| { | |
| "best_metric": 0.07241930067539215, | |
| "best_model_checkpoint": "saves/psy-course-doc/Ministral-8B-Instruct-2410/train/fold4/checkpoint-80", | |
| "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": 0.2759062647819519, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.1991, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.07901234567901234, | |
| "grad_norm": 0.23504853248596191, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.2013, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.11851851851851852, | |
| "grad_norm": 0.2665434181690216, | |
| "learning_rate": 2.307692307692308e-05, | |
| "loss": 0.2276, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.1580246913580247, | |
| "grad_norm": 0.28634434938430786, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.1577, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.19753086419753085, | |
| "grad_norm": 0.272355318069458, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 0.2437, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.23703703703703705, | |
| "grad_norm": 0.2578892409801483, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 0.1929, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.2765432098765432, | |
| "grad_norm": 0.31411013007164, | |
| "learning_rate": 5.384615384615385e-05, | |
| "loss": 0.2263, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.3160493827160494, | |
| "grad_norm": 0.22777129709720612, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.1575, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 0.215264230966568, | |
| "learning_rate": 6.923076923076924e-05, | |
| "loss": 0.1778, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "grad_norm": 0.2228308916091919, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.1224, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "eval_loss": 0.139120414853096, | |
| "eval_runtime": 22.8332, | |
| "eval_samples_per_second": 1.971, | |
| "eval_steps_per_second": 1.971, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.4345679012345679, | |
| "grad_norm": 0.23304517567157745, | |
| "learning_rate": 8.461538461538461e-05, | |
| "loss": 0.1216, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.4740740740740741, | |
| "grad_norm": 0.2072160243988037, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.1138, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.5135802469135803, | |
| "grad_norm": 0.18203678727149963, | |
| "learning_rate": 0.0001, | |
| "loss": 0.1307, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.5530864197530864, | |
| "grad_norm": 0.20423011481761932, | |
| "learning_rate": 9.998033131915266e-05, | |
| "loss": 0.1128, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 0.18817876279354095, | |
| "learning_rate": 9.992134075089084e-05, | |
| "loss": 0.1039, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.6320987654320988, | |
| "grad_norm": 0.169025719165802, | |
| "learning_rate": 9.982307470588098e-05, | |
| "loss": 0.0786, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.671604938271605, | |
| "grad_norm": 0.15945960581302643, | |
| "learning_rate": 9.968561049466214e-05, | |
| "loss": 0.1556, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 0.11714024841785431, | |
| "learning_rate": 9.950905626682228e-05, | |
| "loss": 0.1383, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.7506172839506173, | |
| "grad_norm": 0.11941976100206375, | |
| "learning_rate": 9.92935509259118e-05, | |
| "loss": 0.0782, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "grad_norm": 0.1326676309108734, | |
| "learning_rate": 9.903926402016153e-05, | |
| "loss": 0.1347, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "eval_loss": 0.09002771228551865, | |
| "eval_runtime": 22.9862, | |
| "eval_samples_per_second": 1.958, | |
| "eval_steps_per_second": 1.958, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.8296296296296296, | |
| "grad_norm": 0.11474258452653885, | |
| "learning_rate": 9.874639560909117e-05, | |
| "loss": 0.0798, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.8691358024691358, | |
| "grad_norm": 0.13447517156600952, | |
| "learning_rate": 9.841517610611309e-05, | |
| "loss": 0.0479, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.908641975308642, | |
| "grad_norm": 0.13942432403564453, | |
| "learning_rate": 9.804586609725499e-05, | |
| "loss": 0.0962, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.9481481481481482, | |
| "grad_norm": 0.1423346847295761, | |
| "learning_rate": 9.763875613614482e-05, | |
| "loss": 0.1014, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.9876543209876543, | |
| "grad_norm": 0.10825008898973465, | |
| "learning_rate": 9.719416651541839e-05, | |
| "loss": 0.0727, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.0271604938271606, | |
| "grad_norm": 0.14513178169727325, | |
| "learning_rate": 9.671244701472999e-05, | |
| "loss": 0.083, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.0666666666666667, | |
| "grad_norm": 0.12260837852954865, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 0.1091, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.106172839506173, | |
| "grad_norm": 0.10488468408584595, | |
| "learning_rate": 9.563916325306594e-05, | |
| "loss": 0.0823, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.145679012345679, | |
| "grad_norm": 0.13578175008296967, | |
| "learning_rate": 9.504844339512095e-05, | |
| "loss": 0.0723, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "grad_norm": 0.1257593184709549, | |
| "learning_rate": 9.442228179894362e-05, | |
| "loss": 0.0905, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "eval_loss": 0.0794222429394722, | |
| "eval_runtime": 23.0452, | |
| "eval_samples_per_second": 1.953, | |
| "eval_steps_per_second": 1.953, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.2246913580246914, | |
| "grad_norm": 0.1313965618610382, | |
| "learning_rate": 9.376117109543769e-05, | |
| "loss": 0.0944, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.2641975308641975, | |
| "grad_norm": 0.13362035155296326, | |
| "learning_rate": 9.306563141162046e-05, | |
| "loss": 0.0595, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.3037037037037038, | |
| "grad_norm": 0.11342867463827133, | |
| "learning_rate": 9.233620996141421e-05, | |
| "loss": 0.04, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.34320987654321, | |
| "grad_norm": 0.11743198335170746, | |
| "learning_rate": 9.157348061512727e-05, | |
| "loss": 0.0846, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.382716049382716, | |
| "grad_norm": 0.09698523581027985, | |
| "learning_rate": 9.077804344796302e-05, | |
| "loss": 0.0334, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.4222222222222223, | |
| "grad_norm": 0.11977563798427582, | |
| "learning_rate": 8.995052426791247e-05, | |
| "loss": 0.0763, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.4617283950617284, | |
| "grad_norm": 0.13985329866409302, | |
| "learning_rate": 8.90915741234015e-05, | |
| "loss": 0.1221, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.5012345679012347, | |
| "grad_norm": 0.13668206334114075, | |
| "learning_rate": 8.820186879108038e-05, | |
| "loss": 0.076, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.5407407407407407, | |
| "grad_norm": 0.08504356443881989, | |
| "learning_rate": 8.728210824415827e-05, | |
| "loss": 0.0702, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.5802469135802468, | |
| "grad_norm": 0.10752835869789124, | |
| "learning_rate": 8.633301610170135e-05, | |
| "loss": 0.0727, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.5802469135802468, | |
| "eval_loss": 0.07614804059267044, | |
| "eval_runtime": 23.0771, | |
| "eval_samples_per_second": 1.95, | |
| "eval_steps_per_second": 1.95, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.6197530864197531, | |
| "grad_norm": 0.14249514043331146, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 0.087, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.6592592592592592, | |
| "grad_norm": 0.09899597615003586, | |
| "learning_rate": 8.434984630174509e-05, | |
| "loss": 0.0498, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.6987654320987655, | |
| "grad_norm": 0.09157754480838776, | |
| "learning_rate": 8.33173288976002e-05, | |
| "loss": 0.0498, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.7382716049382716, | |
| "grad_norm": 0.10479103773832321, | |
| "learning_rate": 8.225859917710439e-05, | |
| "loss": 0.0656, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.11315307021141052, | |
| "learning_rate": 8.117449009293668e-05, | |
| "loss": 0.0578, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.817283950617284, | |
| "grad_norm": 0.10360617935657501, | |
| "learning_rate": 8.006585456492029e-05, | |
| "loss": 0.0918, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.8567901234567903, | |
| "grad_norm": 0.09683068096637726, | |
| "learning_rate": 7.89335648089903e-05, | |
| "loss": 0.1001, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.8962962962962964, | |
| "grad_norm": 0.10199464857578278, | |
| "learning_rate": 7.777851165098012e-05, | |
| "loss": 0.0845, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.9358024691358025, | |
| "grad_norm": 0.09505069255828857, | |
| "learning_rate": 7.660160382576683e-05, | |
| "loss": 0.0452, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "grad_norm": 0.08954370766878128, | |
| "learning_rate": 7.540376726232648e-05, | |
| "loss": 0.0255, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "eval_loss": 0.07384096086025238, | |
| "eval_runtime": 23.0735, | |
| "eval_samples_per_second": 1.95, | |
| "eval_steps_per_second": 1.95, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.0148148148148146, | |
| "grad_norm": 0.08641216158866882, | |
| "learning_rate": 7.4185944355262e-05, | |
| "loss": 0.0667, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 2.054320987654321, | |
| "grad_norm": 0.08992330729961395, | |
| "learning_rate": 7.294909322337689e-05, | |
| "loss": 0.0909, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 2.093827160493827, | |
| "grad_norm": 0.09697853773832321, | |
| "learning_rate": 7.169418695587791e-05, | |
| "loss": 0.0767, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.1333333333333333, | |
| "grad_norm": 0.07834423333406448, | |
| "learning_rate": 7.042221284679982e-05, | |
| "loss": 0.0388, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.1728395061728394, | |
| "grad_norm": 0.0919504389166832, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 0.0459, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.212345679012346, | |
| "grad_norm": 0.1273844838142395, | |
| "learning_rate": 6.783107663311565e-05, | |
| "loss": 0.0693, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.251851851851852, | |
| "grad_norm": 0.10880222171545029, | |
| "learning_rate": 6.651395309775837e-05, | |
| "loss": 0.1114, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.291358024691358, | |
| "grad_norm": 0.19058740139007568, | |
| "learning_rate": 6.518383725548074e-05, | |
| "loss": 0.1112, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.330864197530864, | |
| "grad_norm": 0.08417613059282303, | |
| "learning_rate": 6.384177557124247e-05, | |
| "loss": 0.0457, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "grad_norm": 0.0867931917309761, | |
| "learning_rate": 6.248882390836135e-05, | |
| "loss": 0.0285, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "eval_loss": 0.07262618094682693, | |
| "eval_runtime": 23.0996, | |
| "eval_samples_per_second": 1.948, | |
| "eval_steps_per_second": 1.948, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.4098765432098768, | |
| "grad_norm": 0.08550772070884705, | |
| "learning_rate": 6.112604669781572e-05, | |
| "loss": 0.0518, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.449382716049383, | |
| "grad_norm": 0.0737406313419342, | |
| "learning_rate": 5.9754516100806423e-05, | |
| "loss": 0.0328, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.488888888888889, | |
| "grad_norm": 0.09175753593444824, | |
| "learning_rate": 5.837531116523682e-05, | |
| "loss": 0.0461, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.528395061728395, | |
| "grad_norm": 0.10679838061332703, | |
| "learning_rate": 5.698951697677498e-05, | |
| "loss": 0.0636, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.567901234567901, | |
| "grad_norm": 0.15195432305335999, | |
| "learning_rate": 5.559822380516539e-05, | |
| "loss": 0.0824, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.6074074074074076, | |
| "grad_norm": 0.10324520617723465, | |
| "learning_rate": 5.420252624646238e-05, | |
| "loss": 0.0433, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.6469135802469137, | |
| "grad_norm": 0.10497809201478958, | |
| "learning_rate": 5.2803522361859594e-05, | |
| "loss": 0.0798, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.68641975308642, | |
| "grad_norm": 0.12391440570354462, | |
| "learning_rate": 5.140231281379345e-05, | |
| "loss": 0.0599, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.725925925925926, | |
| "grad_norm": 0.11880259215831757, | |
| "learning_rate": 5e-05, | |
| "loss": 0.0432, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.765432098765432, | |
| "grad_norm": 0.12748436629772186, | |
| "learning_rate": 4.859768718620656e-05, | |
| "loss": 0.04, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.765432098765432, | |
| "eval_loss": 0.0727304071187973, | |
| "eval_runtime": 23.1172, | |
| "eval_samples_per_second": 1.947, | |
| "eval_steps_per_second": 1.947, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.8049382716049385, | |
| "grad_norm": 0.11074857413768768, | |
| "learning_rate": 4.7196477638140404e-05, | |
| "loss": 0.0289, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.8444444444444446, | |
| "grad_norm": 0.13451829552650452, | |
| "learning_rate": 4.579747375353763e-05, | |
| "loss": 0.0555, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.8839506172839506, | |
| "grad_norm": 0.11675214767456055, | |
| "learning_rate": 4.4401776194834613e-05, | |
| "loss": 0.0476, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.9234567901234567, | |
| "grad_norm": 0.11922719329595566, | |
| "learning_rate": 4.3010483023225045e-05, | |
| "loss": 0.0652, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.962962962962963, | |
| "grad_norm": 0.1386060118675232, | |
| "learning_rate": 4.162468883476319e-05, | |
| "loss": 0.1127, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 3.0024691358024693, | |
| "grad_norm": 0.10950551927089691, | |
| "learning_rate": 4.0245483899193595e-05, | |
| "loss": 0.0581, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 3.0419753086419754, | |
| "grad_norm": 0.09093952924013138, | |
| "learning_rate": 3.887395330218429e-05, | |
| "loss": 0.0488, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 3.0814814814814815, | |
| "grad_norm": 0.10329506546258926, | |
| "learning_rate": 3.7511176091638653e-05, | |
| "loss": 0.0625, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 3.1209876543209876, | |
| "grad_norm": 0.10993586480617523, | |
| "learning_rate": 3.6158224428757535e-05, | |
| "loss": 0.0735, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 3.1604938271604937, | |
| "grad_norm": 0.12944099307060242, | |
| "learning_rate": 3.4816162744519263e-05, | |
| "loss": 0.0868, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.1604938271604937, | |
| "eval_loss": 0.07241930067539215, | |
| "eval_runtime": 23.1391, | |
| "eval_samples_per_second": 1.945, | |
| "eval_steps_per_second": 1.945, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.10336323827505112, | |
| "learning_rate": 3.3486046902241664e-05, | |
| "loss": 0.0735, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 3.2395061728395063, | |
| "grad_norm": 0.09130258858203888, | |
| "learning_rate": 3.216892336688435e-05, | |
| "loss": 0.0308, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 3.2790123456790123, | |
| "grad_norm": 0.10697068274021149, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 0.0701, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 3.3185185185185184, | |
| "grad_norm": 0.09463872760534286, | |
| "learning_rate": 2.9577787153200197e-05, | |
| "loss": 0.0354, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 3.3580246913580245, | |
| "grad_norm": 0.11037807166576385, | |
| "learning_rate": 2.8305813044122097e-05, | |
| "loss": 0.0433, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 3.397530864197531, | |
| "grad_norm": 0.1262778341770172, | |
| "learning_rate": 2.705090677662311e-05, | |
| "loss": 0.0817, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 3.437037037037037, | |
| "grad_norm": 0.10730726271867752, | |
| "learning_rate": 2.581405564473801e-05, | |
| "loss": 0.0595, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 3.476543209876543, | |
| "grad_norm": 0.13701090216636658, | |
| "learning_rate": 2.459623273767354e-05, | |
| "loss": 0.0606, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 3.5160493827160493, | |
| "grad_norm": 0.10309536010026932, | |
| "learning_rate": 2.3398396174233178e-05, | |
| "loss": 0.0489, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "grad_norm": 0.14796820282936096, | |
| "learning_rate": 2.2221488349019903e-05, | |
| "loss": 0.0345, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "eval_loss": 0.07342290133237839, | |
| "eval_runtime": 23.1747, | |
| "eval_samples_per_second": 1.942, | |
| "eval_steps_per_second": 1.942, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.595061728395062, | |
| "grad_norm": 0.13698893785476685, | |
| "learning_rate": 2.1066435191009715e-05, | |
| "loss": 0.0757, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 3.634567901234568, | |
| "grad_norm": 0.14148332178592682, | |
| "learning_rate": 1.9934145435079702e-05, | |
| "loss": 0.0664, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 3.674074074074074, | |
| "grad_norm": 0.11668102443218231, | |
| "learning_rate": 1.8825509907063327e-05, | |
| "loss": 0.0494, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 3.71358024691358, | |
| "grad_norm": 0.10136590152978897, | |
| "learning_rate": 1.774140082289563e-05, | |
| "loss": 0.0354, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 3.753086419753086, | |
| "grad_norm": 0.09533093869686127, | |
| "learning_rate": 1.6682671102399805e-05, | |
| "loss": 0.0276, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 3.7925925925925927, | |
| "grad_norm": 0.08561649918556213, | |
| "learning_rate": 1.5650153698254916e-05, | |
| "loss": 0.0147, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 3.832098765432099, | |
| "grad_norm": 0.12960147857666016, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 0.0681, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 3.871604938271605, | |
| "grad_norm": 0.11553506553173065, | |
| "learning_rate": 1.3666983898298657e-05, | |
| "loss": 0.04, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 3.911111111111111, | |
| "grad_norm": 0.1409192532300949, | |
| "learning_rate": 1.2717891755841722e-05, | |
| "loss": 0.0773, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 3.950617283950617, | |
| "grad_norm": 0.11355224251747131, | |
| "learning_rate": 1.1798131208919627e-05, | |
| "loss": 0.0422, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.950617283950617, | |
| "eval_loss": 0.0737890675663948, | |
| "eval_runtime": 23.1469, | |
| "eval_samples_per_second": 1.944, | |
| "eval_steps_per_second": 1.944, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.9901234567901236, | |
| "grad_norm": 0.10671416670084, | |
| "learning_rate": 1.090842587659851e-05, | |
| "loss": 0.0363, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 4.029629629629629, | |
| "grad_norm": 0.14172357320785522, | |
| "learning_rate": 1.004947573208756e-05, | |
| "loss": 0.0999, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 4.069135802469136, | |
| "grad_norm": 0.10976026952266693, | |
| "learning_rate": 9.221956552036992e-06, | |
| "loss": 0.0414, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 4.108641975308642, | |
| "grad_norm": 0.10173307359218597, | |
| "learning_rate": 8.426519384872733e-06, | |
| "loss": 0.0218, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 4.148148148148148, | |
| "grad_norm": 0.1254478394985199, | |
| "learning_rate": 7.663790038585793e-06, | |
| "loss": 0.0684, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 4.187654320987654, | |
| "grad_norm": 0.08631468564271927, | |
| "learning_rate": 6.934368588379553e-06, | |
| "loss": 0.017, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 4.22716049382716, | |
| "grad_norm": 0.1372748166322708, | |
| "learning_rate": 6.238828904562316e-06, | |
| "loss": 0.0738, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 4.266666666666667, | |
| "grad_norm": 0.10199376940727234, | |
| "learning_rate": 5.577718201056392e-06, | |
| "loss": 0.0259, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 4.306172839506173, | |
| "grad_norm": 0.11291414499282837, | |
| "learning_rate": 4.951556604879048e-06, | |
| "loss": 0.0471, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 4.345679012345679, | |
| "grad_norm": 0.12336195260286331, | |
| "learning_rate": 4.360836746934055e-06, | |
| "loss": 0.0526, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.345679012345679, | |
| "eval_loss": 0.07390189915895462, | |
| "eval_runtime": 23.1576, | |
| "eval_samples_per_second": 1.943, | |
| "eval_steps_per_second": 1.943, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.385185185185185, | |
| "grad_norm": 0.11578144878149033, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 0.0513, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 4.424691358024692, | |
| "grad_norm": 0.11670077592134476, | |
| "learning_rate": 3.2875529852700147e-06, | |
| "loss": 0.0461, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 4.4641975308641975, | |
| "grad_norm": 0.11630329489707947, | |
| "learning_rate": 2.8058334845816213e-06, | |
| "loss": 0.0487, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 4.503703703703704, | |
| "grad_norm": 0.11543247848749161, | |
| "learning_rate": 2.361243863855184e-06, | |
| "loss": 0.0489, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 4.54320987654321, | |
| "grad_norm": 0.11423362046480179, | |
| "learning_rate": 1.9541339027450256e-06, | |
| "loss": 0.0505, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 4.582716049382716, | |
| "grad_norm": 0.13560818135738373, | |
| "learning_rate": 1.584823893886933e-06, | |
| "loss": 0.0766, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 4.622222222222222, | |
| "grad_norm": 0.13224148750305176, | |
| "learning_rate": 1.2536043909088191e-06, | |
| "loss": 0.0551, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 4.661728395061728, | |
| "grad_norm": 0.09404325485229492, | |
| "learning_rate": 9.607359798384785e-07, | |
| "loss": 0.0187, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 4.701234567901235, | |
| "grad_norm": 0.14365845918655396, | |
| "learning_rate": 7.064490740882057e-07, | |
| "loss": 0.0736, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 4.7407407407407405, | |
| "grad_norm": 0.11218550056219101, | |
| "learning_rate": 4.909437331777179e-07, | |
| "loss": 0.0395, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.7407407407407405, | |
| "eval_loss": 0.07408212870359421, | |
| "eval_runtime": 23.1587, | |
| "eval_samples_per_second": 1.943, | |
| "eval_steps_per_second": 1.943, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.780246913580247, | |
| "grad_norm": 0.09783432632684708, | |
| "learning_rate": 3.143895053378698e-07, | |
| "loss": 0.0313, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 4.8197530864197535, | |
| "grad_norm": 0.12950460612773895, | |
| "learning_rate": 1.7692529411904578e-07, | |
| "loss": 0.0539, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 4.859259259259259, | |
| "grad_norm": 0.121174156665802, | |
| "learning_rate": 7.865924910916977e-08, | |
| "loss": 0.0474, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 4.898765432098766, | |
| "grad_norm": 0.06541449576616287, | |
| "learning_rate": 1.9668680847356735e-08, | |
| "loss": 0.0078, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 4.938271604938271, | |
| "grad_norm": 0.11377643793821335, | |
| "learning_rate": 0.0, | |
| "loss": 0.0424, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 4.938271604938271, | |
| "step": 125, | |
| "total_flos": 1.5470265948615475e+17, | |
| "train_loss": 0.07499962299317121, | |
| "train_runtime": 3855.4334, | |
| "train_samples_per_second": 0.525, | |
| "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": 1.5470265948615475e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |