| { | |
| "best_metric": 0.1026734858751297, | |
| "best_model_checkpoint": "saves/psy-course-doc/Qwen2.5-7B-Instruct/train/fold7/checkpoint-120", | |
| "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": 0.18254302442073822, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.249, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.07901234567901234, | |
| "grad_norm": 0.19255198538303375, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.2546, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.11851851851851852, | |
| "grad_norm": 0.21811692416667938, | |
| "learning_rate": 2.307692307692308e-05, | |
| "loss": 0.2027, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.1580246913580247, | |
| "grad_norm": 0.18568770587444305, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.2394, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.19753086419753085, | |
| "grad_norm": 0.20259220898151398, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 0.2332, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.23703703703703705, | |
| "grad_norm": 0.1811724156141281, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 0.2588, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.2765432098765432, | |
| "grad_norm": 0.1681864857673645, | |
| "learning_rate": 5.384615384615385e-05, | |
| "loss": 0.3108, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.3160493827160494, | |
| "grad_norm": 0.20345968008041382, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.2401, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 0.16792084276676178, | |
| "learning_rate": 6.923076923076924e-05, | |
| "loss": 0.1706, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "grad_norm": 0.16801084578037262, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.2238, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "eval_loss": 0.21720756590366364, | |
| "eval_runtime": 21.4425, | |
| "eval_samples_per_second": 2.099, | |
| "eval_steps_per_second": 2.099, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.4345679012345679, | |
| "grad_norm": 0.16402322053909302, | |
| "learning_rate": 8.461538461538461e-05, | |
| "loss": 0.16, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.4740740740740741, | |
| "grad_norm": 0.13382259011268616, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.2222, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.5135802469135803, | |
| "grad_norm": 0.10727668553590775, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1305, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.5530864197530864, | |
| "grad_norm": 0.10486454516649246, | |
| "learning_rate": 9.998033131915266e-05, | |
| "loss": 0.2377, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 0.09054844826459885, | |
| "learning_rate": 9.992134075089084e-05, | |
| "loss": 0.1441, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.6320987654320988, | |
| "grad_norm": 0.11283206939697266, | |
| "learning_rate": 9.982307470588098e-05, | |
| "loss": 0.1083, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.671604938271605, | |
| "grad_norm": 0.10074640810489655, | |
| "learning_rate": 9.968561049466214e-05, | |
| "loss": 0.1328, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 0.09744096547365189, | |
| "learning_rate": 9.950905626682228e-05, | |
| "loss": 0.1598, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.7506172839506173, | |
| "grad_norm": 0.1034897044301033, | |
| "learning_rate": 9.92935509259118e-05, | |
| "loss": 0.1876, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "grad_norm": 0.0838492214679718, | |
| "learning_rate": 9.903926402016153e-05, | |
| "loss": 0.0965, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "eval_loss": 0.1522609144449234, | |
| "eval_runtime": 21.4629, | |
| "eval_samples_per_second": 2.097, | |
| "eval_steps_per_second": 2.097, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.8296296296296296, | |
| "grad_norm": 0.09693557769060135, | |
| "learning_rate": 9.874639560909117e-05, | |
| "loss": 0.1093, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.8691358024691358, | |
| "grad_norm": 0.09759286791086197, | |
| "learning_rate": 9.841517610611309e-05, | |
| "loss": 0.1077, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.908641975308642, | |
| "grad_norm": 0.10823535919189453, | |
| "learning_rate": 9.804586609725499e-05, | |
| "loss": 0.1482, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.9481481481481482, | |
| "grad_norm": 0.11989855021238327, | |
| "learning_rate": 9.763875613614482e-05, | |
| "loss": 0.1383, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.9876543209876543, | |
| "grad_norm": 0.08927088230848312, | |
| "learning_rate": 9.719416651541839e-05, | |
| "loss": 0.1433, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.0271604938271606, | |
| "grad_norm": 0.14619426429271698, | |
| "learning_rate": 9.671244701472999e-05, | |
| "loss": 0.225, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.0666666666666667, | |
| "grad_norm": 0.07501505315303802, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 0.1316, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.106172839506173, | |
| "grad_norm": 0.0710507184267044, | |
| "learning_rate": 9.563916325306594e-05, | |
| "loss": 0.1046, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.145679012345679, | |
| "grad_norm": 0.06751503050327301, | |
| "learning_rate": 9.504844339512095e-05, | |
| "loss": 0.0878, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "grad_norm": 0.08156848698854446, | |
| "learning_rate": 9.442228179894362e-05, | |
| "loss": 0.1337, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "eval_loss": 0.12314464896917343, | |
| "eval_runtime": 21.4534, | |
| "eval_samples_per_second": 2.098, | |
| "eval_steps_per_second": 2.098, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.2246913580246914, | |
| "grad_norm": 0.06688881665468216, | |
| "learning_rate": 9.376117109543769e-05, | |
| "loss": 0.0864, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.2641975308641975, | |
| "grad_norm": 0.08303424715995789, | |
| "learning_rate": 9.306563141162046e-05, | |
| "loss": 0.1349, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.3037037037037038, | |
| "grad_norm": 0.07813991606235504, | |
| "learning_rate": 9.233620996141421e-05, | |
| "loss": 0.0617, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.34320987654321, | |
| "grad_norm": 0.06176231428980827, | |
| "learning_rate": 9.157348061512727e-05, | |
| "loss": 0.0448, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.382716049382716, | |
| "grad_norm": 0.09021872282028198, | |
| "learning_rate": 9.077804344796302e-05, | |
| "loss": 0.135, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.4222222222222223, | |
| "grad_norm": 0.07563682645559311, | |
| "learning_rate": 8.995052426791247e-05, | |
| "loss": 0.0867, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.4617283950617284, | |
| "grad_norm": 0.08462163060903549, | |
| "learning_rate": 8.90915741234015e-05, | |
| "loss": 0.1201, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.5012345679012347, | |
| "grad_norm": 0.0880044475197792, | |
| "learning_rate": 8.820186879108038e-05, | |
| "loss": 0.0703, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.5407407407407407, | |
| "grad_norm": 0.08864600956439972, | |
| "learning_rate": 8.728210824415827e-05, | |
| "loss": 0.0951, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.5802469135802468, | |
| "grad_norm": 0.0742451623082161, | |
| "learning_rate": 8.633301610170135e-05, | |
| "loss": 0.0717, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.5802469135802468, | |
| "eval_loss": 0.11394847184419632, | |
| "eval_runtime": 21.4575, | |
| "eval_samples_per_second": 2.097, | |
| "eval_steps_per_second": 2.097, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.6197530864197531, | |
| "grad_norm": 0.08356467634439468, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 0.0822, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.6592592592592592, | |
| "grad_norm": 0.05892196297645569, | |
| "learning_rate": 8.434984630174509e-05, | |
| "loss": 0.0749, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.6987654320987655, | |
| "grad_norm": 0.07285855710506439, | |
| "learning_rate": 8.33173288976002e-05, | |
| "loss": 0.0966, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.7382716049382716, | |
| "grad_norm": 0.06478698551654816, | |
| "learning_rate": 8.225859917710439e-05, | |
| "loss": 0.1146, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.08161158859729767, | |
| "learning_rate": 8.117449009293668e-05, | |
| "loss": 0.0784, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.817283950617284, | |
| "grad_norm": 0.10112543404102325, | |
| "learning_rate": 8.006585456492029e-05, | |
| "loss": 0.0687, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.8567901234567903, | |
| "grad_norm": 0.07533222436904907, | |
| "learning_rate": 7.89335648089903e-05, | |
| "loss": 0.1616, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.8962962962962964, | |
| "grad_norm": 0.06019189953804016, | |
| "learning_rate": 7.777851165098012e-05, | |
| "loss": 0.0681, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.9358024691358025, | |
| "grad_norm": 0.07877020537853241, | |
| "learning_rate": 7.660160382576683e-05, | |
| "loss": 0.1099, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "grad_norm": 0.07968045771121979, | |
| "learning_rate": 7.540376726232648e-05, | |
| "loss": 0.1129, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "eval_loss": 0.10940594971179962, | |
| "eval_runtime": 21.4667, | |
| "eval_samples_per_second": 2.096, | |
| "eval_steps_per_second": 2.096, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.0148148148148146, | |
| "grad_norm": 0.11678477376699448, | |
| "learning_rate": 7.4185944355262e-05, | |
| "loss": 0.0936, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 2.054320987654321, | |
| "grad_norm": 0.06708088517189026, | |
| "learning_rate": 7.294909322337689e-05, | |
| "loss": 0.0969, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 2.093827160493827, | |
| "grad_norm": 0.05962765961885452, | |
| "learning_rate": 7.169418695587791e-05, | |
| "loss": 0.0714, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.1333333333333333, | |
| "grad_norm": 0.07481212168931961, | |
| "learning_rate": 7.042221284679982e-05, | |
| "loss": 0.1051, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.1728395061728394, | |
| "grad_norm": 0.084474578499794, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 0.1517, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.212345679012346, | |
| "grad_norm": 0.058137793093919754, | |
| "learning_rate": 6.783107663311565e-05, | |
| "loss": 0.1043, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.251851851851852, | |
| "grad_norm": 0.07785182446241379, | |
| "learning_rate": 6.651395309775837e-05, | |
| "loss": 0.1568, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.291358024691358, | |
| "grad_norm": 0.06690663844347, | |
| "learning_rate": 6.518383725548074e-05, | |
| "loss": 0.1159, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.330864197530864, | |
| "grad_norm": 0.055363718420267105, | |
| "learning_rate": 6.384177557124247e-05, | |
| "loss": 0.0656, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "grad_norm": 0.06924538314342499, | |
| "learning_rate": 6.248882390836135e-05, | |
| "loss": 0.0416, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "eval_loss": 0.1061331033706665, | |
| "eval_runtime": 21.4499, | |
| "eval_samples_per_second": 2.098, | |
| "eval_steps_per_second": 2.098, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.4098765432098768, | |
| "grad_norm": 0.06964460015296936, | |
| "learning_rate": 6.112604669781572e-05, | |
| "loss": 0.1018, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.449382716049383, | |
| "grad_norm": 0.059162598103284836, | |
| "learning_rate": 5.9754516100806423e-05, | |
| "loss": 0.0692, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.488888888888889, | |
| "grad_norm": 0.06618830561637878, | |
| "learning_rate": 5.837531116523682e-05, | |
| "loss": 0.0624, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.528395061728395, | |
| "grad_norm": 0.06807179003953934, | |
| "learning_rate": 5.698951697677498e-05, | |
| "loss": 0.0739, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.567901234567901, | |
| "grad_norm": 0.09343565255403519, | |
| "learning_rate": 5.559822380516539e-05, | |
| "loss": 0.1511, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.6074074074074076, | |
| "grad_norm": 0.07874759286642075, | |
| "learning_rate": 5.420252624646238e-05, | |
| "loss": 0.0667, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.6469135802469137, | |
| "grad_norm": 0.0627421885728836, | |
| "learning_rate": 5.2803522361859594e-05, | |
| "loss": 0.0811, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.68641975308642, | |
| "grad_norm": 0.05986185371875763, | |
| "learning_rate": 5.140231281379345e-05, | |
| "loss": 0.0449, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.725925925925926, | |
| "grad_norm": 0.06414607912302017, | |
| "learning_rate": 5e-05, | |
| "loss": 0.0241, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.765432098765432, | |
| "grad_norm": 0.08229418098926544, | |
| "learning_rate": 4.859768718620656e-05, | |
| "loss": 0.0564, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.765432098765432, | |
| "eval_loss": 0.10523843765258789, | |
| "eval_runtime": 21.457, | |
| "eval_samples_per_second": 2.097, | |
| "eval_steps_per_second": 2.097, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.8049382716049385, | |
| "grad_norm": 0.07865197956562042, | |
| "learning_rate": 4.7196477638140404e-05, | |
| "loss": 0.0607, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.8444444444444446, | |
| "grad_norm": 0.0636526495218277, | |
| "learning_rate": 4.579747375353763e-05, | |
| "loss": 0.0499, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.8839506172839506, | |
| "grad_norm": 0.06069548800587654, | |
| "learning_rate": 4.4401776194834613e-05, | |
| "loss": 0.0649, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.9234567901234567, | |
| "grad_norm": 0.08193910121917725, | |
| "learning_rate": 4.3010483023225045e-05, | |
| "loss": 0.0521, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.962962962962963, | |
| "grad_norm": 0.05500645190477371, | |
| "learning_rate": 4.162468883476319e-05, | |
| "loss": 0.0675, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 3.0024691358024693, | |
| "grad_norm": 0.148792564868927, | |
| "learning_rate": 4.0245483899193595e-05, | |
| "loss": 0.1372, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 3.0419753086419754, | |
| "grad_norm": 0.07226116955280304, | |
| "learning_rate": 3.887395330218429e-05, | |
| "loss": 0.1222, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 3.0814814814814815, | |
| "grad_norm": 0.05812438949942589, | |
| "learning_rate": 3.7511176091638653e-05, | |
| "loss": 0.0919, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 3.1209876543209876, | |
| "grad_norm": 0.061713822185993195, | |
| "learning_rate": 3.6158224428757535e-05, | |
| "loss": 0.0607, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 3.1604938271604937, | |
| "grad_norm": 0.07338086515665054, | |
| "learning_rate": 3.4816162744519263e-05, | |
| "loss": 0.1234, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.1604938271604937, | |
| "eval_loss": 0.10379262268543243, | |
| "eval_runtime": 21.4768, | |
| "eval_samples_per_second": 2.095, | |
| "eval_steps_per_second": 2.095, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.07400018721818924, | |
| "learning_rate": 3.3486046902241664e-05, | |
| "loss": 0.0843, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 3.2395061728395063, | |
| "grad_norm": 0.06821998208761215, | |
| "learning_rate": 3.216892336688435e-05, | |
| "loss": 0.0912, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 3.2790123456790123, | |
| "grad_norm": 0.0646594688296318, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 0.0998, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 3.3185185185185184, | |
| "grad_norm": 0.0627719834446907, | |
| "learning_rate": 2.9577787153200197e-05, | |
| "loss": 0.1013, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 3.3580246913580245, | |
| "grad_norm": 0.083916574716568, | |
| "learning_rate": 2.8305813044122097e-05, | |
| "loss": 0.0534, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 3.397530864197531, | |
| "grad_norm": 0.06759930402040482, | |
| "learning_rate": 2.705090677662311e-05, | |
| "loss": 0.0633, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 3.437037037037037, | |
| "grad_norm": 0.06876127421855927, | |
| "learning_rate": 2.581405564473801e-05, | |
| "loss": 0.0712, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 3.476543209876543, | |
| "grad_norm": 0.06895948946475983, | |
| "learning_rate": 2.459623273767354e-05, | |
| "loss": 0.0869, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 3.5160493827160493, | |
| "grad_norm": 0.06265565007925034, | |
| "learning_rate": 2.3398396174233178e-05, | |
| "loss": 0.0729, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "grad_norm": 0.07499140501022339, | |
| "learning_rate": 2.2221488349019903e-05, | |
| "loss": 0.0923, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "eval_loss": 0.10331548005342484, | |
| "eval_runtime": 21.4637, | |
| "eval_samples_per_second": 2.097, | |
| "eval_steps_per_second": 2.097, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.595061728395062, | |
| "grad_norm": 0.06450280547142029, | |
| "learning_rate": 2.1066435191009715e-05, | |
| "loss": 0.0699, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 3.634567901234568, | |
| "grad_norm": 0.06280648708343506, | |
| "learning_rate": 1.9934145435079702e-05, | |
| "loss": 0.0735, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 3.674074074074074, | |
| "grad_norm": 0.0673796758055687, | |
| "learning_rate": 1.8825509907063327e-05, | |
| "loss": 0.0398, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 3.71358024691358, | |
| "grad_norm": 0.07176631689071655, | |
| "learning_rate": 1.774140082289563e-05, | |
| "loss": 0.0454, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 3.753086419753086, | |
| "grad_norm": 0.07445023208856583, | |
| "learning_rate": 1.6682671102399805e-05, | |
| "loss": 0.0441, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 3.7925925925925927, | |
| "grad_norm": 0.06666605919599533, | |
| "learning_rate": 1.5650153698254916e-05, | |
| "loss": 0.0146, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 3.832098765432099, | |
| "grad_norm": 0.06168539449572563, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 0.0527, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 3.871604938271605, | |
| "grad_norm": 0.06918106228113174, | |
| "learning_rate": 1.3666983898298657e-05, | |
| "loss": 0.1021, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 3.911111111111111, | |
| "grad_norm": 0.07920930534601212, | |
| "learning_rate": 1.2717891755841722e-05, | |
| "loss": 0.0898, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 3.950617283950617, | |
| "grad_norm": 0.08788114786148071, | |
| "learning_rate": 1.1798131208919627e-05, | |
| "loss": 0.0648, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.950617283950617, | |
| "eval_loss": 0.1033155769109726, | |
| "eval_runtime": 21.4905, | |
| "eval_samples_per_second": 2.094, | |
| "eval_steps_per_second": 2.094, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.9901234567901236, | |
| "grad_norm": 0.0996319055557251, | |
| "learning_rate": 1.090842587659851e-05, | |
| "loss": 0.1188, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 4.029629629629629, | |
| "grad_norm": 0.09905175864696503, | |
| "learning_rate": 1.004947573208756e-05, | |
| "loss": 0.0801, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 4.069135802469136, | |
| "grad_norm": 0.06089180335402489, | |
| "learning_rate": 9.221956552036992e-06, | |
| "loss": 0.0868, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 4.108641975308642, | |
| "grad_norm": 0.06779298931360245, | |
| "learning_rate": 8.426519384872733e-06, | |
| "loss": 0.105, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 4.148148148148148, | |
| "grad_norm": 0.06558673828840256, | |
| "learning_rate": 7.663790038585793e-06, | |
| "loss": 0.0895, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 4.187654320987654, | |
| "grad_norm": 0.05850369855761528, | |
| "learning_rate": 6.934368588379553e-06, | |
| "loss": 0.0593, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 4.22716049382716, | |
| "grad_norm": 0.0688764899969101, | |
| "learning_rate": 6.238828904562316e-06, | |
| "loss": 0.1019, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 4.266666666666667, | |
| "grad_norm": 0.06864060461521149, | |
| "learning_rate": 5.577718201056392e-06, | |
| "loss": 0.0553, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 4.306172839506173, | |
| "grad_norm": 0.06651941686868668, | |
| "learning_rate": 4.951556604879048e-06, | |
| "loss": 0.1117, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 4.345679012345679, | |
| "grad_norm": 0.06021948158740997, | |
| "learning_rate": 4.360836746934055e-06, | |
| "loss": 0.0595, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.345679012345679, | |
| "eval_loss": 0.10285650938749313, | |
| "eval_runtime": 21.5008, | |
| "eval_samples_per_second": 2.093, | |
| "eval_steps_per_second": 2.093, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.385185185185185, | |
| "grad_norm": 0.05625497177243233, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 0.0385, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 4.424691358024692, | |
| "grad_norm": 0.06548874080181122, | |
| "learning_rate": 3.2875529852700147e-06, | |
| "loss": 0.0541, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 4.4641975308641975, | |
| "grad_norm": 0.07624312490224838, | |
| "learning_rate": 2.8058334845816213e-06, | |
| "loss": 0.0539, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 4.503703703703704, | |
| "grad_norm": 0.06362943351268768, | |
| "learning_rate": 2.361243863855184e-06, | |
| "loss": 0.0753, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 4.54320987654321, | |
| "grad_norm": 0.07078000158071518, | |
| "learning_rate": 1.9541339027450256e-06, | |
| "loss": 0.0601, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 4.582716049382716, | |
| "grad_norm": 0.06137111410498619, | |
| "learning_rate": 1.584823893886933e-06, | |
| "loss": 0.0924, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 4.622222222222222, | |
| "grad_norm": 0.07251935452222824, | |
| "learning_rate": 1.2536043909088191e-06, | |
| "loss": 0.1117, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 4.661728395061728, | |
| "grad_norm": 0.053561560809612274, | |
| "learning_rate": 9.607359798384785e-07, | |
| "loss": 0.0454, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 4.701234567901235, | |
| "grad_norm": 0.06559063494205475, | |
| "learning_rate": 7.064490740882057e-07, | |
| "loss": 0.0771, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 4.7407407407407405, | |
| "grad_norm": 0.08432595431804657, | |
| "learning_rate": 4.909437331777179e-07, | |
| "loss": 0.1126, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.7407407407407405, | |
| "eval_loss": 0.1026734858751297, | |
| "eval_runtime": 21.4591, | |
| "eval_samples_per_second": 2.097, | |
| "eval_steps_per_second": 2.097, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.780246913580247, | |
| "grad_norm": 0.06667386740446091, | |
| "learning_rate": 3.143895053378698e-07, | |
| "loss": 0.073, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 4.8197530864197535, | |
| "grad_norm": 0.056891944259405136, | |
| "learning_rate": 1.7692529411904578e-07, | |
| "loss": 0.0455, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 4.859259259259259, | |
| "grad_norm": 0.06197427958250046, | |
| "learning_rate": 7.865924910916977e-08, | |
| "loss": 0.062, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 4.898765432098766, | |
| "grad_norm": 0.061731986701488495, | |
| "learning_rate": 1.9668680847356735e-08, | |
| "loss": 0.062, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 4.938271604938271, | |
| "grad_norm": 0.0614369660615921, | |
| "learning_rate": 0.0, | |
| "loss": 0.0807, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 4.938271604938271, | |
| "step": 125, | |
| "total_flos": 1.439468427860951e+17, | |
| "train_loss": 0.10445326153188944, | |
| "train_runtime": 3451.1861, | |
| "train_samples_per_second": 0.587, | |
| "train_steps_per_second": 0.036 | |
| } | |
| ], | |
| "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": 1.439468427860951e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |