| { | |
| "best_metric": 0.13968898355960846, | |
| "best_model_checkpoint": "saves/psy-course-info/Llama3-OpenBioLLM-8B/train/fold8/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": 3.1236395835876465, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.8715, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.07901234567901234, | |
| "grad_norm": 2.6785013675689697, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 1.0293, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.11851851851851852, | |
| "grad_norm": 3.6286182403564453, | |
| "learning_rate": 2.307692307692308e-05, | |
| "loss": 0.982, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.1580246913580247, | |
| "grad_norm": 3.2573342323303223, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.9556, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.19753086419753085, | |
| "grad_norm": 2.9582366943359375, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 0.901, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.23703703703703705, | |
| "grad_norm": 3.263794422149658, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 0.8752, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.2765432098765432, | |
| "grad_norm": 2.6993696689605713, | |
| "learning_rate": 5.384615384615385e-05, | |
| "loss": 0.8065, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.3160493827160494, | |
| "grad_norm": 2.087186336517334, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.5756, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 1.6285301446914673, | |
| "learning_rate": 6.923076923076924e-05, | |
| "loss": 0.5358, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "grad_norm": 1.5133540630340576, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.5756, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "eval_loss": 0.4140697121620178, | |
| "eval_runtime": 20.6497, | |
| "eval_samples_per_second": 2.179, | |
| "eval_steps_per_second": 2.179, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.4345679012345679, | |
| "grad_norm": 1.2356938123703003, | |
| "learning_rate": 8.461538461538461e-05, | |
| "loss": 0.3745, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.4740740740740741, | |
| "grad_norm": 1.2282729148864746, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.5346, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.5135802469135803, | |
| "grad_norm": 1.1528704166412354, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3195, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.5530864197530864, | |
| "grad_norm": 0.9593966007232666, | |
| "learning_rate": 9.998033131915266e-05, | |
| "loss": 0.2525, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 0.9236360192298889, | |
| "learning_rate": 9.992134075089084e-05, | |
| "loss": 0.3034, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.6320987654320988, | |
| "grad_norm": 0.8577955365180969, | |
| "learning_rate": 9.982307470588098e-05, | |
| "loss": 0.2742, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.671604938271605, | |
| "grad_norm": 0.903700590133667, | |
| "learning_rate": 9.968561049466214e-05, | |
| "loss": 0.3221, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 0.7016167044639587, | |
| "learning_rate": 9.950905626682228e-05, | |
| "loss": 0.255, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.7506172839506173, | |
| "grad_norm": 1.0111489295959473, | |
| "learning_rate": 9.92935509259118e-05, | |
| "loss": 0.2668, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "grad_norm": 0.8751689791679382, | |
| "learning_rate": 9.903926402016153e-05, | |
| "loss": 0.3165, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "eval_loss": 0.22876985371112823, | |
| "eval_runtime": 20.5516, | |
| "eval_samples_per_second": 2.19, | |
| "eval_steps_per_second": 2.19, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.8296296296296296, | |
| "grad_norm": 0.9389567375183105, | |
| "learning_rate": 9.874639560909117e-05, | |
| "loss": 0.3185, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.8691358024691358, | |
| "grad_norm": 0.815375804901123, | |
| "learning_rate": 9.841517610611309e-05, | |
| "loss": 0.2159, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.908641975308642, | |
| "grad_norm": 0.6323608756065369, | |
| "learning_rate": 9.804586609725499e-05, | |
| "loss": 0.2124, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.9481481481481482, | |
| "grad_norm": 0.9609124660491943, | |
| "learning_rate": 9.763875613614482e-05, | |
| "loss": 0.334, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.9876543209876543, | |
| "grad_norm": 0.9019918441772461, | |
| "learning_rate": 9.719416651541839e-05, | |
| "loss": 0.2839, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.0271604938271606, | |
| "grad_norm": 1.4859174489974976, | |
| "learning_rate": 9.671244701472999e-05, | |
| "loss": 0.4747, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.0666666666666667, | |
