| { | |
| "best_metric": 0.9274308681488037, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-100", | |
| "epoch": 3.0, | |
| "eval_steps": 50, | |
| "global_step": 141, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02127659574468085, | |
| "grad_norm": 13.848388671875, | |
| "learning_rate": 1.02e-05, | |
| "loss": 4.4616, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02127659574468085, | |
| "eval_loss": 3.8836147785186768, | |
| "eval_runtime": 3.5795, | |
| "eval_samples_per_second": 176.841, | |
| "eval_steps_per_second": 5.587, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0425531914893617, | |
| "grad_norm": 6.909480094909668, | |
| "learning_rate": 2.04e-05, | |
| "loss": 3.8433, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.06382978723404255, | |
| "grad_norm": 5.8276777267456055, | |
| "learning_rate": 3.06e-05, | |
| "loss": 3.5629, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0851063829787234, | |
| "grad_norm": 5.752169132232666, | |
| "learning_rate": 4.08e-05, | |
| "loss": 3.5849, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.10638297872340426, | |
| "grad_norm": 6.22681999206543, | |
| "learning_rate": 5.1e-05, | |
| "loss": 3.1647, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.1276595744680851, | |
| "grad_norm": 9.792737007141113, | |
| "learning_rate": 6.12e-05, | |
| "loss": 2.7336, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.14893617021276595, | |
| "grad_norm": 8.102272033691406, | |
| "learning_rate": 7.14e-05, | |
| "loss": 2.2142, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1702127659574468, | |
| "grad_norm": 5.091486930847168, | |
| "learning_rate": 8.16e-05, | |
| "loss": 1.9189, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.19148936170212766, | |
| "grad_norm": 2.87353515625, | |
| "learning_rate": 9.18e-05, | |
| "loss": 1.9533, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.2127659574468085, | |
| "grad_norm": 2.3648200035095215, | |
| "learning_rate": 0.000102, | |
| "loss": 1.5369, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.23404255319148937, | |
| "grad_norm": 1.6724998950958252, | |
| "learning_rate": 0.00010198533518731099, | |
| "loss": 1.4629, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.2553191489361702, | |
| "grad_norm": 1.2147119045257568, | |
| "learning_rate": 0.0001019413491828413, | |
| "loss": 1.1376, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.2765957446808511, | |
| "grad_norm": 7.048776626586914, | |
| "learning_rate": 0.00010186806728253272, | |
| "loss": 0.8492, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.2978723404255319, | |
| "grad_norm": 5.654258728027344, | |
| "learning_rate": 0.00010176553163012415, | |
| "loss": 0.5646, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.3191489361702128, | |
| "grad_norm": 4.756352424621582, | |
| "learning_rate": 0.00010163380119291505, | |
| "loss": 1.6847, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.3404255319148936, | |
| "grad_norm": 1.8449368476867676, | |
| "learning_rate": 0.00010147295172785395, | |
| "loss": 1.4681, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.3617021276595745, | |
| "grad_norm": 0.6776601672172546, | |
| "learning_rate": 0.00010128307573797129, | |
| "loss": 1.244, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.3829787234042553, | |
| "grad_norm": 0.5467591285705566, | |
| "learning_rate": 0.00010106428241918177, | |
| "loss": 1.1554, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.40425531914893614, | |
| "grad_norm": 0.8386530876159668, | |
| "learning_rate": 0.00010081669759748692, | |
| "loss": 0.7783, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.425531914893617, | |
| "grad_norm": 0.9994120597839355, | |
| "learning_rate": 0.00010054046365661356, | |
| "loss": 0.1474, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.44680851063829785, | |
| "grad_norm": 0.9148034453392029, | |
| "learning_rate": 0.00010023573945613038, | |
| "loss": 1.3979, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.46808510638297873, | |
| "grad_norm": 0.6885150074958801, | |
| "learning_rate": 9.99027002400892e-05, | |
| "loss": 1.4452, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.48936170212765956, | |
| "grad_norm": 0.5044928193092346, | |
| "learning_rate": 9.954153753624383e-05, | |
| "loss": 1.2974, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.5106382978723404, | |
| "grad_norm": 0.517201840877533, | |
| "learning_rate": 9.915245904590414e-05, | |
| "loss": 1.2173, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.5319148936170213, | |
| "grad_norm": 0.7127615809440613, | |
| "learning_rate": 9.873568852448903e-05, | |
| "loss": 1.0097, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.5531914893617021, | |
| "grad_norm": 1.4452153444290161, | |
| "learning_rate": 9.829146565284679e-05, | |
| "loss": 0.0585, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.574468085106383, | |
| "grad_norm": 0.6424712538719177, | |
| "learning_rate": 9.782004589941682e-05, | |
| "loss": 1.1097, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.5957446808510638, | |
| "grad_norm": 0.4880043864250183, | |
| "learning_rate": 9.732170037331209e-05, | |
| "loss": 1.4546, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.6170212765957447, | |
| "grad_norm": 0.4043918251991272, | |
| "learning_rate": 9.679671566840698e-05, | |
| "loss": 1.2941, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.6382978723404256, | |
| "grad_norm": 0.4223072826862335, | |
| "learning_rate": 9.624539369851954e-05, | |
| "loss": 1.14, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.6595744680851063, | |
| "grad_norm": 0.5029892325401306, | |
| "learning_rate": 9.566805152378394e-05, | |
| "loss": 0.9826, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.6808510638297872, | |
| "grad_norm": 0.5004958510398865, | |
| "learning_rate": 9.50650211683119e-05, | |
| "loss": 0.3189, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.7021276595744681, | |
| "grad_norm": 0.46429964900016785, | |
| "learning_rate": 9.443664942924885e-05, | |
| "loss": 0.9003, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.723404255319149, | |
| "grad_norm": 0.5286682844161987, | |
| "learning_rate": 9.378329767733415e-05, | |
| "loss": 1.4447, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.7446808510638298, | |
| "grad_norm": 0.4269276559352875, | |
| "learning_rate": 9.310534164908e-05, | |
| "loss": 1.3137, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.7659574468085106, | |
| "grad_norm": 0.370991587638855, | |
| "learning_rate": 9.240317123068899e-05, | |
| "loss": 1.1297, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.7872340425531915, | |
| "grad_norm": 0.4747011065483093, | |
| "learning_rate": 9.167719023383408e-05, | |
| "loss": 1.0179, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.8085106382978723, | |
| "grad_norm": 0.4905516803264618, | |
| "learning_rate": 9.09278161634304e-05, | |
| "loss": 0.4583, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.8297872340425532, | |
| "grad_norm": 0.43988707661628723, | |
| "learning_rate": 9.015547997753193e-05, | |
| "loss": 0.6616, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.851063829787234, | |
| "grad_norm": 0.45796748995780945, | |
| "learning_rate": 8.936062583949154e-05, | |
| "loss": 1.4275, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.8723404255319149, | |
| "grad_norm": 0.36554864048957825, | |
| "learning_rate": 8.854371086252688e-05, | |
| "loss": 1.2779, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.8936170212765957, | |
| "grad_norm": 0.3677642345428467, | |
| "learning_rate": 8.770520484683873e-05, | |
| "loss": 1.1917, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.9148936170212766, | |
| "grad_norm": 0.35996830463409424, | |
| "learning_rate": 8.68455900094333e-05, | |
| "loss": 1.0847, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.9361702127659575, | |
| "grad_norm": 0.4477192163467407, | |
| "learning_rate": 8.596536070680378e-05, | |
| "loss": 0.7112, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.9574468085106383, | |
| "grad_norm": 0.4123137891292572, | |
| "learning_rate": 8.506502315063037e-05, | |
| "loss": 0.9964, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.9787234042553191, | |
| "grad_norm": 0.358024001121521, | |
| "learning_rate": 8.414509511666283e-05, | |
| "loss": 1.2649, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.37631550431251526, | |
| "learning_rate": 8.320610564695234e-05, | |
| "loss": 0.9995, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.0212765957446808, | |
| "grad_norm": 0.22452251613140106, | |
| "learning_rate": 8.224859474560443e-05, | |
| "loss": 0.1898, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.0425531914893618, | |
| "grad_norm": 0.4229590594768524, | |
| "learning_rate": 8.127311306822753e-05, | |
| "loss": 1.1368, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.0638297872340425, | |
| "grad_norm": 0.37615370750427246, | |
| "learning_rate": 8.028022160525618e-05, | |
| "loss": 1.3321, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.0638297872340425, | |
| "eval_loss": 0.9616568088531494, | |
| "eval_runtime": 3.2062, | |
| "eval_samples_per_second": 197.427, | |
| "eval_steps_per_second": 6.238, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.0851063829787233, | |
| "grad_norm": 0.32277727127075195, | |
| "learning_rate": 7.927049135933059e-05, | |
| "loss": 1.1556, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.1063829787234043, | |
| "grad_norm": 0.4011160135269165, | |
| "learning_rate": 7.82445030169183e-05, | |
| "loss": 1.0646, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.127659574468085, | |
| "grad_norm": 0.5565645098686218, | |
| "learning_rate": 7.720284661436687e-05, | |
| "loss": 0.8884, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.148936170212766, | |
| "grad_norm": 0.30222636461257935, | |
| "learning_rate": 7.614612119857942e-05, | |
| "loss": 0.2506, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.1702127659574468, | |
| "grad_norm": 0.3559470772743225, | |
| "learning_rate": 7.507493448250836e-05, | |
| "loss": 0.722, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.1914893617021276, | |
| "grad_norm": 0.4607730805873871, | |
| "learning_rate": 7.398990249566532e-05, | |
| "loss": 1.3531, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.2127659574468086, | |
| "grad_norm": 0.39202919602394104, | |
| "learning_rate": 7.289164922984824e-05, | |
| "loss": 1.2476, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.2340425531914894, | |
| "grad_norm": 0.35366523265838623, | |
| "learning_rate": 7.178080628028965e-05, | |
| "loss": 1.1099, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.2553191489361701, | |
| "grad_norm": 0.40545791387557983, | |
| "learning_rate": 7.065801248243196e-05, | |
| "loss": 0.9506, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.2765957446808511, | |
| "grad_norm": 0.3662566542625427, | |
| "learning_rate": 6.952391354453924e-05, | |
| "loss": 0.4258, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.297872340425532, | |
| "grad_norm": 0.28656280040740967, | |
| "learning_rate": 6.837916167635644e-05, | |
| "loss": 0.4806, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.3191489361702127, | |
| "grad_norm": 0.47951042652130127, | |
| "learning_rate": 6.722441521402946e-05, | |
| "loss": 1.3409, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.3404255319148937, | |
| "grad_norm": 0.4036313593387604, | |
| "learning_rate": 6.606033824150241e-05, | |
| "loss": 1.272, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.3617021276595744, | |
| "grad_norm": 0.372051477432251, | |
| "learning_rate": 6.48876002086089e-05, | |
| "loss": 1.0842, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.3829787234042552, | |
| "grad_norm": 0.4357682466506958, | |
| "learning_rate": 6.37068755460778e-05, | |
| "loss": 1.0105, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.4042553191489362, | |
| "grad_norm": 0.5092247128486633, | |
| "learning_rate": 6.251884327767429e-05, | |
| "loss": 0.6371, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.425531914893617, | |
| "grad_norm": 0.202036052942276, | |
| "learning_rate": 6.132418662969977e-05, | |
| "loss": 0.2384, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.4468085106382977, | |
| "grad_norm": 0.4590073227882385, | |
| "learning_rate": 6.012359263807463e-05, | |
| "loss": 1.2431, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.4680851063829787, | |
| "grad_norm": 0.395398885011673, | |
| "learning_rate": 5.891775175323035e-05, | |
| "loss": 1.2642, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.4893617021276595, | |
| "grad_norm": 0.4025956690311432, | |
| "learning_rate": 5.770735744303787e-05, | |
| "loss": 1.1548, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.5106382978723403, | |
| "grad_norm": 0.402270644903183, | |
| "learning_rate": 5.6493105794000665e-05, | |
| "loss": 1.0174, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.5319148936170213, | |
| "grad_norm": 0.5136646032333374, | |
| "learning_rate": 5.52756951109419e-05, | |
| "loss": 0.8706, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.5531914893617023, | |
| "grad_norm": 0.0806485190987587, | |
| "learning_rate": 5.405582551541579e-05, | |
| "loss": 0.0626, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.574468085106383, | |
| "grad_norm": 0.42770785093307495, | |
| "learning_rate": 5.283419854307425e-05, | |
| "loss": 1.0383, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.5957446808510638, | |
| "grad_norm": 0.4656476080417633, | |
| "learning_rate": 5.16115167402202e-05, | |
| "loss": 1.293, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.6170212765957448, | |
| "grad_norm": 0.38593193888664246, | |
| "learning_rate": 5.0388483259779815e-05, | |
| "loss": 1.2081, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.6382978723404256, | |
| "grad_norm": 0.3994680643081665, | |
| "learning_rate": 4.916580145692577e-05, | |
| "loss": 1.0481, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.6595744680851063, | |
| "grad_norm": 0.506732702255249, | |
| "learning_rate": 4.794417448458422e-05, | |
| "loss": 0.8817, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.6808510638297873, | |
| "grad_norm": 0.272098571062088, | |
| "learning_rate": 4.67243048890581e-05, | |
| "loss": 0.2088, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.702127659574468, | |
| "grad_norm": 0.35723525285720825, | |
| "learning_rate": 4.5506894205999334e-05, | |
| "loss": 0.6819, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.7234042553191489, | |
| "grad_norm": 0.4605822265148163, | |
| "learning_rate": 4.429264255696214e-05, | |
| "loss": 1.3524, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.7446808510638299, | |
| "grad_norm": 0.38222458958625793, | |
| "learning_rate": 4.308224824676965e-05, | |
| "loss": 1.1625, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.7659574468085106, | |
| "grad_norm": 0.3701620399951935, | |
| "learning_rate": 4.187640736192537e-05, | |
| "loss": 1.064, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.7872340425531914, | |
| "grad_norm": 0.4499792456626892, | |
| "learning_rate": 4.067581337030022e-05, | |
| "loss": 0.9158, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.8085106382978724, | |
| "grad_norm": 0.3936365842819214, | |
| "learning_rate": 3.948115672232572e-05, | |
| "loss": 0.4121, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.8297872340425532, | |
| "grad_norm": 0.27606216073036194, | |
| "learning_rate": 3.8293124453922226e-05, | |
| "loss": 0.4227, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.851063829787234, | |
| "grad_norm": 0.48591378331184387, | |
| "learning_rate": 3.711239979139111e-05, | |
| "loss": 1.3174, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.872340425531915, | |
| "grad_norm": 0.3974682092666626, | |
| "learning_rate": 3.593966175849759e-05, | |
| "loss": 1.2122, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.8936170212765957, | |
| "grad_norm": 0.3816875219345093, | |
| "learning_rate": 3.477558478597054e-05, | |
| "loss": 1.1615, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.9148936170212765, | |
| "grad_norm": 0.4059462547302246, | |
| "learning_rate": 3.362083832364357e-05, | |
| "loss": 0.9824, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.9361702127659575, | |
| "grad_norm": 0.4912261366844177, | |
| "learning_rate": 3.247608645546074e-05, | |
| "loss": 0.683, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.9574468085106385, | |
| "grad_norm": 0.4084428548812866, | |
| "learning_rate": 3.134198751756804e-05, | |
| "loss": 1.0533, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.978723404255319, | |
| "grad_norm": 0.38703203201293945, | |
| "learning_rate": 3.0219193719710368e-05, | |
| "loss": 1.1633, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.4213656783103943, | |
| "learning_rate": 2.910835077015177e-05, | |
| "loss": 0.9266, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 2.021276595744681, | |
| "grad_norm": 0.08342910557985306, | |
| "learning_rate": 2.8010097504334692e-05, | |
| "loss": 0.0614, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.0425531914893615, | |
| "grad_norm": 0.46658873558044434, | |
| "learning_rate": 2.692506551749165e-05, | |
| "loss": 1.0568, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.0638297872340425, | |
| "grad_norm": 0.4580381512641907, | |
| "learning_rate": 2.5853878801420582e-05, | |
| "loss": 1.2753, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 2.0851063829787235, | |
| "grad_norm": 0.3629004955291748, | |
| "learning_rate": 2.4797153385633147e-05, | |
| "loss": 1.0902, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 2.106382978723404, | |
| "grad_norm": 0.3911716938018799, | |
| "learning_rate": 2.3755496983081708e-05, | |
| "loss": 1.0178, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 2.127659574468085, | |
| "grad_norm": 0.475111186504364, | |
| "learning_rate": 2.2729508640669428e-05, | |
| "loss": 0.8545, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.127659574468085, | |
| "eval_loss": 0.9274308681488037, | |
| "eval_runtime": 3.2361, | |
| "eval_samples_per_second": 195.604, | |
| "eval_steps_per_second": 6.18, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.148936170212766, | |
| "grad_norm": 0.2774420976638794, | |
| "learning_rate": 2.1719778394743813e-05, | |
| "loss": 0.212, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 2.1702127659574466, | |
| "grad_norm": 0.3814822733402252, | |
| "learning_rate": 2.0726886931772476e-05, | |
| "loss": 0.7554, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 2.1914893617021276, | |
| "grad_norm": 0.423093318939209, | |
| "learning_rate": 1.9751405254395587e-05, | |
| "loss": 1.3008, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 2.2127659574468086, | |
| "grad_norm": 0.38136741518974304, | |
| "learning_rate": 1.879389435304766e-05, | |
| "loss": 1.1388, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 2.2340425531914896, | |
| "grad_norm": 0.39356493949890137, | |
| "learning_rate": 1.7854904883337184e-05, | |
| "loss": 0.9796, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.25531914893617, | |
| "grad_norm": 0.47382405400276184, | |
| "learning_rate": 1.693497684936963e-05, | |
| "loss": 0.8731, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 2.276595744680851, | |
| "grad_norm": 0.4495427906513214, | |
| "learning_rate": 1.6034639293196224e-05, | |
| "loss": 0.4196, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 2.297872340425532, | |
| "grad_norm": 0.26799333095550537, | |
| "learning_rate": 1.515440999056669e-05, | |
| "loss": 0.338, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 2.3191489361702127, | |
| "grad_norm": 0.4940812587738037, | |
| "learning_rate": 1.429479515316127e-05, | |
| "loss": 1.1992, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 2.3404255319148937, | |
| "grad_norm": 0.4212472140789032, | |
| "learning_rate": 1.3456289137473124e-05, | |
| "loss": 1.153, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.3617021276595747, | |
| "grad_norm": 0.4149324893951416, | |
| "learning_rate": 1.263937416050847e-05, | |
| "loss": 1.05, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 2.382978723404255, | |
| "grad_norm": 0.4490218460559845, | |
| "learning_rate": 1.1844520022468092e-05, | |
| "loss": 0.9362, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 2.404255319148936, | |
| "grad_norm": 0.5155778527259827, | |
| "learning_rate": 1.1072183836569599e-05, | |
| "loss": 0.615, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 2.425531914893617, | |
| "grad_norm": 0.1996319442987442, | |
| "learning_rate": 1.0322809766165916e-05, | |
| "loss": 0.1577, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 2.4468085106382977, | |
| "grad_norm": 0.5449104905128479, | |
| "learning_rate": 9.596828769311028e-06, | |
| "loss": 1.2252, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.4680851063829787, | |
| "grad_norm": 0.4638878405094147, | |
| "learning_rate": 8.894658350919999e-06, | |
| "loss": 1.2128, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.4893617021276597, | |
| "grad_norm": 0.4165002107620239, | |
| "learning_rate": 8.216702322665849e-06, | |
| "loss": 1.0802, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.5106382978723403, | |
| "grad_norm": 0.42553117871284485, | |
| "learning_rate": 7.563350570751137e-06, | |
| "loss": 0.9043, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.5319148936170213, | |
| "grad_norm": 0.5580489039421082, | |
| "learning_rate": 6.934978831688112e-06, | |
| "loss": 0.7406, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.5531914893617023, | |
| "grad_norm": 0.14457367360591888, | |
| "learning_rate": 6.331948476216073e-06, | |
| "loss": 0.0878, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.574468085106383, | |
| "grad_norm": 0.46198827028274536, | |
| "learning_rate": 5.754606301480452e-06, | |
| "loss": 1.0055, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.595744680851064, | |
| "grad_norm": 0.4651603102684021, | |
| "learning_rate": 5.2032843315930305e-06, | |
| "loss": 1.2232, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.617021276595745, | |
| "grad_norm": 0.41709625720977783, | |
| "learning_rate": 4.678299626687903e-06, | |
| "loss": 1.0644, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.6382978723404253, | |
| "grad_norm": 0.42545390129089355, | |
| "learning_rate": 4.179954100583199e-06, | |
| "loss": 0.9916, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.6595744680851063, | |
| "grad_norm": 0.5188893675804138, | |
| "learning_rate": 3.708534347153212e-06, | |
| "loss": 0.8641, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.6808510638297873, | |
| "grad_norm": 0.30658212304115295, | |
| "learning_rate": 3.26431147551097e-06, | |
| "loss": 0.1864, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.702127659574468, | |
| "grad_norm": 0.4011896252632141, | |
| "learning_rate": 2.8475409540958616e-06, | |
| "loss": 0.6371, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.723404255319149, | |
| "grad_norm": 0.4772135615348816, | |
| "learning_rate": 2.45846246375617e-06, | |
| "loss": 1.2719, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.74468085106383, | |
| "grad_norm": 0.44009700417518616, | |
| "learning_rate": 2.097299759910797e-06, | |
| "loss": 1.1568, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.7659574468085104, | |
| "grad_norm": 0.42015552520751953, | |
| "learning_rate": 1.7642605438696306e-06, | |
| "loss": 1.0539, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.7872340425531914, | |
| "grad_norm": 0.4982571303844452, | |
| "learning_rate": 1.4595363433864484e-06, | |
| "loss": 0.8517, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.8085106382978724, | |
| "grad_norm": 0.4780231714248657, | |
| "learning_rate": 1.1833024025130858e-06, | |
| "loss": 0.3794, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.829787234042553, | |
| "grad_norm": 0.3601129949092865, | |
| "learning_rate": 9.357175808182305e-07, | |
| "loss": 0.5229, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.851063829787234, | |
| "grad_norm": 0.5078785419464111, | |
| "learning_rate": 7.169242620287227e-07, | |
| "loss": 1.2511, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.872340425531915, | |
| "grad_norm": 0.42775991559028625, | |
| "learning_rate": 5.270482721460563e-07, | |
| "loss": 1.1546, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.8936170212765955, | |
| "grad_norm": 0.43165627121925354, | |
| "learning_rate": 3.6619880708494724e-07, | |
| "loss": 1.0596, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.9148936170212765, | |
| "grad_norm": 0.4631091356277466, | |
| "learning_rate": 2.3446836987585295e-07, | |
| "loss": 0.9146, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.9361702127659575, | |
| "grad_norm": 0.5159528851509094, | |
| "learning_rate": 1.319327174672832e-07, | |
| "loss": 0.5763, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.9574468085106385, | |
| "grad_norm": 0.426923006772995, | |
| "learning_rate": 5.865081715870424e-08, | |
| "loss": 0.7757, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.978723404255319, | |
| "grad_norm": 0.4215574264526367, | |
| "learning_rate": 1.4664812689001438e-08, | |
| "loss": 1.14, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.5064499378204346, | |
| "learning_rate": 0.0, | |
| "loss": 0.8367, | |
| "step": 141 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 141, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.466930952990884e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |