| { | |
| "best_metric": 0.15611045062541962, | |
| "best_model_checkpoint": "saves/psy-course-info/Llama3-OpenBioLLM-8B/train/fold4/checkpoint-90", | |
| "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": 3.150212287902832, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.9269, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.07901234567901234, | |
| "grad_norm": 2.839662551879883, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.9561, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.11851851851851852, | |
| "grad_norm": 3.2966983318328857, | |
| "learning_rate": 2.307692307692308e-05, | |
| "loss": 1.0853, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.1580246913580247, | |
| "grad_norm": 3.952324390411377, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 1.059, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.19753086419753085, | |
| "grad_norm": 3.5561017990112305, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 1.0555, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.23703703703703705, | |
| "grad_norm": 3.1896321773529053, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 0.766, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.2765432098765432, | |
| "grad_norm": 2.819338083267212, | |
| "learning_rate": 5.384615384615385e-05, | |
| "loss": 0.8993, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.3160493827160494, | |
| "grad_norm": 1.9059076309204102, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.6707, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 2.153079032897949, | |
| "learning_rate": 6.923076923076924e-05, | |
| "loss": 0.7418, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "grad_norm": 1.368592619895935, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.4899, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "eval_loss": 0.44628024101257324, | |
| "eval_runtime": 20.1675, | |
| "eval_samples_per_second": 2.231, | |
| "eval_steps_per_second": 2.231, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.4345679012345679, | |
| "grad_norm": 1.4414180517196655, | |
| "learning_rate": 8.461538461538461e-05, | |
| "loss": 0.4691, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.4740740740740741, | |
| "grad_norm": 1.1040608882904053, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.4109, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.5135802469135803, | |
| "grad_norm": 1.2567126750946045, | |
| "learning_rate": 0.0001, | |
| "loss": 0.4641, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.5530864197530864, | |
| "grad_norm": 0.8284547328948975, | |
| "learning_rate": 9.998033131915266e-05, | |
| "loss": 0.2527, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 1.0195739269256592, | |
| "learning_rate": 9.992134075089084e-05, | |
| "loss": 0.414, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.6320987654320988, | |
| "grad_norm": 0.8992851376533508, | |
| "learning_rate": 9.982307470588098e-05, | |
| "loss": 0.3324, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.671604938271605, | |
| "grad_norm": 0.762380063533783, | |
| "learning_rate": 9.968561049466214e-05, | |
| "loss": 0.2944, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 0.8036015629768372, | |
| "learning_rate": 9.950905626682228e-05, | |
| "loss": 0.3612, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.7506172839506173, | |
| "grad_norm": 0.6959598660469055, | |
| "learning_rate": 9.92935509259118e-05, | |
| "loss": 0.2385, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "grad_norm": 0.7877092957496643, | |
| "learning_rate": 9.903926402016153e-05, | |
| "loss": 0.2516, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "eval_loss": 0.24635133147239685, | |
| "eval_runtime": 20.0843, | |
| "eval_samples_per_second": 2.241, | |
| "eval_steps_per_second": 2.241, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.8296296296296296, | |
| "grad_norm": 0.7724449038505554, | |
| "learning_rate": 9.874639560909117e-05, | |
| "loss": 0.2287, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.8691358024691358, | |
| "grad_norm": 0.7435916066169739, | |
| "learning_rate": 9.841517610611309e-05, | |
| "loss": 0.1786, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.908641975308642, | |
| "grad_norm": 0.7090814113616943, | |
| "learning_rate": 9.804586609725499e-05, | |
| "loss": 0.1971, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.9481481481481482, | |
| "grad_norm": 0.6660599112510681, | |
| "learning_rate": 9.763875613614482e-05, | |
| "loss": 0.2183, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.9876543209876543, | |
| "grad_norm": 0.609691858291626, | |
| "learning_rate": 9.719416651541839e-05, | |
| "loss": 0.171, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.0271604938271606, | |
| "grad_norm": 1.3448481559753418, | |
| "learning_rate": 9.671244701472999e-05, | |
| "loss": 0.2797, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.0666666666666667, | |
| "grad_norm": 0.7494949102401733, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 0.2367, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.106172839506173, | |
| "grad_norm": 0.5752710700035095, | |
| "learning_rate": 9.563916325306594e-05, | |
| "loss": 0.1544, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.145679012345679, | |
| "grad_norm": 0.5666768550872803, | |
| "learning_rate": 9.504844339512095e-05, | |
| "loss": 0.1491, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "grad_norm": 0.8121204376220703, | |
| "learning_rate": 9.442228179894362e-05, | |
| "loss": 0.1562, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "eval_loss": 0.199776753783226, | |
| "eval_runtime": 20.1087, | |
| "eval_samples_per_second": 2.238, | |
| "eval_steps_per_second": 2.238, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.2246913580246914, | |
| "grad_norm": 0.7211824655532837, | |
| "learning_rate": 9.376117109543769e-05, | |
| "loss": 0.1424, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.2641975308641975, | |
| "grad_norm": 0.7016025185585022, | |
| "learning_rate": 9.306563141162046e-05, | |
| "loss": 0.1799, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.3037037037037038, | |
| "grad_norm": 0.8877782821655273, | |
| "learning_rate": 9.233620996141421e-05, | |
| "loss": 0.1673, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.34320987654321, | |
| "grad_norm": 0.9606795907020569, | |
| "learning_rate": 9.157348061512727e-05, | |
| "loss": 0.1775, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.382716049382716, | |
| "grad_norm": 0.7455890774726868, | |
| "learning_rate": 9.077804344796302e-05, | |
| "loss": 0.1574, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.4222222222222223, | |
| "grad_norm": 0.9921693205833435, | |
| "learning_rate": 8.995052426791247e-05, | |
| "loss": 0.192, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.4617283950617284, | |
| "grad_norm": 0.7622290253639221, | |
| "learning_rate": 8.90915741234015e-05, | |
| "loss": 0.1832, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.5012345679012347, | |
| "grad_norm": 0.6994752883911133, | |
| "learning_rate": 8.820186879108038e-05, | |
| "loss": 0.1774, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.5407407407407407, | |
| "grad_norm": 1.0304056406021118, | |
| "learning_rate": 8.728210824415827e-05, | |
| "loss": 0.1955, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.5802469135802468, | |
| "grad_norm": 0.8595938086509705, | |
| "learning_rate": 8.633301610170135e-05, | |
| "loss": 0.1462, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.5802469135802468, | |
| "eval_loss": 0.1835593581199646, | |
| "eval_runtime": 20.0807, | |
| "eval_samples_per_second": 2.241, | |
| "eval_steps_per_second": 2.241, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.6197530864197531, | |
| "grad_norm": 0.703360915184021, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 0.1753, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.6592592592592592, | |
| "grad_norm": 0.8332892060279846, | |
| "learning_rate": 8.434984630174509e-05, | |
| "loss": 0.2153, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.6987654320987655, | |
| "grad_norm": 0.8192476034164429, | |
| "learning_rate": 8.33173288976002e-05, | |
| "loss": 0.1542, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.7382716049382716, | |
| "grad_norm": 0.7894420623779297, | |
| "learning_rate": 8.225859917710439e-05, | |
| "loss": 0.1788, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.568242609500885, | |
| "learning_rate": 8.117449009293668e-05, | |
| "loss": 0.095, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.817283950617284, | |
| "grad_norm": 0.6856135129928589, | |
| "learning_rate": 8.006585456492029e-05, | |
| "loss": 0.1592, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.8567901234567903, | |
| "grad_norm": 0.7471094727516174, | |
| "learning_rate": 7.89335648089903e-05, | |
| "loss": 0.1995, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.8962962962962964, | |
| "grad_norm": 0.6927120089530945, | |
| "learning_rate": 7.777851165098012e-05, | |
| "loss": 0.1429, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.9358024691358025, | |
| "grad_norm": 0.8563644289970398, | |
| "learning_rate": 7.660160382576683e-05, | |
| "loss": 0.1784, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "grad_norm": 0.5290800929069519, | |
| "learning_rate": 7.540376726232648e-05, | |
| "loss": 0.1466, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "eval_loss": 0.16728538274765015, | |
| "eval_runtime": 20.125, | |
| "eval_samples_per_second": 2.236, | |
| "eval_steps_per_second": 2.236, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.0148148148148146, | |
| "grad_norm": 1.4483234882354736, | |
| "learning_rate": 7.4185944355262e-05, | |
| "loss": 0.3706, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 2.054320987654321, | |
| "grad_norm": 0.5486336946487427, | |
| "learning_rate": 7.294909322337689e-05, | |
| "loss": 0.1317, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 2.093827160493827, | |
| "grad_norm": 0.6281251311302185, | |
| "learning_rate": 7.169418695587791e-05, | |
| "loss": 0.1212, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.1333333333333333, | |
| "grad_norm": 0.43187186121940613, | |
| "learning_rate": 7.042221284679982e-05, | |
| "loss": 0.0873, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.1728395061728394, | |
| "grad_norm": 0.8371054530143738, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 0.1257, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.212345679012346, | |
| "grad_norm": 0.7313236594200134, | |
| "learning_rate": 6.783107663311565e-05, | |
| "loss": 0.1247, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.251851851851852, | |
| "grad_norm": 0.5864114165306091, | |
| "learning_rate": 6.651395309775837e-05, | |
| "loss": 0.1553, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.291358024691358, | |
| "grad_norm": 0.7349463701248169, | |
| "learning_rate": 6.518383725548074e-05, | |
| "loss": 0.153, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.330864197530864, | |
| "grad_norm": 0.564865231513977, | |
| "learning_rate": 6.384177557124247e-05, | |
| "loss": 0.1269, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "grad_norm": 0.5237666368484497, | |
| "learning_rate": 6.248882390836135e-05, | |
| "loss": 0.0851, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "eval_loss": 0.1626947969198227, | |
| "eval_runtime": 20.1567, | |
| "eval_samples_per_second": 2.233, | |
| "eval_steps_per_second": 2.233, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.4098765432098768, | |
| "grad_norm": 0.5278699994087219, | |
| "learning_rate": 6.112604669781572e-05, | |
| "loss": 0.1099, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.449382716049383, | |
| "grad_norm": 0.6749888062477112, | |
| "learning_rate": 5.9754516100806423e-05, | |
| "loss": 0.1072, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.488888888888889, | |
| "grad_norm": 0.46553850173950195, | |
| "learning_rate": 5.837531116523682e-05, | |
| "loss": 0.1026, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.528395061728395, | |
| "grad_norm": 0.6362727284431458, | |
| "learning_rate": 5.698951697677498e-05, | |
| "loss": 0.1384, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.567901234567901, | |
| "grad_norm": 0.5537588596343994, | |
| "learning_rate": 5.559822380516539e-05, | |
| "loss": 0.131, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.6074074074074076, | |
| "grad_norm": 0.6617990136146545, | |
| "learning_rate": 5.420252624646238e-05, | |
| "loss": 0.1277, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.6469135802469137, | |
| "grad_norm": 0.5553164482116699, | |
| "learning_rate": 5.2803522361859594e-05, | |
| "loss": 0.074, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.68641975308642, | |
| "grad_norm": 0.6082383990287781, | |
| "learning_rate": 5.140231281379345e-05, | |
| "loss": 0.1429, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.725925925925926, | |
| "grad_norm": 0.5445216298103333, | |
| "learning_rate": 5e-05, | |
| "loss": 0.0789, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.765432098765432, | |
| "grad_norm": 0.40841108560562134, | |
| "learning_rate": 4.859768718620656e-05, | |
| "loss": 0.0661, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.765432098765432, | |
| "eval_loss": 0.15659214556217194, | |
| "eval_runtime": 20.1351, | |
| "eval_samples_per_second": 2.235, | |
| "eval_steps_per_second": 2.235, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.8049382716049385, | |
| "grad_norm": 0.632308840751648, | |
| "learning_rate": 4.7196477638140404e-05, | |
| "loss": 0.1275, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.8444444444444446, | |
| "grad_norm": 0.517742931842804, | |
| "learning_rate": 4.579747375353763e-05, | |
| "loss": 0.0758, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.8839506172839506, | |
| "grad_norm": 0.6730695962905884, | |
| "learning_rate": 4.4401776194834613e-05, | |
| "loss": 0.1134, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.9234567901234567, | |
| "grad_norm": 0.6757080554962158, | |
| "learning_rate": 4.3010483023225045e-05, | |
| "loss": 0.1307, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.962962962962963, | |
| "grad_norm": 0.5498260855674744, | |
| "learning_rate": 4.162468883476319e-05, | |
| "loss": 0.1043, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 3.0024691358024693, | |
| "grad_norm": 1.0815857648849487, | |
| "learning_rate": 4.0245483899193595e-05, | |
| "loss": 0.1586, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 3.0419753086419754, | |
| "grad_norm": 0.4448467493057251, | |
| "learning_rate": 3.887395330218429e-05, | |
| "loss": 0.0588, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 3.0814814814814815, | |
| "grad_norm": 0.5207440853118896, | |
| "learning_rate": 3.7511176091638653e-05, | |
| "loss": 0.0914, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 3.1209876543209876, | |
| "grad_norm": 0.5485023856163025, | |
| "learning_rate": 3.6158224428757535e-05, | |
| "loss": 0.0805, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 3.1604938271604937, | |
| "grad_norm": 0.5623069405555725, | |
| "learning_rate": 3.4816162744519263e-05, | |
| "loss": 0.0802, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.1604938271604937, | |
| "eval_loss": 0.15811502933502197, | |
| "eval_runtime": 20.1282, | |
| "eval_samples_per_second": 2.236, | |
| "eval_steps_per_second": 2.236, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.486346036195755, | |
| "learning_rate": 3.3486046902241664e-05, | |
| "loss": 0.0841, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 3.2395061728395063, | |
| "grad_norm": 0.5264397859573364, | |
| "learning_rate": 3.216892336688435e-05, | |
| "loss": 0.0507, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 3.2790123456790123, | |
| "grad_norm": 0.5107821226119995, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 0.0695, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 3.3185185185185184, | |
| "grad_norm": 0.793722927570343, | |
| "learning_rate": 2.9577787153200197e-05, | |
| "loss": 0.1244, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 3.3580246913580245, | |
| "grad_norm": 0.5132105946540833, | |
| "learning_rate": 2.8305813044122097e-05, | |
| "loss": 0.0772, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 3.397530864197531, | |
| "grad_norm": 0.6101983189582825, | |
| "learning_rate": 2.705090677662311e-05, | |
| "loss": 0.094, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 3.437037037037037, | |
| "grad_norm": 0.556900143623352, | |
| "learning_rate": 2.581405564473801e-05, | |
| "loss": 0.0744, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 3.476543209876543, | |
| "grad_norm": 0.5194830298423767, | |
| "learning_rate": 2.459623273767354e-05, | |
| "loss": 0.0901, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 3.5160493827160493, | |
| "grad_norm": 0.7026291489601135, | |
| "learning_rate": 2.3398396174233178e-05, | |
| "loss": 0.0924, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "grad_norm": 0.4749229848384857, | |
| "learning_rate": 2.2221488349019903e-05, | |
| "loss": 0.0617, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "eval_loss": 0.15611045062541962, | |
| "eval_runtime": 20.137, | |
| "eval_samples_per_second": 2.235, | |
| "eval_steps_per_second": 2.235, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.595061728395062, | |
| "grad_norm": 0.5764184594154358, | |
| "learning_rate": 2.1066435191009715e-05, | |
| "loss": 0.0665, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 3.634567901234568, | |
| "grad_norm": 0.5767927765846252, | |
| "learning_rate": 1.9934145435079702e-05, | |
| "loss": 0.0857, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 3.674074074074074, | |
| "grad_norm": 0.5262015461921692, | |
| "learning_rate": 1.8825509907063327e-05, | |
| "loss": 0.0915, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 3.71358024691358, | |
| "grad_norm": 0.7448687553405762, | |
| "learning_rate": 1.774140082289563e-05, | |
| "loss": 0.083, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 3.753086419753086, | |
| "grad_norm": 0.6409696936607361, | |
| "learning_rate": 1.6682671102399805e-05, | |
| "loss": 0.0573, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 3.7925925925925927, | |
| "grad_norm": 0.6057963371276855, | |
| "learning_rate": 1.5650153698254916e-05, | |
| "loss": 0.0706, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 3.832098765432099, | |
| "grad_norm": 0.8120406866073608, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 0.0894, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 3.871604938271605, | |
| "grad_norm": 0.5812767744064331, | |
| "learning_rate": 1.3666983898298657e-05, | |
| "loss": 0.0723, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 3.911111111111111, | |
| "grad_norm": 0.6257267594337463, | |
| "learning_rate": 1.2717891755841722e-05, | |
| "loss": 0.0764, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 3.950617283950617, | |
| "grad_norm": 0.6552925109863281, | |
| "learning_rate": 1.1798131208919627e-05, | |
| "loss": 0.0708, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.950617283950617, | |
| "eval_loss": 0.15689979493618011, | |
| "eval_runtime": 20.18, | |
| "eval_samples_per_second": 2.23, | |
| "eval_steps_per_second": 2.23, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.9901234567901236, | |
| "grad_norm": 0.8006722331047058, | |
| "learning_rate": 1.090842587659851e-05, | |
| "loss": 0.1015, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 4.029629629629629, | |
| "grad_norm": 1.1077641248703003, | |
| "learning_rate": 1.004947573208756e-05, | |
| "loss": 0.1159, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 4.069135802469136, | |
| "grad_norm": 0.5113848447799683, | |
| "learning_rate": 9.221956552036992e-06, | |
| "loss": 0.0637, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 4.108641975308642, | |
| "grad_norm": 0.42762231826782227, | |
| "learning_rate": 8.426519384872733e-06, | |
| "loss": 0.0557, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 4.148148148148148, | |
| "grad_norm": 0.5600584149360657, | |
| "learning_rate": 7.663790038585793e-06, | |
| "loss": 0.0692, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 4.187654320987654, | |
| "grad_norm": 0.37912362813949585, | |
| "learning_rate": 6.934368588379553e-06, | |
| "loss": 0.0342, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 4.22716049382716, | |
| "grad_norm": 0.5694722533226013, | |
| "learning_rate": 6.238828904562316e-06, | |
| "loss": 0.0864, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 4.266666666666667, | |
| "grad_norm": 0.5017801523208618, | |
| "learning_rate": 5.577718201056392e-06, | |
| "loss": 0.0533, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 4.306172839506173, | |
| "grad_norm": 0.7008262276649475, | |
| "learning_rate": 4.951556604879048e-06, | |
| "loss": 0.0838, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 4.345679012345679, | |
| "grad_norm": 0.5300676226615906, | |
| "learning_rate": 4.360836746934055e-06, | |
| "loss": 0.0713, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.345679012345679, | |
| "eval_loss": 0.15776211023330688, | |
| "eval_runtime": 20.1692, | |
| "eval_samples_per_second": 2.231, | |
| "eval_steps_per_second": 2.231, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.385185185185185, | |
| "grad_norm": 0.6056603193283081, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 0.0704, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 4.424691358024692, | |
| "grad_norm": 0.910595178604126, | |
| "learning_rate": 3.2875529852700147e-06, | |
| "loss": 0.0679, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 4.4641975308641975, | |
| "grad_norm": 0.6534622311592102, | |
| "learning_rate": 2.8058334845816213e-06, | |
| "loss": 0.0583, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 4.503703703703704, | |
| "grad_norm": 0.5863792300224304, | |
| "learning_rate": 2.361243863855184e-06, | |
| "loss": 0.0584, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 4.54320987654321, | |
| "grad_norm": 0.5510096549987793, | |
| "learning_rate": 1.9541339027450256e-06, | |
| "loss": 0.0511, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 4.582716049382716, | |
| "grad_norm": 0.5553609728813171, | |
| "learning_rate": 1.584823893886933e-06, | |
| "loss": 0.0782, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 4.622222222222222, | |
| "grad_norm": 0.602301836013794, | |
| "learning_rate": 1.2536043909088191e-06, | |
| "loss": 0.0897, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 4.661728395061728, | |
| "grad_norm": 0.5180858969688416, | |
| "learning_rate": 9.607359798384785e-07, | |
| "loss": 0.0451, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 4.701234567901235, | |
| "grad_norm": 0.5263860821723938, | |
| "learning_rate": 7.064490740882057e-07, | |
| "loss": 0.065, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 4.7407407407407405, | |
| "grad_norm": 0.7415157556533813, | |
| "learning_rate": 4.909437331777179e-07, | |
| "loss": 0.088, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.7407407407407405, | |
| "eval_loss": 0.15884658694267273, | |
| "eval_runtime": 20.1383, | |
| "eval_samples_per_second": 2.235, | |
| "eval_steps_per_second": 2.235, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.780246913580247, | |
| "grad_norm": 0.480844646692276, | |
| "learning_rate": 3.143895053378698e-07, | |
| "loss": 0.0613, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 4.8197530864197535, | |
| "grad_norm": 0.6222435832023621, | |
| "learning_rate": 1.7692529411904578e-07, | |
| "loss": 0.078, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 4.859259259259259, | |
| "grad_norm": 0.4616582989692688, | |
| "learning_rate": 7.865924910916977e-08, | |
| "loss": 0.052, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 4.898765432098766, | |
| "grad_norm": 0.661733090877533, | |
| "learning_rate": 1.9668680847356735e-08, | |
| "loss": 0.0555, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 4.938271604938271, | |
| "grad_norm": 0.370050311088562, | |
| "learning_rate": 0.0, | |
| "loss": 0.0335, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 4.938271604938271, | |
| "step": 125, | |
| "total_flos": 1.2628766887457587e+17, | |
| "train_loss": 0.19458558702468873, | |
| "train_runtime": 3405.889, | |
| "train_samples_per_second": 0.595, | |
| "train_steps_per_second": 0.037 | |
| } | |
| ], | |
| "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": 1.2628766887457587e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |