diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7406 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 1052,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0019011406844106464,
+      "grad_norm": 3.7587154010656363,
+      "learning_rate": 1.886792452830189e-07,
+      "loss": 0.6968,
+      "step": 1
+    },
+    {
+      "epoch": 0.0038022813688212928,
+      "grad_norm": 3.362741889686214,
+      "learning_rate": 3.773584905660378e-07,
+      "loss": 0.7135,
+      "step": 2
+    },
+    {
+      "epoch": 0.005703422053231939,
+      "grad_norm": 3.5896130033094114,
+      "learning_rate": 5.660377358490567e-07,
+      "loss": 0.6929,
+      "step": 3
+    },
+    {
+      "epoch": 0.0076045627376425855,
+      "grad_norm": 3.129115337194726,
+      "learning_rate": 7.547169811320755e-07,
+      "loss": 0.6637,
+      "step": 4
+    },
+    {
+      "epoch": 0.009505703422053232,
+      "grad_norm": 3.4664521332726212,
+      "learning_rate": 9.433962264150944e-07,
+      "loss": 0.7198,
+      "step": 5
+    },
+    {
+      "epoch": 0.011406844106463879,
+      "grad_norm": 3.148856383737707,
+      "learning_rate": 1.1320754716981133e-06,
+      "loss": 0.6665,
+      "step": 6
+    },
+    {
+      "epoch": 0.013307984790874524,
+      "grad_norm": 3.2635539186157323,
+      "learning_rate": 1.3207547169811322e-06,
+      "loss": 0.692,
+      "step": 7
+    },
+    {
+      "epoch": 0.015209125475285171,
+      "grad_norm": 3.178967596263942,
+      "learning_rate": 1.509433962264151e-06,
+      "loss": 0.7163,
+      "step": 8
+    },
+    {
+      "epoch": 0.017110266159695818,
+      "grad_norm": 2.709915317767805,
+      "learning_rate": 1.6981132075471698e-06,
+      "loss": 0.6834,
+      "step": 9
+    },
+    {
+      "epoch": 0.019011406844106463,
+      "grad_norm": 2.6001939523532878,
+      "learning_rate": 1.8867924528301889e-06,
+      "loss": 0.6529,
+      "step": 10
+    },
+    {
+      "epoch": 0.02091254752851711,
+      "grad_norm": 1.8169707835844755,
+      "learning_rate": 2.075471698113208e-06,
+      "loss": 0.627,
+      "step": 11
+    },
+    {
+      "epoch": 0.022813688212927757,
+      "grad_norm": 1.6687313319074908,
+      "learning_rate": 2.2641509433962266e-06,
+      "loss": 0.5863,
+      "step": 12
+    },
+    {
+      "epoch": 0.024714828897338403,
+      "grad_norm": 1.612437922136354,
+      "learning_rate": 2.4528301886792453e-06,
+      "loss": 0.5872,
+      "step": 13
+    },
+    {
+      "epoch": 0.026615969581749048,
+      "grad_norm": 1.4489343505128192,
+      "learning_rate": 2.6415094339622644e-06,
+      "loss": 0.5514,
+      "step": 14
+    },
+    {
+      "epoch": 0.028517110266159697,
+      "grad_norm": 1.8584327601242208,
+      "learning_rate": 2.830188679245283e-06,
+      "loss": 0.5739,
+      "step": 15
+    },
+    {
+      "epoch": 0.030418250950570342,
+      "grad_norm": 2.2720710394200436,
+      "learning_rate": 3.018867924528302e-06,
+      "loss": 0.5628,
+      "step": 16
+    },
+    {
+      "epoch": 0.03231939163498099,
+      "grad_norm": 2.1986920489076756,
+      "learning_rate": 3.207547169811321e-06,
+      "loss": 0.5692,
+      "step": 17
+    },
+    {
+      "epoch": 0.034220532319391636,
+      "grad_norm": 1.8282542796474943,
+      "learning_rate": 3.3962264150943395e-06,
+      "loss": 0.5516,
+      "step": 18
+    },
+    {
+      "epoch": 0.03612167300380228,
+      "grad_norm": 1.5661235507670834,
+      "learning_rate": 3.5849056603773586e-06,
+      "loss": 0.5442,
+      "step": 19
+    },
+    {
+      "epoch": 0.03802281368821293,
+      "grad_norm": 1.3878179965965174,
+      "learning_rate": 3.7735849056603777e-06,
+      "loss": 0.535,
+      "step": 20
+    },
+    {
+      "epoch": 0.039923954372623575,
+      "grad_norm": 1.6041640537056563,
+      "learning_rate": 3.962264150943396e-06,
+      "loss": 0.5379,
+      "step": 21
+    },
+    {
+      "epoch": 0.04182509505703422,
+      "grad_norm": 1.4871921337178629,
+      "learning_rate": 4.150943396226416e-06,
+      "loss": 0.5383,
+      "step": 22
+    },
+    {
+      "epoch": 0.043726235741444866,
+      "grad_norm": 1.4050204633759933,
+      "learning_rate": 4.339622641509435e-06,
+      "loss": 0.5366,
+      "step": 23
+    },
+    {
+      "epoch": 0.045627376425855515,
+      "grad_norm": 1.1513347181094025,
+      "learning_rate": 4.528301886792453e-06,
+      "loss": 0.5246,
+      "step": 24
+    },
+    {
+      "epoch": 0.04752851711026616,
+      "grad_norm": 1.1332551448500334,
+      "learning_rate": 4.716981132075472e-06,
+      "loss": 0.5386,
+      "step": 25
+    },
+    {
+      "epoch": 0.049429657794676805,
+      "grad_norm": 1.1517027519762522,
+      "learning_rate": 4.905660377358491e-06,
+      "loss": 0.5601,
+      "step": 26
+    },
+    {
+      "epoch": 0.051330798479087454,
+      "grad_norm": 1.0689019420342514,
+      "learning_rate": 5.09433962264151e-06,
+      "loss": 0.5327,
+      "step": 27
+    },
+    {
+      "epoch": 0.053231939163498096,
+      "grad_norm": 1.0501053721662716,
+      "learning_rate": 5.283018867924529e-06,
+      "loss": 0.5293,
+      "step": 28
+    },
+    {
+      "epoch": 0.055133079847908745,
+      "grad_norm": 1.0515665048262461,
+      "learning_rate": 5.4716981132075475e-06,
+      "loss": 0.4955,
+      "step": 29
+    },
+    {
+      "epoch": 0.057034220532319393,
+      "grad_norm": 0.9298091255316036,
+      "learning_rate": 5.660377358490566e-06,
+      "loss": 0.5081,
+      "step": 30
+    },
+    {
+      "epoch": 0.058935361216730035,
+      "grad_norm": 0.9408762894041993,
+      "learning_rate": 5.849056603773585e-06,
+      "loss": 0.5104,
+      "step": 31
+    },
+    {
+      "epoch": 0.060836501901140684,
+      "grad_norm": 0.8935338868366678,
+      "learning_rate": 6.037735849056604e-06,
+      "loss": 0.4639,
+      "step": 32
+    },
+    {
+      "epoch": 0.06273764258555133,
+      "grad_norm": 0.9721755899098623,
+      "learning_rate": 6.226415094339623e-06,
+      "loss": 0.4911,
+      "step": 33
+    },
+    {
+      "epoch": 0.06463878326996197,
+      "grad_norm": 0.9581768289411134,
+      "learning_rate": 6.415094339622642e-06,
+      "loss": 0.4936,
+      "step": 34
+    },
+    {
+      "epoch": 0.06653992395437262,
+      "grad_norm": 0.9752276454649679,
+      "learning_rate": 6.60377358490566e-06,
+      "loss": 0.5007,
+      "step": 35
+    },
+    {
+      "epoch": 0.06844106463878327,
+      "grad_norm": 0.8733291259301061,
+      "learning_rate": 6.792452830188679e-06,
+      "loss": 0.479,
+      "step": 36
+    },
+    {
+      "epoch": 0.07034220532319392,
+      "grad_norm": 0.8904954990051419,
+      "learning_rate": 6.981132075471699e-06,
+      "loss": 0.4715,
+      "step": 37
+    },
+    {
+      "epoch": 0.07224334600760456,
+      "grad_norm": 0.967859573018954,
+      "learning_rate": 7.169811320754717e-06,
+      "loss": 0.4843,
+      "step": 38
+    },
+    {
+      "epoch": 0.0741444866920152,
+      "grad_norm": 0.9269812233280805,
+      "learning_rate": 7.358490566037736e-06,
+      "loss": 0.4892,
+      "step": 39
+    },
+    {
+      "epoch": 0.07604562737642585,
+      "grad_norm": 0.9108140699204871,
+      "learning_rate": 7.5471698113207555e-06,
+      "loss": 0.4642,
+      "step": 40
+    },
+    {
+      "epoch": 0.0779467680608365,
+      "grad_norm": 0.9660717497951608,
+      "learning_rate": 7.735849056603775e-06,
+      "loss": 0.4828,
+      "step": 41
+    },
+    {
+      "epoch": 0.07984790874524715,
+      "grad_norm": 0.8734941727606541,
+      "learning_rate": 7.924528301886793e-06,
+      "loss": 0.4643,
+      "step": 42
+    },
+    {
+      "epoch": 0.0817490494296578,
+      "grad_norm": 1.0123926941455381,
+      "learning_rate": 8.113207547169812e-06,
+      "loss": 0.4656,
+      "step": 43
+    },
+    {
+      "epoch": 0.08365019011406843,
+      "grad_norm": 0.8793753916842706,
+      "learning_rate": 8.301886792452832e-06,
+      "loss": 0.4568,
+      "step": 44
+    },
+    {
+      "epoch": 0.08555133079847908,
+      "grad_norm": 0.8856372795938922,
+      "learning_rate": 8.49056603773585e-06,
+      "loss": 0.4658,
+      "step": 45
+    },
+    {
+      "epoch": 0.08745247148288973,
+      "grad_norm": 0.8965193046559465,
+      "learning_rate": 8.67924528301887e-06,
+      "loss": 0.4674,
+      "step": 46
+    },
+    {
+      "epoch": 0.08935361216730038,
+      "grad_norm": 0.8699265702868152,
+      "learning_rate": 8.867924528301887e-06,
+      "loss": 0.4699,
+      "step": 47
+    },
+    {
+      "epoch": 0.09125475285171103,
+      "grad_norm": 0.8683493743735327,
+      "learning_rate": 9.056603773584907e-06,
+      "loss": 0.471,
+      "step": 48
+    },
+    {
+      "epoch": 0.09315589353612168,
+      "grad_norm": 0.9349939464306016,
+      "learning_rate": 9.245283018867926e-06,
+      "loss": 0.4767,
+      "step": 49
+    },
+    {
+      "epoch": 0.09505703422053231,
+      "grad_norm": 0.8645137385128104,
+      "learning_rate": 9.433962264150944e-06,
+      "loss": 0.4699,
+      "step": 50
+    },
+    {
+      "epoch": 0.09695817490494296,
+      "grad_norm": 0.9274273429803022,
+      "learning_rate": 9.622641509433963e-06,
+      "loss": 0.4634,
+      "step": 51
+    },
+    {
+      "epoch": 0.09885931558935361,
+      "grad_norm": 0.9230697151987387,
+      "learning_rate": 9.811320754716981e-06,
+      "loss": 0.4713,
+      "step": 52
+    },
+    {
+      "epoch": 0.10076045627376426,
+      "grad_norm": 1.1764052162479033,
+      "learning_rate": 1e-05,
+      "loss": 0.4634,
+      "step": 53
+    },
+    {
+      "epoch": 0.10266159695817491,
+      "grad_norm": 0.9156676260966872,
+      "learning_rate": 1.018867924528302e-05,
+      "loss": 0.4793,
+      "step": 54
+    },
+    {
+      "epoch": 0.10456273764258556,
+      "grad_norm": 1.1346300246553258,
+      "learning_rate": 1.0377358490566038e-05,
+      "loss": 0.4794,
+      "step": 55
+    },
+    {
+      "epoch": 0.10646387832699619,
+      "grad_norm": 0.9634324140489856,
+      "learning_rate": 1.0566037735849058e-05,
+      "loss": 0.4643,
+      "step": 56
+    },
+    {
+      "epoch": 0.10836501901140684,
+      "grad_norm": 0.8630802969654159,
+      "learning_rate": 1.0754716981132076e-05,
+      "loss": 0.4409,
+      "step": 57
+    },
+    {
+      "epoch": 0.11026615969581749,
+      "grad_norm": 0.9763312718222703,
+      "learning_rate": 1.0943396226415095e-05,
+      "loss": 0.4656,
+      "step": 58
+    },
+    {
+      "epoch": 0.11216730038022814,
+      "grad_norm": 0.9816001471655289,
+      "learning_rate": 1.1132075471698115e-05,
+      "loss": 0.4513,
+      "step": 59
+    },
+    {
+      "epoch": 0.11406844106463879,
+      "grad_norm": 0.9695065756889812,
+      "learning_rate": 1.1320754716981132e-05,
+      "loss": 0.4777,
+      "step": 60
+    },
+    {
+      "epoch": 0.11596958174904944,
+      "grad_norm": 1.0977384638649899,
+      "learning_rate": 1.1509433962264152e-05,
+      "loss": 0.4391,
+      "step": 61
+    },
+    {
+      "epoch": 0.11787072243346007,
+      "grad_norm": 0.9593913599738026,
+      "learning_rate": 1.169811320754717e-05,
+      "loss": 0.4652,
+      "step": 62
+    },
+    {
+      "epoch": 0.11977186311787072,
+      "grad_norm": 0.8957672419005775,
+      "learning_rate": 1.188679245283019e-05,
+      "loss": 0.4409,
+      "step": 63
+    },
+    {
+      "epoch": 0.12167300380228137,
+      "grad_norm": 1.036557734592175,
+      "learning_rate": 1.2075471698113209e-05,
+      "loss": 0.4902,
+      "step": 64
+    },
+    {
+      "epoch": 0.12357414448669202,
+      "grad_norm": 0.9159986109184152,
+      "learning_rate": 1.2264150943396227e-05,
+      "loss": 0.4711,
+      "step": 65
+    },
+    {
+      "epoch": 0.12547528517110265,
+      "grad_norm": 0.9157436424614411,
+      "learning_rate": 1.2452830188679246e-05,
+      "loss": 0.4468,
+      "step": 66
+    },
+    {
+      "epoch": 0.12737642585551331,
+      "grad_norm": 0.9531601255168413,
+      "learning_rate": 1.2641509433962264e-05,
+      "loss": 0.4563,
+      "step": 67
+    },
+    {
+      "epoch": 0.12927756653992395,
+      "grad_norm": 0.8928606504716562,
+      "learning_rate": 1.2830188679245283e-05,
+      "loss": 0.4648,
+      "step": 68
+    },
+    {
+      "epoch": 0.1311787072243346,
+      "grad_norm": 1.294929519516281,
+      "learning_rate": 1.3018867924528303e-05,
+      "loss": 0.4845,
+      "step": 69
+    },
+    {
+      "epoch": 0.13307984790874525,
+      "grad_norm": 0.9095331465783749,
+      "learning_rate": 1.320754716981132e-05,
+      "loss": 0.457,
+      "step": 70
+    },
+    {
+      "epoch": 0.13498098859315588,
+      "grad_norm": 0.9759136641098354,
+      "learning_rate": 1.339622641509434e-05,
+      "loss": 0.4535,
+      "step": 71
+    },
+    {
+      "epoch": 0.13688212927756654,
+      "grad_norm": 1.0288586843539438,
+      "learning_rate": 1.3584905660377358e-05,
+      "loss": 0.4617,
+      "step": 72
+    },
+    {
+      "epoch": 0.13878326996197718,
+      "grad_norm": 0.8657516488562751,
+      "learning_rate": 1.3773584905660378e-05,
+      "loss": 0.4704,
+      "step": 73
+    },
+    {
+      "epoch": 0.14068441064638784,
+      "grad_norm": 0.9753927198358316,
+      "learning_rate": 1.3962264150943397e-05,
+      "loss": 0.4306,
+      "step": 74
+    },
+    {
+      "epoch": 0.14258555133079848,
+      "grad_norm": 0.9090424913964088,
+      "learning_rate": 1.4150943396226415e-05,
+      "loss": 0.4414,
+      "step": 75
+    },
+    {
+      "epoch": 0.1444866920152091,
+      "grad_norm": 0.8707480642407023,
+      "learning_rate": 1.4339622641509435e-05,
+      "loss": 0.4598,
+      "step": 76
+    },
+    {
+      "epoch": 0.14638783269961977,
+      "grad_norm": 1.0069627349198476,
+      "learning_rate": 1.4528301886792452e-05,
+      "loss": 0.4808,
+      "step": 77
+    },
+    {
+      "epoch": 0.1482889733840304,
+      "grad_norm": 0.9105067107270177,
+      "learning_rate": 1.4716981132075472e-05,
+      "loss": 0.445,
+      "step": 78
+    },
+    {
+      "epoch": 0.15019011406844107,
+      "grad_norm": 0.9218190841283949,
+      "learning_rate": 1.4905660377358491e-05,
+      "loss": 0.4554,
+      "step": 79
+    },
+    {
+      "epoch": 0.1520912547528517,
+      "grad_norm": 1.020573291542075,
+      "learning_rate": 1.5094339622641511e-05,
+      "loss": 0.4761,
+      "step": 80
+    },
+    {
+      "epoch": 0.15399239543726237,
+      "grad_norm": 0.8435667653721475,
+      "learning_rate": 1.5283018867924532e-05,
+      "loss": 0.4659,
+      "step": 81
+    },
+    {
+      "epoch": 0.155893536121673,
+      "grad_norm": 0.9569290276016456,
+      "learning_rate": 1.547169811320755e-05,
+      "loss": 0.4407,
+      "step": 82
+    },
+    {
+      "epoch": 0.15779467680608364,
+      "grad_norm": 0.9398905495824446,
+      "learning_rate": 1.5660377358490568e-05,
+      "loss": 0.4645,
+      "step": 83
+    },
+    {
+      "epoch": 0.1596958174904943,
+      "grad_norm": 1.1018200288551356,
+      "learning_rate": 1.5849056603773586e-05,
+      "loss": 0.4708,
+      "step": 84
+    },
+    {
+      "epoch": 0.16159695817490494,
+      "grad_norm": 1.0624840571756866,
+      "learning_rate": 1.6037735849056607e-05,
+      "loss": 0.4432,
+      "step": 85
+    },
+    {
+      "epoch": 0.1634980988593156,
+      "grad_norm": 1.0706654234468287,
+      "learning_rate": 1.6226415094339625e-05,
+      "loss": 0.4339,
+      "step": 86
+    },
+    {
+      "epoch": 0.16539923954372623,
+      "grad_norm": 0.9832014045257631,
+      "learning_rate": 1.6415094339622643e-05,
+      "loss": 0.4598,
+      "step": 87
+    },
+    {
+      "epoch": 0.16730038022813687,
+      "grad_norm": 0.8606511269860765,
+      "learning_rate": 1.6603773584905664e-05,
+      "loss": 0.4491,
+      "step": 88
+    },
+    {
+      "epoch": 0.16920152091254753,
+      "grad_norm": 0.9758920091640942,
+      "learning_rate": 1.679245283018868e-05,
+      "loss": 0.4734,
+      "step": 89
+    },
+    {
+      "epoch": 0.17110266159695817,
+      "grad_norm": 0.988312178116455,
+      "learning_rate": 1.69811320754717e-05,
+      "loss": 0.465,
+      "step": 90
+    },
+    {
+      "epoch": 0.17300380228136883,
+      "grad_norm": 0.9669130125766525,
+      "learning_rate": 1.716981132075472e-05,
+      "loss": 0.4671,
+      "step": 91
+    },
+    {
+      "epoch": 0.17490494296577946,
+      "grad_norm": 1.274369798841813,
+      "learning_rate": 1.735849056603774e-05,
+      "loss": 0.4829,
+      "step": 92
+    },
+    {
+      "epoch": 0.17680608365019013,
+      "grad_norm": 0.8098281325891173,
+      "learning_rate": 1.7547169811320756e-05,
+      "loss": 0.4362,
+      "step": 93
+    },
+    {
+      "epoch": 0.17870722433460076,
+      "grad_norm": 1.0052478570017351,
+      "learning_rate": 1.7735849056603774e-05,
+      "loss": 0.4574,
+      "step": 94
+    },
+    {
+      "epoch": 0.1806083650190114,
+      "grad_norm": 1.0812592927912872,
+      "learning_rate": 1.7924528301886795e-05,
+      "loss": 0.4703,
+      "step": 95
+    },
+    {
+      "epoch": 0.18250950570342206,
+      "grad_norm": 0.896344657500704,
+      "learning_rate": 1.8113207547169813e-05,
+      "loss": 0.4524,
+      "step": 96
+    },
+    {
+      "epoch": 0.1844106463878327,
+      "grad_norm": 1.1592992056764342,
+      "learning_rate": 1.830188679245283e-05,
+      "loss": 0.4644,
+      "step": 97
+    },
+    {
+      "epoch": 0.18631178707224336,
+      "grad_norm": 0.8739226310543647,
+      "learning_rate": 1.8490566037735852e-05,
+      "loss": 0.4224,
+      "step": 98
+    },
+    {
+      "epoch": 0.188212927756654,
+      "grad_norm": 0.8922170029831241,
+      "learning_rate": 1.867924528301887e-05,
+      "loss": 0.4565,
+      "step": 99
+    },
+    {
+      "epoch": 0.19011406844106463,
+      "grad_norm": 1.1833096225117925,
+      "learning_rate": 1.8867924528301888e-05,
+      "loss": 0.4548,
+      "step": 100
+    },
+    {
+      "epoch": 0.1920152091254753,
+      "grad_norm": 0.8774292577559527,
+      "learning_rate": 1.905660377358491e-05,
+      "loss": 0.4713,
+      "step": 101
+    },
+    {
+      "epoch": 0.19391634980988592,
+      "grad_norm": 1.011387332375806,
+      "learning_rate": 1.9245283018867927e-05,
+      "loss": 0.4431,
+      "step": 102
+    },
+    {
+      "epoch": 0.1958174904942966,
+      "grad_norm": 0.821841067276584,
+      "learning_rate": 1.9433962264150945e-05,
+      "loss": 0.4786,
+      "step": 103
+    },
+    {
+      "epoch": 0.19771863117870722,
+      "grad_norm": 0.9293558236221588,
+      "learning_rate": 1.9622641509433963e-05,
+      "loss": 0.4425,
+      "step": 104
+    },
+    {
+      "epoch": 0.19961977186311788,
+      "grad_norm": 0.9272021075252388,
+      "learning_rate": 1.9811320754716984e-05,
+      "loss": 0.4667,
+      "step": 105
+    },
+    {
+      "epoch": 0.20152091254752852,
+      "grad_norm": 0.834595094641444,
+      "learning_rate": 2e-05,
+      "loss": 0.4571,
+      "step": 106
+    },
+    {
+      "epoch": 0.20342205323193915,
+      "grad_norm": 0.9335242186265841,
+      "learning_rate": 1.9999944857420527e-05,
+      "loss": 0.4365,
+      "step": 107
+    },
+    {
+      "epoch": 0.20532319391634982,
+      "grad_norm": 0.8700522970125515,
+      "learning_rate": 1.9999779430290247e-05,
+      "loss": 0.4491,
+      "step": 108
+    },
+    {
+      "epoch": 0.20722433460076045,
+      "grad_norm": 0.9238091269927581,
+      "learning_rate": 1.9999503720433575e-05,
+      "loss": 0.4617,
+      "step": 109
+    },
+    {
+      "epoch": 0.20912547528517111,
+      "grad_norm": 0.9084779069556055,
+      "learning_rate": 1.999911773089118e-05,
+      "loss": 0.4399,
+      "step": 110
+    },
+    {
+      "epoch": 0.21102661596958175,
+      "grad_norm": 0.867661445176709,
+      "learning_rate": 1.999862146591996e-05,
+      "loss": 0.4377,
+      "step": 111
+    },
+    {
+      "epoch": 0.21292775665399238,
+      "grad_norm": 0.9440495894897893,
+      "learning_rate": 1.9998014930992976e-05,
+      "loss": 0.4478,
+      "step": 112
+    },
+    {
+      "epoch": 0.21482889733840305,
+      "grad_norm": 0.9160613556830745,
+      "learning_rate": 1.9997298132799408e-05,
+      "loss": 0.4682,
+      "step": 113
+    },
+    {
+      "epoch": 0.21673003802281368,
+      "grad_norm": 0.9204449054828825,
+      "learning_rate": 1.9996471079244477e-05,
+      "loss": 0.4798,
+      "step": 114
+    },
+    {
+      "epoch": 0.21863117870722434,
+      "grad_norm": 1.0162735057047352,
+      "learning_rate": 1.999553377944936e-05,
+      "loss": 0.4386,
+      "step": 115
+    },
+    {
+      "epoch": 0.22053231939163498,
+      "grad_norm": 0.8387932169822102,
+      "learning_rate": 1.9994486243751076e-05,
+      "loss": 0.4644,
+      "step": 116
+    },
+    {
+      "epoch": 0.2224334600760456,
+      "grad_norm": 0.9805193927104513,
+      "learning_rate": 1.9993328483702393e-05,
+      "loss": 0.4509,
+      "step": 117
+    },
+    {
+      "epoch": 0.22433460076045628,
+      "grad_norm": 0.8794007625259703,
+      "learning_rate": 1.999206051207169e-05,
+      "loss": 0.4663,
+      "step": 118
+    },
+    {
+      "epoch": 0.2262357414448669,
+      "grad_norm": 0.8715342892907946,
+      "learning_rate": 1.9990682342842805e-05,
+      "loss": 0.4566,
+      "step": 119
+    },
+    {
+      "epoch": 0.22813688212927757,
+      "grad_norm": 0.9487738970190361,
+      "learning_rate": 1.99891939912149e-05,
+      "loss": 0.4396,
+      "step": 120
+    },
+    {
+      "epoch": 0.2300380228136882,
+      "grad_norm": 1.1915130795368503,
+      "learning_rate": 1.9987595473602292e-05,
+      "loss": 0.4644,
+      "step": 121
+    },
+    {
+      "epoch": 0.23193916349809887,
+      "grad_norm": 0.8943531299965776,
+      "learning_rate": 1.9985886807634246e-05,
+      "loss": 0.4862,
+      "step": 122
+    },
+    {
+      "epoch": 0.2338403041825095,
+      "grad_norm": 0.9505181701096156,
+      "learning_rate": 1.9984068012154824e-05,
+      "loss": 0.4382,
+      "step": 123
+    },
+    {
+      "epoch": 0.23574144486692014,
+      "grad_norm": 1.0213990009679197,
+      "learning_rate": 1.9982139107222634e-05,
+      "loss": 0.4561,
+      "step": 124
+    },
+    {
+      "epoch": 0.2376425855513308,
+      "grad_norm": 1.0001814161308735,
+      "learning_rate": 1.9980100114110637e-05,
+      "loss": 0.4505,
+      "step": 125
+    },
+    {
+      "epoch": 0.23954372623574144,
+      "grad_norm": 1.0965677177589834,
+      "learning_rate": 1.99779510553059e-05,
+      "loss": 0.4584,
+      "step": 126
+    },
+    {
+      "epoch": 0.2414448669201521,
+      "grad_norm": 0.8305227098716941,
+      "learning_rate": 1.9975691954509347e-05,
+      "loss": 0.4751,
+      "step": 127
+    },
+    {
+      "epoch": 0.24334600760456274,
+      "grad_norm": 1.0328532350738455,
+      "learning_rate": 1.9973322836635517e-05,
+      "loss": 0.4454,
+      "step": 128
+    },
+    {
+      "epoch": 0.24524714828897337,
+      "grad_norm": 0.9131979487268771,
+      "learning_rate": 1.997084372781226e-05,
+      "loss": 0.4765,
+      "step": 129
+    },
+    {
+      "epoch": 0.24714828897338403,
+      "grad_norm": 0.8997989471521403,
+      "learning_rate": 1.9968254655380465e-05,
+      "loss": 0.4382,
+      "step": 130
+    },
+    {
+      "epoch": 0.24904942965779467,
+      "grad_norm": 1.022913186430162,
+      "learning_rate": 1.996555564789376e-05,
+      "loss": 0.455,
+      "step": 131
+    },
+    {
+      "epoch": 0.2509505703422053,
+      "grad_norm": 0.8454141967612432,
+      "learning_rate": 1.996274673511819e-05,
+      "loss": 0.4257,
+      "step": 132
+    },
+    {
+      "epoch": 0.25285171102661597,
+      "grad_norm": 0.9496397628574214,
+      "learning_rate": 1.99598279480319e-05,
+      "loss": 0.4527,
+      "step": 133
+    },
+    {
+      "epoch": 0.25475285171102663,
+      "grad_norm": 0.8672892296978352,
+      "learning_rate": 1.9956799318824776e-05,
+      "loss": 0.4354,
+      "step": 134
+    },
+    {
+      "epoch": 0.25665399239543724,
+      "grad_norm": 0.9602041893539319,
+      "learning_rate": 1.99536608808981e-05,
+      "loss": 0.4619,
+      "step": 135
+    },
+    {
+      "epoch": 0.2585551330798479,
+      "grad_norm": 1.0102999738185805,
+      "learning_rate": 1.995041266886419e-05,
+      "loss": 0.4575,
+      "step": 136
+    },
+    {
+      "epoch": 0.26045627376425856,
+      "grad_norm": 0.8676619014978241,
+      "learning_rate": 1.9947054718545996e-05,
+      "loss": 0.4545,
+      "step": 137
+    },
+    {
+      "epoch": 0.2623574144486692,
+      "grad_norm": 0.9435908473741579,
+      "learning_rate": 1.994358706697674e-05,
+      "loss": 0.4631,
+      "step": 138
+    },
+    {
+      "epoch": 0.26425855513307983,
+      "grad_norm": 0.9011850195999636,
+      "learning_rate": 1.9940009752399462e-05,
+      "loss": 0.4673,
+      "step": 139
+    },
+    {
+      "epoch": 0.2661596958174905,
+      "grad_norm": 1.0201988433006477,
+      "learning_rate": 1.9936322814266634e-05,
+      "loss": 0.4661,
+      "step": 140
+    },
+    {
+      "epoch": 0.26806083650190116,
+      "grad_norm": 0.9070023246839101,
+      "learning_rate": 1.9932526293239713e-05,
+      "loss": 0.4503,
+      "step": 141
+    },
+    {
+      "epoch": 0.26996197718631176,
+      "grad_norm": 0.833565267357631,
+      "learning_rate": 1.9928620231188694e-05,
+      "loss": 0.4419,
+      "step": 142
+    },
+    {
+      "epoch": 0.2718631178707224,
+      "grad_norm": 1.0825464254239512,
+      "learning_rate": 1.992460467119164e-05,
+      "loss": 0.4639,
+      "step": 143
+    },
+    {
+      "epoch": 0.2737642585551331,
+      "grad_norm": 0.8848149746232583,
+      "learning_rate": 1.992047965753422e-05,
+      "loss": 0.4767,
+      "step": 144
+    },
+    {
+      "epoch": 0.27566539923954375,
+      "grad_norm": 0.8922043871545696,
+      "learning_rate": 1.991624523570922e-05,
+      "loss": 0.4585,
+      "step": 145
+    },
+    {
+      "epoch": 0.27756653992395436,
+      "grad_norm": 1.0049154506291247,
+      "learning_rate": 1.9911901452416012e-05,
+      "loss": 0.4424,
+      "step": 146
+    },
+    {
+      "epoch": 0.279467680608365,
+      "grad_norm": 0.8240094093567454,
+      "learning_rate": 1.9907448355560094e-05,
+      "loss": 0.4452,
+      "step": 147
+    },
+    {
+      "epoch": 0.2813688212927757,
+      "grad_norm": 1.0600074463699127,
+      "learning_rate": 1.9902885994252506e-05,
+      "loss": 0.4578,
+      "step": 148
+    },
+    {
+      "epoch": 0.2832699619771863,
+      "grad_norm": 0.8846660827117652,
+      "learning_rate": 1.989821441880933e-05,
+      "loss": 0.4564,
+      "step": 149
+    },
+    {
+      "epoch": 0.28517110266159695,
+      "grad_norm": 0.9961956949468344,
+      "learning_rate": 1.9893433680751105e-05,
+      "loss": 0.4374,
+      "step": 150
+    },
+    {
+      "epoch": 0.2870722433460076,
+      "grad_norm": 0.8310760138947617,
+      "learning_rate": 1.9888543832802277e-05,
+      "loss": 0.4263,
+      "step": 151
+    },
+    {
+      "epoch": 0.2889733840304182,
+      "grad_norm": 0.8268687504317589,
+      "learning_rate": 1.9883544928890612e-05,
+      "loss": 0.4508,
+      "step": 152
+    },
+    {
+      "epoch": 0.2908745247148289,
+      "grad_norm": 0.8841602005584022,
+      "learning_rate": 1.9878437024146603e-05,
+      "loss": 0.4885,
+      "step": 153
+    },
+    {
+      "epoch": 0.29277566539923955,
+      "grad_norm": 0.9321271483481541,
+      "learning_rate": 1.9873220174902857e-05,
+      "loss": 0.4457,
+      "step": 154
+    },
+    {
+      "epoch": 0.2946768060836502,
+      "grad_norm": 0.9873670559376374,
+      "learning_rate": 1.986789443869348e-05,
+      "loss": 0.4576,
+      "step": 155
+    },
+    {
+      "epoch": 0.2965779467680608,
+      "grad_norm": 0.8900834790852145,
+      "learning_rate": 1.9862459874253438e-05,
+      "loss": 0.4228,
+      "step": 156
+    },
+    {
+      "epoch": 0.2984790874524715,
+      "grad_norm": 0.9059448973047894,
+      "learning_rate": 1.985691654151791e-05,
+      "loss": 0.4338,
+      "step": 157
+    },
+    {
+      "epoch": 0.30038022813688214,
+      "grad_norm": 0.7832858462149278,
+      "learning_rate": 1.9851264501621635e-05,
+      "loss": 0.4578,
+      "step": 158
+    },
+    {
+      "epoch": 0.30228136882129275,
+      "grad_norm": 0.936383302395317,
+      "learning_rate": 1.984550381689822e-05,
+      "loss": 0.4588,
+      "step": 159
+    },
+    {
+      "epoch": 0.3041825095057034,
+      "grad_norm": 0.8217133091416859,
+      "learning_rate": 1.983963455087946e-05,
+      "loss": 0.4566,
+      "step": 160
+    },
+    {
+      "epoch": 0.3060836501901141,
+      "grad_norm": 0.8498115958843986,
+      "learning_rate": 1.983365676829466e-05,
+      "loss": 0.4558,
+      "step": 161
+    },
+    {
+      "epoch": 0.30798479087452474,
+      "grad_norm": 0.8016402808974186,
+      "learning_rate": 1.982757053506989e-05,
+      "loss": 0.4489,
+      "step": 162
+    },
+    {
+      "epoch": 0.30988593155893535,
+      "grad_norm": 0.7751229367129853,
+      "learning_rate": 1.9821375918327268e-05,
+      "loss": 0.443,
+      "step": 163
+    },
+    {
+      "epoch": 0.311787072243346,
+      "grad_norm": 0.8059104159920085,
+      "learning_rate": 1.981507298638422e-05,
+      "loss": 0.4147,
+      "step": 164
+    },
+    {
+      "epoch": 0.31368821292775667,
+      "grad_norm": 0.8536694610287676,
+      "learning_rate": 1.9808661808752735e-05,
+      "loss": 0.4574,
+      "step": 165
+    },
+    {
+      "epoch": 0.3155893536121673,
+      "grad_norm": 0.9206053777136621,
+      "learning_rate": 1.980214245613858e-05,
+      "loss": 0.4484,
+      "step": 166
+    },
+    {
+      "epoch": 0.31749049429657794,
+      "grad_norm": 0.8391503913062355,
+      "learning_rate": 1.979551500044055e-05,
+      "loss": 0.4528,
+      "step": 167
+    },
+    {
+      "epoch": 0.3193916349809886,
+      "grad_norm": 0.8623820469757025,
+      "learning_rate": 1.9788779514749635e-05,
+      "loss": 0.4686,
+      "step": 168
+    },
+    {
+      "epoch": 0.32129277566539927,
+      "grad_norm": 0.886429911271713,
+      "learning_rate": 1.978193607334826e-05,
+      "loss": 0.4385,
+      "step": 169
+    },
+    {
+      "epoch": 0.3231939163498099,
+      "grad_norm": 1.1241448343196239,
+      "learning_rate": 1.977498475170941e-05,
+      "loss": 0.4603,
+      "step": 170
+    },
+    {
+      "epoch": 0.32509505703422054,
+      "grad_norm": 0.8133920725428255,
+      "learning_rate": 1.9767925626495857e-05,
+      "loss": 0.4475,
+      "step": 171
+    },
+    {
+      "epoch": 0.3269961977186312,
+      "grad_norm": 0.9167370292690682,
+      "learning_rate": 1.9760758775559275e-05,
+      "loss": 0.4469,
+      "step": 172
+    },
+    {
+      "epoch": 0.3288973384030418,
+      "grad_norm": 0.9596197488154138,
+      "learning_rate": 1.975348427793939e-05,
+      "loss": 0.4588,
+      "step": 173
+    },
+    {
+      "epoch": 0.33079847908745247,
+      "grad_norm": 0.8172335619675806,
+      "learning_rate": 1.9746102213863113e-05,
+      "loss": 0.4518,
+      "step": 174
+    },
+    {
+      "epoch": 0.33269961977186313,
+      "grad_norm": 1.0364959651182555,
+      "learning_rate": 1.973861266474366e-05,
+      "loss": 0.4525,
+      "step": 175
+    },
+    {
+      "epoch": 0.33460076045627374,
+      "grad_norm": 0.7896657449125826,
+      "learning_rate": 1.9731015713179643e-05,
+      "loss": 0.4544,
+      "step": 176
+    },
+    {
+      "epoch": 0.3365019011406844,
+      "grad_norm": 0.9376942547994319,
+      "learning_rate": 1.9723311442954163e-05,
+      "loss": 0.4588,
+      "step": 177
+    },
+    {
+      "epoch": 0.33840304182509506,
+      "grad_norm": 0.8415897873276912,
+      "learning_rate": 1.9715499939033883e-05,
+      "loss": 0.4329,
+      "step": 178
+    },
+    {
+      "epoch": 0.3403041825095057,
+      "grad_norm": 0.7491819624029536,
+      "learning_rate": 1.9707581287568094e-05,
+      "loss": 0.4336,
+      "step": 179
+    },
+    {
+      "epoch": 0.34220532319391633,
+      "grad_norm": 1.069033618481511,
+      "learning_rate": 1.969955557588778e-05,
+      "loss": 0.4327,
+      "step": 180
+    },
+    {
+      "epoch": 0.344106463878327,
+      "grad_norm": 0.7702422975501193,
+      "learning_rate": 1.9691422892504626e-05,
+      "loss": 0.4306,
+      "step": 181
+    },
+    {
+      "epoch": 0.34600760456273766,
+      "grad_norm": 1.033680894215674,
+      "learning_rate": 1.968318332711006e-05,
+      "loss": 0.46,
+      "step": 182
+    },
+    {
+      "epoch": 0.34790874524714827,
+      "grad_norm": 0.9378116290695766,
+      "learning_rate": 1.9674836970574253e-05,
+      "loss": 0.4818,
+      "step": 183
+    },
+    {
+      "epoch": 0.34980988593155893,
+      "grad_norm": 0.8387791048659992,
+      "learning_rate": 1.966638391494514e-05,
+      "loss": 0.4475,
+      "step": 184
+    },
+    {
+      "epoch": 0.3517110266159696,
+      "grad_norm": 1.0621849806400059,
+      "learning_rate": 1.9657824253447378e-05,
+      "loss": 0.4277,
+      "step": 185
+    },
+    {
+      "epoch": 0.35361216730038025,
+      "grad_norm": 0.8872428493389745,
+      "learning_rate": 1.9649158080481327e-05,
+      "loss": 0.4584,
+      "step": 186
+    },
+    {
+      "epoch": 0.35551330798479086,
+      "grad_norm": 0.8385107279025317,
+      "learning_rate": 1.964038549162201e-05,
+      "loss": 0.4253,
+      "step": 187
+    },
+    {
+      "epoch": 0.3574144486692015,
+      "grad_norm": 1.1342322954222368,
+      "learning_rate": 1.963150658361807e-05,
+      "loss": 0.4383,
+      "step": 188
+    },
+    {
+      "epoch": 0.3593155893536122,
+      "grad_norm": 0.7710295766471317,
+      "learning_rate": 1.962252145439068e-05,
+      "loss": 0.4272,
+      "step": 189
+    },
+    {
+      "epoch": 0.3612167300380228,
+      "grad_norm": 0.943614976507202,
+      "learning_rate": 1.9613430203032486e-05,
+      "loss": 0.4345,
+      "step": 190
+    },
+    {
+      "epoch": 0.36311787072243346,
+      "grad_norm": 1.0747882871628456,
+      "learning_rate": 1.9604232929806493e-05,
+      "loss": 0.4407,
+      "step": 191
+    },
+    {
+      "epoch": 0.3650190114068441,
+      "grad_norm": 0.8703814589438721,
+      "learning_rate": 1.9594929736144978e-05,
+      "loss": 0.4321,
+      "step": 192
+    },
+    {
+      "epoch": 0.3669201520912547,
+      "grad_norm": 0.8964800239866031,
+      "learning_rate": 1.9585520724648354e-05,
+      "loss": 0.4337,
+      "step": 193
+    },
+    {
+      "epoch": 0.3688212927756654,
+      "grad_norm": 0.846801285526771,
+      "learning_rate": 1.957600599908406e-05,
+      "loss": 0.4368,
+      "step": 194
+    },
+    {
+      "epoch": 0.37072243346007605,
+      "grad_norm": 0.9190940443815219,
+      "learning_rate": 1.95663856643854e-05,
+      "loss": 0.4438,
+      "step": 195
+    },
+    {
+      "epoch": 0.3726235741444867,
+      "grad_norm": 0.8852933595722409,
+      "learning_rate": 1.955665982665038e-05,
+      "loss": 0.4278,
+      "step": 196
+    },
+    {
+      "epoch": 0.3745247148288973,
+      "grad_norm": 0.7338028012229248,
+      "learning_rate": 1.9546828593140565e-05,
+      "loss": 0.4204,
+      "step": 197
+    },
+    {
+      "epoch": 0.376425855513308,
+      "grad_norm": 0.8880827920619041,
+      "learning_rate": 1.9536892072279863e-05,
+      "loss": 0.4505,
+      "step": 198
+    },
+    {
+      "epoch": 0.37832699619771865,
+      "grad_norm": 0.8055465213716715,
+      "learning_rate": 1.9526850373653356e-05,
+      "loss": 0.4692,
+      "step": 199
+    },
+    {
+      "epoch": 0.38022813688212925,
+      "grad_norm": 0.7385860605150167,
+      "learning_rate": 1.9516703608006074e-05,
+      "loss": 0.3969,
+      "step": 200
+    },
+    {
+      "epoch": 0.3821292775665399,
+      "grad_norm": 0.7998871780527864,
+      "learning_rate": 1.9506451887241787e-05,
+      "loss": 0.43,
+      "step": 201
+    },
+    {
+      "epoch": 0.3840304182509506,
+      "grad_norm": 0.7729734293304378,
+      "learning_rate": 1.949609532442176e-05,
+      "loss": 0.4433,
+      "step": 202
+    },
+    {
+      "epoch": 0.38593155893536124,
+      "grad_norm": 0.8075690292130411,
+      "learning_rate": 1.9485634033763507e-05,
+      "loss": 0.4199,
+      "step": 203
+    },
+    {
+      "epoch": 0.38783269961977185,
+      "grad_norm": 0.8006024972032324,
+      "learning_rate": 1.9475068130639543e-05,
+      "loss": 0.4157,
+      "step": 204
+    },
+    {
+      "epoch": 0.3897338403041825,
+      "grad_norm": 0.805198471242757,
+      "learning_rate": 1.9464397731576093e-05,
+      "loss": 0.448,
+      "step": 205
+    },
+    {
+      "epoch": 0.3916349809885932,
+      "grad_norm": 0.8198823759777583,
+      "learning_rate": 1.945362295425183e-05,
+      "loss": 0.4411,
+      "step": 206
+    },
+    {
+      "epoch": 0.3935361216730038,
+      "grad_norm": 0.8376863963373794,
+      "learning_rate": 1.944274391749655e-05,
+      "loss": 0.4281,
+      "step": 207
+    },
+    {
+      "epoch": 0.39543726235741444,
+      "grad_norm": 0.8182021117836741,
+      "learning_rate": 1.9431760741289886e-05,
+      "loss": 0.4209,
+      "step": 208
+    },
+    {
+      "epoch": 0.3973384030418251,
+      "grad_norm": 0.7919911464163376,
+      "learning_rate": 1.942067354675997e-05,
+      "loss": 0.4439,
+      "step": 209
+    },
+    {
+      "epoch": 0.39923954372623577,
+      "grad_norm": 0.7881627781393306,
+      "learning_rate": 1.9409482456182105e-05,
+      "loss": 0.4284,
+      "step": 210
+    },
+    {
+      "epoch": 0.4011406844106464,
+      "grad_norm": 0.7655466240593265,
+      "learning_rate": 1.939818759297741e-05,
+      "loss": 0.4241,
+      "step": 211
+    },
+    {
+      "epoch": 0.40304182509505704,
+      "grad_norm": 0.7874252908247497,
+      "learning_rate": 1.9386789081711465e-05,
+      "loss": 0.4321,
+      "step": 212
+    },
+    {
+      "epoch": 0.4049429657794677,
+      "grad_norm": 0.7823202283489393,
+      "learning_rate": 1.9375287048092927e-05,
+      "loss": 0.448,
+      "step": 213
+    },
+    {
+      "epoch": 0.4068441064638783,
+      "grad_norm": 0.7545510398894911,
+      "learning_rate": 1.9363681618972166e-05,
+      "loss": 0.4328,
+      "step": 214
+    },
+    {
+      "epoch": 0.40874524714828897,
+      "grad_norm": 0.7889241019066625,
+      "learning_rate": 1.9351972922339835e-05,
+      "loss": 0.4367,
+      "step": 215
+    },
+    {
+      "epoch": 0.41064638783269963,
+      "grad_norm": 0.7483946443939699,
+      "learning_rate": 1.9340161087325483e-05,
+      "loss": 0.4397,
+      "step": 216
+    },
+    {
+      "epoch": 0.41254752851711024,
+      "grad_norm": 0.7660466250314436,
+      "learning_rate": 1.9328246244196117e-05,
+      "loss": 0.4303,
+      "step": 217
+    },
+    {
+      "epoch": 0.4144486692015209,
+      "grad_norm": 0.776619474237903,
+      "learning_rate": 1.931622852435478e-05,
+      "loss": 0.4416,
+      "step": 218
+    },
+    {
+      "epoch": 0.41634980988593157,
+      "grad_norm": 0.7616351223608897,
+      "learning_rate": 1.930410806033908e-05,
+      "loss": 0.4345,
+      "step": 219
+    },
+    {
+      "epoch": 0.41825095057034223,
+      "grad_norm": 0.9384742141181847,
+      "learning_rate": 1.929188498581975e-05,
+      "loss": 0.432,
+      "step": 220
+    },
+    {
+      "epoch": 0.42015209125475284,
+      "grad_norm": 0.8855190946497012,
+      "learning_rate": 1.9279559435599164e-05,
+      "loss": 0.447,
+      "step": 221
+    },
+    {
+      "epoch": 0.4220532319391635,
+      "grad_norm": 0.9801225564460324,
+      "learning_rate": 1.926713154560984e-05,
+      "loss": 0.453,
+      "step": 222
+    },
+    {
+      "epoch": 0.42395437262357416,
+      "grad_norm": 0.8039052999022875,
+      "learning_rate": 1.9254601452912972e-05,
+      "loss": 0.4596,
+      "step": 223
+    },
+    {
+      "epoch": 0.42585551330798477,
+      "grad_norm": 1.026094464979351,
+      "learning_rate": 1.924196929569688e-05,
+      "loss": 0.4348,
+      "step": 224
+    },
+    {
+      "epoch": 0.42775665399239543,
+      "grad_norm": 0.7601287164420519,
+      "learning_rate": 1.922923521327551e-05,
+      "loss": 0.4407,
+      "step": 225
+    },
+    {
+      "epoch": 0.4296577946768061,
+      "grad_norm": 0.9912810781784912,
+      "learning_rate": 1.9216399346086893e-05,
+      "loss": 0.4362,
+      "step": 226
+    },
+    {
+      "epoch": 0.43155893536121676,
+      "grad_norm": 0.7607922604900124,
+      "learning_rate": 1.9203461835691596e-05,
+      "loss": 0.44,
+      "step": 227
+    },
+    {
+      "epoch": 0.43346007604562736,
+      "grad_norm": 0.8824871863136691,
+      "learning_rate": 1.9190422824771158e-05,
+      "loss": 0.435,
+      "step": 228
+    },
+    {
+      "epoch": 0.435361216730038,
+      "grad_norm": 0.7919343571024645,
+      "learning_rate": 1.9177282457126515e-05,
+      "loss": 0.4158,
+      "step": 229
+    },
+    {
+      "epoch": 0.4372623574144487,
+      "grad_norm": 0.768369683318518,
+      "learning_rate": 1.9164040877676425e-05,
+      "loss": 0.4294,
+      "step": 230
+    },
+    {
+      "epoch": 0.4391634980988593,
+      "grad_norm": 1.016935123578636,
+      "learning_rate": 1.9150698232455853e-05,
+      "loss": 0.3963,
+      "step": 231
+    },
+    {
+      "epoch": 0.44106463878326996,
+      "grad_norm": 0.7582803051677965,
+      "learning_rate": 1.913725466861438e-05,
+      "loss": 0.4278,
+      "step": 232
+    },
+    {
+      "epoch": 0.4429657794676806,
+      "grad_norm": 0.8946850217393366,
+      "learning_rate": 1.9123710334414552e-05,
+      "loss": 0.4585,
+      "step": 233
+    },
+    {
+      "epoch": 0.4448669201520912,
+      "grad_norm": 0.7805131342935635,
+      "learning_rate": 1.911006537923029e-05,
+      "loss": 0.4274,
+      "step": 234
+    },
+    {
+      "epoch": 0.4467680608365019,
+      "grad_norm": 0.7342344209154682,
+      "learning_rate": 1.9096319953545186e-05,
+      "loss": 0.4356,
+      "step": 235
+    },
+    {
+      "epoch": 0.44866920152091255,
+      "grad_norm": 0.8179223681628749,
+      "learning_rate": 1.908247420895089e-05,
+      "loss": 0.4306,
+      "step": 236
+    },
+    {
+      "epoch": 0.4505703422053232,
+      "grad_norm": 0.9193102485800642,
+      "learning_rate": 1.9068528298145418e-05,
+      "loss": 0.424,
+      "step": 237
+    },
+    {
+      "epoch": 0.4524714828897338,
+      "grad_norm": 0.7656374155303313,
+      "learning_rate": 1.905448237493147e-05,
+      "loss": 0.4416,
+      "step": 238
+    },
+    {
+      "epoch": 0.4543726235741445,
+      "grad_norm": 0.9885795674447727,
+      "learning_rate": 1.9040336594214727e-05,
+      "loss": 0.4073,
+      "step": 239
+    },
+    {
+      "epoch": 0.45627376425855515,
+      "grad_norm": 0.7984749132802632,
+      "learning_rate": 1.9026091112002163e-05,
+      "loss": 0.4249,
+      "step": 240
+    },
+    {
+      "epoch": 0.45817490494296575,
+      "grad_norm": 0.9282796120941919,
+      "learning_rate": 1.90117460854003e-05,
+      "loss": 0.4299,
+      "step": 241
+    },
+    {
+      "epoch": 0.4600760456273764,
+      "grad_norm": 0.9516443166975507,
+      "learning_rate": 1.8997301672613496e-05,
+      "loss": 0.4605,
+      "step": 242
+    },
+    {
+      "epoch": 0.4619771863117871,
+      "grad_norm": 0.8851837571478424,
+      "learning_rate": 1.8982758032942184e-05,
+      "loss": 0.4431,
+      "step": 243
+    },
+    {
+      "epoch": 0.46387832699619774,
+      "grad_norm": 0.8934412546734641,
+      "learning_rate": 1.896811532678113e-05,
+      "loss": 0.4486,
+      "step": 244
+    },
+    {
+      "epoch": 0.46577946768060835,
+      "grad_norm": 0.7280065680634561,
+      "learning_rate": 1.8953373715617646e-05,
+      "loss": 0.4369,
+      "step": 245
+    },
+    {
+      "epoch": 0.467680608365019,
+      "grad_norm": 0.8010993716940912,
+      "learning_rate": 1.893853336202983e-05,
+      "loss": 0.4235,
+      "step": 246
+    },
+    {
+      "epoch": 0.4695817490494297,
+      "grad_norm": 0.8899457650418342,
+      "learning_rate": 1.892359442968475e-05,
+      "loss": 0.405,
+      "step": 247
+    },
+    {
+      "epoch": 0.4714828897338403,
+      "grad_norm": 0.7812008992593449,
+      "learning_rate": 1.8908557083336668e-05,
+      "loss": 0.4266,
+      "step": 248
+    },
+    {
+      "epoch": 0.47338403041825095,
+      "grad_norm": 0.8451147767780671,
+      "learning_rate": 1.889342148882519e-05,
+      "loss": 0.4221,
+      "step": 249
+    },
+    {
+      "epoch": 0.4752851711026616,
+      "grad_norm": 0.7808305640950098,
+      "learning_rate": 1.8878187813073465e-05,
+      "loss": 0.437,
+      "step": 250
+    },
+    {
+      "epoch": 0.47718631178707227,
+      "grad_norm": 0.7494563001346748,
+      "learning_rate": 1.886285622408633e-05,
+      "loss": 0.4195,
+      "step": 251
+    },
+    {
+      "epoch": 0.4790874524714829,
+      "grad_norm": 0.8477515641050383,
+      "learning_rate": 1.8847426890948447e-05,
+      "loss": 0.4271,
+      "step": 252
+    },
+    {
+      "epoch": 0.48098859315589354,
+      "grad_norm": 0.754403803693056,
+      "learning_rate": 1.8831899983822475e-05,
+      "loss": 0.4437,
+      "step": 253
+    },
+    {
+      "epoch": 0.4828897338403042,
+      "grad_norm": 0.8086680469988717,
+      "learning_rate": 1.8816275673947148e-05,
+      "loss": 0.4107,
+      "step": 254
+    },
+    {
+      "epoch": 0.4847908745247148,
+      "grad_norm": 0.7925479551334843,
+      "learning_rate": 1.8800554133635417e-05,
+      "loss": 0.4152,
+      "step": 255
+    },
+    {
+      "epoch": 0.4866920152091255,
+      "grad_norm": 0.7626048336664352,
+      "learning_rate": 1.8784735536272543e-05,
+      "loss": 0.433,
+      "step": 256
+    },
+    {
+      "epoch": 0.48859315589353614,
+      "grad_norm": 0.8603560876138445,
+      "learning_rate": 1.8768820056314173e-05,
+      "loss": 0.4142,
+      "step": 257
+    },
+    {
+      "epoch": 0.49049429657794674,
+      "grad_norm": 0.736530696765543,
+      "learning_rate": 1.875280786928444e-05,
+      "loss": 0.4366,
+      "step": 258
+    },
+    {
+      "epoch": 0.4923954372623574,
+      "grad_norm": 0.8741756327011199,
+      "learning_rate": 1.873669915177399e-05,
+      "loss": 0.4361,
+      "step": 259
+    },
+    {
+      "epoch": 0.49429657794676807,
+      "grad_norm": 0.8008511094673902,
+      "learning_rate": 1.872049408143808e-05,
+      "loss": 0.4142,
+      "step": 260
+    },
+    {
+      "epoch": 0.49619771863117873,
+      "grad_norm": 0.7782774124949869,
+      "learning_rate": 1.8704192836994578e-05,
+      "loss": 0.4267,
+      "step": 261
+    },
+    {
+      "epoch": 0.49809885931558934,
+      "grad_norm": 0.835772295394541,
+      "learning_rate": 1.8687795598222024e-05,
+      "loss": 0.4484,
+      "step": 262
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7161910677629045,
+      "learning_rate": 1.8671302545957628e-05,
+      "loss": 0.433,
+      "step": 263
+    },
+    {
+      "epoch": 0.5019011406844106,
+      "grad_norm": 0.7129409060749992,
+      "learning_rate": 1.8654713862095272e-05,
+      "loss": 0.4268,
+      "step": 264
+    },
+    {
+      "epoch": 0.5038022813688213,
+      "grad_norm": 0.8746936224004868,
+      "learning_rate": 1.8638029729583524e-05,
+      "loss": 0.426,
+      "step": 265
+    },
+    {
+      "epoch": 0.5057034220532319,
+      "grad_norm": 0.7199661881084182,
+      "learning_rate": 1.8621250332423603e-05,
+      "loss": 0.4206,
+      "step": 266
+    },
+    {
+      "epoch": 0.5076045627376425,
+      "grad_norm": 0.8776639159854762,
+      "learning_rate": 1.860437585566736e-05,
+      "loss": 0.4209,
+      "step": 267
+    },
+    {
+      "epoch": 0.5095057034220533,
+      "grad_norm": 0.7038553668788641,
+      "learning_rate": 1.8587406485415226e-05,
+      "loss": 0.4275,
+      "step": 268
+    },
+    {
+      "epoch": 0.5114068441064639,
+      "grad_norm": 0.8246952846777812,
+      "learning_rate": 1.8570342408814173e-05,
+      "loss": 0.4359,
+      "step": 269
+    },
+    {
+      "epoch": 0.5133079847908745,
+      "grad_norm": 0.8420390392108025,
+      "learning_rate": 1.855318381405564e-05,
+      "loss": 0.4478,
+      "step": 270
+    },
+    {
+      "epoch": 0.5152091254752852,
+      "grad_norm": 0.8281462911560388,
+      "learning_rate": 1.8535930890373467e-05,
+      "loss": 0.4264,
+      "step": 271
+    },
+    {
+      "epoch": 0.5171102661596958,
+      "grad_norm": 0.7843657243577942,
+      "learning_rate": 1.8518583828041787e-05,
+      "loss": 0.4364,
+      "step": 272
+    },
+    {
+      "epoch": 0.5190114068441065,
+      "grad_norm": 0.7765503284681777,
+      "learning_rate": 1.8501142818372964e-05,
+      "loss": 0.4195,
+      "step": 273
+    },
+    {
+      "epoch": 0.5209125475285171,
+      "grad_norm": 0.7744347795836504,
+      "learning_rate": 1.848360805371544e-05,
+      "loss": 0.4371,
+      "step": 274
+    },
+    {
+      "epoch": 0.5228136882129277,
+      "grad_norm": 0.7931467086003418,
+      "learning_rate": 1.8465979727451653e-05,
+      "loss": 0.4192,
+      "step": 275
+    },
+    {
+      "epoch": 0.5247148288973384,
+      "grad_norm": 0.759121399714454,
+      "learning_rate": 1.8448258033995877e-05,
+      "loss": 0.4266,
+      "step": 276
+    },
+    {
+      "epoch": 0.526615969581749,
+      "grad_norm": 0.7701454848778623,
+      "learning_rate": 1.8430443168792087e-05,
+      "loss": 0.4029,
+      "step": 277
+    },
+    {
+      "epoch": 0.5285171102661597,
+      "grad_norm": 0.8947519037395579,
+      "learning_rate": 1.8412535328311813e-05,
+      "loss": 0.4264,
+      "step": 278
+    },
+    {
+      "epoch": 0.5304182509505704,
+      "grad_norm": 0.764683578264793,
+      "learning_rate": 1.8394534710051956e-05,
+      "loss": 0.4293,
+      "step": 279
+    },
+    {
+      "epoch": 0.532319391634981,
+      "grad_norm": 0.9986327813906521,
+      "learning_rate": 1.8376441512532617e-05,
+      "loss": 0.4304,
+      "step": 280
+    },
+    {
+      "epoch": 0.5342205323193916,
+      "grad_norm": 0.7530244817990461,
+      "learning_rate": 1.835825593529492e-05,
+      "loss": 0.4105,
+      "step": 281
+    },
+    {
+      "epoch": 0.5361216730038023,
+      "grad_norm": 0.8782003043844607,
+      "learning_rate": 1.833997817889878e-05,
+      "loss": 0.417,
+      "step": 282
+    },
+    {
+      "epoch": 0.5380228136882129,
+      "grad_norm": 0.8916898317005598,
+      "learning_rate": 1.8321608444920738e-05,
+      "loss": 0.4089,
+      "step": 283
+    },
+    {
+      "epoch": 0.5399239543726235,
+      "grad_norm": 0.7931488574412902,
+      "learning_rate": 1.830314693595169e-05,
+      "loss": 0.4084,
+      "step": 284
+    },
+    {
+      "epoch": 0.5418250950570342,
+      "grad_norm": 0.8743392401753514,
+      "learning_rate": 1.828459385559468e-05,
+      "loss": 0.4163,
+      "step": 285
+    },
+    {
+      "epoch": 0.5437262357414449,
+      "grad_norm": 0.6993379251712359,
+      "learning_rate": 1.8265949408462657e-05,
+      "loss": 0.4215,
+      "step": 286
+    },
+    {
+      "epoch": 0.5456273764258555,
+      "grad_norm": 0.8115000089946518,
+      "learning_rate": 1.8247213800176192e-05,
+      "loss": 0.4206,
+      "step": 287
+    },
+    {
+      "epoch": 0.5475285171102662,
+      "grad_norm": 0.6832345985766087,
+      "learning_rate": 1.8228387237361245e-05,
+      "loss": 0.4184,
+      "step": 288
+    },
+    {
+      "epoch": 0.5494296577946768,
+      "grad_norm": 0.6885893557235931,
+      "learning_rate": 1.8209469927646863e-05,
+      "loss": 0.4129,
+      "step": 289
+    },
+    {
+      "epoch": 0.5513307984790875,
+      "grad_norm": 0.6951032366993848,
+      "learning_rate": 1.8190462079662897e-05,
+      "loss": 0.4276,
+      "step": 290
+    },
+    {
+      "epoch": 0.5532319391634981,
+      "grad_norm": 0.7208573115005421,
+      "learning_rate": 1.81713639030377e-05,
+      "loss": 0.4158,
+      "step": 291
+    },
+    {
+      "epoch": 0.5551330798479087,
+      "grad_norm": 0.7151189411118184,
+      "learning_rate": 1.8152175608395814e-05,
+      "loss": 0.4063,
+      "step": 292
+    },
+    {
+      "epoch": 0.5570342205323194,
+      "grad_norm": 0.7618151350052784,
+      "learning_rate": 1.8132897407355657e-05,
+      "loss": 0.4228,
+      "step": 293
+    },
+    {
+      "epoch": 0.55893536121673,
+      "grad_norm": 0.7280925750642289,
+      "learning_rate": 1.811352951252717e-05,
+      "loss": 0.4118,
+      "step": 294
+    },
+    {
+      "epoch": 0.5608365019011406,
+      "grad_norm": 0.6801705394482089,
+      "learning_rate": 1.809407213750949e-05,
+      "loss": 0.4211,
+      "step": 295
+    },
+    {
+      "epoch": 0.5627376425855514,
+      "grad_norm": 0.7856044506860411,
+      "learning_rate": 1.807452549688859e-05,
+      "loss": 0.4388,
+      "step": 296
+    },
+    {
+      "epoch": 0.564638783269962,
+      "grad_norm": 0.794800402247415,
+      "learning_rate": 1.8054889806234906e-05,
+      "loss": 0.4566,
+      "step": 297
+    },
+    {
+      "epoch": 0.5665399239543726,
+      "grad_norm": 0.8131356260561907,
+      "learning_rate": 1.8035165282100963e-05,
+      "loss": 0.4248,
+      "step": 298
+    },
+    {
+      "epoch": 0.5684410646387833,
+      "grad_norm": 0.7464654000588856,
+      "learning_rate": 1.8015352142018984e-05,
+      "loss": 0.421,
+      "step": 299
+    },
+    {
+      "epoch": 0.5703422053231939,
+      "grad_norm": 0.7131841313678775,
+      "learning_rate": 1.799545060449851e-05,
+      "loss": 0.4211,
+      "step": 300
+    },
+    {
+      "epoch": 0.5722433460076045,
+      "grad_norm": 0.7387170406076679,
+      "learning_rate": 1.797546088902396e-05,
+      "loss": 0.4081,
+      "step": 301
+    },
+    {
+      "epoch": 0.5741444866920152,
+      "grad_norm": 0.792108751509653,
+      "learning_rate": 1.7955383216052224e-05,
+      "loss": 0.4276,
+      "step": 302
+    },
+    {
+      "epoch": 0.5760456273764258,
+      "grad_norm": 0.7300887999445866,
+      "learning_rate": 1.7935217807010238e-05,
+      "loss": 0.4246,
+      "step": 303
+    },
+    {
+      "epoch": 0.5779467680608364,
+      "grad_norm": 0.8271445399660962,
+      "learning_rate": 1.7914964884292543e-05,
+      "loss": 0.4373,
+      "step": 304
+    },
+    {
+      "epoch": 0.5798479087452472,
+      "grad_norm": 0.7178525468830346,
+      "learning_rate": 1.7894624671258813e-05,
+      "loss": 0.4353,
+      "step": 305
+    },
+    {
+      "epoch": 0.5817490494296578,
+      "grad_norm": 0.7525552514804682,
+      "learning_rate": 1.7874197392231414e-05,
+      "loss": 0.4231,
+      "step": 306
+    },
+    {
+      "epoch": 0.5836501901140685,
+      "grad_norm": 0.8362327721180475,
+      "learning_rate": 1.7853683272492913e-05,
+      "loss": 0.4215,
+      "step": 307
+    },
+    {
+      "epoch": 0.5855513307984791,
+      "grad_norm": 0.839930364021564,
+      "learning_rate": 1.7833082538283615e-05,
+      "loss": 0.4125,
+      "step": 308
+    },
+    {
+      "epoch": 0.5874524714828897,
+      "grad_norm": 0.776281344866687,
+      "learning_rate": 1.7812395416799034e-05,
+      "loss": 0.4047,
+      "step": 309
+    },
+    {
+      "epoch": 0.5893536121673004,
+      "grad_norm": 0.7637957154344026,
+      "learning_rate": 1.7791622136187422e-05,
+      "loss": 0.4522,
+      "step": 310
+    },
+    {
+      "epoch": 0.591254752851711,
+      "grad_norm": 0.8049321143956409,
+      "learning_rate": 1.7770762925547235e-05,
+      "loss": 0.422,
+      "step": 311
+    },
+    {
+      "epoch": 0.5931558935361216,
+      "grad_norm": 0.7253015248660504,
+      "learning_rate": 1.7749818014924612e-05,
+      "loss": 0.4286,
+      "step": 312
+    },
+    {
+      "epoch": 0.5950570342205324,
+      "grad_norm": 0.7059888261108643,
+      "learning_rate": 1.7728787635310828e-05,
+      "loss": 0.4047,
+      "step": 313
+    },
+    {
+      "epoch": 0.596958174904943,
+      "grad_norm": 0.7824517701597381,
+      "learning_rate": 1.770767201863976e-05,
+      "loss": 0.4212,
+      "step": 314
+    },
+    {
+      "epoch": 0.5988593155893536,
+      "grad_norm": 0.7268105386751071,
+      "learning_rate": 1.7686471397785322e-05,
+      "loss": 0.414,
+      "step": 315
+    },
+    {
+      "epoch": 0.6007604562737643,
+      "grad_norm": 0.7465031606207332,
+      "learning_rate": 1.76651860065589e-05,
+      "loss": 0.4212,
+      "step": 316
+    },
+    {
+      "epoch": 0.6026615969581749,
+      "grad_norm": 0.7793014868165488,
+      "learning_rate": 1.764381607970677e-05,
+      "loss": 0.4199,
+      "step": 317
+    },
+    {
+      "epoch": 0.6045627376425855,
+      "grad_norm": 0.776969130282458,
+      "learning_rate": 1.7622361852907506e-05,
+      "loss": 0.414,
+      "step": 318
+    },
+    {
+      "epoch": 0.6064638783269962,
+      "grad_norm": 0.8146335121895256,
+      "learning_rate": 1.760082356276939e-05,
+      "loss": 0.4403,
+      "step": 319
+    },
+    {
+      "epoch": 0.6083650190114068,
+      "grad_norm": 0.7441862022739212,
+      "learning_rate": 1.75792014468278e-05,
+      "loss": 0.419,
+      "step": 320
+    },
+    {
+      "epoch": 0.6102661596958175,
+      "grad_norm": 0.7901745117506392,
+      "learning_rate": 1.7557495743542586e-05,
+      "loss": 0.4398,
+      "step": 321
+    },
+    {
+      "epoch": 0.6121673003802282,
+      "grad_norm": 0.739303646326241,
+      "learning_rate": 1.7535706692295436e-05,
+      "loss": 0.419,
+      "step": 322
+    },
+    {
+      "epoch": 0.6140684410646388,
+      "grad_norm": 0.7224930356712501,
+      "learning_rate": 1.7513834533387256e-05,
+      "loss": 0.4136,
+      "step": 323
+    },
+    {
+      "epoch": 0.6159695817490495,
+      "grad_norm": 0.7570371896775817,
+      "learning_rate": 1.7491879508035488e-05,
+      "loss": 0.4197,
+      "step": 324
+    },
+    {
+      "epoch": 0.6178707224334601,
+      "grad_norm": 0.6797463913573002,
+      "learning_rate": 1.746984185837149e-05,
+      "loss": 0.4065,
+      "step": 325
+    },
+    {
+      "epoch": 0.6197718631178707,
+      "grad_norm": 0.8680955704544816,
+      "learning_rate": 1.744772182743782e-05,
+      "loss": 0.4197,
+      "step": 326
+    },
+    {
+      "epoch": 0.6216730038022814,
+      "grad_norm": 0.6916049420985639,
+      "learning_rate": 1.7425519659185596e-05,
+      "loss": 0.4101,
+      "step": 327
+    },
+    {
+      "epoch": 0.623574144486692,
+      "grad_norm": 0.8117898782267594,
+      "learning_rate": 1.740323559847179e-05,
+      "loss": 0.4132,
+      "step": 328
+    },
+    {
+      "epoch": 0.6254752851711026,
+      "grad_norm": 0.6860477601327285,
+      "learning_rate": 1.738086989105651e-05,
+      "loss": 0.4079,
+      "step": 329
+    },
+    {
+      "epoch": 0.6273764258555133,
+      "grad_norm": 0.7354356036807334,
+      "learning_rate": 1.735842278360032e-05,
+      "loss": 0.435,
+      "step": 330
+    },
+    {
+      "epoch": 0.629277566539924,
+      "grad_norm": 0.7191376650164161,
+      "learning_rate": 1.73358945236615e-05,
+      "loss": 0.4161,
+      "step": 331
+    },
+    {
+      "epoch": 0.6311787072243346,
+      "grad_norm": 0.6842360700055289,
+      "learning_rate": 1.7313285359693322e-05,
+      "loss": 0.4302,
+      "step": 332
+    },
+    {
+      "epoch": 0.6330798479087453,
+      "grad_norm": 0.6472886708852534,
+      "learning_rate": 1.7290595541041312e-05,
+      "loss": 0.4031,
+      "step": 333
+    },
+    {
+      "epoch": 0.6349809885931559,
+      "grad_norm": 0.703931673583881,
+      "learning_rate": 1.7267825317940494e-05,
+      "loss": 0.4346,
+      "step": 334
+    },
+    {
+      "epoch": 0.6368821292775665,
+      "grad_norm": 0.7678042705630014,
+      "learning_rate": 1.724497494151264e-05,
+      "loss": 0.4104,
+      "step": 335
+    },
+    {
+      "epoch": 0.6387832699619772,
+      "grad_norm": 0.7413059337967789,
+      "learning_rate": 1.7222044663763484e-05,
+      "loss": 0.4203,
+      "step": 336
+    },
+    {
+      "epoch": 0.6406844106463878,
+      "grad_norm": 0.7367710114399577,
+      "learning_rate": 1.7199034737579962e-05,
+      "loss": 0.4083,
+      "step": 337
+    },
+    {
+      "epoch": 0.6425855513307985,
+      "grad_norm": 0.6881715276339009,
+      "learning_rate": 1.7175945416727405e-05,
+      "loss": 0.4141,
+      "step": 338
+    },
+    {
+      "epoch": 0.6444866920152091,
+      "grad_norm": 0.7662058535912438,
+      "learning_rate": 1.7152776955846768e-05,
+      "loss": 0.4131,
+      "step": 339
+    },
+    {
+      "epoch": 0.6463878326996197,
+      "grad_norm": 0.7061712057750617,
+      "learning_rate": 1.7129529610451775e-05,
+      "loss": 0.412,
+      "step": 340
+    },
+    {
+      "epoch": 0.6482889733840305,
+      "grad_norm": 0.7289089945193655,
+      "learning_rate": 1.7106203636926154e-05,
+      "loss": 0.4251,
+      "step": 341
+    },
+    {
+      "epoch": 0.6501901140684411,
+      "grad_norm": 0.7324639550070794,
+      "learning_rate": 1.7082799292520767e-05,
+      "loss": 0.4155,
+      "step": 342
+    },
+    {
+      "epoch": 0.6520912547528517,
+      "grad_norm": 0.791922279082062,
+      "learning_rate": 1.7059316835350806e-05,
+      "loss": 0.4273,
+      "step": 343
+    },
+    {
+      "epoch": 0.6539923954372624,
+      "grad_norm": 0.7705902947783712,
+      "learning_rate": 1.7035756524392924e-05,
+      "loss": 0.4219,
+      "step": 344
+    },
+    {
+      "epoch": 0.655893536121673,
+      "grad_norm": 0.8763456771850822,
+      "learning_rate": 1.7012118619482376e-05,
+      "loss": 0.4233,
+      "step": 345
+    },
+    {
+      "epoch": 0.6577946768060836,
+      "grad_norm": 0.732442433980309,
+      "learning_rate": 1.6988403381310177e-05,
+      "loss": 0.4351,
+      "step": 346
+    },
+    {
+      "epoch": 0.6596958174904943,
+      "grad_norm": 0.9826668310848153,
+      "learning_rate": 1.696461107142021e-05,
+      "loss": 0.4244,
+      "step": 347
+    },
+    {
+      "epoch": 0.6615969581749049,
+      "grad_norm": 0.6852179914988137,
+      "learning_rate": 1.6940741952206342e-05,
+      "loss": 0.4023,
+      "step": 348
+    },
+    {
+      "epoch": 0.6634980988593155,
+      "grad_norm": 0.7440795370888574,
+      "learning_rate": 1.691679628690953e-05,
+      "loss": 0.4253,
+      "step": 349
+    },
+    {
+      "epoch": 0.6653992395437263,
+      "grad_norm": 0.7193784150797833,
+      "learning_rate": 1.6892774339614927e-05,
+      "loss": 0.4033,
+      "step": 350
+    },
+    {
+      "epoch": 0.6673003802281369,
+      "grad_norm": 0.8156600376301576,
+      "learning_rate": 1.686867637524896e-05,
+      "loss": 0.4495,
+      "step": 351
+    },
+    {
+      "epoch": 0.6692015209125475,
+      "grad_norm": 0.7176786429913768,
+      "learning_rate": 1.6844502659576414e-05,
+      "loss": 0.4208,
+      "step": 352
+    },
+    {
+      "epoch": 0.6711026615969582,
+      "grad_norm": 0.7659504090904091,
+      "learning_rate": 1.6820253459197493e-05,
+      "loss": 0.4388,
+      "step": 353
+    },
+    {
+      "epoch": 0.6730038022813688,
+      "grad_norm": 0.7978319359054401,
+      "learning_rate": 1.679592904154489e-05,
+      "loss": 0.4365,
+      "step": 354
+    },
+    {
+      "epoch": 0.6749049429657795,
+      "grad_norm": 0.644489027594137,
+      "learning_rate": 1.677152967488084e-05,
+      "loss": 0.4148,
+      "step": 355
+    },
+    {
+      "epoch": 0.6768060836501901,
+      "grad_norm": 0.7495183754977875,
+      "learning_rate": 1.6747055628294134e-05,
+      "loss": 0.4212,
+      "step": 356
+    },
+    {
+      "epoch": 0.6787072243346007,
+      "grad_norm": 0.6912626966302459,
+      "learning_rate": 1.6722507171697184e-05,
+      "loss": 0.4184,
+      "step": 357
+    },
+    {
+      "epoch": 0.6806083650190115,
+      "grad_norm": 0.6655628781553221,
+      "learning_rate": 1.669788457582304e-05,
+      "loss": 0.4249,
+      "step": 358
+    },
+    {
+      "epoch": 0.6825095057034221,
+      "grad_norm": 0.791689238520468,
+      "learning_rate": 1.6673188112222394e-05,
+      "loss": 0.4084,
+      "step": 359
+    },
+    {
+      "epoch": 0.6844106463878327,
+      "grad_norm": 0.7498636055362748,
+      "learning_rate": 1.6648418053260585e-05,
+      "loss": 0.4202,
+      "step": 360
+    },
+    {
+      "epoch": 0.6863117870722434,
+      "grad_norm": 0.6809544908914978,
+      "learning_rate": 1.6623574672114596e-05,
+      "loss": 0.4311,
+      "step": 361
+    },
+    {
+      "epoch": 0.688212927756654,
+      "grad_norm": 0.707924479551307,
+      "learning_rate": 1.6598658242770054e-05,
+      "loss": 0.4226,
+      "step": 362
+    },
+    {
+      "epoch": 0.6901140684410646,
+      "grad_norm": 0.7875192296853099,
+      "learning_rate": 1.6573669040018202e-05,
+      "loss": 0.4035,
+      "step": 363
+    },
+    {
+      "epoch": 0.6920152091254753,
+      "grad_norm": 0.6989927093509163,
+      "learning_rate": 1.6548607339452853e-05,
+      "loss": 0.4015,
+      "step": 364
+    },
+    {
+      "epoch": 0.6939163498098859,
+      "grad_norm": 0.7613004712226819,
+      "learning_rate": 1.652347341746737e-05,
+      "loss": 0.4196,
+      "step": 365
+    },
+    {
+      "epoch": 0.6958174904942965,
+      "grad_norm": 0.7801064614311649,
+      "learning_rate": 1.6498267551251618e-05,
+      "loss": 0.415,
+      "step": 366
+    },
+    {
+      "epoch": 0.6977186311787072,
+      "grad_norm": 0.7143915414824872,
+      "learning_rate": 1.6472990018788884e-05,
+      "loss": 0.4007,
+      "step": 367
+    },
+    {
+      "epoch": 0.6996197718631179,
+      "grad_norm": 0.7213556382982281,
+      "learning_rate": 1.644764109885284e-05,
+      "loss": 0.435,
+      "step": 368
+    },
+    {
+      "epoch": 0.7015209125475285,
+      "grad_norm": 0.7937361641182753,
+      "learning_rate": 1.642222107100446e-05,
+      "loss": 0.415,
+      "step": 369
+    },
+    {
+      "epoch": 0.7034220532319392,
+      "grad_norm": 0.7077243928080609,
+      "learning_rate": 1.6396730215588913e-05,
+      "loss": 0.4128,
+      "step": 370
+    },
+    {
+      "epoch": 0.7053231939163498,
+      "grad_norm": 0.7891653499369282,
+      "learning_rate": 1.6371168813732514e-05,
+      "loss": 0.4179,
+      "step": 371
+    },
+    {
+      "epoch": 0.7072243346007605,
+      "grad_norm": 0.764693769594645,
+      "learning_rate": 1.6345537147339578e-05,
+      "loss": 0.4165,
+      "step": 372
+    },
+    {
+      "epoch": 0.7091254752851711,
+      "grad_norm": 0.8271146906155411,
+      "learning_rate": 1.6319835499089358e-05,
+      "loss": 0.4042,
+      "step": 373
+    },
+    {
+      "epoch": 0.7110266159695817,
+      "grad_norm": 0.7509192905087371,
+      "learning_rate": 1.6294064152432878e-05,
+      "loss": 0.4061,
+      "step": 374
+    },
+    {
+      "epoch": 0.7129277566539924,
+      "grad_norm": 0.8345464270506578,
+      "learning_rate": 1.626822339158985e-05,
+      "loss": 0.4025,
+      "step": 375
+    },
+    {
+      "epoch": 0.714828897338403,
+      "grad_norm": 0.7072083557074277,
+      "learning_rate": 1.6242313501545522e-05,
+      "loss": 0.4173,
+      "step": 376
+    },
+    {
+      "epoch": 0.7167300380228137,
+      "grad_norm": 0.7293814090642944,
+      "learning_rate": 1.621633476804752e-05,
+      "loss": 0.4166,
+      "step": 377
+    },
+    {
+      "epoch": 0.7186311787072244,
+      "grad_norm": 0.846495558869141,
+      "learning_rate": 1.6190287477602716e-05,
+      "loss": 0.416,
+      "step": 378
+    },
+    {
+      "epoch": 0.720532319391635,
+      "grad_norm": 0.7085045302214626,
+      "learning_rate": 1.6164171917474078e-05,
+      "loss": 0.4009,
+      "step": 379
+    },
+    {
+      "epoch": 0.7224334600760456,
+      "grad_norm": 0.8925969184626525,
+      "learning_rate": 1.6137988375677466e-05,
+      "loss": 0.4268,
+      "step": 380
+    },
+    {
+      "epoch": 0.7243346007604563,
+      "grad_norm": 0.7253513104874315,
+      "learning_rate": 1.6111737140978495e-05,
+      "loss": 0.4089,
+      "step": 381
+    },
+    {
+      "epoch": 0.7262357414448669,
+      "grad_norm": 0.8538365232177512,
+      "learning_rate": 1.6085418502889315e-05,
+      "loss": 0.4171,
+      "step": 382
+    },
+    {
+      "epoch": 0.7281368821292775,
+      "grad_norm": 0.884407137887338,
+      "learning_rate": 1.6059032751665454e-05,
+      "loss": 0.4368,
+      "step": 383
+    },
+    {
+      "epoch": 0.7300380228136882,
+      "grad_norm": 0.7597979823976849,
+      "learning_rate": 1.6032580178302585e-05,
+      "loss": 0.4313,
+      "step": 384
+    },
+    {
+      "epoch": 0.7319391634980988,
+      "grad_norm": 0.8665276836369039,
+      "learning_rate": 1.600606107453333e-05,
+      "loss": 0.411,
+      "step": 385
+    },
+    {
+      "epoch": 0.7338403041825095,
+      "grad_norm": 0.706717491048344,
+      "learning_rate": 1.597947573282405e-05,
+      "loss": 0.4067,
+      "step": 386
+    },
+    {
+      "epoch": 0.7357414448669202,
+      "grad_norm": 0.6879678689920503,
+      "learning_rate": 1.5952824446371608e-05,
+      "loss": 0.4252,
+      "step": 387
+    },
+    {
+      "epoch": 0.7376425855513308,
+      "grad_norm": 0.7214214636724359,
+      "learning_rate": 1.592610750910014e-05,
+      "loss": 0.4136,
+      "step": 388
+    },
+    {
+      "epoch": 0.7395437262357415,
+      "grad_norm": 0.721849313528591,
+      "learning_rate": 1.589932521565781e-05,
+      "loss": 0.4,
+      "step": 389
+    },
+    {
+      "epoch": 0.7414448669201521,
+      "grad_norm": 0.6864079089694056,
+      "learning_rate": 1.587247786141358e-05,
+      "loss": 0.4107,
+      "step": 390
+    },
+    {
+      "epoch": 0.7433460076045627,
+      "grad_norm": 0.7168944825196677,
+      "learning_rate": 1.5845565742453906e-05,
+      "loss": 0.4338,
+      "step": 391
+    },
+    {
+      "epoch": 0.7452471482889734,
+      "grad_norm": 0.7182661630266403,
+      "learning_rate": 1.581858915557953e-05,
+      "loss": 0.4103,
+      "step": 392
+    },
+    {
+      "epoch": 0.747148288973384,
+      "grad_norm": 0.7213064699225278,
+      "learning_rate": 1.5791548398302167e-05,
+      "loss": 0.4349,
+      "step": 393
+    },
+    {
+      "epoch": 0.7490494296577946,
+      "grad_norm": 0.6656926915900984,
+      "learning_rate": 1.5764443768841234e-05,
+      "loss": 0.3977,
+      "step": 394
+    },
+    {
+      "epoch": 0.7509505703422054,
+      "grad_norm": 0.6587251971187607,
+      "learning_rate": 1.5737275566120577e-05,
+      "loss": 0.4188,
+      "step": 395
+    },
+    {
+      "epoch": 0.752851711026616,
+      "grad_norm": 0.6587010876487651,
+      "learning_rate": 1.5710044089765144e-05,
+      "loss": 0.4182,
+      "step": 396
+    },
+    {
+      "epoch": 0.7547528517110266,
+      "grad_norm": 0.6971765373848173,
+      "learning_rate": 1.5682749640097708e-05,
+      "loss": 0.4139,
+      "step": 397
+    },
+    {
+      "epoch": 0.7566539923954373,
+      "grad_norm": 0.6814783148658188,
+      "learning_rate": 1.565539251813554e-05,
+      "loss": 0.3984,
+      "step": 398
+    },
+    {
+      "epoch": 0.7585551330798479,
+      "grad_norm": 0.6987036480754459,
+      "learning_rate": 1.5627973025587093e-05,
+      "loss": 0.4254,
+      "step": 399
+    },
+    {
+      "epoch": 0.7604562737642585,
+      "grad_norm": 0.6913577141743426,
+      "learning_rate": 1.560049146484868e-05,
+      "loss": 0.4134,
+      "step": 400
+    },
+    {
+      "epoch": 0.7623574144486692,
+      "grad_norm": 0.7249966354524016,
+      "learning_rate": 1.5572948139001128e-05,
+      "loss": 0.4174,
+      "step": 401
+    },
+    {
+      "epoch": 0.7642585551330798,
+      "grad_norm": 0.6864586824036938,
+      "learning_rate": 1.5545343351806443e-05,
+      "loss": 0.3987,
+      "step": 402
+    },
+    {
+      "epoch": 0.7661596958174905,
+      "grad_norm": 0.6909163969072585,
+      "learning_rate": 1.551767740770446e-05,
+      "loss": 0.4028,
+      "step": 403
+    },
+    {
+      "epoch": 0.7680608365019012,
+      "grad_norm": 0.7665854158952898,
+      "learning_rate": 1.5489950611809484e-05,
+      "loss": 0.4036,
+      "step": 404
+    },
+    {
+      "epoch": 0.7699619771863118,
+      "grad_norm": 0.6746980733298985,
+      "learning_rate": 1.5462163269906928e-05,
+      "loss": 0.4171,
+      "step": 405
+    },
+    {
+      "epoch": 0.7718631178707225,
+      "grad_norm": 0.6888109049161354,
+      "learning_rate": 1.5434315688449924e-05,
+      "loss": 0.4192,
+      "step": 406
+    },
+    {
+      "epoch": 0.7737642585551331,
+      "grad_norm": 0.680295967985405,
+      "learning_rate": 1.5406408174555978e-05,
+      "loss": 0.388,
+      "step": 407
+    },
+    {
+      "epoch": 0.7756653992395437,
+      "grad_norm": 0.7573720244568496,
+      "learning_rate": 1.5378441036003543e-05,
+      "loss": 0.4571,
+      "step": 408
+    },
+    {
+      "epoch": 0.7775665399239544,
+      "grad_norm": 0.7263352251395035,
+      "learning_rate": 1.535041458122865e-05,
+      "loss": 0.4103,
+      "step": 409
+    },
+    {
+      "epoch": 0.779467680608365,
+      "grad_norm": 0.6611458759741928,
+      "learning_rate": 1.5322329119321508e-05,
+      "loss": 0.396,
+      "step": 410
+    },
+    {
+      "epoch": 0.7813688212927756,
+      "grad_norm": 0.6702451623154632,
+      "learning_rate": 1.529418496002308e-05,
+      "loss": 0.4085,
+      "step": 411
+    },
+    {
+      "epoch": 0.7832699619771863,
+      "grad_norm": 0.6708900730058422,
+      "learning_rate": 1.5265982413721662e-05,
+      "loss": 0.3866,
+      "step": 412
+    },
+    {
+      "epoch": 0.785171102661597,
+      "grad_norm": 0.7202699620582907,
+      "learning_rate": 1.5237721791449497e-05,
+      "loss": 0.4223,
+      "step": 413
+    },
+    {
+      "epoch": 0.7870722433460076,
+      "grad_norm": 0.7183698093396678,
+      "learning_rate": 1.5209403404879305e-05,
+      "loss": 0.4028,
+      "step": 414
+    },
+    {
+      "epoch": 0.7889733840304183,
+      "grad_norm": 0.6544471597225477,
+      "learning_rate": 1.5181027566320858e-05,
+      "loss": 0.4041,
+      "step": 415
+    },
+    {
+      "epoch": 0.7908745247148289,
+      "grad_norm": 0.7083722634196089,
+      "learning_rate": 1.5152594588717544e-05,
+      "loss": 0.42,
+      "step": 416
+    },
+    {
+      "epoch": 0.7927756653992395,
+      "grad_norm": 0.712188607476995,
+      "learning_rate": 1.5124104785642909e-05,
+      "loss": 0.4121,
+      "step": 417
+    },
+    {
+      "epoch": 0.7946768060836502,
+      "grad_norm": 0.706428015235297,
+      "learning_rate": 1.5095558471297196e-05,
+      "loss": 0.407,
+      "step": 418
+    },
+    {
+      "epoch": 0.7965779467680608,
+      "grad_norm": 0.7290748424935284,
+      "learning_rate": 1.5066955960503893e-05,
+      "loss": 0.3899,
+      "step": 419
+    },
+    {
+      "epoch": 0.7984790874524715,
+      "grad_norm": 0.6700744527803965,
+      "learning_rate": 1.5038297568706244e-05,
+      "loss": 0.3986,
+      "step": 420
+    },
+    {
+      "epoch": 0.8003802281368821,
+      "grad_norm": 0.7369358259318646,
+      "learning_rate": 1.5009583611963772e-05,
+      "loss": 0.409,
+      "step": 421
+    },
+    {
+      "epoch": 0.8022813688212928,
+      "grad_norm": 0.6219125276106352,
+      "learning_rate": 1.4980814406948806e-05,
+      "loss": 0.4077,
+      "step": 422
+    },
+    {
+      "epoch": 0.8041825095057035,
+      "grad_norm": 0.7010378671600722,
+      "learning_rate": 1.4951990270942991e-05,
+      "loss": 0.4077,
+      "step": 423
+    },
+    {
+      "epoch": 0.8060836501901141,
+      "grad_norm": 0.6726478792952649,
+      "learning_rate": 1.492311152183376e-05,
+      "loss": 0.3896,
+      "step": 424
+    },
+    {
+      "epoch": 0.8079847908745247,
+      "grad_norm": 0.7064600715429418,
+      "learning_rate": 1.4894178478110856e-05,
+      "loss": 0.4135,
+      "step": 425
+    },
+    {
+      "epoch": 0.8098859315589354,
+      "grad_norm": 0.6333196467789542,
+      "learning_rate": 1.4865191458862816e-05,
+      "loss": 0.3838,
+      "step": 426
+    },
+    {
+      "epoch": 0.811787072243346,
+      "grad_norm": 0.7082130527531607,
+      "learning_rate": 1.4836150783773442e-05,
+      "loss": 0.4014,
+      "step": 427
+    },
+    {
+      "epoch": 0.8136882129277566,
+      "grad_norm": 0.6875888043485031,
+      "learning_rate": 1.4807056773118276e-05,
+      "loss": 0.4244,
+      "step": 428
+    },
+    {
+      "epoch": 0.8155893536121673,
+      "grad_norm": 0.6320161894647373,
+      "learning_rate": 1.4777909747761085e-05,
+      "loss": 0.3772,
+      "step": 429
+    },
+    {
+      "epoch": 0.8174904942965779,
+      "grad_norm": 0.688853241624807,
+      "learning_rate": 1.4748710029150296e-05,
+      "loss": 0.4249,
+      "step": 430
+    },
+    {
+      "epoch": 0.8193916349809885,
+      "grad_norm": 0.6952309844328408,
+      "learning_rate": 1.4719457939315468e-05,
+      "loss": 0.4095,
+      "step": 431
+    },
+    {
+      "epoch": 0.8212927756653993,
+      "grad_norm": 0.6926853800220879,
+      "learning_rate": 1.4690153800863743e-05,
+      "loss": 0.3959,
+      "step": 432
+    },
+    {
+      "epoch": 0.8231939163498099,
+      "grad_norm": 0.6882725839401531,
+      "learning_rate": 1.4660797936976278e-05,
+      "loss": 0.3914,
+      "step": 433
+    },
+    {
+      "epoch": 0.8250950570342205,
+      "grad_norm": 0.6937884472100325,
+      "learning_rate": 1.4631390671404682e-05,
+      "loss": 0.3954,
+      "step": 434
+    },
+    {
+      "epoch": 0.8269961977186312,
+      "grad_norm": 0.6983089912012388,
+      "learning_rate": 1.460193232846745e-05,
+      "loss": 0.3944,
+      "step": 435
+    },
+    {
+      "epoch": 0.8288973384030418,
+      "grad_norm": 0.7195173335044036,
+      "learning_rate": 1.4572423233046386e-05,
+      "loss": 0.4331,
+      "step": 436
+    },
+    {
+      "epoch": 0.8307984790874525,
+      "grad_norm": 0.6580077754873673,
+      "learning_rate": 1.4542863710583022e-05,
+      "loss": 0.3942,
+      "step": 437
+    },
+    {
+      "epoch": 0.8326996197718631,
+      "grad_norm": 0.6733246723166247,
+      "learning_rate": 1.4513254087075015e-05,
+      "loss": 0.4085,
+      "step": 438
+    },
+    {
+      "epoch": 0.8346007604562737,
+      "grad_norm": 0.6657800290029222,
+      "learning_rate": 1.4483594689072571e-05,
+      "loss": 0.4053,
+      "step": 439
+    },
+    {
+      "epoch": 0.8365019011406845,
+      "grad_norm": 0.6746881212167681,
+      "learning_rate": 1.4453885843674837e-05,
+      "loss": 0.3814,
+      "step": 440
+    },
+    {
+      "epoch": 0.8384030418250951,
+      "grad_norm": 0.7547054593243069,
+      "learning_rate": 1.4424127878526278e-05,
+      "loss": 0.4032,
+      "step": 441
+    },
+    {
+      "epoch": 0.8403041825095057,
+      "grad_norm": 0.6725366331373717,
+      "learning_rate": 1.4394321121813093e-05,
+      "loss": 0.4104,
+      "step": 442
+    },
+    {
+      "epoch": 0.8422053231939164,
+      "grad_norm": 0.7565414606466715,
+      "learning_rate": 1.436446590225957e-05,
+      "loss": 0.4035,
+      "step": 443
+    },
+    {
+      "epoch": 0.844106463878327,
+      "grad_norm": 0.713176979263959,
+      "learning_rate": 1.433456254912447e-05,
+      "loss": 0.3911,
+      "step": 444
+    },
+    {
+      "epoch": 0.8460076045627376,
+      "grad_norm": 0.6509118241865428,
+      "learning_rate": 1.4304611392197399e-05,
+      "loss": 0.3984,
+      "step": 445
+    },
+    {
+      "epoch": 0.8479087452471483,
+      "grad_norm": 0.8340268492547279,
+      "learning_rate": 1.427461276179517e-05,
+      "loss": 0.4056,
+      "step": 446
+    },
+    {
+      "epoch": 0.8498098859315589,
+      "grad_norm": 0.6758952421736762,
+      "learning_rate": 1.4244566988758152e-05,
+      "loss": 0.3999,
+      "step": 447
+    },
+    {
+      "epoch": 0.8517110266159695,
+      "grad_norm": 0.6827895953297911,
+      "learning_rate": 1.4214474404446633e-05,
+      "loss": 0.4017,
+      "step": 448
+    },
+    {
+      "epoch": 0.8536121673003803,
+      "grad_norm": 0.7711863434469166,
+      "learning_rate": 1.4184335340737158e-05,
+      "loss": 0.4071,
+      "step": 449
+    },
+    {
+      "epoch": 0.8555133079847909,
+      "grad_norm": 0.6781663142258388,
+      "learning_rate": 1.4154150130018867e-05,
+      "loss": 0.4044,
+      "step": 450
+    },
+    {
+      "epoch": 0.8574144486692015,
+      "grad_norm": 0.7551604766767213,
+      "learning_rate": 1.4123919105189836e-05,
+      "loss": 0.4213,
+      "step": 451
+    },
+    {
+      "epoch": 0.8593155893536122,
+      "grad_norm": 0.7250490782553661,
+      "learning_rate": 1.4093642599653406e-05,
+      "loss": 0.3896,
+      "step": 452
+    },
+    {
+      "epoch": 0.8612167300380228,
+      "grad_norm": 0.6783847587289916,
+      "learning_rate": 1.40633209473145e-05,
+      "loss": 0.4228,
+      "step": 453
+    },
+    {
+      "epoch": 0.8631178707224335,
+      "grad_norm": 0.6679146169714046,
+      "learning_rate": 1.4032954482575938e-05,
+      "loss": 0.4171,
+      "step": 454
+    },
+    {
+      "epoch": 0.8650190114068441,
+      "grad_norm": 0.6940343012185746,
+      "learning_rate": 1.4002543540334766e-05,
+      "loss": 0.4151,
+      "step": 455
+    },
+    {
+      "epoch": 0.8669201520912547,
+      "grad_norm": 0.6984027602326602,
+      "learning_rate": 1.3972088455978537e-05,
+      "loss": 0.3984,
+      "step": 456
+    },
+    {
+      "epoch": 0.8688212927756654,
+      "grad_norm": 0.6610983027702386,
+      "learning_rate": 1.3941589565381635e-05,
+      "loss": 0.407,
+      "step": 457
+    },
+    {
+      "epoch": 0.870722433460076,
+      "grad_norm": 0.7624260952295645,
+      "learning_rate": 1.391104720490156e-05,
+      "loss": 0.4152,
+      "step": 458
+    },
+    {
+      "epoch": 0.8726235741444867,
+      "grad_norm": 0.6725609036577359,
+      "learning_rate": 1.3880461711375224e-05,
+      "loss": 0.3933,
+      "step": 459
+    },
+    {
+      "epoch": 0.8745247148288974,
+      "grad_norm": 0.718319776967452,
+      "learning_rate": 1.3849833422115221e-05,
+      "loss": 0.4071,
+      "step": 460
+    },
+    {
+      "epoch": 0.876425855513308,
+      "grad_norm": 0.6981976290895487,
+      "learning_rate": 1.3819162674906134e-05,
+      "loss": 0.4017,
+      "step": 461
+    },
+    {
+      "epoch": 0.8783269961977186,
+      "grad_norm": 0.6494752933923883,
+      "learning_rate": 1.378844980800078e-05,
+      "loss": 0.392,
+      "step": 462
+    },
+    {
+      "epoch": 0.8802281368821293,
+      "grad_norm": 0.6854761697231652,
+      "learning_rate": 1.3757695160116502e-05,
+      "loss": 0.4106,
+      "step": 463
+    },
+    {
+      "epoch": 0.8821292775665399,
+      "grad_norm": 0.6635941524739034,
+      "learning_rate": 1.3726899070431423e-05,
+      "loss": 0.3912,
+      "step": 464
+    },
+    {
+      "epoch": 0.8840304182509505,
+      "grad_norm": 0.6926983482111053,
+      "learning_rate": 1.3696061878580707e-05,
+      "loss": 0.4148,
+      "step": 465
+    },
+    {
+      "epoch": 0.8859315589353612,
+      "grad_norm": 0.773661723974928,
+      "learning_rate": 1.3665183924652817e-05,
+      "loss": 0.4141,
+      "step": 466
+    },
+    {
+      "epoch": 0.8878326996197718,
+      "grad_norm": 0.7205839110020779,
+      "learning_rate": 1.3634265549185755e-05,
+      "loss": 0.3836,
+      "step": 467
+    },
+    {
+      "epoch": 0.8897338403041825,
+      "grad_norm": 0.6732299481307472,
+      "learning_rate": 1.3603307093163319e-05,
+      "loss": 0.42,
+      "step": 468
+    },
+    {
+      "epoch": 0.8916349809885932,
+      "grad_norm": 0.7612253415936112,
+      "learning_rate": 1.3572308898011328e-05,
+      "loss": 0.4097,
+      "step": 469
+    },
+    {
+      "epoch": 0.8935361216730038,
+      "grad_norm": 0.6546109810554149,
+      "learning_rate": 1.3541271305593878e-05,
+      "loss": 0.3992,
+      "step": 470
+    },
+    {
+      "epoch": 0.8954372623574145,
+      "grad_norm": 0.6585874637257052,
+      "learning_rate": 1.3510194658209547e-05,
+      "loss": 0.4027,
+      "step": 471
+    },
+    {
+      "epoch": 0.8973384030418251,
+      "grad_norm": 0.7652282223913617,
+      "learning_rate": 1.3479079298587634e-05,
+      "loss": 0.4005,
+      "step": 472
+    },
+    {
+      "epoch": 0.8992395437262357,
+      "grad_norm": 0.6933073840200412,
+      "learning_rate": 1.3447925569884374e-05,
+      "loss": 0.4087,
+      "step": 473
+    },
+    {
+      "epoch": 0.9011406844106464,
+      "grad_norm": 0.6550862975055665,
+      "learning_rate": 1.3416733815679166e-05,
+      "loss": 0.3978,
+      "step": 474
+    },
+    {
+      "epoch": 0.903041825095057,
+      "grad_norm": 0.7256952053007989,
+      "learning_rate": 1.3385504379970764e-05,
+      "loss": 0.4046,
+      "step": 475
+    },
+    {
+      "epoch": 0.9049429657794676,
+      "grad_norm": 0.7148585643351413,
+      "learning_rate": 1.3354237607173494e-05,
+      "loss": 0.3978,
+      "step": 476
+    },
+    {
+      "epoch": 0.9068441064638784,
+      "grad_norm": 0.7142015421348241,
+      "learning_rate": 1.3322933842113457e-05,
+      "loss": 0.4209,
+      "step": 477
+    },
+    {
+      "epoch": 0.908745247148289,
+      "grad_norm": 0.687730106015921,
+      "learning_rate": 1.3291593430024727e-05,
+      "loss": 0.4043,
+      "step": 478
+    },
+    {
+      "epoch": 0.9106463878326996,
+      "grad_norm": 0.6822137552970915,
+      "learning_rate": 1.3260216716545534e-05,
+      "loss": 0.4171,
+      "step": 479
+    },
+    {
+      "epoch": 0.9125475285171103,
+      "grad_norm": 0.6377657140279679,
+      "learning_rate": 1.3228804047714462e-05,
+      "loss": 0.3887,
+      "step": 480
+    },
+    {
+      "epoch": 0.9144486692015209,
+      "grad_norm": 0.6966064429186608,
+      "learning_rate": 1.319735576996663e-05,
+      "loss": 0.422,
+      "step": 481
+    },
+    {
+      "epoch": 0.9163498098859315,
+      "grad_norm": 0.6902117617603477,
+      "learning_rate": 1.3165872230129869e-05,
+      "loss": 0.404,
+      "step": 482
+    },
+    {
+      "epoch": 0.9182509505703422,
+      "grad_norm": 0.6325022995547926,
+      "learning_rate": 1.3134353775420895e-05,
+      "loss": 0.3899,
+      "step": 483
+    },
+    {
+      "epoch": 0.9201520912547528,
+      "grad_norm": 0.6311490054113847,
+      "learning_rate": 1.3102800753441488e-05,
+      "loss": 0.404,
+      "step": 484
+    },
+    {
+      "epoch": 0.9220532319391636,
+      "grad_norm": 0.7119891602684172,
+      "learning_rate": 1.3071213512174655e-05,
+      "loss": 0.3847,
+      "step": 485
+    },
+    {
+      "epoch": 0.9239543726235742,
+      "grad_norm": 0.7140520982978685,
+      "learning_rate": 1.3039592399980785e-05,
+      "loss": 0.389,
+      "step": 486
+    },
+    {
+      "epoch": 0.9258555133079848,
+      "grad_norm": 0.6648479626315674,
+      "learning_rate": 1.3007937765593818e-05,
+      "loss": 0.4076,
+      "step": 487
+    },
+    {
+      "epoch": 0.9277566539923955,
+      "grad_norm": 0.7207022349815417,
+      "learning_rate": 1.2976249958117395e-05,
+      "loss": 0.4197,
+      "step": 488
+    },
+    {
+      "epoch": 0.9296577946768061,
+      "grad_norm": 0.6538093572467635,
+      "learning_rate": 1.2944529327021002e-05,
+      "loss": 0.4003,
+      "step": 489
+    },
+    {
+      "epoch": 0.9315589353612167,
+      "grad_norm": 0.6145457797778117,
+      "learning_rate": 1.291277622213612e-05,
+      "loss": 0.3861,
+      "step": 490
+    },
+    {
+      "epoch": 0.9334600760456274,
+      "grad_norm": 0.6717081075163996,
+      "learning_rate": 1.2880990993652379e-05,
+      "loss": 0.404,
+      "step": 491
+    },
+    {
+      "epoch": 0.935361216730038,
+      "grad_norm": 0.6381375708758967,
+      "learning_rate": 1.2849173992113669e-05,
+      "loss": 0.3907,
+      "step": 492
+    },
+    {
+      "epoch": 0.9372623574144486,
+      "grad_norm": 0.6405334578995517,
+      "learning_rate": 1.2817325568414299e-05,
+      "loss": 0.3966,
+      "step": 493
+    },
+    {
+      "epoch": 0.9391634980988594,
+      "grad_norm": 0.6731779594600235,
+      "learning_rate": 1.2785446073795118e-05,
+      "loss": 0.4034,
+      "step": 494
+    },
+    {
+      "epoch": 0.94106463878327,
+      "grad_norm": 0.6969180914718714,
+      "learning_rate": 1.2753535859839638e-05,
+      "loss": 0.4121,
+      "step": 495
+    },
+    {
+      "epoch": 0.9429657794676806,
+      "grad_norm": 0.618180483581813,
+      "learning_rate": 1.272159527847016e-05,
+      "loss": 0.3816,
+      "step": 496
+    },
+    {
+      "epoch": 0.9448669201520913,
+      "grad_norm": 0.6952759330447784,
+      "learning_rate": 1.2689624681943897e-05,
+      "loss": 0.413,
+      "step": 497
+    },
+    {
+      "epoch": 0.9467680608365019,
+      "grad_norm": 0.6898427307299725,
+      "learning_rate": 1.2657624422849077e-05,
+      "loss": 0.4179,
+      "step": 498
+    },
+    {
+      "epoch": 0.9486692015209125,
+      "grad_norm": 0.699291230631293,
+      "learning_rate": 1.2625594854101066e-05,
+      "loss": 0.4157,
+      "step": 499
+    },
+    {
+      "epoch": 0.9505703422053232,
+      "grad_norm": 0.6694458644588648,
+      "learning_rate": 1.2593536328938471e-05,
+      "loss": 0.406,
+      "step": 500
+    },
+    {
+      "epoch": 0.9524714828897338,
+      "grad_norm": 0.6470287009480223,
+      "learning_rate": 1.2561449200919253e-05,
+      "loss": 0.3892,
+      "step": 501
+    },
+    {
+      "epoch": 0.9543726235741445,
+      "grad_norm": 0.6077026389317716,
+      "learning_rate": 1.2529333823916807e-05,
+      "loss": 0.385,
+      "step": 502
+    },
+    {
+      "epoch": 0.9562737642585551,
+      "grad_norm": 0.7033970824785918,
+      "learning_rate": 1.2497190552116082e-05,
+      "loss": 0.4175,
+      "step": 503
+    },
+    {
+      "epoch": 0.9581749049429658,
+      "grad_norm": 0.6803712575864149,
+      "learning_rate": 1.2465019740009662e-05,
+      "loss": 0.3934,
+      "step": 504
+    },
+    {
+      "epoch": 0.9600760456273765,
+      "grad_norm": 0.6965934029353575,
+      "learning_rate": 1.2432821742393854e-05,
+      "loss": 0.412,
+      "step": 505
+    },
+    {
+      "epoch": 0.9619771863117871,
+      "grad_norm": 0.6163645226860306,
+      "learning_rate": 1.2400596914364792e-05,
+      "loss": 0.3975,
+      "step": 506
+    },
+    {
+      "epoch": 0.9638783269961977,
+      "grad_norm": 0.7311609087122628,
+      "learning_rate": 1.2368345611314508e-05,
+      "loss": 0.4082,
+      "step": 507
+    },
+    {
+      "epoch": 0.9657794676806084,
+      "grad_norm": 0.6183545409406597,
+      "learning_rate": 1.2336068188927002e-05,
+      "loss": 0.3789,
+      "step": 508
+    },
+    {
+      "epoch": 0.967680608365019,
+      "grad_norm": 0.6792795167354,
+      "learning_rate": 1.2303765003174342e-05,
+      "loss": 0.4054,
+      "step": 509
+    },
+    {
+      "epoch": 0.9695817490494296,
+      "grad_norm": 0.6667740298389744,
+      "learning_rate": 1.2271436410312727e-05,
+      "loss": 0.4252,
+      "step": 510
+    },
+    {
+      "epoch": 0.9714828897338403,
+      "grad_norm": 0.7651718493141103,
+      "learning_rate": 1.2239082766878557e-05,
+      "loss": 0.4111,
+      "step": 511
+    },
+    {
+      "epoch": 0.973384030418251,
+      "grad_norm": 0.7086287990509619,
+      "learning_rate": 1.2206704429684504e-05,
+      "loss": 0.405,
+      "step": 512
+    },
+    {
+      "epoch": 0.9752851711026616,
+      "grad_norm": 0.6852672674776774,
+      "learning_rate": 1.2174301755815572e-05,
+      "loss": 0.414,
+      "step": 513
+    },
+    {
+      "epoch": 0.9771863117870723,
+      "grad_norm": 0.7193466666893342,
+      "learning_rate": 1.2141875102625166e-05,
+      "loss": 0.3903,
+      "step": 514
+    },
+    {
+      "epoch": 0.9790874524714829,
+      "grad_norm": 0.6858267717011416,
+      "learning_rate": 1.2109424827731144e-05,
+      "loss": 0.4138,
+      "step": 515
+    },
+    {
+      "epoch": 0.9809885931558935,
+      "grad_norm": 0.7580356802405223,
+      "learning_rate": 1.2076951289011884e-05,
+      "loss": 0.4244,
+      "step": 516
+    },
+    {
+      "epoch": 0.9828897338403042,
+      "grad_norm": 0.6622004562904025,
+      "learning_rate": 1.204445484460232e-05,
+      "loss": 0.3984,
+      "step": 517
+    },
+    {
+      "epoch": 0.9847908745247148,
+      "grad_norm": 0.6636607447233084,
+      "learning_rate": 1.2011935852890004e-05,
+      "loss": 0.421,
+      "step": 518
+    },
+    {
+      "epoch": 0.9866920152091255,
+      "grad_norm": 0.6495045702865927,
+      "learning_rate": 1.1979394672511156e-05,
+      "loss": 0.3729,
+      "step": 519
+    },
+    {
+      "epoch": 0.9885931558935361,
+      "grad_norm": 0.6981543503300218,
+      "learning_rate": 1.19468316623467e-05,
+      "loss": 0.4029,
+      "step": 520
+    },
+    {
+      "epoch": 0.9904942965779467,
+      "grad_norm": 0.6662456471440278,
+      "learning_rate": 1.1914247181518312e-05,
+      "loss": 0.4002,
+      "step": 521
+    },
+    {
+      "epoch": 0.9923954372623575,
+      "grad_norm": 0.6604787309214006,
+      "learning_rate": 1.1881641589384456e-05,
+      "loss": 0.4031,
+      "step": 522
+    },
+    {
+      "epoch": 0.9942965779467681,
+      "grad_norm": 0.650902110560888,
+      "learning_rate": 1.1849015245536424e-05,
+      "loss": 0.4031,
+      "step": 523
+    },
+    {
+      "epoch": 0.9961977186311787,
+      "grad_norm": 0.6798864680139955,
+      "learning_rate": 1.1816368509794365e-05,
+      "loss": 0.4123,
+      "step": 524
+    },
+    {
+      "epoch": 0.9980988593155894,
+      "grad_norm": 0.628617009647784,
+      "learning_rate": 1.1783701742203326e-05,
+      "loss": 0.3965,
+      "step": 525
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8265708108728387,
+      "learning_rate": 1.1751015303029272e-05,
+      "loss": 0.385,
+      "step": 526
+    },
+    {
+      "epoch": 1.0019011406844107,
+      "grad_norm": 0.7680727922669428,
+      "learning_rate": 1.1718309552755118e-05,
+      "loss": 0.3102,
+      "step": 527
+    },
+    {
+      "epoch": 1.0038022813688212,
+      "grad_norm": 1.0059228784897538,
+      "learning_rate": 1.1685584852076746e-05,
+      "loss": 0.3157,
+      "step": 528
+    },
+    {
+      "epoch": 1.005703422053232,
+      "grad_norm": 0.6815057663621917,
+      "learning_rate": 1.1652841561899042e-05,
+      "loss": 0.307,
+      "step": 529
+    },
+    {
+      "epoch": 1.0076045627376427,
+      "grad_norm": 0.6843106380799396,
+      "learning_rate": 1.1620080043331901e-05,
+      "loss": 0.3048,
+      "step": 530
+    },
+    {
+      "epoch": 1.0095057034220531,
+      "grad_norm": 0.8839345706214902,
+      "learning_rate": 1.1587300657686254e-05,
+      "loss": 0.3108,
+      "step": 531
+    },
+    {
+      "epoch": 1.0114068441064639,
+      "grad_norm": 0.8303137006693243,
+      "learning_rate": 1.1554503766470069e-05,
+      "loss": 0.3047,
+      "step": 532
+    },
+    {
+      "epoch": 1.0133079847908746,
+      "grad_norm": 0.8892874126831852,
+      "learning_rate": 1.1521689731384391e-05,
+      "loss": 0.3063,
+      "step": 533
+    },
+    {
+      "epoch": 1.015209125475285,
+      "grad_norm": 0.8054007880181805,
+      "learning_rate": 1.1488858914319321e-05,
+      "loss": 0.3027,
+      "step": 534
+    },
+    {
+      "epoch": 1.0171102661596958,
+      "grad_norm": 0.7287944687516933,
+      "learning_rate": 1.1456011677350052e-05,
+      "loss": 0.2995,
+      "step": 535
+    },
+    {
+      "epoch": 1.0190114068441065,
+      "grad_norm": 0.7348972384405209,
+      "learning_rate": 1.1423148382732854e-05,
+      "loss": 0.3088,
+      "step": 536
+    },
+    {
+      "epoch": 1.020912547528517,
+      "grad_norm": 0.8489682812305904,
+      "learning_rate": 1.1390269392901096e-05,
+      "loss": 0.307,
+      "step": 537
+    },
+    {
+      "epoch": 1.0228136882129277,
+      "grad_norm": 0.7904156627221454,
+      "learning_rate": 1.1357375070461241e-05,
+      "loss": 0.2947,
+      "step": 538
+    },
+    {
+      "epoch": 1.0247148288973384,
+      "grad_norm": 0.6982214896146977,
+      "learning_rate": 1.1324465778188846e-05,
+      "loss": 0.2902,
+      "step": 539
+    },
+    {
+      "epoch": 1.026615969581749,
+      "grad_norm": 0.7784279498569209,
+      "learning_rate": 1.1291541879024568e-05,
+      "loss": 0.3027,
+      "step": 540
+    },
+    {
+      "epoch": 1.0285171102661597,
+      "grad_norm": 0.7453010549356432,
+      "learning_rate": 1.1258603736070145e-05,
+      "loss": 0.2725,
+      "step": 541
+    },
+    {
+      "epoch": 1.0304182509505704,
+      "grad_norm": 0.759587479503962,
+      "learning_rate": 1.1225651712584413e-05,
+      "loss": 0.3005,
+      "step": 542
+    },
+    {
+      "epoch": 1.032319391634981,
+      "grad_norm": 0.7533765897520451,
+      "learning_rate": 1.1192686171979288e-05,
+      "loss": 0.2931,
+      "step": 543
+    },
+    {
+      "epoch": 1.0342205323193916,
+      "grad_norm": 0.9132088527957141,
+      "learning_rate": 1.1159707477815756e-05,
+      "loss": 0.2883,
+      "step": 544
+    },
+    {
+      "epoch": 1.0361216730038023,
+      "grad_norm": 0.7770234792600628,
+      "learning_rate": 1.1126715993799875e-05,
+      "loss": 0.295,
+      "step": 545
+    },
+    {
+      "epoch": 1.038022813688213,
+      "grad_norm": 0.7090004784228509,
+      "learning_rate": 1.1093712083778748e-05,
+      "loss": 0.2812,
+      "step": 546
+    },
+    {
+      "epoch": 1.0399239543726235,
+      "grad_norm": 0.9581263392106095,
+      "learning_rate": 1.1060696111736515e-05,
+      "loss": 0.3059,
+      "step": 547
+    },
+    {
+      "epoch": 1.0418250950570342,
+      "grad_norm": 0.8609157729184757,
+      "learning_rate": 1.1027668441790358e-05,
+      "loss": 0.2977,
+      "step": 548
+    },
+    {
+      "epoch": 1.043726235741445,
+      "grad_norm": 0.8034824100554511,
+      "learning_rate": 1.099462943818646e-05,
+      "loss": 0.3047,
+      "step": 549
+    },
+    {
+      "epoch": 1.0456273764258555,
+      "grad_norm": 0.8362078393611865,
+      "learning_rate": 1.0961579465295987e-05,
+      "loss": 0.2854,
+      "step": 550
+    },
+    {
+      "epoch": 1.0475285171102662,
+      "grad_norm": 0.8052304462734117,
+      "learning_rate": 1.0928518887611099e-05,
+      "loss": 0.3048,
+      "step": 551
+    },
+    {
+      "epoch": 1.049429657794677,
+      "grad_norm": 0.6757313280701089,
+      "learning_rate": 1.0895448069740902e-05,
+      "loss": 0.2926,
+      "step": 552
+    },
+    {
+      "epoch": 1.0513307984790874,
+      "grad_norm": 0.757814413923201,
+      "learning_rate": 1.0862367376407433e-05,
+      "loss": 0.2967,
+      "step": 553
+    },
+    {
+      "epoch": 1.053231939163498,
+      "grad_norm": 0.8824935621625047,
+      "learning_rate": 1.0829277172441648e-05,
+      "loss": 0.3089,
+      "step": 554
+    },
+    {
+      "epoch": 1.0551330798479088,
+      "grad_norm": 0.6561496691480129,
+      "learning_rate": 1.0796177822779384e-05,
+      "loss": 0.2763,
+      "step": 555
+    },
+    {
+      "epoch": 1.0570342205323193,
+      "grad_norm": 0.7447663394849013,
+      "learning_rate": 1.0763069692457346e-05,
+      "loss": 0.3005,
+      "step": 556
+    },
+    {
+      "epoch": 1.05893536121673,
+      "grad_norm": 0.7583657632247124,
+      "learning_rate": 1.0729953146609076e-05,
+      "loss": 0.2945,
+      "step": 557
+    },
+    {
+      "epoch": 1.0608365019011408,
+      "grad_norm": 0.7254406369200254,
+      "learning_rate": 1.0696828550460928e-05,
+      "loss": 0.2984,
+      "step": 558
+    },
+    {
+      "epoch": 1.0627376425855513,
+      "grad_norm": 0.7149689970744459,
+      "learning_rate": 1.0663696269328034e-05,
+      "loss": 0.3029,
+      "step": 559
+    },
+    {
+      "epoch": 1.064638783269962,
+      "grad_norm": 0.7554356406107401,
+      "learning_rate": 1.0630556668610286e-05,
+      "loss": 0.2799,
+      "step": 560
+    },
+    {
+      "epoch": 1.0665399239543727,
+      "grad_norm": 0.7162776578629291,
+      "learning_rate": 1.059741011378829e-05,
+      "loss": 0.2908,
+      "step": 561
+    },
+    {
+      "epoch": 1.0684410646387832,
+      "grad_norm": 0.6767578441666678,
+      "learning_rate": 1.0564256970419367e-05,
+      "loss": 0.2894,
+      "step": 562
+    },
+    {
+      "epoch": 1.070342205323194,
+      "grad_norm": 0.7246406918900237,
+      "learning_rate": 1.0531097604133473e-05,
+      "loss": 0.2963,
+      "step": 563
+    },
+    {
+      "epoch": 1.0722433460076046,
+      "grad_norm": 0.7819364408744047,
+      "learning_rate": 1.0497932380629207e-05,
+      "loss": 0.2869,
+      "step": 564
+    },
+    {
+      "epoch": 1.0741444866920151,
+      "grad_norm": 0.707692605716939,
+      "learning_rate": 1.0464761665669771e-05,
+      "loss": 0.306,
+      "step": 565
+    },
+    {
+      "epoch": 1.0760456273764258,
+      "grad_norm": 0.6881486143947666,
+      "learning_rate": 1.0431585825078916e-05,
+      "loss": 0.2921,
+      "step": 566
+    },
+    {
+      "epoch": 1.0779467680608366,
+      "grad_norm": 0.6996755754699221,
+      "learning_rate": 1.0398405224736927e-05,
+      "loss": 0.301,
+      "step": 567
+    },
+    {
+      "epoch": 1.079847908745247,
+      "grad_norm": 0.6813043003304055,
+      "learning_rate": 1.0365220230576592e-05,
+      "loss": 0.2834,
+      "step": 568
+    },
+    {
+      "epoch": 1.0817490494296578,
+      "grad_norm": 0.728597382505159,
+      "learning_rate": 1.0332031208579133e-05,
+      "loss": 0.2955,
+      "step": 569
+    },
+    {
+      "epoch": 1.0836501901140685,
+      "grad_norm": 0.7013527462079502,
+      "learning_rate": 1.0298838524770212e-05,
+      "loss": 0.2953,
+      "step": 570
+    },
+    {
+      "epoch": 1.085551330798479,
+      "grad_norm": 0.6876135582500057,
+      "learning_rate": 1.0265642545215872e-05,
+      "loss": 0.2958,
+      "step": 571
+    },
+    {
+      "epoch": 1.0874524714828897,
+      "grad_norm": 0.6923169491384853,
+      "learning_rate": 1.0232443636018502e-05,
+      "loss": 0.2975,
+      "step": 572
+    },
+    {
+      "epoch": 1.0893536121673004,
+      "grad_norm": 0.6632548937249962,
+      "learning_rate": 1.0199242163312794e-05,
+      "loss": 0.2855,
+      "step": 573
+    },
+    {
+      "epoch": 1.091254752851711,
+      "grad_norm": 0.6494008105848307,
+      "learning_rate": 1.0166038493261723e-05,
+      "loss": 0.2957,
+      "step": 574
+    },
+    {
+      "epoch": 1.0931558935361216,
+      "grad_norm": 0.735892211231415,
+      "learning_rate": 1.013283299205249e-05,
+      "loss": 0.2976,
+      "step": 575
+    },
+    {
+      "epoch": 1.0950570342205324,
+      "grad_norm": 0.6959908109311861,
+      "learning_rate": 1.0099626025892491e-05,
+      "loss": 0.2874,
+      "step": 576
+    },
+    {
+      "epoch": 1.0969581749049429,
+      "grad_norm": 0.7177111387542888,
+      "learning_rate": 1.0066417961005283e-05,
+      "loss": 0.2979,
+      "step": 577
+    },
+    {
+      "epoch": 1.0988593155893536,
+      "grad_norm": 0.6953888719227289,
+      "learning_rate": 1.0033209163626539e-05,
+      "loss": 0.292,
+      "step": 578
+    },
+    {
+      "epoch": 1.1007604562737643,
+      "grad_norm": 0.7083766299450723,
+      "learning_rate": 1e-05,
+      "loss": 0.2829,
+      "step": 579
+    },
+    {
+      "epoch": 1.102661596958175,
+      "grad_norm": 0.7080381925816993,
+      "learning_rate": 9.966790836373465e-06,
+      "loss": 0.2865,
+      "step": 580
+    },
+    {
+      "epoch": 1.1045627376425855,
+      "grad_norm": 0.7286544209683984,
+      "learning_rate": 9.933582038994719e-06,
+      "loss": 0.2997,
+      "step": 581
+    },
+    {
+      "epoch": 1.1064638783269962,
+      "grad_norm": 0.6761518102078841,
+      "learning_rate": 9.90037397410751e-06,
+      "loss": 0.2906,
+      "step": 582
+    },
+    {
+      "epoch": 1.108365019011407,
+      "grad_norm": 0.7230980131346857,
+      "learning_rate": 9.867167007947511e-06,
+      "loss": 0.2963,
+      "step": 583
+    },
+    {
+      "epoch": 1.1102661596958174,
+      "grad_norm": 0.7364339625878032,
+      "learning_rate": 9.833961506738282e-06,
+      "loss": 0.2972,
+      "step": 584
+    },
+    {
+      "epoch": 1.1121673003802282,
+      "grad_norm": 0.73189078228728,
+      "learning_rate": 9.80075783668721e-06,
+      "loss": 0.2954,
+      "step": 585
+    },
+    {
+      "epoch": 1.1140684410646389,
+      "grad_norm": 0.7421538721938589,
+      "learning_rate": 9.767556363981503e-06,
+      "loss": 0.2892,
+      "step": 586
+    },
+    {
+      "epoch": 1.1159695817490494,
+      "grad_norm": 0.709171276811597,
+      "learning_rate": 9.734357454784131e-06,
+      "loss": 0.2805,
+      "step": 587
+    },
+    {
+      "epoch": 1.11787072243346,
+      "grad_norm": 0.6904742526194513,
+      "learning_rate": 9.701161475229791e-06,
+      "loss": 0.2913,
+      "step": 588
+    },
+    {
+      "epoch": 1.1197718631178708,
+      "grad_norm": 0.7150803253416469,
+      "learning_rate": 9.66796879142087e-06,
+      "loss": 0.2978,
+      "step": 589
+    },
+    {
+      "epoch": 1.1216730038022813,
+      "grad_norm": 0.6739329092291103,
+      "learning_rate": 9.634779769423412e-06,
+      "loss": 0.2815,
+      "step": 590
+    },
+    {
+      "epoch": 1.123574144486692,
+      "grad_norm": 0.7250322303380577,
+      "learning_rate": 9.601594775263073e-06,
+      "loss": 0.286,
+      "step": 591
+    },
+    {
+      "epoch": 1.1254752851711027,
+      "grad_norm": 0.8094477868153702,
+      "learning_rate": 9.568414174921085e-06,
+      "loss": 0.3124,
+      "step": 592
+    },
+    {
+      "epoch": 1.1273764258555132,
+      "grad_norm": 0.7502812677192022,
+      "learning_rate": 9.535238334330234e-06,
+      "loss": 0.2988,
+      "step": 593
+    },
+    {
+      "epoch": 1.129277566539924,
+      "grad_norm": 0.7119909198968319,
+      "learning_rate": 9.502067619370794e-06,
+      "loss": 0.2992,
+      "step": 594
+    },
+    {
+      "epoch": 1.1311787072243347,
+      "grad_norm": 0.7105064977926391,
+      "learning_rate": 9.468902395866532e-06,
+      "loss": 0.2989,
+      "step": 595
+    },
+    {
+      "epoch": 1.1330798479087452,
+      "grad_norm": 0.7321763340548684,
+      "learning_rate": 9.435743029580638e-06,
+      "loss": 0.2899,
+      "step": 596
+    },
+    {
+      "epoch": 1.1349809885931559,
+      "grad_norm": 0.6961265075411925,
+      "learning_rate": 9.402589886211711e-06,
+      "loss": 0.2888,
+      "step": 597
+    },
+    {
+      "epoch": 1.1368821292775666,
+      "grad_norm": 0.7258014284775477,
+      "learning_rate": 9.369443331389718e-06,
+      "loss": 0.3089,
+      "step": 598
+    },
+    {
+      "epoch": 1.138783269961977,
+      "grad_norm": 0.7634271257053743,
+      "learning_rate": 9.336303730671968e-06,
+      "loss": 0.2938,
+      "step": 599
+    },
+    {
+      "epoch": 1.1406844106463878,
+      "grad_norm": 0.73614464279472,
+      "learning_rate": 9.303171449539074e-06,
+      "loss": 0.3038,
+      "step": 600
+    },
+    {
+      "epoch": 1.1425855513307985,
+      "grad_norm": 0.6797341272857729,
+      "learning_rate": 9.270046853390924e-06,
+      "loss": 0.2935,
+      "step": 601
+    },
+    {
+      "epoch": 1.144486692015209,
+      "grad_norm": 0.7643628334850742,
+      "learning_rate": 9.236930307542654e-06,
+      "loss": 0.2926,
+      "step": 602
+    },
+    {
+      "epoch": 1.1463878326996197,
+      "grad_norm": 0.6955099869929693,
+      "learning_rate": 9.203822177220621e-06,
+      "loss": 0.2827,
+      "step": 603
+    },
+    {
+      "epoch": 1.1482889733840305,
+      "grad_norm": 0.7538104752903337,
+      "learning_rate": 9.170722827558357e-06,
+      "loss": 0.295,
+      "step": 604
+    },
+    {
+      "epoch": 1.1501901140684412,
+      "grad_norm": 0.7443764594532538,
+      "learning_rate": 9.13763262359257e-06,
+      "loss": 0.3024,
+      "step": 605
+    },
+    {
+      "epoch": 1.1520912547528517,
+      "grad_norm": 0.7400515219435118,
+      "learning_rate": 9.104551930259101e-06,
+      "loss": 0.2975,
+      "step": 606
+    },
+    {
+      "epoch": 1.1539923954372624,
+      "grad_norm": 0.7377989638091142,
+      "learning_rate": 9.071481112388905e-06,
+      "loss": 0.3021,
+      "step": 607
+    },
+    {
+      "epoch": 1.1558935361216731,
+      "grad_norm": 0.6351873542370496,
+      "learning_rate": 9.038420534704015e-06,
+      "loss": 0.287,
+      "step": 608
+    },
+    {
+      "epoch": 1.1577946768060836,
+      "grad_norm": 0.7257461591215666,
+      "learning_rate": 9.005370561813545e-06,
+      "loss": 0.3081,
+      "step": 609
+    },
+    {
+      "epoch": 1.1596958174904943,
+      "grad_norm": 0.671370248912275,
+      "learning_rate": 8.972331558209644e-06,
+      "loss": 0.3062,
+      "step": 610
+    },
+    {
+      "epoch": 1.161596958174905,
+      "grad_norm": 0.6493761967149301,
+      "learning_rate": 8.939303888263485e-06,
+      "loss": 0.2962,
+      "step": 611
+    },
+    {
+      "epoch": 1.1634980988593155,
+      "grad_norm": 0.686426789947725,
+      "learning_rate": 8.906287916221259e-06,
+      "loss": 0.2906,
+      "step": 612
+    },
+    {
+      "epoch": 1.1653992395437263,
+      "grad_norm": 0.690026487488721,
+      "learning_rate": 8.873284006200129e-06,
+      "loss": 0.2896,
+      "step": 613
+    },
+    {
+      "epoch": 1.167300380228137,
+      "grad_norm": 0.6872226384582601,
+      "learning_rate": 8.840292522184247e-06,
+      "loss": 0.2817,
+      "step": 614
+    },
+    {
+      "epoch": 1.1692015209125475,
+      "grad_norm": 0.7004489925928739,
+      "learning_rate": 8.807313828020715e-06,
+      "loss": 0.2936,
+      "step": 615
+    },
+    {
+      "epoch": 1.1711026615969582,
+      "grad_norm": 0.8125254022073283,
+      "learning_rate": 8.774348287415589e-06,
+      "loss": 0.3087,
+      "step": 616
+    },
+    {
+      "epoch": 1.173003802281369,
+      "grad_norm": 0.7569157165063869,
+      "learning_rate": 8.74139626392986e-06,
+      "loss": 0.305,
+      "step": 617
+    },
+    {
+      "epoch": 1.1749049429657794,
+      "grad_norm": 0.8607321690785082,
+      "learning_rate": 8.708458120975436e-06,
+      "loss": 0.3007,
+      "step": 618
+    },
+    {
+      "epoch": 1.1768060836501901,
+      "grad_norm": 0.7337026909841357,
+      "learning_rate": 8.675534221811156e-06,
+      "loss": 0.2888,
+      "step": 619
+    },
+    {
+      "epoch": 1.1787072243346008,
+      "grad_norm": 0.7420738957496412,
+      "learning_rate": 8.64262492953876e-06,
+      "loss": 0.292,
+      "step": 620
+    },
+    {
+      "epoch": 1.1806083650190113,
+      "grad_norm": 0.7658329152469303,
+      "learning_rate": 8.60973060709891e-06,
+      "loss": 0.3158,
+      "step": 621
+    },
+    {
+      "epoch": 1.182509505703422,
+      "grad_norm": 0.6977889118897987,
+      "learning_rate": 8.576851617267151e-06,
+      "loss": 0.2907,
+      "step": 622
+    },
+    {
+      "epoch": 1.1844106463878328,
+      "grad_norm": 0.7337642630477222,
+      "learning_rate": 8.543988322649954e-06,
+      "loss": 0.3118,
+      "step": 623
+    },
+    {
+      "epoch": 1.1863117870722433,
+      "grad_norm": 0.6925312820115364,
+      "learning_rate": 8.511141085680684e-06,
+      "loss": 0.2954,
+      "step": 624
+    },
+    {
+      "epoch": 1.188212927756654,
+      "grad_norm": 0.6946774043184807,
+      "learning_rate": 8.478310268615612e-06,
+      "loss": 0.302,
+      "step": 625
+    },
+    {
+      "epoch": 1.1901140684410647,
+      "grad_norm": 0.6768648494708821,
+      "learning_rate": 8.445496233529934e-06,
+      "loss": 0.2923,
+      "step": 626
+    },
+    {
+      "epoch": 1.1920152091254752,
+      "grad_norm": 0.707499524009905,
+      "learning_rate": 8.41269934231375e-06,
+      "loss": 0.307,
+      "step": 627
+    },
+    {
+      "epoch": 1.193916349809886,
+      "grad_norm": 0.6693483006844595,
+      "learning_rate": 8.3799199566681e-06,
+      "loss": 0.2835,
+      "step": 628
+    },
+    {
+      "epoch": 1.1958174904942966,
+      "grad_norm": 0.6756297972031414,
+      "learning_rate": 8.34715843810096e-06,
+      "loss": 0.2804,
+      "step": 629
+    },
+    {
+      "epoch": 1.1977186311787071,
+      "grad_norm": 0.7185793007102922,
+      "learning_rate": 8.314415147923254e-06,
+      "loss": 0.3013,
+      "step": 630
+    },
+    {
+      "epoch": 1.1996197718631179,
+      "grad_norm": 0.6917942502617963,
+      "learning_rate": 8.281690447244887e-06,
+      "loss": 0.2901,
+      "step": 631
+    },
+    {
+      "epoch": 1.2015209125475286,
+      "grad_norm": 0.6922299225747153,
+      "learning_rate": 8.248984696970732e-06,
+      "loss": 0.2912,
+      "step": 632
+    },
+    {
+      "epoch": 1.203422053231939,
+      "grad_norm": 0.7215625334990756,
+      "learning_rate": 8.216298257796677e-06,
+      "loss": 0.3058,
+      "step": 633
+    },
+    {
+      "epoch": 1.2053231939163498,
+      "grad_norm": 0.7039222410582643,
+      "learning_rate": 8.183631490205636e-06,
+      "loss": 0.292,
+      "step": 634
+    },
+    {
+      "epoch": 1.2072243346007605,
+      "grad_norm": 0.7190859502492053,
+      "learning_rate": 8.150984754463578e-06,
+      "loss": 0.2948,
+      "step": 635
+    },
+    {
+      "epoch": 1.209125475285171,
+      "grad_norm": 0.7334102633794286,
+      "learning_rate": 8.118358410615545e-06,
+      "loss": 0.2811,
+      "step": 636
+    },
+    {
+      "epoch": 1.2110266159695817,
+      "grad_norm": 0.7643201091864527,
+      "learning_rate": 8.08575281848169e-06,
+      "loss": 0.2824,
+      "step": 637
+    },
+    {
+      "epoch": 1.2129277566539924,
+      "grad_norm": 0.774557710183706,
+      "learning_rate": 8.0531683376533e-06,
+      "loss": 0.2779,
+      "step": 638
+    },
+    {
+      "epoch": 1.214828897338403,
+      "grad_norm": 0.6902011960902502,
+      "learning_rate": 8.020605327488846e-06,
+      "loss": 0.2956,
+      "step": 639
+    },
+    {
+      "epoch": 1.2167300380228137,
+      "grad_norm": 0.7179399279179142,
+      "learning_rate": 7.988064147110001e-06,
+      "loss": 0.3033,
+      "step": 640
+    },
+    {
+      "epoch": 1.2186311787072244,
+      "grad_norm": 0.7016361082390848,
+      "learning_rate": 7.955545155397684e-06,
+      "loss": 0.2949,
+      "step": 641
+    },
+    {
+      "epoch": 1.2205323193916349,
+      "grad_norm": 0.703951517850484,
+      "learning_rate": 7.923048710988119e-06,
+      "loss": 0.2971,
+      "step": 642
+    },
+    {
+      "epoch": 1.2224334600760456,
+      "grad_norm": 0.7484726688095609,
+      "learning_rate": 7.890575172268858e-06,
+      "loss": 0.3055,
+      "step": 643
+    },
+    {
+      "epoch": 1.2243346007604563,
+      "grad_norm": 0.6904067527237151,
+      "learning_rate": 7.858124897374837e-06,
+      "loss": 0.2832,
+      "step": 644
+    },
+    {
+      "epoch": 1.2262357414448668,
+      "grad_norm": 0.7925625613106448,
+      "learning_rate": 7.825698244184432e-06,
+      "loss": 0.2889,
+      "step": 645
+    },
+    {
+      "epoch": 1.2281368821292775,
+      "grad_norm": 0.7082163840176895,
+      "learning_rate": 7.7932955703155e-06,
+      "loss": 0.2843,
+      "step": 646
+    },
+    {
+      "epoch": 1.2300380228136882,
+      "grad_norm": 0.7486971273363527,
+      "learning_rate": 7.760917233121443e-06,
+      "loss": 0.3093,
+      "step": 647
+    },
+    {
+      "epoch": 1.231939163498099,
+      "grad_norm": 0.6885731853188624,
+      "learning_rate": 7.728563589687275e-06,
+      "loss": 0.2835,
+      "step": 648
+    },
+    {
+      "epoch": 1.2338403041825095,
+      "grad_norm": 0.6929242766183379,
+      "learning_rate": 7.696234996825663e-06,
+      "loss": 0.2962,
+      "step": 649
+    },
+    {
+      "epoch": 1.2357414448669202,
+      "grad_norm": 0.6789650292657143,
+      "learning_rate": 7.663931811073003e-06,
+      "loss": 0.2919,
+      "step": 650
+    },
+    {
+      "epoch": 1.2376425855513309,
+      "grad_norm": 0.7371899002650942,
+      "learning_rate": 7.631654388685496e-06,
+      "loss": 0.2894,
+      "step": 651
+    },
+    {
+      "epoch": 1.2395437262357414,
+      "grad_norm": 0.7411950998073552,
+      "learning_rate": 7.599403085635208e-06,
+      "loss": 0.2963,
+      "step": 652
+    },
+    {
+      "epoch": 1.241444866920152,
+      "grad_norm": 0.6843358181587386,
+      "learning_rate": 7.567178257606147e-06,
+      "loss": 0.2942,
+      "step": 653
+    },
+    {
+      "epoch": 1.2433460076045628,
+      "grad_norm": 0.6836482839157375,
+      "learning_rate": 7.534980259990341e-06,
+      "loss": 0.2838,
+      "step": 654
+    },
+    {
+      "epoch": 1.2452471482889733,
+      "grad_norm": 0.7044286897986738,
+      "learning_rate": 7.50280944788392e-06,
+      "loss": 0.2863,
+      "step": 655
+    },
+    {
+      "epoch": 1.247148288973384,
+      "grad_norm": 0.7200508714497933,
+      "learning_rate": 7.470666176083193e-06,
+      "loss": 0.2999,
+      "step": 656
+    },
+    {
+      "epoch": 1.2490494296577948,
+      "grad_norm": 0.7326826806293245,
+      "learning_rate": 7.438550799080746e-06,
+      "loss": 0.2864,
+      "step": 657
+    },
+    {
+      "epoch": 1.2509505703422052,
+      "grad_norm": 0.6929056989440776,
+      "learning_rate": 7.40646367106153e-06,
+      "loss": 0.2922,
+      "step": 658
+    },
+    {
+      "epoch": 1.252851711026616,
+      "grad_norm": 0.6754973509672555,
+      "learning_rate": 7.3744051458989395e-06,
+      "loss": 0.2868,
+      "step": 659
+    },
+    {
+      "epoch": 1.2547528517110267,
+      "grad_norm": 0.6897474799851169,
+      "learning_rate": 7.342375577150928e-06,
+      "loss": 0.2909,
+      "step": 660
+    },
+    {
+      "epoch": 1.2566539923954372,
+      "grad_norm": 0.7147415653676243,
+      "learning_rate": 7.310375318056107e-06,
+      "loss": 0.3053,
+      "step": 661
+    },
+    {
+      "epoch": 1.258555133079848,
+      "grad_norm": 0.7080936922840922,
+      "learning_rate": 7.278404721529843e-06,
+      "loss": 0.2968,
+      "step": 662
+    },
+    {
+      "epoch": 1.2604562737642586,
+      "grad_norm": 0.6755973914703102,
+      "learning_rate": 7.246464140160365e-06,
+      "loss": 0.2871,
+      "step": 663
+    },
+    {
+      "epoch": 1.2623574144486693,
+      "grad_norm": 0.75565239191557,
+      "learning_rate": 7.214553926204884e-06,
+      "loss": 0.2836,
+      "step": 664
+    },
+    {
+      "epoch": 1.2642585551330798,
+      "grad_norm": 0.7053307489953805,
+      "learning_rate": 7.182674431585703e-06,
+      "loss": 0.2928,
+      "step": 665
+    },
+    {
+      "epoch": 1.2661596958174905,
+      "grad_norm": 0.7672803521959862,
+      "learning_rate": 7.150826007886334e-06,
+      "loss": 0.2954,
+      "step": 666
+    },
+    {
+      "epoch": 1.2680608365019013,
+      "grad_norm": 0.7767735171999255,
+      "learning_rate": 7.119009006347625e-06,
+      "loss": 0.2814,
+      "step": 667
+    },
+    {
+      "epoch": 1.2699619771863118,
+      "grad_norm": 0.742984890915227,
+      "learning_rate": 7.087223777863883e-06,
+      "loss": 0.2922,
+      "step": 668
+    },
+    {
+      "epoch": 1.2718631178707225,
+      "grad_norm": 0.711514011396354,
+      "learning_rate": 7.055470672979003e-06,
+      "loss": 0.3058,
+      "step": 669
+    },
+    {
+      "epoch": 1.2737642585551332,
+      "grad_norm": 0.7325340265134281,
+      "learning_rate": 7.023750041882609e-06,
+      "loss": 0.2978,
+      "step": 670
+    },
+    {
+      "epoch": 1.2756653992395437,
+      "grad_norm": 0.7465630777746666,
+      "learning_rate": 6.992062234406185e-06,
+      "loss": 0.2933,
+      "step": 671
+    },
+    {
+      "epoch": 1.2775665399239544,
+      "grad_norm": 0.689387074966669,
+      "learning_rate": 6.960407600019217e-06,
+      "loss": 0.2834,
+      "step": 672
+    },
+    {
+      "epoch": 1.2794676806083651,
+      "grad_norm": 0.7366198182791822,
+      "learning_rate": 6.9287864878253475e-06,
+      "loss": 0.3013,
+      "step": 673
+    },
+    {
+      "epoch": 1.2813688212927756,
+      "grad_norm": 0.7136083886736968,
+      "learning_rate": 6.897199246558515e-06,
+      "loss": 0.2873,
+      "step": 674
+    },
+    {
+      "epoch": 1.2832699619771863,
+      "grad_norm": 0.7592020918059184,
+      "learning_rate": 6.865646224579108e-06,
+      "loss": 0.2865,
+      "step": 675
+    },
+    {
+      "epoch": 1.285171102661597,
+      "grad_norm": 0.7112598768252201,
+      "learning_rate": 6.834127769870134e-06,
+      "loss": 0.2846,
+      "step": 676
+    },
+    {
+      "epoch": 1.2870722433460076,
+      "grad_norm": 0.7584893612586059,
+      "learning_rate": 6.802644230033373e-06,
+      "loss": 0.3095,
+      "step": 677
+    },
+    {
+      "epoch": 1.2889733840304183,
+      "grad_norm": 0.7977539723711234,
+      "learning_rate": 6.771195952285541e-06,
+      "loss": 0.2873,
+      "step": 678
+    },
+    {
+      "epoch": 1.290874524714829,
+      "grad_norm": 0.7676232332340589,
+      "learning_rate": 6.739783283454469e-06,
+      "loss": 0.2966,
+      "step": 679
+    },
+    {
+      "epoch": 1.2927756653992395,
+      "grad_norm": 0.6893092059340077,
+      "learning_rate": 6.708406569975274e-06,
+      "loss": 0.2738,
+      "step": 680
+    },
+    {
+      "epoch": 1.2946768060836502,
+      "grad_norm": 0.7246494882694059,
+      "learning_rate": 6.6770661578865444e-06,
+      "loss": 0.2896,
+      "step": 681
+    },
+    {
+      "epoch": 1.296577946768061,
+      "grad_norm": 0.7073493095396471,
+      "learning_rate": 6.645762392826509e-06,
+      "loss": 0.3019,
+      "step": 682
+    },
+    {
+      "epoch": 1.2984790874524714,
+      "grad_norm": 0.7007386925363137,
+      "learning_rate": 6.614495620029238e-06,
+      "loss": 0.2922,
+      "step": 683
+    },
+    {
+      "epoch": 1.3003802281368821,
+      "grad_norm": 0.6917901925134944,
+      "learning_rate": 6.583266184320836e-06,
+      "loss": 0.2827,
+      "step": 684
+    },
+    {
+      "epoch": 1.3022813688212929,
+      "grad_norm": 0.7369661907640651,
+      "learning_rate": 6.552074430115624e-06,
+      "loss": 0.2981,
+      "step": 685
+    },
+    {
+      "epoch": 1.3041825095057034,
+      "grad_norm": 0.6922292999782869,
+      "learning_rate": 6.520920701412371e-06,
+      "loss": 0.2755,
+      "step": 686
+    },
+    {
+      "epoch": 1.306083650190114,
+      "grad_norm": 0.7056760086531454,
+      "learning_rate": 6.489805341790456e-06,
+      "loss": 0.2773,
+      "step": 687
+    },
+    {
+      "epoch": 1.3079847908745248,
+      "grad_norm": 0.7434266181863763,
+      "learning_rate": 6.458728694406124e-06,
+      "loss": 0.2986,
+      "step": 688
+    },
+    {
+      "epoch": 1.3098859315589353,
+      "grad_norm": 0.687168409457717,
+      "learning_rate": 6.427691101988673e-06,
+      "loss": 0.2927,
+      "step": 689
+    },
+    {
+      "epoch": 1.311787072243346,
+      "grad_norm": 0.7282233391544028,
+      "learning_rate": 6.396692906836686e-06,
+      "loss": 0.2913,
+      "step": 690
+    },
+    {
+      "epoch": 1.3136882129277567,
+      "grad_norm": 0.6961182692812412,
+      "learning_rate": 6.3657344508142495e-06,
+      "loss": 0.2816,
+      "step": 691
+    },
+    {
+      "epoch": 1.3155893536121672,
+      "grad_norm": 0.6760834185922154,
+      "learning_rate": 6.334816075347185e-06,
+      "loss": 0.2849,
+      "step": 692
+    },
+    {
+      "epoch": 1.317490494296578,
+      "grad_norm": 0.7275066890851415,
+      "learning_rate": 6.303938121419295e-06,
+      "loss": 0.2808,
+      "step": 693
+    },
+    {
+      "epoch": 1.3193916349809887,
+      "grad_norm": 0.7354558968528108,
+      "learning_rate": 6.273100929568579e-06,
+      "loss": 0.289,
+      "step": 694
+    },
+    {
+      "epoch": 1.3212927756653992,
+      "grad_norm": 0.6948206887971208,
+      "learning_rate": 6.242304839883502e-06,
+      "loss": 0.294,
+      "step": 695
+    },
+    {
+      "epoch": 1.3231939163498099,
+      "grad_norm": 0.7047087277063151,
+      "learning_rate": 6.211550191999223e-06,
+      "loss": 0.2932,
+      "step": 696
+    },
+    {
+      "epoch": 1.3250950570342206,
+      "grad_norm": 0.7558936018582135,
+      "learning_rate": 6.18083732509387e-06,
+      "loss": 0.3002,
+      "step": 697
+    },
+    {
+      "epoch": 1.326996197718631,
+      "grad_norm": 0.7273211841524094,
+      "learning_rate": 6.150166577884781e-06,
+      "loss": 0.286,
+      "step": 698
+    },
+    {
+      "epoch": 1.3288973384030418,
+      "grad_norm": 0.7205144056055376,
+      "learning_rate": 6.119538288624778e-06,
+      "loss": 0.2849,
+      "step": 699
+    },
+    {
+      "epoch": 1.3307984790874525,
+      "grad_norm": 0.7245637955873974,
+      "learning_rate": 6.088952795098442e-06,
+      "loss": 0.2871,
+      "step": 700
+    },
+    {
+      "epoch": 1.332699619771863,
+      "grad_norm": 0.7456434412563985,
+      "learning_rate": 6.058410434618367e-06,
+      "loss": 0.2865,
+      "step": 701
+    },
+    {
+      "epoch": 1.3346007604562737,
+      "grad_norm": 0.6755876877519845,
+      "learning_rate": 6.027911544021465e-06,
+      "loss": 0.2771,
+      "step": 702
+    },
+    {
+      "epoch": 1.3365019011406845,
+      "grad_norm": 0.683988241876236,
+      "learning_rate": 5.997456459665237e-06,
+      "loss": 0.2905,
+      "step": 703
+    },
+    {
+      "epoch": 1.338403041825095,
+      "grad_norm": 0.7099361523830153,
+      "learning_rate": 5.967045517424062e-06,
+      "loss": 0.2874,
+      "step": 704
+    },
+    {
+      "epoch": 1.3403041825095057,
+      "grad_norm": 0.7815235818349345,
+      "learning_rate": 5.936679052685505e-06,
+      "loss": 0.2866,
+      "step": 705
+    },
+    {
+      "epoch": 1.3422053231939164,
+      "grad_norm": 0.7752856826992063,
+      "learning_rate": 5.906357400346596e-06,
+      "loss": 0.297,
+      "step": 706
+    },
+    {
+      "epoch": 1.3441064638783269,
+      "grad_norm": 0.7234830739038847,
+      "learning_rate": 5.876080894810167e-06,
+      "loss": 0.2933,
+      "step": 707
+    },
+    {
+      "epoch": 1.3460076045627376,
+      "grad_norm": 0.9255515792091867,
+      "learning_rate": 5.845849869981137e-06,
+      "loss": 0.2933,
+      "step": 708
+    },
+    {
+      "epoch": 1.3479087452471483,
+      "grad_norm": 0.666046326625046,
+      "learning_rate": 5.815664659262845e-06,
+      "loss": 0.2904,
+      "step": 709
+    },
+    {
+      "epoch": 1.3498098859315588,
+      "grad_norm": 0.7364163343078383,
+      "learning_rate": 5.78552559555337e-06,
+      "loss": 0.2867,
+      "step": 710
+    },
+    {
+      "epoch": 1.3517110266159695,
+      "grad_norm": 0.6726645429514998,
+      "learning_rate": 5.755433011241851e-06,
+      "loss": 0.2762,
+      "step": 711
+    },
+    {
+      "epoch": 1.3536121673003803,
+      "grad_norm": 0.7059335896205419,
+      "learning_rate": 5.725387238204831e-06,
+      "loss": 0.2968,
+      "step": 712
+    },
+    {
+      "epoch": 1.3555133079847907,
+      "grad_norm": 0.705725744890944,
+      "learning_rate": 5.695388607802603e-06,
+      "loss": 0.3011,
+      "step": 713
+    },
+    {
+      "epoch": 1.3574144486692015,
+      "grad_norm": 0.726735540484456,
+      "learning_rate": 5.665437450875534e-06,
+      "loss": 0.287,
+      "step": 714
+    },
+    {
+      "epoch": 1.3593155893536122,
+      "grad_norm": 0.6840462972087888,
+      "learning_rate": 5.635534097740435e-06,
+      "loss": 0.2904,
+      "step": 715
+    },
+    {
+      "epoch": 1.3612167300380227,
+      "grad_norm": 0.7166382524728736,
+      "learning_rate": 5.605678878186911e-06,
+      "loss": 0.289,
+      "step": 716
+    },
+    {
+      "epoch": 1.3631178707224334,
+      "grad_norm": 0.6751991675844237,
+      "learning_rate": 5.575872121473722e-06,
+      "loss": 0.2777,
+      "step": 717
+    },
+    {
+      "epoch": 1.3650190114068441,
+      "grad_norm": 0.7585194497054609,
+      "learning_rate": 5.546114156325166e-06,
+      "loss": 0.2858,
+      "step": 718
+    },
+    {
+      "epoch": 1.3669201520912546,
+      "grad_norm": 0.6801020689899001,
+      "learning_rate": 5.516405310927431e-06,
+      "loss": 0.2811,
+      "step": 719
+    },
+    {
+      "epoch": 1.3688212927756653,
+      "grad_norm": 0.680530139917202,
+      "learning_rate": 5.4867459129249846e-06,
+      "loss": 0.284,
+      "step": 720
+    },
+    {
+      "epoch": 1.370722433460076,
+      "grad_norm": 0.7120069344474718,
+      "learning_rate": 5.4571362894169795e-06,
+      "loss": 0.2939,
+      "step": 721
+    },
+    {
+      "epoch": 1.3726235741444868,
+      "grad_norm": 0.7423374877845463,
+      "learning_rate": 5.427576766953615e-06,
+      "loss": 0.288,
+      "step": 722
+    },
+    {
+      "epoch": 1.3745247148288973,
+      "grad_norm": 0.7218822981199374,
+      "learning_rate": 5.398067671532554e-06,
+      "loss": 0.2901,
+      "step": 723
+    },
+    {
+      "epoch": 1.376425855513308,
+      "grad_norm": 0.6735131115089157,
+      "learning_rate": 5.368609328595323e-06,
+      "loss": 0.2828,
+      "step": 724
+    },
+    {
+      "epoch": 1.3783269961977187,
+      "grad_norm": 0.6879069283917294,
+      "learning_rate": 5.339202063023727e-06,
+      "loss": 0.2786,
+      "step": 725
+    },
+    {
+      "epoch": 1.3802281368821292,
+      "grad_norm": 0.6924030127198797,
+      "learning_rate": 5.309846199136258e-06,
+      "loss": 0.28,
+      "step": 726
+    },
+    {
+      "epoch": 1.38212927756654,
+      "grad_norm": 0.7095484592211323,
+      "learning_rate": 5.280542060684535e-06,
+      "loss": 0.2873,
+      "step": 727
+    },
+    {
+      "epoch": 1.3840304182509506,
+      "grad_norm": 0.7279589483349861,
+      "learning_rate": 5.2512899708497086e-06,
+      "loss": 0.2922,
+      "step": 728
+    },
+    {
+      "epoch": 1.3859315589353614,
+      "grad_norm": 0.6812029663780734,
+      "learning_rate": 5.222090252238916e-06,
+      "loss": 0.2757,
+      "step": 729
+    },
+    {
+      "epoch": 1.3878326996197718,
+      "grad_norm": 0.6910962379153492,
+      "learning_rate": 5.192943226881724e-06,
+      "loss": 0.2745,
+      "step": 730
+    },
+    {
+      "epoch": 1.3897338403041826,
+      "grad_norm": 0.7532695913142263,
+      "learning_rate": 5.163849216226562e-06,
+      "loss": 0.2953,
+      "step": 731
+    },
+    {
+      "epoch": 1.3916349809885933,
+      "grad_norm": 0.7092334688343299,
+      "learning_rate": 5.134808541137183e-06,
+      "loss": 0.2863,
+      "step": 732
+    },
+    {
+      "epoch": 1.3935361216730038,
+      "grad_norm": 0.7265062119412851,
+      "learning_rate": 5.105821521889147e-06,
+      "loss": 0.2991,
+      "step": 733
+    },
+    {
+      "epoch": 1.3954372623574145,
+      "grad_norm": 0.7138008594382279,
+      "learning_rate": 5.076888478166247e-06,
+      "loss": 0.2915,
+      "step": 734
+    },
+    {
+      "epoch": 1.3973384030418252,
+      "grad_norm": 0.7288896607533202,
+      "learning_rate": 5.048009729057012e-06,
+      "loss": 0.2996,
+      "step": 735
+    },
+    {
+      "epoch": 1.3992395437262357,
+      "grad_norm": 0.7475825325755282,
+      "learning_rate": 5.0191855930511946e-06,
+      "loss": 0.2983,
+      "step": 736
+    },
+    {
+      "epoch": 1.4011406844106464,
+      "grad_norm": 0.7141287075640784,
+      "learning_rate": 4.990416388036233e-06,
+      "loss": 0.2834,
+      "step": 737
+    },
+    {
+      "epoch": 1.4030418250950571,
+      "grad_norm": 0.7364723630267682,
+      "learning_rate": 4.961702431293759e-06,
+      "loss": 0.2985,
+      "step": 738
+    },
+    {
+      "epoch": 1.4049429657794676,
+      "grad_norm": 0.7264122514305142,
+      "learning_rate": 4.933044039496107e-06,
+      "loss": 0.2813,
+      "step": 739
+    },
+    {
+      "epoch": 1.4068441064638784,
+      "grad_norm": 0.673044585290331,
+      "learning_rate": 4.904441528702806e-06,
+      "loss": 0.2793,
+      "step": 740
+    },
+    {
+      "epoch": 1.408745247148289,
+      "grad_norm": 0.7388415667669372,
+      "learning_rate": 4.875895214357093e-06,
+      "loss": 0.2887,
+      "step": 741
+    },
+    {
+      "epoch": 1.4106463878326996,
+      "grad_norm": 0.7683401629002008,
+      "learning_rate": 4.847405411282462e-06,
+      "loss": 0.3025,
+      "step": 742
+    },
+    {
+      "epoch": 1.4125475285171103,
+      "grad_norm": 0.7170938336542718,
+      "learning_rate": 4.818972433679145e-06,
+      "loss": 0.285,
+      "step": 743
+    },
+    {
+      "epoch": 1.414448669201521,
+      "grad_norm": 0.7158403517137417,
+      "learning_rate": 4.790596595120699e-06,
+      "loss": 0.2801,
+      "step": 744
+    },
+    {
+      "epoch": 1.4163498098859315,
+      "grad_norm": 0.649447466407859,
+      "learning_rate": 4.762278208550505e-06,
+      "loss": 0.2689,
+      "step": 745
+    },
+    {
+      "epoch": 1.4182509505703422,
+      "grad_norm": 0.7082614566074693,
+      "learning_rate": 4.734017586278337e-06,
+      "loss": 0.2844,
+      "step": 746
+    },
+    {
+      "epoch": 1.420152091254753,
+      "grad_norm": 0.7085031748073999,
+      "learning_rate": 4.7058150399769245e-06,
+      "loss": 0.2906,
+      "step": 747
+    },
+    {
+      "epoch": 1.4220532319391634,
+      "grad_norm": 0.6732158714717617,
+      "learning_rate": 4.677670880678493e-06,
+      "loss": 0.2855,
+      "step": 748
+    },
+    {
+      "epoch": 1.4239543726235742,
+      "grad_norm": 0.6984936344041431,
+      "learning_rate": 4.649585418771348e-06,
+      "loss": 0.2796,
+      "step": 749
+    },
+    {
+      "epoch": 1.4258555133079849,
+      "grad_norm": 0.7338928129777204,
+      "learning_rate": 4.621558963996458e-06,
+      "loss": 0.2902,
+      "step": 750
+    },
+    {
+      "epoch": 1.4277566539923954,
+      "grad_norm": 0.6773644133879478,
+      "learning_rate": 4.593591825444028e-06,
+      "loss": 0.2784,
+      "step": 751
+    },
+    {
+      "epoch": 1.429657794676806,
+      "grad_norm": 0.7142651502999355,
+      "learning_rate": 4.565684311550077e-06,
+      "loss": 0.2789,
+      "step": 752
+    },
+    {
+      "epoch": 1.4315589353612168,
+      "grad_norm": 0.7078564802768612,
+      "learning_rate": 4.537836730093077e-06,
+      "loss": 0.2931,
+      "step": 753
+    },
+    {
+      "epoch": 1.4334600760456273,
+      "grad_norm": 0.6625652539973749,
+      "learning_rate": 4.510049388190518e-06,
+      "loss": 0.2896,
+      "step": 754
+    },
+    {
+      "epoch": 1.435361216730038,
+      "grad_norm": 0.6765165356102307,
+      "learning_rate": 4.482322592295541e-06,
+      "loss": 0.2835,
+      "step": 755
+    },
+    {
+      "epoch": 1.4372623574144487,
+      "grad_norm": 0.7166068338877377,
+      "learning_rate": 4.454656648193559e-06,
+      "loss": 0.2729,
+      "step": 756
+    },
+    {
+      "epoch": 1.4391634980988592,
+      "grad_norm": 0.6723192572008214,
+      "learning_rate": 4.427051860998877e-06,
+      "loss": 0.2803,
+      "step": 757
+    },
+    {
+      "epoch": 1.44106463878327,
+      "grad_norm": 0.6384592830220657,
+      "learning_rate": 4.399508535151321e-06,
+      "loss": 0.2701,
+      "step": 758
+    },
+    {
+      "epoch": 1.4429657794676807,
+      "grad_norm": 0.7055705261507379,
+      "learning_rate": 4.372026974412907e-06,
+      "loss": 0.284,
+      "step": 759
+    },
+    {
+      "epoch": 1.4448669201520912,
+      "grad_norm": 0.7358982346816677,
+      "learning_rate": 4.344607481864466e-06,
+      "loss": 0.2977,
+      "step": 760
+    },
+    {
+      "epoch": 1.446768060836502,
+      "grad_norm": 0.7124418132493656,
+      "learning_rate": 4.317250359902295e-06,
+      "loss": 0.2891,
+      "step": 761
+    },
+    {
+      "epoch": 1.4486692015209126,
+      "grad_norm": 0.7158845956833674,
+      "learning_rate": 4.2899559102348585e-06,
+      "loss": 0.2914,
+      "step": 762
+    },
+    {
+      "epoch": 1.450570342205323,
+      "grad_norm": 0.7182051849255634,
+      "learning_rate": 4.262724433879427e-06,
+      "loss": 0.2826,
+      "step": 763
+    },
+    {
+      "epoch": 1.4524714828897338,
+      "grad_norm": 0.7784690277833876,
+      "learning_rate": 4.235556231158765e-06,
+      "loss": 0.2919,
+      "step": 764
+    },
+    {
+      "epoch": 1.4543726235741445,
+      "grad_norm": 0.7023453496978478,
+      "learning_rate": 4.208451601697836e-06,
+      "loss": 0.2805,
+      "step": 765
+    },
+    {
+      "epoch": 1.456273764258555,
+      "grad_norm": 0.6788665299316395,
+      "learning_rate": 4.181410844420473e-06,
+      "loss": 0.29,
+      "step": 766
+    },
+    {
+      "epoch": 1.4581749049429658,
+      "grad_norm": 0.7817956027920046,
+      "learning_rate": 4.154434257546095e-06,
+      "loss": 0.2798,
+      "step": 767
+    },
+    {
+      "epoch": 1.4600760456273765,
+      "grad_norm": 0.7501150341927527,
+      "learning_rate": 4.127522138586424e-06,
+      "loss": 0.2769,
+      "step": 768
+    },
+    {
+      "epoch": 1.461977186311787,
+      "grad_norm": 0.6871700574697513,
+      "learning_rate": 4.10067478434219e-06,
+      "loss": 0.2792,
+      "step": 769
+    },
+    {
+      "epoch": 1.4638783269961977,
+      "grad_norm": 0.6919539570178334,
+      "learning_rate": 4.073892490899865e-06,
+      "loss": 0.2793,
+      "step": 770
+    },
+    {
+      "epoch": 1.4657794676806084,
+      "grad_norm": 0.6988102328053417,
+      "learning_rate": 4.047175553628397e-06,
+      "loss": 0.2804,
+      "step": 771
+    },
+    {
+      "epoch": 1.467680608365019,
+      "grad_norm": 0.713755478290317,
+      "learning_rate": 4.020524267175954e-06,
+      "loss": 0.2886,
+      "step": 772
+    },
+    {
+      "epoch": 1.4695817490494296,
+      "grad_norm": 0.6885537137147855,
+      "learning_rate": 3.993938925466674e-06,
+      "loss": 0.2752,
+      "step": 773
+    },
+    {
+      "epoch": 1.4714828897338403,
+      "grad_norm": 0.68133424767042,
+      "learning_rate": 3.96741982169742e-06,
+      "loss": 0.2868,
+      "step": 774
+    },
+    {
+      "epoch": 1.4733840304182508,
+      "grad_norm": 0.7283991398915899,
+      "learning_rate": 3.9409672483345465e-06,
+      "loss": 0.279,
+      "step": 775
+    },
+    {
+      "epoch": 1.4752851711026616,
+      "grad_norm": 0.7407407654354737,
+      "learning_rate": 3.914581497110684e-06,
+      "loss": 0.2806,
+      "step": 776
+    },
+    {
+      "epoch": 1.4771863117870723,
+      "grad_norm": 0.7227652333140067,
+      "learning_rate": 3.888262859021508e-06,
+      "loss": 0.2756,
+      "step": 777
+    },
+    {
+      "epoch": 1.4790874524714828,
+      "grad_norm": 0.7012898936512341,
+      "learning_rate": 3.862011624322534e-06,
+      "loss": 0.2871,
+      "step": 778
+    },
+    {
+      "epoch": 1.4809885931558935,
+      "grad_norm": 0.7217796654034508,
+      "learning_rate": 3.835828082525925e-06,
+      "loss": 0.2929,
+      "step": 779
+    },
+    {
+      "epoch": 1.4828897338403042,
+      "grad_norm": 0.7111580824215556,
+      "learning_rate": 3.8097125223972864e-06,
+      "loss": 0.2839,
+      "step": 780
+    },
+    {
+      "epoch": 1.4847908745247147,
+      "grad_norm": 0.7100534623335487,
+      "learning_rate": 3.7836652319524835e-06,
+      "loss": 0.2927,
+      "step": 781
+    },
+    {
+      "epoch": 1.4866920152091254,
+      "grad_norm": 0.7324444142897687,
+      "learning_rate": 3.7576864984544814e-06,
+      "loss": 0.294,
+      "step": 782
+    },
+    {
+      "epoch": 1.4885931558935361,
+      "grad_norm": 0.7158797850703152,
+      "learning_rate": 3.73177660841015e-06,
+      "loss": 0.3031,
+      "step": 783
+    },
+    {
+      "epoch": 1.4904942965779466,
+      "grad_norm": 0.702168210462647,
+      "learning_rate": 3.7059358475671225e-06,
+      "loss": 0.2787,
+      "step": 784
+    },
+    {
+      "epoch": 1.4923954372623573,
+      "grad_norm": 0.7088407589813892,
+      "learning_rate": 3.680164500910646e-06,
+      "loss": 0.2786,
+      "step": 785
+    },
+    {
+      "epoch": 1.494296577946768,
+      "grad_norm": 0.8644030284119277,
+      "learning_rate": 3.654462852660423e-06,
+      "loss": 0.2823,
+      "step": 786
+    },
+    {
+      "epoch": 1.4961977186311788,
+      "grad_norm": 0.6952639368659915,
+      "learning_rate": 3.6288311862674885e-06,
+      "loss": 0.2823,
+      "step": 787
+    },
+    {
+      "epoch": 1.4980988593155893,
+      "grad_norm": 0.7096763794742813,
+      "learning_rate": 3.6032697844110896e-06,
+      "loss": 0.2934,
+      "step": 788
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6695426045719942,
+      "learning_rate": 3.5777789289955454e-06,
+      "loss": 0.2784,
+      "step": 789
+    },
+    {
+      "epoch": 1.5019011406844105,
+      "grad_norm": 0.6524566363145925,
+      "learning_rate": 3.5523589011471592e-06,
+      "loss": 0.2699,
+      "step": 790
+    },
+    {
+      "epoch": 1.5038022813688214,
+      "grad_norm": 0.7532424129615138,
+      "learning_rate": 3.527009981211119e-06,
+      "loss": 0.2929,
+      "step": 791
+    },
+    {
+      "epoch": 1.505703422053232,
+      "grad_norm": 0.681946779446385,
+      "learning_rate": 3.5017324487483873e-06,
+      "loss": 0.2805,
+      "step": 792
+    },
+    {
+      "epoch": 1.5076045627376424,
+      "grad_norm": 0.6903860675980712,
+      "learning_rate": 3.47652658253263e-06,
+      "loss": 0.2846,
+      "step": 793
+    },
+    {
+      "epoch": 1.5095057034220534,
+      "grad_norm": 0.7017392608830051,
+      "learning_rate": 3.4513926605471504e-06,
+      "loss": 0.2832,
+      "step": 794
+    },
+    {
+      "epoch": 1.5114068441064639,
+      "grad_norm": 0.6995603939424366,
+      "learning_rate": 3.4263309599818017e-06,
+      "loss": 0.2776,
+      "step": 795
+    },
+    {
+      "epoch": 1.5133079847908744,
+      "grad_norm": 0.7174754922676291,
+      "learning_rate": 3.4013417572299446e-06,
+      "loss": 0.2759,
+      "step": 796
+    },
+    {
+      "epoch": 1.5152091254752853,
+      "grad_norm": 0.7064539732128455,
+      "learning_rate": 3.37642532788541e-06,
+      "loss": 0.2713,
+      "step": 797
+    },
+    {
+      "epoch": 1.5171102661596958,
+      "grad_norm": 0.7418356973993784,
+      "learning_rate": 3.3515819467394184e-06,
+      "loss": 0.2945,
+      "step": 798
+    },
+    {
+      "epoch": 1.5190114068441065,
+      "grad_norm": 0.6988116146786837,
+      "learning_rate": 3.326811887777607e-06,
+      "loss": 0.295,
+      "step": 799
+    },
+    {
+      "epoch": 1.5209125475285172,
+      "grad_norm": 0.7017162668187764,
+      "learning_rate": 3.3021154241769606e-06,
+      "loss": 0.2894,
+      "step": 800
+    },
+    {
+      "epoch": 1.5228136882129277,
+      "grad_norm": 0.7237575766380469,
+      "learning_rate": 3.2774928283028153e-06,
+      "loss": 0.2843,
+      "step": 801
+    },
+    {
+      "epoch": 1.5247148288973384,
+      "grad_norm": 0.7235878172992064,
+      "learning_rate": 3.2529443717058693e-06,
+      "loss": 0.2897,
+      "step": 802
+    },
+    {
+      "epoch": 1.5266159695817492,
+      "grad_norm": 0.692253906482968,
+      "learning_rate": 3.228470325119164e-06,
+      "loss": 0.2806,
+      "step": 803
+    },
+    {
+      "epoch": 1.5285171102661597,
+      "grad_norm": 0.6956965527154768,
+      "learning_rate": 3.20407095845511e-06,
+      "loss": 0.2753,
+      "step": 804
+    },
+    {
+      "epoch": 1.5304182509505704,
+      "grad_norm": 0.7506609358953361,
+      "learning_rate": 3.179746540802506e-06,
+      "loss": 0.2887,
+      "step": 805
+    },
+    {
+      "epoch": 1.532319391634981,
+      "grad_norm": 0.6857963089357372,
+      "learning_rate": 3.155497340423588e-06,
+      "loss": 0.2762,
+      "step": 806
+    },
+    {
+      "epoch": 1.5342205323193916,
+      "grad_norm": 0.7186855429598759,
+      "learning_rate": 3.1313236247510414e-06,
+      "loss": 0.2919,
+      "step": 807
+    },
+    {
+      "epoch": 1.5361216730038023,
+      "grad_norm": 0.7857081103211361,
+      "learning_rate": 3.107225660385077e-06,
+      "loss": 0.3134,
+      "step": 808
+    },
+    {
+      "epoch": 1.538022813688213,
+      "grad_norm": 0.7220170069663582,
+      "learning_rate": 3.0832037130904748e-06,
+      "loss": 0.2913,
+      "step": 809
+    },
+    {
+      "epoch": 1.5399239543726235,
+      "grad_norm": 0.7613384595428253,
+      "learning_rate": 3.0592580477936606e-06,
+      "loss": 0.2808,
+      "step": 810
+    },
+    {
+      "epoch": 1.5418250950570342,
+      "grad_norm": 0.7333226548617051,
+      "learning_rate": 3.035388928579792e-06,
+      "loss": 0.2873,
+      "step": 811
+    },
+    {
+      "epoch": 1.543726235741445,
+      "grad_norm": 0.692625551876295,
+      "learning_rate": 3.011596618689825e-06,
+      "loss": 0.2841,
+      "step": 812
+    },
+    {
+      "epoch": 1.5456273764258555,
+      "grad_norm": 0.696623210581126,
+      "learning_rate": 2.9878813805176252e-06,
+      "loss": 0.2696,
+      "step": 813
+    },
+    {
+      "epoch": 1.5475285171102662,
+      "grad_norm": 0.7282669107396988,
+      "learning_rate": 2.9642434756070793e-06,
+      "loss": 0.2771,
+      "step": 814
+    },
+    {
+      "epoch": 1.549429657794677,
+      "grad_norm": 0.6877414577334098,
+      "learning_rate": 2.940683164649194e-06,
+      "loss": 0.2836,
+      "step": 815
+    },
+    {
+      "epoch": 1.5513307984790874,
+      "grad_norm": 0.6952013837873315,
+      "learning_rate": 2.9172007074792342e-06,
+      "loss": 0.2754,
+      "step": 816
+    },
+    {
+      "epoch": 1.553231939163498,
+      "grad_norm": 0.6787420113133055,
+      "learning_rate": 2.8937963630738517e-06,
+      "loss": 0.2625,
+      "step": 817
+    },
+    {
+      "epoch": 1.5551330798479088,
+      "grad_norm": 0.6922945134021774,
+      "learning_rate": 2.87047038954823e-06,
+      "loss": 0.2911,
+      "step": 818
+    },
+    {
+      "epoch": 1.5570342205323193,
+      "grad_norm": 0.8402896970167123,
+      "learning_rate": 2.8472230441532365e-06,
+      "loss": 0.2703,
+      "step": 819
+    },
+    {
+      "epoch": 1.55893536121673,
+      "grad_norm": 0.6678746930340207,
+      "learning_rate": 2.8240545832725963e-06,
+      "loss": 0.2695,
+      "step": 820
+    },
+    {
+      "epoch": 1.5608365019011408,
+      "grad_norm": 0.705078676312885,
+      "learning_rate": 2.8009652624200436e-06,
+      "loss": 0.2825,
+      "step": 821
+    },
+    {
+      "epoch": 1.5627376425855513,
+      "grad_norm": 0.6677130225262102,
+      "learning_rate": 2.7779553362365184e-06,
+      "loss": 0.2613,
+      "step": 822
+    },
+    {
+      "epoch": 1.564638783269962,
+      "grad_norm": 0.7412097663609457,
+      "learning_rate": 2.755025058487364e-06,
+      "loss": 0.2884,
+      "step": 823
+    },
+    {
+      "epoch": 1.5665399239543727,
+      "grad_norm": 0.7406514147916663,
+      "learning_rate": 2.7321746820595084e-06,
+      "loss": 0.2843,
+      "step": 824
+    },
+    {
+      "epoch": 1.5684410646387832,
+      "grad_norm": 0.7676929694787699,
+      "learning_rate": 2.709404458958693e-06,
+      "loss": 0.3027,
+      "step": 825
+    },
+    {
+      "epoch": 1.570342205323194,
+      "grad_norm": 0.7081642929814832,
+      "learning_rate": 2.6867146403066833e-06,
+      "loss": 0.2777,
+      "step": 826
+    },
+    {
+      "epoch": 1.5722433460076046,
+      "grad_norm": 0.6897624695748474,
+      "learning_rate": 2.6641054763385044e-06,
+      "loss": 0.2766,
+      "step": 827
+    },
+    {
+      "epoch": 1.5741444866920151,
+      "grad_norm": 0.7230487976603425,
+      "learning_rate": 2.6415772163996845e-06,
+      "loss": 0.2856,
+      "step": 828
+    },
+    {
+      "epoch": 1.5760456273764258,
+      "grad_norm": 0.8022793519023697,
+      "learning_rate": 2.619130108943494e-06,
+      "loss": 0.3006,
+      "step": 829
+    },
+    {
+      "epoch": 1.5779467680608366,
+      "grad_norm": 0.6924872584474147,
+      "learning_rate": 2.5967644015282146e-06,
+      "loss": 0.2647,
+      "step": 830
+    },
+    {
+      "epoch": 1.579847908745247,
+      "grad_norm": 0.7422641131977431,
+      "learning_rate": 2.5744803408144026e-06,
+      "loss": 0.2843,
+      "step": 831
+    },
+    {
+      "epoch": 1.5817490494296578,
+      "grad_norm": 0.7068176008466083,
+      "learning_rate": 2.5522781725621814e-06,
+      "loss": 0.2945,
+      "step": 832
+    },
+    {
+      "epoch": 1.5836501901140685,
+      "grad_norm": 0.685027307753865,
+      "learning_rate": 2.530158141628515e-06,
+      "loss": 0.2746,
+      "step": 833
+    },
+    {
+      "epoch": 1.585551330798479,
+      "grad_norm": 0.6945132808889379,
+      "learning_rate": 2.508120491964512e-06,
+      "loss": 0.285,
+      "step": 834
+    },
+    {
+      "epoch": 1.5874524714828897,
+      "grad_norm": 0.6941931524929925,
+      "learning_rate": 2.486165466612751e-06,
+      "loss": 0.2882,
+      "step": 835
+    },
+    {
+      "epoch": 1.5893536121673004,
+      "grad_norm": 0.6868828496922105,
+      "learning_rate": 2.464293307704566e-06,
+      "loss": 0.2757,
+      "step": 836
+    },
+    {
+      "epoch": 1.591254752851711,
+      "grad_norm": 0.6880910265615507,
+      "learning_rate": 2.4425042564574186e-06,
+      "loss": 0.2762,
+      "step": 837
+    },
+    {
+      "epoch": 1.5931558935361216,
+      "grad_norm": 0.6676074344861813,
+      "learning_rate": 2.4207985531722034e-06,
+      "loss": 0.276,
+      "step": 838
+    },
+    {
+      "epoch": 1.5950570342205324,
+      "grad_norm": 0.7059312393161081,
+      "learning_rate": 2.3991764372306113e-06,
+      "loss": 0.2816,
+      "step": 839
+    },
+    {
+      "epoch": 1.5969581749049429,
+      "grad_norm": 0.630962071331813,
+      "learning_rate": 2.377638147092497e-06,
+      "loss": 0.2641,
+      "step": 840
+    },
+    {
+      "epoch": 1.5988593155893536,
+      "grad_norm": 0.7035563239120735,
+      "learning_rate": 2.3561839202932344e-06,
+      "loss": 0.277,
+      "step": 841
+    },
+    {
+      "epoch": 1.6007604562737643,
+      "grad_norm": 0.7083571423625011,
+      "learning_rate": 2.3348139934411008e-06,
+      "loss": 0.2835,
+      "step": 842
+    },
+    {
+      "epoch": 1.6026615969581748,
+      "grad_norm": 0.6631571631543133,
+      "learning_rate": 2.3135286022146785e-06,
+      "loss": 0.2815,
+      "step": 843
+    },
+    {
+      "epoch": 1.6045627376425855,
+      "grad_norm": 0.6388015083937685,
+      "learning_rate": 2.292327981360245e-06,
+      "loss": 0.2573,
+      "step": 844
+    },
+    {
+      "epoch": 1.6064638783269962,
+      "grad_norm": 0.715955776983333,
+      "learning_rate": 2.271212364689176e-06,
+      "loss": 0.2774,
+      "step": 845
+    },
+    {
+      "epoch": 1.6083650190114067,
+      "grad_norm": 0.711586708269853,
+      "learning_rate": 2.2501819850753925e-06,
+      "loss": 0.2885,
+      "step": 846
+    },
+    {
+      "epoch": 1.6102661596958177,
+      "grad_norm": 0.7393571861519403,
+      "learning_rate": 2.229237074452768e-06,
+      "loss": 0.2938,
+      "step": 847
+    },
+    {
+      "epoch": 1.6121673003802282,
+      "grad_norm": 0.6554387256904496,
+      "learning_rate": 2.2083778638125796e-06,
+      "loss": 0.274,
+      "step": 848
+    },
+    {
+      "epoch": 1.6140684410646386,
+      "grad_norm": 0.7014285964200045,
+      "learning_rate": 2.1876045832009694e-06,
+      "loss": 0.27,
+      "step": 849
+    },
+    {
+      "epoch": 1.6159695817490496,
+      "grad_norm": 0.6924990428725116,
+      "learning_rate": 2.16691746171639e-06,
+      "loss": 0.2904,
+      "step": 850
+    },
+    {
+      "epoch": 1.61787072243346,
+      "grad_norm": 0.6916299465747496,
+      "learning_rate": 2.1463167275070863e-06,
+      "loss": 0.2791,
+      "step": 851
+    },
+    {
+      "epoch": 1.6197718631178706,
+      "grad_norm": 0.6887309552468869,
+      "learning_rate": 2.125802607768588e-06,
+      "loss": 0.2705,
+      "step": 852
+    },
+    {
+      "epoch": 1.6216730038022815,
+      "grad_norm": 0.706954336046035,
+      "learning_rate": 2.1053753287411895e-06,
+      "loss": 0.2775,
+      "step": 853
+    },
+    {
+      "epoch": 1.623574144486692,
+      "grad_norm": 0.698665188650369,
+      "learning_rate": 2.08503511570746e-06,
+      "loss": 0.2847,
+      "step": 854
+    },
+    {
+      "epoch": 1.6254752851711025,
+      "grad_norm": 0.6584186947310746,
+      "learning_rate": 2.064782192989765e-06,
+      "loss": 0.2686,
+      "step": 855
+    },
+    {
+      "epoch": 1.6273764258555135,
+      "grad_norm": 0.6969233044708365,
+      "learning_rate": 2.0446167839477815e-06,
+      "loss": 0.2719,
+      "step": 856
+    },
+    {
+      "epoch": 1.629277566539924,
+      "grad_norm": 0.7144170802033729,
+      "learning_rate": 2.0245391109760437e-06,
+      "loss": 0.285,
+      "step": 857
+    },
+    {
+      "epoch": 1.6311787072243344,
+      "grad_norm": 0.6953686650834182,
+      "learning_rate": 2.0045493955014915e-06,
+      "loss": 0.2679,
+      "step": 858
+    },
+    {
+      "epoch": 1.6330798479087454,
+      "grad_norm": 0.6990107420847762,
+      "learning_rate": 1.984647857981017e-06,
+      "loss": 0.2852,
+      "step": 859
+    },
+    {
+      "epoch": 1.6349809885931559,
+      "grad_norm": 0.7331158331376239,
+      "learning_rate": 1.96483471789904e-06,
+      "loss": 0.2902,
+      "step": 860
+    },
+    {
+      "epoch": 1.6368821292775664,
+      "grad_norm": 0.7131005917254934,
+      "learning_rate": 1.9451101937650963e-06,
+      "loss": 0.2816,
+      "step": 861
+    },
+    {
+      "epoch": 1.6387832699619773,
+      "grad_norm": 0.7251317144048253,
+      "learning_rate": 1.925474503111412e-06,
+      "loss": 0.2874,
+      "step": 862
+    },
+    {
+      "epoch": 1.6406844106463878,
+      "grad_norm": 0.7088245879322422,
+      "learning_rate": 1.905927862490512e-06,
+      "loss": 0.2954,
+      "step": 863
+    },
+    {
+      "epoch": 1.6425855513307985,
+      "grad_norm": 0.701752862243773,
+      "learning_rate": 1.8864704874728346e-06,
+      "loss": 0.2734,
+      "step": 864
+    },
+    {
+      "epoch": 1.6444866920152093,
+      "grad_norm": 0.7362344870791694,
+      "learning_rate": 1.8671025926443464e-06,
+      "loss": 0.2863,
+      "step": 865
+    },
+    {
+      "epoch": 1.6463878326996197,
+      "grad_norm": 0.7021293700111677,
+      "learning_rate": 1.8478243916041882e-06,
+      "loss": 0.2716,
+      "step": 866
+    },
+    {
+      "epoch": 1.6482889733840305,
+      "grad_norm": 0.7356801901576552,
+      "learning_rate": 1.828636096962304e-06,
+      "loss": 0.2746,
+      "step": 867
+    },
+    {
+      "epoch": 1.6501901140684412,
+      "grad_norm": 0.6870830602302619,
+      "learning_rate": 1.8095379203371044e-06,
+      "loss": 0.2822,
+      "step": 868
+    },
+    {
+      "epoch": 1.6520912547528517,
+      "grad_norm": 0.6979492816520024,
+      "learning_rate": 1.7905300723531393e-06,
+      "loss": 0.2754,
+      "step": 869
+    },
+    {
+      "epoch": 1.6539923954372624,
+      "grad_norm": 0.7065837476677297,
+      "learning_rate": 1.771612762638758e-06,
+      "loss": 0.2816,
+      "step": 870
+    },
+    {
+      "epoch": 1.6558935361216731,
+      "grad_norm": 0.7444380019012486,
+      "learning_rate": 1.7527861998238094e-06,
+      "loss": 0.2795,
+      "step": 871
+    },
+    {
+      "epoch": 1.6577946768060836,
+      "grad_norm": 0.686869282932523,
+      "learning_rate": 1.7340505915373495e-06,
+      "loss": 0.2693,
+      "step": 872
+    },
+    {
+      "epoch": 1.6596958174904943,
+      "grad_norm": 0.7112165686359406,
+      "learning_rate": 1.7154061444053239e-06,
+      "loss": 0.2863,
+      "step": 873
+    },
+    {
+      "epoch": 1.661596958174905,
+      "grad_norm": 0.7040250455743816,
+      "learning_rate": 1.6968530640483126e-06,
+      "loss": 0.2837,
+      "step": 874
+    },
+    {
+      "epoch": 1.6634980988593155,
+      "grad_norm": 0.7086799345623529,
+      "learning_rate": 1.6783915550792652e-06,
+      "loss": 0.276,
+      "step": 875
+    },
+    {
+      "epoch": 1.6653992395437263,
+      "grad_norm": 0.7269400961962234,
+      "learning_rate": 1.660021821101222e-06,
+      "loss": 0.2903,
+      "step": 876
+    },
+    {
+      "epoch": 1.667300380228137,
+      "grad_norm": 0.6533572900538692,
+      "learning_rate": 1.6417440647050853e-06,
+      "loss": 0.2695,
+      "step": 877
+    },
+    {
+      "epoch": 1.6692015209125475,
+      "grad_norm": 0.7040682654486197,
+      "learning_rate": 1.6235584874673848e-06,
+      "loss": 0.2697,
+      "step": 878
+    },
+    {
+      "epoch": 1.6711026615969582,
+      "grad_norm": 0.6967149920205667,
+      "learning_rate": 1.6054652899480472e-06,
+      "loss": 0.2625,
+      "step": 879
+    },
+    {
+      "epoch": 1.673003802281369,
+      "grad_norm": 0.6683823969010846,
+      "learning_rate": 1.587464671688187e-06,
+      "loss": 0.2588,
+      "step": 880
+    },
+    {
+      "epoch": 1.6749049429657794,
+      "grad_norm": 0.6988335123776637,
+      "learning_rate": 1.5695568312079156e-06,
+      "loss": 0.2752,
+      "step": 881
+    },
+    {
+      "epoch": 1.6768060836501901,
+      "grad_norm": 0.7159400201149252,
+      "learning_rate": 1.5517419660041277e-06,
+      "loss": 0.2655,
+      "step": 882
+    },
+    {
+      "epoch": 1.6787072243346008,
+      "grad_norm": 0.7191911719471372,
+      "learning_rate": 1.534020272548349e-06,
+      "loss": 0.2876,
+      "step": 883
+    },
+    {
+      "epoch": 1.6806083650190113,
+      "grad_norm": 0.7199125152529207,
+      "learning_rate": 1.5163919462845622e-06,
+      "loss": 0.2826,
+      "step": 884
+    },
+    {
+      "epoch": 1.682509505703422,
+      "grad_norm": 0.7246025443648172,
+      "learning_rate": 1.4988571816270402e-06,
+      "loss": 0.2934,
+      "step": 885
+    },
+    {
+      "epoch": 1.6844106463878328,
+      "grad_norm": 0.7370028183061983,
+      "learning_rate": 1.4814161719582132e-06,
+      "loss": 0.2798,
+      "step": 886
+    },
+    {
+      "epoch": 1.6863117870722433,
+      "grad_norm": 0.7266892496508501,
+      "learning_rate": 1.4640691096265358e-06,
+      "loss": 0.277,
+      "step": 887
+    },
+    {
+      "epoch": 1.688212927756654,
+      "grad_norm": 0.6873680993908002,
+      "learning_rate": 1.4468161859443609e-06,
+      "loss": 0.2739,
+      "step": 888
+    },
+    {
+      "epoch": 1.6901140684410647,
+      "grad_norm": 0.6703885359620109,
+      "learning_rate": 1.4296575911858268e-06,
+      "loss": 0.2592,
+      "step": 889
+    },
+    {
+      "epoch": 1.6920152091254752,
+      "grad_norm": 0.6922176530568109,
+      "learning_rate": 1.412593514584777e-06,
+      "loss": 0.2772,
+      "step": 890
+    },
+    {
+      "epoch": 1.693916349809886,
+      "grad_norm": 0.7288533894027804,
+      "learning_rate": 1.3956241443326423e-06,
+      "loss": 0.2781,
+      "step": 891
+    },
+    {
+      "epoch": 1.6958174904942966,
+      "grad_norm": 0.7254147952557574,
+      "learning_rate": 1.378749667576399e-06,
+      "loss": 0.2868,
+      "step": 892
+    },
+    {
+      "epoch": 1.6977186311787071,
+      "grad_norm": 0.678166963002887,
+      "learning_rate": 1.3619702704164783e-06,
+      "loss": 0.2708,
+      "step": 893
+    },
+    {
+      "epoch": 1.6996197718631179,
+      "grad_norm": 0.7245499335550314,
+      "learning_rate": 1.3452861379047289e-06,
+      "loss": 0.2801,
+      "step": 894
+    },
+    {
+      "epoch": 1.7015209125475286,
+      "grad_norm": 0.7022238999485703,
+      "learning_rate": 1.3286974540423747e-06,
+      "loss": 0.2758,
+      "step": 895
+    },
+    {
+      "epoch": 1.703422053231939,
+      "grad_norm": 0.7200323997934382,
+      "learning_rate": 1.3122044017779768e-06,
+      "loss": 0.2873,
+      "step": 896
+    },
+    {
+      "epoch": 1.7053231939163498,
+      "grad_norm": 0.6857166277134318,
+      "learning_rate": 1.2958071630054214e-06,
+      "loss": 0.2732,
+      "step": 897
+    },
+    {
+      "epoch": 1.7072243346007605,
+      "grad_norm": 0.7163910985950618,
+      "learning_rate": 1.279505918561923e-06,
+      "loss": 0.2916,
+      "step": 898
+    },
+    {
+      "epoch": 1.709125475285171,
+      "grad_norm": 0.7542787111966438,
+      "learning_rate": 1.2633008482260146e-06,
+      "loss": 0.2842,
+      "step": 899
+    },
+    {
+      "epoch": 1.7110266159695817,
+      "grad_norm": 0.726413975854438,
+      "learning_rate": 1.2471921307155655e-06,
+      "loss": 0.2761,
+      "step": 900
+    },
+    {
+      "epoch": 1.7129277566539924,
+      "grad_norm": 0.6987579133055972,
+      "learning_rate": 1.2311799436858275e-06,
+      "loss": 0.2799,
+      "step": 901
+    },
+    {
+      "epoch": 1.714828897338403,
+      "grad_norm": 0.6885924970107223,
+      "learning_rate": 1.2152644637274603e-06,
+      "loss": 0.2684,
+      "step": 902
+    },
+    {
+      "epoch": 1.7167300380228137,
+      "grad_norm": 0.6988132337614955,
+      "learning_rate": 1.1994458663645836e-06,
+      "loss": 0.2759,
+      "step": 903
+    },
+    {
+      "epoch": 1.7186311787072244,
+      "grad_norm": 0.7309714586927603,
+      "learning_rate": 1.1837243260528542e-06,
+      "loss": 0.2751,
+      "step": 904
+    },
+    {
+      "epoch": 1.7205323193916349,
+      "grad_norm": 0.76461243677071,
+      "learning_rate": 1.168100016177528e-06,
+      "loss": 0.2694,
+      "step": 905
+    },
+    {
+      "epoch": 1.7224334600760456,
+      "grad_norm": 0.7742088444378231,
+      "learning_rate": 1.1525731090515536e-06,
+      "loss": 0.2799,
+      "step": 906
+    },
+    {
+      "epoch": 1.7243346007604563,
+      "grad_norm": 0.7322199982287021,
+      "learning_rate": 1.137143775913675e-06,
+      "loss": 0.2767,
+      "step": 907
+    },
+    {
+      "epoch": 1.7262357414448668,
+      "grad_norm": 0.6860769792038816,
+      "learning_rate": 1.1218121869265365e-06,
+      "loss": 0.2705,
+      "step": 908
+    },
+    {
+      "epoch": 1.7281368821292775,
+      "grad_norm": 0.6448650288419475,
+      "learning_rate": 1.1065785111748117e-06,
+      "loss": 0.2606,
+      "step": 909
+    },
+    {
+      "epoch": 1.7300380228136882,
+      "grad_norm": 0.6707110996002345,
+      "learning_rate": 1.0914429166633355e-06,
+      "loss": 0.2754,
+      "step": 910
+    },
+    {
+      "epoch": 1.7319391634980987,
+      "grad_norm": 0.722794817855145,
+      "learning_rate": 1.076405570315252e-06,
+      "loss": 0.2795,
+      "step": 911
+    },
+    {
+      "epoch": 1.7338403041825095,
+      "grad_norm": 0.709790203721322,
+      "learning_rate": 1.0614666379701732e-06,
+      "loss": 0.2777,
+      "step": 912
+    },
+    {
+      "epoch": 1.7357414448669202,
+      "grad_norm": 0.7314261458203519,
+      "learning_rate": 1.046626284382356e-06,
+      "loss": 0.2755,
+      "step": 913
+    },
+    {
+      "epoch": 1.7376425855513307,
+      "grad_norm": 0.7144173568357892,
+      "learning_rate": 1.0318846732188737e-06,
+      "loss": 0.2741,
+      "step": 914
+    },
+    {
+      "epoch": 1.7395437262357416,
+      "grad_norm": 0.7091709277961338,
+      "learning_rate": 1.017241967057816e-06,
+      "loss": 0.2761,
+      "step": 915
+    },
+    {
+      "epoch": 1.741444866920152,
+      "grad_norm": 0.6816043757736597,
+      "learning_rate": 1.0026983273865055e-06,
+      "loss": 0.274,
+      "step": 916
+    },
+    {
+      "epoch": 1.7433460076045626,
+      "grad_norm": 0.6870586809823216,
+      "learning_rate": 9.882539145997027e-07,
+      "loss": 0.2749,
+      "step": 917
+    },
+    {
+      "epoch": 1.7452471482889735,
+      "grad_norm": 0.7063464281048318,
+      "learning_rate": 9.739088879978409e-07,
+      "loss": 0.2775,
+      "step": 918
+    },
+    {
+      "epoch": 1.747148288973384,
+      "grad_norm": 0.6836699610582385,
+      "learning_rate": 9.59663405785277e-07,
+      "loss": 0.2602,
+      "step": 919
+    },
+    {
+      "epoch": 1.7490494296577945,
+      "grad_norm": 0.7176367294846602,
+      "learning_rate": 9.455176250685338e-07,
+      "loss": 0.2804,
+      "step": 920
+    },
+    {
+      "epoch": 1.7509505703422055,
+      "grad_norm": 0.6884051954183766,
+      "learning_rate": 9.314717018545838e-07,
+      "loss": 0.2698,
+      "step": 921
+    },
+    {
+      "epoch": 1.752851711026616,
+      "grad_norm": 0.6867023937088869,
+      "learning_rate": 9.17525791049112e-07,
+      "loss": 0.2575,
+      "step": 922
+    },
+    {
+      "epoch": 1.7547528517110265,
+      "grad_norm": 0.6955579788978207,
+      "learning_rate": 9.036800464548157e-07,
+      "loss": 0.2624,
+      "step": 923
+    },
+    {
+      "epoch": 1.7566539923954374,
+      "grad_norm": 0.7355570512216763,
+      "learning_rate": 8.899346207697135e-07,
+      "loss": 0.2762,
+      "step": 924
+    },
+    {
+      "epoch": 1.758555133079848,
+      "grad_norm": 0.7214698905495001,
+      "learning_rate": 8.762896655854481e-07,
+      "loss": 0.2868,
+      "step": 925
+    },
+    {
+      "epoch": 1.7604562737642584,
+      "grad_norm": 0.711700910954847,
+      "learning_rate": 8.627453313856249e-07,
+      "loss": 0.2709,
+      "step": 926
+    },
+    {
+      "epoch": 1.7623574144486693,
+      "grad_norm": 0.6929007543937056,
+      "learning_rate": 8.493017675441495e-07,
+      "loss": 0.2775,
+      "step": 927
+    },
+    {
+      "epoch": 1.7642585551330798,
+      "grad_norm": 0.723721391366007,
+      "learning_rate": 8.359591223235785e-07,
+      "loss": 0.2837,
+      "step": 928
+    },
+    {
+      "epoch": 1.7661596958174905,
+      "grad_norm": 0.703758893010897,
+      "learning_rate": 8.227175428734868e-07,
+      "loss": 0.2791,
+      "step": 929
+    },
+    {
+      "epoch": 1.7680608365019013,
+      "grad_norm": 0.7001962710879965,
+      "learning_rate": 8.095771752288451e-07,
+      "loss": 0.2671,
+      "step": 930
+    },
+    {
+      "epoch": 1.7699619771863118,
+      "grad_norm": 0.781253006375267,
+      "learning_rate": 7.965381643084069e-07,
+      "loss": 0.2694,
+      "step": 931
+    },
+    {
+      "epoch": 1.7718631178707225,
+      "grad_norm": 0.6852863424227412,
+      "learning_rate": 7.83600653913108e-07,
+      "loss": 0.2851,
+      "step": 932
+    },
+    {
+      "epoch": 1.7737642585551332,
+      "grad_norm": 0.6758302354581138,
+      "learning_rate": 7.707647867244927e-07,
+      "loss": 0.2753,
+      "step": 933
+    },
+    {
+      "epoch": 1.7756653992395437,
+      "grad_norm": 0.7442328586008942,
+      "learning_rate": 7.580307043031232e-07,
+      "loss": 0.2834,
+      "step": 934
+    },
+    {
+      "epoch": 1.7775665399239544,
+      "grad_norm": 0.7032226140198627,
+      "learning_rate": 7.453985470870284e-07,
+      "loss": 0.2764,
+      "step": 935
+    },
+    {
+      "epoch": 1.7794676806083651,
+      "grad_norm": 0.6860196036777416,
+      "learning_rate": 7.328684543901598e-07,
+      "loss": 0.2805,
+      "step": 936
+    },
+    {
+      "epoch": 1.7813688212927756,
+      "grad_norm": 0.6888415995303877,
+      "learning_rate": 7.204405644008416e-07,
+      "loss": 0.2605,
+      "step": 937
+    },
+    {
+      "epoch": 1.7832699619771863,
+      "grad_norm": 0.7035418458893344,
+      "learning_rate": 7.081150141802518e-07,
+      "loss": 0.2847,
+      "step": 938
+    },
+    {
+      "epoch": 1.785171102661597,
+      "grad_norm": 0.6511714601457113,
+      "learning_rate": 6.958919396609231e-07,
+      "loss": 0.2719,
+      "step": 939
+    },
+    {
+      "epoch": 1.7870722433460076,
+      "grad_norm": 0.7105385241153644,
+      "learning_rate": 6.837714756452241e-07,
+      "loss": 0.275,
+      "step": 940
+    },
+    {
+      "epoch": 1.7889733840304183,
+      "grad_norm": 0.6944218324612568,
+      "learning_rate": 6.717537558038845e-07,
+      "loss": 0.2769,
+      "step": 941
+    },
+    {
+      "epoch": 1.790874524714829,
+      "grad_norm": 0.715529180640508,
+      "learning_rate": 6.598389126745209e-07,
+      "loss": 0.2914,
+      "step": 942
+    },
+    {
+      "epoch": 1.7927756653992395,
+      "grad_norm": 0.6868017579514627,
+      "learning_rate": 6.480270776601682e-07,
+      "loss": 0.2776,
+      "step": 943
+    },
+    {
+      "epoch": 1.7946768060836502,
+      "grad_norm": 0.6805099573128464,
+      "learning_rate": 6.36318381027835e-07,
+      "loss": 0.2767,
+      "step": 944
+    },
+    {
+      "epoch": 1.796577946768061,
+      "grad_norm": 0.6639962817866456,
+      "learning_rate": 6.247129519070728e-07,
+      "loss": 0.2652,
+      "step": 945
+    },
+    {
+      "epoch": 1.7984790874524714,
+      "grad_norm": 0.6769414476687063,
+      "learning_rate": 6.132109182885382e-07,
+      "loss": 0.2833,
+      "step": 946
+    },
+    {
+      "epoch": 1.8003802281368821,
+      "grad_norm": 0.6563560713528813,
+      "learning_rate": 6.018124070225928e-07,
+      "loss": 0.2802,
+      "step": 947
+    },
+    {
+      "epoch": 1.8022813688212929,
+      "grad_norm": 0.6425101989048673,
+      "learning_rate": 5.905175438178979e-07,
+      "loss": 0.2744,
+      "step": 948
+    },
+    {
+      "epoch": 1.8041825095057034,
+      "grad_norm": 0.7010609348778838,
+      "learning_rate": 5.793264532400311e-07,
+      "loss": 0.2775,
+      "step": 949
+    },
+    {
+      "epoch": 1.806083650190114,
+      "grad_norm": 0.6949554370002154,
+      "learning_rate": 5.68239258710116e-07,
+      "loss": 0.271,
+      "step": 950
+    },
+    {
+      "epoch": 1.8079847908745248,
+      "grad_norm": 0.6771239362569724,
+      "learning_rate": 5.572560825034523e-07,
+      "loss": 0.2763,
+      "step": 951
+    },
+    {
+      "epoch": 1.8098859315589353,
+      "grad_norm": 0.7246258021509412,
+      "learning_rate": 5.463770457481732e-07,
+      "loss": 0.2862,
+      "step": 952
+    },
+    {
+      "epoch": 1.811787072243346,
+      "grad_norm": 0.6931117078118749,
+      "learning_rate": 5.35602268423906e-07,
+      "loss": 0.2751,
+      "step": 953
+    },
+    {
+      "epoch": 1.8136882129277567,
+      "grad_norm": 1.5629788121151869,
+      "learning_rate": 5.249318693604577e-07,
+      "loss": 0.294,
+      "step": 954
+    },
+    {
+      "epoch": 1.8155893536121672,
+      "grad_norm": 0.7245605831996964,
+      "learning_rate": 5.143659662364931e-07,
+      "loss": 0.2952,
+      "step": 955
+    },
+    {
+      "epoch": 1.817490494296578,
+      "grad_norm": 0.8273796207350173,
+      "learning_rate": 5.039046755782417e-07,
+      "loss": 0.2782,
+      "step": 956
+    },
+    {
+      "epoch": 1.8193916349809887,
+      "grad_norm": 0.6605671600196621,
+      "learning_rate": 4.935481127582131e-07,
+      "loss": 0.2558,
+      "step": 957
+    },
+    {
+      "epoch": 1.8212927756653992,
+      "grad_norm": 0.7049666552091358,
+      "learning_rate": 4.83296391993926e-07,
+      "loss": 0.2887,
+      "step": 958
+    },
+    {
+      "epoch": 1.8231939163498099,
+      "grad_norm": 0.6683582082239524,
+      "learning_rate": 4.7314962634664616e-07,
+      "loss": 0.2678,
+      "step": 959
+    },
+    {
+      "epoch": 1.8250950570342206,
+      "grad_norm": 0.666511372401851,
+      "learning_rate": 4.631079277201389e-07,
+      "loss": 0.2684,
+      "step": 960
+    },
+    {
+      "epoch": 1.826996197718631,
+      "grad_norm": 0.713076805831552,
+      "learning_rate": 4.5317140685943726e-07,
+      "loss": 0.2822,
+      "step": 961
+    },
+    {
+      "epoch": 1.8288973384030418,
+      "grad_norm": 0.6840380305959163,
+      "learning_rate": 4.433401733496201e-07,
+      "loss": 0.2746,
+      "step": 962
+    },
+    {
+      "epoch": 1.8307984790874525,
+      "grad_norm": 0.6861255531559266,
+      "learning_rate": 4.3361433561460274e-07,
+      "loss": 0.2835,
+      "step": 963
+    },
+    {
+      "epoch": 1.832699619771863,
+      "grad_norm": 0.7108197439095701,
+      "learning_rate": 4.2399400091594154e-07,
+      "loss": 0.2749,
+      "step": 964
+    },
+    {
+      "epoch": 1.8346007604562737,
+      "grad_norm": 0.6443379601457143,
+      "learning_rate": 4.14479275351648e-07,
+      "loss": 0.2629,
+      "step": 965
+    },
+    {
+      "epoch": 1.8365019011406845,
+      "grad_norm": 0.656392467160194,
+      "learning_rate": 4.0507026385502747e-07,
+      "loss": 0.2754,
+      "step": 966
+    },
+    {
+      "epoch": 1.838403041825095,
+      "grad_norm": 0.6815368148193814,
+      "learning_rate": 3.9576707019350903e-07,
+      "loss": 0.2783,
+      "step": 967
+    },
+    {
+      "epoch": 1.8403041825095057,
+      "grad_norm": 0.7310484437764095,
+      "learning_rate": 3.865697969675164e-07,
+      "loss": 0.2785,
+      "step": 968
+    },
+    {
+      "epoch": 1.8422053231939164,
+      "grad_norm": 0.7053270490582717,
+      "learning_rate": 3.7747854560931996e-07,
+      "loss": 0.2791,
+      "step": 969
+    },
+    {
+      "epoch": 1.8441064638783269,
+      "grad_norm": 0.6834121858358295,
+      "learning_rate": 3.684934163819309e-07,
+      "loss": 0.2819,
+      "step": 970
+    },
+    {
+      "epoch": 1.8460076045627376,
+      "grad_norm": 0.6587444216756789,
+      "learning_rate": 3.596145083779912e-07,
+      "loss": 0.2697,
+      "step": 971
+    },
+    {
+      "epoch": 1.8479087452471483,
+      "grad_norm": 0.668894674972671,
+      "learning_rate": 3.508419195186774e-07,
+      "loss": 0.2604,
+      "step": 972
+    },
+    {
+      "epoch": 1.8498098859315588,
+      "grad_norm": 0.7078173054603617,
+      "learning_rate": 3.421757465526243e-07,
+      "loss": 0.2787,
+      "step": 973
+    },
+    {
+      "epoch": 1.8517110266159695,
+      "grad_norm": 0.7042714991904963,
+      "learning_rate": 3.33616085054862e-07,
+      "loss": 0.2745,
+      "step": 974
+    },
+    {
+      "epoch": 1.8536121673003803,
+      "grad_norm": 0.7054790800557664,
+      "learning_rate": 3.2516302942574794e-07,
+      "loss": 0.2837,
+      "step": 975
+    },
+    {
+      "epoch": 1.8555133079847907,
+      "grad_norm": 0.6946526683774901,
+      "learning_rate": 3.1681667288994353e-07,
+      "loss": 0.286,
+      "step": 976
+    },
+    {
+      "epoch": 1.8574144486692015,
+      "grad_norm": 0.672029880394973,
+      "learning_rate": 3.0857710749537585e-07,
+      "loss": 0.2745,
+      "step": 977
+    },
+    {
+      "epoch": 1.8593155893536122,
+      "grad_norm": 0.6453329118204936,
+      "learning_rate": 3.0044442411222066e-07,
+      "loss": 0.2669,
+      "step": 978
+    },
+    {
+      "epoch": 1.8612167300380227,
+      "grad_norm": 0.6817068127909786,
+      "learning_rate": 2.9241871243190555e-07,
+      "loss": 0.2743,
+      "step": 979
+    },
+    {
+      "epoch": 1.8631178707224336,
+      "grad_norm": 0.6781979485553478,
+      "learning_rate": 2.845000609661208e-07,
+      "loss": 0.2784,
+      "step": 980
+    },
+    {
+      "epoch": 1.8650190114068441,
+      "grad_norm": 0.6777472732815648,
+      "learning_rate": 2.7668855704583997e-07,
+      "loss": 0.2697,
+      "step": 981
+    },
+    {
+      "epoch": 1.8669201520912546,
+      "grad_norm": 0.6600893491559898,
+      "learning_rate": 2.689842868203563e-07,
+      "loss": 0.2632,
+      "step": 982
+    },
+    {
+      "epoch": 1.8688212927756656,
+      "grad_norm": 0.6848691236860588,
+      "learning_rate": 2.6138733525633896e-07,
+      "loss": 0.271,
+      "step": 983
+    },
+    {
+      "epoch": 1.870722433460076,
+      "grad_norm": 0.713711477238897,
+      "learning_rate": 2.5389778613688744e-07,
+      "loss": 0.2853,
+      "step": 984
+    },
+    {
+      "epoch": 1.8726235741444865,
+      "grad_norm": 0.6712765682266242,
+      "learning_rate": 2.46515722060614e-07,
+      "loss": 0.2729,
+      "step": 985
+    },
+    {
+      "epoch": 1.8745247148288975,
+      "grad_norm": 0.6492014723950595,
+      "learning_rate": 2.392412244407294e-07,
+      "loss": 0.2547,
+      "step": 986
+    },
+    {
+      "epoch": 1.876425855513308,
+      "grad_norm": 0.6821474827315953,
+      "learning_rate": 2.3207437350414418e-07,
+      "loss": 0.2722,
+      "step": 987
+    },
+    {
+      "epoch": 1.8783269961977185,
+      "grad_norm": 0.6935198347935035,
+      "learning_rate": 2.2501524829059208e-07,
+      "loss": 0.2811,
+      "step": 988
+    },
+    {
+      "epoch": 1.8802281368821294,
+      "grad_norm": 0.7166980434502301,
+      "learning_rate": 2.180639266517448e-07,
+      "loss": 0.2757,
+      "step": 989
+    },
+    {
+      "epoch": 1.88212927756654,
+      "grad_norm": 0.6994391420116437,
+      "learning_rate": 2.1122048525036409e-07,
+      "loss": 0.2789,
+      "step": 990
+    },
+    {
+      "epoch": 1.8840304182509504,
+      "grad_norm": 0.7312907428692164,
+      "learning_rate": 2.0448499955945223e-07,
+      "loss": 0.2791,
+      "step": 991
+    },
+    {
+      "epoch": 1.8859315589353614,
+      "grad_norm": 0.710070025719305,
+      "learning_rate": 1.9785754386142164e-07,
+      "loss": 0.2813,
+      "step": 992
+    },
+    {
+      "epoch": 1.8878326996197718,
+      "grad_norm": 0.668768603271782,
+      "learning_rate": 1.9133819124727003e-07,
+      "loss": 0.2757,
+      "step": 993
+    },
+    {
+      "epoch": 1.8897338403041823,
+      "grad_norm": 0.7583087524168939,
+      "learning_rate": 1.8492701361578326e-07,
+      "loss": 0.2844,
+      "step": 994
+    },
+    {
+      "epoch": 1.8916349809885933,
+      "grad_norm": 0.6935207627941306,
+      "learning_rate": 1.7862408167273472e-07,
+      "loss": 0.2924,
+      "step": 995
+    },
+    {
+      "epoch": 1.8935361216730038,
+      "grad_norm": 0.7123814823636084,
+      "learning_rate": 1.724294649301095e-07,
+      "loss": 0.2807,
+      "step": 996
+    },
+    {
+      "epoch": 1.8954372623574145,
+      "grad_norm": 0.6968075678141445,
+      "learning_rate": 1.6634323170533928e-07,
+      "loss": 0.2787,
+      "step": 997
+    },
+    {
+      "epoch": 1.8973384030418252,
+      "grad_norm": 0.6720523523906995,
+      "learning_rate": 1.6036544912054087e-07,
+      "loss": 0.271,
+      "step": 998
+    },
+    {
+      "epoch": 1.8992395437262357,
+      "grad_norm": 0.7150382906764732,
+      "learning_rate": 1.544961831017855e-07,
+      "loss": 0.2881,
+      "step": 999
+    },
+    {
+      "epoch": 1.9011406844106464,
+      "grad_norm": 0.7261200240290979,
+      "learning_rate": 1.487354983783673e-07,
+      "loss": 0.2798,
+      "step": 1000
+    },
+    {
+      "epoch": 1.9030418250950571,
+      "grad_norm": 0.6801206319161398,
+      "learning_rate": 1.430834584820895e-07,
+      "loss": 0.2761,
+      "step": 1001
+    },
+    {
+      "epoch": 1.9049429657794676,
+      "grad_norm": 0.6928480609200753,
+      "learning_rate": 1.375401257465625e-07,
+      "loss": 0.2798,
+      "step": 1002
+    },
+    {
+      "epoch": 1.9068441064638784,
+      "grad_norm": 0.7066673420549063,
+      "learning_rate": 1.3210556130652031e-07,
+      "loss": 0.2753,
+      "step": 1003
+    },
+    {
+      "epoch": 1.908745247148289,
+      "grad_norm": 0.6938024176715012,
+      "learning_rate": 1.2677982509714415e-07,
+      "loss": 0.2696,
+      "step": 1004
+    },
+    {
+      "epoch": 1.9106463878326996,
+      "grad_norm": 0.6832162725081329,
+      "learning_rate": 1.2156297585339872e-07,
+      "loss": 0.2702,
+      "step": 1005
+    },
+    {
+      "epoch": 1.9125475285171103,
+      "grad_norm": 0.6868544179685439,
+      "learning_rate": 1.1645507110938925e-07,
+      "loss": 0.2759,
+      "step": 1006
+    },
+    {
+      "epoch": 1.914448669201521,
+      "grad_norm": 0.7009678907368356,
+      "learning_rate": 1.1145616719772545e-07,
+      "loss": 0.2818,
+      "step": 1007
+    },
+    {
+      "epoch": 1.9163498098859315,
+      "grad_norm": 0.6906500283728016,
+      "learning_rate": 1.0656631924889749e-07,
+      "loss": 0.2781,
+      "step": 1008
+    },
+    {
+      "epoch": 1.9182509505703422,
+      "grad_norm": 0.7178300123778346,
+      "learning_rate": 1.0178558119067316e-07,
+      "loss": 0.2766,
+      "step": 1009
+    },
+    {
+      "epoch": 1.920152091254753,
+      "grad_norm": 0.6780159664421509,
+      "learning_rate": 9.711400574749507e-08,
+      "loss": 0.265,
+      "step": 1010
+    },
+    {
+      "epoch": 1.9220532319391634,
+      "grad_norm": 0.6496202474796194,
+      "learning_rate": 9.255164443990994e-08,
+      "loss": 0.2748,
+      "step": 1011
+    },
+    {
+      "epoch": 1.9239543726235742,
+      "grad_norm": 0.7352653366632927,
+      "learning_rate": 8.809854758399017e-08,
+      "loss": 0.2752,
+      "step": 1012
+    },
+    {
+      "epoch": 1.9258555133079849,
+      "grad_norm": 0.6820168331591996,
+      "learning_rate": 8.375476429078543e-08,
+      "loss": 0.2903,
+      "step": 1013
+    },
+    {
+      "epoch": 1.9277566539923954,
+      "grad_norm": 0.6685324652949994,
+      "learning_rate": 7.952034246577977e-08,
+      "loss": 0.2707,
+      "step": 1014
+    },
+    {
+      "epoch": 1.929657794676806,
+      "grad_norm": 0.6938832969532969,
+      "learning_rate": 7.539532880836087e-08,
+      "loss": 0.2821,
+      "step": 1015
+    },
+    {
+      "epoch": 1.9315589353612168,
+      "grad_norm": 0.6434289753672234,
+      "learning_rate": 7.137976881130826e-08,
+      "loss": 0.2655,
+      "step": 1016
+    },
+    {
+      "epoch": 1.9334600760456273,
+      "grad_norm": 0.7086701935734794,
+      "learning_rate": 6.747370676028819e-08,
+      "loss": 0.2704,
+      "step": 1017
+    },
+    {
+      "epoch": 1.935361216730038,
+      "grad_norm": 0.7360043690034902,
+      "learning_rate": 6.367718573336845e-08,
+      "loss": 0.2794,
+      "step": 1018
+    },
+    {
+      "epoch": 1.9372623574144487,
+      "grad_norm": 0.7306312833478155,
+      "learning_rate": 5.999024760054095e-08,
+      "loss": 0.2775,
+      "step": 1019
+    },
+    {
+      "epoch": 1.9391634980988592,
+      "grad_norm": 0.6660644821121462,
+      "learning_rate": 5.641293302326323e-08,
+      "loss": 0.2709,
+      "step": 1020
+    },
+    {
+      "epoch": 1.94106463878327,
+      "grad_norm": 0.6685425836634464,
+      "learning_rate": 5.2945281454003236e-08,
+      "loss": 0.2791,
+      "step": 1021
+    },
+    {
+      "epoch": 1.9429657794676807,
+      "grad_norm": 0.6848757248963188,
+      "learning_rate": 4.958733113581415e-08,
+      "loss": 0.2758,
+      "step": 1022
+    },
+    {
+      "epoch": 1.9448669201520912,
+      "grad_norm": 0.7092741245479027,
+      "learning_rate": 4.6339119101902475e-08,
+      "loss": 0.269,
+      "step": 1023
+    },
+    {
+      "epoch": 1.946768060836502,
+      "grad_norm": 0.6752088458139953,
+      "learning_rate": 4.320068117522835e-08,
+      "loss": 0.2813,
+      "step": 1024
+    },
+    {
+      "epoch": 1.9486692015209126,
+      "grad_norm": 0.6822127675124053,
+      "learning_rate": 4.0172051968101474e-08,
+      "loss": 0.277,
+      "step": 1025
+    },
+    {
+      "epoch": 1.950570342205323,
+      "grad_norm": 0.6551400841235119,
+      "learning_rate": 3.7253264881809137e-08,
+      "loss": 0.2634,
+      "step": 1026
+    },
+    {
+      "epoch": 1.9524714828897338,
+      "grad_norm": 0.6589303753374033,
+      "learning_rate": 3.4444352106242086e-08,
+      "loss": 0.2632,
+      "step": 1027
+    },
+    {
+      "epoch": 1.9543726235741445,
+      "grad_norm": 0.6674874277029278,
+      "learning_rate": 3.174534461953593e-08,
+      "loss": 0.2662,
+      "step": 1028
+    },
+    {
+      "epoch": 1.956273764258555,
+      "grad_norm": 0.7391127060114976,
+      "learning_rate": 2.915627218774142e-08,
+      "loss": 0.267,
+      "step": 1029
+    },
+    {
+      "epoch": 1.9581749049429658,
+      "grad_norm": 0.713977297670967,
+      "learning_rate": 2.667716336448356e-08,
+      "loss": 0.2827,
+      "step": 1030
+    },
+    {
+      "epoch": 1.9600760456273765,
+      "grad_norm": 0.668278900879979,
+      "learning_rate": 2.430804549065302e-08,
+      "loss": 0.2664,
+      "step": 1031
+    },
+    {
+      "epoch": 1.961977186311787,
+      "grad_norm": 0.6680015728275445,
+      "learning_rate": 2.2048944694104123e-08,
+      "loss": 0.2657,
+      "step": 1032
+    },
+    {
+      "epoch": 1.9638783269961977,
+      "grad_norm": 0.7243982885229068,
+      "learning_rate": 1.989988588936509e-08,
+      "loss": 0.2933,
+      "step": 1033
+    },
+    {
+      "epoch": 1.9657794676806084,
+      "grad_norm": 0.6947306471412228,
+      "learning_rate": 1.7860892777367133e-08,
+      "loss": 0.2675,
+      "step": 1034
+    },
+    {
+      "epoch": 1.967680608365019,
+      "grad_norm": 0.6625040156740044,
+      "learning_rate": 1.5931987845176912e-08,
+      "loss": 0.2588,
+      "step": 1035
+    },
+    {
+      "epoch": 1.9695817490494296,
+      "grad_norm": 0.6748019757125813,
+      "learning_rate": 1.411319236575337e-08,
+      "loss": 0.2701,
+      "step": 1036
+    },
+    {
+      "epoch": 1.9714828897338403,
+      "grad_norm": 0.7437194669718353,
+      "learning_rate": 1.2404526397711281e-08,
+      "loss": 0.2904,
+      "step": 1037
+    },
+    {
+      "epoch": 1.9733840304182508,
+      "grad_norm": 0.6601772381612543,
+      "learning_rate": 1.0806008785100297e-08,
+      "loss": 0.2644,
+      "step": 1038
+    },
+    {
+      "epoch": 1.9752851711026616,
+      "grad_norm": 0.7614110492574241,
+      "learning_rate": 9.317657157197347e-09,
+      "loss": 0.2899,
+      "step": 1039
+    },
+    {
+      "epoch": 1.9771863117870723,
+      "grad_norm": 0.7584653405527474,
+      "learning_rate": 7.93948792831234e-09,
+      "loss": 0.2845,
+      "step": 1040
+    },
+    {
+      "epoch": 1.9790874524714828,
+      "grad_norm": 0.707319769410083,
+      "learning_rate": 6.671516297606095e-09,
+      "loss": 0.2767,
+      "step": 1041
+    },
+    {
+      "epoch": 1.9809885931558935,
+      "grad_norm": 0.7191158662108926,
+      "learning_rate": 5.513756248924917e-09,
+      "loss": 0.2665,
+      "step": 1042
+    },
+    {
+      "epoch": 1.9828897338403042,
+      "grad_norm": 0.6804447955325379,
+      "learning_rate": 4.466220550641831e-09,
+      "loss": 0.267,
+      "step": 1043
+    },
+    {
+      "epoch": 1.9847908745247147,
+      "grad_norm": 0.707744828388341,
+      "learning_rate": 3.528920755523357e-09,
+      "loss": 0.281,
+      "step": 1044
+    },
+    {
+      "epoch": 1.9866920152091256,
+      "grad_norm": 0.7064534939100834,
+      "learning_rate": 2.701867200592956e-09,
+      "loss": 0.2781,
+      "step": 1045
+    },
+    {
+      "epoch": 1.9885931558935361,
+      "grad_norm": 0.6585852925605958,
+      "learning_rate": 1.9850690070266633e-09,
+      "loss": 0.2742,
+      "step": 1046
+    },
+    {
+      "epoch": 1.9904942965779466,
+      "grad_norm": 0.692491061678846,
+      "learning_rate": 1.378534080042071e-09,
+      "loss": 0.2748,
+      "step": 1047
+    },
+    {
+      "epoch": 1.9923954372623576,
+      "grad_norm": 0.7017954683716567,
+      "learning_rate": 8.822691088195001e-10,
+      "loss": 0.2717,
+      "step": 1048
+    },
+    {
+      "epoch": 1.994296577946768,
+      "grad_norm": 0.6978590940080142,
+      "learning_rate": 4.962795664265052e-10,
+      "loss": 0.268,
+      "step": 1049
+    },
+    {
+      "epoch": 1.9961977186311786,
+      "grad_norm": 0.6984542211561577,
+      "learning_rate": 2.2056970975459223e-10,
+      "loss": 0.266,
+      "step": 1050
+    },
+    {
+      "epoch": 1.9980988593155895,
+      "grad_norm": 0.7222912986809148,
+      "learning_rate": 5.514257947369928e-11,
+      "loss": 0.2866,
+      "step": 1051
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6501804000315491,
+      "learning_rate": 0.0,
+      "loss": 0.2698,
+      "step": 1052
+    },
+    {
+      "epoch": 2.0,
+      "step": 1052,
+      "total_flos": 100181286666240.0,
+      "train_loss": 0.36181722887341966,
+      "train_runtime": 3413.9469,
+      "train_samples_per_second": 39.443,
+      "train_steps_per_second": 0.308
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1052,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 100181286666240.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}