| "grad_norm": 0.8583764433860779, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 0.1644, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.106172839506173, | |
| "grad_norm": 0.747122049331665, | |
| "learning_rate": 9.563916325306594e-05, | |
| "loss": 0.206, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.145679012345679, | |
| "grad_norm": 0.5967144966125488, | |
| "learning_rate": 9.504844339512095e-05, | |
| "loss": 0.186, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "grad_norm": 0.7630570530891418, | |
| "learning_rate": 9.442228179894362e-05, | |
| "loss": 0.2057, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "eval_loss": 0.1790097951889038, | |
| "eval_runtime": 20.5269, | |
| "eval_samples_per_second": 2.192, | |
| "eval_steps_per_second": 2.192, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.2246913580246914, | |
| "grad_norm": 0.5888400077819824, | |
| "learning_rate": 9.376117109543769e-05, | |
| "loss": 0.1741, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.2641975308641975, | |
| "grad_norm": 0.7971126437187195, | |
| "learning_rate": 9.306563141162046e-05, | |
| "loss": 0.2153, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.3037037037037038, | |
| "grad_norm": 0.8883408904075623, | |
| "learning_rate": 9.233620996141421e-05, | |
| "loss": 0.224, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.34320987654321, | |
| "grad_norm": 0.6711205840110779, | |
| "learning_rate": 9.157348061512727e-05, | |
| "loss": 0.1894, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.382716049382716, | |
| "grad_norm": 0.7529289126396179, | |
| "learning_rate": 9.077804344796302e-05, | |
| "loss": 0.1571, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.4222222222222223, | |
| "grad_norm": 0.5837678909301758, | |
| "learning_rate": 8.995052426791247e-05, | |
| "loss": 0.1339, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.4617283950617284, | |
| "grad_norm": 0.5557311773300171, | |
| "learning_rate": 8.90915741234015e-05, | |
| "loss": 0.1372, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.5012345679012347, | |
| "grad_norm": 0.9485201835632324, | |
| "learning_rate": 8.820186879108038e-05, | |
| "loss": 0.1594, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.5407407407407407, | |
| "grad_norm": 0.7968395352363586, | |
| "learning_rate": 8.728210824415827e-05, | |
| "loss": 0.1949, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.5802469135802468, | |
| "grad_norm": 0.8298124074935913, | |
| "learning_rate": 8.633301610170135e-05, | |
| "loss": 0.1753, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.5802469135802468, | |
| "eval_loss": 0.1530352234840393, | |
| "eval_runtime": 20.5406, | |
| "eval_samples_per_second": 2.191, | |
| "eval_steps_per_second": 2.191, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.6197530864197531, | |
| "grad_norm": 0.6871436834335327, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 0.1813, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.6592592592592592, | |
| "grad_norm": 0.6357116103172302, | |
| "learning_rate": 8.434984630174509e-05, | |
| "loss": 0.1148, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.6987654320987655, | |
| "grad_norm": 0.7641391158103943, | |
| "learning_rate": 8.33173288976002e-05, | |
| "loss": 0.188, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.7382716049382716, | |
| "grad_norm": 0.7134028077125549, | |
| "learning_rate": 8.225859917710439e-05, | |
| "loss": 0.1127, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.5360139608383179, | |
| "learning_rate": 8.117449009293668e-05, | |
| "loss": 0.1299, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.817283950617284, | |
| "grad_norm": 0.6457854509353638, | |
| "learning_rate": 8.006585456492029e-05, | |
| "loss": 0.1467, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.8567901234567903, | |
| "grad_norm": 0.7671291828155518, | |
| "learning_rate": 7.89335648089903e-05, | |
| "loss": 0.1861, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.8962962962962964, | |
| "grad_norm": 0.8006156086921692, | |
| "learning_rate": 7.777851165098012e-05, | |
| "loss": 0.1645, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.9358024691358025, | |
| "grad_norm": 0.5853805541992188, | |
| "learning_rate": 7.660160382576683e-05, | |
| "loss": 0.1304, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "grad_norm": 0.617660403251648, | |
| "learning_rate": 7.540376726232648e-05, | |
| "loss": 0.12, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "eval_loss": 0.14782202243804932, | |
| "eval_runtime": 20.5425, | |
| "eval_samples_per_second": 2.191, | |
| "eval_steps_per_second": 2.191, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.0148148148148146, | |
| "grad_norm": 1.5551854372024536, | |
| "learning_rate": 7.4185944355262e-05, | |
| "loss": 0.1831, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 2.054320987654321, | |
| "grad_norm": 0.4419083297252655, | |
| "learning_rate": 7.294909322337689e-05, | |
| "loss": 0.0911, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 2.093827160493827, | |
| "grad_norm": 0.5179771184921265, | |
| "learning_rate": 7.169418695587791e-05, | |
| "loss": 0.1113, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.1333333333333333, | |
| "grad_norm": 0.6325857639312744, | |
| "learning_rate": 7.042221284679982e-05, | |
| "loss": 0.1367, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.1728395061728394, | |
| "grad_norm": 0.5268187522888184, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 0.1126, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.212345679012346, | |
| "grad_norm": 0.5794491171836853, | |
| "learning_rate": 6.783107663311565e-05, | |
| "loss": 0.0919, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.251851851851852, | |
| "grad_norm": 0.59217369556427, | |
| "learning_rate": 6.651395309775837e-05, | |
| "loss": 0.1075, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.291358024691358, | |
| "grad_norm": 0.8060773611068726, | |
| "learning_rate": 6.518383725548074e-05, | |
| "loss": 0.1575, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.330864197530864, | |
| "grad_norm": 0.6405791640281677, | |
| "learning_rate": 6.384177557124247e-05, | |
| "loss": 0.1139, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "grad_norm": 0.510718822479248, | |
| "learning_rate": 6.248882390836135e-05, | |
| "loss": 0.1055, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "eval_loss": 0.1452956199645996, | |
| "eval_runtime": 20.5379, | |
| "eval_samples_per_second": 2.191, | |
| "eval_steps_per_second": 2.191, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.4098765432098768, | |
| "grad_norm": 0.7461621165275574, | |
| "learning_rate": 6.112604669781572e-05, | |
| "loss": 0.1049, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.449382716049383, | |
| "grad_norm": 0.7133863568305969, | |
| "learning_rate": 5.9754516100806423e-05, | |
| "loss": 0.1197, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.488888888888889, | |
| "grad_norm": 0.5839875936508179, | |
| "learning_rate": 5.837531116523682e-05, | |
| "loss": 0.1, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.528395061728395, | |
| "grad_norm": 0.7603824138641357, | |
| "learning_rate": 5.698951697677498e-05, | |
| "loss": 0.1361, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.567901234567901, | |
| "grad_norm": 0.8560308814048767, | |
| "learning_rate": 5.559822380516539e-05, | |
| "loss": 0.141, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.6074074074074076, | |
| "grad_norm": 0.5927801132202148, | |
| "learning_rate": 5.420252624646238e-05, | |
| "loss": 0.0921, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.6469135802469137, | |
| "grad_norm": 0.6076046228408813, | |
| "learning_rate": 5.2803522361859594e-05, | |
| "loss": 0.1114, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.68641975308642, | |
| "grad_norm": 0.490334153175354, | |
| "learning_rate": 5.140231281379345e-05, | |
| "loss": 0.0607, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.725925925925926, | |
| "grad_norm": 0.666424036026001, | |
| "learning_rate": 5e-05, | |
| "loss": 0.1302, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.765432098765432, | |
| "grad_norm": 0.7501330971717834, | |
| "learning_rate": 4.859768718620656e-05, | |
| "loss": 0.1206, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.765432098765432, | |
| "eval_loss": 0.14114373922348022, | |
| "eval_runtime": 20.5445, | |
| "eval_samples_per_second": 2.19, | |
| "eval_steps_per_second": 2.19, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.8049382716049385, | |
| "grad_norm": 0.5758643746376038, | |
| "learning_rate": 4.7196477638140404e-05, | |
| "loss": 0.093, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.8444444444444446, | |
| "grad_norm": 0.6395496726036072, | |
| "learning_rate": 4.579747375353763e-05, | |
| "loss": 0.083, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.8839506172839506, | |
| "grad_norm": 0.7084903717041016, | |
| "learning_rate": 4.4401776194834613e-05, | |
| "loss": 0.122, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.9234567901234567, | |
| "grad_norm": 0.6841776371002197, | |
| "learning_rate": 4.3010483023225045e-05, | |
| "loss": 0.1016, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.962962962962963, | |
| "grad_norm": 0.7126519083976746, | |
| "learning_rate": 4.162468883476319e-05, | |
| "loss": 0.0974, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 3.0024691358024693, | |
| "grad_norm": 1.506861686706543, | |
| "learning_rate": 4.0245483899193595e-05, | |
| "loss": 0.1674, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 3.0419753086419754, | |
| "grad_norm": 0.5017262697219849, | |
| "learning_rate": 3.887395330218429e-05, | |
| "loss": 0.0703, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 3.0814814814814815, | |
| "grad_norm": 0.43486133217811584, | |
| "learning_rate": 3.7511176091638653e-05, | |
| "loss": 0.0634, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 3.1209876543209876, | |
| "grad_norm": 0.5067691802978516, | |
| "learning_rate": 3.6158224428757535e-05, | |
| "loss": 0.0818, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 3.1604938271604937, | |
| "grad_norm": 0.6629949808120728, | |
| "learning_rate": 3.4816162744519263e-05, | |
| "loss": 0.0681, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.1604938271604937, | |
| "eval_loss": 0.13968898355960846, | |
| "eval_runtime": 20.5853, | |
| "eval_samples_per_second": 2.186, | |
| "eval_steps_per_second": 2.186, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.7635481357574463, | |
| "learning_rate": 3.3486046902241664e-05, | |
| "loss": 0.0726, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 3.2395061728395063, | |
| "grad_norm": 0.7752376198768616, | |
| "learning_rate": 3.216892336688435e-05, | |
| "loss": 0.0831, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 3.2790123456790123, | |
| "grad_norm": 0.643225371837616, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 0.0848, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 3.3185185185185184, | |
| "grad_norm": 0.6511841416358948, | |
| "learning_rate": 2.9577787153200197e-05, | |
| "loss": 0.0814, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 3.3580246913580245, | |
| "grad_norm": 0.6683458089828491, | |
| "learning_rate": 2.8305813044122097e-05, | |
| "loss": 0.094, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 3.397530864197531, | |
| "grad_norm": 0.5880411267280579, | |
| "learning_rate": 2.705090677662311e-05, | |
| "loss": 0.0759, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 3.437037037037037, | |
| "grad_norm": 0.6802661418914795, | |
| "learning_rate": 2.581405564473801e-05, | |
| "loss": 0.0992, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 3.476543209876543, | |
| "grad_norm": 0.6112740635871887, | |
| "learning_rate": 2.459623273767354e-05, | |
| "loss": 0.0751, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 3.5160493827160493, | |
| "grad_norm": 0.5678839087486267, | |
| "learning_rate": 2.3398396174233178e-05, | |
| "loss": 0.079, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "grad_norm": 0.6539812684059143, | |
| "learning_rate": 2.2221488349019903e-05, | |
| "loss": 0.0998, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "eval_loss": 0.1477135270833969, | |
| "eval_runtime": 20.5746, | |
| "eval_samples_per_second": 2.187, | |
| "eval_steps_per_second": 2.187, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.595061728395062, | |
| "grad_norm": 0.4667294919490814, | |
| "learning_rate": 2.1066435191009715e-05, | |
| "loss": 0.071, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 3.634567901234568, | |
| "grad_norm": 0.657560408115387, | |
| "learning_rate": 1.9934145435079702e-05, | |
| "loss": 0.0676, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 3.674074074074074, | |
| "grad_norm": 0.6763099431991577, | |
| "learning_rate": 1.8825509907063327e-05, | |
| "loss": 0.0456, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 3.71358024691358, | |
| "grad_norm": 0.7919898629188538, | |
| "learning_rate": 1.774140082289563e-05, | |
| "loss": 0.0808, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 3.753086419753086, | |
| "grad_norm": 0.7872672080993652, | |
| "learning_rate": 1.6682671102399805e-05, | |
| "loss": 0.0791, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 3.7925925925925927, | |
| "grad_norm": 0.5582082271575928, | |
| "learning_rate": 1.5650153698254916e-05, | |
| "loss": 0.0441, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 3.832098765432099, | |
| "grad_norm": 0.4519927501678467, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 0.0517, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 3.871604938271605, | |
| "grad_norm": 0.6880300641059875, | |
| "learning_rate": 1.3666983898298657e-05, | |
| "loss": 0.0709, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 3.911111111111111, | |
| "grad_norm": 0.6515712141990662, | |
| "learning_rate": 1.2717891755841722e-05, | |
| "loss": 0.0755, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 3.950617283950617, | |
| "grad_norm": 0.7855006456375122, | |
| "learning_rate": 1.1798131208919627e-05, | |
| "loss": 0.0925, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.950617283950617, | |
| "eval_loss": 0.14526966214179993, | |
| "eval_runtime": 20.546, | |
| "eval_samples_per_second": 2.19, | |
| "eval_steps_per_second": 2.19, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.9901234567901236, | |
| "grad_norm": 0.8963893055915833, | |
| "learning_rate": 1.090842587659851e-05, | |
| "loss": 0.123, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 4.029629629629629, | |
| "grad_norm": 1.4712413549423218, | |
| "learning_rate": 1.004947573208756e-05, | |
| "loss": 0.1398, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 4.069135802469136, | |
| "grad_norm": 0.43849503993988037, | |
| "learning_rate": 9.221956552036992e-06, | |
| "loss": 0.0375, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 4.108641975308642, | |
| "grad_norm": 0.4816286563873291, | |
| "learning_rate": 8.426519384872733e-06, | |
| "loss": 0.0499, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 4.148148148148148, | |
| "grad_norm": 0.6433464884757996, | |
| "learning_rate": 7.663790038585793e-06, | |
| "loss": 0.0568, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 4.187654320987654, | |
| "grad_norm": 0.6620740294456482, | |
| "learning_rate": 6.934368588379553e-06, | |
| "loss": 0.0773, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 4.22716049382716, | |
| "grad_norm": 0.5736699104309082, | |
| "learning_rate": 6.238828904562316e-06, | |
| "loss": 0.068, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 4.266666666666667, | |
| "grad_norm": 0.5692580342292786, | |
| "learning_rate": 5.577718201056392e-06, | |
| "loss": 0.0579, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 4.306172839506173, | |
| "grad_norm": 0.7503917217254639, | |
| "learning_rate": 4.951556604879048e-06, | |
| "loss": 0.0817, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 4.345679012345679, | |
| "grad_norm": 0.5718289613723755, | |
| "learning_rate": 4.360836746934055e-06, | |
| "loss": 0.0629, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.345679012345679, | |
| "eval_loss": 0.14678055047988892, | |
| "eval_runtime": 20.5757, | |
| "eval_samples_per_second": 2.187, | |
| "eval_steps_per_second": 2.187, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.385185185185185, | |
| "grad_norm": 0.501072108745575, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 0.0509, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 4.424691358024692, | |
| "grad_norm": 0.6218196749687195, | |
| "learning_rate": 3.2875529852700147e-06, | |
| "loss": 0.0588, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 4.4641975308641975, | |
| "grad_norm": 0.5222399234771729, | |
| "learning_rate": 2.8058334845816213e-06, | |
| "loss": 0.0486, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 4.503703703703704, | |
| "grad_norm": 0.5513294339179993, | |
| "learning_rate": 2.361243863855184e-06, | |
| "loss": 0.0584, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 4.54320987654321, | |
| "grad_norm": 0.5653938055038452, | |
| "learning_rate": 1.9541339027450256e-06, | |
| "loss": 0.0461, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 4.582716049382716, | |
| "grad_norm": 0.5166693925857544, | |
| "learning_rate": 1.584823893886933e-06, | |
| "loss": 0.06, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 4.622222222222222, | |
| "grad_norm": 0.5390232801437378, | |
| "learning_rate": 1.2536043909088191e-06, | |
| "loss": 0.0515, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 4.661728395061728, | |
| "grad_norm": 0.5809735655784607, | |
| "learning_rate": 9.607359798384785e-07, | |
| "loss": 0.0739, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 4.701234567901235, | |
| "grad_norm": 0.6697901487350464, | |
| "learning_rate": 7.064490740882057e-07, | |
| "loss": 0.0599, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 4.7407407407407405, | |
| "grad_norm": 0.5547673106193542, | |
| "learning_rate": 4.909437331777179e-07, | |
| "loss": 0.0494, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.7407407407407405, | |
| "eval_loss": 0.1466922163963318, | |
| "eval_runtime": 20.5445, | |
| "eval_samples_per_second": 2.19, | |
| "eval_steps_per_second": 2.19, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.780246913580247, | |
| "grad_norm": 0.671406090259552, | |
| "learning_rate": 3.143895053378698e-07, | |
| "loss": 0.0545, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 4.8197530864197535, | |
| "grad_norm": 0.6312061548233032, | |
| "learning_rate": 1.7692529411904578e-07, | |
| "loss": 0.0768, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 4.859259259259259, | |
| "grad_norm": 0.7700077295303345, | |
| "learning_rate": 7.865924910916977e-08, | |
| "loss": 0.0795, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 4.898765432098766, | |
| "grad_norm": 0.6207782030105591, | |
| "learning_rate": 1.9668680847356735e-08, | |
| "loss": 0.0631, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 4.938271604938271, | |
| "grad_norm": 0.6642947793006897, | |
| "learning_rate": 0.0, | |
| "loss": 0.0721, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 4.938271604938271, | |
| "step": 125, | |
| "total_flos": 1.2645167320439194e+17, | |
| "train_loss": 0.1889761376976967, | |
| "train_runtime": 3434.1648, | |
| "train_samples_per_second": 0.59, | |
| "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": 1.2645167320439194e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |