021, + "step": 26104 + }, + { + "epoch": 2.830117085862966, + "grad_norm": 0.8985130190849304, + "learning_rate": 1.1366603800957494e-06, + "loss": 0.002, + "step": 26105 + }, + { + "epoch": 2.830225498699046, + "grad_norm": 2.361208915710449, + "learning_rate": 1.1359350065283624e-06, + "loss": 0.0246, + "step": 26106 + }, + { + "epoch": 2.830333911535126, + "grad_norm": 1.3003244400024414, + "learning_rate": 1.1352096329609751e-06, + "loss": 0.0051, + "step": 26107 + }, + { + "epoch": 2.8304423243712056, + "grad_norm": 3.9936177730560303, + "learning_rate": 1.1344842593935877e-06, + "loss": 0.0094, + "step": 26108 + }, + { + "epoch": 2.830550737207285, + "grad_norm": 1.6030343770980835, + "learning_rate": 1.1337588858262005e-06, + "loss": 0.0102, + "step": 26109 + }, + { + "epoch": 2.830659150043365, + "grad_norm": 3.7613391876220703, + "learning_rate": 1.1330335122588133e-06, + "loss": 0.0635, + "step": 26110 + }, + { + "epoch": 2.830767562879445, + "grad_norm": 3.3111519813537598, + "learning_rate": 1.132308138691426e-06, + "loss": 0.0378, + "step": 26111 + }, + { + "epoch": 2.8308759757155246, + "grad_norm": 1.3367016315460205, + "learning_rate": 1.131582765124039e-06, + "loss": 0.0018, + "step": 26112 + }, + { + "epoch": 2.8309843885516046, + "grad_norm": 0.21121108531951904, + "learning_rate": 1.1308573915566518e-06, + "loss": 0.001, + "step": 26113 + }, + { + "epoch": 2.831092801387684, + "grad_norm": 0.7055564522743225, + "learning_rate": 1.1301320179892646e-06, + "loss": 0.0095, + "step": 26114 + }, + { + "epoch": 2.831201214223764, + "grad_norm": 1.065787434577942, + "learning_rate": 1.1294066444218774e-06, + "loss": 0.0037, + "step": 26115 + }, + { + "epoch": 2.831309627059844, + "grad_norm": 0.12316679209470749, + "learning_rate": 1.1286812708544901e-06, + "loss": 0.0006, + "step": 26116 + }, + { + "epoch": 2.8314180398959237, + "grad_norm": 2.8620388507843018, + "learning_rate": 1.127955897287103e-06, + "loss": 0.0046, + "step": 26117 + }, + { + "epoch": 2.831526452732003, + "grad_norm": 2.9064900875091553, + "learning_rate": 1.1272305237197157e-06, + "loss": 0.0098, + "step": 26118 + }, + { + "epoch": 2.831634865568083, + "grad_norm": 1.3970714807510376, + "learning_rate": 1.1265051501523287e-06, + "loss": 0.0008, + "step": 26119 + }, + { + "epoch": 2.831743278404163, + "grad_norm": 0.9432534575462341, + "learning_rate": 1.1257797765849415e-06, + "loss": 0.0007, + "step": 26120 + }, + { + "epoch": 2.8318516912402427, + "grad_norm": 6.076910018920898, + "learning_rate": 1.1250544030175542e-06, + "loss": 0.0212, + "step": 26121 + }, + { + "epoch": 2.8319601040763227, + "grad_norm": 8.400269508361816, + "learning_rate": 1.1243290294501668e-06, + "loss": 0.0285, + "step": 26122 + }, + { + "epoch": 2.8320685169124022, + "grad_norm": 0.8006654381752014, + "learning_rate": 1.1236036558827796e-06, + "loss": 0.003, + "step": 26123 + }, + { + "epoch": 2.8321769297484822, + "grad_norm": 16.64331817626953, + "learning_rate": 1.1228782823153926e-06, + "loss": 0.031, + "step": 26124 + }, + { + "epoch": 2.832285342584562, + "grad_norm": 0.09115003794431686, + "learning_rate": 1.1221529087480054e-06, + "loss": 0.0007, + "step": 26125 + }, + { + "epoch": 2.8323937554206418, + "grad_norm": 0.954840362071991, + "learning_rate": 1.1214275351806181e-06, + "loss": 0.0207, + "step": 26126 + }, + { + "epoch": 2.8325021682567217, + "grad_norm": 2.3884048461914062, + "learning_rate": 1.120702161613231e-06, + "loss": 0.032, + "step": 26127 + }, + { + "epoch": 2.8326105810928013, + "grad_norm": 3.3178741931915283, + "learning_rate": 1.1199767880458437e-06, + "loss": 0.0211, + "step": 26128 + }, + { + "epoch": 2.8327189939288813, + "grad_norm": 14.210948944091797, + "learning_rate": 1.1192514144784565e-06, + "loss": 0.0102, + "step": 26129 + }, + { + "epoch": 2.832827406764961, + "grad_norm": 0.055468685925006866, + "learning_rate": 1.1185260409110693e-06, + "loss": 0.0002, + "step": 26130 + }, + { + "epoch": 2.832935819601041, + "grad_norm": 0.23430871963500977, + "learning_rate": 1.117800667343682e-06, + "loss": 0.0011, + "step": 26131 + }, + { + "epoch": 2.8330442324371203, + "grad_norm": 0.08132006973028183, + "learning_rate": 1.117075293776295e-06, + "loss": 0.0007, + "step": 26132 + }, + { + "epoch": 2.8331526452732003, + "grad_norm": 0.09057407081127167, + "learning_rate": 1.1163499202089078e-06, + "loss": 0.0005, + "step": 26133 + }, + { + "epoch": 2.8332610581092803, + "grad_norm": 0.05009591579437256, + "learning_rate": 1.1156245466415206e-06, + "loss": 0.0003, + "step": 26134 + }, + { + "epoch": 2.83336947094536, + "grad_norm": 3.8111445903778076, + "learning_rate": 1.1148991730741331e-06, + "loss": 0.0102, + "step": 26135 + }, + { + "epoch": 2.83347788378144, + "grad_norm": 0.04297400265932083, + "learning_rate": 1.114173799506746e-06, + "loss": 0.0005, + "step": 26136 + }, + { + "epoch": 2.8335862966175194, + "grad_norm": 0.08203590661287308, + "learning_rate": 1.113448425939359e-06, + "loss": 0.0006, + "step": 26137 + }, + { + "epoch": 2.8336947094535994, + "grad_norm": 2.027742624282837, + "learning_rate": 1.1127230523719717e-06, + "loss": 0.0138, + "step": 26138 + }, + { + "epoch": 2.8338031222896793, + "grad_norm": 0.03312915563583374, + "learning_rate": 1.1119976788045845e-06, + "loss": 0.0003, + "step": 26139 + }, + { + "epoch": 2.833911535125759, + "grad_norm": 0.16475088894367218, + "learning_rate": 1.1112723052371973e-06, + "loss": 0.001, + "step": 26140 + }, + { + "epoch": 2.8340199479618384, + "grad_norm": 2.860370397567749, + "learning_rate": 1.11054693166981e-06, + "loss": 0.0022, + "step": 26141 + }, + { + "epoch": 2.8341283607979184, + "grad_norm": 0.02121610753238201, + "learning_rate": 1.1098215581024228e-06, + "loss": 0.0002, + "step": 26142 + }, + { + "epoch": 2.8342367736339984, + "grad_norm": 0.4722980558872223, + "learning_rate": 1.1090961845350356e-06, + "loss": 0.0013, + "step": 26143 + }, + { + "epoch": 2.834345186470078, + "grad_norm": 3.354046583175659, + "learning_rate": 1.1083708109676484e-06, + "loss": 0.0144, + "step": 26144 + }, + { + "epoch": 2.834453599306158, + "grad_norm": 5.018072605133057, + "learning_rate": 1.1076454374002614e-06, + "loss": 0.0136, + "step": 26145 + }, + { + "epoch": 2.8345620121422375, + "grad_norm": 0.45239362120628357, + "learning_rate": 1.1069200638328741e-06, + "loss": 0.002, + "step": 26146 + }, + { + "epoch": 2.8346704249783174, + "grad_norm": 2.969076156616211, + "learning_rate": 1.106194690265487e-06, + "loss": 0.0223, + "step": 26147 + }, + { + "epoch": 2.8347788378143974, + "grad_norm": 2.6091878414154053, + "learning_rate": 1.1054693166980997e-06, + "loss": 0.0306, + "step": 26148 + }, + { + "epoch": 2.834887250650477, + "grad_norm": 0.0529077984392643, + "learning_rate": 1.1047439431307123e-06, + "loss": 0.0003, + "step": 26149 + }, + { + "epoch": 2.834995663486557, + "grad_norm": 1.075883150100708, + "learning_rate": 1.1040185695633252e-06, + "loss": 0.0021, + "step": 26150 + }, + { + "epoch": 2.8351040763226365, + "grad_norm": 3.069047689437866, + "learning_rate": 1.103293195995938e-06, + "loss": 0.0208, + "step": 26151 + }, + { + "epoch": 2.8352124891587165, + "grad_norm": 2.870972156524658, + "learning_rate": 1.1025678224285508e-06, + "loss": 0.0486, + "step": 26152 + }, + { + "epoch": 2.835320901994796, + "grad_norm": 0.23393666744232178, + "learning_rate": 1.1018424488611636e-06, + "loss": 0.0011, + "step": 26153 + }, + { + "epoch": 2.835429314830876, + "grad_norm": 0.21916721761226654, + "learning_rate": 1.1011170752937764e-06, + "loss": 0.0005, + "step": 26154 + }, + { + "epoch": 2.8355377276669556, + "grad_norm": 0.2691790461540222, + "learning_rate": 1.1003917017263891e-06, + "loss": 0.0022, + "step": 26155 + }, + { + "epoch": 2.8356461405030355, + "grad_norm": 0.07995254546403885, + "learning_rate": 1.099666328159002e-06, + "loss": 0.0007, + "step": 26156 + }, + { + "epoch": 2.8357545533391155, + "grad_norm": 2.2122509479522705, + "learning_rate": 1.0989409545916147e-06, + "loss": 0.0067, + "step": 26157 + }, + { + "epoch": 2.835862966175195, + "grad_norm": 3.476583957672119, + "learning_rate": 1.0982155810242277e-06, + "loss": 0.0093, + "step": 26158 + }, + { + "epoch": 2.835971379011275, + "grad_norm": 2.4911327362060547, + "learning_rate": 1.0974902074568405e-06, + "loss": 0.0081, + "step": 26159 + }, + { + "epoch": 2.8360797918473546, + "grad_norm": 1.4757721424102783, + "learning_rate": 1.0967648338894532e-06, + "loss": 0.0283, + "step": 26160 + }, + { + "epoch": 2.8361882046834346, + "grad_norm": 9.58710765838623, + "learning_rate": 1.096039460322066e-06, + "loss": 0.029, + "step": 26161 + }, + { + "epoch": 2.8362966175195146, + "grad_norm": 3.0202603340148926, + "learning_rate": 1.0953140867546788e-06, + "loss": 0.006, + "step": 26162 + }, + { + "epoch": 2.836405030355594, + "grad_norm": 1.5497596263885498, + "learning_rate": 1.0945887131872916e-06, + "loss": 0.0058, + "step": 26163 + }, + { + "epoch": 2.8365134431916736, + "grad_norm": 0.05164797231554985, + "learning_rate": 1.0938633396199044e-06, + "loss": 0.0003, + "step": 26164 + }, + { + "epoch": 2.8366218560277536, + "grad_norm": 1.9262068271636963, + "learning_rate": 1.0931379660525171e-06, + "loss": 0.0052, + "step": 26165 + }, + { + "epoch": 2.8367302688638336, + "grad_norm": 0.0217727180570364, + "learning_rate": 1.09241259248513e-06, + "loss": 0.0002, + "step": 26166 + }, + { + "epoch": 2.836838681699913, + "grad_norm": 0.10871723294258118, + "learning_rate": 1.0916872189177427e-06, + "loss": 0.0008, + "step": 26167 + }, + { + "epoch": 2.836947094535993, + "grad_norm": 0.601698100566864, + "learning_rate": 1.0909618453503555e-06, + "loss": 0.0008, + "step": 26168 + }, + { + "epoch": 2.8370555073720727, + "grad_norm": 3.9869937896728516, + "learning_rate": 1.0902364717829683e-06, + "loss": 0.0495, + "step": 26169 + }, + { + "epoch": 2.8371639202081527, + "grad_norm": 0.03609265014529228, + "learning_rate": 1.089511098215581e-06, + "loss": 0.0003, + "step": 26170 + }, + { + "epoch": 2.8372723330442327, + "grad_norm": 0.012265280820429325, + "learning_rate": 1.088785724648194e-06, + "loss": 0.0001, + "step": 26171 + }, + { + "epoch": 2.837380745880312, + "grad_norm": 5.8138933181762695, + "learning_rate": 1.0880603510808068e-06, + "loss": 0.0182, + "step": 26172 + }, + { + "epoch": 2.837489158716392, + "grad_norm": 3.437516689300537, + "learning_rate": 1.0873349775134196e-06, + "loss": 0.0175, + "step": 26173 + }, + { + "epoch": 2.8375975715524717, + "grad_norm": 6.049077987670898, + "learning_rate": 1.0866096039460324e-06, + "loss": 0.0084, + "step": 26174 + }, + { + "epoch": 2.8377059843885517, + "grad_norm": 3.2663073539733887, + "learning_rate": 1.0858842303786451e-06, + "loss": 0.0451, + "step": 26175 + }, + { + "epoch": 2.8378143972246312, + "grad_norm": 0.10468343645334244, + "learning_rate": 1.085158856811258e-06, + "loss": 0.0005, + "step": 26176 + }, + { + "epoch": 2.8379228100607112, + "grad_norm": 3.5926010608673096, + "learning_rate": 1.0844334832438707e-06, + "loss": 0.0104, + "step": 26177 + }, + { + "epoch": 2.8380312228967908, + "grad_norm": 2.879709243774414, + "learning_rate": 1.0837081096764835e-06, + "loss": 0.0079, + "step": 26178 + }, + { + "epoch": 2.8381396357328708, + "grad_norm": 2.338181495666504, + "learning_rate": 1.0829827361090963e-06, + "loss": 0.0339, + "step": 26179 + }, + { + "epoch": 2.8382480485689507, + "grad_norm": 0.019611084833741188, + "learning_rate": 1.082257362541709e-06, + "loss": 0.0002, + "step": 26180 + }, + { + "epoch": 2.8383564614050303, + "grad_norm": 3.7984235286712646, + "learning_rate": 1.0815319889743218e-06, + "loss": 0.0024, + "step": 26181 + }, + { + "epoch": 2.8384648742411103, + "grad_norm": 1.3874807357788086, + "learning_rate": 1.0808066154069346e-06, + "loss": 0.0159, + "step": 26182 + }, + { + "epoch": 2.83857328707719, + "grad_norm": 9.900873184204102, + "learning_rate": 1.0800812418395474e-06, + "loss": 0.0112, + "step": 26183 + }, + { + "epoch": 2.83868169991327, + "grad_norm": 0.09164442867040634, + "learning_rate": 1.0793558682721604e-06, + "loss": 0.0002, + "step": 26184 + }, + { + "epoch": 2.83879011274935, + "grad_norm": 1.5246798992156982, + "learning_rate": 1.0786304947047731e-06, + "loss": 0.0034, + "step": 26185 + }, + { + "epoch": 2.8388985255854293, + "grad_norm": 0.35043850541114807, + "learning_rate": 1.077905121137386e-06, + "loss": 0.0014, + "step": 26186 + }, + { + "epoch": 2.839006938421509, + "grad_norm": 0.29483187198638916, + "learning_rate": 1.0771797475699987e-06, + "loss": 0.0006, + "step": 26187 + }, + { + "epoch": 2.839115351257589, + "grad_norm": 3.0327303409576416, + "learning_rate": 1.0764543740026115e-06, + "loss": 0.032, + "step": 26188 + }, + { + "epoch": 2.839223764093669, + "grad_norm": 0.011410507373511791, + "learning_rate": 1.0757290004352243e-06, + "loss": 0.0002, + "step": 26189 + }, + { + "epoch": 2.8393321769297484, + "grad_norm": 0.07547932863235474, + "learning_rate": 1.075003626867837e-06, + "loss": 0.0007, + "step": 26190 + }, + { + "epoch": 2.8394405897658284, + "grad_norm": 1.4541376829147339, + "learning_rate": 1.0742782533004498e-06, + "loss": 0.0067, + "step": 26191 + }, + { + "epoch": 2.839549002601908, + "grad_norm": 0.15192347764968872, + "learning_rate": 1.0735528797330626e-06, + "loss": 0.001, + "step": 26192 + }, + { + "epoch": 2.839657415437988, + "grad_norm": 0.3384363651275635, + "learning_rate": 1.0728275061656754e-06, + "loss": 0.0016, + "step": 26193 + }, + { + "epoch": 2.839765828274068, + "grad_norm": 1.768707275390625, + "learning_rate": 1.0721021325982881e-06, + "loss": 0.0061, + "step": 26194 + }, + { + "epoch": 2.8398742411101474, + "grad_norm": 5.957396030426025, + "learning_rate": 1.071376759030901e-06, + "loss": 0.0198, + "step": 26195 + }, + { + "epoch": 2.8399826539462274, + "grad_norm": 5.644471168518066, + "learning_rate": 1.0706513854635137e-06, + "loss": 0.041, + "step": 26196 + }, + { + "epoch": 2.840091066782307, + "grad_norm": 5.145005702972412, + "learning_rate": 1.0699260118961267e-06, + "loss": 0.0061, + "step": 26197 + }, + { + "epoch": 2.840199479618387, + "grad_norm": 1.2755506038665771, + "learning_rate": 1.0692006383287395e-06, + "loss": 0.0041, + "step": 26198 + }, + { + "epoch": 2.8403078924544665, + "grad_norm": 0.22275184094905853, + "learning_rate": 1.0684752647613523e-06, + "loss": 0.0007, + "step": 26199 + }, + { + "epoch": 2.8404163052905465, + "grad_norm": 2.0424723625183105, + "learning_rate": 1.067749891193965e-06, + "loss": 0.0098, + "step": 26200 + }, + { + "epoch": 2.840524718126626, + "grad_norm": 0.5381215810775757, + "learning_rate": 1.0670245176265778e-06, + "loss": 0.0019, + "step": 26201 + }, + { + "epoch": 2.840633130962706, + "grad_norm": 0.04603903740644455, + "learning_rate": 1.0662991440591906e-06, + "loss": 0.0005, + "step": 26202 + }, + { + "epoch": 2.840741543798786, + "grad_norm": 1.7064595222473145, + "learning_rate": 1.0655737704918034e-06, + "loss": 0.006, + "step": 26203 + }, + { + "epoch": 2.8408499566348655, + "grad_norm": 1.8332117795944214, + "learning_rate": 1.0648483969244161e-06, + "loss": 0.0329, + "step": 26204 + }, + { + "epoch": 2.8409583694709455, + "grad_norm": 0.09128683805465698, + "learning_rate": 1.064123023357029e-06, + "loss": 0.0008, + "step": 26205 + }, + { + "epoch": 2.841066782307025, + "grad_norm": 0.15346087515354156, + "learning_rate": 1.0633976497896417e-06, + "loss": 0.0005, + "step": 26206 + }, + { + "epoch": 2.841175195143105, + "grad_norm": 3.2217657566070557, + "learning_rate": 1.0626722762222545e-06, + "loss": 0.0244, + "step": 26207 + }, + { + "epoch": 2.841283607979185, + "grad_norm": 1.7819372415542603, + "learning_rate": 1.0619469026548673e-06, + "loss": 0.0069, + "step": 26208 + }, + { + "epoch": 2.8413920208152645, + "grad_norm": 3.9383533000946045, + "learning_rate": 1.06122152908748e-06, + "loss": 0.0047, + "step": 26209 + }, + { + "epoch": 2.841500433651344, + "grad_norm": 0.1433878391981125, + "learning_rate": 1.060496155520093e-06, + "loss": 0.0013, + "step": 26210 + }, + { + "epoch": 2.841608846487424, + "grad_norm": 0.24627585709095, + "learning_rate": 1.0597707819527058e-06, + "loss": 0.0009, + "step": 26211 + }, + { + "epoch": 2.841717259323504, + "grad_norm": 3.887722969055176, + "learning_rate": 1.0590454083853186e-06, + "loss": 0.0296, + "step": 26212 + }, + { + "epoch": 2.8418256721595836, + "grad_norm": 3.8281431198120117, + "learning_rate": 1.0583200348179314e-06, + "loss": 0.0184, + "step": 26213 + }, + { + "epoch": 2.8419340849956636, + "grad_norm": 0.0497012659907341, + "learning_rate": 1.0575946612505441e-06, + "loss": 0.0004, + "step": 26214 + }, + { + "epoch": 2.842042497831743, + "grad_norm": 0.014323534443974495, + "learning_rate": 1.056869287683157e-06, + "loss": 0.0001, + "step": 26215 + }, + { + "epoch": 2.842150910667823, + "grad_norm": 0.07645178586244583, + "learning_rate": 1.0561439141157697e-06, + "loss": 0.0003, + "step": 26216 + }, + { + "epoch": 2.842259323503903, + "grad_norm": 1.3656119108200073, + "learning_rate": 1.0554185405483825e-06, + "loss": 0.0157, + "step": 26217 + }, + { + "epoch": 2.8423677363399826, + "grad_norm": 0.3976196348667145, + "learning_rate": 1.0546931669809953e-06, + "loss": 0.0023, + "step": 26218 + }, + { + "epoch": 2.842476149176062, + "grad_norm": 3.8198535442352295, + "learning_rate": 1.053967793413608e-06, + "loss": 0.022, + "step": 26219 + }, + { + "epoch": 2.842584562012142, + "grad_norm": 1.168726921081543, + "learning_rate": 1.0532424198462208e-06, + "loss": 0.0051, + "step": 26220 + }, + { + "epoch": 2.842692974848222, + "grad_norm": 18.69605827331543, + "learning_rate": 1.0525170462788336e-06, + "loss": 0.0623, + "step": 26221 + }, + { + "epoch": 2.8428013876843017, + "grad_norm": 0.030784038826823235, + "learning_rate": 1.0517916727114464e-06, + "loss": 0.0003, + "step": 26222 + }, + { + "epoch": 2.8429098005203817, + "grad_norm": 0.3000393211841583, + "learning_rate": 1.0510662991440594e-06, + "loss": 0.0006, + "step": 26223 + }, + { + "epoch": 2.843018213356461, + "grad_norm": 0.3337700068950653, + "learning_rate": 1.0503409255766721e-06, + "loss": 0.0016, + "step": 26224 + }, + { + "epoch": 2.843126626192541, + "grad_norm": 9.53876781463623, + "learning_rate": 1.049615552009285e-06, + "loss": 0.0397, + "step": 26225 + }, + { + "epoch": 2.843235039028621, + "grad_norm": 0.11554192006587982, + "learning_rate": 1.0488901784418977e-06, + "loss": 0.0007, + "step": 26226 + }, + { + "epoch": 2.8433434518647007, + "grad_norm": 0.5295475125312805, + "learning_rate": 1.0481648048745105e-06, + "loss": 0.0018, + "step": 26227 + }, + { + "epoch": 2.8434518647007807, + "grad_norm": 0.7961758375167847, + "learning_rate": 1.0474394313071233e-06, + "loss": 0.0048, + "step": 26228 + }, + { + "epoch": 2.8435602775368602, + "grad_norm": 0.23072905838489532, + "learning_rate": 1.046714057739736e-06, + "loss": 0.0008, + "step": 26229 + }, + { + "epoch": 2.8436686903729402, + "grad_norm": 0.895939290523529, + "learning_rate": 1.0459886841723488e-06, + "loss": 0.0019, + "step": 26230 + }, + { + "epoch": 2.84377710320902, + "grad_norm": 2.0615782737731934, + "learning_rate": 1.0452633106049616e-06, + "loss": 0.0139, + "step": 26231 + }, + { + "epoch": 2.8438855160450998, + "grad_norm": 0.025251422077417374, + "learning_rate": 1.0445379370375744e-06, + "loss": 0.0002, + "step": 26232 + }, + { + "epoch": 2.8439939288811793, + "grad_norm": 0.08304394036531448, + "learning_rate": 1.0438125634701871e-06, + "loss": 0.0005, + "step": 26233 + }, + { + "epoch": 2.8441023417172593, + "grad_norm": 0.020520107820630074, + "learning_rate": 1.0430871899028e-06, + "loss": 0.0001, + "step": 26234 + }, + { + "epoch": 2.8442107545533393, + "grad_norm": 0.018692946061491966, + "learning_rate": 1.0423618163354127e-06, + "loss": 0.0002, + "step": 26235 + }, + { + "epoch": 2.844319167389419, + "grad_norm": 2.4313535690307617, + "learning_rate": 1.0416364427680257e-06, + "loss": 0.0182, + "step": 26236 + }, + { + "epoch": 2.844427580225499, + "grad_norm": 3.199561357498169, + "learning_rate": 1.0409110692006385e-06, + "loss": 0.0044, + "step": 26237 + }, + { + "epoch": 2.8445359930615783, + "grad_norm": 0.03161505237221718, + "learning_rate": 1.0401856956332513e-06, + "loss": 0.0003, + "step": 26238 + }, + { + "epoch": 2.8446444058976583, + "grad_norm": 4.167767524719238, + "learning_rate": 1.039460322065864e-06, + "loss": 0.0074, + "step": 26239 + }, + { + "epoch": 2.8447528187337383, + "grad_norm": 0.5730230212211609, + "learning_rate": 1.0387349484984768e-06, + "loss": 0.0012, + "step": 26240 + }, + { + "epoch": 2.844861231569818, + "grad_norm": 0.3259164094924927, + "learning_rate": 1.0380095749310896e-06, + "loss": 0.0023, + "step": 26241 + }, + { + "epoch": 2.8449696444058974, + "grad_norm": 2.7493252754211426, + "learning_rate": 1.0372842013637024e-06, + "loss": 0.0198, + "step": 26242 + }, + { + "epoch": 2.8450780572419774, + "grad_norm": 0.5913376808166504, + "learning_rate": 1.0365588277963151e-06, + "loss": 0.0019, + "step": 26243 + }, + { + "epoch": 2.8451864700780574, + "grad_norm": 0.02272322215139866, + "learning_rate": 1.035833454228928e-06, + "loss": 0.0002, + "step": 26244 + }, + { + "epoch": 2.845294882914137, + "grad_norm": 0.033111587166786194, + "learning_rate": 1.0351080806615407e-06, + "loss": 0.0003, + "step": 26245 + }, + { + "epoch": 2.845403295750217, + "grad_norm": 1.9463227987289429, + "learning_rate": 1.0343827070941535e-06, + "loss": 0.0112, + "step": 26246 + }, + { + "epoch": 2.8455117085862964, + "grad_norm": 0.3315405249595642, + "learning_rate": 1.0336573335267663e-06, + "loss": 0.0013, + "step": 26247 + }, + { + "epoch": 2.8456201214223764, + "grad_norm": 0.0245517585426569, + "learning_rate": 1.032931959959379e-06, + "loss": 0.0001, + "step": 26248 + }, + { + "epoch": 2.8457285342584564, + "grad_norm": 6.4189605712890625, + "learning_rate": 1.032206586391992e-06, + "loss": 0.0058, + "step": 26249 + }, + { + "epoch": 2.845836947094536, + "grad_norm": 1.4291707277297974, + "learning_rate": 1.0314812128246048e-06, + "loss": 0.0325, + "step": 26250 + }, + { + "epoch": 2.845945359930616, + "grad_norm": 0.0605664886534214, + "learning_rate": 1.0307558392572176e-06, + "loss": 0.0005, + "step": 26251 + }, + { + "epoch": 2.8460537727666955, + "grad_norm": 3.6680283546447754, + "learning_rate": 1.0300304656898304e-06, + "loss": 0.0064, + "step": 26252 + }, + { + "epoch": 2.8461621856027755, + "grad_norm": 0.09650271385908127, + "learning_rate": 1.0293050921224431e-06, + "loss": 0.0006, + "step": 26253 + }, + { + "epoch": 2.8462705984388554, + "grad_norm": 3.1450421810150146, + "learning_rate": 1.028579718555056e-06, + "loss": 0.0146, + "step": 26254 + }, + { + "epoch": 2.846379011274935, + "grad_norm": 0.08486256748437881, + "learning_rate": 1.0278543449876687e-06, + "loss": 0.0006, + "step": 26255 + }, + { + "epoch": 2.8464874241110145, + "grad_norm": 0.9777550101280212, + "learning_rate": 1.0271289714202817e-06, + "loss": 0.0019, + "step": 26256 + }, + { + "epoch": 2.8465958369470945, + "grad_norm": 3.4595115184783936, + "learning_rate": 1.0264035978528943e-06, + "loss": 0.0353, + "step": 26257 + }, + { + "epoch": 2.8467042497831745, + "grad_norm": 0.29749342799186707, + "learning_rate": 1.025678224285507e-06, + "loss": 0.0018, + "step": 26258 + }, + { + "epoch": 2.846812662619254, + "grad_norm": 4.092766284942627, + "learning_rate": 1.0249528507181198e-06, + "loss": 0.0126, + "step": 26259 + }, + { + "epoch": 2.846921075455334, + "grad_norm": 2.1681735515594482, + "learning_rate": 1.0242274771507326e-06, + "loss": 0.0157, + "step": 26260 + }, + { + "epoch": 2.8470294882914136, + "grad_norm": 0.15898959338665009, + "learning_rate": 1.0235021035833454e-06, + "loss": 0.0008, + "step": 26261 + }, + { + "epoch": 2.8471379011274935, + "grad_norm": 7.711094856262207, + "learning_rate": 1.0227767300159584e-06, + "loss": 0.0468, + "step": 26262 + }, + { + "epoch": 2.8472463139635735, + "grad_norm": 0.5855044722557068, + "learning_rate": 1.0220513564485711e-06, + "loss": 0.0021, + "step": 26263 + }, + { + "epoch": 2.847354726799653, + "grad_norm": 4.301718711853027, + "learning_rate": 1.021325982881184e-06, + "loss": 0.0135, + "step": 26264 + }, + { + "epoch": 2.8474631396357326, + "grad_norm": 6.196464538574219, + "learning_rate": 1.0206006093137967e-06, + "loss": 0.0366, + "step": 26265 + }, + { + "epoch": 2.8475715524718126, + "grad_norm": 1.3975238800048828, + "learning_rate": 1.0198752357464095e-06, + "loss": 0.0302, + "step": 26266 + }, + { + "epoch": 2.8476799653078926, + "grad_norm": 0.2649327218532562, + "learning_rate": 1.0191498621790223e-06, + "loss": 0.0012, + "step": 26267 + }, + { + "epoch": 2.847788378143972, + "grad_norm": 0.047713227570056915, + "learning_rate": 1.018424488611635e-06, + "loss": 0.0005, + "step": 26268 + }, + { + "epoch": 2.847896790980052, + "grad_norm": 0.32622474431991577, + "learning_rate": 1.017699115044248e-06, + "loss": 0.0007, + "step": 26269 + }, + { + "epoch": 2.8480052038161316, + "grad_norm": 0.44505223631858826, + "learning_rate": 1.0169737414768608e-06, + "loss": 0.0021, + "step": 26270 + }, + { + "epoch": 2.8481136166522116, + "grad_norm": 2.7558116912841797, + "learning_rate": 1.0162483679094734e-06, + "loss": 0.0054, + "step": 26271 + }, + { + "epoch": 2.8482220294882916, + "grad_norm": 0.07711987942457199, + "learning_rate": 1.0155229943420862e-06, + "loss": 0.0007, + "step": 26272 + }, + { + "epoch": 2.848330442324371, + "grad_norm": 4.331855773925781, + "learning_rate": 1.014797620774699e-06, + "loss": 0.0026, + "step": 26273 + }, + { + "epoch": 2.848438855160451, + "grad_norm": 8.99905014038086, + "learning_rate": 1.014072247207312e-06, + "loss": 0.0268, + "step": 26274 + }, + { + "epoch": 2.8485472679965307, + "grad_norm": 0.638943612575531, + "learning_rate": 1.0133468736399247e-06, + "loss": 0.0503, + "step": 26275 + }, + { + "epoch": 2.8486556808326107, + "grad_norm": 0.41623544692993164, + "learning_rate": 1.0126215000725375e-06, + "loss": 0.0013, + "step": 26276 + }, + { + "epoch": 2.84876409366869, + "grad_norm": 3.3398661613464355, + "learning_rate": 1.0118961265051503e-06, + "loss": 0.0047, + "step": 26277 + }, + { + "epoch": 2.84887250650477, + "grad_norm": 0.37122979760169983, + "learning_rate": 1.011170752937763e-06, + "loss": 0.0015, + "step": 26278 + }, + { + "epoch": 2.8489809193408497, + "grad_norm": 3.4860751628875732, + "learning_rate": 1.0104453793703758e-06, + "loss": 0.039, + "step": 26279 + }, + { + "epoch": 2.8490893321769297, + "grad_norm": 0.5976065397262573, + "learning_rate": 1.0097200058029886e-06, + "loss": 0.0004, + "step": 26280 + }, + { + "epoch": 2.8491977450130097, + "grad_norm": 1.103468418121338, + "learning_rate": 1.0089946322356014e-06, + "loss": 0.0033, + "step": 26281 + }, + { + "epoch": 2.8493061578490892, + "grad_norm": 1.282541275024414, + "learning_rate": 1.0082692586682144e-06, + "loss": 0.0408, + "step": 26282 + }, + { + "epoch": 2.8494145706851692, + "grad_norm": 6.004312038421631, + "learning_rate": 1.0075438851008271e-06, + "loss": 0.0098, + "step": 26283 + }, + { + "epoch": 2.8495229835212488, + "grad_norm": 0.7866176962852478, + "learning_rate": 1.0068185115334397e-06, + "loss": 0.0032, + "step": 26284 + }, + { + "epoch": 2.8496313963573288, + "grad_norm": 0.020734017714858055, + "learning_rate": 1.0060931379660525e-06, + "loss": 0.0001, + "step": 26285 + }, + { + "epoch": 2.8497398091934087, + "grad_norm": 0.08891039341688156, + "learning_rate": 1.0053677643986653e-06, + "loss": 0.0004, + "step": 26286 + }, + { + "epoch": 2.8498482220294883, + "grad_norm": 8.265947341918945, + "learning_rate": 1.0046423908312783e-06, + "loss": 0.0235, + "step": 26287 + }, + { + "epoch": 2.849956634865568, + "grad_norm": 0.04791382700204849, + "learning_rate": 1.003917017263891e-06, + "loss": 0.0005, + "step": 26288 + }, + { + "epoch": 2.850065047701648, + "grad_norm": 1.399598240852356, + "learning_rate": 1.0031916436965038e-06, + "loss": 0.0027, + "step": 26289 + }, + { + "epoch": 2.850173460537728, + "grad_norm": 0.30193328857421875, + "learning_rate": 1.0024662701291166e-06, + "loss": 0.0019, + "step": 26290 + }, + { + "epoch": 2.8502818733738073, + "grad_norm": 0.12296568602323532, + "learning_rate": 1.0017408965617294e-06, + "loss": 0.0008, + "step": 26291 + }, + { + "epoch": 2.8503902862098873, + "grad_norm": 0.01960105262696743, + "learning_rate": 1.0010155229943421e-06, + "loss": 0.0002, + "step": 26292 + }, + { + "epoch": 2.850498699045967, + "grad_norm": 1.8622620105743408, + "learning_rate": 1.000290149426955e-06, + "loss": 0.0351, + "step": 26293 + }, + { + "epoch": 2.850607111882047, + "grad_norm": 6.772964000701904, + "learning_rate": 9.995647758595677e-07, + "loss": 0.0172, + "step": 26294 + }, + { + "epoch": 2.850715524718127, + "grad_norm": 0.6761320233345032, + "learning_rate": 9.988394022921807e-07, + "loss": 0.0024, + "step": 26295 + }, + { + "epoch": 2.8508239375542064, + "grad_norm": 0.381622850894928, + "learning_rate": 9.981140287247935e-07, + "loss": 0.0017, + "step": 26296 + }, + { + "epoch": 2.8509323503902864, + "grad_norm": 0.13121286034584045, + "learning_rate": 9.973886551574063e-07, + "loss": 0.0009, + "step": 26297 + }, + { + "epoch": 2.851040763226366, + "grad_norm": 2.887648344039917, + "learning_rate": 9.966632815900188e-07, + "loss": 0.0352, + "step": 26298 + }, + { + "epoch": 2.851149176062446, + "grad_norm": 0.027220943942666054, + "learning_rate": 9.959379080226316e-07, + "loss": 0.0003, + "step": 26299 + }, + { + "epoch": 2.8512575888985254, + "grad_norm": 0.01661336049437523, + "learning_rate": 9.952125344552446e-07, + "loss": 0.0002, + "step": 26300 + }, + { + "epoch": 2.8513660017346054, + "grad_norm": 19.903234481811523, + "learning_rate": 9.944871608878574e-07, + "loss": 0.0076, + "step": 26301 + }, + { + "epoch": 2.851474414570685, + "grad_norm": 2.7054011821746826, + "learning_rate": 9.937617873204701e-07, + "loss": 0.0261, + "step": 26302 + }, + { + "epoch": 2.851582827406765, + "grad_norm": 15.77989673614502, + "learning_rate": 9.93036413753083e-07, + "loss": 0.0364, + "step": 26303 + }, + { + "epoch": 2.851691240242845, + "grad_norm": 0.9606302976608276, + "learning_rate": 9.923110401856957e-07, + "loss": 0.031, + "step": 26304 + }, + { + "epoch": 2.8517996530789245, + "grad_norm": 0.14700262248516083, + "learning_rate": 9.915856666183085e-07, + "loss": 0.0008, + "step": 26305 + }, + { + "epoch": 2.8519080659150045, + "grad_norm": 2.1877944469451904, + "learning_rate": 9.908602930509213e-07, + "loss": 0.0037, + "step": 26306 + }, + { + "epoch": 2.852016478751084, + "grad_norm": 0.021997366100549698, + "learning_rate": 9.90134919483534e-07, + "loss": 0.0001, + "step": 26307 + }, + { + "epoch": 2.852124891587164, + "grad_norm": 0.12913483381271362, + "learning_rate": 9.89409545916147e-07, + "loss": 0.0004, + "step": 26308 + }, + { + "epoch": 2.852233304423244, + "grad_norm": 0.021332889795303345, + "learning_rate": 9.886841723487598e-07, + "loss": 0.0002, + "step": 26309 + }, + { + "epoch": 2.8523417172593235, + "grad_norm": 0.33010169863700867, + "learning_rate": 9.879587987813726e-07, + "loss": 0.0007, + "step": 26310 + }, + { + "epoch": 2.852450130095403, + "grad_norm": 0.24155119061470032, + "learning_rate": 9.872334252139854e-07, + "loss": 0.0008, + "step": 26311 + }, + { + "epoch": 2.852558542931483, + "grad_norm": 8.378458023071289, + "learning_rate": 9.86508051646598e-07, + "loss": 0.0146, + "step": 26312 + }, + { + "epoch": 2.852666955767563, + "grad_norm": 0.9299297332763672, + "learning_rate": 9.85782678079211e-07, + "loss": 0.0009, + "step": 26313 + }, + { + "epoch": 2.8527753686036426, + "grad_norm": 0.25002145767211914, + "learning_rate": 9.850573045118237e-07, + "loss": 0.0015, + "step": 26314 + }, + { + "epoch": 2.8528837814397225, + "grad_norm": 0.027237333357334137, + "learning_rate": 9.843319309444365e-07, + "loss": 0.0003, + "step": 26315 + }, + { + "epoch": 2.852992194275802, + "grad_norm": 0.07735610753297806, + "learning_rate": 9.836065573770493e-07, + "loss": 0.0006, + "step": 26316 + }, + { + "epoch": 2.853100607111882, + "grad_norm": 0.16948658227920532, + "learning_rate": 9.82881183809662e-07, + "loss": 0.0006, + "step": 26317 + }, + { + "epoch": 2.853209019947962, + "grad_norm": 0.04635543376207352, + "learning_rate": 9.821558102422748e-07, + "loss": 0.0004, + "step": 26318 + }, + { + "epoch": 2.8533174327840416, + "grad_norm": 0.6725022196769714, + "learning_rate": 9.814304366748876e-07, + "loss": 0.0025, + "step": 26319 + }, + { + "epoch": 2.8534258456201216, + "grad_norm": 1.2461808919906616, + "learning_rate": 9.807050631075004e-07, + "loss": 0.0274, + "step": 26320 + }, + { + "epoch": 2.853534258456201, + "grad_norm": 4.9329633712768555, + "learning_rate": 9.799796895401134e-07, + "loss": 0.023, + "step": 26321 + }, + { + "epoch": 2.853642671292281, + "grad_norm": 1.2395697832107544, + "learning_rate": 9.792543159727261e-07, + "loss": 0.0318, + "step": 26322 + }, + { + "epoch": 2.8537510841283606, + "grad_norm": 0.0785578265786171, + "learning_rate": 9.78528942405339e-07, + "loss": 0.0004, + "step": 26323 + }, + { + "epoch": 2.8538594969644406, + "grad_norm": 0.06067240610718727, + "learning_rate": 9.778035688379517e-07, + "loss": 0.0007, + "step": 26324 + }, + { + "epoch": 2.85396790980052, + "grad_norm": 0.47054001688957214, + "learning_rate": 9.770781952705643e-07, + "loss": 0.0008, + "step": 26325 + }, + { + "epoch": 2.8540763226366, + "grad_norm": 1.027938961982727, + "learning_rate": 9.763528217031773e-07, + "loss": 0.0015, + "step": 26326 + }, + { + "epoch": 2.85418473547268, + "grad_norm": 0.028407851234078407, + "learning_rate": 9.7562744813579e-07, + "loss": 0.0003, + "step": 26327 + }, + { + "epoch": 2.8542931483087597, + "grad_norm": 0.12614767253398895, + "learning_rate": 9.749020745684028e-07, + "loss": 0.0006, + "step": 26328 + }, + { + "epoch": 2.8544015611448397, + "grad_norm": 3.002594232559204, + "learning_rate": 9.741767010010156e-07, + "loss": 0.038, + "step": 26329 + }, + { + "epoch": 2.854509973980919, + "grad_norm": 0.06931837648153305, + "learning_rate": 9.734513274336284e-07, + "loss": 0.0004, + "step": 26330 + }, + { + "epoch": 2.854618386816999, + "grad_norm": 0.21317113935947418, + "learning_rate": 9.727259538662412e-07, + "loss": 0.001, + "step": 26331 + }, + { + "epoch": 2.854726799653079, + "grad_norm": 0.5600626468658447, + "learning_rate": 9.72000580298854e-07, + "loss": 0.0032, + "step": 26332 + }, + { + "epoch": 2.8548352124891587, + "grad_norm": 0.34528279304504395, + "learning_rate": 9.712752067314667e-07, + "loss": 0.0014, + "step": 26333 + }, + { + "epoch": 2.8549436253252383, + "grad_norm": 0.017342790961265564, + "learning_rate": 9.705498331640797e-07, + "loss": 0.0002, + "step": 26334 + }, + { + "epoch": 2.8550520381613183, + "grad_norm": 0.7266972064971924, + "learning_rate": 9.698244595966925e-07, + "loss": 0.0016, + "step": 26335 + }, + { + "epoch": 2.8551604509973982, + "grad_norm": 0.6347476840019226, + "learning_rate": 9.690990860293053e-07, + "loss": 0.001, + "step": 26336 + }, + { + "epoch": 2.8552688638334778, + "grad_norm": 0.22718849778175354, + "learning_rate": 9.68373712461918e-07, + "loss": 0.0005, + "step": 26337 + }, + { + "epoch": 2.8553772766695578, + "grad_norm": 0.14427824318408966, + "learning_rate": 9.676483388945308e-07, + "loss": 0.0007, + "step": 26338 + }, + { + "epoch": 2.8554856895056373, + "grad_norm": 0.1395386904478073, + "learning_rate": 9.669229653271436e-07, + "loss": 0.0007, + "step": 26339 + }, + { + "epoch": 2.8555941023417173, + "grad_norm": 0.03472168371081352, + "learning_rate": 9.661975917597564e-07, + "loss": 0.0003, + "step": 26340 + }, + { + "epoch": 2.8557025151777973, + "grad_norm": 10.76657772064209, + "learning_rate": 9.654722181923692e-07, + "loss": 0.0081, + "step": 26341 + }, + { + "epoch": 2.855810928013877, + "grad_norm": 0.03692436218261719, + "learning_rate": 9.64746844624982e-07, + "loss": 0.0002, + "step": 26342 + }, + { + "epoch": 2.855919340849957, + "grad_norm": 0.2728940546512604, + "learning_rate": 9.640214710575947e-07, + "loss": 0.0009, + "step": 26343 + }, + { + "epoch": 2.8560277536860363, + "grad_norm": 1.3226313591003418, + "learning_rate": 9.632960974902075e-07, + "loss": 0.0072, + "step": 26344 + }, + { + "epoch": 2.8561361665221163, + "grad_norm": 0.7762860655784607, + "learning_rate": 9.625707239228203e-07, + "loss": 0.0016, + "step": 26345 + }, + { + "epoch": 2.856244579358196, + "grad_norm": 0.11634309589862823, + "learning_rate": 9.61845350355433e-07, + "loss": 0.0007, + "step": 26346 + }, + { + "epoch": 2.856352992194276, + "grad_norm": 2.8590307235717773, + "learning_rate": 9.61119976788046e-07, + "loss": 0.0465, + "step": 26347 + }, + { + "epoch": 2.8564614050303554, + "grad_norm": 0.8117227554321289, + "learning_rate": 9.603946032206588e-07, + "loss": 0.0029, + "step": 26348 + }, + { + "epoch": 2.8565698178664354, + "grad_norm": 7.501105785369873, + "learning_rate": 9.596692296532716e-07, + "loss": 0.0171, + "step": 26349 + }, + { + "epoch": 2.8566782307025154, + "grad_norm": 0.1257782280445099, + "learning_rate": 9.589438560858844e-07, + "loss": 0.0006, + "step": 26350 + }, + { + "epoch": 2.856786643538595, + "grad_norm": 0.24845415353775024, + "learning_rate": 9.582184825184971e-07, + "loss": 0.0017, + "step": 26351 + }, + { + "epoch": 2.856895056374675, + "grad_norm": 3.7918338775634766, + "learning_rate": 9.5749310895111e-07, + "loss": 0.0665, + "step": 26352 + }, + { + "epoch": 2.8570034692107544, + "grad_norm": 2.412121057510376, + "learning_rate": 9.567677353837227e-07, + "loss": 0.0325, + "step": 26353 + }, + { + "epoch": 2.8571118820468344, + "grad_norm": 16.613306045532227, + "learning_rate": 9.560423618163355e-07, + "loss": 0.0162, + "step": 26354 + }, + { + "epoch": 2.8572202948829144, + "grad_norm": 0.1378326117992401, + "learning_rate": 9.553169882489483e-07, + "loss": 0.0005, + "step": 26355 + }, + { + "epoch": 2.857328707718994, + "grad_norm": 0.059113115072250366, + "learning_rate": 9.54591614681561e-07, + "loss": 0.0005, + "step": 26356 + }, + { + "epoch": 2.8574371205550735, + "grad_norm": 2.1255781650543213, + "learning_rate": 9.538662411141738e-07, + "loss": 0.0125, + "step": 26357 + }, + { + "epoch": 2.8575455333911535, + "grad_norm": 0.030830074101686478, + "learning_rate": 9.531408675467867e-07, + "loss": 0.0002, + "step": 26358 + }, + { + "epoch": 2.8576539462272335, + "grad_norm": 1.0923622846603394, + "learning_rate": 9.524154939793995e-07, + "loss": 0.0033, + "step": 26359 + }, + { + "epoch": 2.857762359063313, + "grad_norm": 4.8771162033081055, + "learning_rate": 9.516901204120123e-07, + "loss": 0.0292, + "step": 26360 + }, + { + "epoch": 2.857870771899393, + "grad_norm": 1.8439682722091675, + "learning_rate": 9.50964746844625e-07, + "loss": 0.0294, + "step": 26361 + }, + { + "epoch": 2.8579791847354725, + "grad_norm": 3.2582054138183594, + "learning_rate": 9.502393732772379e-07, + "loss": 0.0176, + "step": 26362 + }, + { + "epoch": 2.8580875975715525, + "grad_norm": 2.8872437477111816, + "learning_rate": 9.495139997098507e-07, + "loss": 0.0132, + "step": 26363 + }, + { + "epoch": 2.8581960104076325, + "grad_norm": 4.346555709838867, + "learning_rate": 9.487886261424635e-07, + "loss": 0.0095, + "step": 26364 + }, + { + "epoch": 2.858304423243712, + "grad_norm": 1.8958481550216675, + "learning_rate": 9.480632525750763e-07, + "loss": 0.003, + "step": 26365 + }, + { + "epoch": 2.8584128360797916, + "grad_norm": 0.029713820666074753, + "learning_rate": 9.473378790076889e-07, + "loss": 0.0002, + "step": 26366 + }, + { + "epoch": 2.8585212489158716, + "grad_norm": 0.042906492948532104, + "learning_rate": 9.466125054403018e-07, + "loss": 0.0006, + "step": 26367 + }, + { + "epoch": 2.8586296617519515, + "grad_norm": 0.025103144347667694, + "learning_rate": 9.458871318729146e-07, + "loss": 0.0001, + "step": 26368 + }, + { + "epoch": 2.858738074588031, + "grad_norm": 1.7617018222808838, + "learning_rate": 9.451617583055274e-07, + "loss": 0.032, + "step": 26369 + }, + { + "epoch": 2.858846487424111, + "grad_norm": 0.03850479796528816, + "learning_rate": 9.444363847381402e-07, + "loss": 0.0003, + "step": 26370 + }, + { + "epoch": 2.8589549002601906, + "grad_norm": 2.737611770629883, + "learning_rate": 9.43711011170753e-07, + "loss": 0.023, + "step": 26371 + }, + { + "epoch": 2.8590633130962706, + "grad_norm": 5.426394939422607, + "learning_rate": 9.429856376033658e-07, + "loss": 0.0165, + "step": 26372 + }, + { + "epoch": 2.8591717259323506, + "grad_norm": 8.805907249450684, + "learning_rate": 9.422602640359786e-07, + "loss": 0.0353, + "step": 26373 + }, + { + "epoch": 2.85928013876843, + "grad_norm": 2.685234308242798, + "learning_rate": 9.415348904685914e-07, + "loss": 0.0294, + "step": 26374 + }, + { + "epoch": 2.85938855160451, + "grad_norm": 0.16669932007789612, + "learning_rate": 9.408095169012043e-07, + "loss": 0.0007, + "step": 26375 + }, + { + "epoch": 2.8594969644405897, + "grad_norm": 3.087266445159912, + "learning_rate": 9.40084143333817e-07, + "loss": 0.0168, + "step": 26376 + }, + { + "epoch": 2.8596053772766696, + "grad_norm": 0.02314305678009987, + "learning_rate": 9.393587697664298e-07, + "loss": 0.0002, + "step": 26377 + }, + { + "epoch": 2.8597137901127496, + "grad_norm": 0.49935638904571533, + "learning_rate": 9.386333961990426e-07, + "loss": 0.0026, + "step": 26378 + }, + { + "epoch": 2.859822202948829, + "grad_norm": 5.028162956237793, + "learning_rate": 9.379080226316553e-07, + "loss": 0.0714, + "step": 26379 + }, + { + "epoch": 2.8599306157849087, + "grad_norm": 0.17695829272270203, + "learning_rate": 9.371826490642682e-07, + "loss": 0.0012, + "step": 26380 + }, + { + "epoch": 2.8600390286209887, + "grad_norm": 0.3086180090904236, + "learning_rate": 9.364572754968809e-07, + "loss": 0.0014, + "step": 26381 + }, + { + "epoch": 2.8601474414570687, + "grad_norm": 0.13828317821025848, + "learning_rate": 9.357319019294937e-07, + "loss": 0.0005, + "step": 26382 + }, + { + "epoch": 2.860255854293148, + "grad_norm": 1.7702184915542603, + "learning_rate": 9.350065283621065e-07, + "loss": 0.0085, + "step": 26383 + }, + { + "epoch": 2.860364267129228, + "grad_norm": 0.14143982529640198, + "learning_rate": 9.342811547947194e-07, + "loss": 0.0008, + "step": 26384 + }, + { + "epoch": 2.8604726799653077, + "grad_norm": 0.04708152636885643, + "learning_rate": 9.335557812273322e-07, + "loss": 0.0004, + "step": 26385 + }, + { + "epoch": 2.8605810928013877, + "grad_norm": 1.1508522033691406, + "learning_rate": 9.328304076599449e-07, + "loss": 0.003, + "step": 26386 + }, + { + "epoch": 2.8606895056374677, + "grad_norm": 0.07862761616706848, + "learning_rate": 9.321050340925577e-07, + "loss": 0.0004, + "step": 26387 + }, + { + "epoch": 2.8607979184735473, + "grad_norm": 0.017023775726556778, + "learning_rate": 9.313796605251706e-07, + "loss": 0.0002, + "step": 26388 + }, + { + "epoch": 2.860906331309627, + "grad_norm": 0.02477080188691616, + "learning_rate": 9.306542869577834e-07, + "loss": 0.0002, + "step": 26389 + }, + { + "epoch": 2.861014744145707, + "grad_norm": 4.377696990966797, + "learning_rate": 9.299289133903962e-07, + "loss": 0.0207, + "step": 26390 + }, + { + "epoch": 2.8611231569817868, + "grad_norm": 0.6203030943870544, + "learning_rate": 9.292035398230089e-07, + "loss": 0.0007, + "step": 26391 + }, + { + "epoch": 2.8612315698178663, + "grad_norm": 2.8041248321533203, + "learning_rate": 9.284781662556218e-07, + "loss": 0.0073, + "step": 26392 + }, + { + "epoch": 2.8613399826539463, + "grad_norm": 4.2498650550842285, + "learning_rate": 9.277527926882345e-07, + "loss": 0.0324, + "step": 26393 + }, + { + "epoch": 2.861448395490026, + "grad_norm": 0.09014928340911865, + "learning_rate": 9.270274191208473e-07, + "loss": 0.0006, + "step": 26394 + }, + { + "epoch": 2.861556808326106, + "grad_norm": 14.594766616821289, + "learning_rate": 9.2630204555346e-07, + "loss": 0.0094, + "step": 26395 + }, + { + "epoch": 2.861665221162186, + "grad_norm": 2.1852834224700928, + "learning_rate": 9.255766719860728e-07, + "loss": 0.0181, + "step": 26396 + }, + { + "epoch": 2.8617736339982653, + "grad_norm": 1.0032963752746582, + "learning_rate": 9.248512984186857e-07, + "loss": 0.0321, + "step": 26397 + }, + { + "epoch": 2.8618820468343453, + "grad_norm": 0.07423661649227142, + "learning_rate": 9.241259248512985e-07, + "loss": 0.0003, + "step": 26398 + }, + { + "epoch": 2.861990459670425, + "grad_norm": 4.135098457336426, + "learning_rate": 9.234005512839113e-07, + "loss": 0.049, + "step": 26399 + }, + { + "epoch": 2.862098872506505, + "grad_norm": 0.20677024126052856, + "learning_rate": 9.22675177716524e-07, + "loss": 0.0008, + "step": 26400 + }, + { + "epoch": 2.862207285342585, + "grad_norm": 0.44720450043678284, + "learning_rate": 9.219498041491369e-07, + "loss": 0.0009, + "step": 26401 + }, + { + "epoch": 2.8623156981786644, + "grad_norm": 0.2643100321292877, + "learning_rate": 9.212244305817497e-07, + "loss": 0.0013, + "step": 26402 + }, + { + "epoch": 2.862424111014744, + "grad_norm": 0.10439251363277435, + "learning_rate": 9.204990570143625e-07, + "loss": 0.0006, + "step": 26403 + }, + { + "epoch": 2.862532523850824, + "grad_norm": 1.02808678150177, + "learning_rate": 9.197736834469753e-07, + "loss": 0.0106, + "step": 26404 + }, + { + "epoch": 2.862640936686904, + "grad_norm": 0.15081699192523956, + "learning_rate": 9.190483098795882e-07, + "loss": 0.0012, + "step": 26405 + }, + { + "epoch": 2.8627493495229834, + "grad_norm": 0.10753151774406433, + "learning_rate": 9.183229363122008e-07, + "loss": 0.0007, + "step": 26406 + }, + { + "epoch": 2.8628577623590634, + "grad_norm": 0.019794540479779243, + "learning_rate": 9.175975627448136e-07, + "loss": 0.0002, + "step": 26407 + }, + { + "epoch": 2.862966175195143, + "grad_norm": 3.195155620574951, + "learning_rate": 9.168721891774264e-07, + "loss": 0.0508, + "step": 26408 + }, + { + "epoch": 2.863074588031223, + "grad_norm": 0.1078764796257019, + "learning_rate": 9.161468156100392e-07, + "loss": 0.0005, + "step": 26409 + }, + { + "epoch": 2.863183000867303, + "grad_norm": 0.7867595553398132, + "learning_rate": 9.15421442042652e-07, + "loss": 0.0032, + "step": 26410 + }, + { + "epoch": 2.8632914137033825, + "grad_norm": 0.05374639108777046, + "learning_rate": 9.146960684752648e-07, + "loss": 0.0005, + "step": 26411 + }, + { + "epoch": 2.863399826539462, + "grad_norm": 0.05138583481311798, + "learning_rate": 9.139706949078776e-07, + "loss": 0.0006, + "step": 26412 + }, + { + "epoch": 2.863508239375542, + "grad_norm": 0.9113348126411438, + "learning_rate": 9.132453213404904e-07, + "loss": 0.0031, + "step": 26413 + }, + { + "epoch": 2.863616652211622, + "grad_norm": 0.020480871200561523, + "learning_rate": 9.125199477731033e-07, + "loss": 0.0002, + "step": 26414 + }, + { + "epoch": 2.8637250650477015, + "grad_norm": 0.31076350808143616, + "learning_rate": 9.11794574205716e-07, + "loss": 0.0017, + "step": 26415 + }, + { + "epoch": 2.8638334778837815, + "grad_norm": 4.2866740226745605, + "learning_rate": 9.110692006383288e-07, + "loss": 0.009, + "step": 26416 + }, + { + "epoch": 2.863941890719861, + "grad_norm": 0.038366980850696564, + "learning_rate": 9.103438270709417e-07, + "loss": 0.0004, + "step": 26417 + }, + { + "epoch": 2.864050303555941, + "grad_norm": 0.29740050435066223, + "learning_rate": 9.096184535035545e-07, + "loss": 0.0017, + "step": 26418 + }, + { + "epoch": 2.864158716392021, + "grad_norm": 0.12492506206035614, + "learning_rate": 9.088930799361673e-07, + "loss": 0.0011, + "step": 26419 + }, + { + "epoch": 2.8642671292281006, + "grad_norm": 4.982802391052246, + "learning_rate": 9.081677063687799e-07, + "loss": 0.0104, + "step": 26420 + }, + { + "epoch": 2.8643755420641805, + "grad_norm": 0.25490379333496094, + "learning_rate": 9.074423328013927e-07, + "loss": 0.0017, + "step": 26421 + }, + { + "epoch": 2.86448395490026, + "grad_norm": 0.7777912616729736, + "learning_rate": 9.067169592340055e-07, + "loss": 0.0026, + "step": 26422 + }, + { + "epoch": 2.86459236773634, + "grad_norm": 1.2457153797149658, + "learning_rate": 9.059915856666184e-07, + "loss": 0.002, + "step": 26423 + }, + { + "epoch": 2.8647007805724196, + "grad_norm": 1.2770113945007324, + "learning_rate": 9.052662120992312e-07, + "loss": 0.0272, + "step": 26424 + }, + { + "epoch": 2.8648091934084996, + "grad_norm": 0.0297947209328413, + "learning_rate": 9.045408385318439e-07, + "loss": 0.0003, + "step": 26425 + }, + { + "epoch": 2.864917606244579, + "grad_norm": 0.1406306028366089, + "learning_rate": 9.038154649644568e-07, + "loss": 0.0006, + "step": 26426 + }, + { + "epoch": 2.865026019080659, + "grad_norm": 2.5017542839050293, + "learning_rate": 9.030900913970696e-07, + "loss": 0.0313, + "step": 26427 + }, + { + "epoch": 2.865134431916739, + "grad_norm": 3.5907092094421387, + "learning_rate": 9.023647178296824e-07, + "loss": 0.0084, + "step": 26428 + }, + { + "epoch": 2.8652428447528187, + "grad_norm": 0.5089981555938721, + "learning_rate": 9.016393442622952e-07, + "loss": 0.0018, + "step": 26429 + }, + { + "epoch": 2.8653512575888986, + "grad_norm": 0.422095388174057, + "learning_rate": 9.00913970694908e-07, + "loss": 0.0005, + "step": 26430 + }, + { + "epoch": 2.865459670424978, + "grad_norm": 0.1458650529384613, + "learning_rate": 9.001885971275208e-07, + "loss": 0.0017, + "step": 26431 + }, + { + "epoch": 2.865568083261058, + "grad_norm": 0.005578933749347925, + "learning_rate": 8.994632235601336e-07, + "loss": 0.0001, + "step": 26432 + }, + { + "epoch": 2.865676496097138, + "grad_norm": 0.19591212272644043, + "learning_rate": 8.987378499927463e-07, + "loss": 0.0008, + "step": 26433 + }, + { + "epoch": 2.8657849089332177, + "grad_norm": 5.723886489868164, + "learning_rate": 8.98012476425359e-07, + "loss": 0.0165, + "step": 26434 + }, + { + "epoch": 2.8658933217692972, + "grad_norm": 3.046076536178589, + "learning_rate": 8.972871028579719e-07, + "loss": 0.0772, + "step": 26435 + }, + { + "epoch": 2.866001734605377, + "grad_norm": 1.1288660764694214, + "learning_rate": 8.965617292905847e-07, + "loss": 0.0034, + "step": 26436 + }, + { + "epoch": 2.866110147441457, + "grad_norm": 2.276651620864868, + "learning_rate": 8.958363557231975e-07, + "loss": 0.0422, + "step": 26437 + }, + { + "epoch": 2.8662185602775367, + "grad_norm": 0.463340699672699, + "learning_rate": 8.951109821558103e-07, + "loss": 0.002, + "step": 26438 + }, + { + "epoch": 2.8663269731136167, + "grad_norm": 0.6125754117965698, + "learning_rate": 8.943856085884232e-07, + "loss": 0.0032, + "step": 26439 + }, + { + "epoch": 2.8664353859496963, + "grad_norm": 0.09080293774604797, + "learning_rate": 8.936602350210359e-07, + "loss": 0.0007, + "step": 26440 + }, + { + "epoch": 2.8665437987857763, + "grad_norm": 9.034358024597168, + "learning_rate": 8.929348614536487e-07, + "loss": 0.0129, + "step": 26441 + }, + { + "epoch": 2.8666522116218562, + "grad_norm": 0.35691022872924805, + "learning_rate": 8.922094878862615e-07, + "loss": 0.0015, + "step": 26442 + }, + { + "epoch": 2.866760624457936, + "grad_norm": 0.07624217867851257, + "learning_rate": 8.914841143188744e-07, + "loss": 0.0005, + "step": 26443 + }, + { + "epoch": 2.8668690372940158, + "grad_norm": 3.848374843597412, + "learning_rate": 8.907587407514872e-07, + "loss": 0.0105, + "step": 26444 + }, + { + "epoch": 2.8669774501300953, + "grad_norm": 1.7224810123443604, + "learning_rate": 8.900333671840999e-07, + "loss": 0.0529, + "step": 26445 + }, + { + "epoch": 2.8670858629661753, + "grad_norm": 0.04076980799436569, + "learning_rate": 8.893079936167127e-07, + "loss": 0.0004, + "step": 26446 + }, + { + "epoch": 2.867194275802255, + "grad_norm": 0.3488345146179199, + "learning_rate": 8.885826200493254e-07, + "loss": 0.0014, + "step": 26447 + }, + { + "epoch": 2.867302688638335, + "grad_norm": 0.7542880177497864, + "learning_rate": 8.878572464819383e-07, + "loss": 0.002, + "step": 26448 + }, + { + "epoch": 2.8674111014744144, + "grad_norm": 4.680333614349365, + "learning_rate": 8.87131872914551e-07, + "loss": 0.0049, + "step": 26449 + }, + { + "epoch": 2.8675195143104943, + "grad_norm": 0.01575358957052231, + "learning_rate": 8.864064993471638e-07, + "loss": 0.0002, + "step": 26450 + }, + { + "epoch": 2.8676279271465743, + "grad_norm": 3.4329421520233154, + "learning_rate": 8.856811257797766e-07, + "loss": 0.0592, + "step": 26451 + }, + { + "epoch": 2.867736339982654, + "grad_norm": 2.3551647663116455, + "learning_rate": 8.849557522123895e-07, + "loss": 0.0236, + "step": 26452 + }, + { + "epoch": 2.867844752818734, + "grad_norm": 1.5157124996185303, + "learning_rate": 8.842303786450023e-07, + "loss": 0.0174, + "step": 26453 + }, + { + "epoch": 2.8679531656548134, + "grad_norm": 0.9926807284355164, + "learning_rate": 8.83505005077615e-07, + "loss": 0.0023, + "step": 26454 + }, + { + "epoch": 2.8680615784908934, + "grad_norm": 5.554308891296387, + "learning_rate": 8.827796315102278e-07, + "loss": 0.0154, + "step": 26455 + }, + { + "epoch": 2.8681699913269734, + "grad_norm": 0.8253231048583984, + "learning_rate": 8.820542579428407e-07, + "loss": 0.0025, + "step": 26456 + }, + { + "epoch": 2.868278404163053, + "grad_norm": 1.0838165283203125, + "learning_rate": 8.813288843754535e-07, + "loss": 0.0025, + "step": 26457 + }, + { + "epoch": 2.8683868169991324, + "grad_norm": 0.20403416454792023, + "learning_rate": 8.806035108080663e-07, + "loss": 0.0009, + "step": 26458 + }, + { + "epoch": 2.8684952298352124, + "grad_norm": 0.18002113699913025, + "learning_rate": 8.79878137240679e-07, + "loss": 0.0012, + "step": 26459 + }, + { + "epoch": 2.8686036426712924, + "grad_norm": 0.05836395174264908, + "learning_rate": 8.791527636732917e-07, + "loss": 0.0006, + "step": 26460 + }, + { + "epoch": 2.868712055507372, + "grad_norm": 0.4804096221923828, + "learning_rate": 8.784273901059046e-07, + "loss": 0.0016, + "step": 26461 + }, + { + "epoch": 2.868820468343452, + "grad_norm": 0.17980726063251495, + "learning_rate": 8.777020165385174e-07, + "loss": 0.001, + "step": 26462 + }, + { + "epoch": 2.8689288811795315, + "grad_norm": 0.20860224962234497, + "learning_rate": 8.769766429711302e-07, + "loss": 0.0007, + "step": 26463 + }, + { + "epoch": 2.8690372940156115, + "grad_norm": 1.1314867734909058, + "learning_rate": 8.762512694037429e-07, + "loss": 0.0328, + "step": 26464 + }, + { + "epoch": 2.8691457068516915, + "grad_norm": 0.3271941542625427, + "learning_rate": 8.755258958363558e-07, + "loss": 0.002, + "step": 26465 + }, + { + "epoch": 2.869254119687771, + "grad_norm": 0.07134118676185608, + "learning_rate": 8.748005222689686e-07, + "loss": 0.0003, + "step": 26466 + }, + { + "epoch": 2.869362532523851, + "grad_norm": 0.23063941299915314, + "learning_rate": 8.740751487015814e-07, + "loss": 0.0016, + "step": 26467 + }, + { + "epoch": 2.8694709453599305, + "grad_norm": 0.42510348558425903, + "learning_rate": 8.733497751341942e-07, + "loss": 0.003, + "step": 26468 + }, + { + "epoch": 2.8695793581960105, + "grad_norm": 0.010203447192907333, + "learning_rate": 8.72624401566807e-07, + "loss": 0.0001, + "step": 26469 + }, + { + "epoch": 2.86968777103209, + "grad_norm": 0.34015727043151855, + "learning_rate": 8.718990279994198e-07, + "loss": 0.0017, + "step": 26470 + }, + { + "epoch": 2.86979618386817, + "grad_norm": 0.822404682636261, + "learning_rate": 8.711736544320326e-07, + "loss": 0.0032, + "step": 26471 + }, + { + "epoch": 2.8699045967042496, + "grad_norm": 0.051541734486818314, + "learning_rate": 8.704482808646454e-07, + "loss": 0.0005, + "step": 26472 + }, + { + "epoch": 2.8700130095403296, + "grad_norm": 0.06263592094182968, + "learning_rate": 8.697229072972583e-07, + "loss": 0.0004, + "step": 26473 + }, + { + "epoch": 2.8701214223764095, + "grad_norm": 0.024545658379793167, + "learning_rate": 8.689975337298709e-07, + "loss": 0.0003, + "step": 26474 + }, + { + "epoch": 2.870229835212489, + "grad_norm": 0.03677219897508621, + "learning_rate": 8.682721601624837e-07, + "loss": 0.0004, + "step": 26475 + }, + { + "epoch": 2.870338248048569, + "grad_norm": 4.293689727783203, + "learning_rate": 8.675467865950965e-07, + "loss": 0.0218, + "step": 26476 + }, + { + "epoch": 2.8704466608846486, + "grad_norm": 1.2314553260803223, + "learning_rate": 8.668214130277093e-07, + "loss": 0.0298, + "step": 26477 + }, + { + "epoch": 2.8705550737207286, + "grad_norm": 2.8493995666503906, + "learning_rate": 8.660960394603222e-07, + "loss": 0.0381, + "step": 26478 + }, + { + "epoch": 2.8706634865568086, + "grad_norm": 0.3079712390899658, + "learning_rate": 8.653706658929349e-07, + "loss": 0.0013, + "step": 26479 + }, + { + "epoch": 2.870771899392888, + "grad_norm": 0.09045502543449402, + "learning_rate": 8.646452923255477e-07, + "loss": 0.0008, + "step": 26480 + }, + { + "epoch": 2.8708803122289677, + "grad_norm": 0.045913469046354294, + "learning_rate": 8.639199187581605e-07, + "loss": 0.0005, + "step": 26481 + }, + { + "epoch": 2.8709887250650477, + "grad_norm": 5.508289813995361, + "learning_rate": 8.631945451907734e-07, + "loss": 0.0223, + "step": 26482 + }, + { + "epoch": 2.8710971379011276, + "grad_norm": 1.1376888751983643, + "learning_rate": 8.624691716233862e-07, + "loss": 0.0056, + "step": 26483 + }, + { + "epoch": 2.871205550737207, + "grad_norm": 0.011487810872495174, + "learning_rate": 8.617437980559989e-07, + "loss": 0.0002, + "step": 26484 + }, + { + "epoch": 2.871313963573287, + "grad_norm": 3.9369654655456543, + "learning_rate": 8.610184244886117e-07, + "loss": 0.0144, + "step": 26485 + }, + { + "epoch": 2.8714223764093667, + "grad_norm": 2.6764371395111084, + "learning_rate": 8.602930509212246e-07, + "loss": 0.0484, + "step": 26486 + }, + { + "epoch": 2.8715307892454467, + "grad_norm": 16.07872772216797, + "learning_rate": 8.595676773538374e-07, + "loss": 0.0606, + "step": 26487 + }, + { + "epoch": 2.8716392020815267, + "grad_norm": 3.048429250717163, + "learning_rate": 8.5884230378645e-07, + "loss": 0.0191, + "step": 26488 + }, + { + "epoch": 2.871747614917606, + "grad_norm": 2.9917354583740234, + "learning_rate": 8.581169302190628e-07, + "loss": 0.0322, + "step": 26489 + }, + { + "epoch": 2.871856027753686, + "grad_norm": 3.4956135749816895, + "learning_rate": 8.573915566516756e-07, + "loss": 0.0266, + "step": 26490 + }, + { + "epoch": 2.8719644405897657, + "grad_norm": 0.648070752620697, + "learning_rate": 8.566661830842885e-07, + "loss": 0.0023, + "step": 26491 + }, + { + "epoch": 2.8720728534258457, + "grad_norm": 3.0840277671813965, + "learning_rate": 8.559408095169013e-07, + "loss": 0.0049, + "step": 26492 + }, + { + "epoch": 2.8721812662619253, + "grad_norm": 0.24503542482852936, + "learning_rate": 8.55215435949514e-07, + "loss": 0.0009, + "step": 26493 + }, + { + "epoch": 2.8722896790980053, + "grad_norm": 0.05296620726585388, + "learning_rate": 8.544900623821268e-07, + "loss": 0.0003, + "step": 26494 + }, + { + "epoch": 2.872398091934085, + "grad_norm": 6.060492515563965, + "learning_rate": 8.537646888147397e-07, + "loss": 0.041, + "step": 26495 + }, + { + "epoch": 2.872506504770165, + "grad_norm": 8.633379936218262, + "learning_rate": 8.530393152473525e-07, + "loss": 0.0274, + "step": 26496 + }, + { + "epoch": 2.8726149176062448, + "grad_norm": 4.726291656494141, + "learning_rate": 8.523139416799653e-07, + "loss": 0.0234, + "step": 26497 + }, + { + "epoch": 2.8727233304423243, + "grad_norm": 3.5221402645111084, + "learning_rate": 8.51588568112578e-07, + "loss": 0.0132, + "step": 26498 + }, + { + "epoch": 2.8728317432784043, + "grad_norm": 9.544782638549805, + "learning_rate": 8.508631945451909e-07, + "loss": 0.0438, + "step": 26499 + }, + { + "epoch": 2.872940156114484, + "grad_norm": 0.34177473187446594, + "learning_rate": 8.501378209778037e-07, + "loss": 0.0009, + "step": 26500 + }, + { + "epoch": 2.873048568950564, + "grad_norm": 0.2105325609445572, + "learning_rate": 8.494124474104164e-07, + "loss": 0.0008, + "step": 26501 + }, + { + "epoch": 2.873156981786644, + "grad_norm": 0.06025814265012741, + "learning_rate": 8.486870738430292e-07, + "loss": 0.0003, + "step": 26502 + }, + { + "epoch": 2.8732653946227233, + "grad_norm": 4.354841232299805, + "learning_rate": 8.479617002756419e-07, + "loss": 0.0204, + "step": 26503 + }, + { + "epoch": 2.873373807458803, + "grad_norm": 3.851608991622925, + "learning_rate": 8.472363267082548e-07, + "loss": 0.0143, + "step": 26504 + }, + { + "epoch": 2.873482220294883, + "grad_norm": 0.1670377254486084, + "learning_rate": 8.465109531408676e-07, + "loss": 0.0006, + "step": 26505 + }, + { + "epoch": 2.873590633130963, + "grad_norm": 3.9624059200286865, + "learning_rate": 8.457855795734804e-07, + "loss": 0.0263, + "step": 26506 + }, + { + "epoch": 2.8736990459670424, + "grad_norm": 14.747564315795898, + "learning_rate": 8.450602060060932e-07, + "loss": 0.0172, + "step": 26507 + }, + { + "epoch": 2.8738074588031224, + "grad_norm": 2.229736089706421, + "learning_rate": 8.44334832438706e-07, + "loss": 0.0254, + "step": 26508 + }, + { + "epoch": 2.873915871639202, + "grad_norm": 0.45613712072372437, + "learning_rate": 8.436094588713188e-07, + "loss": 0.0015, + "step": 26509 + }, + { + "epoch": 2.874024284475282, + "grad_norm": 3.0842673778533936, + "learning_rate": 8.428840853039316e-07, + "loss": 0.0107, + "step": 26510 + }, + { + "epoch": 2.874132697311362, + "grad_norm": 0.38195282220840454, + "learning_rate": 8.421587117365444e-07, + "loss": 0.0013, + "step": 26511 + }, + { + "epoch": 2.8742411101474414, + "grad_norm": 0.18336857855319977, + "learning_rate": 8.414333381691573e-07, + "loss": 0.0012, + "step": 26512 + }, + { + "epoch": 2.874349522983521, + "grad_norm": 0.2226163148880005, + "learning_rate": 8.4070796460177e-07, + "loss": 0.001, + "step": 26513 + }, + { + "epoch": 2.874457935819601, + "grad_norm": 2.752913236618042, + "learning_rate": 8.399825910343828e-07, + "loss": 0.007, + "step": 26514 + }, + { + "epoch": 2.874566348655681, + "grad_norm": 0.14814217388629913, + "learning_rate": 8.392572174669955e-07, + "loss": 0.0009, + "step": 26515 + }, + { + "epoch": 2.8746747614917605, + "grad_norm": 0.8400922417640686, + "learning_rate": 8.385318438996083e-07, + "loss": 0.0045, + "step": 26516 + }, + { + "epoch": 2.8747831743278405, + "grad_norm": 0.08152595162391663, + "learning_rate": 8.378064703322212e-07, + "loss": 0.0006, + "step": 26517 + }, + { + "epoch": 2.87489158716392, + "grad_norm": 0.29158616065979004, + "learning_rate": 8.370810967648339e-07, + "loss": 0.0011, + "step": 26518 + }, + { + "epoch": 2.875, + "grad_norm": 0.4901926517486572, + "learning_rate": 8.363557231974467e-07, + "loss": 0.0008, + "step": 26519 + }, + { + "epoch": 2.87510841283608, + "grad_norm": 0.06069103255867958, + "learning_rate": 8.356303496300595e-07, + "loss": 0.0006, + "step": 26520 + }, + { + "epoch": 2.8752168256721595, + "grad_norm": 0.08546659350395203, + "learning_rate": 8.349049760626724e-07, + "loss": 0.0006, + "step": 26521 + }, + { + "epoch": 2.8753252385082395, + "grad_norm": 0.13249200582504272, + "learning_rate": 8.341796024952852e-07, + "loss": 0.0009, + "step": 26522 + }, + { + "epoch": 2.875433651344319, + "grad_norm": 0.3263591229915619, + "learning_rate": 8.334542289278979e-07, + "loss": 0.0024, + "step": 26523 + }, + { + "epoch": 2.875542064180399, + "grad_norm": 0.14882947504520416, + "learning_rate": 8.327288553605107e-07, + "loss": 0.0003, + "step": 26524 + }, + { + "epoch": 2.875650477016479, + "grad_norm": 0.0163232684135437, + "learning_rate": 8.320034817931236e-07, + "loss": 0.0002, + "step": 26525 + }, + { + "epoch": 2.8757588898525586, + "grad_norm": 0.07501086592674255, + "learning_rate": 8.312781082257364e-07, + "loss": 0.0008, + "step": 26526 + }, + { + "epoch": 2.875867302688638, + "grad_norm": 0.1197299212217331, + "learning_rate": 8.305527346583492e-07, + "loss": 0.001, + "step": 26527 + }, + { + "epoch": 2.875975715524718, + "grad_norm": 5.2906084060668945, + "learning_rate": 8.298273610909618e-07, + "loss": 0.0296, + "step": 26528 + }, + { + "epoch": 2.876084128360798, + "grad_norm": 1.7345436811447144, + "learning_rate": 8.291019875235746e-07, + "loss": 0.0051, + "step": 26529 + }, + { + "epoch": 2.8761925411968776, + "grad_norm": 0.09356959909200668, + "learning_rate": 8.283766139561875e-07, + "loss": 0.0008, + "step": 26530 + }, + { + "epoch": 2.8763009540329576, + "grad_norm": 14.930047988891602, + "learning_rate": 8.276512403888003e-07, + "loss": 0.0094, + "step": 26531 + }, + { + "epoch": 2.876409366869037, + "grad_norm": 0.02738683857023716, + "learning_rate": 8.26925866821413e-07, + "loss": 0.0003, + "step": 26532 + }, + { + "epoch": 2.876517779705117, + "grad_norm": 12.219181060791016, + "learning_rate": 8.262004932540258e-07, + "loss": 0.0169, + "step": 26533 + }, + { + "epoch": 2.876626192541197, + "grad_norm": 0.040720123797655106, + "learning_rate": 8.254751196866387e-07, + "loss": 0.0003, + "step": 26534 + }, + { + "epoch": 2.8767346053772767, + "grad_norm": 1.1707261800765991, + "learning_rate": 8.247497461192515e-07, + "loss": 0.018, + "step": 26535 + }, + { + "epoch": 2.876843018213356, + "grad_norm": 2.6267497539520264, + "learning_rate": 8.240243725518643e-07, + "loss": 0.0074, + "step": 26536 + }, + { + "epoch": 2.876951431049436, + "grad_norm": 2.7576253414154053, + "learning_rate": 8.23298998984477e-07, + "loss": 0.0358, + "step": 26537 + }, + { + "epoch": 2.877059843885516, + "grad_norm": 2.8112854957580566, + "learning_rate": 8.225736254170899e-07, + "loss": 0.0037, + "step": 26538 + }, + { + "epoch": 2.8771682567215957, + "grad_norm": 1.4306586980819702, + "learning_rate": 8.218482518497027e-07, + "loss": 0.0095, + "step": 26539 + }, + { + "epoch": 2.8772766695576757, + "grad_norm": 0.3982287049293518, + "learning_rate": 8.211228782823155e-07, + "loss": 0.0034, + "step": 26540 + }, + { + "epoch": 2.8773850823937552, + "grad_norm": 0.2310444414615631, + "learning_rate": 8.203975047149283e-07, + "loss": 0.0009, + "step": 26541 + }, + { + "epoch": 2.877493495229835, + "grad_norm": 1.0304070711135864, + "learning_rate": 8.196721311475409e-07, + "loss": 0.0021, + "step": 26542 + }, + { + "epoch": 2.877601908065915, + "grad_norm": 2.0453054904937744, + "learning_rate": 8.189467575801538e-07, + "loss": 0.0202, + "step": 26543 + }, + { + "epoch": 2.8777103209019947, + "grad_norm": 0.134065642952919, + "learning_rate": 8.182213840127666e-07, + "loss": 0.0007, + "step": 26544 + }, + { + "epoch": 2.8778187337380747, + "grad_norm": 0.3326126039028168, + "learning_rate": 8.174960104453794e-07, + "loss": 0.0014, + "step": 26545 + }, + { + "epoch": 2.8779271465741543, + "grad_norm": 3.5109968185424805, + "learning_rate": 8.167706368779922e-07, + "loss": 0.0235, + "step": 26546 + }, + { + "epoch": 2.8780355594102343, + "grad_norm": 0.6043915748596191, + "learning_rate": 8.16045263310605e-07, + "loss": 0.0401, + "step": 26547 + }, + { + "epoch": 2.878143972246314, + "grad_norm": 0.12046290934085846, + "learning_rate": 8.153198897432178e-07, + "loss": 0.0006, + "step": 26548 + }, + { + "epoch": 2.878252385082394, + "grad_norm": 1.1898281574249268, + "learning_rate": 8.145945161758306e-07, + "loss": 0.0026, + "step": 26549 + }, + { + "epoch": 2.8783607979184733, + "grad_norm": 0.029352055862545967, + "learning_rate": 8.138691426084434e-07, + "loss": 0.0004, + "step": 26550 + }, + { + "epoch": 2.8784692107545533, + "grad_norm": 1.9128878116607666, + "learning_rate": 8.131437690410563e-07, + "loss": 0.0082, + "step": 26551 + }, + { + "epoch": 2.8785776235906333, + "grad_norm": 0.014071131125092506, + "learning_rate": 8.12418395473669e-07, + "loss": 0.0002, + "step": 26552 + }, + { + "epoch": 2.878686036426713, + "grad_norm": 0.623175323009491, + "learning_rate": 8.116930219062818e-07, + "loss": 0.0008, + "step": 26553 + }, + { + "epoch": 2.878794449262793, + "grad_norm": 2.9300825595855713, + "learning_rate": 8.109676483388946e-07, + "loss": 0.0653, + "step": 26554 + }, + { + "epoch": 2.8789028620988724, + "grad_norm": 2.9608078002929688, + "learning_rate": 8.102422747715073e-07, + "loss": 0.0149, + "step": 26555 + }, + { + "epoch": 2.8790112749349523, + "grad_norm": 8.864386558532715, + "learning_rate": 8.095169012041202e-07, + "loss": 0.0486, + "step": 26556 + }, + { + "epoch": 2.8791196877710323, + "grad_norm": 0.08394281566143036, + "learning_rate": 8.087915276367329e-07, + "loss": 0.0007, + "step": 26557 + }, + { + "epoch": 2.879228100607112, + "grad_norm": 0.16274897754192352, + "learning_rate": 8.080661540693457e-07, + "loss": 0.0004, + "step": 26558 + }, + { + "epoch": 2.8793365134431914, + "grad_norm": 0.5172501802444458, + "learning_rate": 8.073407805019585e-07, + "loss": 0.0023, + "step": 26559 + }, + { + "epoch": 2.8794449262792714, + "grad_norm": 6.165109157562256, + "learning_rate": 8.066154069345714e-07, + "loss": 0.0079, + "step": 26560 + }, + { + "epoch": 2.8795533391153514, + "grad_norm": 2.4201643466949463, + "learning_rate": 8.058900333671842e-07, + "loss": 0.0179, + "step": 26561 + }, + { + "epoch": 2.879661751951431, + "grad_norm": 0.28368595242500305, + "learning_rate": 8.051646597997969e-07, + "loss": 0.0011, + "step": 26562 + }, + { + "epoch": 2.879770164787511, + "grad_norm": 0.047824397683143616, + "learning_rate": 8.044392862324097e-07, + "loss": 0.0004, + "step": 26563 + }, + { + "epoch": 2.8798785776235905, + "grad_norm": 7.478035926818848, + "learning_rate": 8.037139126650226e-07, + "loss": 0.0303, + "step": 26564 + }, + { + "epoch": 2.8799869904596704, + "grad_norm": 2.872392177581787, + "learning_rate": 8.029885390976354e-07, + "loss": 0.0258, + "step": 26565 + }, + { + "epoch": 2.8800954032957504, + "grad_norm": 1.8213294744491577, + "learning_rate": 8.022631655302482e-07, + "loss": 0.002, + "step": 26566 + }, + { + "epoch": 2.88020381613183, + "grad_norm": 0.2011466771364212, + "learning_rate": 8.01537791962861e-07, + "loss": 0.0012, + "step": 26567 + }, + { + "epoch": 2.88031222896791, + "grad_norm": 1.4289618730545044, + "learning_rate": 8.008124183954738e-07, + "loss": 0.0048, + "step": 26568 + }, + { + "epoch": 2.8804206418039895, + "grad_norm": 1.773483395576477, + "learning_rate": 8.000870448280865e-07, + "loss": 0.0099, + "step": 26569 + }, + { + "epoch": 2.8805290546400695, + "grad_norm": 3.4859259128570557, + "learning_rate": 7.993616712606993e-07, + "loss": 0.0447, + "step": 26570 + }, + { + "epoch": 2.880637467476149, + "grad_norm": 1.9938671588897705, + "learning_rate": 7.986362976933121e-07, + "loss": 0.0203, + "step": 26571 + }, + { + "epoch": 2.880745880312229, + "grad_norm": 5.502124309539795, + "learning_rate": 7.979109241259248e-07, + "loss": 0.0118, + "step": 26572 + }, + { + "epoch": 2.8808542931483085, + "grad_norm": 1.7376545667648315, + "learning_rate": 7.971855505585377e-07, + "loss": 0.0062, + "step": 26573 + }, + { + "epoch": 2.8809627059843885, + "grad_norm": 0.3519327640533447, + "learning_rate": 7.964601769911505e-07, + "loss": 0.001, + "step": 26574 + }, + { + "epoch": 2.8810711188204685, + "grad_norm": 1.1166636943817139, + "learning_rate": 7.957348034237633e-07, + "loss": 0.0037, + "step": 26575 + }, + { + "epoch": 2.881179531656548, + "grad_norm": 0.04062279686331749, + "learning_rate": 7.950094298563762e-07, + "loss": 0.0004, + "step": 26576 + }, + { + "epoch": 2.881287944492628, + "grad_norm": 0.14153291285037994, + "learning_rate": 7.942840562889889e-07, + "loss": 0.0009, + "step": 26577 + }, + { + "epoch": 2.8813963573287076, + "grad_norm": 0.05018132925033569, + "learning_rate": 7.935586827216017e-07, + "loss": 0.0006, + "step": 26578 + }, + { + "epoch": 2.8815047701647876, + "grad_norm": 3.6878604888916016, + "learning_rate": 7.928333091542145e-07, + "loss": 0.028, + "step": 26579 + }, + { + "epoch": 2.8816131830008676, + "grad_norm": 0.05229448527097702, + "learning_rate": 7.921079355868274e-07, + "loss": 0.0005, + "step": 26580 + }, + { + "epoch": 2.881721595836947, + "grad_norm": 4.423407077789307, + "learning_rate": 7.913825620194402e-07, + "loss": 0.0084, + "step": 26581 + }, + { + "epoch": 2.8818300086730266, + "grad_norm": 2.909520387649536, + "learning_rate": 7.906571884520528e-07, + "loss": 0.0053, + "step": 26582 + }, + { + "epoch": 2.8819384215091066, + "grad_norm": 4.3845930099487305, + "learning_rate": 7.899318148846656e-07, + "loss": 0.0134, + "step": 26583 + }, + { + "epoch": 2.8820468343451866, + "grad_norm": 1.3488574028015137, + "learning_rate": 7.892064413172784e-07, + "loss": 0.0013, + "step": 26584 + }, + { + "epoch": 2.882155247181266, + "grad_norm": 6.337797164916992, + "learning_rate": 7.884810677498912e-07, + "loss": 0.0244, + "step": 26585 + }, + { + "epoch": 2.882263660017346, + "grad_norm": 2.600351333618164, + "learning_rate": 7.877556941825041e-07, + "loss": 0.0083, + "step": 26586 + }, + { + "epoch": 2.8823720728534257, + "grad_norm": 0.040067460387945175, + "learning_rate": 7.870303206151168e-07, + "loss": 0.0005, + "step": 26587 + }, + { + "epoch": 2.8824804856895057, + "grad_norm": 0.08447299897670746, + "learning_rate": 7.863049470477296e-07, + "loss": 0.0006, + "step": 26588 + }, + { + "epoch": 2.8825888985255856, + "grad_norm": 0.3646599352359772, + "learning_rate": 7.855795734803425e-07, + "loss": 0.001, + "step": 26589 + }, + { + "epoch": 2.882697311361665, + "grad_norm": 0.26255932450294495, + "learning_rate": 7.848541999129553e-07, + "loss": 0.0015, + "step": 26590 + }, + { + "epoch": 2.882805724197745, + "grad_norm": 0.27018189430236816, + "learning_rate": 7.84128826345568e-07, + "loss": 0.0023, + "step": 26591 + }, + { + "epoch": 2.8829141370338247, + "grad_norm": 1.4734275341033936, + "learning_rate": 7.834034527781808e-07, + "loss": 0.0059, + "step": 26592 + }, + { + "epoch": 2.8830225498699047, + "grad_norm": 0.4925231337547302, + "learning_rate": 7.826780792107937e-07, + "loss": 0.0007, + "step": 26593 + }, + { + "epoch": 2.8831309627059842, + "grad_norm": 0.01921992003917694, + "learning_rate": 7.819527056434065e-07, + "loss": 0.0002, + "step": 26594 + }, + { + "epoch": 2.883239375542064, + "grad_norm": 4.008617877960205, + "learning_rate": 7.812273320760193e-07, + "loss": 0.0504, + "step": 26595 + }, + { + "epoch": 2.8833477883781438, + "grad_norm": 6.968096733093262, + "learning_rate": 7.805019585086319e-07, + "loss": 0.0292, + "step": 26596 + }, + { + "epoch": 2.8834562012142237, + "grad_norm": 0.10509148240089417, + "learning_rate": 7.797765849412447e-07, + "loss": 0.0005, + "step": 26597 + }, + { + "epoch": 2.8835646140503037, + "grad_norm": 1.8223203420639038, + "learning_rate": 7.790512113738576e-07, + "loss": 0.0053, + "step": 26598 + }, + { + "epoch": 2.8836730268863833, + "grad_norm": 0.2711750864982605, + "learning_rate": 7.783258378064704e-07, + "loss": 0.002, + "step": 26599 + }, + { + "epoch": 2.8837814397224633, + "grad_norm": 0.15759244561195374, + "learning_rate": 7.776004642390832e-07, + "loss": 0.0004, + "step": 26600 + }, + { + "epoch": 2.883889852558543, + "grad_norm": 0.9593262076377869, + "learning_rate": 7.768750906716959e-07, + "loss": 0.0017, + "step": 26601 + }, + { + "epoch": 2.883998265394623, + "grad_norm": 0.06311024725437164, + "learning_rate": 7.761497171043088e-07, + "loss": 0.0004, + "step": 26602 + }, + { + "epoch": 2.8841066782307028, + "grad_norm": 0.026839077472686768, + "learning_rate": 7.754243435369216e-07, + "loss": 0.0003, + "step": 26603 + }, + { + "epoch": 2.8842150910667823, + "grad_norm": 0.03714523836970329, + "learning_rate": 7.746989699695344e-07, + "loss": 0.0002, + "step": 26604 + }, + { + "epoch": 2.884323503902862, + "grad_norm": 2.1266605854034424, + "learning_rate": 7.739735964021472e-07, + "loss": 0.025, + "step": 26605 + }, + { + "epoch": 2.884431916738942, + "grad_norm": 0.8356961607933044, + "learning_rate": 7.7324822283476e-07, + "loss": 0.0024, + "step": 26606 + }, + { + "epoch": 2.884540329575022, + "grad_norm": 6.143948078155518, + "learning_rate": 7.725228492673728e-07, + "loss": 0.019, + "step": 26607 + }, + { + "epoch": 2.8846487424111014, + "grad_norm": 0.1322392225265503, + "learning_rate": 7.717974756999856e-07, + "loss": 0.0009, + "step": 26608 + }, + { + "epoch": 2.8847571552471813, + "grad_norm": 4.724921226501465, + "learning_rate": 7.710721021325983e-07, + "loss": 0.0294, + "step": 26609 + }, + { + "epoch": 2.884865568083261, + "grad_norm": 1.839795708656311, + "learning_rate": 7.703467285652111e-07, + "loss": 0.0066, + "step": 26610 + }, + { + "epoch": 2.884973980919341, + "grad_norm": 0.11673883348703384, + "learning_rate": 7.696213549978239e-07, + "loss": 0.0009, + "step": 26611 + }, + { + "epoch": 2.885082393755421, + "grad_norm": 0.27452170848846436, + "learning_rate": 7.688959814304367e-07, + "loss": 0.0013, + "step": 26612 + }, + { + "epoch": 2.8851908065915004, + "grad_norm": 0.1933584064245224, + "learning_rate": 7.681706078630495e-07, + "loss": 0.0011, + "step": 26613 + }, + { + "epoch": 2.8852992194275804, + "grad_norm": 0.13220247626304626, + "learning_rate": 7.674452342956623e-07, + "loss": 0.0007, + "step": 26614 + }, + { + "epoch": 2.88540763226366, + "grad_norm": 0.6293020844459534, + "learning_rate": 7.667198607282752e-07, + "loss": 0.0024, + "step": 26615 + }, + { + "epoch": 2.88551604509974, + "grad_norm": 0.1142512783408165, + "learning_rate": 7.659944871608879e-07, + "loss": 0.0005, + "step": 26616 + }, + { + "epoch": 2.8856244579358195, + "grad_norm": 0.030331473797559738, + "learning_rate": 7.652691135935007e-07, + "loss": 0.0003, + "step": 26617 + }, + { + "epoch": 2.8857328707718994, + "grad_norm": 0.09120509028434753, + "learning_rate": 7.645437400261135e-07, + "loss": 0.0006, + "step": 26618 + }, + { + "epoch": 2.885841283607979, + "grad_norm": 2.9351086616516113, + "learning_rate": 7.638183664587264e-07, + "loss": 0.0019, + "step": 26619 + }, + { + "epoch": 2.885949696444059, + "grad_norm": 0.02663334272801876, + "learning_rate": 7.630929928913392e-07, + "loss": 0.0002, + "step": 26620 + }, + { + "epoch": 2.886058109280139, + "grad_norm": 5.358060836791992, + "learning_rate": 7.623676193239519e-07, + "loss": 0.0196, + "step": 26621 + }, + { + "epoch": 2.8861665221162185, + "grad_norm": 4.4967474937438965, + "learning_rate": 7.616422457565647e-07, + "loss": 0.0367, + "step": 26622 + }, + { + "epoch": 2.8862749349522985, + "grad_norm": 0.08314495533704758, + "learning_rate": 7.609168721891774e-07, + "loss": 0.0005, + "step": 26623 + }, + { + "epoch": 2.886383347788378, + "grad_norm": 0.039423707872629166, + "learning_rate": 7.601914986217903e-07, + "loss": 0.0002, + "step": 26624 + }, + { + "epoch": 2.886491760624458, + "grad_norm": 0.040546517819166183, + "learning_rate": 7.594661250544031e-07, + "loss": 0.0003, + "step": 26625 + }, + { + "epoch": 2.886600173460538, + "grad_norm": 1.4722918272018433, + "learning_rate": 7.587407514870158e-07, + "loss": 0.0025, + "step": 26626 + }, + { + "epoch": 2.8867085862966175, + "grad_norm": 2.2374391555786133, + "learning_rate": 7.580153779196286e-07, + "loss": 0.0028, + "step": 26627 + }, + { + "epoch": 2.886816999132697, + "grad_norm": 0.5985956192016602, + "learning_rate": 7.572900043522415e-07, + "loss": 0.0021, + "step": 26628 + }, + { + "epoch": 2.886925411968777, + "grad_norm": 1.1486570835113525, + "learning_rate": 7.565646307848543e-07, + "loss": 0.0038, + "step": 26629 + }, + { + "epoch": 2.887033824804857, + "grad_norm": 2.6106255054473877, + "learning_rate": 7.558392572174671e-07, + "loss": 0.0052, + "step": 26630 + }, + { + "epoch": 2.8871422376409366, + "grad_norm": 3.127779483795166, + "learning_rate": 7.551138836500798e-07, + "loss": 0.0521, + "step": 26631 + }, + { + "epoch": 2.8872506504770166, + "grad_norm": 1.1723394393920898, + "learning_rate": 7.543885100826927e-07, + "loss": 0.0007, + "step": 26632 + }, + { + "epoch": 2.887359063313096, + "grad_norm": 8.100726127624512, + "learning_rate": 7.536631365153055e-07, + "loss": 0.0138, + "step": 26633 + }, + { + "epoch": 2.887467476149176, + "grad_norm": 0.08633379638195038, + "learning_rate": 7.529377629479183e-07, + "loss": 0.0006, + "step": 26634 + }, + { + "epoch": 2.887575888985256, + "grad_norm": 0.14629186689853668, + "learning_rate": 7.522123893805311e-07, + "loss": 0.0009, + "step": 26635 + }, + { + "epoch": 2.8876843018213356, + "grad_norm": 2.5574114322662354, + "learning_rate": 7.514870158131439e-07, + "loss": 0.002, + "step": 26636 + }, + { + "epoch": 2.887792714657415, + "grad_norm": 0.13795077800750732, + "learning_rate": 7.507616422457566e-07, + "loss": 0.0005, + "step": 26637 + }, + { + "epoch": 2.887901127493495, + "grad_norm": 0.2825482487678528, + "learning_rate": 7.500362686783694e-07, + "loss": 0.0023, + "step": 26638 + }, + { + "epoch": 2.888009540329575, + "grad_norm": 2.207482099533081, + "learning_rate": 7.493108951109822e-07, + "loss": 0.007, + "step": 26639 + }, + { + "epoch": 2.8881179531656547, + "grad_norm": 0.18461263179779053, + "learning_rate": 7.48585521543595e-07, + "loss": 0.0007, + "step": 26640 + }, + { + "epoch": 2.8882263660017347, + "grad_norm": 1.2268832921981812, + "learning_rate": 7.478601479762078e-07, + "loss": 0.0017, + "step": 26641 + }, + { + "epoch": 2.888334778837814, + "grad_norm": 2.514202356338501, + "learning_rate": 7.471347744088206e-07, + "loss": 0.0184, + "step": 26642 + }, + { + "epoch": 2.888443191673894, + "grad_norm": 5.097575664520264, + "learning_rate": 7.464094008414334e-07, + "loss": 0.0082, + "step": 26643 + }, + { + "epoch": 2.888551604509974, + "grad_norm": 0.49270209670066833, + "learning_rate": 7.456840272740462e-07, + "loss": 0.0027, + "step": 26644 + }, + { + "epoch": 2.8886600173460537, + "grad_norm": 0.15699031949043274, + "learning_rate": 7.449586537066591e-07, + "loss": 0.001, + "step": 26645 + }, + { + "epoch": 2.8887684301821337, + "grad_norm": 1.4569123983383179, + "learning_rate": 7.442332801392718e-07, + "loss": 0.0019, + "step": 26646 + }, + { + "epoch": 2.8888768430182132, + "grad_norm": 2.442491292953491, + "learning_rate": 7.435079065718846e-07, + "loss": 0.0077, + "step": 26647 + }, + { + "epoch": 2.888985255854293, + "grad_norm": 0.25278475880622864, + "learning_rate": 7.427825330044974e-07, + "loss": 0.002, + "step": 26648 + }, + { + "epoch": 2.889093668690373, + "grad_norm": 0.9481766223907471, + "learning_rate": 7.420571594371103e-07, + "loss": 0.004, + "step": 26649 + }, + { + "epoch": 2.8892020815264527, + "grad_norm": 0.013811242766678333, + "learning_rate": 7.413317858697229e-07, + "loss": 0.0002, + "step": 26650 + }, + { + "epoch": 2.8893104943625323, + "grad_norm": 1.1567420959472656, + "learning_rate": 7.406064123023357e-07, + "loss": 0.002, + "step": 26651 + }, + { + "epoch": 2.8894189071986123, + "grad_norm": 3.076026439666748, + "learning_rate": 7.398810387349485e-07, + "loss": 0.0207, + "step": 26652 + }, + { + "epoch": 2.8895273200346923, + "grad_norm": 0.2834305167198181, + "learning_rate": 7.391556651675613e-07, + "loss": 0.0018, + "step": 26653 + }, + { + "epoch": 2.889635732870772, + "grad_norm": 3.427888870239258, + "learning_rate": 7.384302916001742e-07, + "loss": 0.0063, + "step": 26654 + }, + { + "epoch": 2.889744145706852, + "grad_norm": 0.08566103875637054, + "learning_rate": 7.377049180327869e-07, + "loss": 0.0007, + "step": 26655 + }, + { + "epoch": 2.8898525585429313, + "grad_norm": 5.7867207527160645, + "learning_rate": 7.369795444653997e-07, + "loss": 0.0198, + "step": 26656 + }, + { + "epoch": 2.8899609713790113, + "grad_norm": 0.022674771025776863, + "learning_rate": 7.362541708980125e-07, + "loss": 0.0002, + "step": 26657 + }, + { + "epoch": 2.8900693842150913, + "grad_norm": 0.4021631181240082, + "learning_rate": 7.355287973306254e-07, + "loss": 0.0027, + "step": 26658 + }, + { + "epoch": 2.890177797051171, + "grad_norm": 7.293490409851074, + "learning_rate": 7.348034237632382e-07, + "loss": 0.0215, + "step": 26659 + }, + { + "epoch": 2.8902862098872504, + "grad_norm": 0.008161524310708046, + "learning_rate": 7.340780501958509e-07, + "loss": 0.0001, + "step": 26660 + }, + { + "epoch": 2.8903946227233304, + "grad_norm": 0.07478100061416626, + "learning_rate": 7.333526766284637e-07, + "loss": 0.0004, + "step": 26661 + }, + { + "epoch": 2.8905030355594103, + "grad_norm": 0.11543452739715576, + "learning_rate": 7.326273030610766e-07, + "loss": 0.0002, + "step": 26662 + }, + { + "epoch": 2.89061144839549, + "grad_norm": 8.631308555603027, + "learning_rate": 7.319019294936894e-07, + "loss": 0.0489, + "step": 26663 + }, + { + "epoch": 2.89071986123157, + "grad_norm": 1.1596883535385132, + "learning_rate": 7.311765559263021e-07, + "loss": 0.0038, + "step": 26664 + }, + { + "epoch": 2.8908282740676494, + "grad_norm": 3.2482924461364746, + "learning_rate": 7.304511823589148e-07, + "loss": 0.032, + "step": 26665 + }, + { + "epoch": 2.8909366869037294, + "grad_norm": 3.5877249240875244, + "learning_rate": 7.297258087915276e-07, + "loss": 0.0283, + "step": 26666 + }, + { + "epoch": 2.8910450997398094, + "grad_norm": 2.4239532947540283, + "learning_rate": 7.290004352241405e-07, + "loss": 0.006, + "step": 26667 + }, + { + "epoch": 2.891153512575889, + "grad_norm": 0.20272988080978394, + "learning_rate": 7.282750616567533e-07, + "loss": 0.0004, + "step": 26668 + }, + { + "epoch": 2.891261925411969, + "grad_norm": 0.024233438074588776, + "learning_rate": 7.275496880893661e-07, + "loss": 0.0003, + "step": 26669 + }, + { + "epoch": 2.8913703382480485, + "grad_norm": 0.06386455148458481, + "learning_rate": 7.268243145219788e-07, + "loss": 0.0006, + "step": 26670 + }, + { + "epoch": 2.8914787510841284, + "grad_norm": 0.049441754817962646, + "learning_rate": 7.260989409545917e-07, + "loss": 0.0006, + "step": 26671 + }, + { + "epoch": 2.8915871639202084, + "grad_norm": 0.6409540176391602, + "learning_rate": 7.253735673872045e-07, + "loss": 0.0058, + "step": 26672 + }, + { + "epoch": 2.891695576756288, + "grad_norm": 4.276674747467041, + "learning_rate": 7.246481938198173e-07, + "loss": 0.0075, + "step": 26673 + }, + { + "epoch": 2.8918039895923675, + "grad_norm": 0.29249027371406555, + "learning_rate": 7.239228202524301e-07, + "loss": 0.0017, + "step": 26674 + }, + { + "epoch": 2.8919124024284475, + "grad_norm": 0.9951590299606323, + "learning_rate": 7.231974466850429e-07, + "loss": 0.0015, + "step": 26675 + }, + { + "epoch": 2.8920208152645275, + "grad_norm": 0.1846672147512436, + "learning_rate": 7.224720731176557e-07, + "loss": 0.0011, + "step": 26676 + }, + { + "epoch": 2.892129228100607, + "grad_norm": 0.08506114035844803, + "learning_rate": 7.217466995502684e-07, + "loss": 0.0006, + "step": 26677 + }, + { + "epoch": 2.892237640936687, + "grad_norm": 0.39147496223449707, + "learning_rate": 7.210213259828812e-07, + "loss": 0.0012, + "step": 26678 + }, + { + "epoch": 2.8923460537727665, + "grad_norm": 2.23957896232605, + "learning_rate": 7.20295952415494e-07, + "loss": 0.0055, + "step": 26679 + }, + { + "epoch": 2.8924544666088465, + "grad_norm": 3.4391372203826904, + "learning_rate": 7.195705788481068e-07, + "loss": 0.0052, + "step": 26680 + }, + { + "epoch": 2.8925628794449265, + "grad_norm": 0.03624405711889267, + "learning_rate": 7.188452052807196e-07, + "loss": 0.0004, + "step": 26681 + }, + { + "epoch": 2.892671292281006, + "grad_norm": 2.8562679290771484, + "learning_rate": 7.181198317133324e-07, + "loss": 0.0261, + "step": 26682 + }, + { + "epoch": 2.8927797051170856, + "grad_norm": 0.010746016167104244, + "learning_rate": 7.173944581459452e-07, + "loss": 0.0001, + "step": 26683 + }, + { + "epoch": 2.8928881179531656, + "grad_norm": 9.772360801696777, + "learning_rate": 7.166690845785581e-07, + "loss": 0.0073, + "step": 26684 + }, + { + "epoch": 2.8929965307892456, + "grad_norm": 3.4693052768707275, + "learning_rate": 7.159437110111708e-07, + "loss": 0.0197, + "step": 26685 + }, + { + "epoch": 2.893104943625325, + "grad_norm": 0.010820978321135044, + "learning_rate": 7.152183374437836e-07, + "loss": 0.0001, + "step": 26686 + }, + { + "epoch": 2.893213356461405, + "grad_norm": 10.708521842956543, + "learning_rate": 7.144929638763964e-07, + "loss": 0.0121, + "step": 26687 + }, + { + "epoch": 2.8933217692974846, + "grad_norm": 0.24228064715862274, + "learning_rate": 7.137675903090093e-07, + "loss": 0.0011, + "step": 26688 + }, + { + "epoch": 2.8934301821335646, + "grad_norm": 2.205662488937378, + "learning_rate": 7.130422167416221e-07, + "loss": 0.0149, + "step": 26689 + }, + { + "epoch": 2.8935385949696446, + "grad_norm": 0.014877332374453545, + "learning_rate": 7.123168431742348e-07, + "loss": 0.0002, + "step": 26690 + }, + { + "epoch": 2.893647007805724, + "grad_norm": 3.4694294929504395, + "learning_rate": 7.115914696068475e-07, + "loss": 0.0247, + "step": 26691 + }, + { + "epoch": 2.893755420641804, + "grad_norm": 18.09864616394043, + "learning_rate": 7.108660960394603e-07, + "loss": 0.0217, + "step": 26692 + }, + { + "epoch": 2.8938638334778837, + "grad_norm": 1.4255926609039307, + "learning_rate": 7.101407224720732e-07, + "loss": 0.0031, + "step": 26693 + }, + { + "epoch": 2.8939722463139637, + "grad_norm": 0.05530387908220291, + "learning_rate": 7.09415348904686e-07, + "loss": 0.0003, + "step": 26694 + }, + { + "epoch": 2.894080659150043, + "grad_norm": 2.6939585208892822, + "learning_rate": 7.086899753372987e-07, + "loss": 0.0164, + "step": 26695 + }, + { + "epoch": 2.894189071986123, + "grad_norm": 2.5948376655578613, + "learning_rate": 7.079646017699115e-07, + "loss": 0.0066, + "step": 26696 + }, + { + "epoch": 2.8942974848222027, + "grad_norm": 2.060793161392212, + "learning_rate": 7.072392282025244e-07, + "loss": 0.0337, + "step": 26697 + }, + { + "epoch": 2.8944058976582827, + "grad_norm": 0.02082987129688263, + "learning_rate": 7.065138546351372e-07, + "loss": 0.0003, + "step": 26698 + }, + { + "epoch": 2.8945143104943627, + "grad_norm": 2.6682612895965576, + "learning_rate": 7.0578848106775e-07, + "loss": 0.0244, + "step": 26699 + }, + { + "epoch": 2.8946227233304422, + "grad_norm": 0.046332091093063354, + "learning_rate": 7.050631075003627e-07, + "loss": 0.0004, + "step": 26700 + }, + { + "epoch": 2.8947311361665222, + "grad_norm": 14.053250312805176, + "learning_rate": 7.043377339329756e-07, + "loss": 0.0093, + "step": 26701 + }, + { + "epoch": 2.8948395490026018, + "grad_norm": 0.08833445608615875, + "learning_rate": 7.036123603655884e-07, + "loss": 0.0002, + "step": 26702 + }, + { + "epoch": 2.8949479618386817, + "grad_norm": 0.146876722574234, + "learning_rate": 7.028869867982012e-07, + "loss": 0.001, + "step": 26703 + }, + { + "epoch": 2.8950563746747617, + "grad_norm": 0.004140613600611687, + "learning_rate": 7.021616132308138e-07, + "loss": 0.0, + "step": 26704 + }, + { + "epoch": 2.8951647875108413, + "grad_norm": 9.496896743774414, + "learning_rate": 7.014362396634266e-07, + "loss": 0.0274, + "step": 26705 + }, + { + "epoch": 2.895273200346921, + "grad_norm": 0.13116006553173065, + "learning_rate": 7.007108660960395e-07, + "loss": 0.001, + "step": 26706 + }, + { + "epoch": 2.895381613183001, + "grad_norm": 4.418795108795166, + "learning_rate": 6.999854925286523e-07, + "loss": 0.0196, + "step": 26707 + }, + { + "epoch": 2.895490026019081, + "grad_norm": 0.1587880700826645, + "learning_rate": 6.992601189612651e-07, + "loss": 0.0013, + "step": 26708 + }, + { + "epoch": 2.8955984388551603, + "grad_norm": 0.0374373234808445, + "learning_rate": 6.985347453938778e-07, + "loss": 0.0003, + "step": 26709 + }, + { + "epoch": 2.8957068516912403, + "grad_norm": 0.060466788709163666, + "learning_rate": 6.978093718264907e-07, + "loss": 0.0005, + "step": 26710 + }, + { + "epoch": 2.89581526452732, + "grad_norm": 0.06448277831077576, + "learning_rate": 6.970839982591035e-07, + "loss": 0.0004, + "step": 26711 + }, + { + "epoch": 2.8959236773634, + "grad_norm": 3.0062506198883057, + "learning_rate": 6.963586246917163e-07, + "loss": 0.0389, + "step": 26712 + }, + { + "epoch": 2.89603209019948, + "grad_norm": 0.023716533556580544, + "learning_rate": 6.956332511243291e-07, + "loss": 0.0001, + "step": 26713 + }, + { + "epoch": 2.8961405030355594, + "grad_norm": 0.07530722767114639, + "learning_rate": 6.949078775569419e-07, + "loss": 0.0002, + "step": 26714 + }, + { + "epoch": 2.8962489158716394, + "grad_norm": 0.08723174035549164, + "learning_rate": 6.941825039895547e-07, + "loss": 0.0004, + "step": 26715 + }, + { + "epoch": 2.896357328707719, + "grad_norm": 0.009876027703285217, + "learning_rate": 6.934571304221675e-07, + "loss": 0.0001, + "step": 26716 + }, + { + "epoch": 2.896465741543799, + "grad_norm": 13.547168731689453, + "learning_rate": 6.927317568547804e-07, + "loss": 0.0411, + "step": 26717 + }, + { + "epoch": 2.8965741543798784, + "grad_norm": 0.2131434977054596, + "learning_rate": 6.92006383287393e-07, + "loss": 0.0008, + "step": 26718 + }, + { + "epoch": 2.8966825672159584, + "grad_norm": 0.05202654004096985, + "learning_rate": 6.912810097200058e-07, + "loss": 0.0004, + "step": 26719 + }, + { + "epoch": 2.896790980052038, + "grad_norm": 1.8233624696731567, + "learning_rate": 6.905556361526186e-07, + "loss": 0.0035, + "step": 26720 + }, + { + "epoch": 2.896899392888118, + "grad_norm": 2.3269472122192383, + "learning_rate": 6.898302625852314e-07, + "loss": 0.0049, + "step": 26721 + }, + { + "epoch": 2.897007805724198, + "grad_norm": 0.5976357460021973, + "learning_rate": 6.891048890178442e-07, + "loss": 0.002, + "step": 26722 + }, + { + "epoch": 2.8971162185602775, + "grad_norm": 0.08572588860988617, + "learning_rate": 6.883795154504571e-07, + "loss": 0.0006, + "step": 26723 + }, + { + "epoch": 2.8972246313963574, + "grad_norm": 1.3592607975006104, + "learning_rate": 6.876541418830698e-07, + "loss": 0.0029, + "step": 26724 + }, + { + "epoch": 2.897333044232437, + "grad_norm": 0.07649064809083939, + "learning_rate": 6.869287683156826e-07, + "loss": 0.0005, + "step": 26725 + }, + { + "epoch": 2.897441457068517, + "grad_norm": 0.27126234769821167, + "learning_rate": 6.862033947482954e-07, + "loss": 0.0022, + "step": 26726 + }, + { + "epoch": 2.897549869904597, + "grad_norm": 5.135932445526123, + "learning_rate": 6.854780211809083e-07, + "loss": 0.0189, + "step": 26727 + }, + { + "epoch": 2.8976582827406765, + "grad_norm": 1.6795690059661865, + "learning_rate": 6.847526476135211e-07, + "loss": 0.0413, + "step": 26728 + }, + { + "epoch": 2.897766695576756, + "grad_norm": 0.08056370168924332, + "learning_rate": 6.840272740461338e-07, + "loss": 0.0006, + "step": 26729 + }, + { + "epoch": 2.897875108412836, + "grad_norm": 0.022576013579964638, + "learning_rate": 6.833019004787467e-07, + "loss": 0.0002, + "step": 26730 + }, + { + "epoch": 2.897983521248916, + "grad_norm": 0.7064093947410583, + "learning_rate": 6.825765269113593e-07, + "loss": 0.003, + "step": 26731 + }, + { + "epoch": 2.8980919340849955, + "grad_norm": 0.482403039932251, + "learning_rate": 6.818511533439722e-07, + "loss": 0.0014, + "step": 26732 + }, + { + "epoch": 2.8982003469210755, + "grad_norm": 0.015187062323093414, + "learning_rate": 6.81125779776585e-07, + "loss": 0.0002, + "step": 26733 + }, + { + "epoch": 2.898308759757155, + "grad_norm": 0.07379856705665588, + "learning_rate": 6.804004062091977e-07, + "loss": 0.0005, + "step": 26734 + }, + { + "epoch": 2.898417172593235, + "grad_norm": 0.09827899187803268, + "learning_rate": 6.796750326418105e-07, + "loss": 0.0005, + "step": 26735 + }, + { + "epoch": 2.898525585429315, + "grad_norm": 6.912657260894775, + "learning_rate": 6.789496590744234e-07, + "loss": 0.0221, + "step": 26736 + }, + { + "epoch": 2.8986339982653946, + "grad_norm": 3.0690693855285645, + "learning_rate": 6.782242855070362e-07, + "loss": 0.0389, + "step": 26737 + }, + { + "epoch": 2.8987424111014746, + "grad_norm": 1.9650670289993286, + "learning_rate": 6.77498911939649e-07, + "loss": 0.028, + "step": 26738 + }, + { + "epoch": 2.898850823937554, + "grad_norm": 2.389157772064209, + "learning_rate": 6.767735383722618e-07, + "loss": 0.0112, + "step": 26739 + }, + { + "epoch": 2.898959236773634, + "grad_norm": 1.2812745571136475, + "learning_rate": 6.760481648048746e-07, + "loss": 0.0029, + "step": 26740 + }, + { + "epoch": 2.8990676496097136, + "grad_norm": 0.4060352146625519, + "learning_rate": 6.753227912374874e-07, + "loss": 0.0027, + "step": 26741 + }, + { + "epoch": 2.8991760624457936, + "grad_norm": 1.0616792440414429, + "learning_rate": 6.745974176701002e-07, + "loss": 0.0052, + "step": 26742 + }, + { + "epoch": 2.899284475281873, + "grad_norm": 3.8626651763916016, + "learning_rate": 6.738720441027131e-07, + "loss": 0.0057, + "step": 26743 + }, + { + "epoch": 2.899392888117953, + "grad_norm": 0.03916621580719948, + "learning_rate": 6.731466705353258e-07, + "loss": 0.0003, + "step": 26744 + }, + { + "epoch": 2.899501300954033, + "grad_norm": 0.7892897129058838, + "learning_rate": 6.724212969679385e-07, + "loss": 0.0046, + "step": 26745 + }, + { + "epoch": 2.8996097137901127, + "grad_norm": 0.05357372388243675, + "learning_rate": 6.716959234005513e-07, + "loss": 0.0003, + "step": 26746 + }, + { + "epoch": 2.8997181266261927, + "grad_norm": 0.020774655044078827, + "learning_rate": 6.709705498331641e-07, + "loss": 0.0002, + "step": 26747 + }, + { + "epoch": 2.899826539462272, + "grad_norm": 8.193653106689453, + "learning_rate": 6.70245176265777e-07, + "loss": 0.0458, + "step": 26748 + }, + { + "epoch": 2.899934952298352, + "grad_norm": 0.12243334949016571, + "learning_rate": 6.695198026983897e-07, + "loss": 0.0006, + "step": 26749 + }, + { + "epoch": 2.900043365134432, + "grad_norm": 1.175504207611084, + "learning_rate": 6.687944291310025e-07, + "loss": 0.0035, + "step": 26750 + }, + { + "epoch": 2.9001517779705117, + "grad_norm": 0.02450639382004738, + "learning_rate": 6.680690555636153e-07, + "loss": 0.0002, + "step": 26751 + }, + { + "epoch": 2.9002601908065913, + "grad_norm": 13.557029724121094, + "learning_rate": 6.673436819962282e-07, + "loss": 0.0139, + "step": 26752 + }, + { + "epoch": 2.9003686036426712, + "grad_norm": 0.2734416723251343, + "learning_rate": 6.66618308428841e-07, + "loss": 0.0005, + "step": 26753 + }, + { + "epoch": 2.9004770164787512, + "grad_norm": 0.18233869969844818, + "learning_rate": 6.658929348614537e-07, + "loss": 0.0006, + "step": 26754 + }, + { + "epoch": 2.9005854293148308, + "grad_norm": 0.027514567598700523, + "learning_rate": 6.651675612940665e-07, + "loss": 0.0002, + "step": 26755 + }, + { + "epoch": 2.9006938421509108, + "grad_norm": 2.064876079559326, + "learning_rate": 6.644421877266794e-07, + "loss": 0.0287, + "step": 26756 + }, + { + "epoch": 2.9008022549869903, + "grad_norm": 0.14167898893356323, + "learning_rate": 6.637168141592922e-07, + "loss": 0.0008, + "step": 26757 + }, + { + "epoch": 2.9009106678230703, + "grad_norm": 0.36405831575393677, + "learning_rate": 6.629914405919048e-07, + "loss": 0.0013, + "step": 26758 + }, + { + "epoch": 2.9010190806591503, + "grad_norm": 0.4729389250278473, + "learning_rate": 6.622660670245176e-07, + "loss": 0.001, + "step": 26759 + }, + { + "epoch": 2.90112749349523, + "grad_norm": 0.32054662704467773, + "learning_rate": 6.615406934571304e-07, + "loss": 0.0014, + "step": 26760 + }, + { + "epoch": 2.90123590633131, + "grad_norm": 0.08773749321699142, + "learning_rate": 6.608153198897433e-07, + "loss": 0.0005, + "step": 26761 + }, + { + "epoch": 2.9013443191673893, + "grad_norm": 14.975913047790527, + "learning_rate": 6.600899463223561e-07, + "loss": 0.0828, + "step": 26762 + }, + { + "epoch": 2.9014527320034693, + "grad_norm": 0.9784658551216125, + "learning_rate": 6.593645727549688e-07, + "loss": 0.0016, + "step": 26763 + }, + { + "epoch": 2.901561144839549, + "grad_norm": 0.1509406715631485, + "learning_rate": 6.586391991875816e-07, + "loss": 0.0014, + "step": 26764 + }, + { + "epoch": 2.901669557675629, + "grad_norm": 8.433503150939941, + "learning_rate": 6.579138256201945e-07, + "loss": 0.0184, + "step": 26765 + }, + { + "epoch": 2.9017779705117084, + "grad_norm": 2.485529661178589, + "learning_rate": 6.571884520528073e-07, + "loss": 0.0036, + "step": 26766 + }, + { + "epoch": 2.9018863833477884, + "grad_norm": 0.15611396729946136, + "learning_rate": 6.564630784854201e-07, + "loss": 0.0011, + "step": 26767 + }, + { + "epoch": 2.9019947961838684, + "grad_norm": 0.27824801206588745, + "learning_rate": 6.557377049180328e-07, + "loss": 0.0016, + "step": 26768 + }, + { + "epoch": 2.902103209019948, + "grad_norm": 0.44923165440559387, + "learning_rate": 6.550123313506457e-07, + "loss": 0.0019, + "step": 26769 + }, + { + "epoch": 2.902211621856028, + "grad_norm": 12.99138069152832, + "learning_rate": 6.542869577832585e-07, + "loss": 0.0536, + "step": 26770 + }, + { + "epoch": 2.9023200346921074, + "grad_norm": 0.997506320476532, + "learning_rate": 6.535615842158713e-07, + "loss": 0.0024, + "step": 26771 + }, + { + "epoch": 2.9024284475281874, + "grad_norm": 1.207179307937622, + "learning_rate": 6.52836210648484e-07, + "loss": 0.0071, + "step": 26772 + }, + { + "epoch": 2.9025368603642674, + "grad_norm": 4.772152423858643, + "learning_rate": 6.521108370810967e-07, + "loss": 0.0147, + "step": 26773 + }, + { + "epoch": 2.902645273200347, + "grad_norm": 0.061893854290246964, + "learning_rate": 6.513854635137096e-07, + "loss": 0.0005, + "step": 26774 + }, + { + "epoch": 2.9027536860364265, + "grad_norm": 2.7481696605682373, + "learning_rate": 6.506600899463224e-07, + "loss": 0.0041, + "step": 26775 + }, + { + "epoch": 2.9028620988725065, + "grad_norm": 1.3688080310821533, + "learning_rate": 6.499347163789352e-07, + "loss": 0.0222, + "step": 26776 + }, + { + "epoch": 2.9029705117085864, + "grad_norm": 1.2722922563552856, + "learning_rate": 6.49209342811548e-07, + "loss": 0.0184, + "step": 26777 + }, + { + "epoch": 2.903078924544666, + "grad_norm": 0.10830378532409668, + "learning_rate": 6.484839692441608e-07, + "loss": 0.0012, + "step": 26778 + }, + { + "epoch": 2.903187337380746, + "grad_norm": 1.2245879173278809, + "learning_rate": 6.477585956767736e-07, + "loss": 0.0039, + "step": 26779 + }, + { + "epoch": 2.9032957502168255, + "grad_norm": 0.07070425152778625, + "learning_rate": 6.470332221093864e-07, + "loss": 0.0004, + "step": 26780 + }, + { + "epoch": 2.9034041630529055, + "grad_norm": 4.918841361999512, + "learning_rate": 6.463078485419992e-07, + "loss": 0.0074, + "step": 26781 + }, + { + "epoch": 2.9035125758889855, + "grad_norm": 1.2979358434677124, + "learning_rate": 6.455824749746121e-07, + "loss": 0.0057, + "step": 26782 + }, + { + "epoch": 2.903620988725065, + "grad_norm": 0.0443132258951664, + "learning_rate": 6.448571014072248e-07, + "loss": 0.0004, + "step": 26783 + }, + { + "epoch": 2.9037294015611446, + "grad_norm": 0.018578339368104935, + "learning_rate": 6.441317278398376e-07, + "loss": 0.0001, + "step": 26784 + }, + { + "epoch": 2.9038378143972245, + "grad_norm": 0.15821336209774017, + "learning_rate": 6.434063542724503e-07, + "loss": 0.0013, + "step": 26785 + }, + { + "epoch": 2.9039462272333045, + "grad_norm": 10.45956039428711, + "learning_rate": 6.426809807050631e-07, + "loss": 0.0055, + "step": 26786 + }, + { + "epoch": 2.904054640069384, + "grad_norm": 0.3953110873699188, + "learning_rate": 6.41955607137676e-07, + "loss": 0.0018, + "step": 26787 + }, + { + "epoch": 2.904163052905464, + "grad_norm": 1.3090379238128662, + "learning_rate": 6.412302335702887e-07, + "loss": 0.0045, + "step": 26788 + }, + { + "epoch": 2.9042714657415436, + "grad_norm": 0.85032719373703, + "learning_rate": 6.405048600029015e-07, + "loss": 0.0031, + "step": 26789 + }, + { + "epoch": 2.9043798785776236, + "grad_norm": 1.5225907564163208, + "learning_rate": 6.397794864355143e-07, + "loss": 0.0258, + "step": 26790 + }, + { + "epoch": 2.9044882914137036, + "grad_norm": 3.4696552753448486, + "learning_rate": 6.390541128681272e-07, + "loss": 0.0992, + "step": 26791 + }, + { + "epoch": 2.904596704249783, + "grad_norm": 0.06861323118209839, + "learning_rate": 6.3832873930074e-07, + "loss": 0.0005, + "step": 26792 + }, + { + "epoch": 2.904705117085863, + "grad_norm": 2.5349926948547363, + "learning_rate": 6.376033657333527e-07, + "loss": 0.0079, + "step": 26793 + }, + { + "epoch": 2.9048135299219426, + "grad_norm": 4.556452751159668, + "learning_rate": 6.368779921659655e-07, + "loss": 0.0845, + "step": 26794 + }, + { + "epoch": 2.9049219427580226, + "grad_norm": 2.732600688934326, + "learning_rate": 6.361526185985784e-07, + "loss": 0.024, + "step": 26795 + }, + { + "epoch": 2.9050303555941026, + "grad_norm": 10.914010047912598, + "learning_rate": 6.354272450311912e-07, + "loss": 0.0448, + "step": 26796 + }, + { + "epoch": 2.905138768430182, + "grad_norm": 0.9362132549285889, + "learning_rate": 6.34701871463804e-07, + "loss": 0.0021, + "step": 26797 + }, + { + "epoch": 2.9052471812662617, + "grad_norm": 2.58404541015625, + "learning_rate": 6.339764978964167e-07, + "loss": 0.0297, + "step": 26798 + }, + { + "epoch": 2.9053555941023417, + "grad_norm": 2.385930299758911, + "learning_rate": 6.332511243290294e-07, + "loss": 0.0115, + "step": 26799 + }, + { + "epoch": 2.9054640069384217, + "grad_norm": 2.717496395111084, + "learning_rate": 6.325257507616423e-07, + "loss": 0.0033, + "step": 26800 + }, + { + "epoch": 2.905572419774501, + "grad_norm": 3.0931522846221924, + "learning_rate": 6.318003771942551e-07, + "loss": 0.0049, + "step": 26801 + }, + { + "epoch": 2.905680832610581, + "grad_norm": 0.2613234519958496, + "learning_rate": 6.310750036268678e-07, + "loss": 0.0012, + "step": 26802 + }, + { + "epoch": 2.9057892454466607, + "grad_norm": 0.0697375014424324, + "learning_rate": 6.303496300594806e-07, + "loss": 0.0006, + "step": 26803 + }, + { + "epoch": 2.9058976582827407, + "grad_norm": 6.439998626708984, + "learning_rate": 6.296242564920935e-07, + "loss": 0.0112, + "step": 26804 + }, + { + "epoch": 2.9060060711188207, + "grad_norm": 0.3333207368850708, + "learning_rate": 6.288988829247063e-07, + "loss": 0.0015, + "step": 26805 + }, + { + "epoch": 2.9061144839549002, + "grad_norm": 0.036486707627773285, + "learning_rate": 6.281735093573191e-07, + "loss": 0.0002, + "step": 26806 + }, + { + "epoch": 2.90622289679098, + "grad_norm": 0.5020928978919983, + "learning_rate": 6.274481357899318e-07, + "loss": 0.0026, + "step": 26807 + }, + { + "epoch": 2.9063313096270598, + "grad_norm": 0.830138087272644, + "learning_rate": 6.267227622225447e-07, + "loss": 0.0038, + "step": 26808 + }, + { + "epoch": 2.9064397224631398, + "grad_norm": 3.059711456298828, + "learning_rate": 6.259973886551575e-07, + "loss": 0.0203, + "step": 26809 + }, + { + "epoch": 2.9065481352992193, + "grad_norm": 0.03844796121120453, + "learning_rate": 6.252720150877703e-07, + "loss": 0.0003, + "step": 26810 + }, + { + "epoch": 2.9066565481352993, + "grad_norm": 0.26168644428253174, + "learning_rate": 6.245466415203831e-07, + "loss": 0.0016, + "step": 26811 + }, + { + "epoch": 2.906764960971379, + "grad_norm": 0.02572896145284176, + "learning_rate": 6.238212679529958e-07, + "loss": 0.0002, + "step": 26812 + }, + { + "epoch": 2.906873373807459, + "grad_norm": 0.03847105801105499, + "learning_rate": 6.230958943856086e-07, + "loss": 0.0003, + "step": 26813 + }, + { + "epoch": 2.906981786643539, + "grad_norm": 1.9573312997817993, + "learning_rate": 6.223705208182215e-07, + "loss": 0.0099, + "step": 26814 + }, + { + "epoch": 2.9070901994796183, + "grad_norm": 0.04379737749695778, + "learning_rate": 6.216451472508343e-07, + "loss": 0.0003, + "step": 26815 + }, + { + "epoch": 2.9071986123156983, + "grad_norm": 3.995711088180542, + "learning_rate": 6.20919773683447e-07, + "loss": 0.0148, + "step": 26816 + }, + { + "epoch": 2.907307025151778, + "grad_norm": 2.284353256225586, + "learning_rate": 6.201944001160598e-07, + "loss": 0.0421, + "step": 26817 + }, + { + "epoch": 2.907415437987858, + "grad_norm": 0.05393950268626213, + "learning_rate": 6.194690265486726e-07, + "loss": 0.0005, + "step": 26818 + }, + { + "epoch": 2.907523850823938, + "grad_norm": 7.172441482543945, + "learning_rate": 6.187436529812854e-07, + "loss": 0.0434, + "step": 26819 + }, + { + "epoch": 2.9076322636600174, + "grad_norm": 0.0629732683300972, + "learning_rate": 6.180182794138982e-07, + "loss": 0.0005, + "step": 26820 + }, + { + "epoch": 2.907740676496097, + "grad_norm": 0.03710359334945679, + "learning_rate": 6.172929058465111e-07, + "loss": 0.0003, + "step": 26821 + }, + { + "epoch": 2.907849089332177, + "grad_norm": 0.021599125117063522, + "learning_rate": 6.165675322791238e-07, + "loss": 0.0002, + "step": 26822 + }, + { + "epoch": 2.907957502168257, + "grad_norm": 0.3229139745235443, + "learning_rate": 6.158421587117366e-07, + "loss": 0.0015, + "step": 26823 + }, + { + "epoch": 2.9080659150043364, + "grad_norm": 1.8161967992782593, + "learning_rate": 6.151167851443494e-07, + "loss": 0.0194, + "step": 26824 + }, + { + "epoch": 2.9081743278404164, + "grad_norm": 1.297333836555481, + "learning_rate": 6.143914115769622e-07, + "loss": 0.0017, + "step": 26825 + }, + { + "epoch": 2.908282740676496, + "grad_norm": 0.009846153669059277, + "learning_rate": 6.13666038009575e-07, + "loss": 0.0001, + "step": 26826 + }, + { + "epoch": 2.908391153512576, + "grad_norm": 0.045305173844099045, + "learning_rate": 6.129406644421878e-07, + "loss": 0.0003, + "step": 26827 + }, + { + "epoch": 2.908499566348656, + "grad_norm": 3.672184944152832, + "learning_rate": 6.122152908748006e-07, + "loss": 0.0134, + "step": 26828 + }, + { + "epoch": 2.9086079791847355, + "grad_norm": 0.9490107893943787, + "learning_rate": 6.114899173074133e-07, + "loss": 0.0402, + "step": 26829 + }, + { + "epoch": 2.908716392020815, + "grad_norm": 0.1021624207496643, + "learning_rate": 6.107645437400262e-07, + "loss": 0.0005, + "step": 26830 + }, + { + "epoch": 2.908824804856895, + "grad_norm": 10.656961441040039, + "learning_rate": 6.10039170172639e-07, + "loss": 0.0413, + "step": 26831 + }, + { + "epoch": 2.908933217692975, + "grad_norm": 0.4849393963813782, + "learning_rate": 6.093137966052517e-07, + "loss": 0.0006, + "step": 26832 + }, + { + "epoch": 2.9090416305290545, + "grad_norm": 13.44057559967041, + "learning_rate": 6.085884230378645e-07, + "loss": 0.0071, + "step": 26833 + }, + { + "epoch": 2.9091500433651345, + "grad_norm": 0.1413939893245697, + "learning_rate": 6.078630494704774e-07, + "loss": 0.001, + "step": 26834 + }, + { + "epoch": 2.909258456201214, + "grad_norm": 0.5288137793540955, + "learning_rate": 6.071376759030902e-07, + "loss": 0.0018, + "step": 26835 + }, + { + "epoch": 2.909366869037294, + "grad_norm": 4.46058464050293, + "learning_rate": 6.06412302335703e-07, + "loss": 0.0501, + "step": 26836 + }, + { + "epoch": 2.909475281873374, + "grad_norm": 12.823922157287598, + "learning_rate": 6.056869287683157e-07, + "loss": 0.0268, + "step": 26837 + }, + { + "epoch": 2.9095836947094535, + "grad_norm": 0.2228090763092041, + "learning_rate": 6.049615552009285e-07, + "loss": 0.0013, + "step": 26838 + }, + { + "epoch": 2.9096921075455335, + "grad_norm": 3.314204692840576, + "learning_rate": 6.042361816335413e-07, + "loss": 0.0123, + "step": 26839 + }, + { + "epoch": 2.909800520381613, + "grad_norm": 0.045835476368665695, + "learning_rate": 6.035108080661542e-07, + "loss": 0.0004, + "step": 26840 + }, + { + "epoch": 2.909908933217693, + "grad_norm": 2.049476385116577, + "learning_rate": 6.02785434498767e-07, + "loss": 0.0367, + "step": 26841 + }, + { + "epoch": 2.9100173460537726, + "grad_norm": 0.11641906946897507, + "learning_rate": 6.020600609313797e-07, + "loss": 0.0009, + "step": 26842 + }, + { + "epoch": 2.9101257588898526, + "grad_norm": 0.4150746464729309, + "learning_rate": 6.013346873639925e-07, + "loss": 0.0026, + "step": 26843 + }, + { + "epoch": 2.910234171725932, + "grad_norm": 0.013777420856058598, + "learning_rate": 6.006093137966053e-07, + "loss": 0.0001, + "step": 26844 + }, + { + "epoch": 2.910342584562012, + "grad_norm": 12.702127456665039, + "learning_rate": 5.998839402292181e-07, + "loss": 0.044, + "step": 26845 + }, + { + "epoch": 2.910450997398092, + "grad_norm": 2.972032308578491, + "learning_rate": 5.991585666618308e-07, + "loss": 0.0275, + "step": 26846 + }, + { + "epoch": 2.9105594102341716, + "grad_norm": 1.2742626667022705, + "learning_rate": 5.984331930944437e-07, + "loss": 0.0175, + "step": 26847 + }, + { + "epoch": 2.9106678230702516, + "grad_norm": 1.255127191543579, + "learning_rate": 5.977078195270565e-07, + "loss": 0.0321, + "step": 26848 + }, + { + "epoch": 2.910776235906331, + "grad_norm": 0.07734121382236481, + "learning_rate": 5.969824459596693e-07, + "loss": 0.0006, + "step": 26849 + }, + { + "epoch": 2.910884648742411, + "grad_norm": 2.87817645072937, + "learning_rate": 5.962570723922821e-07, + "loss": 0.0127, + "step": 26850 + }, + { + "epoch": 2.910993061578491, + "grad_norm": 2.7054851055145264, + "learning_rate": 5.955316988248948e-07, + "loss": 0.004, + "step": 26851 + }, + { + "epoch": 2.9111014744145707, + "grad_norm": 1.368321180343628, + "learning_rate": 5.948063252575076e-07, + "loss": 0.0036, + "step": 26852 + }, + { + "epoch": 2.91120988725065, + "grad_norm": 0.014339905232191086, + "learning_rate": 5.940809516901205e-07, + "loss": 0.0002, + "step": 26853 + }, + { + "epoch": 2.91131830008673, + "grad_norm": 0.03493437170982361, + "learning_rate": 5.933555781227333e-07, + "loss": 0.0004, + "step": 26854 + }, + { + "epoch": 2.91142671292281, + "grad_norm": 0.13677288591861725, + "learning_rate": 5.926302045553461e-07, + "loss": 0.001, + "step": 26855 + }, + { + "epoch": 2.9115351257588897, + "grad_norm": 7.686297416687012, + "learning_rate": 5.919048309879588e-07, + "loss": 0.0066, + "step": 26856 + }, + { + "epoch": 2.9116435385949697, + "grad_norm": 0.6524405479431152, + "learning_rate": 5.911794574205716e-07, + "loss": 0.0022, + "step": 26857 + }, + { + "epoch": 2.9117519514310493, + "grad_norm": 1.626819372177124, + "learning_rate": 5.904540838531844e-07, + "loss": 0.0022, + "step": 26858 + }, + { + "epoch": 2.9118603642671292, + "grad_norm": 0.03383341804146767, + "learning_rate": 5.897287102857972e-07, + "loss": 0.0004, + "step": 26859 + }, + { + "epoch": 2.9119687771032092, + "grad_norm": 0.05017358809709549, + "learning_rate": 5.890033367184101e-07, + "loss": 0.0004, + "step": 26860 + }, + { + "epoch": 2.9120771899392888, + "grad_norm": 0.17368493974208832, + "learning_rate": 5.882779631510228e-07, + "loss": 0.0007, + "step": 26861 + }, + { + "epoch": 2.9121856027753688, + "grad_norm": 0.5591144561767578, + "learning_rate": 5.875525895836356e-07, + "loss": 0.0025, + "step": 26862 + }, + { + "epoch": 2.9122940156114483, + "grad_norm": 0.3035433888435364, + "learning_rate": 5.868272160162484e-07, + "loss": 0.0021, + "step": 26863 + }, + { + "epoch": 2.9124024284475283, + "grad_norm": 3.0605552196502686, + "learning_rate": 5.861018424488612e-07, + "loss": 0.0065, + "step": 26864 + }, + { + "epoch": 2.912510841283608, + "grad_norm": 0.01992388442158699, + "learning_rate": 5.85376468881474e-07, + "loss": 0.0001, + "step": 26865 + }, + { + "epoch": 2.912619254119688, + "grad_norm": 1.7998651266098022, + "learning_rate": 5.846510953140868e-07, + "loss": 0.0176, + "step": 26866 + }, + { + "epoch": 2.9127276669557673, + "grad_norm": 0.29168394207954407, + "learning_rate": 5.839257217466996e-07, + "loss": 0.001, + "step": 26867 + }, + { + "epoch": 2.9128360797918473, + "grad_norm": 0.24744915962219238, + "learning_rate": 5.832003481793124e-07, + "loss": 0.0012, + "step": 26868 + }, + { + "epoch": 2.9129444926279273, + "grad_norm": 0.49937450885772705, + "learning_rate": 5.824749746119252e-07, + "loss": 0.0038, + "step": 26869 + }, + { + "epoch": 2.913052905464007, + "grad_norm": 5.163775444030762, + "learning_rate": 5.81749601044538e-07, + "loss": 0.0164, + "step": 26870 + }, + { + "epoch": 2.913161318300087, + "grad_norm": 0.5608192682266235, + "learning_rate": 5.810242274771507e-07, + "loss": 0.0013, + "step": 26871 + }, + { + "epoch": 2.9132697311361664, + "grad_norm": 0.04905320703983307, + "learning_rate": 5.802988539097635e-07, + "loss": 0.0005, + "step": 26872 + }, + { + "epoch": 2.9133781439722464, + "grad_norm": 2.2318637371063232, + "learning_rate": 5.795734803423764e-07, + "loss": 0.0064, + "step": 26873 + }, + { + "epoch": 2.9134865568083264, + "grad_norm": 0.25971120595932007, + "learning_rate": 5.788481067749892e-07, + "loss": 0.001, + "step": 26874 + }, + { + "epoch": 2.913594969644406, + "grad_norm": 5.689784526824951, + "learning_rate": 5.78122733207602e-07, + "loss": 0.0071, + "step": 26875 + }, + { + "epoch": 2.9137033824804854, + "grad_norm": 3.999053478240967, + "learning_rate": 5.773973596402147e-07, + "loss": 0.035, + "step": 26876 + }, + { + "epoch": 2.9138117953165654, + "grad_norm": 7.843182563781738, + "learning_rate": 5.766719860728275e-07, + "loss": 0.0039, + "step": 26877 + }, + { + "epoch": 2.9139202081526454, + "grad_norm": 1.785484790802002, + "learning_rate": 5.759466125054403e-07, + "loss": 0.0097, + "step": 26878 + }, + { + "epoch": 2.914028620988725, + "grad_norm": 0.01949550211429596, + "learning_rate": 5.752212389380532e-07, + "loss": 0.0002, + "step": 26879 + }, + { + "epoch": 2.914137033824805, + "grad_norm": 0.3230697512626648, + "learning_rate": 5.74495865370666e-07, + "loss": 0.0011, + "step": 26880 + }, + { + "epoch": 2.9142454466608845, + "grad_norm": 1.3730840682983398, + "learning_rate": 5.737704918032787e-07, + "loss": 0.0122, + "step": 26881 + }, + { + "epoch": 2.9143538594969645, + "grad_norm": 11.175516128540039, + "learning_rate": 5.730451182358915e-07, + "loss": 0.0258, + "step": 26882 + }, + { + "epoch": 2.9144622723330444, + "grad_norm": 0.8497040271759033, + "learning_rate": 5.723197446685044e-07, + "loss": 0.0033, + "step": 26883 + }, + { + "epoch": 2.914570685169124, + "grad_norm": 3.57845139503479, + "learning_rate": 5.715943711011171e-07, + "loss": 0.0069, + "step": 26884 + }, + { + "epoch": 2.914679098005204, + "grad_norm": 1.7427055835723877, + "learning_rate": 5.708689975337299e-07, + "loss": 0.0025, + "step": 26885 + }, + { + "epoch": 2.9147875108412835, + "grad_norm": 0.060001298785209656, + "learning_rate": 5.701436239663427e-07, + "loss": 0.0008, + "step": 26886 + }, + { + "epoch": 2.9148959236773635, + "grad_norm": 0.09447606652975082, + "learning_rate": 5.694182503989555e-07, + "loss": 0.0007, + "step": 26887 + }, + { + "epoch": 2.915004336513443, + "grad_norm": 0.13856114447116852, + "learning_rate": 5.686928768315683e-07, + "loss": 0.0008, + "step": 26888 + }, + { + "epoch": 2.915112749349523, + "grad_norm": 1.8285714387893677, + "learning_rate": 5.679675032641812e-07, + "loss": 0.0055, + "step": 26889 + }, + { + "epoch": 2.9152211621856026, + "grad_norm": 0.18600475788116455, + "learning_rate": 5.672421296967938e-07, + "loss": 0.0002, + "step": 26890 + }, + { + "epoch": 2.9153295750216826, + "grad_norm": 0.0878511369228363, + "learning_rate": 5.665167561294066e-07, + "loss": 0.0004, + "step": 26891 + }, + { + "epoch": 2.9154379878577625, + "grad_norm": 0.361189067363739, + "learning_rate": 5.657913825620195e-07, + "loss": 0.0013, + "step": 26892 + }, + { + "epoch": 2.915546400693842, + "grad_norm": 0.008848106488585472, + "learning_rate": 5.650660089946323e-07, + "loss": 0.0, + "step": 26893 + }, + { + "epoch": 2.915654813529922, + "grad_norm": 2.622804641723633, + "learning_rate": 5.643406354272451e-07, + "loss": 0.0178, + "step": 26894 + }, + { + "epoch": 2.9157632263660016, + "grad_norm": 0.3120042383670807, + "learning_rate": 5.636152618598578e-07, + "loss": 0.0018, + "step": 26895 + }, + { + "epoch": 2.9158716392020816, + "grad_norm": 1.058725357055664, + "learning_rate": 5.628898882924707e-07, + "loss": 0.0026, + "step": 26896 + }, + { + "epoch": 2.9159800520381616, + "grad_norm": 2.5262246131896973, + "learning_rate": 5.621645147250834e-07, + "loss": 0.0093, + "step": 26897 + }, + { + "epoch": 2.916088464874241, + "grad_norm": 1.4349602460861206, + "learning_rate": 5.614391411576963e-07, + "loss": 0.0446, + "step": 26898 + }, + { + "epoch": 2.9161968777103207, + "grad_norm": 0.12476015090942383, + "learning_rate": 5.607137675903091e-07, + "loss": 0.0006, + "step": 26899 + }, + { + "epoch": 2.9163052905464006, + "grad_norm": 12.385324478149414, + "learning_rate": 5.599883940229218e-07, + "loss": 0.0301, + "step": 26900 + }, + { + "epoch": 2.9164137033824806, + "grad_norm": 4.891871452331543, + "learning_rate": 5.592630204555346e-07, + "loss": 0.0133, + "step": 26901 + }, + { + "epoch": 2.91652211621856, + "grad_norm": 0.027039822190999985, + "learning_rate": 5.585376468881475e-07, + "loss": 0.0004, + "step": 26902 + }, + { + "epoch": 2.91663052905464, + "grad_norm": 18.14236831665039, + "learning_rate": 5.578122733207603e-07, + "loss": 0.0254, + "step": 26903 + }, + { + "epoch": 2.9167389418907197, + "grad_norm": 0.4883624315261841, + "learning_rate": 5.57086899753373e-07, + "loss": 0.0018, + "step": 26904 + }, + { + "epoch": 2.9168473547267997, + "grad_norm": 3.7133331298828125, + "learning_rate": 5.563615261859858e-07, + "loss": 0.0124, + "step": 26905 + }, + { + "epoch": 2.9169557675628797, + "grad_norm": 1.7770640850067139, + "learning_rate": 5.556361526185986e-07, + "loss": 0.0534, + "step": 26906 + }, + { + "epoch": 2.917064180398959, + "grad_norm": 0.04509255290031433, + "learning_rate": 5.549107790512114e-07, + "loss": 0.0004, + "step": 26907 + }, + { + "epoch": 2.917172593235039, + "grad_norm": 0.09847188740968704, + "learning_rate": 5.541854054838242e-07, + "loss": 0.0009, + "step": 26908 + }, + { + "epoch": 2.9172810060711187, + "grad_norm": 1.6070518493652344, + "learning_rate": 5.534600319164371e-07, + "loss": 0.0295, + "step": 26909 + }, + { + "epoch": 2.9173894189071987, + "grad_norm": 1.675759196281433, + "learning_rate": 5.527346583490498e-07, + "loss": 0.0015, + "step": 26910 + }, + { + "epoch": 2.9174978317432783, + "grad_norm": 4.542179107666016, + "learning_rate": 5.520092847816626e-07, + "loss": 0.0263, + "step": 26911 + }, + { + "epoch": 2.9176062445793582, + "grad_norm": 4.6630072593688965, + "learning_rate": 5.512839112142754e-07, + "loss": 0.0051, + "step": 26912 + }, + { + "epoch": 2.917714657415438, + "grad_norm": 0.051764052361249924, + "learning_rate": 5.505585376468882e-07, + "loss": 0.0004, + "step": 26913 + }, + { + "epoch": 2.9178230702515178, + "grad_norm": 3.2983646392822266, + "learning_rate": 5.49833164079501e-07, + "loss": 0.0117, + "step": 26914 + }, + { + "epoch": 2.9179314830875978, + "grad_norm": 3.919212579727173, + "learning_rate": 5.491077905121138e-07, + "loss": 0.0188, + "step": 26915 + }, + { + "epoch": 2.9180398959236773, + "grad_norm": 1.6940547227859497, + "learning_rate": 5.483824169447266e-07, + "loss": 0.0066, + "step": 26916 + }, + { + "epoch": 2.9181483087597573, + "grad_norm": 0.6733911633491516, + "learning_rate": 5.476570433773394e-07, + "loss": 0.002, + "step": 26917 + }, + { + "epoch": 2.918256721595837, + "grad_norm": 0.020643625408411026, + "learning_rate": 5.469316698099522e-07, + "loss": 0.0002, + "step": 26918 + }, + { + "epoch": 2.918365134431917, + "grad_norm": 2.1453638076782227, + "learning_rate": 5.46206296242565e-07, + "loss": 0.0036, + "step": 26919 + }, + { + "epoch": 2.918473547267997, + "grad_norm": 0.07271428406238556, + "learning_rate": 5.454809226751777e-07, + "loss": 0.0005, + "step": 26920 + }, + { + "epoch": 2.9185819601040763, + "grad_norm": 2.7112085819244385, + "learning_rate": 5.447555491077905e-07, + "loss": 0.0213, + "step": 26921 + }, + { + "epoch": 2.918690372940156, + "grad_norm": 0.048232316970825195, + "learning_rate": 5.440301755404034e-07, + "loss": 0.0004, + "step": 26922 + }, + { + "epoch": 2.918798785776236, + "grad_norm": 0.05613983795046806, + "learning_rate": 5.433048019730162e-07, + "loss": 0.0004, + "step": 26923 + }, + { + "epoch": 2.918907198612316, + "grad_norm": 0.00798211619257927, + "learning_rate": 5.42579428405629e-07, + "loss": 0.0001, + "step": 26924 + }, + { + "epoch": 2.9190156114483954, + "grad_norm": 3.23964262008667, + "learning_rate": 5.418540548382417e-07, + "loss": 0.0033, + "step": 26925 + }, + { + "epoch": 2.9191240242844754, + "grad_norm": 0.06605904549360275, + "learning_rate": 5.411286812708545e-07, + "loss": 0.0007, + "step": 26926 + }, + { + "epoch": 2.919232437120555, + "grad_norm": 0.7590980529785156, + "learning_rate": 5.404033077034673e-07, + "loss": 0.0017, + "step": 26927 + }, + { + "epoch": 2.919340849956635, + "grad_norm": 0.047010574489831924, + "learning_rate": 5.396779341360802e-07, + "loss": 0.0005, + "step": 26928 + }, + { + "epoch": 2.919449262792715, + "grad_norm": 1.0206022262573242, + "learning_rate": 5.38952560568693e-07, + "loss": 0.0019, + "step": 26929 + }, + { + "epoch": 2.9195576756287944, + "grad_norm": 3.1723010540008545, + "learning_rate": 5.382271870013057e-07, + "loss": 0.0347, + "step": 26930 + }, + { + "epoch": 2.919666088464874, + "grad_norm": 2.49696946144104, + "learning_rate": 5.375018134339185e-07, + "loss": 0.0035, + "step": 26931 + }, + { + "epoch": 2.919774501300954, + "grad_norm": 0.5107040405273438, + "learning_rate": 5.367764398665313e-07, + "loss": 0.003, + "step": 26932 + }, + { + "epoch": 2.919882914137034, + "grad_norm": 3.915604591369629, + "learning_rate": 5.360510662991441e-07, + "loss": 0.0356, + "step": 26933 + }, + { + "epoch": 2.9199913269731135, + "grad_norm": 0.033632274717092514, + "learning_rate": 5.353256927317569e-07, + "loss": 0.0002, + "step": 26934 + }, + { + "epoch": 2.9200997398091935, + "grad_norm": 0.005306858103722334, + "learning_rate": 5.346003191643697e-07, + "loss": 0.0001, + "step": 26935 + }, + { + "epoch": 2.920208152645273, + "grad_norm": 0.48774218559265137, + "learning_rate": 5.338749455969825e-07, + "loss": 0.001, + "step": 26936 + }, + { + "epoch": 2.920316565481353, + "grad_norm": 1.6006567478179932, + "learning_rate": 5.331495720295953e-07, + "loss": 0.0052, + "step": 26937 + }, + { + "epoch": 2.920424978317433, + "grad_norm": 11.62883472442627, + "learning_rate": 5.324241984622081e-07, + "loss": 0.0084, + "step": 26938 + }, + { + "epoch": 2.9205333911535125, + "grad_norm": 0.07519375532865524, + "learning_rate": 5.316988248948209e-07, + "loss": 0.0003, + "step": 26939 + }, + { + "epoch": 2.9206418039895925, + "grad_norm": 1.880653738975525, + "learning_rate": 5.309734513274336e-07, + "loss": 0.0125, + "step": 26940 + }, + { + "epoch": 2.920750216825672, + "grad_norm": 0.04445802792906761, + "learning_rate": 5.302480777600465e-07, + "loss": 0.0003, + "step": 26941 + }, + { + "epoch": 2.920858629661752, + "grad_norm": 10.487580299377441, + "learning_rate": 5.295227041926593e-07, + "loss": 0.0073, + "step": 26942 + }, + { + "epoch": 2.920967042497832, + "grad_norm": 0.021829722449183464, + "learning_rate": 5.287973306252721e-07, + "loss": 0.0001, + "step": 26943 + }, + { + "epoch": 2.9210754553339116, + "grad_norm": 5.900472164154053, + "learning_rate": 5.280719570578849e-07, + "loss": 0.018, + "step": 26944 + }, + { + "epoch": 2.921183868169991, + "grad_norm": 2.4139316082000732, + "learning_rate": 5.273465834904976e-07, + "loss": 0.055, + "step": 26945 + }, + { + "epoch": 2.921292281006071, + "grad_norm": 0.12579584121704102, + "learning_rate": 5.266212099231104e-07, + "loss": 0.0012, + "step": 26946 + }, + { + "epoch": 2.921400693842151, + "grad_norm": 1.1590962409973145, + "learning_rate": 5.258958363557232e-07, + "loss": 0.0084, + "step": 26947 + }, + { + "epoch": 2.9215091066782306, + "grad_norm": 0.28112101554870605, + "learning_rate": 5.251704627883361e-07, + "loss": 0.0009, + "step": 26948 + }, + { + "epoch": 2.9216175195143106, + "grad_norm": 0.5401988625526428, + "learning_rate": 5.244450892209488e-07, + "loss": 0.0013, + "step": 26949 + }, + { + "epoch": 2.92172593235039, + "grad_norm": 3.063474178314209, + "learning_rate": 5.237197156535616e-07, + "loss": 0.023, + "step": 26950 + }, + { + "epoch": 2.92183434518647, + "grad_norm": 5.384069919586182, + "learning_rate": 5.229943420861744e-07, + "loss": 0.0157, + "step": 26951 + }, + { + "epoch": 2.92194275802255, + "grad_norm": 3.0771517753601074, + "learning_rate": 5.222689685187872e-07, + "loss": 0.0059, + "step": 26952 + }, + { + "epoch": 2.9220511708586296, + "grad_norm": 0.019286278635263443, + "learning_rate": 5.215435949514e-07, + "loss": 0.0002, + "step": 26953 + }, + { + "epoch": 2.922159583694709, + "grad_norm": 0.02347339130938053, + "learning_rate": 5.208182213840128e-07, + "loss": 0.0002, + "step": 26954 + }, + { + "epoch": 2.922267996530789, + "grad_norm": 0.015477489680051804, + "learning_rate": 5.200928478166256e-07, + "loss": 0.0002, + "step": 26955 + }, + { + "epoch": 2.922376409366869, + "grad_norm": 2.432685375213623, + "learning_rate": 5.193674742492384e-07, + "loss": 0.0029, + "step": 26956 + }, + { + "epoch": 2.9224848222029487, + "grad_norm": 0.26789411902427673, + "learning_rate": 5.186421006818512e-07, + "loss": 0.0009, + "step": 26957 + }, + { + "epoch": 2.9225932350390287, + "grad_norm": 0.06462398171424866, + "learning_rate": 5.17916727114464e-07, + "loss": 0.0003, + "step": 26958 + }, + { + "epoch": 2.922701647875108, + "grad_norm": 0.025463197380304337, + "learning_rate": 5.171913535470767e-07, + "loss": 0.0002, + "step": 26959 + }, + { + "epoch": 2.922810060711188, + "grad_norm": 0.16579769551753998, + "learning_rate": 5.164659799796895e-07, + "loss": 0.0013, + "step": 26960 + }, + { + "epoch": 2.922918473547268, + "grad_norm": 0.06837236881256104, + "learning_rate": 5.157406064123024e-07, + "loss": 0.0003, + "step": 26961 + }, + { + "epoch": 2.9230268863833477, + "grad_norm": 0.13090047240257263, + "learning_rate": 5.150152328449152e-07, + "loss": 0.0009, + "step": 26962 + }, + { + "epoch": 2.9231352992194277, + "grad_norm": 7.791872501373291, + "learning_rate": 5.14289859277528e-07, + "loss": 0.0536, + "step": 26963 + }, + { + "epoch": 2.9232437120555073, + "grad_norm": 1.2927414178848267, + "learning_rate": 5.135644857101408e-07, + "loss": 0.0027, + "step": 26964 + }, + { + "epoch": 2.9233521248915872, + "grad_norm": 4.255307674407959, + "learning_rate": 5.128391121427535e-07, + "loss": 0.005, + "step": 26965 + }, + { + "epoch": 2.9234605377276672, + "grad_norm": 5.7741570472717285, + "learning_rate": 5.121137385753663e-07, + "loss": 0.0275, + "step": 26966 + }, + { + "epoch": 2.9235689505637468, + "grad_norm": 0.07241856306791306, + "learning_rate": 5.113883650079792e-07, + "loss": 0.0005, + "step": 26967 + }, + { + "epoch": 2.9236773633998263, + "grad_norm": 4.3795037269592285, + "learning_rate": 5.10662991440592e-07, + "loss": 0.0228, + "step": 26968 + }, + { + "epoch": 2.9237857762359063, + "grad_norm": 0.015175165608525276, + "learning_rate": 5.099376178732047e-07, + "loss": 0.0001, + "step": 26969 + }, + { + "epoch": 2.9238941890719863, + "grad_norm": 0.25625771284103394, + "learning_rate": 5.092122443058175e-07, + "loss": 0.0012, + "step": 26970 + }, + { + "epoch": 2.924002601908066, + "grad_norm": 3.1793479919433594, + "learning_rate": 5.084868707384304e-07, + "loss": 0.0184, + "step": 26971 + }, + { + "epoch": 2.924111014744146, + "grad_norm": 0.17441758513450623, + "learning_rate": 5.077614971710431e-07, + "loss": 0.0016, + "step": 26972 + }, + { + "epoch": 2.9242194275802254, + "grad_norm": 0.015920771285891533, + "learning_rate": 5.07036123603656e-07, + "loss": 0.0002, + "step": 26973 + }, + { + "epoch": 2.9243278404163053, + "grad_norm": 0.02836981602013111, + "learning_rate": 5.063107500362687e-07, + "loss": 0.0001, + "step": 26974 + }, + { + "epoch": 2.9244362532523853, + "grad_norm": 0.7329725027084351, + "learning_rate": 5.055853764688815e-07, + "loss": 0.0065, + "step": 26975 + }, + { + "epoch": 2.924544666088465, + "grad_norm": 1.1552170515060425, + "learning_rate": 5.048600029014943e-07, + "loss": 0.0046, + "step": 26976 + }, + { + "epoch": 2.9246530789245444, + "grad_norm": 4.925846099853516, + "learning_rate": 5.041346293341072e-07, + "loss": 0.0668, + "step": 26977 + }, + { + "epoch": 2.9247614917606244, + "grad_norm": 2.9531030654907227, + "learning_rate": 5.034092557667199e-07, + "loss": 0.0335, + "step": 26978 + }, + { + "epoch": 2.9248699045967044, + "grad_norm": 1.0949339866638184, + "learning_rate": 5.026838821993326e-07, + "loss": 0.0163, + "step": 26979 + }, + { + "epoch": 2.924978317432784, + "grad_norm": 0.041054073721170425, + "learning_rate": 5.019585086319455e-07, + "loss": 0.0003, + "step": 26980 + }, + { + "epoch": 2.925086730268864, + "grad_norm": 3.6528964042663574, + "learning_rate": 5.012331350645583e-07, + "loss": 0.0081, + "step": 26981 + }, + { + "epoch": 2.9251951431049434, + "grad_norm": 2.2933948040008545, + "learning_rate": 5.005077614971711e-07, + "loss": 0.0074, + "step": 26982 + }, + { + "epoch": 2.9253035559410234, + "grad_norm": 0.06758727878332138, + "learning_rate": 4.997823879297839e-07, + "loss": 0.0004, + "step": 26983 + }, + { + "epoch": 2.9254119687771034, + "grad_norm": 6.070744037628174, + "learning_rate": 4.990570143623967e-07, + "loss": 0.0367, + "step": 26984 + }, + { + "epoch": 2.925520381613183, + "grad_norm": 0.4625746011734009, + "learning_rate": 4.983316407950094e-07, + "loss": 0.0026, + "step": 26985 + }, + { + "epoch": 2.925628794449263, + "grad_norm": 8.512067794799805, + "learning_rate": 4.976062672276223e-07, + "loss": 0.0314, + "step": 26986 + }, + { + "epoch": 2.9257372072853425, + "grad_norm": 0.39522799849510193, + "learning_rate": 4.968808936602351e-07, + "loss": 0.0016, + "step": 26987 + }, + { + "epoch": 2.9258456201214225, + "grad_norm": 5.568704128265381, + "learning_rate": 4.961555200928479e-07, + "loss": 0.0165, + "step": 26988 + }, + { + "epoch": 2.925954032957502, + "grad_norm": 0.16049568355083466, + "learning_rate": 4.954301465254606e-07, + "loss": 0.001, + "step": 26989 + }, + { + "epoch": 2.926062445793582, + "grad_norm": 0.04410437494516373, + "learning_rate": 4.947047729580735e-07, + "loss": 0.0004, + "step": 26990 + }, + { + "epoch": 2.9261708586296615, + "grad_norm": 2.514455795288086, + "learning_rate": 4.939793993906863e-07, + "loss": 0.0318, + "step": 26991 + }, + { + "epoch": 2.9262792714657415, + "grad_norm": 6.279289722442627, + "learning_rate": 4.93254025823299e-07, + "loss": 0.0122, + "step": 26992 + }, + { + "epoch": 2.9263876843018215, + "grad_norm": 0.14081700146198273, + "learning_rate": 4.925286522559119e-07, + "loss": 0.0006, + "step": 26993 + }, + { + "epoch": 2.926496097137901, + "grad_norm": 0.27355486154556274, + "learning_rate": 4.918032786885246e-07, + "loss": 0.0009, + "step": 26994 + }, + { + "epoch": 2.926604509973981, + "grad_norm": 10.654217720031738, + "learning_rate": 4.910779051211374e-07, + "loss": 0.02, + "step": 26995 + }, + { + "epoch": 2.9267129228100606, + "grad_norm": 3.9625742435455322, + "learning_rate": 4.903525315537502e-07, + "loss": 0.0267, + "step": 26996 + }, + { + "epoch": 2.9268213356461406, + "grad_norm": 12.666926383972168, + "learning_rate": 4.896271579863631e-07, + "loss": 0.0436, + "step": 26997 + }, + { + "epoch": 2.9269297484822205, + "grad_norm": 1.2061164379119873, + "learning_rate": 4.889017844189759e-07, + "loss": 0.0014, + "step": 26998 + }, + { + "epoch": 2.9270381613183, + "grad_norm": 3.0712313652038574, + "learning_rate": 4.881764108515886e-07, + "loss": 0.0152, + "step": 26999 + }, + { + "epoch": 2.9271465741543796, + "grad_norm": 0.20201687514781952, + "learning_rate": 4.874510372842014e-07, + "loss": 0.0008, + "step": 27000 + }, + { + "epoch": 2.9272549869904596, + "grad_norm": 0.27966800332069397, + "learning_rate": 4.867256637168142e-07, + "loss": 0.0011, + "step": 27001 + }, + { + "epoch": 2.9273633998265396, + "grad_norm": 2.4325482845306396, + "learning_rate": 4.86000290149427e-07, + "loss": 0.0038, + "step": 27002 + }, + { + "epoch": 2.927471812662619, + "grad_norm": 1.8965866565704346, + "learning_rate": 4.852749165820399e-07, + "loss": 0.0082, + "step": 27003 + }, + { + "epoch": 2.927580225498699, + "grad_norm": 9.329872131347656, + "learning_rate": 4.845495430146526e-07, + "loss": 0.0062, + "step": 27004 + }, + { + "epoch": 2.9276886383347787, + "grad_norm": 10.153425216674805, + "learning_rate": 4.838241694472654e-07, + "loss": 0.0313, + "step": 27005 + }, + { + "epoch": 2.9277970511708586, + "grad_norm": 9.056111335754395, + "learning_rate": 4.830987958798782e-07, + "loss": 0.0161, + "step": 27006 + }, + { + "epoch": 2.9279054640069386, + "grad_norm": 0.04194505885243416, + "learning_rate": 4.82373422312491e-07, + "loss": 0.0003, + "step": 27007 + }, + { + "epoch": 2.928013876843018, + "grad_norm": 1.15150785446167, + "learning_rate": 4.816480487451037e-07, + "loss": 0.026, + "step": 27008 + }, + { + "epoch": 2.928122289679098, + "grad_norm": 0.02087576687335968, + "learning_rate": 4.809226751777165e-07, + "loss": 0.0001, + "step": 27009 + }, + { + "epoch": 2.9282307025151777, + "grad_norm": 0.033887121826410294, + "learning_rate": 4.801973016103294e-07, + "loss": 0.0003, + "step": 27010 + }, + { + "epoch": 2.9283391153512577, + "grad_norm": 0.026018846780061722, + "learning_rate": 4.794719280429422e-07, + "loss": 0.0003, + "step": 27011 + }, + { + "epoch": 2.9284475281873372, + "grad_norm": 0.08653596043586731, + "learning_rate": 4.78746554475555e-07, + "loss": 0.0006, + "step": 27012 + }, + { + "epoch": 2.928555941023417, + "grad_norm": 0.3692929446697235, + "learning_rate": 4.780211809081677e-07, + "loss": 0.0012, + "step": 27013 + }, + { + "epoch": 2.9286643538594967, + "grad_norm": 1.9812885522842407, + "learning_rate": 4.772958073407805e-07, + "loss": 0.017, + "step": 27014 + }, + { + "epoch": 2.9287727666955767, + "grad_norm": 0.14937293529510498, + "learning_rate": 4.7657043377339335e-07, + "loss": 0.0009, + "step": 27015 + }, + { + "epoch": 2.9288811795316567, + "grad_norm": 3.74521541595459, + "learning_rate": 4.7584506020600613e-07, + "loss": 0.0074, + "step": 27016 + }, + { + "epoch": 2.9289895923677363, + "grad_norm": 0.4053057134151459, + "learning_rate": 4.7511968663861896e-07, + "loss": 0.0005, + "step": 27017 + }, + { + "epoch": 2.9290980052038162, + "grad_norm": 2.2749736309051514, + "learning_rate": 4.7439431307123174e-07, + "loss": 0.0104, + "step": 27018 + }, + { + "epoch": 2.929206418039896, + "grad_norm": 2.0666658878326416, + "learning_rate": 4.7366893950384447e-07, + "loss": 0.004, + "step": 27019 + }, + { + "epoch": 2.9293148308759758, + "grad_norm": 20.093427658081055, + "learning_rate": 4.729435659364573e-07, + "loss": 0.0125, + "step": 27020 + }, + { + "epoch": 2.9294232437120558, + "grad_norm": 0.440672904253006, + "learning_rate": 4.722181923690701e-07, + "loss": 0.0025, + "step": 27021 + }, + { + "epoch": 2.9295316565481353, + "grad_norm": 3.40927791595459, + "learning_rate": 4.714928188016829e-07, + "loss": 0.0292, + "step": 27022 + }, + { + "epoch": 2.929640069384215, + "grad_norm": 0.22187836468219757, + "learning_rate": 4.707674452342957e-07, + "loss": 0.0006, + "step": 27023 + }, + { + "epoch": 2.929748482220295, + "grad_norm": 2.7464616298675537, + "learning_rate": 4.700420716669085e-07, + "loss": 0.0135, + "step": 27024 + }, + { + "epoch": 2.929856895056375, + "grad_norm": 1.4383727312088013, + "learning_rate": 4.693166980995213e-07, + "loss": 0.0501, + "step": 27025 + }, + { + "epoch": 2.9299653078924544, + "grad_norm": 0.260336309671402, + "learning_rate": 4.685913245321341e-07, + "loss": 0.0015, + "step": 27026 + }, + { + "epoch": 2.9300737207285343, + "grad_norm": 1.9800820350646973, + "learning_rate": 4.6786595096474686e-07, + "loss": 0.0161, + "step": 27027 + }, + { + "epoch": 2.930182133564614, + "grad_norm": 0.09645397961139679, + "learning_rate": 4.671405773973597e-07, + "loss": 0.0009, + "step": 27028 + }, + { + "epoch": 2.930290546400694, + "grad_norm": 0.047650862485170364, + "learning_rate": 4.6641520382997247e-07, + "loss": 0.0002, + "step": 27029 + }, + { + "epoch": 2.930398959236774, + "grad_norm": 3.9981725215911865, + "learning_rate": 4.656898302625853e-07, + "loss": 0.015, + "step": 27030 + }, + { + "epoch": 2.9305073720728534, + "grad_norm": 1.6920782327651978, + "learning_rate": 4.649644566951981e-07, + "loss": 0.0041, + "step": 27031 + }, + { + "epoch": 2.9306157849089334, + "grad_norm": 0.5445091128349304, + "learning_rate": 4.642390831278109e-07, + "loss": 0.0038, + "step": 27032 + }, + { + "epoch": 2.930724197745013, + "grad_norm": 0.05058101564645767, + "learning_rate": 4.6351370956042363e-07, + "loss": 0.0002, + "step": 27033 + }, + { + "epoch": 2.930832610581093, + "grad_norm": 1.6378774642944336, + "learning_rate": 4.627883359930364e-07, + "loss": 0.0018, + "step": 27034 + }, + { + "epoch": 2.9309410234171724, + "grad_norm": 0.03121504932641983, + "learning_rate": 4.6206296242564924e-07, + "loss": 0.0004, + "step": 27035 + }, + { + "epoch": 2.9310494362532524, + "grad_norm": 1.5711637735366821, + "learning_rate": 4.61337588858262e-07, + "loss": 0.0022, + "step": 27036 + }, + { + "epoch": 2.931157849089332, + "grad_norm": 1.7781963348388672, + "learning_rate": 4.6061221529087485e-07, + "loss": 0.0052, + "step": 27037 + }, + { + "epoch": 2.931266261925412, + "grad_norm": 5.309787750244141, + "learning_rate": 4.5988684172348763e-07, + "loss": 0.0382, + "step": 27038 + }, + { + "epoch": 2.931374674761492, + "grad_norm": 5.530873775482178, + "learning_rate": 4.591614681561004e-07, + "loss": 0.0083, + "step": 27039 + }, + { + "epoch": 2.9314830875975715, + "grad_norm": 0.06961692124605179, + "learning_rate": 4.584360945887132e-07, + "loss": 0.0005, + "step": 27040 + }, + { + "epoch": 2.9315915004336515, + "grad_norm": 0.08982890099287033, + "learning_rate": 4.57710721021326e-07, + "loss": 0.0003, + "step": 27041 + }, + { + "epoch": 2.931699913269731, + "grad_norm": 3.9730257987976074, + "learning_rate": 4.569853474539388e-07, + "loss": 0.0034, + "step": 27042 + }, + { + "epoch": 2.931808326105811, + "grad_norm": 7.9339985847473145, + "learning_rate": 4.5625997388655163e-07, + "loss": 0.0272, + "step": 27043 + }, + { + "epoch": 2.931916738941891, + "grad_norm": 1.8090304136276245, + "learning_rate": 4.555346003191644e-07, + "loss": 0.0063, + "step": 27044 + }, + { + "epoch": 2.9320251517779705, + "grad_norm": 0.1621026247739792, + "learning_rate": 4.5480922675177724e-07, + "loss": 0.001, + "step": 27045 + }, + { + "epoch": 2.93213356461405, + "grad_norm": 0.1386643946170807, + "learning_rate": 4.5408385318438997e-07, + "loss": 0.0007, + "step": 27046 + }, + { + "epoch": 2.93224197745013, + "grad_norm": 0.15233924984931946, + "learning_rate": 4.5335847961700275e-07, + "loss": 0.0008, + "step": 27047 + }, + { + "epoch": 2.93235039028621, + "grad_norm": 1.1416264772415161, + "learning_rate": 4.526331060496156e-07, + "loss": 0.0024, + "step": 27048 + }, + { + "epoch": 2.9324588031222896, + "grad_norm": 1.1642780303955078, + "learning_rate": 4.519077324822284e-07, + "loss": 0.0039, + "step": 27049 + }, + { + "epoch": 2.9325672159583696, + "grad_norm": 0.5409708023071289, + "learning_rate": 4.511823589148412e-07, + "loss": 0.0025, + "step": 27050 + }, + { + "epoch": 2.932675628794449, + "grad_norm": 12.070107460021973, + "learning_rate": 4.50456985347454e-07, + "loss": 0.0162, + "step": 27051 + }, + { + "epoch": 2.932784041630529, + "grad_norm": 0.04413892328739166, + "learning_rate": 4.497316117800668e-07, + "loss": 0.0002, + "step": 27052 + }, + { + "epoch": 2.932892454466609, + "grad_norm": 1.6876976490020752, + "learning_rate": 4.490062382126795e-07, + "loss": 0.0018, + "step": 27053 + }, + { + "epoch": 2.9330008673026886, + "grad_norm": 0.3290573060512543, + "learning_rate": 4.4828086464529236e-07, + "loss": 0.0012, + "step": 27054 + }, + { + "epoch": 2.9331092801387686, + "grad_norm": 0.5085753202438354, + "learning_rate": 4.4755549107790513e-07, + "loss": 0.0014, + "step": 27055 + }, + { + "epoch": 2.933217692974848, + "grad_norm": 0.3955126106739044, + "learning_rate": 4.4683011751051797e-07, + "loss": 0.0018, + "step": 27056 + }, + { + "epoch": 2.933326105810928, + "grad_norm": 0.2625862658023834, + "learning_rate": 4.4610474394313075e-07, + "loss": 0.001, + "step": 27057 + }, + { + "epoch": 2.9334345186470077, + "grad_norm": 0.03896129131317139, + "learning_rate": 4.453793703757436e-07, + "loss": 0.0004, + "step": 27058 + }, + { + "epoch": 2.9335429314830876, + "grad_norm": 2.363111972808838, + "learning_rate": 4.4465399680835636e-07, + "loss": 0.0191, + "step": 27059 + }, + { + "epoch": 2.933651344319167, + "grad_norm": 2.3948748111724854, + "learning_rate": 4.4392862324096913e-07, + "loss": 0.019, + "step": 27060 + }, + { + "epoch": 2.933759757155247, + "grad_norm": 2.475959539413452, + "learning_rate": 4.432032496735819e-07, + "loss": 0.0311, + "step": 27061 + }, + { + "epoch": 2.933868169991327, + "grad_norm": 6.042241096496582, + "learning_rate": 4.4247787610619474e-07, + "loss": 0.0419, + "step": 27062 + }, + { + "epoch": 2.9339765828274067, + "grad_norm": 7.241573810577393, + "learning_rate": 4.417525025388075e-07, + "loss": 0.0312, + "step": 27063 + }, + { + "epoch": 2.9340849956634867, + "grad_norm": 1.482003927230835, + "learning_rate": 4.4102712897142035e-07, + "loss": 0.0032, + "step": 27064 + }, + { + "epoch": 2.9341934084995662, + "grad_norm": 0.059348322451114655, + "learning_rate": 4.4030175540403313e-07, + "loss": 0.0005, + "step": 27065 + }, + { + "epoch": 2.934301821335646, + "grad_norm": 0.036200713366270065, + "learning_rate": 4.3957638183664586e-07, + "loss": 0.0004, + "step": 27066 + }, + { + "epoch": 2.934410234171726, + "grad_norm": 6.325877666473389, + "learning_rate": 4.388510082692587e-07, + "loss": 0.0311, + "step": 27067 + }, + { + "epoch": 2.9345186470078057, + "grad_norm": 0.01990659534931183, + "learning_rate": 4.3812563470187147e-07, + "loss": 0.0001, + "step": 27068 + }, + { + "epoch": 2.9346270598438853, + "grad_norm": 2.2365241050720215, + "learning_rate": 4.374002611344843e-07, + "loss": 0.0021, + "step": 27069 + }, + { + "epoch": 2.9347354726799653, + "grad_norm": 1.835990071296692, + "learning_rate": 4.366748875670971e-07, + "loss": 0.0034, + "step": 27070 + }, + { + "epoch": 2.9348438855160452, + "grad_norm": 0.12379803508520126, + "learning_rate": 4.359495139997099e-07, + "loss": 0.0009, + "step": 27071 + }, + { + "epoch": 2.934952298352125, + "grad_norm": 5.203895092010498, + "learning_rate": 4.352241404323227e-07, + "loss": 0.0277, + "step": 27072 + }, + { + "epoch": 2.9350607111882048, + "grad_norm": 0.3624671995639801, + "learning_rate": 4.3449876686493547e-07, + "loss": 0.0031, + "step": 27073 + }, + { + "epoch": 2.9351691240242843, + "grad_norm": 0.03448858857154846, + "learning_rate": 4.3377339329754825e-07, + "loss": 0.0002, + "step": 27074 + }, + { + "epoch": 2.9352775368603643, + "grad_norm": 0.29588887095451355, + "learning_rate": 4.330480197301611e-07, + "loss": 0.0016, + "step": 27075 + }, + { + "epoch": 2.9353859496964443, + "grad_norm": 1.077552080154419, + "learning_rate": 4.3232264616277386e-07, + "loss": 0.0047, + "step": 27076 + }, + { + "epoch": 2.935494362532524, + "grad_norm": 1.113783597946167, + "learning_rate": 4.315972725953867e-07, + "loss": 0.0033, + "step": 27077 + }, + { + "epoch": 2.9356027753686034, + "grad_norm": 0.12317410856485367, + "learning_rate": 4.3087189902799947e-07, + "loss": 0.0006, + "step": 27078 + }, + { + "epoch": 2.9357111882046834, + "grad_norm": 0.03953960910439491, + "learning_rate": 4.301465254606123e-07, + "loss": 0.0002, + "step": 27079 + }, + { + "epoch": 2.9358196010407633, + "grad_norm": 0.08086550980806351, + "learning_rate": 4.29421151893225e-07, + "loss": 0.0004, + "step": 27080 + }, + { + "epoch": 2.935928013876843, + "grad_norm": 4.250651836395264, + "learning_rate": 4.286957783258378e-07, + "loss": 0.004, + "step": 27081 + }, + { + "epoch": 2.936036426712923, + "grad_norm": 0.2561190724372864, + "learning_rate": 4.2797040475845064e-07, + "loss": 0.0007, + "step": 27082 + }, + { + "epoch": 2.9361448395490024, + "grad_norm": 0.7630582451820374, + "learning_rate": 4.272450311910634e-07, + "loss": 0.0026, + "step": 27083 + }, + { + "epoch": 2.9362532523850824, + "grad_norm": 7.600354194641113, + "learning_rate": 4.2651965762367625e-07, + "loss": 0.0109, + "step": 27084 + }, + { + "epoch": 2.9363616652211624, + "grad_norm": 1.015820860862732, + "learning_rate": 4.25794284056289e-07, + "loss": 0.003, + "step": 27085 + }, + { + "epoch": 2.936470078057242, + "grad_norm": 0.6071223616600037, + "learning_rate": 4.2506891048890186e-07, + "loss": 0.0017, + "step": 27086 + }, + { + "epoch": 2.936578490893322, + "grad_norm": 3.8857462406158447, + "learning_rate": 4.243435369215146e-07, + "loss": 0.0054, + "step": 27087 + }, + { + "epoch": 2.9366869037294014, + "grad_norm": 1.0827891826629639, + "learning_rate": 4.236181633541274e-07, + "loss": 0.0444, + "step": 27088 + }, + { + "epoch": 2.9367953165654814, + "grad_norm": 1.1686315536499023, + "learning_rate": 4.228927897867402e-07, + "loss": 0.0033, + "step": 27089 + }, + { + "epoch": 2.9369037294015614, + "grad_norm": 1.7404475212097168, + "learning_rate": 4.22167416219353e-07, + "loss": 0.0328, + "step": 27090 + }, + { + "epoch": 2.937012142237641, + "grad_norm": 1.7517863512039185, + "learning_rate": 4.214420426519658e-07, + "loss": 0.0045, + "step": 27091 + }, + { + "epoch": 2.9371205550737205, + "grad_norm": 0.20657570660114288, + "learning_rate": 4.2071666908457863e-07, + "loss": 0.0014, + "step": 27092 + }, + { + "epoch": 2.9372289679098005, + "grad_norm": 1.5133585929870605, + "learning_rate": 4.199912955171914e-07, + "loss": 0.0235, + "step": 27093 + }, + { + "epoch": 2.9373373807458805, + "grad_norm": 0.45409467816352844, + "learning_rate": 4.1926592194980414e-07, + "loss": 0.0006, + "step": 27094 + }, + { + "epoch": 2.93744579358196, + "grad_norm": 1.1438727378845215, + "learning_rate": 4.1854054838241697e-07, + "loss": 0.0016, + "step": 27095 + }, + { + "epoch": 2.93755420641804, + "grad_norm": 0.29716795682907104, + "learning_rate": 4.1781517481502975e-07, + "loss": 0.0018, + "step": 27096 + }, + { + "epoch": 2.9376626192541195, + "grad_norm": 0.27224573493003845, + "learning_rate": 4.170898012476426e-07, + "loss": 0.0006, + "step": 27097 + }, + { + "epoch": 2.9377710320901995, + "grad_norm": 6.005338191986084, + "learning_rate": 4.1636442768025536e-07, + "loss": 0.0223, + "step": 27098 + }, + { + "epoch": 2.9378794449262795, + "grad_norm": 20.531356811523438, + "learning_rate": 4.156390541128682e-07, + "loss": 0.0384, + "step": 27099 + }, + { + "epoch": 2.937987857762359, + "grad_norm": 2.8189990520477295, + "learning_rate": 4.149136805454809e-07, + "loss": 0.028, + "step": 27100 + }, + { + "epoch": 2.9380962705984386, + "grad_norm": 4.353042125701904, + "learning_rate": 4.1418830697809375e-07, + "loss": 0.025, + "step": 27101 + }, + { + "epoch": 2.9382046834345186, + "grad_norm": 1.5345067977905273, + "learning_rate": 4.134629334107065e-07, + "loss": 0.0797, + "step": 27102 + }, + { + "epoch": 2.9383130962705986, + "grad_norm": 0.03492831066250801, + "learning_rate": 4.1273755984331936e-07, + "loss": 0.0003, + "step": 27103 + }, + { + "epoch": 2.938421509106678, + "grad_norm": 1.1610664129257202, + "learning_rate": 4.1201218627593214e-07, + "loss": 0.0062, + "step": 27104 + }, + { + "epoch": 2.938529921942758, + "grad_norm": 0.03933045268058777, + "learning_rate": 4.1128681270854497e-07, + "loss": 0.0003, + "step": 27105 + }, + { + "epoch": 2.9386383347788376, + "grad_norm": 2.620562791824341, + "learning_rate": 4.1056143914115775e-07, + "loss": 0.0043, + "step": 27106 + }, + { + "epoch": 2.9387467476149176, + "grad_norm": 9.774600982666016, + "learning_rate": 4.0983606557377047e-07, + "loss": 0.0185, + "step": 27107 + }, + { + "epoch": 2.9388551604509976, + "grad_norm": 0.3949384093284607, + "learning_rate": 4.091106920063833e-07, + "loss": 0.0012, + "step": 27108 + }, + { + "epoch": 2.938963573287077, + "grad_norm": 0.517851710319519, + "learning_rate": 4.083853184389961e-07, + "loss": 0.003, + "step": 27109 + }, + { + "epoch": 2.939071986123157, + "grad_norm": 0.9638906717300415, + "learning_rate": 4.076599448716089e-07, + "loss": 0.0023, + "step": 27110 + }, + { + "epoch": 2.9391803989592367, + "grad_norm": 2.294834852218628, + "learning_rate": 4.069345713042217e-07, + "loss": 0.0401, + "step": 27111 + }, + { + "epoch": 2.9392888117953166, + "grad_norm": 0.0428360216319561, + "learning_rate": 4.062091977368345e-07, + "loss": 0.0005, + "step": 27112 + }, + { + "epoch": 2.9393972246313966, + "grad_norm": 0.1252135932445526, + "learning_rate": 4.054838241694473e-07, + "loss": 0.0008, + "step": 27113 + }, + { + "epoch": 2.939505637467476, + "grad_norm": 2.4131710529327393, + "learning_rate": 4.047584506020601e-07, + "loss": 0.0059, + "step": 27114 + }, + { + "epoch": 2.9396140503035557, + "grad_norm": 0.010546919889748096, + "learning_rate": 4.0403307703467286e-07, + "loss": 0.0001, + "step": 27115 + }, + { + "epoch": 2.9397224631396357, + "grad_norm": 2.0412044525146484, + "learning_rate": 4.033077034672857e-07, + "loss": 0.0444, + "step": 27116 + }, + { + "epoch": 2.9398308759757157, + "grad_norm": 0.2259720414876938, + "learning_rate": 4.0258232989989847e-07, + "loss": 0.0005, + "step": 27117 + }, + { + "epoch": 2.9399392888117952, + "grad_norm": 5.749305248260498, + "learning_rate": 4.018569563325113e-07, + "loss": 0.0038, + "step": 27118 + }, + { + "epoch": 2.940047701647875, + "grad_norm": 5.039578914642334, + "learning_rate": 4.011315827651241e-07, + "loss": 0.0341, + "step": 27119 + }, + { + "epoch": 2.9401561144839548, + "grad_norm": 0.06608037650585175, + "learning_rate": 4.004062091977369e-07, + "loss": 0.0005, + "step": 27120 + }, + { + "epoch": 2.9402645273200347, + "grad_norm": 1.725531816482544, + "learning_rate": 3.9968083563034964e-07, + "loss": 0.0068, + "step": 27121 + }, + { + "epoch": 2.9403729401561147, + "grad_norm": 3.2723987102508545, + "learning_rate": 3.989554620629624e-07, + "loss": 0.0237, + "step": 27122 + }, + { + "epoch": 2.9404813529921943, + "grad_norm": 0.14605607092380524, + "learning_rate": 3.9823008849557525e-07, + "loss": 0.0006, + "step": 27123 + }, + { + "epoch": 2.940589765828274, + "grad_norm": 0.39702898263931274, + "learning_rate": 3.975047149281881e-07, + "loss": 0.0026, + "step": 27124 + }, + { + "epoch": 2.940698178664354, + "grad_norm": 0.21129731833934784, + "learning_rate": 3.9677934136080086e-07, + "loss": 0.0006, + "step": 27125 + }, + { + "epoch": 2.9408065915004338, + "grad_norm": 0.20259904861450195, + "learning_rate": 3.960539677934137e-07, + "loss": 0.0016, + "step": 27126 + }, + { + "epoch": 2.9409150043365133, + "grad_norm": 4.603806495666504, + "learning_rate": 3.953285942260264e-07, + "loss": 0.0394, + "step": 27127 + }, + { + "epoch": 2.9410234171725933, + "grad_norm": 9.280555725097656, + "learning_rate": 3.946032206586392e-07, + "loss": 0.0109, + "step": 27128 + }, + { + "epoch": 2.941131830008673, + "grad_norm": 4.760382175445557, + "learning_rate": 3.9387784709125203e-07, + "loss": 0.0097, + "step": 27129 + }, + { + "epoch": 2.941240242844753, + "grad_norm": 0.0784415453672409, + "learning_rate": 3.931524735238648e-07, + "loss": 0.0003, + "step": 27130 + }, + { + "epoch": 2.941348655680833, + "grad_norm": 0.18920481204986572, + "learning_rate": 3.9242709995647764e-07, + "loss": 0.0009, + "step": 27131 + }, + { + "epoch": 2.9414570685169124, + "grad_norm": 0.16080786287784576, + "learning_rate": 3.917017263890904e-07, + "loss": 0.0008, + "step": 27132 + }, + { + "epoch": 2.9415654813529923, + "grad_norm": 0.19172607362270355, + "learning_rate": 3.9097635282170325e-07, + "loss": 0.0011, + "step": 27133 + }, + { + "epoch": 2.941673894189072, + "grad_norm": 0.3068205416202545, + "learning_rate": 3.9025097925431597e-07, + "loss": 0.0013, + "step": 27134 + }, + { + "epoch": 2.941782307025152, + "grad_norm": 0.0796893909573555, + "learning_rate": 3.895256056869288e-07, + "loss": 0.0006, + "step": 27135 + }, + { + "epoch": 2.9418907198612314, + "grad_norm": 5.430551052093506, + "learning_rate": 3.888002321195416e-07, + "loss": 0.0016, + "step": 27136 + }, + { + "epoch": 2.9419991326973114, + "grad_norm": 3.0582683086395264, + "learning_rate": 3.880748585521544e-07, + "loss": 0.0608, + "step": 27137 + }, + { + "epoch": 2.942107545533391, + "grad_norm": 5.862838268280029, + "learning_rate": 3.873494849847672e-07, + "loss": 0.014, + "step": 27138 + }, + { + "epoch": 2.942215958369471, + "grad_norm": 12.598569869995117, + "learning_rate": 3.8662411141738e-07, + "loss": 0.0285, + "step": 27139 + }, + { + "epoch": 2.942324371205551, + "grad_norm": 1.3637497425079346, + "learning_rate": 3.858987378499928e-07, + "loss": 0.002, + "step": 27140 + }, + { + "epoch": 2.9424327840416304, + "grad_norm": 1.432892084121704, + "learning_rate": 3.8517336428260553e-07, + "loss": 0.0275, + "step": 27141 + }, + { + "epoch": 2.9425411968777104, + "grad_norm": 1.99241042137146, + "learning_rate": 3.8444799071521836e-07, + "loss": 0.0009, + "step": 27142 + }, + { + "epoch": 2.94264960971379, + "grad_norm": 2.9531710147857666, + "learning_rate": 3.8372261714783114e-07, + "loss": 0.006, + "step": 27143 + }, + { + "epoch": 2.94275802254987, + "grad_norm": 0.6783053874969482, + "learning_rate": 3.8299724358044397e-07, + "loss": 0.0023, + "step": 27144 + }, + { + "epoch": 2.94286643538595, + "grad_norm": 1.2150331735610962, + "learning_rate": 3.8227187001305675e-07, + "loss": 0.006, + "step": 27145 + }, + { + "epoch": 2.9429748482220295, + "grad_norm": 0.12091359496116638, + "learning_rate": 3.815464964456696e-07, + "loss": 0.0007, + "step": 27146 + }, + { + "epoch": 2.943083261058109, + "grad_norm": 7.928192615509033, + "learning_rate": 3.8082112287828236e-07, + "loss": 0.006, + "step": 27147 + }, + { + "epoch": 2.943191673894189, + "grad_norm": 2.945091724395752, + "learning_rate": 3.8009574931089514e-07, + "loss": 0.0451, + "step": 27148 + }, + { + "epoch": 2.943300086730269, + "grad_norm": 14.716750144958496, + "learning_rate": 3.793703757435079e-07, + "loss": 0.0123, + "step": 27149 + }, + { + "epoch": 2.9434084995663485, + "grad_norm": 0.4086664617061615, + "learning_rate": 3.7864500217612075e-07, + "loss": 0.0017, + "step": 27150 + }, + { + "epoch": 2.9435169124024285, + "grad_norm": 0.05910641327500343, + "learning_rate": 3.7791962860873353e-07, + "loss": 0.0003, + "step": 27151 + }, + { + "epoch": 2.943625325238508, + "grad_norm": 3.1221976280212402, + "learning_rate": 3.7719425504134636e-07, + "loss": 0.0315, + "step": 27152 + }, + { + "epoch": 2.943733738074588, + "grad_norm": 1.7423259019851685, + "learning_rate": 3.7646888147395914e-07, + "loss": 0.0036, + "step": 27153 + }, + { + "epoch": 2.943842150910668, + "grad_norm": 1.252463936805725, + "learning_rate": 3.7574350790657197e-07, + "loss": 0.0268, + "step": 27154 + }, + { + "epoch": 2.9439505637467476, + "grad_norm": 7.43550443649292, + "learning_rate": 3.750181343391847e-07, + "loss": 0.0237, + "step": 27155 + }, + { + "epoch": 2.9440589765828276, + "grad_norm": 0.6779254674911499, + "learning_rate": 3.742927607717975e-07, + "loss": 0.0022, + "step": 27156 + }, + { + "epoch": 2.944167389418907, + "grad_norm": 4.097908020019531, + "learning_rate": 3.735673872044103e-07, + "loss": 0.0103, + "step": 27157 + }, + { + "epoch": 2.944275802254987, + "grad_norm": 0.048346392810344696, + "learning_rate": 3.728420136370231e-07, + "loss": 0.0003, + "step": 27158 + }, + { + "epoch": 2.9443842150910666, + "grad_norm": 4.854616641998291, + "learning_rate": 3.721166400696359e-07, + "loss": 0.0278, + "step": 27159 + }, + { + "epoch": 2.9444926279271466, + "grad_norm": 2.7060959339141846, + "learning_rate": 3.713912665022487e-07, + "loss": 0.0027, + "step": 27160 + }, + { + "epoch": 2.944601040763226, + "grad_norm": 0.9900509119033813, + "learning_rate": 3.7066589293486147e-07, + "loss": 0.0034, + "step": 27161 + }, + { + "epoch": 2.944709453599306, + "grad_norm": 0.1224762424826622, + "learning_rate": 3.6994051936747425e-07, + "loss": 0.0008, + "step": 27162 + }, + { + "epoch": 2.944817866435386, + "grad_norm": 1.8007019758224487, + "learning_rate": 3.692151458000871e-07, + "loss": 0.0067, + "step": 27163 + }, + { + "epoch": 2.9449262792714657, + "grad_norm": 5.385641574859619, + "learning_rate": 3.6848977223269986e-07, + "loss": 0.0108, + "step": 27164 + }, + { + "epoch": 2.9450346921075456, + "grad_norm": 2.3675291538238525, + "learning_rate": 3.677643986653127e-07, + "loss": 0.0043, + "step": 27165 + }, + { + "epoch": 2.945143104943625, + "grad_norm": 2.023930311203003, + "learning_rate": 3.6703902509792547e-07, + "loss": 0.0463, + "step": 27166 + }, + { + "epoch": 2.945251517779705, + "grad_norm": 0.2842613458633423, + "learning_rate": 3.663136515305383e-07, + "loss": 0.0019, + "step": 27167 + }, + { + "epoch": 2.945359930615785, + "grad_norm": 4.481351852416992, + "learning_rate": 3.6558827796315103e-07, + "loss": 0.0068, + "step": 27168 + }, + { + "epoch": 2.9454683434518647, + "grad_norm": 0.10584912449121475, + "learning_rate": 3.648629043957638e-07, + "loss": 0.0006, + "step": 27169 + }, + { + "epoch": 2.9455767562879442, + "grad_norm": 11.198163986206055, + "learning_rate": 3.6413753082837664e-07, + "loss": 0.0049, + "step": 27170 + }, + { + "epoch": 2.9456851691240242, + "grad_norm": 0.09681900590658188, + "learning_rate": 3.634121572609894e-07, + "loss": 0.0007, + "step": 27171 + }, + { + "epoch": 2.945793581960104, + "grad_norm": 1.614296555519104, + "learning_rate": 3.6268678369360225e-07, + "loss": 0.0043, + "step": 27172 + }, + { + "epoch": 2.9459019947961838, + "grad_norm": 1.2772377729415894, + "learning_rate": 3.6196141012621503e-07, + "loss": 0.0031, + "step": 27173 + }, + { + "epoch": 2.9460104076322637, + "grad_norm": 0.02439403347671032, + "learning_rate": 3.6123603655882786e-07, + "loss": 0.0002, + "step": 27174 + }, + { + "epoch": 2.9461188204683433, + "grad_norm": 0.04139251261949539, + "learning_rate": 3.605106629914406e-07, + "loss": 0.0002, + "step": 27175 + }, + { + "epoch": 2.9462272333044233, + "grad_norm": 0.3711796998977661, + "learning_rate": 3.597852894240534e-07, + "loss": 0.002, + "step": 27176 + }, + { + "epoch": 2.9463356461405033, + "grad_norm": 2.0319669246673584, + "learning_rate": 3.590599158566662e-07, + "loss": 0.0016, + "step": 27177 + }, + { + "epoch": 2.946444058976583, + "grad_norm": 0.08861920982599258, + "learning_rate": 3.5833454228927903e-07, + "loss": 0.0006, + "step": 27178 + }, + { + "epoch": 2.9465524718126628, + "grad_norm": 3.528947353363037, + "learning_rate": 3.576091687218918e-07, + "loss": 0.0308, + "step": 27179 + }, + { + "epoch": 2.9466608846487423, + "grad_norm": 0.621052622795105, + "learning_rate": 3.5688379515450464e-07, + "loss": 0.0022, + "step": 27180 + }, + { + "epoch": 2.9467692974848223, + "grad_norm": 0.0816555842757225, + "learning_rate": 3.561584215871174e-07, + "loss": 0.0006, + "step": 27181 + }, + { + "epoch": 2.946877710320902, + "grad_norm": 0.026510607451200485, + "learning_rate": 3.5543304801973014e-07, + "loss": 0.0003, + "step": 27182 + }, + { + "epoch": 2.946986123156982, + "grad_norm": 0.23875078558921814, + "learning_rate": 3.54707674452343e-07, + "loss": 0.0009, + "step": 27183 + }, + { + "epoch": 2.9470945359930614, + "grad_norm": 0.46486496925354004, + "learning_rate": 3.5398230088495575e-07, + "loss": 0.001, + "step": 27184 + }, + { + "epoch": 2.9472029488291414, + "grad_norm": 1.764094352722168, + "learning_rate": 3.532569273175686e-07, + "loss": 0.037, + "step": 27185 + }, + { + "epoch": 2.9473113616652213, + "grad_norm": 0.04566163942217827, + "learning_rate": 3.5253155375018136e-07, + "loss": 0.0003, + "step": 27186 + }, + { + "epoch": 2.947419774501301, + "grad_norm": 5.316723346710205, + "learning_rate": 3.518061801827942e-07, + "loss": 0.0303, + "step": 27187 + }, + { + "epoch": 2.947528187337381, + "grad_norm": 0.2365449219942093, + "learning_rate": 3.510808066154069e-07, + "loss": 0.0008, + "step": 27188 + }, + { + "epoch": 2.9476366001734604, + "grad_norm": 0.5599775910377502, + "learning_rate": 3.5035543304801975e-07, + "loss": 0.001, + "step": 27189 + }, + { + "epoch": 2.9477450130095404, + "grad_norm": 0.20699813961982727, + "learning_rate": 3.4963005948063253e-07, + "loss": 0.0008, + "step": 27190 + }, + { + "epoch": 2.9478534258456204, + "grad_norm": 6.127868175506592, + "learning_rate": 3.4890468591324536e-07, + "loss": 0.0249, + "step": 27191 + }, + { + "epoch": 2.9479618386817, + "grad_norm": 1.2928341627120972, + "learning_rate": 3.4817931234585814e-07, + "loss": 0.0031, + "step": 27192 + }, + { + "epoch": 2.9480702515177795, + "grad_norm": 2.7452170848846436, + "learning_rate": 3.4745393877847097e-07, + "loss": 0.002, + "step": 27193 + }, + { + "epoch": 2.9481786643538594, + "grad_norm": 0.026279963552951813, + "learning_rate": 3.4672856521108375e-07, + "loss": 0.0003, + "step": 27194 + }, + { + "epoch": 2.9482870771899394, + "grad_norm": 2.0604946613311768, + "learning_rate": 3.460031916436965e-07, + "loss": 0.0181, + "step": 27195 + }, + { + "epoch": 2.948395490026019, + "grad_norm": 0.45418232679367065, + "learning_rate": 3.452778180763093e-07, + "loss": 0.0017, + "step": 27196 + }, + { + "epoch": 2.948503902862099, + "grad_norm": 6.2881178855896, + "learning_rate": 3.445524445089221e-07, + "loss": 0.0431, + "step": 27197 + }, + { + "epoch": 2.9486123156981785, + "grad_norm": 0.5077844262123108, + "learning_rate": 3.438270709415349e-07, + "loss": 0.0026, + "step": 27198 + }, + { + "epoch": 2.9487207285342585, + "grad_norm": 3.1792142391204834, + "learning_rate": 3.431016973741477e-07, + "loss": 0.0413, + "step": 27199 + }, + { + "epoch": 2.9488291413703385, + "grad_norm": 0.06628753244876862, + "learning_rate": 3.4237632380676053e-07, + "loss": 0.0006, + "step": 27200 + }, + { + "epoch": 2.948937554206418, + "grad_norm": 7.974021911621094, + "learning_rate": 3.4165095023937336e-07, + "loss": 0.01, + "step": 27201 + }, + { + "epoch": 2.949045967042498, + "grad_norm": 4.659401893615723, + "learning_rate": 3.409255766719861e-07, + "loss": 0.0083, + "step": 27202 + }, + { + "epoch": 2.9491543798785775, + "grad_norm": 0.14531110227108002, + "learning_rate": 3.4020020310459887e-07, + "loss": 0.0004, + "step": 27203 + }, + { + "epoch": 2.9492627927146575, + "grad_norm": 10.449695587158203, + "learning_rate": 3.394748295372117e-07, + "loss": 0.0139, + "step": 27204 + }, + { + "epoch": 2.949371205550737, + "grad_norm": 1.1488354206085205, + "learning_rate": 3.387494559698245e-07, + "loss": 0.0337, + "step": 27205 + }, + { + "epoch": 2.949479618386817, + "grad_norm": 1.8592004776000977, + "learning_rate": 3.380240824024373e-07, + "loss": 0.0157, + "step": 27206 + }, + { + "epoch": 2.9495880312228966, + "grad_norm": 3.802434206008911, + "learning_rate": 3.372987088350501e-07, + "loss": 0.019, + "step": 27207 + }, + { + "epoch": 2.9496964440589766, + "grad_norm": 1.6338295936584473, + "learning_rate": 3.365733352676629e-07, + "loss": 0.0044, + "step": 27208 + }, + { + "epoch": 2.9498048568950566, + "grad_norm": 14.08557415008545, + "learning_rate": 3.3584796170027564e-07, + "loss": 0.0354, + "step": 27209 + }, + { + "epoch": 2.949913269731136, + "grad_norm": 0.5759569406509399, + "learning_rate": 3.351225881328885e-07, + "loss": 0.002, + "step": 27210 + }, + { + "epoch": 2.950021682567216, + "grad_norm": 4.4775800704956055, + "learning_rate": 3.3439721456550125e-07, + "loss": 0.0265, + "step": 27211 + }, + { + "epoch": 2.9501300954032956, + "grad_norm": 0.17458085715770721, + "learning_rate": 3.336718409981141e-07, + "loss": 0.0006, + "step": 27212 + }, + { + "epoch": 2.9502385082393756, + "grad_norm": 4.532145023345947, + "learning_rate": 3.3294646743072686e-07, + "loss": 0.0049, + "step": 27213 + }, + { + "epoch": 2.9503469210754556, + "grad_norm": 6.207526683807373, + "learning_rate": 3.322210938633397e-07, + "loss": 0.0194, + "step": 27214 + }, + { + "epoch": 2.950455333911535, + "grad_norm": 3.7764079570770264, + "learning_rate": 3.314957202959524e-07, + "loss": 0.0177, + "step": 27215 + }, + { + "epoch": 2.9505637467476147, + "grad_norm": 0.2441713660955429, + "learning_rate": 3.307703467285652e-07, + "loss": 0.001, + "step": 27216 + }, + { + "epoch": 2.9506721595836947, + "grad_norm": 0.7460384368896484, + "learning_rate": 3.3004497316117803e-07, + "loss": 0.002, + "step": 27217 + }, + { + "epoch": 2.9507805724197746, + "grad_norm": 0.24826882779598236, + "learning_rate": 3.293195995937908e-07, + "loss": 0.0088, + "step": 27218 + }, + { + "epoch": 2.950888985255854, + "grad_norm": 1.4224131107330322, + "learning_rate": 3.2859422602640364e-07, + "loss": 0.0048, + "step": 27219 + }, + { + "epoch": 2.950997398091934, + "grad_norm": 0.08251930773258209, + "learning_rate": 3.278688524590164e-07, + "loss": 0.0008, + "step": 27220 + }, + { + "epoch": 2.9511058109280137, + "grad_norm": 0.09334186464548111, + "learning_rate": 3.2714347889162925e-07, + "loss": 0.0006, + "step": 27221 + }, + { + "epoch": 2.9512142237640937, + "grad_norm": 0.7038339376449585, + "learning_rate": 3.26418105324242e-07, + "loss": 0.0011, + "step": 27222 + }, + { + "epoch": 2.9513226366001737, + "grad_norm": 0.029168004170060158, + "learning_rate": 3.256927317568548e-07, + "loss": 0.0002, + "step": 27223 + }, + { + "epoch": 2.9514310494362532, + "grad_norm": 4.712735176086426, + "learning_rate": 3.249673581894676e-07, + "loss": 0.0339, + "step": 27224 + }, + { + "epoch": 2.9515394622723328, + "grad_norm": 0.43887844681739807, + "learning_rate": 3.242419846220804e-07, + "loss": 0.0006, + "step": 27225 + }, + { + "epoch": 2.9516478751084128, + "grad_norm": 0.049467023462057114, + "learning_rate": 3.235166110546932e-07, + "loss": 0.0002, + "step": 27226 + }, + { + "epoch": 2.9517562879444927, + "grad_norm": 0.012567861005663872, + "learning_rate": 3.2279123748730603e-07, + "loss": 0.0002, + "step": 27227 + }, + { + "epoch": 2.9518647007805723, + "grad_norm": 0.4974375069141388, + "learning_rate": 3.220658639199188e-07, + "loss": 0.0028, + "step": 27228 + }, + { + "epoch": 2.9519731136166523, + "grad_norm": 0.08901778608560562, + "learning_rate": 3.2134049035253153e-07, + "loss": 0.0005, + "step": 27229 + }, + { + "epoch": 2.952081526452732, + "grad_norm": 2.0749762058258057, + "learning_rate": 3.2061511678514437e-07, + "loss": 0.0315, + "step": 27230 + }, + { + "epoch": 2.952189939288812, + "grad_norm": 4.218719005584717, + "learning_rate": 3.1988974321775714e-07, + "loss": 0.0076, + "step": 27231 + }, + { + "epoch": 2.952298352124892, + "grad_norm": 6.15308141708374, + "learning_rate": 3.1916436965037e-07, + "loss": 0.0096, + "step": 27232 + }, + { + "epoch": 2.9524067649609713, + "grad_norm": 0.012526293285191059, + "learning_rate": 3.1843899608298275e-07, + "loss": 0.0001, + "step": 27233 + }, + { + "epoch": 2.9525151777970513, + "grad_norm": 0.8578619956970215, + "learning_rate": 3.177136225155956e-07, + "loss": 0.0016, + "step": 27234 + }, + { + "epoch": 2.952623590633131, + "grad_norm": 2.1572341918945312, + "learning_rate": 3.1698824894820836e-07, + "loss": 0.0031, + "step": 27235 + }, + { + "epoch": 2.952732003469211, + "grad_norm": 1.9503742456436157, + "learning_rate": 3.1626287538082114e-07, + "loss": 0.0154, + "step": 27236 + }, + { + "epoch": 2.952840416305291, + "grad_norm": 1.911254644393921, + "learning_rate": 3.155375018134339e-07, + "loss": 0.0171, + "step": 27237 + }, + { + "epoch": 2.9529488291413704, + "grad_norm": 30.284887313842773, + "learning_rate": 3.1481212824604675e-07, + "loss": 0.0544, + "step": 27238 + }, + { + "epoch": 2.95305724197745, + "grad_norm": 0.28427162766456604, + "learning_rate": 3.1408675467865953e-07, + "loss": 0.0021, + "step": 27239 + }, + { + "epoch": 2.95316565481353, + "grad_norm": 3.6873912811279297, + "learning_rate": 3.1336138111127236e-07, + "loss": 0.024, + "step": 27240 + }, + { + "epoch": 2.95327406764961, + "grad_norm": 1.1204848289489746, + "learning_rate": 3.1263600754388514e-07, + "loss": 0.0066, + "step": 27241 + }, + { + "epoch": 2.9533824804856894, + "grad_norm": 0.3437441289424896, + "learning_rate": 3.119106339764979e-07, + "loss": 0.0014, + "step": 27242 + }, + { + "epoch": 2.9534908933217694, + "grad_norm": 0.015208825469017029, + "learning_rate": 3.1118526040911075e-07, + "loss": 0.0001, + "step": 27243 + }, + { + "epoch": 2.953599306157849, + "grad_norm": 0.46407029032707214, + "learning_rate": 3.104598868417235e-07, + "loss": 0.0006, + "step": 27244 + }, + { + "epoch": 2.953707718993929, + "grad_norm": 0.06313348561525345, + "learning_rate": 3.097345132743363e-07, + "loss": 0.0005, + "step": 27245 + }, + { + "epoch": 2.953816131830009, + "grad_norm": 0.11481307446956635, + "learning_rate": 3.090091397069491e-07, + "loss": 0.0006, + "step": 27246 + }, + { + "epoch": 2.9539245446660884, + "grad_norm": 0.06556840240955353, + "learning_rate": 3.082837661395619e-07, + "loss": 0.0005, + "step": 27247 + }, + { + "epoch": 2.954032957502168, + "grad_norm": 1.4082679748535156, + "learning_rate": 3.075583925721747e-07, + "loss": 0.0046, + "step": 27248 + }, + { + "epoch": 2.954141370338248, + "grad_norm": 0.03459394350647926, + "learning_rate": 3.068330190047875e-07, + "loss": 0.0002, + "step": 27249 + }, + { + "epoch": 2.954249783174328, + "grad_norm": 0.26599183678627014, + "learning_rate": 3.061076454374003e-07, + "loss": 0.0007, + "step": 27250 + }, + { + "epoch": 2.9543581960104075, + "grad_norm": 1.1431547403335571, + "learning_rate": 3.053822718700131e-07, + "loss": 0.0018, + "step": 27251 + }, + { + "epoch": 2.9544666088464875, + "grad_norm": 1.4760345220565796, + "learning_rate": 3.0465689830262587e-07, + "loss": 0.0027, + "step": 27252 + }, + { + "epoch": 2.954575021682567, + "grad_norm": 0.0674978718161583, + "learning_rate": 3.039315247352387e-07, + "loss": 0.0008, + "step": 27253 + }, + { + "epoch": 2.954683434518647, + "grad_norm": 0.5320195555686951, + "learning_rate": 3.032061511678515e-07, + "loss": 0.0009, + "step": 27254 + }, + { + "epoch": 2.954791847354727, + "grad_norm": 0.24089586734771729, + "learning_rate": 3.0248077760046426e-07, + "loss": 0.0008, + "step": 27255 + }, + { + "epoch": 2.9549002601908065, + "grad_norm": 2.681908369064331, + "learning_rate": 3.017554040330771e-07, + "loss": 0.0112, + "step": 27256 + }, + { + "epoch": 2.9550086730268865, + "grad_norm": 0.18053270876407623, + "learning_rate": 3.0103003046568987e-07, + "loss": 0.0009, + "step": 27257 + }, + { + "epoch": 2.955117085862966, + "grad_norm": 1.7291544675827026, + "learning_rate": 3.0030465689830264e-07, + "loss": 0.0283, + "step": 27258 + }, + { + "epoch": 2.955225498699046, + "grad_norm": 6.283664703369141, + "learning_rate": 2.995792833309154e-07, + "loss": 0.0101, + "step": 27259 + }, + { + "epoch": 2.955333911535126, + "grad_norm": 5.591394901275635, + "learning_rate": 2.9885390976352826e-07, + "loss": 0.02, + "step": 27260 + }, + { + "epoch": 2.9554423243712056, + "grad_norm": 4.79437255859375, + "learning_rate": 2.9812853619614103e-07, + "loss": 0.0386, + "step": 27261 + }, + { + "epoch": 2.955550737207285, + "grad_norm": 0.2724479138851166, + "learning_rate": 2.974031626287538e-07, + "loss": 0.0008, + "step": 27262 + }, + { + "epoch": 2.955659150043365, + "grad_norm": 2.685685157775879, + "learning_rate": 2.9667778906136664e-07, + "loss": 0.0309, + "step": 27263 + }, + { + "epoch": 2.955767562879445, + "grad_norm": 0.7268714308738708, + "learning_rate": 2.959524154939794e-07, + "loss": 0.0024, + "step": 27264 + }, + { + "epoch": 2.9558759757155246, + "grad_norm": 0.31511425971984863, + "learning_rate": 2.952270419265922e-07, + "loss": 0.0013, + "step": 27265 + }, + { + "epoch": 2.9559843885516046, + "grad_norm": 4.185439109802246, + "learning_rate": 2.9450166835920503e-07, + "loss": 0.0093, + "step": 27266 + }, + { + "epoch": 2.956092801387684, + "grad_norm": 0.08576653897762299, + "learning_rate": 2.937762947918178e-07, + "loss": 0.0007, + "step": 27267 + }, + { + "epoch": 2.956201214223764, + "grad_norm": 0.026560883969068527, + "learning_rate": 2.930509212244306e-07, + "loss": 0.0002, + "step": 27268 + }, + { + "epoch": 2.956309627059844, + "grad_norm": 0.3269653618335724, + "learning_rate": 2.923255476570434e-07, + "loss": 0.0012, + "step": 27269 + }, + { + "epoch": 2.9564180398959237, + "grad_norm": 3.810513496398926, + "learning_rate": 2.916001740896562e-07, + "loss": 0.0029, + "step": 27270 + }, + { + "epoch": 2.956526452732003, + "grad_norm": 1.378448247909546, + "learning_rate": 2.90874800522269e-07, + "loss": 0.0036, + "step": 27271 + }, + { + "epoch": 2.956634865568083, + "grad_norm": 0.5954757928848267, + "learning_rate": 2.9014942695488176e-07, + "loss": 0.002, + "step": 27272 + }, + { + "epoch": 2.956743278404163, + "grad_norm": 0.14178478717803955, + "learning_rate": 2.894240533874946e-07, + "loss": 0.0006, + "step": 27273 + }, + { + "epoch": 2.9568516912402427, + "grad_norm": 1.1465933322906494, + "learning_rate": 2.8869867982010737e-07, + "loss": 0.0257, + "step": 27274 + }, + { + "epoch": 2.9569601040763227, + "grad_norm": 0.7399756908416748, + "learning_rate": 2.8797330625272015e-07, + "loss": 0.0012, + "step": 27275 + }, + { + "epoch": 2.9570685169124022, + "grad_norm": 0.01290583610534668, + "learning_rate": 2.87247932685333e-07, + "loss": 0.0001, + "step": 27276 + }, + { + "epoch": 2.9571769297484822, + "grad_norm": 0.07605700194835663, + "learning_rate": 2.8652255911794576e-07, + "loss": 0.0006, + "step": 27277 + }, + { + "epoch": 2.957285342584562, + "grad_norm": 2.5640652179718018, + "learning_rate": 2.8579718555055854e-07, + "loss": 0.002, + "step": 27278 + }, + { + "epoch": 2.9573937554206418, + "grad_norm": 1.1033517122268677, + "learning_rate": 2.8507181198317137e-07, + "loss": 0.0025, + "step": 27279 + }, + { + "epoch": 2.9575021682567217, + "grad_norm": 0.012131727300584316, + "learning_rate": 2.8434643841578415e-07, + "loss": 0.0001, + "step": 27280 + }, + { + "epoch": 2.9576105810928013, + "grad_norm": 0.07145832479000092, + "learning_rate": 2.836210648483969e-07, + "loss": 0.0006, + "step": 27281 + }, + { + "epoch": 2.9577189939288813, + "grad_norm": 4.797216892242432, + "learning_rate": 2.8289569128100976e-07, + "loss": 0.0145, + "step": 27282 + }, + { + "epoch": 2.957827406764961, + "grad_norm": 2.700777769088745, + "learning_rate": 2.8217031771362253e-07, + "loss": 0.0269, + "step": 27283 + }, + { + "epoch": 2.957935819601041, + "grad_norm": 1.6787809133529663, + "learning_rate": 2.8144494414623537e-07, + "loss": 0.003, + "step": 27284 + }, + { + "epoch": 2.9580442324371203, + "grad_norm": 7.615391731262207, + "learning_rate": 2.8071957057884815e-07, + "loss": 0.0476, + "step": 27285 + }, + { + "epoch": 2.9581526452732003, + "grad_norm": 0.15592767298221588, + "learning_rate": 2.799941970114609e-07, + "loss": 0.0011, + "step": 27286 + }, + { + "epoch": 2.9582610581092803, + "grad_norm": 0.09872270375490189, + "learning_rate": 2.7926882344407376e-07, + "loss": 0.0004, + "step": 27287 + }, + { + "epoch": 2.95836947094536, + "grad_norm": 2.6921157836914062, + "learning_rate": 2.785434498766865e-07, + "loss": 0.0078, + "step": 27288 + }, + { + "epoch": 2.95847788378144, + "grad_norm": 0.8222032785415649, + "learning_rate": 2.778180763092993e-07, + "loss": 0.0034, + "step": 27289 + }, + { + "epoch": 2.9585862966175194, + "grad_norm": 6.75652551651001, + "learning_rate": 2.770927027419121e-07, + "loss": 0.0259, + "step": 27290 + }, + { + "epoch": 2.9586947094535994, + "grad_norm": 13.753278732299805, + "learning_rate": 2.763673291745249e-07, + "loss": 0.0807, + "step": 27291 + }, + { + "epoch": 2.9588031222896793, + "grad_norm": 3.872405767440796, + "learning_rate": 2.756419556071377e-07, + "loss": 0.0048, + "step": 27292 + }, + { + "epoch": 2.958911535125759, + "grad_norm": 0.05430327355861664, + "learning_rate": 2.749165820397505e-07, + "loss": 0.0004, + "step": 27293 + }, + { + "epoch": 2.9590199479618384, + "grad_norm": 2.4944233894348145, + "learning_rate": 2.741912084723633e-07, + "loss": 0.0572, + "step": 27294 + }, + { + "epoch": 2.9591283607979184, + "grad_norm": 0.6446067690849304, + "learning_rate": 2.734658349049761e-07, + "loss": 0.0019, + "step": 27295 + }, + { + "epoch": 2.9592367736339984, + "grad_norm": 0.7316578030586243, + "learning_rate": 2.7274046133758887e-07, + "loss": 0.0027, + "step": 27296 + }, + { + "epoch": 2.959345186470078, + "grad_norm": 5.763298511505127, + "learning_rate": 2.720150877702017e-07, + "loss": 0.0194, + "step": 27297 + }, + { + "epoch": 2.959453599306158, + "grad_norm": 0.1299491971731186, + "learning_rate": 2.712897142028145e-07, + "loss": 0.0008, + "step": 27298 + }, + { + "epoch": 2.9595620121422375, + "grad_norm": 0.15518128871917725, + "learning_rate": 2.7056434063542726e-07, + "loss": 0.001, + "step": 27299 + }, + { + "epoch": 2.9596704249783174, + "grad_norm": 0.04845162108540535, + "learning_rate": 2.698389670680401e-07, + "loss": 0.0002, + "step": 27300 + }, + { + "epoch": 2.9597788378143974, + "grad_norm": 4.254210948944092, + "learning_rate": 2.6911359350065287e-07, + "loss": 0.0579, + "step": 27301 + }, + { + "epoch": 2.959887250650477, + "grad_norm": 3.0622572898864746, + "learning_rate": 2.6838821993326565e-07, + "loss": 0.0529, + "step": 27302 + }, + { + "epoch": 2.959995663486557, + "grad_norm": 0.026323385536670685, + "learning_rate": 2.676628463658784e-07, + "loss": 0.0002, + "step": 27303 + }, + { + "epoch": 2.9601040763226365, + "grad_norm": 2.8617584705352783, + "learning_rate": 2.6693747279849126e-07, + "loss": 0.031, + "step": 27304 + }, + { + "epoch": 2.9602124891587165, + "grad_norm": 0.28856155276298523, + "learning_rate": 2.6621209923110404e-07, + "loss": 0.001, + "step": 27305 + }, + { + "epoch": 2.960320901994796, + "grad_norm": 0.07230586558580399, + "learning_rate": 2.654867256637168e-07, + "loss": 0.0005, + "step": 27306 + }, + { + "epoch": 2.960429314830876, + "grad_norm": 0.37441736459732056, + "learning_rate": 2.6476135209632965e-07, + "loss": 0.0014, + "step": 27307 + }, + { + "epoch": 2.9605377276669556, + "grad_norm": 0.06179642304778099, + "learning_rate": 2.640359785289424e-07, + "loss": 0.0003, + "step": 27308 + }, + { + "epoch": 2.9606461405030355, + "grad_norm": 0.07708027958869934, + "learning_rate": 2.633106049615552e-07, + "loss": 0.0002, + "step": 27309 + }, + { + "epoch": 2.9607545533391155, + "grad_norm": 7.667144298553467, + "learning_rate": 2.6258523139416804e-07, + "loss": 0.0091, + "step": 27310 + }, + { + "epoch": 2.960862966175195, + "grad_norm": 0.005850865505635738, + "learning_rate": 2.618598578267808e-07, + "loss": 0.0001, + "step": 27311 + }, + { + "epoch": 2.960971379011275, + "grad_norm": 0.2506358027458191, + "learning_rate": 2.611344842593936e-07, + "loss": 0.0005, + "step": 27312 + }, + { + "epoch": 2.9610797918473546, + "grad_norm": 3.719691276550293, + "learning_rate": 2.604091106920064e-07, + "loss": 0.0511, + "step": 27313 + }, + { + "epoch": 2.9611882046834346, + "grad_norm": 0.031203363090753555, + "learning_rate": 2.596837371246192e-07, + "loss": 0.0003, + "step": 27314 + }, + { + "epoch": 2.9612966175195146, + "grad_norm": 5.78015661239624, + "learning_rate": 2.58958363557232e-07, + "loss": 0.0206, + "step": 27315 + }, + { + "epoch": 2.961405030355594, + "grad_norm": 5.6153082847595215, + "learning_rate": 2.5823298998984476e-07, + "loss": 0.0073, + "step": 27316 + }, + { + "epoch": 2.9615134431916736, + "grad_norm": 0.027763815596699715, + "learning_rate": 2.575076164224576e-07, + "loss": 0.0002, + "step": 27317 + }, + { + "epoch": 2.9616218560277536, + "grad_norm": 0.08166281878948212, + "learning_rate": 2.567822428550704e-07, + "loss": 0.0005, + "step": 27318 + }, + { + "epoch": 2.9617302688638336, + "grad_norm": 1.4625076055526733, + "learning_rate": 2.5605686928768315e-07, + "loss": 0.0029, + "step": 27319 + }, + { + "epoch": 2.961838681699913, + "grad_norm": 0.047837577760219574, + "learning_rate": 2.55331495720296e-07, + "loss": 0.0002, + "step": 27320 + }, + { + "epoch": 2.961947094535993, + "grad_norm": 0.07811019569635391, + "learning_rate": 2.5460612215290876e-07, + "loss": 0.0005, + "step": 27321 + }, + { + "epoch": 2.9620555073720727, + "grad_norm": 0.10774100571870804, + "learning_rate": 2.5388074858552154e-07, + "loss": 0.0004, + "step": 27322 + }, + { + "epoch": 2.9621639202081527, + "grad_norm": 3.636648654937744, + "learning_rate": 2.5315537501813437e-07, + "loss": 0.0272, + "step": 27323 + }, + { + "epoch": 2.9622723330442327, + "grad_norm": 0.1190570816397667, + "learning_rate": 2.5243000145074715e-07, + "loss": 0.0005, + "step": 27324 + }, + { + "epoch": 2.962380745880312, + "grad_norm": 0.09158974140882492, + "learning_rate": 2.5170462788335993e-07, + "loss": 0.0006, + "step": 27325 + }, + { + "epoch": 2.962489158716392, + "grad_norm": 4.795073986053467, + "learning_rate": 2.5097925431597276e-07, + "loss": 0.0203, + "step": 27326 + }, + { + "epoch": 2.9625975715524717, + "grad_norm": 0.5439139008522034, + "learning_rate": 2.5025388074858554e-07, + "loss": 0.0012, + "step": 27327 + }, + { + "epoch": 2.9627059843885517, + "grad_norm": 19.840145111083984, + "learning_rate": 2.4952850718119837e-07, + "loss": 0.0084, + "step": 27328 + }, + { + "epoch": 2.9628143972246312, + "grad_norm": 1.5950571298599243, + "learning_rate": 2.4880313361381115e-07, + "loss": 0.0057, + "step": 27329 + }, + { + "epoch": 2.9629228100607112, + "grad_norm": 4.179330825805664, + "learning_rate": 2.480777600464239e-07, + "loss": 0.0397, + "step": 27330 + }, + { + "epoch": 2.9630312228967908, + "grad_norm": 7.303222179412842, + "learning_rate": 2.4735238647903676e-07, + "loss": 0.0252, + "step": 27331 + }, + { + "epoch": 2.9631396357328708, + "grad_norm": 0.1381135731935501, + "learning_rate": 2.466270129116495e-07, + "loss": 0.0008, + "step": 27332 + }, + { + "epoch": 2.9632480485689507, + "grad_norm": 0.055115360766649246, + "learning_rate": 2.459016393442623e-07, + "loss": 0.0006, + "step": 27333 + }, + { + "epoch": 2.9633564614050303, + "grad_norm": 0.19239473342895508, + "learning_rate": 2.451762657768751e-07, + "loss": 0.0011, + "step": 27334 + }, + { + "epoch": 2.9634648742411103, + "grad_norm": 6.079241752624512, + "learning_rate": 2.444508922094879e-07, + "loss": 0.0349, + "step": 27335 + }, + { + "epoch": 2.96357328707719, + "grad_norm": 0.11060057580471039, + "learning_rate": 2.437255186421007e-07, + "loss": 0.0007, + "step": 27336 + }, + { + "epoch": 2.96368169991327, + "grad_norm": 5.282808303833008, + "learning_rate": 2.430001450747135e-07, + "loss": 0.0172, + "step": 27337 + }, + { + "epoch": 2.96379011274935, + "grad_norm": 1.25961172580719, + "learning_rate": 2.422747715073263e-07, + "loss": 0.0022, + "step": 27338 + }, + { + "epoch": 2.9638985255854293, + "grad_norm": 0.22542428970336914, + "learning_rate": 2.415493979399391e-07, + "loss": 0.0015, + "step": 27339 + }, + { + "epoch": 2.964006938421509, + "grad_norm": 4.187071800231934, + "learning_rate": 2.4082402437255187e-07, + "loss": 0.018, + "step": 27340 + }, + { + "epoch": 2.964115351257589, + "grad_norm": 0.050487857311964035, + "learning_rate": 2.400986508051647e-07, + "loss": 0.0003, + "step": 27341 + }, + { + "epoch": 2.964223764093669, + "grad_norm": 0.9190403819084167, + "learning_rate": 2.393732772377775e-07, + "loss": 0.0017, + "step": 27342 + }, + { + "epoch": 2.9643321769297484, + "grad_norm": 0.3665851652622223, + "learning_rate": 2.3864790367039026e-07, + "loss": 0.0009, + "step": 27343 + }, + { + "epoch": 2.9644405897658284, + "grad_norm": 3.3086464405059814, + "learning_rate": 2.3792253010300307e-07, + "loss": 0.0333, + "step": 27344 + }, + { + "epoch": 2.964549002601908, + "grad_norm": 9.204329490661621, + "learning_rate": 2.3719715653561587e-07, + "loss": 0.0062, + "step": 27345 + }, + { + "epoch": 2.964657415437988, + "grad_norm": 0.016751829534769058, + "learning_rate": 2.3647178296822865e-07, + "loss": 0.0002, + "step": 27346 + }, + { + "epoch": 2.964765828274068, + "grad_norm": 1.3164265155792236, + "learning_rate": 2.3574640940084145e-07, + "loss": 0.0015, + "step": 27347 + }, + { + "epoch": 2.9648742411101474, + "grad_norm": 2.466935634613037, + "learning_rate": 2.3502103583345426e-07, + "loss": 0.0081, + "step": 27348 + }, + { + "epoch": 2.9649826539462274, + "grad_norm": 2.0030322074890137, + "learning_rate": 2.3429566226606704e-07, + "loss": 0.0388, + "step": 27349 + }, + { + "epoch": 2.965091066782307, + "grad_norm": 3.1394591331481934, + "learning_rate": 2.3357028869867984e-07, + "loss": 0.0081, + "step": 27350 + }, + { + "epoch": 2.965199479618387, + "grad_norm": 1.0515074729919434, + "learning_rate": 2.3284491513129265e-07, + "loss": 0.0041, + "step": 27351 + }, + { + "epoch": 2.9653078924544665, + "grad_norm": 2.069697856903076, + "learning_rate": 2.3211954156390545e-07, + "loss": 0.0067, + "step": 27352 + }, + { + "epoch": 2.9654163052905465, + "grad_norm": 6.747673988342285, + "learning_rate": 2.313941679965182e-07, + "loss": 0.0387, + "step": 27353 + }, + { + "epoch": 2.965524718126626, + "grad_norm": 1.0863031148910522, + "learning_rate": 2.30668794429131e-07, + "loss": 0.0042, + "step": 27354 + }, + { + "epoch": 2.965633130962706, + "grad_norm": 0.2589901089668274, + "learning_rate": 2.2994342086174382e-07, + "loss": 0.0013, + "step": 27355 + }, + { + "epoch": 2.965741543798786, + "grad_norm": 3.2909233570098877, + "learning_rate": 2.292180472943566e-07, + "loss": 0.0162, + "step": 27356 + }, + { + "epoch": 2.9658499566348655, + "grad_norm": 10.248318672180176, + "learning_rate": 2.284926737269694e-07, + "loss": 0.0443, + "step": 27357 + }, + { + "epoch": 2.9659583694709455, + "grad_norm": 3.1295857429504395, + "learning_rate": 2.277673001595822e-07, + "loss": 0.0145, + "step": 27358 + }, + { + "epoch": 2.966066782307025, + "grad_norm": 1.623734951019287, + "learning_rate": 2.2704192659219498e-07, + "loss": 0.0043, + "step": 27359 + }, + { + "epoch": 2.966175195143105, + "grad_norm": 0.06474211812019348, + "learning_rate": 2.263165530248078e-07, + "loss": 0.0003, + "step": 27360 + }, + { + "epoch": 2.966283607979185, + "grad_norm": 4.390343189239502, + "learning_rate": 2.255911794574206e-07, + "loss": 0.0033, + "step": 27361 + }, + { + "epoch": 2.9663920208152645, + "grad_norm": 0.12452136725187302, + "learning_rate": 2.248658058900334e-07, + "loss": 0.0006, + "step": 27362 + }, + { + "epoch": 2.966500433651344, + "grad_norm": 3.122464179992676, + "learning_rate": 2.2414043232264618e-07, + "loss": 0.0279, + "step": 27363 + }, + { + "epoch": 2.966608846487424, + "grad_norm": 2.821812629699707, + "learning_rate": 2.2341505875525898e-07, + "loss": 0.0322, + "step": 27364 + }, + { + "epoch": 2.966717259323504, + "grad_norm": 0.5652546882629395, + "learning_rate": 2.226896851878718e-07, + "loss": 0.0024, + "step": 27365 + }, + { + "epoch": 2.9668256721595836, + "grad_norm": 0.049571406096220016, + "learning_rate": 2.2196431162048457e-07, + "loss": 0.0004, + "step": 27366 + }, + { + "epoch": 2.9669340849956636, + "grad_norm": 2.1189072132110596, + "learning_rate": 2.2123893805309737e-07, + "loss": 0.0069, + "step": 27367 + }, + { + "epoch": 2.967042497831743, + "grad_norm": 0.7448372840881348, + "learning_rate": 2.2051356448571018e-07, + "loss": 0.0015, + "step": 27368 + }, + { + "epoch": 2.967150910667823, + "grad_norm": 0.5246762633323669, + "learning_rate": 2.1978819091832293e-07, + "loss": 0.0005, + "step": 27369 + }, + { + "epoch": 2.967259323503903, + "grad_norm": 0.07757432758808136, + "learning_rate": 2.1906281735093573e-07, + "loss": 0.0005, + "step": 27370 + }, + { + "epoch": 2.9673677363399826, + "grad_norm": 1.0019516944885254, + "learning_rate": 2.1833744378354854e-07, + "loss": 0.0017, + "step": 27371 + }, + { + "epoch": 2.967476149176062, + "grad_norm": 5.2770304679870605, + "learning_rate": 2.1761207021616134e-07, + "loss": 0.031, + "step": 27372 + }, + { + "epoch": 2.967584562012142, + "grad_norm": 0.14634551107883453, + "learning_rate": 2.1688669664877412e-07, + "loss": 0.0006, + "step": 27373 + }, + { + "epoch": 2.967692974848222, + "grad_norm": 7.82118034362793, + "learning_rate": 2.1616132308138693e-07, + "loss": 0.0076, + "step": 27374 + }, + { + "epoch": 2.9678013876843017, + "grad_norm": 2.708670139312744, + "learning_rate": 2.1543594951399973e-07, + "loss": 0.0241, + "step": 27375 + }, + { + "epoch": 2.9679098005203817, + "grad_norm": 0.13520486652851105, + "learning_rate": 2.147105759466125e-07, + "loss": 0.0011, + "step": 27376 + }, + { + "epoch": 2.968018213356461, + "grad_norm": 0.02158515155315399, + "learning_rate": 2.1398520237922532e-07, + "loss": 0.0002, + "step": 27377 + }, + { + "epoch": 2.968126626192541, + "grad_norm": 1.741455078125, + "learning_rate": 2.1325982881183812e-07, + "loss": 0.0356, + "step": 27378 + }, + { + "epoch": 2.968235039028621, + "grad_norm": 2.6409778594970703, + "learning_rate": 2.1253445524445093e-07, + "loss": 0.0022, + "step": 27379 + }, + { + "epoch": 2.9683434518647007, + "grad_norm": 0.20157112181186676, + "learning_rate": 2.118090816770637e-07, + "loss": 0.001, + "step": 27380 + }, + { + "epoch": 2.9684518647007807, + "grad_norm": 0.4902688264846802, + "learning_rate": 2.110837081096765e-07, + "loss": 0.001, + "step": 27381 + }, + { + "epoch": 2.9685602775368602, + "grad_norm": 0.11040527373552322, + "learning_rate": 2.1035833454228932e-07, + "loss": 0.0007, + "step": 27382 + }, + { + "epoch": 2.9686686903729402, + "grad_norm": 0.059848252683877945, + "learning_rate": 2.0963296097490207e-07, + "loss": 0.0006, + "step": 27383 + }, + { + "epoch": 2.96877710320902, + "grad_norm": 0.32745620608329773, + "learning_rate": 2.0890758740751487e-07, + "loss": 0.0008, + "step": 27384 + }, + { + "epoch": 2.9688855160450998, + "grad_norm": 0.4940800368785858, + "learning_rate": 2.0818221384012768e-07, + "loss": 0.0017, + "step": 27385 + }, + { + "epoch": 2.9689939288811793, + "grad_norm": 10.947332382202148, + "learning_rate": 2.0745684027274046e-07, + "loss": 0.0615, + "step": 27386 + }, + { + "epoch": 2.9691023417172593, + "grad_norm": 0.19054847955703735, + "learning_rate": 2.0673146670535326e-07, + "loss": 0.001, + "step": 27387 + }, + { + "epoch": 2.9692107545533393, + "grad_norm": 2.2103044986724854, + "learning_rate": 2.0600609313796607e-07, + "loss": 0.0128, + "step": 27388 + }, + { + "epoch": 2.969319167389419, + "grad_norm": 0.030540330335497856, + "learning_rate": 2.0528071957057887e-07, + "loss": 0.0003, + "step": 27389 + }, + { + "epoch": 2.969427580225499, + "grad_norm": 0.19361433386802673, + "learning_rate": 2.0455534600319165e-07, + "loss": 0.0007, + "step": 27390 + }, + { + "epoch": 2.9695359930615783, + "grad_norm": 0.004168248735368252, + "learning_rate": 2.0382997243580446e-07, + "loss": 0.0, + "step": 27391 + }, + { + "epoch": 2.9696444058976583, + "grad_norm": 0.21504725515842438, + "learning_rate": 2.0310459886841726e-07, + "loss": 0.0011, + "step": 27392 + }, + { + "epoch": 2.9697528187337383, + "grad_norm": 0.03869282454252243, + "learning_rate": 2.0237922530103004e-07, + "loss": 0.0004, + "step": 27393 + }, + { + "epoch": 2.969861231569818, + "grad_norm": 4.745090961456299, + "learning_rate": 2.0165385173364285e-07, + "loss": 0.0389, + "step": 27394 + }, + { + "epoch": 2.9699696444058974, + "grad_norm": 0.04711189866065979, + "learning_rate": 2.0092847816625565e-07, + "loss": 0.0005, + "step": 27395 + }, + { + "epoch": 2.9700780572419774, + "grad_norm": 0.11280985176563263, + "learning_rate": 2.0020310459886846e-07, + "loss": 0.0004, + "step": 27396 + }, + { + "epoch": 2.9701864700780574, + "grad_norm": 0.1423349380493164, + "learning_rate": 1.994777310314812e-07, + "loss": 0.0006, + "step": 27397 + }, + { + "epoch": 2.970294882914137, + "grad_norm": 0.10282297432422638, + "learning_rate": 1.9875235746409404e-07, + "loss": 0.0004, + "step": 27398 + }, + { + "epoch": 2.970403295750217, + "grad_norm": 0.05995297431945801, + "learning_rate": 1.9802698389670685e-07, + "loss": 0.0004, + "step": 27399 + }, + { + "epoch": 2.9705117085862964, + "grad_norm": 6.6007080078125, + "learning_rate": 1.973016103293196e-07, + "loss": 0.0411, + "step": 27400 + }, + { + "epoch": 2.9706201214223764, + "grad_norm": 0.0885082557797432, + "learning_rate": 1.965762367619324e-07, + "loss": 0.0007, + "step": 27401 + }, + { + "epoch": 2.9707285342584564, + "grad_norm": 1.4297001361846924, + "learning_rate": 1.958508631945452e-07, + "loss": 0.0158, + "step": 27402 + }, + { + "epoch": 2.970836947094536, + "grad_norm": 0.6627113819122314, + "learning_rate": 1.9512548962715799e-07, + "loss": 0.0012, + "step": 27403 + }, + { + "epoch": 2.970945359930616, + "grad_norm": 0.13157808780670166, + "learning_rate": 1.944001160597708e-07, + "loss": 0.0004, + "step": 27404 + }, + { + "epoch": 2.9710537727666955, + "grad_norm": 0.04667180776596069, + "learning_rate": 1.936747424923836e-07, + "loss": 0.0004, + "step": 27405 + }, + { + "epoch": 2.9711621856027755, + "grad_norm": 0.04020010307431221, + "learning_rate": 1.929493689249964e-07, + "loss": 0.0004, + "step": 27406 + }, + { + "epoch": 2.9712705984388554, + "grad_norm": 0.052539024502038956, + "learning_rate": 1.9222399535760918e-07, + "loss": 0.0003, + "step": 27407 + }, + { + "epoch": 2.971379011274935, + "grad_norm": 3.685614824295044, + "learning_rate": 1.9149862179022199e-07, + "loss": 0.0072, + "step": 27408 + }, + { + "epoch": 2.9714874241110145, + "grad_norm": 0.1434304416179657, + "learning_rate": 1.907732482228348e-07, + "loss": 0.0006, + "step": 27409 + }, + { + "epoch": 2.9715958369470945, + "grad_norm": 2.430716037750244, + "learning_rate": 1.9004787465544757e-07, + "loss": 0.013, + "step": 27410 + }, + { + "epoch": 2.9717042497831745, + "grad_norm": 0.043243952095508575, + "learning_rate": 1.8932250108806037e-07, + "loss": 0.0002, + "step": 27411 + }, + { + "epoch": 2.971812662619254, + "grad_norm": 0.655093789100647, + "learning_rate": 1.8859712752067318e-07, + "loss": 0.0016, + "step": 27412 + }, + { + "epoch": 2.971921075455334, + "grad_norm": 0.04702567309141159, + "learning_rate": 1.8787175395328598e-07, + "loss": 0.0003, + "step": 27413 + }, + { + "epoch": 2.9720294882914136, + "grad_norm": 4.362457275390625, + "learning_rate": 1.8714638038589874e-07, + "loss": 0.0345, + "step": 27414 + }, + { + "epoch": 2.9721379011274935, + "grad_norm": 4.58318567276001, + "learning_rate": 1.8642100681851154e-07, + "loss": 0.0267, + "step": 27415 + }, + { + "epoch": 2.9722463139635735, + "grad_norm": 0.36602145433425903, + "learning_rate": 1.8569563325112435e-07, + "loss": 0.0024, + "step": 27416 + }, + { + "epoch": 2.972354726799653, + "grad_norm": 3.8725686073303223, + "learning_rate": 1.8497025968373713e-07, + "loss": 0.0179, + "step": 27417 + }, + { + "epoch": 2.9724631396357326, + "grad_norm": 17.163501739501953, + "learning_rate": 1.8424488611634993e-07, + "loss": 0.0124, + "step": 27418 + }, + { + "epoch": 2.9725715524718126, + "grad_norm": 0.033830128610134125, + "learning_rate": 1.8351951254896274e-07, + "loss": 0.0005, + "step": 27419 + }, + { + "epoch": 2.9726799653078926, + "grad_norm": 0.06863768398761749, + "learning_rate": 1.8279413898157551e-07, + "loss": 0.0004, + "step": 27420 + }, + { + "epoch": 2.972788378143972, + "grad_norm": 0.07782428711652756, + "learning_rate": 1.8206876541418832e-07, + "loss": 0.0007, + "step": 27421 + }, + { + "epoch": 2.972896790980052, + "grad_norm": 0.7416461706161499, + "learning_rate": 1.8134339184680113e-07, + "loss": 0.0031, + "step": 27422 + }, + { + "epoch": 2.9730052038161316, + "grad_norm": 20.044876098632812, + "learning_rate": 1.8061801827941393e-07, + "loss": 0.0326, + "step": 27423 + }, + { + "epoch": 2.9731136166522116, + "grad_norm": 0.3299593925476074, + "learning_rate": 1.798926447120267e-07, + "loss": 0.0013, + "step": 27424 + }, + { + "epoch": 2.9732220294882916, + "grad_norm": 0.07539357990026474, + "learning_rate": 1.7916727114463951e-07, + "loss": 0.0004, + "step": 27425 + }, + { + "epoch": 2.973330442324371, + "grad_norm": 0.09216637164354324, + "learning_rate": 1.7844189757725232e-07, + "loss": 0.0008, + "step": 27426 + }, + { + "epoch": 2.973438855160451, + "grad_norm": 2.214121103286743, + "learning_rate": 1.7771652400986507e-07, + "loss": 0.014, + "step": 27427 + }, + { + "epoch": 2.9735472679965307, + "grad_norm": 0.14402776956558228, + "learning_rate": 1.7699115044247788e-07, + "loss": 0.001, + "step": 27428 + }, + { + "epoch": 2.9736556808326107, + "grad_norm": 3.256415843963623, + "learning_rate": 1.7626577687509068e-07, + "loss": 0.0428, + "step": 27429 + }, + { + "epoch": 2.97376409366869, + "grad_norm": 0.22895458340644836, + "learning_rate": 1.7554040330770346e-07, + "loss": 0.0013, + "step": 27430 + }, + { + "epoch": 2.97387250650477, + "grad_norm": 0.4765581786632538, + "learning_rate": 1.7481502974031627e-07, + "loss": 0.0028, + "step": 27431 + }, + { + "epoch": 2.9739809193408497, + "grad_norm": 0.05245077610015869, + "learning_rate": 1.7408965617292907e-07, + "loss": 0.0005, + "step": 27432 + }, + { + "epoch": 2.9740893321769297, + "grad_norm": 0.5358343124389648, + "learning_rate": 1.7336428260554188e-07, + "loss": 0.0016, + "step": 27433 + }, + { + "epoch": 2.9741977450130097, + "grad_norm": 0.11497023701667786, + "learning_rate": 1.7263890903815465e-07, + "loss": 0.0006, + "step": 27434 + }, + { + "epoch": 2.9743061578490892, + "grad_norm": 0.017100073397159576, + "learning_rate": 1.7191353547076746e-07, + "loss": 0.0003, + "step": 27435 + }, + { + "epoch": 2.9744145706851692, + "grad_norm": 2.52278733253479, + "learning_rate": 1.7118816190338026e-07, + "loss": 0.0091, + "step": 27436 + }, + { + "epoch": 2.9745229835212488, + "grad_norm": 2.3638839721679688, + "learning_rate": 1.7046278833599304e-07, + "loss": 0.0033, + "step": 27437 + }, + { + "epoch": 2.9746313963573288, + "grad_norm": 2.6050631999969482, + "learning_rate": 1.6973741476860585e-07, + "loss": 0.0063, + "step": 27438 + }, + { + "epoch": 2.9747398091934087, + "grad_norm": 1.324377417564392, + "learning_rate": 1.6901204120121865e-07, + "loss": 0.0076, + "step": 27439 + }, + { + "epoch": 2.9748482220294883, + "grad_norm": 1.1286803483963013, + "learning_rate": 1.6828666763383146e-07, + "loss": 0.0026, + "step": 27440 + }, + { + "epoch": 2.974956634865568, + "grad_norm": 0.8280332684516907, + "learning_rate": 1.6756129406644424e-07, + "loss": 0.0019, + "step": 27441 + }, + { + "epoch": 2.975065047701648, + "grad_norm": 0.061759352684020996, + "learning_rate": 1.6683592049905704e-07, + "loss": 0.0002, + "step": 27442 + }, + { + "epoch": 2.975173460537728, + "grad_norm": 2.7493038177490234, + "learning_rate": 1.6611054693166985e-07, + "loss": 0.0135, + "step": 27443 + }, + { + "epoch": 2.9752818733738073, + "grad_norm": 0.48763930797576904, + "learning_rate": 1.653851733642826e-07, + "loss": 0.0021, + "step": 27444 + }, + { + "epoch": 2.9753902862098873, + "grad_norm": 0.020004048943519592, + "learning_rate": 1.646597997968954e-07, + "loss": 0.0001, + "step": 27445 + }, + { + "epoch": 2.975498699045967, + "grad_norm": 2.3168656826019287, + "learning_rate": 1.639344262295082e-07, + "loss": 0.025, + "step": 27446 + }, + { + "epoch": 2.975607111882047, + "grad_norm": 1.4692414999008179, + "learning_rate": 1.63209052662121e-07, + "loss": 0.0039, + "step": 27447 + }, + { + "epoch": 2.975715524718127, + "grad_norm": 0.46329808235168457, + "learning_rate": 1.624836790947338e-07, + "loss": 0.0021, + "step": 27448 + }, + { + "epoch": 2.9758239375542064, + "grad_norm": 5.526555061340332, + "learning_rate": 1.617583055273466e-07, + "loss": 0.0285, + "step": 27449 + }, + { + "epoch": 2.9759323503902864, + "grad_norm": 0.4919157326221466, + "learning_rate": 1.610329319599594e-07, + "loss": 0.0012, + "step": 27450 + }, + { + "epoch": 2.976040763226366, + "grad_norm": 1.9052815437316895, + "learning_rate": 1.6030755839257218e-07, + "loss": 0.031, + "step": 27451 + }, + { + "epoch": 2.976149176062446, + "grad_norm": 3.5807151794433594, + "learning_rate": 1.59582184825185e-07, + "loss": 0.0194, + "step": 27452 + }, + { + "epoch": 2.9762575888985254, + "grad_norm": 0.13285861909389496, + "learning_rate": 1.588568112577978e-07, + "loss": 0.001, + "step": 27453 + }, + { + "epoch": 2.9763660017346054, + "grad_norm": 0.2805815637111664, + "learning_rate": 1.5813143769041057e-07, + "loss": 0.0012, + "step": 27454 + }, + { + "epoch": 2.976474414570685, + "grad_norm": 0.048505231738090515, + "learning_rate": 1.5740606412302338e-07, + "loss": 0.0003, + "step": 27455 + }, + { + "epoch": 2.976582827406765, + "grad_norm": 1.3789809942245483, + "learning_rate": 1.5668069055563618e-07, + "loss": 0.0043, + "step": 27456 + }, + { + "epoch": 2.976691240242845, + "grad_norm": 0.0830155462026596, + "learning_rate": 1.5595531698824896e-07, + "loss": 0.0004, + "step": 27457 + }, + { + "epoch": 2.9767996530789245, + "grad_norm": 1.5872306823730469, + "learning_rate": 1.5522994342086174e-07, + "loss": 0.0013, + "step": 27458 + }, + { + "epoch": 2.9769080659150045, + "grad_norm": 1.3403550386428833, + "learning_rate": 1.5450456985347454e-07, + "loss": 0.0015, + "step": 27459 + }, + { + "epoch": 2.977016478751084, + "grad_norm": 3.130786657333374, + "learning_rate": 1.5377919628608735e-07, + "loss": 0.0308, + "step": 27460 + }, + { + "epoch": 2.977124891587164, + "grad_norm": 2.78941011428833, + "learning_rate": 1.5305382271870015e-07, + "loss": 0.0184, + "step": 27461 + }, + { + "epoch": 2.977233304423244, + "grad_norm": 2.1631970405578613, + "learning_rate": 1.5232844915131293e-07, + "loss": 0.0288, + "step": 27462 + }, + { + "epoch": 2.9773417172593235, + "grad_norm": 1.3536547422409058, + "learning_rate": 1.5160307558392574e-07, + "loss": 0.0069, + "step": 27463 + }, + { + "epoch": 2.977450130095403, + "grad_norm": 8.874406814575195, + "learning_rate": 1.5087770201653854e-07, + "loss": 0.0261, + "step": 27464 + }, + { + "epoch": 2.977558542931483, + "grad_norm": 2.2003395557403564, + "learning_rate": 1.5015232844915132e-07, + "loss": 0.0097, + "step": 27465 + }, + { + "epoch": 2.977666955767563, + "grad_norm": 0.3395247161388397, + "learning_rate": 1.4942695488176413e-07, + "loss": 0.0021, + "step": 27466 + }, + { + "epoch": 2.9777753686036426, + "grad_norm": 1.086348295211792, + "learning_rate": 1.487015813143769e-07, + "loss": 0.0016, + "step": 27467 + }, + { + "epoch": 2.9778837814397225, + "grad_norm": 2.305243492126465, + "learning_rate": 1.479762077469897e-07, + "loss": 0.0434, + "step": 27468 + }, + { + "epoch": 2.977992194275802, + "grad_norm": 0.05995041877031326, + "learning_rate": 1.4725083417960252e-07, + "loss": 0.0004, + "step": 27469 + }, + { + "epoch": 2.978100607111882, + "grad_norm": 2.785330295562744, + "learning_rate": 1.465254606122153e-07, + "loss": 0.0034, + "step": 27470 + }, + { + "epoch": 2.978209019947962, + "grad_norm": 1.2749643325805664, + "learning_rate": 1.458000870448281e-07, + "loss": 0.0013, + "step": 27471 + }, + { + "epoch": 2.9783174327840416, + "grad_norm": 1.7555568218231201, + "learning_rate": 1.4507471347744088e-07, + "loss": 0.0204, + "step": 27472 + }, + { + "epoch": 2.9784258456201216, + "grad_norm": 0.2954530417919159, + "learning_rate": 1.4434933991005368e-07, + "loss": 0.0009, + "step": 27473 + }, + { + "epoch": 2.978534258456201, + "grad_norm": 0.08493651449680328, + "learning_rate": 1.436239663426665e-07, + "loss": 0.0006, + "step": 27474 + }, + { + "epoch": 2.978642671292281, + "grad_norm": 5.362784385681152, + "learning_rate": 1.4289859277527927e-07, + "loss": 0.0127, + "step": 27475 + }, + { + "epoch": 2.9787510841283606, + "grad_norm": 0.06930667161941528, + "learning_rate": 1.4217321920789207e-07, + "loss": 0.0006, + "step": 27476 + }, + { + "epoch": 2.9788594969644406, + "grad_norm": 0.07996077835559845, + "learning_rate": 1.4144784564050488e-07, + "loss": 0.0004, + "step": 27477 + }, + { + "epoch": 2.97896790980052, + "grad_norm": 0.0634566992521286, + "learning_rate": 1.4072247207311768e-07, + "loss": 0.0003, + "step": 27478 + }, + { + "epoch": 2.9790763226366, + "grad_norm": 0.16732093691825867, + "learning_rate": 1.3999709850573046e-07, + "loss": 0.0006, + "step": 27479 + }, + { + "epoch": 2.97918473547268, + "grad_norm": 0.12075357884168625, + "learning_rate": 1.3927172493834324e-07, + "loss": 0.0006, + "step": 27480 + }, + { + "epoch": 2.9792931483087597, + "grad_norm": 0.09448935836553574, + "learning_rate": 1.3854635137095605e-07, + "loss": 0.0004, + "step": 27481 + }, + { + "epoch": 2.9794015611448397, + "grad_norm": 0.18486711382865906, + "learning_rate": 1.3782097780356885e-07, + "loss": 0.0007, + "step": 27482 + }, + { + "epoch": 2.979509973980919, + "grad_norm": 0.052091363817453384, + "learning_rate": 1.3709560423618166e-07, + "loss": 0.0002, + "step": 27483 + }, + { + "epoch": 2.979618386816999, + "grad_norm": 4.528409481048584, + "learning_rate": 1.3637023066879443e-07, + "loss": 0.0116, + "step": 27484 + }, + { + "epoch": 2.979726799653079, + "grad_norm": 0.013153156265616417, + "learning_rate": 1.3564485710140724e-07, + "loss": 0.0002, + "step": 27485 + }, + { + "epoch": 2.9798352124891587, + "grad_norm": 0.04121920093894005, + "learning_rate": 1.3491948353402004e-07, + "loss": 0.0002, + "step": 27486 + }, + { + "epoch": 2.9799436253252383, + "grad_norm": 0.20791344344615936, + "learning_rate": 1.3419410996663282e-07, + "loss": 0.0008, + "step": 27487 + }, + { + "epoch": 2.9800520381613183, + "grad_norm": 3.925642967224121, + "learning_rate": 1.3346873639924563e-07, + "loss": 0.0034, + "step": 27488 + }, + { + "epoch": 2.9801604509973982, + "grad_norm": 7.048496246337891, + "learning_rate": 1.327433628318584e-07, + "loss": 0.0043, + "step": 27489 + }, + { + "epoch": 2.9802688638334778, + "grad_norm": 0.18738943338394165, + "learning_rate": 1.320179892644712e-07, + "loss": 0.0005, + "step": 27490 + }, + { + "epoch": 2.9803772766695578, + "grad_norm": 0.01997983828186989, + "learning_rate": 1.3129261569708402e-07, + "loss": 0.0002, + "step": 27491 + }, + { + "epoch": 2.9804856895056373, + "grad_norm": 10.412262916564941, + "learning_rate": 1.305672421296968e-07, + "loss": 0.0383, + "step": 27492 + }, + { + "epoch": 2.9805941023417173, + "grad_norm": 0.5881668329238892, + "learning_rate": 1.298418685623096e-07, + "loss": 0.0007, + "step": 27493 + }, + { + "epoch": 2.9807025151777973, + "grad_norm": 1.4938794374465942, + "learning_rate": 1.2911649499492238e-07, + "loss": 0.0096, + "step": 27494 + }, + { + "epoch": 2.980810928013877, + "grad_norm": 2.4352753162384033, + "learning_rate": 1.283911214275352e-07, + "loss": 0.009, + "step": 27495 + }, + { + "epoch": 2.980919340849957, + "grad_norm": 6.20747709274292, + "learning_rate": 1.27665747860148e-07, + "loss": 0.0242, + "step": 27496 + }, + { + "epoch": 2.9810277536860363, + "grad_norm": 0.14965178072452545, + "learning_rate": 1.2694037429276077e-07, + "loss": 0.0013, + "step": 27497 + }, + { + "epoch": 2.9811361665221163, + "grad_norm": 0.1484709233045578, + "learning_rate": 1.2621500072537357e-07, + "loss": 0.0008, + "step": 27498 + }, + { + "epoch": 2.981244579358196, + "grad_norm": 1.813516616821289, + "learning_rate": 1.2548962715798638e-07, + "loss": 0.0098, + "step": 27499 + }, + { + "epoch": 2.981352992194276, + "grad_norm": 2.3200161457061768, + "learning_rate": 1.2476425359059918e-07, + "loss": 0.0112, + "step": 27500 + }, + { + "epoch": 2.9814614050303554, + "grad_norm": 0.27255913615226746, + "learning_rate": 1.2403888002321196e-07, + "loss": 0.0012, + "step": 27501 + }, + { + "epoch": 2.9815698178664354, + "grad_norm": 1.1818876266479492, + "learning_rate": 1.2331350645582474e-07, + "loss": 0.0031, + "step": 27502 + }, + { + "epoch": 2.9816782307025154, + "grad_norm": 0.1487366259098053, + "learning_rate": 1.2258813288843755e-07, + "loss": 0.001, + "step": 27503 + }, + { + "epoch": 2.981786643538595, + "grad_norm": 0.16208869218826294, + "learning_rate": 1.2186275932105035e-07, + "loss": 0.0008, + "step": 27504 + }, + { + "epoch": 2.981895056374675, + "grad_norm": 0.7818003296852112, + "learning_rate": 1.2113738575366316e-07, + "loss": 0.0017, + "step": 27505 + }, + { + "epoch": 2.9820034692107544, + "grad_norm": 0.3491099774837494, + "learning_rate": 1.2041201218627594e-07, + "loss": 0.0013, + "step": 27506 + }, + { + "epoch": 2.9821118820468344, + "grad_norm": 2.8342063426971436, + "learning_rate": 1.1968663861888874e-07, + "loss": 0.0022, + "step": 27507 + }, + { + "epoch": 2.9822202948829144, + "grad_norm": 2.03847074508667, + "learning_rate": 1.1896126505150153e-07, + "loss": 0.0138, + "step": 27508 + }, + { + "epoch": 2.982328707718994, + "grad_norm": 0.3565134108066559, + "learning_rate": 1.1823589148411432e-07, + "loss": 0.002, + "step": 27509 + }, + { + "epoch": 2.9824371205550735, + "grad_norm": 0.17949549853801727, + "learning_rate": 1.1751051791672713e-07, + "loss": 0.0013, + "step": 27510 + }, + { + "epoch": 2.9825455333911535, + "grad_norm": 0.7828460335731506, + "learning_rate": 1.1678514434933992e-07, + "loss": 0.003, + "step": 27511 + }, + { + "epoch": 2.9826539462272335, + "grad_norm": 1.910038948059082, + "learning_rate": 1.1605977078195273e-07, + "loss": 0.0037, + "step": 27512 + }, + { + "epoch": 2.982762359063313, + "grad_norm": 0.28258636593818665, + "learning_rate": 1.153343972145655e-07, + "loss": 0.0006, + "step": 27513 + }, + { + "epoch": 2.982870771899393, + "grad_norm": 2.30308198928833, + "learning_rate": 1.146090236471783e-07, + "loss": 0.0102, + "step": 27514 + }, + { + "epoch": 2.9829791847354725, + "grad_norm": 3.215670585632324, + "learning_rate": 1.138836500797911e-07, + "loss": 0.0123, + "step": 27515 + }, + { + "epoch": 2.9830875975715525, + "grad_norm": 1.9145547151565552, + "learning_rate": 1.131582765124039e-07, + "loss": 0.0078, + "step": 27516 + }, + { + "epoch": 2.9831960104076325, + "grad_norm": 0.46886661648750305, + "learning_rate": 1.124329029450167e-07, + "loss": 0.0016, + "step": 27517 + }, + { + "epoch": 2.983304423243712, + "grad_norm": 0.9834979772567749, + "learning_rate": 1.1170752937762949e-07, + "loss": 0.0046, + "step": 27518 + }, + { + "epoch": 2.9834128360797916, + "grad_norm": 1.7802397012710571, + "learning_rate": 1.1098215581024228e-07, + "loss": 0.0166, + "step": 27519 + }, + { + "epoch": 2.9835212489158716, + "grad_norm": 0.5279967188835144, + "learning_rate": 1.1025678224285509e-07, + "loss": 0.0039, + "step": 27520 + }, + { + "epoch": 2.9836296617519515, + "grad_norm": 0.15462465584278107, + "learning_rate": 1.0953140867546787e-07, + "loss": 0.0007, + "step": 27521 + }, + { + "epoch": 2.983738074588031, + "grad_norm": 2.2755284309387207, + "learning_rate": 1.0880603510808067e-07, + "loss": 0.0092, + "step": 27522 + }, + { + "epoch": 2.983846487424111, + "grad_norm": 6.680399417877197, + "learning_rate": 1.0808066154069346e-07, + "loss": 0.0239, + "step": 27523 + }, + { + "epoch": 2.9839549002601906, + "grad_norm": 4.27944278717041, + "learning_rate": 1.0735528797330626e-07, + "loss": 0.0754, + "step": 27524 + }, + { + "epoch": 2.9840633130962706, + "grad_norm": 0.06981359422206879, + "learning_rate": 1.0662991440591906e-07, + "loss": 0.0004, + "step": 27525 + }, + { + "epoch": 2.9841717259323506, + "grad_norm": 0.08800007402896881, + "learning_rate": 1.0590454083853185e-07, + "loss": 0.0003, + "step": 27526 + }, + { + "epoch": 2.98428013876843, + "grad_norm": 0.044688791036605835, + "learning_rate": 1.0517916727114466e-07, + "loss": 0.0003, + "step": 27527 + }, + { + "epoch": 2.98438855160451, + "grad_norm": 0.19535841047763824, + "learning_rate": 1.0445379370375744e-07, + "loss": 0.0007, + "step": 27528 + }, + { + "epoch": 2.9844969644405897, + "grad_norm": 33.24535369873047, + "learning_rate": 1.0372842013637023e-07, + "loss": 0.0332, + "step": 27529 + }, + { + "epoch": 2.9846053772766696, + "grad_norm": 8.728246688842773, + "learning_rate": 1.0300304656898303e-07, + "loss": 0.0217, + "step": 27530 + }, + { + "epoch": 2.9847137901127496, + "grad_norm": 0.6039589047431946, + "learning_rate": 1.0227767300159583e-07, + "loss": 0.0023, + "step": 27531 + }, + { + "epoch": 2.984822202948829, + "grad_norm": 6.784130573272705, + "learning_rate": 1.0155229943420863e-07, + "loss": 0.0236, + "step": 27532 + }, + { + "epoch": 2.9849306157849087, + "grad_norm": 0.1756596714258194, + "learning_rate": 1.0082692586682142e-07, + "loss": 0.0064, + "step": 27533 + }, + { + "epoch": 2.9850390286209887, + "grad_norm": 1.835674524307251, + "learning_rate": 1.0010155229943423e-07, + "loss": 0.0692, + "step": 27534 + }, + { + "epoch": 2.9851474414570687, + "grad_norm": 0.07150544226169586, + "learning_rate": 9.937617873204702e-08, + "loss": 0.0005, + "step": 27535 + }, + { + "epoch": 2.985255854293148, + "grad_norm": 4.804615020751953, + "learning_rate": 9.86508051646598e-08, + "loss": 0.006, + "step": 27536 + }, + { + "epoch": 2.985364267129228, + "grad_norm": 2.142094612121582, + "learning_rate": 9.79254315972726e-08, + "loss": 0.0235, + "step": 27537 + }, + { + "epoch": 2.9854726799653077, + "grad_norm": 2.1729938983917236, + "learning_rate": 9.72000580298854e-08, + "loss": 0.0021, + "step": 27538 + }, + { + "epoch": 2.9855810928013877, + "grad_norm": 0.023647300899028778, + "learning_rate": 9.64746844624982e-08, + "loss": 0.0003, + "step": 27539 + }, + { + "epoch": 2.9856895056374677, + "grad_norm": 0.017916100099682808, + "learning_rate": 9.574931089511099e-08, + "loss": 0.0001, + "step": 27540 + }, + { + "epoch": 2.9857979184735473, + "grad_norm": 1.5274401903152466, + "learning_rate": 9.502393732772378e-08, + "loss": 0.0346, + "step": 27541 + }, + { + "epoch": 2.985906331309627, + "grad_norm": 0.8680791854858398, + "learning_rate": 9.429856376033659e-08, + "loss": 0.0021, + "step": 27542 + }, + { + "epoch": 2.986014744145707, + "grad_norm": 0.0461682491004467, + "learning_rate": 9.357319019294937e-08, + "loss": 0.0003, + "step": 27543 + }, + { + "epoch": 2.9861231569817868, + "grad_norm": 3.361844062805176, + "learning_rate": 9.284781662556217e-08, + "loss": 0.0144, + "step": 27544 + }, + { + "epoch": 2.9862315698178663, + "grad_norm": 3.402677297592163, + "learning_rate": 9.212244305817497e-08, + "loss": 0.0064, + "step": 27545 + }, + { + "epoch": 2.9863399826539463, + "grad_norm": 20.780914306640625, + "learning_rate": 9.139706949078776e-08, + "loss": 0.0465, + "step": 27546 + }, + { + "epoch": 2.986448395490026, + "grad_norm": 0.04677814990282059, + "learning_rate": 9.067169592340056e-08, + "loss": 0.0002, + "step": 27547 + }, + { + "epoch": 2.986556808326106, + "grad_norm": 4.845525741577148, + "learning_rate": 8.994632235601335e-08, + "loss": 0.0898, + "step": 27548 + }, + { + "epoch": 2.986665221162186, + "grad_norm": 2.446079969406128, + "learning_rate": 8.922094878862616e-08, + "loss": 0.008, + "step": 27549 + }, + { + "epoch": 2.9867736339982653, + "grad_norm": 0.25824466347694397, + "learning_rate": 8.849557522123894e-08, + "loss": 0.0006, + "step": 27550 + }, + { + "epoch": 2.9868820468343453, + "grad_norm": 4.717284679412842, + "learning_rate": 8.777020165385173e-08, + "loss": 0.0773, + "step": 27551 + }, + { + "epoch": 2.986990459670425, + "grad_norm": 0.15760120749473572, + "learning_rate": 8.704482808646454e-08, + "loss": 0.001, + "step": 27552 + }, + { + "epoch": 2.987098872506505, + "grad_norm": 0.268106609582901, + "learning_rate": 8.631945451907733e-08, + "loss": 0.0016, + "step": 27553 + }, + { + "epoch": 2.987207285342585, + "grad_norm": 4.009091854095459, + "learning_rate": 8.559408095169013e-08, + "loss": 0.0281, + "step": 27554 + }, + { + "epoch": 2.9873156981786644, + "grad_norm": 0.2744939923286438, + "learning_rate": 8.486870738430292e-08, + "loss": 0.001, + "step": 27555 + }, + { + "epoch": 2.987424111014744, + "grad_norm": 1.029527187347412, + "learning_rate": 8.414333381691573e-08, + "loss": 0.0018, + "step": 27556 + }, + { + "epoch": 2.987532523850824, + "grad_norm": 0.01274732407182455, + "learning_rate": 8.341796024952852e-08, + "loss": 0.0002, + "step": 27557 + }, + { + "epoch": 2.987640936686904, + "grad_norm": 1.978183388710022, + "learning_rate": 8.26925866821413e-08, + "loss": 0.0136, + "step": 27558 + }, + { + "epoch": 2.9877493495229834, + "grad_norm": 0.2068641036748886, + "learning_rate": 8.19672131147541e-08, + "loss": 0.0008, + "step": 27559 + }, + { + "epoch": 2.9878577623590634, + "grad_norm": 0.17425081133842468, + "learning_rate": 8.12418395473669e-08, + "loss": 0.0007, + "step": 27560 + }, + { + "epoch": 2.987966175195143, + "grad_norm": 1.104454755783081, + "learning_rate": 8.05164659799797e-08, + "loss": 0.004, + "step": 27561 + }, + { + "epoch": 2.988074588031223, + "grad_norm": 2.8387935161590576, + "learning_rate": 7.97910924125925e-08, + "loss": 0.0161, + "step": 27562 + }, + { + "epoch": 2.988183000867303, + "grad_norm": 0.06851095706224442, + "learning_rate": 7.906571884520529e-08, + "loss": 0.0004, + "step": 27563 + }, + { + "epoch": 2.9882914137033825, + "grad_norm": 0.06736914813518524, + "learning_rate": 7.834034527781809e-08, + "loss": 0.0006, + "step": 27564 + }, + { + "epoch": 2.988399826539462, + "grad_norm": 1.8327112197875977, + "learning_rate": 7.761497171043087e-08, + "loss": 0.0199, + "step": 27565 + }, + { + "epoch": 2.988508239375542, + "grad_norm": 7.433347702026367, + "learning_rate": 7.688959814304367e-08, + "loss": 0.0437, + "step": 27566 + }, + { + "epoch": 2.988616652211622, + "grad_norm": 0.1272507756948471, + "learning_rate": 7.616422457565647e-08, + "loss": 0.0004, + "step": 27567 + }, + { + "epoch": 2.9887250650477015, + "grad_norm": 3.81192946434021, + "learning_rate": 7.543885100826927e-08, + "loss": 0.0028, + "step": 27568 + }, + { + "epoch": 2.9888334778837815, + "grad_norm": 3.082850217819214, + "learning_rate": 7.471347744088206e-08, + "loss": 0.0074, + "step": 27569 + }, + { + "epoch": 2.988941890719861, + "grad_norm": 2.3704516887664795, + "learning_rate": 7.398810387349486e-08, + "loss": 0.0118, + "step": 27570 + }, + { + "epoch": 2.989050303555941, + "grad_norm": 0.013823808170855045, + "learning_rate": 7.326273030610765e-08, + "loss": 0.0002, + "step": 27571 + }, + { + "epoch": 2.989158716392021, + "grad_norm": 3.676536798477173, + "learning_rate": 7.253735673872044e-08, + "loss": 0.007, + "step": 27572 + }, + { + "epoch": 2.9892671292281006, + "grad_norm": 0.1990358829498291, + "learning_rate": 7.181198317133324e-08, + "loss": 0.0007, + "step": 27573 + }, + { + "epoch": 2.9893755420641805, + "grad_norm": 4.1395721435546875, + "learning_rate": 7.108660960394604e-08, + "loss": 0.0351, + "step": 27574 + }, + { + "epoch": 2.98948395490026, + "grad_norm": 0.2627376616001129, + "learning_rate": 7.036123603655884e-08, + "loss": 0.0015, + "step": 27575 + }, + { + "epoch": 2.98959236773634, + "grad_norm": 0.10518259555101395, + "learning_rate": 6.963586246917162e-08, + "loss": 0.0005, + "step": 27576 + }, + { + "epoch": 2.9897007805724196, + "grad_norm": 0.04846774414181709, + "learning_rate": 6.891048890178443e-08, + "loss": 0.0002, + "step": 27577 + }, + { + "epoch": 2.9898091934084996, + "grad_norm": 2.9040637016296387, + "learning_rate": 6.818511533439722e-08, + "loss": 0.0345, + "step": 27578 + }, + { + "epoch": 2.989917606244579, + "grad_norm": 2.050684690475464, + "learning_rate": 6.745974176701002e-08, + "loss": 0.0067, + "step": 27579 + }, + { + "epoch": 2.990026019080659, + "grad_norm": 1.7100918292999268, + "learning_rate": 6.673436819962281e-08, + "loss": 0.0059, + "step": 27580 + }, + { + "epoch": 2.990134431916739, + "grad_norm": 0.3824073076248169, + "learning_rate": 6.60089946322356e-08, + "loss": 0.0008, + "step": 27581 + }, + { + "epoch": 2.9902428447528187, + "grad_norm": 0.18811441957950592, + "learning_rate": 6.52836210648484e-08, + "loss": 0.001, + "step": 27582 + }, + { + "epoch": 2.9903512575888986, + "grad_norm": 0.28363144397735596, + "learning_rate": 6.455824749746119e-08, + "loss": 0.0002, + "step": 27583 + }, + { + "epoch": 2.990459670424978, + "grad_norm": 2.236778497695923, + "learning_rate": 6.3832873930074e-08, + "loss": 0.0291, + "step": 27584 + }, + { + "epoch": 2.990568083261058, + "grad_norm": 0.2172090709209442, + "learning_rate": 6.310750036268679e-08, + "loss": 0.0012, + "step": 27585 + }, + { + "epoch": 2.990676496097138, + "grad_norm": 4.664428234100342, + "learning_rate": 6.238212679529959e-08, + "loss": 0.0354, + "step": 27586 + }, + { + "epoch": 2.9907849089332177, + "grad_norm": 0.11630373448133469, + "learning_rate": 6.165675322791237e-08, + "loss": 0.0008, + "step": 27587 + }, + { + "epoch": 2.9908933217692972, + "grad_norm": 4.866086006164551, + "learning_rate": 6.093137966052518e-08, + "loss": 0.0056, + "step": 27588 + }, + { + "epoch": 2.991001734605377, + "grad_norm": 0.7832564115524292, + "learning_rate": 6.020600609313797e-08, + "loss": 0.0035, + "step": 27589 + }, + { + "epoch": 2.991110147441457, + "grad_norm": 4.70439338684082, + "learning_rate": 5.9480632525750766e-08, + "loss": 0.0068, + "step": 27590 + }, + { + "epoch": 2.9912185602775367, + "grad_norm": 0.6272104382514954, + "learning_rate": 5.8755258958363565e-08, + "loss": 0.0025, + "step": 27591 + }, + { + "epoch": 2.9913269731136167, + "grad_norm": 0.21775294840335846, + "learning_rate": 5.8029885390976363e-08, + "loss": 0.0008, + "step": 27592 + }, + { + "epoch": 2.9914353859496963, + "grad_norm": 0.5552591681480408, + "learning_rate": 5.730451182358915e-08, + "loss": 0.0019, + "step": 27593 + }, + { + "epoch": 2.9915437987857763, + "grad_norm": 3.818486452102661, + "learning_rate": 5.657913825620195e-08, + "loss": 0.0106, + "step": 27594 + }, + { + "epoch": 2.9916522116218562, + "grad_norm": 0.04864152893424034, + "learning_rate": 5.5853764688814746e-08, + "loss": 0.0003, + "step": 27595 + }, + { + "epoch": 2.991760624457936, + "grad_norm": 5.883777141571045, + "learning_rate": 5.5128391121427544e-08, + "loss": 0.066, + "step": 27596 + }, + { + "epoch": 2.9918690372940158, + "grad_norm": 0.3338611423969269, + "learning_rate": 5.4403017554040336e-08, + "loss": 0.0012, + "step": 27597 + }, + { + "epoch": 2.9919774501300953, + "grad_norm": 0.09361086785793304, + "learning_rate": 5.367764398665313e-08, + "loss": 0.0008, + "step": 27598 + }, + { + "epoch": 2.9920858629661753, + "grad_norm": 0.22908742725849152, + "learning_rate": 5.2952270419265927e-08, + "loss": 0.0015, + "step": 27599 + }, + { + "epoch": 2.992194275802255, + "grad_norm": 2.545868396759033, + "learning_rate": 5.222689685187872e-08, + "loss": 0.0335, + "step": 27600 + }, + { + "epoch": 2.992302688638335, + "grad_norm": 0.016105111688375473, + "learning_rate": 5.150152328449152e-08, + "loss": 0.0001, + "step": 27601 + }, + { + "epoch": 2.9924111014744144, + "grad_norm": 28.951189041137695, + "learning_rate": 5.0776149717104316e-08, + "loss": 0.0307, + "step": 27602 + }, + { + "epoch": 2.9925195143104943, + "grad_norm": 0.0594671294093132, + "learning_rate": 5.0050776149717114e-08, + "loss": 0.0004, + "step": 27603 + }, + { + "epoch": 2.9926279271465743, + "grad_norm": 0.028936071321368217, + "learning_rate": 4.93254025823299e-08, + "loss": 0.0002, + "step": 27604 + }, + { + "epoch": 2.992736339982654, + "grad_norm": 7.415746688842773, + "learning_rate": 4.86000290149427e-08, + "loss": 0.0303, + "step": 27605 + }, + { + "epoch": 2.992844752818734, + "grad_norm": 13.451263427734375, + "learning_rate": 4.7874655447555496e-08, + "loss": 0.0052, + "step": 27606 + }, + { + "epoch": 2.9929531656548134, + "grad_norm": 4.200572490692139, + "learning_rate": 4.7149281880168295e-08, + "loss": 0.0415, + "step": 27607 + }, + { + "epoch": 2.9930615784908934, + "grad_norm": 0.40786510705947876, + "learning_rate": 4.642390831278109e-08, + "loss": 0.0023, + "step": 27608 + }, + { + "epoch": 2.9931699913269734, + "grad_norm": 0.4739282727241516, + "learning_rate": 4.569853474539388e-08, + "loss": 0.0035, + "step": 27609 + }, + { + "epoch": 2.993278404163053, + "grad_norm": 0.030527599155902863, + "learning_rate": 4.497316117800668e-08, + "loss": 0.0003, + "step": 27610 + }, + { + "epoch": 2.9933868169991324, + "grad_norm": 0.07082654535770416, + "learning_rate": 4.424778761061947e-08, + "loss": 0.0002, + "step": 27611 + }, + { + "epoch": 2.9934952298352124, + "grad_norm": 25.590768814086914, + "learning_rate": 4.352241404323227e-08, + "loss": 0.021, + "step": 27612 + }, + { + "epoch": 2.9936036426712924, + "grad_norm": 0.16966001689434052, + "learning_rate": 4.2797040475845066e-08, + "loss": 0.0008, + "step": 27613 + }, + { + "epoch": 2.993712055507372, + "grad_norm": 6.028697967529297, + "learning_rate": 4.2071666908457865e-08, + "loss": 0.0064, + "step": 27614 + }, + { + "epoch": 2.993820468343452, + "grad_norm": 0.04315699636936188, + "learning_rate": 4.134629334107065e-08, + "loss": 0.0002, + "step": 27615 + }, + { + "epoch": 2.9939288811795315, + "grad_norm": 4.98809814453125, + "learning_rate": 4.062091977368345e-08, + "loss": 0.01, + "step": 27616 + }, + { + "epoch": 2.9940372940156115, + "grad_norm": 0.11582150310277939, + "learning_rate": 3.989554620629625e-08, + "loss": 0.0008, + "step": 27617 + }, + { + "epoch": 2.9941457068516915, + "grad_norm": 2.9684369564056396, + "learning_rate": 3.9170172638909046e-08, + "loss": 0.0071, + "step": 27618 + }, + { + "epoch": 2.994254119687771, + "grad_norm": 0.06679364293813705, + "learning_rate": 3.844479907152184e-08, + "loss": 0.0005, + "step": 27619 + }, + { + "epoch": 2.994362532523851, + "grad_norm": 0.4361625015735626, + "learning_rate": 3.7719425504134636e-08, + "loss": 0.0019, + "step": 27620 + }, + { + "epoch": 2.9944709453599305, + "grad_norm": 4.9904913902282715, + "learning_rate": 3.699405193674743e-08, + "loss": 0.014, + "step": 27621 + }, + { + "epoch": 2.9945793581960105, + "grad_norm": 2.446260929107666, + "learning_rate": 3.626867836936022e-08, + "loss": 0.0143, + "step": 27622 + }, + { + "epoch": 2.99468777103209, + "grad_norm": 4.855140209197998, + "learning_rate": 3.554330480197302e-08, + "loss": 0.0413, + "step": 27623 + }, + { + "epoch": 2.99479618386817, + "grad_norm": 0.2174970954656601, + "learning_rate": 3.481793123458581e-08, + "loss": 0.001, + "step": 27624 + }, + { + "epoch": 2.9949045967042496, + "grad_norm": 16.23314666748047, + "learning_rate": 3.409255766719861e-08, + "loss": 0.0073, + "step": 27625 + }, + { + "epoch": 2.9950130095403296, + "grad_norm": 0.1815008968114853, + "learning_rate": 3.336718409981141e-08, + "loss": 0.0008, + "step": 27626 + }, + { + "epoch": 2.9951214223764095, + "grad_norm": 1.4453061819076538, + "learning_rate": 3.26418105324242e-08, + "loss": 0.0227, + "step": 27627 + }, + { + "epoch": 2.995229835212489, + "grad_norm": 3.7630820274353027, + "learning_rate": 3.1916436965037e-08, + "loss": 0.0054, + "step": 27628 + }, + { + "epoch": 2.995338248048569, + "grad_norm": 5.778252124786377, + "learning_rate": 3.1191063397649796e-08, + "loss": 0.0393, + "step": 27629 + }, + { + "epoch": 2.9954466608846486, + "grad_norm": 0.23527786135673523, + "learning_rate": 3.046568983026259e-08, + "loss": 0.001, + "step": 27630 + }, + { + "epoch": 2.9955550737207286, + "grad_norm": 0.14057503640651703, + "learning_rate": 2.9740316262875383e-08, + "loss": 0.001, + "step": 27631 + }, + { + "epoch": 2.9956634865568086, + "grad_norm": 0.4078972041606903, + "learning_rate": 2.9014942695488182e-08, + "loss": 0.001, + "step": 27632 + }, + { + "epoch": 2.995771899392888, + "grad_norm": 2.8916168212890625, + "learning_rate": 2.8289569128100974e-08, + "loss": 0.0261, + "step": 27633 + }, + { + "epoch": 2.9958803122289677, + "grad_norm": 0.8942898511886597, + "learning_rate": 2.7564195560713772e-08, + "loss": 0.0021, + "step": 27634 + }, + { + "epoch": 2.9959887250650477, + "grad_norm": 0.07737787812948227, + "learning_rate": 2.6838821993326564e-08, + "loss": 0.0002, + "step": 27635 + }, + { + "epoch": 2.9960971379011276, + "grad_norm": 0.2740829288959503, + "learning_rate": 2.611344842593936e-08, + "loss": 0.0006, + "step": 27636 + }, + { + "epoch": 2.996205550737207, + "grad_norm": 0.032913096249103546, + "learning_rate": 2.5388074858552158e-08, + "loss": 0.0003, + "step": 27637 + }, + { + "epoch": 2.996313963573287, + "grad_norm": 2.961520195007324, + "learning_rate": 2.466270129116495e-08, + "loss": 0.0419, + "step": 27638 + }, + { + "epoch": 2.9964223764093667, + "grad_norm": 0.3413160741329193, + "learning_rate": 2.3937327723777748e-08, + "loss": 0.0012, + "step": 27639 + }, + { + "epoch": 2.9965307892454467, + "grad_norm": 0.40384358167648315, + "learning_rate": 2.3211954156390543e-08, + "loss": 0.0009, + "step": 27640 + }, + { + "epoch": 2.9966392020815267, + "grad_norm": 0.7975634336471558, + "learning_rate": 2.248658058900334e-08, + "loss": 0.0413, + "step": 27641 + }, + { + "epoch": 2.996747614917606, + "grad_norm": 0.26883664727211, + "learning_rate": 2.1761207021616134e-08, + "loss": 0.0021, + "step": 27642 + }, + { + "epoch": 2.996856027753686, + "grad_norm": 0.12379983067512512, + "learning_rate": 2.1035833454228932e-08, + "loss": 0.0006, + "step": 27643 + }, + { + "epoch": 2.9969644405897657, + "grad_norm": 3.9106781482696533, + "learning_rate": 2.0310459886841724e-08, + "loss": 0.0103, + "step": 27644 + }, + { + "epoch": 2.9970728534258457, + "grad_norm": 0.2366141378879547, + "learning_rate": 1.9585086319454523e-08, + "loss": 0.001, + "step": 27645 + }, + { + "epoch": 2.9971812662619253, + "grad_norm": 7.613354206085205, + "learning_rate": 1.8859712752067318e-08, + "loss": 0.0204, + "step": 27646 + }, + { + "epoch": 2.9972896790980053, + "grad_norm": 5.1532464027404785, + "learning_rate": 1.813433918468011e-08, + "loss": 0.0093, + "step": 27647 + }, + { + "epoch": 2.997398091934085, + "grad_norm": 0.7951837182044983, + "learning_rate": 1.7408965617292905e-08, + "loss": 0.0025, + "step": 27648 + }, + { + "epoch": 2.997506504770165, + "grad_norm": 0.256339967250824, + "learning_rate": 1.6683592049905704e-08, + "loss": 0.0005, + "step": 27649 + }, + { + "epoch": 2.9976149176062448, + "grad_norm": 2.745439052581787, + "learning_rate": 1.59582184825185e-08, + "loss": 0.0086, + "step": 27650 + }, + { + "epoch": 2.9977233304423243, + "grad_norm": 0.5774828195571899, + "learning_rate": 1.5232844915131294e-08, + "loss": 0.0021, + "step": 27651 + }, + { + "epoch": 2.9978317432784043, + "grad_norm": 0.24769540131092072, + "learning_rate": 1.4507471347744091e-08, + "loss": 0.0006, + "step": 27652 + }, + { + "epoch": 2.997940156114484, + "grad_norm": 1.480919361114502, + "learning_rate": 1.3782097780356886e-08, + "loss": 0.0033, + "step": 27653 + }, + { + "epoch": 2.998048568950564, + "grad_norm": 0.20756438374519348, + "learning_rate": 1.305672421296968e-08, + "loss": 0.0009, + "step": 27654 + }, + { + "epoch": 2.998156981786644, + "grad_norm": 2.1977462768554688, + "learning_rate": 1.2331350645582475e-08, + "loss": 0.0062, + "step": 27655 + }, + { + "epoch": 2.9982653946227233, + "grad_norm": 0.6773496270179749, + "learning_rate": 1.1605977078195272e-08, + "loss": 0.0021, + "step": 27656 + }, + { + "epoch": 2.998373807458803, + "grad_norm": 0.058111872524023056, + "learning_rate": 1.0880603510808067e-08, + "loss": 0.0005, + "step": 27657 + }, + { + "epoch": 2.998482220294883, + "grad_norm": 0.039474934339523315, + "learning_rate": 1.0155229943420862e-08, + "loss": 0.0004, + "step": 27658 + }, + { + "epoch": 2.998590633130963, + "grad_norm": 0.026758508756756783, + "learning_rate": 9.429856376033659e-09, + "loss": 0.0002, + "step": 27659 + }, + { + "epoch": 2.9986990459670424, + "grad_norm": 0.014908218756318092, + "learning_rate": 8.704482808646453e-09, + "loss": 0.0001, + "step": 27660 + }, + { + "epoch": 2.9988074588031224, + "grad_norm": 0.31263163685798645, + "learning_rate": 7.97910924125925e-09, + "loss": 0.0015, + "step": 27661 + }, + { + "epoch": 2.998915871639202, + "grad_norm": 0.28769901394844055, + "learning_rate": 7.2537356738720454e-09, + "loss": 0.0007, + "step": 27662 + }, + { + "epoch": 2.999024284475282, + "grad_norm": 2.607081890106201, + "learning_rate": 6.52836210648484e-09, + "loss": 0.011, + "step": 27663 + }, + { + "epoch": 2.999132697311362, + "grad_norm": 2.7559611797332764, + "learning_rate": 5.802988539097636e-09, + "loss": 0.0063, + "step": 27664 + }, + { + "epoch": 2.9992411101474414, + "grad_norm": 6.560662746429443, + "learning_rate": 5.077614971710431e-09, + "loss": 0.0413, + "step": 27665 + }, + { + "epoch": 2.999349522983521, + "grad_norm": 6.373676300048828, + "learning_rate": 4.352241404323226e-09, + "loss": 0.0343, + "step": 27666 + }, + { + "epoch": 2.999457935819601, + "grad_norm": 1.665048360824585, + "learning_rate": 3.6268678369360227e-09, + "loss": 0.0252, + "step": 27667 + }, + { + "epoch": 2.999566348655681, + "grad_norm": 0.10561617463827133, + "learning_rate": 2.901494269548818e-09, + "loss": 0.0005, + "step": 27668 + }, + { + "epoch": 2.9996747614917605, + "grad_norm": 0.012634233571588993, + "learning_rate": 2.176120702161613e-09, + "loss": 0.0001, + "step": 27669 + }, + { + "epoch": 2.9997831743278405, + "grad_norm": 0.014291783794760704, + "learning_rate": 1.450747134774409e-09, + "loss": 0.0001, + "step": 27670 + }, + { + "epoch": 2.99989158716392, + "grad_norm": 0.3318418264389038, + "learning_rate": 7.253735673872045e-10, + "loss": 0.0014, + "step": 27671 + }, + { + "epoch": 3.0, + "grad_norm": 0.01911129802465439, + "learning_rate": 0.0, + "loss": 0.0002, + "step": 27672 + }, + { + "epoch": 3.0, + "step": 27672, + "total_flos": 3.556096111225602e+18, + "train_loss": 0.02784311800693059, + "train_runtime": 7197.3424, + "train_samples_per_second": 61.514, + "train_steps_per_second": 3.845 + } + ], + "logging_steps": 1.0, + "max_steps": 27672, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 3.556096111225602e+18, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}