| { | |
| "best_metric": 0.5428289771080017, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-50", | |
| "epoch": 3.0038910505836576, | |
| "eval_steps": 50, | |
| "global_step": 193, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01556420233463035, | |
| "grad_norm": 1.7703475952148438, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6598, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01556420233463035, | |
| "eval_loss": 0.805687665939331, | |
| "eval_runtime": 10.5307, | |
| "eval_samples_per_second": 10.256, | |
| "eval_steps_per_second": 5.128, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0311284046692607, | |
| "grad_norm": 2.244293689727783, | |
| "learning_rate": 1e-05, | |
| "loss": 0.7386, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.04669260700389105, | |
| "grad_norm": 1.9753533601760864, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.709, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0622568093385214, | |
| "grad_norm": 1.581067681312561, | |
| "learning_rate": 2e-05, | |
| "loss": 0.6545, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.07782101167315175, | |
| "grad_norm": 1.4357757568359375, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.6287, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0933852140077821, | |
| "grad_norm": 1.279219627380371, | |
| "learning_rate": 3e-05, | |
| "loss": 0.6593, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.10894941634241245, | |
| "grad_norm": 1.0461534261703491, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.563, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1245136186770428, | |
| "grad_norm": 0.9932262897491455, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6083, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.14007782101167315, | |
| "grad_norm": 1.0002769231796265, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.575, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.1556420233463035, | |
| "grad_norm": 1.071124792098999, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5916, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.17120622568093385, | |
| "grad_norm": 1.0537177324295044, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 0.5642, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1867704280155642, | |
| "grad_norm": 0.9426125884056091, | |
| "learning_rate": 6e-05, | |
| "loss": 0.5533, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.20233463035019456, | |
| "grad_norm": 0.9893934726715088, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 0.5613, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.2178988326848249, | |
| "grad_norm": 0.9386663436889648, | |
| "learning_rate": 7e-05, | |
| "loss": 0.578, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.23346303501945526, | |
| "grad_norm": 0.9610932469367981, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.5623, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2490272373540856, | |
| "grad_norm": 1.0002214908599854, | |
| "learning_rate": 8e-05, | |
| "loss": 0.6155, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.26459143968871596, | |
| "grad_norm": 0.7454622983932495, | |
| "learning_rate": 8.5e-05, | |
| "loss": 0.5654, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2801556420233463, | |
| "grad_norm": 0.7215997576713562, | |
| "learning_rate": 9e-05, | |
| "loss": 0.5547, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.29571984435797666, | |
| "grad_norm": 0.6941145658493042, | |
| "learning_rate": 9.5e-05, | |
| "loss": 0.5509, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.311284046692607, | |
| "grad_norm": 0.7395036220550537, | |
| "learning_rate": 0.0001, | |
| "loss": 0.5265, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.32684824902723736, | |
| "grad_norm": 0.7585510015487671, | |
| "learning_rate": 9.999175604498867e-05, | |
| "loss": 0.584, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.3424124513618677, | |
| "grad_norm": 0.7440906167030334, | |
| "learning_rate": 9.996702689846645e-05, | |
| "loss": 0.5634, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.35797665369649806, | |
| "grad_norm": 0.7914589643478394, | |
| "learning_rate": 9.992582071507216e-05, | |
| "loss": 0.5744, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3735408560311284, | |
| "grad_norm": 0.7631120681762695, | |
| "learning_rate": 9.986815108288272e-05, | |
| "loss": 0.5593, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.38910505836575876, | |
| "grad_norm": 0.68586665391922, | |
| "learning_rate": 9.979403701893226e-05, | |
| "loss": 0.5455, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.4046692607003891, | |
| "grad_norm": 0.7517482042312622, | |
| "learning_rate": 9.970350296294113e-05, | |
| "loss": 0.5978, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.42023346303501946, | |
| "grad_norm": 0.7497158646583557, | |
| "learning_rate": 9.959657876925671e-05, | |
| "loss": 0.5164, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.4357976653696498, | |
| "grad_norm": 0.6995195746421814, | |
| "learning_rate": 9.94732996970087e-05, | |
| "loss": 0.5031, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.45136186770428016, | |
| "grad_norm": 0.7487537264823914, | |
| "learning_rate": 9.933370639848211e-05, | |
| "loss": 0.5705, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.4669260700389105, | |
| "grad_norm": 0.7707574963569641, | |
| "learning_rate": 9.917784490571187e-05, | |
| "loss": 0.5422, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.48249027237354086, | |
| "grad_norm": 0.7618570327758789, | |
| "learning_rate": 9.900576661530335e-05, | |
| "loss": 0.4955, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4980544747081712, | |
| "grad_norm": 0.8625350594520569, | |
| "learning_rate": 9.88175282714839e-05, | |
| "loss": 0.6098, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.5136186770428015, | |
| "grad_norm": 0.5663038492202759, | |
| "learning_rate": 9.861319194739109e-05, | |
| "loss": 0.5213, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.5291828793774319, | |
| "grad_norm": 0.6109879016876221, | |
| "learning_rate": 9.83928250246034e-05, | |
| "loss": 0.555, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.5447470817120622, | |
| "grad_norm": 0.5990862250328064, | |
| "learning_rate": 9.815650017092077e-05, | |
| "loss": 0.5395, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.5603112840466926, | |
| "grad_norm": 0.5716366767883301, | |
| "learning_rate": 9.790429531640161e-05, | |
| "loss": 0.5209, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.5758754863813229, | |
| "grad_norm": 0.6569287776947021, | |
| "learning_rate": 9.763629362766496e-05, | |
| "loss": 0.5587, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.5914396887159533, | |
| "grad_norm": 0.6579481363296509, | |
| "learning_rate": 9.735258348046536e-05, | |
| "loss": 0.5347, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.6070038910505836, | |
| "grad_norm": 0.6650751829147339, | |
| "learning_rate": 9.705325843055045e-05, | |
| "loss": 0.5646, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.622568093385214, | |
| "grad_norm": 0.6674900650978088, | |
| "learning_rate": 9.673841718280999e-05, | |
| "loss": 0.6048, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.6381322957198443, | |
| "grad_norm": 0.6020020246505737, | |
| "learning_rate": 9.64081635587273e-05, | |
| "loss": 0.5133, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.6536964980544747, | |
| "grad_norm": 0.6875168681144714, | |
| "learning_rate": 9.606260646214313e-05, | |
| "loss": 0.5529, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.669260700389105, | |
| "grad_norm": 0.6856683492660522, | |
| "learning_rate": 9.570185984334383e-05, | |
| "loss": 0.5606, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.6848249027237354, | |
| "grad_norm": 0.6488378047943115, | |
| "learning_rate": 9.53260426614852e-05, | |
| "loss": 0.5298, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.7003891050583657, | |
| "grad_norm": 0.6739886403083801, | |
| "learning_rate": 9.493527884536486e-05, | |
| "loss": 0.5314, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.7159533073929961, | |
| "grad_norm": 0.6689781546592712, | |
| "learning_rate": 9.452969725255558e-05, | |
| "loss": 0.5413, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.7315175097276264, | |
| "grad_norm": 0.7617279887199402, | |
| "learning_rate": 9.410943162691359e-05, | |
| "loss": 0.5513, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.7470817120622568, | |
| "grad_norm": 0.7101250886917114, | |
| "learning_rate": 9.367462055447528e-05, | |
| "loss": 0.4799, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.7626459143968871, | |
| "grad_norm": 0.5001528263092041, | |
| "learning_rate": 9.322540741775744e-05, | |
| "loss": 0.5058, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.7782101167315175, | |
| "grad_norm": 0.5632477402687073, | |
| "learning_rate": 9.276194034847566e-05, | |
| "loss": 0.5636, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.7782101167315175, | |
| "eval_loss": 0.5428289771080017, | |
| "eval_runtime": 10.8958, | |
| "eval_samples_per_second": 9.912, | |
| "eval_steps_per_second": 4.956, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.7937743190661478, | |
| "grad_norm": 0.5578231811523438, | |
| "learning_rate": 9.228437217869667e-05, | |
| "loss": 0.5325, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.8093385214007782, | |
| "grad_norm": 0.5764113068580627, | |
| "learning_rate": 9.179286039044073e-05, | |
| "loss": 0.5326, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.8249027237354085, | |
| "grad_norm": 0.6034092903137207, | |
| "learning_rate": 9.128756706375065e-05, | |
| "loss": 0.5717, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.8404669260700389, | |
| "grad_norm": 0.6172077655792236, | |
| "learning_rate": 9.076865882324452e-05, | |
| "loss": 0.5724, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.8560311284046692, | |
| "grad_norm": 0.5496807098388672, | |
| "learning_rate": 9.023630678316995e-05, | |
| "loss": 0.5209, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.8715953307392996, | |
| "grad_norm": 0.583033561706543, | |
| "learning_rate": 8.969068649097766e-05, | |
| "loss": 0.5555, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.8871595330739299, | |
| "grad_norm": 0.5868954062461853, | |
| "learning_rate": 8.913197786943336e-05, | |
| "loss": 0.5197, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.9027237354085603, | |
| "grad_norm": 0.5880308747291565, | |
| "learning_rate": 8.856036515728666e-05, | |
| "loss": 0.5064, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.9182879377431906, | |
| "grad_norm": 0.689042866230011, | |
| "learning_rate": 8.797603684851685e-05, | |
| "loss": 0.5509, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.933852140077821, | |
| "grad_norm": 0.6428240537643433, | |
| "learning_rate": 8.737918563017553e-05, | |
| "loss": 0.5875, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.9494163424124513, | |
| "grad_norm": 0.5802725553512573, | |
| "learning_rate": 8.677000831884638e-05, | |
| "loss": 0.5031, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.9649805447470817, | |
| "grad_norm": 0.6728083491325378, | |
| "learning_rate": 8.614870579574337e-05, | |
| "loss": 0.5216, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.980544747081712, | |
| "grad_norm": 0.620725154876709, | |
| "learning_rate": 8.551548294046843e-05, | |
| "loss": 0.4774, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.9961089494163424, | |
| "grad_norm": 0.6825420260429382, | |
| "learning_rate": 8.487054856345081e-05, | |
| "loss": 0.5513, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.0116731517509727, | |
| "grad_norm": 1.8129425048828125, | |
| "learning_rate": 8.421411533709009e-05, | |
| "loss": 0.9706, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.027237354085603, | |
| "grad_norm": 0.5069100260734558, | |
| "learning_rate": 8.35463997256257e-05, | |
| "loss": 0.4212, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.0428015564202335, | |
| "grad_norm": 0.5340793132781982, | |
| "learning_rate": 8.28676219137561e-05, | |
| "loss": 0.4766, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.0583657587548638, | |
| "grad_norm": 0.5560380816459656, | |
| "learning_rate": 8.217800573403105e-05, | |
| "loss": 0.4478, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.0739299610894941, | |
| "grad_norm": 0.5705240368843079, | |
| "learning_rate": 8.147777859304096e-05, | |
| "loss": 0.395, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.0894941634241244, | |
| "grad_norm": 0.6902859807014465, | |
| "learning_rate": 8.076717139642775e-05, | |
| "loss": 0.4399, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.105058365758755, | |
| "grad_norm": 0.6529071927070618, | |
| "learning_rate": 8.004641847274181e-05, | |
| "loss": 0.3999, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.1206225680933852, | |
| "grad_norm": 0.5854980945587158, | |
| "learning_rate": 7.931575749617026e-05, | |
| "loss": 0.3876, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.1361867704280155, | |
| "grad_norm": 0.5517836809158325, | |
| "learning_rate": 7.857542940816183e-05, | |
| "loss": 0.3721, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.1517509727626458, | |
| "grad_norm": 0.6773819327354431, | |
| "learning_rate": 7.782567833797457e-05, | |
| "loss": 0.3864, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.1673151750972763, | |
| "grad_norm": 0.7298661470413208, | |
| "learning_rate": 7.70667515221722e-05, | |
| "loss": 0.4056, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.1828793774319066, | |
| "grad_norm": 0.6650790572166443, | |
| "learning_rate": 7.629889922309577e-05, | |
| "loss": 0.4164, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.198443579766537, | |
| "grad_norm": 0.7574925422668457, | |
| "learning_rate": 7.552237464633761e-05, | |
| "loss": 0.3818, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.2140077821011672, | |
| "grad_norm": 0.7199667692184448, | |
| "learning_rate": 7.473743385724478e-05, | |
| "loss": 0.3314, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.2295719844357977, | |
| "grad_norm": 0.7638531923294067, | |
| "learning_rate": 7.394433569647934e-05, | |
| "loss": 0.3508, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.245136186770428, | |
| "grad_norm": 0.8220713138580322, | |
| "learning_rate": 7.31433416946636e-05, | |
| "loss": 0.3309, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.2607003891050583, | |
| "grad_norm": 0.7258894443511963, | |
| "learning_rate": 7.233471598613815e-05, | |
| "loss": 0.4268, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.2762645914396886, | |
| "grad_norm": 0.680307149887085, | |
| "learning_rate": 7.151872522186146e-05, | |
| "loss": 0.4124, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.2918287937743191, | |
| "grad_norm": 0.6185837984085083, | |
| "learning_rate": 7.069563848147956e-05, | |
| "loss": 0.4174, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.3073929961089494, | |
| "grad_norm": 0.5700533390045166, | |
| "learning_rate": 6.986572718459479e-05, | |
| "loss": 0.4186, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.3229571984435797, | |
| "grad_norm": 0.6120938062667847, | |
| "learning_rate": 6.902926500126292e-05, | |
| "loss": 0.4139, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.3385214007782102, | |
| "grad_norm": 0.5696559548377991, | |
| "learning_rate": 6.818652776174827e-05, | |
| "loss": 0.3949, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.3540856031128405, | |
| "grad_norm": 0.6349061131477356, | |
| "learning_rate": 6.733779336556642e-05, | |
| "loss": 0.4209, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.3696498054474708, | |
| "grad_norm": 0.6833882331848145, | |
| "learning_rate": 6.648334168984452e-05, | |
| "loss": 0.4119, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.3852140077821011, | |
| "grad_norm": 0.6829484105110168, | |
| "learning_rate": 6.562345449702951e-05, | |
| "loss": 0.3991, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.4007782101167314, | |
| "grad_norm": 0.6942391991615295, | |
| "learning_rate": 6.47584153419747e-05, | |
| "loss": 0.4104, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.416342412451362, | |
| "grad_norm": 0.6838539242744446, | |
| "learning_rate": 6.388850947843517e-05, | |
| "loss": 0.3672, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.4319066147859922, | |
| "grad_norm": 0.7023414373397827, | |
| "learning_rate": 6.301402376500304e-05, | |
| "loss": 0.3579, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.4474708171206225, | |
| "grad_norm": 0.738709032535553, | |
| "learning_rate": 6.213524657051353e-05, | |
| "loss": 0.4006, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.463035019455253, | |
| "grad_norm": 0.724032461643219, | |
| "learning_rate": 6.125246767895286e-05, | |
| "loss": 0.328, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.4785992217898833, | |
| "grad_norm": 0.7786082029342651, | |
| "learning_rate": 6.036597819389972e-05, | |
| "loss": 0.3658, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.4941634241245136, | |
| "grad_norm": 0.9172019958496094, | |
| "learning_rate": 5.947607044253142e-05, | |
| "loss": 0.3407, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.509727626459144, | |
| "grad_norm": 0.6716464161872864, | |
| "learning_rate": 5.858303787922663e-05, | |
| "loss": 0.3548, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.5252918287937742, | |
| "grad_norm": 0.7111585736274719, | |
| "learning_rate": 5.768717498879635e-05, | |
| "loss": 0.4403, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.5408560311284045, | |
| "grad_norm": 0.6962729692459106, | |
| "learning_rate": 5.67887771893752e-05, | |
| "loss": 0.4766, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.556420233463035, | |
| "grad_norm": 0.5779573321342468, | |
| "learning_rate": 5.5888140735004804e-05, | |
| "loss": 0.3564, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.556420233463035, | |
| "eval_loss": 0.5569015145301819, | |
| "eval_runtime": 10.8918, | |
| "eval_samples_per_second": 9.916, | |
| "eval_steps_per_second": 4.958, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.5719844357976653, | |
| "grad_norm": 0.6904802322387695, | |
| "learning_rate": 5.498556261794161e-05, | |
| "loss": 0.4492, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.5875486381322959, | |
| "grad_norm": 0.6333062648773193, | |
| "learning_rate": 5.4081340470721284e-05, | |
| "loss": 0.4208, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.6031128404669261, | |
| "grad_norm": 0.5850105881690979, | |
| "learning_rate": 5.31757724680119e-05, | |
| "loss": 0.3787, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.6186770428015564, | |
| "grad_norm": 0.628386914730072, | |
| "learning_rate": 5.22691572282884e-05, | |
| "loss": 0.4535, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.6342412451361867, | |
| "grad_norm": 0.5983573198318481, | |
| "learning_rate": 5.136179371536076e-05, | |
| "loss": 0.3768, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.649805447470817, | |
| "grad_norm": 0.6634320020675659, | |
| "learning_rate": 5.045398113978817e-05, | |
| "loss": 0.3814, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.6653696498054473, | |
| "grad_norm": 0.5900952219963074, | |
| "learning_rate": 4.9546018860211844e-05, | |
| "loss": 0.3388, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.6809338521400778, | |
| "grad_norm": 0.741165816783905, | |
| "learning_rate": 4.863820628463925e-05, | |
| "loss": 0.359, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.6964980544747081, | |
| "grad_norm": 0.7050174474716187, | |
| "learning_rate": 4.773084277171161e-05, | |
| "loss": 0.3425, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.7120622568093387, | |
| "grad_norm": 0.7681849598884583, | |
| "learning_rate": 4.682422753198812e-05, | |
| "loss": 0.3615, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.727626459143969, | |
| "grad_norm": 0.7492854595184326, | |
| "learning_rate": 4.591865952927873e-05, | |
| "loss": 0.3326, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.7431906614785992, | |
| "grad_norm": 0.8712835907936096, | |
| "learning_rate": 4.501443738205841e-05, | |
| "loss": 0.3441, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.7587548638132295, | |
| "grad_norm": 0.6653512120246887, | |
| "learning_rate": 4.41118592649952e-05, | |
| "loss": 0.4284, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.7743190661478598, | |
| "grad_norm": 0.6634366512298584, | |
| "learning_rate": 4.321122281062481e-05, | |
| "loss": 0.441, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.7898832684824901, | |
| "grad_norm": 0.6530463099479675, | |
| "learning_rate": 4.231282501120366e-05, | |
| "loss": 0.4032, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.8054474708171206, | |
| "grad_norm": 0.6511101126670837, | |
| "learning_rate": 4.1416962120773396e-05, | |
| "loss": 0.4025, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.821011673151751, | |
| "grad_norm": 0.6638808846473694, | |
| "learning_rate": 4.0523929557468594e-05, | |
| "loss": 0.4072, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.8365758754863815, | |
| "grad_norm": 0.6273797750473022, | |
| "learning_rate": 3.9634021806100274e-05, | |
| "loss": 0.424, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.8521400778210118, | |
| "grad_norm": 0.6482309103012085, | |
| "learning_rate": 3.874753232104714e-05, | |
| "loss": 0.4046, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.867704280155642, | |
| "grad_norm": 0.6184787750244141, | |
| "learning_rate": 3.786475342948647e-05, | |
| "loss": 0.3906, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.8832684824902723, | |
| "grad_norm": 0.655604362487793, | |
| "learning_rate": 3.6985976234996954e-05, | |
| "loss": 0.4001, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.8988326848249026, | |
| "grad_norm": 0.6891288757324219, | |
| "learning_rate": 3.611149052156483e-05, | |
| "loss": 0.4013, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.914396887159533, | |
| "grad_norm": 0.7344037890434265, | |
| "learning_rate": 3.524158465802531e-05, | |
| "loss": 0.3742, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.9299610894941635, | |
| "grad_norm": 0.669601559638977, | |
| "learning_rate": 3.437654550297049e-05, | |
| "loss": 0.3722, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.9455252918287937, | |
| "grad_norm": 0.7063509225845337, | |
| "learning_rate": 3.351665831015549e-05, | |
| "loss": 0.3521, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.9610894941634243, | |
| "grad_norm": 0.7894170880317688, | |
| "learning_rate": 3.2662206634433576e-05, | |
| "loss": 0.3634, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.9766536964980546, | |
| "grad_norm": 0.799424946308136, | |
| "learning_rate": 3.181347223825174e-05, | |
| "loss": 0.3357, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.9922178988326849, | |
| "grad_norm": 0.8893400430679321, | |
| "learning_rate": 3.0970734998737095e-05, | |
| "loss": 0.3503, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.007782101167315, | |
| "grad_norm": 2.1515607833862305, | |
| "learning_rate": 3.013427281540523e-05, | |
| "loss": 0.6816, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.0233463035019454, | |
| "grad_norm": 0.48720091581344604, | |
| "learning_rate": 2.9304361518520445e-05, | |
| "loss": 0.2903, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.0389105058365757, | |
| "grad_norm": 0.6416101455688477, | |
| "learning_rate": 2.8481274778138567e-05, | |
| "loss": 0.3466, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.054474708171206, | |
| "grad_norm": 0.6130954623222351, | |
| "learning_rate": 2.766528401386187e-05, | |
| "loss": 0.3311, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.0700389105058368, | |
| "grad_norm": 0.6637570858001709, | |
| "learning_rate": 2.685665830533642e-05, | |
| "loss": 0.2966, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.085603112840467, | |
| "grad_norm": 0.697539210319519, | |
| "learning_rate": 2.6055664303520653e-05, | |
| "loss": 0.2972, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.1011673151750974, | |
| "grad_norm": 0.692152202129364, | |
| "learning_rate": 2.526256614275524e-05, | |
| "loss": 0.3071, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.1167315175097277, | |
| "grad_norm": 0.704645574092865, | |
| "learning_rate": 2.4477625353662398e-05, | |
| "loss": 0.2674, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.132295719844358, | |
| "grad_norm": 0.7347646951675415, | |
| "learning_rate": 2.370110077690425e-05, | |
| "loss": 0.2777, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.1478599221789882, | |
| "grad_norm": 0.7842221260070801, | |
| "learning_rate": 2.2933248477827813e-05, | |
| "loss": 0.2354, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.1634241245136185, | |
| "grad_norm": 0.702087938785553, | |
| "learning_rate": 2.2174321662025427e-05, | |
| "loss": 0.2226, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.178988326848249, | |
| "grad_norm": 0.8129343390464783, | |
| "learning_rate": 2.1424570591838183e-05, | |
| "loss": 0.2312, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.1945525291828796, | |
| "grad_norm": 0.7870627045631409, | |
| "learning_rate": 2.068424250382974e-05, | |
| "loss": 0.2237, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.21011673151751, | |
| "grad_norm": 0.8365576267242432, | |
| "learning_rate": 1.9953581527258182e-05, | |
| "loss": 0.2173, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.22568093385214, | |
| "grad_norm": 0.8475888967514038, | |
| "learning_rate": 1.9232828603572256e-05, | |
| "loss": 0.2003, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.2412451361867705, | |
| "grad_norm": 0.9297432899475098, | |
| "learning_rate": 1.852222140695906e-05, | |
| "loss": 0.1835, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.2568093385214008, | |
| "grad_norm": 0.9988775253295898, | |
| "learning_rate": 1.7821994265968962e-05, | |
| "loss": 0.2748, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.272373540856031, | |
| "grad_norm": 0.8897411823272705, | |
| "learning_rate": 1.7132378086243904e-05, | |
| "loss": 0.2846, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.2879377431906613, | |
| "grad_norm": 1.0305852890014648, | |
| "learning_rate": 1.6453600274374298e-05, | |
| "loss": 0.337, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.3035019455252916, | |
| "grad_norm": 0.9018167853355408, | |
| "learning_rate": 1.5785884662909916e-05, | |
| "loss": 0.2914, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.319066147859922, | |
| "grad_norm": 0.9446598291397095, | |
| "learning_rate": 1.5129451436549203e-05, | |
| "loss": 0.2451, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.3346303501945527, | |
| "grad_norm": 0.8176512718200684, | |
| "learning_rate": 1.4484517059531588e-05, | |
| "loss": 0.2951, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.3346303501945527, | |
| "eval_loss": 0.5986903309822083, | |
| "eval_runtime": 10.8903, | |
| "eval_samples_per_second": 9.917, | |
| "eval_steps_per_second": 4.959, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.350194552529183, | |
| "grad_norm": 0.7516515851020813, | |
| "learning_rate": 1.3851294204256638e-05, | |
| "loss": 0.2704, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.3657587548638133, | |
| "grad_norm": 0.7066805958747864, | |
| "learning_rate": 1.322999168115363e-05, | |
| "loss": 0.2575, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.3813229571984436, | |
| "grad_norm": 0.7425258755683899, | |
| "learning_rate": 1.262081436982448e-05, | |
| "loss": 0.2803, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.396887159533074, | |
| "grad_norm": 0.7812573909759521, | |
| "learning_rate": 1.2023963151483165e-05, | |
| "loss": 0.2676, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.412451361867704, | |
| "grad_norm": 0.7473729848861694, | |
| "learning_rate": 1.143963484271337e-05, | |
| "loss": 0.2523, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.4280155642023344, | |
| "grad_norm": 0.7576512694358826, | |
| "learning_rate": 1.0868022130566651e-05, | |
| "loss": 0.2431, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.443579766536965, | |
| "grad_norm": 0.7459601759910583, | |
| "learning_rate": 1.0309313509022351e-05, | |
| "loss": 0.2187, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.4591439688715955, | |
| "grad_norm": 0.7837074398994446, | |
| "learning_rate": 9.763693216830055e-06, | |
| "loss": 0.209, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.4747081712062258, | |
| "grad_norm": 0.7549136281013489, | |
| "learning_rate": 9.231341176755488e-06, | |
| "loss": 0.1914, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.490272373540856, | |
| "grad_norm": 0.826531708240509, | |
| "learning_rate": 8.712432936249365e-06, | |
| "loss": 0.1852, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.5058365758754864, | |
| "grad_norm": 0.778913140296936, | |
| "learning_rate": 8.207139609559283e-06, | |
| "loss": 0.2547, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.5214007782101167, | |
| "grad_norm": 0.7146987915039062, | |
| "learning_rate": 7.715627821303339e-06, | |
| "loss": 0.3109, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.536964980544747, | |
| "grad_norm": 0.7523741722106934, | |
| "learning_rate": 7.238059651524354e-06, | |
| "loss": 0.308, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.5525291828793772, | |
| "grad_norm": 0.7645745873451233, | |
| "learning_rate": 6.774592582242567e-06, | |
| "loss": 0.266, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.5680933852140075, | |
| "grad_norm": 0.7666687965393066, | |
| "learning_rate": 6.325379445524731e-06, | |
| "loss": 0.2654, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.5836575875486383, | |
| "grad_norm": 0.8015041351318359, | |
| "learning_rate": 5.890568373086425e-06, | |
| "loss": 0.27, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.5992217898832686, | |
| "grad_norm": 0.883911669254303, | |
| "learning_rate": 5.470302747444428e-06, | |
| "loss": 0.283, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.614785992217899, | |
| "grad_norm": 0.9192494750022888, | |
| "learning_rate": 5.064721154635155e-06, | |
| "loss": 0.2714, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.630350194552529, | |
| "grad_norm": 0.9188051819801331, | |
| "learning_rate": 4.673957338514812e-06, | |
| "loss": 0.2363, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.6459143968871595, | |
| "grad_norm": 0.9108679890632629, | |
| "learning_rate": 4.298140156656178e-06, | |
| "loss": 0.2704, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.6614785992217898, | |
| "grad_norm": 0.9353700876235962, | |
| "learning_rate": 3.937393537856871e-06, | |
| "loss": 0.259, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.6770428015564205, | |
| "grad_norm": 0.7926956415176392, | |
| "learning_rate": 3.5918364412727e-06, | |
| "loss": 0.2228, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.692607003891051, | |
| "grad_norm": 0.8170326948165894, | |
| "learning_rate": 3.261582817190023e-06, | |
| "loss": 0.2315, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.708171206225681, | |
| "grad_norm": 0.7552840709686279, | |
| "learning_rate": 2.9467415694495627e-06, | |
| "loss": 0.2082, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.7237354085603114, | |
| "grad_norm": 0.8596252799034119, | |
| "learning_rate": 2.6474165195346346e-06, | |
| "loss": 0.1993, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.7392996108949417, | |
| "grad_norm": 0.8098993301391602, | |
| "learning_rate": 2.363706372335045e-06, | |
| "loss": 0.1707, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.754863813229572, | |
| "grad_norm": 0.8145043253898621, | |
| "learning_rate": 2.095704683598376e-06, | |
| "loss": 0.2661, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.7704280155642023, | |
| "grad_norm": 0.7261046767234802, | |
| "learning_rate": 1.843499829079237e-06, | |
| "loss": 0.3408, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.7859922178988326, | |
| "grad_norm": 0.6709286570549011, | |
| "learning_rate": 1.6071749753965914e-06, | |
| "loss": 0.264, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.801556420233463, | |
| "grad_norm": 0.7309187650680542, | |
| "learning_rate": 1.3868080526089178e-06, | |
| "loss": 0.3079, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.817120622568093, | |
| "grad_norm": 0.729026198387146, | |
| "learning_rate": 1.1824717285160991e-06, | |
| "loss": 0.2465, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.832684824902724, | |
| "grad_norm": 0.847587525844574, | |
| "learning_rate": 9.942333846966746e-07, | |
| "loss": 0.302, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.848249027237354, | |
| "grad_norm": 0.8388762474060059, | |
| "learning_rate": 8.221550942881406e-07, | |
| "loss": 0.2598, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.8638132295719845, | |
| "grad_norm": 0.8012559413909912, | |
| "learning_rate": 6.662936015178978e-07, | |
| "loss": 0.2566, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.8793774319066148, | |
| "grad_norm": 0.749428391456604, | |
| "learning_rate": 5.267003029913065e-07, | |
| "loss": 0.2746, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.894941634241245, | |
| "grad_norm": 0.7644572257995605, | |
| "learning_rate": 4.03421230743295e-07, | |
| "loss": 0.222, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.9105058365758754, | |
| "grad_norm": 0.8482156991958618, | |
| "learning_rate": 2.9649703705887375e-07, | |
| "loss": 0.2298, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.926070038910506, | |
| "grad_norm": 0.7980473041534424, | |
| "learning_rate": 2.0596298106774213e-07, | |
| "loss": 0.2058, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.9416342412451364, | |
| "grad_norm": 0.8403069972991943, | |
| "learning_rate": 1.3184891711727764e-07, | |
| "loss": 0.2168, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.9571984435797667, | |
| "grad_norm": 0.8483723998069763, | |
| "learning_rate": 7.417928492784443e-08, | |
| "loss": 0.1897, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.972762645914397, | |
| "grad_norm": 0.9391508102416992, | |
| "learning_rate": 3.2973101533567695e-08, | |
| "loss": 0.2266, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.9883268482490273, | |
| "grad_norm": 0.7663131952285767, | |
| "learning_rate": 8.243955011333349e-09, | |
| "loss": 0.1504, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 3.0038910505836576, | |
| "grad_norm": 1.7181938886642456, | |
| "learning_rate": 0.0, | |
| "loss": 0.4475, | |
| "step": 193 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 193, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 2 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.4379746513413734e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |