| { | |
| "best_metric": 0.12957406044006348, | |
| "best_model_checkpoint": "saves/psy-course-info-chain/Llama3-OpenBioLLM-8B/train/fold5/checkpoint-50", | |
| "epoch": 4.938271604938271, | |
| "eval_steps": 10, | |
| "global_step": 125, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03950617283950617, | |
| "grad_norm": 1.5741175413131714, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.3695, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.07901234567901234, | |
| "grad_norm": 1.076414704322815, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.2799, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.11851851851851852, | |
| "grad_norm": 1.104756236076355, | |
| "learning_rate": 2.307692307692308e-05, | |
| "loss": 0.34, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.1580246913580247, | |
| "grad_norm": 1.2480536699295044, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.3206, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.19753086419753085, | |
| "grad_norm": 1.4102470874786377, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 0.3089, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.23703703703703705, | |
| "grad_norm": 1.3739464282989502, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 0.2988, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.2765432098765432, | |
| "grad_norm": 1.0986212491989136, | |
| "learning_rate": 5.384615384615385e-05, | |
| "loss": 0.2735, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.3160493827160494, | |
| "grad_norm": 1.2744570970535278, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.2763, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 1.1677888631820679, | |
| "learning_rate": 6.923076923076924e-05, | |
| "loss": 0.2863, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "grad_norm": 0.7967779040336609, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.2295, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "eval_loss": 0.24715404212474823, | |
| "eval_runtime": 25.925, | |
| "eval_samples_per_second": 1.736, | |
| "eval_steps_per_second": 1.736, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.4345679012345679, | |
| "grad_norm": 1.156676173210144, | |
| "learning_rate": 8.461538461538461e-05, | |
| "loss": 0.204, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.4740740740740741, | |
| "grad_norm": 1.0251966714859009, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.1599, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.5135802469135803, | |
| "grad_norm": 0.8487848043441772, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1834, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.5530864197530864, | |
| "grad_norm": 0.9569368362426758, | |
| "learning_rate": 9.998033131915266e-05, | |
| "loss": 0.2157, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 0.7333651781082153, | |
| "learning_rate": 9.992134075089084e-05, | |
| "loss": 0.2021, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.6320987654320988, | |
| "grad_norm": 0.6334995031356812, | |
| "learning_rate": 9.982307470588098e-05, | |
| "loss": 0.1336, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.671604938271605, | |
| "grad_norm": 0.5932794213294983, | |
| "learning_rate": 9.968561049466214e-05, | |
| "loss": 0.1589, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 0.6435730457305908, | |
| "learning_rate": 9.950905626682228e-05, | |
| "loss": 0.1338, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.7506172839506173, | |
| "grad_norm": 0.7178583145141602, | |
| "learning_rate": 9.92935509259118e-05, | |
| "loss": 0.121, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "grad_norm": 0.479437917470932, | |
| "learning_rate": 9.903926402016153e-05, | |
| "loss": 0.1186, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "eval_loss": 0.16460061073303223, | |
| "eval_runtime": 25.7795, | |
| "eval_samples_per_second": 1.746, | |
| "eval_steps_per_second": 1.746, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.8296296296296296, | |
| "grad_norm": 0.4801785945892334, | |
| "learning_rate": 9.874639560909117e-05, | |
| "loss": 0.0798, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.8691358024691358, | |
| "grad_norm": 0.38912343978881836, | |
| "learning_rate": 9.841517610611309e-05, | |
| "loss": 0.0613, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.908641975308642, | |
| "grad_norm": 0.9448972344398499, | |
| "learning_rate": 9.804586609725499e-05, | |
| "loss": 0.1331, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.9481481481481482, | |
| "grad_norm": 0.7427482604980469, | |
| "learning_rate": 9.763875613614482e-05, | |
| "loss": 0.1455, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.9876543209876543, | |
| "grad_norm": 0.6236357092857361, | |
| "learning_rate": 9.719416651541839e-05, | |
| "loss": 0.0844, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.0271604938271606, | |
| "grad_norm": 1.2017364501953125, | |
| "learning_rate": 9.671244701472999e-05, | |
| "loss": 0.1921, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.0666666666666667, | |
| "grad_norm": 0.5268481969833374, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 0.1161, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.106172839506173, | |
| "grad_norm": 0.5642991662025452, | |
| "learning_rate": 9.563916325306594e-05, | |
| "loss": 0.0966, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.145679012345679, | |
| "grad_norm": 0.5303311347961426, | |
| "learning_rate": 9.504844339512095e-05, | |
| "loss": 0.0779, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "grad_norm": 0.7268487215042114, | |
| "learning_rate": 9.442228179894362e-05, | |
| "loss": 0.0857, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "eval_loss": 0.1460779756307602, | |
| "eval_runtime": 25.6938, | |
| "eval_samples_per_second": 1.751, | |
| "eval_steps_per_second": 1.751, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.2246913580246914, | |
| "grad_norm": 0.39900702238082886, | |
| "learning_rate": 9.376117109543769e-05, | |
| "loss": 0.0582, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.2641975308641975, | |
| "grad_norm": 0.39973634481430054, | |
| "learning_rate": 9.306563141162046e-05, | |
| "loss": 0.0632, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.3037037037037038, | |
| "grad_norm": 0.4668061137199402, | |
| "learning_rate": 9.233620996141421e-05, | |
| "loss": 0.0801, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.34320987654321, | |
| "grad_norm": 0.8073009848594666, | |
| "learning_rate": 9.157348061512727e-05, | |
| "loss": 0.125, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.382716049382716, | |
| "grad_norm": 0.5337339043617249, | |
| "learning_rate": 9.077804344796302e-05, | |
| "loss": 0.0999, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.4222222222222223, | |
| "grad_norm": 0.5674287676811218, | |
| "learning_rate": 8.995052426791247e-05, | |
| "loss": 0.1063, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.4617283950617284, | |
| "grad_norm": 0.43140146136283875, | |
| "learning_rate": 8.90915741234015e-05, | |
| "loss": 0.062, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.5012345679012347, | |
| "grad_norm": 0.5618436932563782, | |
| "learning_rate": 8.820186879108038e-05, | |
| "loss": 0.0855, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.5407407407407407, | |
| "grad_norm": 0.6559361219406128, | |
| "learning_rate": 8.728210824415827e-05, | |
| "loss": 0.0946, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.5802469135802468, | |
| "grad_norm": 0.5632725358009338, | |
| "learning_rate": 8.633301610170135e-05, | |
| "loss": 0.0912, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.5802469135802468, | |
| "eval_loss": 0.13206160068511963, | |
| "eval_runtime": 25.8257, | |
| "eval_samples_per_second": 1.742, | |
| "eval_steps_per_second": 1.742, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.6197530864197531, | |
| "grad_norm": 0.4386998414993286, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 0.061, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.6592592592592592, | |
| "grad_norm": 0.5159305930137634, | |
| "learning_rate": 8.434984630174509e-05, | |
| "loss": 0.0773, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.6987654320987655, | |
| "grad_norm": 0.4360441565513611, | |
| "learning_rate": 8.33173288976002e-05, | |
| "loss": 0.0861, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.7382716049382716, | |
| "grad_norm": 0.693144679069519, | |
| "learning_rate": 8.225859917710439e-05, | |
| "loss": 0.1202, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.7602325677871704, | |
| "learning_rate": 8.117449009293668e-05, | |
| "loss": 0.1011, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.817283950617284, | |
| "grad_norm": 0.5033719539642334, | |
| "learning_rate": 8.006585456492029e-05, | |
| "loss": 0.0776, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.8567901234567903, | |
| "grad_norm": 0.43237414956092834, | |
| "learning_rate": 7.89335648089903e-05, | |
| "loss": 0.0665, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.8962962962962964, | |
| "grad_norm": 0.49880075454711914, | |
| "learning_rate": 7.777851165098012e-05, | |
| "loss": 0.1061, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.9358024691358025, | |
| "grad_norm": 0.45825496315956116, | |
| "learning_rate": 7.660160382576683e-05, | |
| "loss": 0.1121, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "grad_norm": 0.49814876914024353, | |
| "learning_rate": 7.540376726232648e-05, | |
| "loss": 0.0842, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "eval_loss": 0.12957406044006348, | |
| "eval_runtime": 25.7094, | |
| "eval_samples_per_second": 1.75, | |
| "eval_steps_per_second": 1.75, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.0148148148148146, | |
| "grad_norm": 1.0445586442947388, | |
| "learning_rate": 7.4185944355262e-05, | |
| "loss": 0.1629, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 2.054320987654321, | |
| "grad_norm": 0.40933534502983093, | |
| "learning_rate": 7.294909322337689e-05, | |
| "loss": 0.066, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 2.093827160493827, | |
| "grad_norm": 0.4381013810634613, | |
| "learning_rate": 7.169418695587791e-05, | |
| "loss": 0.0464, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.1333333333333333, | |
| "grad_norm": 0.347000390291214, | |
| "learning_rate": 7.042221284679982e-05, | |
| "loss": 0.0689, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.1728395061728394, | |
| "grad_norm": 0.3864345848560333, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 0.05, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.212345679012346, | |
| "grad_norm": 0.3639497756958008, | |
| "learning_rate": 6.783107663311565e-05, | |
| "loss": 0.0674, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.251851851851852, | |
| "grad_norm": 0.4984375238418579, | |
| "learning_rate": 6.651395309775837e-05, | |
| "loss": 0.0656, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.291358024691358, | |
| "grad_norm": 0.46161821484565735, | |
| "learning_rate": 6.518383725548074e-05, | |
| "loss": 0.0736, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.330864197530864, | |
| "grad_norm": 0.5982030630111694, | |
| "learning_rate": 6.384177557124247e-05, | |
| "loss": 0.0918, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "grad_norm": 0.5921144485473633, | |
| "learning_rate": 6.248882390836135e-05, | |
| "loss": 0.1157, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "eval_loss": 0.13510963320732117, | |
| "eval_runtime": 25.738, | |
| "eval_samples_per_second": 1.748, | |
| "eval_steps_per_second": 1.748, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.4098765432098768, | |
| "grad_norm": 0.3214099705219269, | |
| "learning_rate": 6.112604669781572e-05, | |
| "loss": 0.0461, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.449382716049383, | |
| "grad_norm": 0.4707167446613312, | |
| "learning_rate": 5.9754516100806423e-05, | |
| "loss": 0.0566, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.488888888888889, | |
| "grad_norm": 0.402936726808548, | |
| "learning_rate": 5.837531116523682e-05, | |
| "loss": 0.057, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.528395061728395, | |
| "grad_norm": 0.6839144825935364, | |
| "learning_rate": 5.698951697677498e-05, | |
| "loss": 0.0669, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.567901234567901, | |
| "grad_norm": 0.44188666343688965, | |
| "learning_rate": 5.559822380516539e-05, | |
| "loss": 0.0387, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.6074074074074076, | |
| "grad_norm": 0.4152555763721466, | |
| "learning_rate": 5.420252624646238e-05, | |
| "loss": 0.065, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.6469135802469137, | |
| "grad_norm": 0.5819364786148071, | |
| "learning_rate": 5.2803522361859594e-05, | |
| "loss": 0.0543, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.68641975308642, | |
| "grad_norm": 0.5173784494400024, | |
| "learning_rate": 5.140231281379345e-05, | |
| "loss": 0.0537, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.725925925925926, | |
| "grad_norm": 0.47724661231040955, | |
| "learning_rate": 5e-05, | |
| "loss": 0.0507, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.765432098765432, | |
| "grad_norm": 0.5026245713233948, | |
| "learning_rate": 4.859768718620656e-05, | |
| "loss": 0.061, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.765432098765432, | |
| "eval_loss": 0.13312244415283203, | |
| "eval_runtime": 25.6777, | |
| "eval_samples_per_second": 1.752, | |
| "eval_steps_per_second": 1.752, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.8049382716049385, | |
| "grad_norm": 0.5522587895393372, | |
| "learning_rate": 4.7196477638140404e-05, | |
| "loss": 0.0409, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.8444444444444446, | |
| "grad_norm": 0.7273226380348206, | |
| "learning_rate": 4.579747375353763e-05, | |
| "loss": 0.0907, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.8839506172839506, | |
| "grad_norm": 0.3048068583011627, | |
| "learning_rate": 4.4401776194834613e-05, | |
| "loss": 0.0301, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.9234567901234567, | |
| "grad_norm": 0.7032773494720459, | |
| "learning_rate": 4.3010483023225045e-05, | |
| "loss": 0.0621, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.962962962962963, | |
| "grad_norm": 0.6828241348266602, | |
| "learning_rate": 4.162468883476319e-05, | |
| "loss": 0.0669, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 3.0024691358024693, | |
| "grad_norm": 1.360464334487915, | |
| "learning_rate": 4.0245483899193595e-05, | |
| "loss": 0.1132, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 3.0419753086419754, | |
| "grad_norm": 0.5620725750923157, | |
| "learning_rate": 3.887395330218429e-05, | |
| "loss": 0.0571, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 3.0814814814814815, | |
| "grad_norm": 0.5476599931716919, | |
| "learning_rate": 3.7511176091638653e-05, | |
| "loss": 0.0577, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 3.1209876543209876, | |
| "grad_norm": 0.7123511433601379, | |
| "learning_rate": 3.6158224428757535e-05, | |
| "loss": 0.0683, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 3.1604938271604937, | |
| "grad_norm": 0.35225990414619446, | |
| "learning_rate": 3.4816162744519263e-05, | |
| "loss": 0.0386, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.1604938271604937, | |
| "eval_loss": 0.13098205626010895, | |
| "eval_runtime": 25.7388, | |
| "eval_samples_per_second": 1.748, | |
| "eval_steps_per_second": 1.748, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.7066668272018433, | |
| "learning_rate": 3.3486046902241664e-05, | |
| "loss": 0.0546, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 3.2395061728395063, | |
| "grad_norm": 0.5175016522407532, | |
| "learning_rate": 3.216892336688435e-05, | |
| "loss": 0.0543, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 3.2790123456790123, | |
| "grad_norm": 0.6760972738265991, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 0.0423, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 3.3185185185185184, | |
| "grad_norm": 0.38699501752853394, | |
| "learning_rate": 2.9577787153200197e-05, | |
| "loss": 0.0443, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 3.3580246913580245, | |
| "grad_norm": 0.5332982540130615, | |
| "learning_rate": 2.8305813044122097e-05, | |
| "loss": 0.0588, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 3.397530864197531, | |
| "grad_norm": 0.33675557374954224, | |
| "learning_rate": 2.705090677662311e-05, | |
| "loss": 0.0527, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 3.437037037037037, | |
| "grad_norm": 0.3769690990447998, | |
| "learning_rate": 2.581405564473801e-05, | |
| "loss": 0.0462, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 3.476543209876543, | |
| "grad_norm": 0.5502105355262756, | |
| "learning_rate": 2.459623273767354e-05, | |
| "loss": 0.0303, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 3.5160493827160493, | |
| "grad_norm": 0.48587530851364136, | |
| "learning_rate": 2.3398396174233178e-05, | |
| "loss": 0.0645, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "grad_norm": 0.5066972970962524, | |
| "learning_rate": 2.2221488349019903e-05, | |
| "loss": 0.0546, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "eval_loss": 0.13408593833446503, | |
| "eval_runtime": 25.7431, | |
| "eval_samples_per_second": 1.748, | |
| "eval_steps_per_second": 1.748, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.595061728395062, | |
| "grad_norm": 0.43133023381233215, | |
| "learning_rate": 2.1066435191009715e-05, | |
| "loss": 0.0363, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 3.634567901234568, | |
| "grad_norm": 0.4554174244403839, | |
| "learning_rate": 1.9934145435079702e-05, | |
| "loss": 0.0448, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 3.674074074074074, | |
| "grad_norm": 0.3631313741207123, | |
| "learning_rate": 1.8825509907063327e-05, | |
| "loss": 0.0339, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 3.71358024691358, | |
| "grad_norm": 0.3978934586048126, | |
| "learning_rate": 1.774140082289563e-05, | |
| "loss": 0.0303, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 3.753086419753086, | |
| "grad_norm": 0.4692234694957733, | |
| "learning_rate": 1.6682671102399805e-05, | |
| "loss": 0.0369, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 3.7925925925925927, | |
| "grad_norm": 0.3033137917518616, | |
| "learning_rate": 1.5650153698254916e-05, | |
| "loss": 0.0294, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 3.832098765432099, | |
| "grad_norm": 0.3886677026748657, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 0.0449, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 3.871604938271605, | |
| "grad_norm": 0.43847882747650146, | |
| "learning_rate": 1.3666983898298657e-05, | |
| "loss": 0.0544, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 3.911111111111111, | |
| "grad_norm": 0.3743950128555298, | |
| "learning_rate": 1.2717891755841722e-05, | |
| "loss": 0.0369, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 3.950617283950617, | |
| "grad_norm": 0.6424919366836548, | |
| "learning_rate": 1.1798131208919627e-05, | |
| "loss": 0.0418, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.950617283950617, | |
| "eval_loss": 0.1339671015739441, | |
| "eval_runtime": 25.6768, | |
| "eval_samples_per_second": 1.753, | |
| "eval_steps_per_second": 1.753, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.9901234567901236, | |
| "grad_norm": 0.6912626028060913, | |
| "learning_rate": 1.090842587659851e-05, | |
| "loss": 0.0581, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 4.029629629629629, | |
| "grad_norm": 1.1715627908706665, | |
| "learning_rate": 1.004947573208756e-05, | |
| "loss": 0.0948, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 4.069135802469136, | |
| "grad_norm": 0.30927592515945435, | |
| "learning_rate": 9.221956552036992e-06, | |
| "loss": 0.0376, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 4.108641975308642, | |
| "grad_norm": 0.36368781328201294, | |
| "learning_rate": 8.426519384872733e-06, | |
| "loss": 0.0296, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 4.148148148148148, | |
| "grad_norm": 0.45377442240715027, | |
| "learning_rate": 7.663790038585793e-06, | |
| "loss": 0.0466, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 4.187654320987654, | |
| "grad_norm": 0.41035357117652893, | |
| "learning_rate": 6.934368588379553e-06, | |
| "loss": 0.0329, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 4.22716049382716, | |
| "grad_norm": 0.5300284624099731, | |
| "learning_rate": 6.238828904562316e-06, | |
| "loss": 0.0384, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 4.266666666666667, | |
| "grad_norm": 0.4160001873970032, | |
| "learning_rate": 5.577718201056392e-06, | |
| "loss": 0.0404, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 4.306172839506173, | |
| "grad_norm": 0.33686503767967224, | |
| "learning_rate": 4.951556604879048e-06, | |
| "loss": 0.0281, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 4.345679012345679, | |
| "grad_norm": 0.29499897360801697, | |
| "learning_rate": 4.360836746934055e-06, | |
| "loss": 0.0287, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.345679012345679, | |
| "eval_loss": 0.1347362995147705, | |
| "eval_runtime": 25.1139, | |
| "eval_samples_per_second": 1.792, | |
| "eval_steps_per_second": 1.792, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.385185185185185, | |
| "grad_norm": 0.3669702410697937, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 0.0367, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 4.424691358024692, | |
| "grad_norm": 0.3981896638870239, | |
| "learning_rate": 3.2875529852700147e-06, | |
| "loss": 0.0475, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 4.4641975308641975, | |
| "grad_norm": 0.39808550477027893, | |
| "learning_rate": 2.8058334845816213e-06, | |
| "loss": 0.0387, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 4.503703703703704, | |
| "grad_norm": 0.4687879979610443, | |
| "learning_rate": 2.361243863855184e-06, | |
| "loss": 0.0432, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 4.54320987654321, | |
| "grad_norm": 0.3455727696418762, | |
| "learning_rate": 1.9541339027450256e-06, | |
| "loss": 0.0293, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 4.582716049382716, | |
| "grad_norm": 0.4003695845603943, | |
| "learning_rate": 1.584823893886933e-06, | |
| "loss": 0.0321, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 4.622222222222222, | |
| "grad_norm": 0.3148975372314453, | |
| "learning_rate": 1.2536043909088191e-06, | |
| "loss": 0.0259, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 4.661728395061728, | |
| "grad_norm": 0.49285128712654114, | |
| "learning_rate": 9.607359798384785e-07, | |
| "loss": 0.0545, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 4.701234567901235, | |
| "grad_norm": 0.2844616770744324, | |
| "learning_rate": 7.064490740882057e-07, | |
| "loss": 0.0246, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 4.7407407407407405, | |
| "grad_norm": 0.4715191125869751, | |
| "learning_rate": 4.909437331777179e-07, | |
| "loss": 0.0504, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.7407407407407405, | |
| "eval_loss": 0.13547781109809875, | |
| "eval_runtime": 24.9411, | |
| "eval_samples_per_second": 1.804, | |
| "eval_steps_per_second": 1.804, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.780246913580247, | |
| "grad_norm": 0.46037256717681885, | |
| "learning_rate": 3.143895053378698e-07, | |
| "loss": 0.0412, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 4.8197530864197535, | |
| "grad_norm": 0.302042692899704, | |
| "learning_rate": 1.7692529411904578e-07, | |
| "loss": 0.0292, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 4.859259259259259, | |
| "grad_norm": 0.45368897914886475, | |
| "learning_rate": 7.865924910916977e-08, | |
| "loss": 0.0333, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 4.898765432098766, | |
| "grad_norm": 0.492509126663208, | |
| "learning_rate": 1.9668680847356735e-08, | |
| "loss": 0.0573, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 4.938271604938271, | |
| "grad_norm": 0.35250523686408997, | |
| "learning_rate": 0.0, | |
| "loss": 0.0437, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 4.938271604938271, | |
| "step": 125, | |
| "total_flos": 1.3189127055763046e+17, | |
| "train_loss": 0.0907540387660265, | |
| "train_runtime": 3926.6465, | |
| "train_samples_per_second": 0.516, | |
| "train_steps_per_second": 0.032 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 125, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 10, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.3189127055763046e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |