| { | |
| "best_global_step": 440, | |
| "best_metric": 0.16359447004608296, | |
| "best_model_checkpoint": "out_qwen_var_sft/checkpoint-440", | |
| "epoch": 7.213114754098361, | |
| "eval_steps": 10, | |
| "global_step": 440, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08196721311475409, | |
| "grad_norm": 2001.4421126814825, | |
| "learning_rate": 4.0816326530612243e-07, | |
| "loss": 42.6172, | |
| "step": 5, | |
| "true_loss": 5.0281 | |
| }, | |
| { | |
| "epoch": 0.16393442622950818, | |
| "grad_norm": 158.37906608693382, | |
| "learning_rate": 9.183673469387756e-07, | |
| "loss": 43.1594, | |
| "step": 10, | |
| "true_loss": 5.1953 | |
| }, | |
| { | |
| "epoch": 0.16393442622950818, | |
| "eval_accuracy": 0.01152073732718894, | |
| "eval_loss": 5.286218166351318, | |
| "eval_runtime": 15.0885, | |
| "eval_samples_per_second": 28.764, | |
| "eval_steps_per_second": 3.645, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.2459016393442623, | |
| "grad_norm": 203.23818895534905, | |
| "learning_rate": 1.4285714285714286e-06, | |
| "loss": 42.6562, | |
| "step": 15, | |
| "true_loss": 5.2555 | |
| }, | |
| { | |
| "epoch": 0.32786885245901637, | |
| "grad_norm": 176.6457829157802, | |
| "learning_rate": 1.938775510204082e-06, | |
| "loss": 42.1543, | |
| "step": 20, | |
| "true_loss": 5.1922 | |
| }, | |
| { | |
| "epoch": 0.32786885245901637, | |
| "eval_accuracy": 0.01152073732718894, | |
| "eval_loss": 5.156322002410889, | |
| "eval_runtime": 14.9518, | |
| "eval_samples_per_second": 29.027, | |
| "eval_steps_per_second": 3.678, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.4098360655737705, | |
| "grad_norm": 157.3725914328133, | |
| "learning_rate": 2.4489795918367347e-06, | |
| "loss": 41.0754, | |
| "step": 25, | |
| "true_loss": 4.9437 | |
| }, | |
| { | |
| "epoch": 0.4918032786885246, | |
| "grad_norm": 177.5762016006674, | |
| "learning_rate": 2.959183673469388e-06, | |
| "loss": 38.793, | |
| "step": 30, | |
| "true_loss": 5.1707 | |
| }, | |
| { | |
| "epoch": 0.4918032786885246, | |
| "eval_accuracy": 0.02304147465437788, | |
| "eval_loss": 4.853038787841797, | |
| "eval_runtime": 14.9796, | |
| "eval_samples_per_second": 28.973, | |
| "eval_steps_per_second": 3.672, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5737704918032787, | |
| "grad_norm": 224.10088573018953, | |
| "learning_rate": 3.469387755102041e-06, | |
| "loss": 38.6121, | |
| "step": 35, | |
| "true_loss": 4.7551 | |
| }, | |
| { | |
| "epoch": 0.6557377049180327, | |
| "grad_norm": 302.8526334743389, | |
| "learning_rate": 3.979591836734694e-06, | |
| "loss": 37.8027, | |
| "step": 40, | |
| "true_loss": 4.9977 | |
| }, | |
| { | |
| "epoch": 0.6557377049180327, | |
| "eval_accuracy": 0.06221198156682028, | |
| "eval_loss": 4.661470413208008, | |
| "eval_runtime": 15.006, | |
| "eval_samples_per_second": 28.922, | |
| "eval_steps_per_second": 3.665, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7377049180327869, | |
| "grad_norm": 177.15837283975523, | |
| "learning_rate": 4.489795918367348e-06, | |
| "loss": 36.4664, | |
| "step": 45, | |
| "true_loss": 4.5766 | |
| }, | |
| { | |
| "epoch": 0.819672131147541, | |
| "grad_norm": 190.75273796127124, | |
| "learning_rate": 5e-06, | |
| "loss": 36.8304, | |
| "step": 50, | |
| "true_loss": 4.7605 | |
| }, | |
| { | |
| "epoch": 0.819672131147541, | |
| "eval_accuracy": 0.052995391705069124, | |
| "eval_loss": 4.522649765014648, | |
| "eval_runtime": 14.8422, | |
| "eval_samples_per_second": 29.241, | |
| "eval_steps_per_second": 3.706, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.9016393442622951, | |
| "grad_norm": 185.07218548511187, | |
| "learning_rate": 4.943052391799545e-06, | |
| "loss": 36.585, | |
| "step": 55, | |
| "true_loss": 4.5461 | |
| }, | |
| { | |
| "epoch": 0.9836065573770492, | |
| "grad_norm": 169.35214138176565, | |
| "learning_rate": 4.886104783599089e-06, | |
| "loss": 35.3022, | |
| "step": 60, | |
| "true_loss": 4.3983 | |
| }, | |
| { | |
| "epoch": 0.9836065573770492, | |
| "eval_accuracy": 0.06221198156682028, | |
| "eval_loss": 4.3924431800842285, | |
| "eval_runtime": 14.9823, | |
| "eval_samples_per_second": 28.967, | |
| "eval_steps_per_second": 3.671, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0655737704918034, | |
| "grad_norm": 186.19471703324027, | |
| "learning_rate": 4.829157175398634e-06, | |
| "loss": 34.9871, | |
| "step": 65, | |
| "true_loss": 4.534 | |
| }, | |
| { | |
| "epoch": 1.1475409836065573, | |
| "grad_norm": 178.50065430802542, | |
| "learning_rate": 4.772209567198178e-06, | |
| "loss": 34.8587, | |
| "step": 70, | |
| "true_loss": 4.0799 | |
| }, | |
| { | |
| "epoch": 1.1475409836065573, | |
| "eval_accuracy": 0.06682027649769585, | |
| "eval_loss": 4.316955089569092, | |
| "eval_runtime": 14.9903, | |
| "eval_samples_per_second": 28.952, | |
| "eval_steps_per_second": 3.669, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.2295081967213115, | |
| "grad_norm": 256.1945567078422, | |
| "learning_rate": 4.7152619589977225e-06, | |
| "loss": 33.4347, | |
| "step": 75, | |
| "true_loss": 3.7832 | |
| }, | |
| { | |
| "epoch": 1.3114754098360657, | |
| "grad_norm": 178.23899873402075, | |
| "learning_rate": 4.658314350797267e-06, | |
| "loss": 33.5913, | |
| "step": 80, | |
| "true_loss": 4.3762 | |
| }, | |
| { | |
| "epoch": 1.3114754098360657, | |
| "eval_accuracy": 0.08064516129032258, | |
| "eval_loss": 4.25508975982666, | |
| "eval_runtime": 14.9497, | |
| "eval_samples_per_second": 29.031, | |
| "eval_steps_per_second": 3.679, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.3934426229508197, | |
| "grad_norm": 176.89010340519602, | |
| "learning_rate": 4.601366742596811e-06, | |
| "loss": 33.1532, | |
| "step": 85, | |
| "true_loss": 4.1607 | |
| }, | |
| { | |
| "epoch": 1.4754098360655736, | |
| "grad_norm": 146.21731787991052, | |
| "learning_rate": 4.544419134396356e-06, | |
| "loss": 33.0214, | |
| "step": 90, | |
| "true_loss": 4.3359 | |
| }, | |
| { | |
| "epoch": 1.4754098360655736, | |
| "eval_accuracy": 0.0967741935483871, | |
| "eval_loss": 4.17578125, | |
| "eval_runtime": 14.9139, | |
| "eval_samples_per_second": 29.1, | |
| "eval_steps_per_second": 3.688, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.5573770491803278, | |
| "grad_norm": 163.96782239209782, | |
| "learning_rate": 4.4874715261959e-06, | |
| "loss": 31.8735, | |
| "step": 95, | |
| "true_loss": 4.1047 | |
| }, | |
| { | |
| "epoch": 1.639344262295082, | |
| "grad_norm": 172.0759692604596, | |
| "learning_rate": 4.4305239179954446e-06, | |
| "loss": 33.1337, | |
| "step": 100, | |
| "true_loss": 4.4055 | |
| }, | |
| { | |
| "epoch": 1.639344262295082, | |
| "eval_accuracy": 0.09907834101382489, | |
| "eval_loss": 4.103317737579346, | |
| "eval_runtime": 15.1423, | |
| "eval_samples_per_second": 28.661, | |
| "eval_steps_per_second": 3.632, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.721311475409836, | |
| "grad_norm": 163.50986481723115, | |
| "learning_rate": 4.373576309794989e-06, | |
| "loss": 30.4056, | |
| "step": 105, | |
| "true_loss": 3.5892 | |
| }, | |
| { | |
| "epoch": 1.8032786885245902, | |
| "grad_norm": 174.28722424027927, | |
| "learning_rate": 4.316628701594533e-06, | |
| "loss": 32.4443, | |
| "step": 110, | |
| "true_loss": 3.8227 | |
| }, | |
| { | |
| "epoch": 1.8032786885245902, | |
| "eval_accuracy": 0.11059907834101383, | |
| "eval_loss": 4.104156970977783, | |
| "eval_runtime": 14.9928, | |
| "eval_samples_per_second": 28.947, | |
| "eval_steps_per_second": 3.668, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.8852459016393444, | |
| "grad_norm": 599.8963091372194, | |
| "learning_rate": 4.259681093394078e-06, | |
| "loss": 32.7365, | |
| "step": 115, | |
| "true_loss": 4.3132 | |
| }, | |
| { | |
| "epoch": 1.9672131147540983, | |
| "grad_norm": 176.15071401035615, | |
| "learning_rate": 4.202733485193622e-06, | |
| "loss": 30.8238, | |
| "step": 120, | |
| "true_loss": 3.9128 | |
| }, | |
| { | |
| "epoch": 1.9672131147540983, | |
| "eval_accuracy": 0.09216589861751152, | |
| "eval_loss": 4.044179439544678, | |
| "eval_runtime": 14.9286, | |
| "eval_samples_per_second": 29.072, | |
| "eval_steps_per_second": 3.684, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.0491803278688523, | |
| "grad_norm": 193.228785849168, | |
| "learning_rate": 4.145785876993167e-06, | |
| "loss": 31.3295, | |
| "step": 125, | |
| "true_loss": 4.0655 | |
| }, | |
| { | |
| "epoch": 2.1311475409836067, | |
| "grad_norm": 167.07209609143695, | |
| "learning_rate": 4.088838268792711e-06, | |
| "loss": 29.0552, | |
| "step": 130, | |
| "true_loss": 3.727 | |
| }, | |
| { | |
| "epoch": 2.1311475409836067, | |
| "eval_accuracy": 0.12211981566820276, | |
| "eval_loss": 4.029242992401123, | |
| "eval_runtime": 15.3137, | |
| "eval_samples_per_second": 28.341, | |
| "eval_steps_per_second": 3.592, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.2131147540983607, | |
| "grad_norm": 208.65171046196372, | |
| "learning_rate": 4.0318906605922555e-06, | |
| "loss": 28.6111, | |
| "step": 135, | |
| "true_loss": 3.5737 | |
| }, | |
| { | |
| "epoch": 2.2950819672131146, | |
| "grad_norm": 198.8811989422508, | |
| "learning_rate": 3.9749430523918e-06, | |
| "loss": 28.6129, | |
| "step": 140, | |
| "true_loss": 3.4636 | |
| }, | |
| { | |
| "epoch": 2.2950819672131146, | |
| "eval_accuracy": 0.1152073732718894, | |
| "eval_loss": 4.056672096252441, | |
| "eval_runtime": 14.9575, | |
| "eval_samples_per_second": 29.016, | |
| "eval_steps_per_second": 3.677, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.3770491803278686, | |
| "grad_norm": 199.51861074084005, | |
| "learning_rate": 3.917995444191344e-06, | |
| "loss": 28.8025, | |
| "step": 145, | |
| "true_loss": 3.5059 | |
| }, | |
| { | |
| "epoch": 2.459016393442623, | |
| "grad_norm": 193.67355565121787, | |
| "learning_rate": 3.861047835990889e-06, | |
| "loss": 28.5974, | |
| "step": 150, | |
| "true_loss": 2.9942 | |
| }, | |
| { | |
| "epoch": 2.459016393442623, | |
| "eval_accuracy": 0.0967741935483871, | |
| "eval_loss": 4.023684978485107, | |
| "eval_runtime": 15.0321, | |
| "eval_samples_per_second": 28.872, | |
| "eval_steps_per_second": 3.659, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.540983606557377, | |
| "grad_norm": 205.5841428370514, | |
| "learning_rate": 3.804100227790433e-06, | |
| "loss": 28.4969, | |
| "step": 155, | |
| "true_loss": 3.5412 | |
| }, | |
| { | |
| "epoch": 2.6229508196721314, | |
| "grad_norm": 182.43691547474708, | |
| "learning_rate": 3.7471526195899776e-06, | |
| "loss": 29.1582, | |
| "step": 160, | |
| "true_loss": 3.5827 | |
| }, | |
| { | |
| "epoch": 2.6229508196721314, | |
| "eval_accuracy": 0.10368663594470046, | |
| "eval_loss": 4.006450176239014, | |
| "eval_runtime": 15.0347, | |
| "eval_samples_per_second": 28.867, | |
| "eval_steps_per_second": 3.658, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.7049180327868854, | |
| "grad_norm": 209.23223838323548, | |
| "learning_rate": 3.690205011389522e-06, | |
| "loss": 28.4554, | |
| "step": 165, | |
| "true_loss": 3.8515 | |
| }, | |
| { | |
| "epoch": 2.7868852459016393, | |
| "grad_norm": 181.70496603318674, | |
| "learning_rate": 3.6332574031890664e-06, | |
| "loss": 28.0033, | |
| "step": 170, | |
| "true_loss": 3.8595 | |
| }, | |
| { | |
| "epoch": 2.7868852459016393, | |
| "eval_accuracy": 0.1336405529953917, | |
| "eval_loss": 3.9438364505767822, | |
| "eval_runtime": 15.1184, | |
| "eval_samples_per_second": 28.707, | |
| "eval_steps_per_second": 3.638, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.8688524590163933, | |
| "grad_norm": 189.43537999115608, | |
| "learning_rate": 3.5763097949886104e-06, | |
| "loss": 27.3218, | |
| "step": 175, | |
| "true_loss": 3.4404 | |
| }, | |
| { | |
| "epoch": 2.9508196721311473, | |
| "grad_norm": 203.83162067201448, | |
| "learning_rate": 3.519362186788155e-06, | |
| "loss": 28.3321, | |
| "step": 180, | |
| "true_loss": 3.1337 | |
| }, | |
| { | |
| "epoch": 2.9508196721311473, | |
| "eval_accuracy": 0.10829493087557604, | |
| "eval_loss": 3.9445700645446777, | |
| "eval_runtime": 15.1698, | |
| "eval_samples_per_second": 28.61, | |
| "eval_steps_per_second": 3.626, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.0327868852459017, | |
| "grad_norm": 188.7733943154734, | |
| "learning_rate": 3.4624145785876997e-06, | |
| "loss": 27.27, | |
| "step": 185, | |
| "true_loss": 3.334 | |
| }, | |
| { | |
| "epoch": 3.1147540983606556, | |
| "grad_norm": 299.860020893839, | |
| "learning_rate": 3.405466970387244e-06, | |
| "loss": 23.8397, | |
| "step": 190, | |
| "true_loss": 2.6139 | |
| }, | |
| { | |
| "epoch": 3.1147540983606556, | |
| "eval_accuracy": 0.11059907834101383, | |
| "eval_loss": 3.9798836708068848, | |
| "eval_runtime": 15.4586, | |
| "eval_samples_per_second": 28.075, | |
| "eval_steps_per_second": 3.558, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.19672131147541, | |
| "grad_norm": 261.2265032273993, | |
| "learning_rate": 3.3485193621867885e-06, | |
| "loss": 24.7364, | |
| "step": 195, | |
| "true_loss": 3.0539 | |
| }, | |
| { | |
| "epoch": 3.278688524590164, | |
| "grad_norm": 244.13814588236283, | |
| "learning_rate": 3.291571753986333e-06, | |
| "loss": 25.6726, | |
| "step": 200, | |
| "true_loss": 3.0328 | |
| }, | |
| { | |
| "epoch": 3.278688524590164, | |
| "eval_accuracy": 0.11981566820276497, | |
| "eval_loss": 3.931368350982666, | |
| "eval_runtime": 14.9343, | |
| "eval_samples_per_second": 29.061, | |
| "eval_steps_per_second": 3.683, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.360655737704918, | |
| "grad_norm": 223.86848695043577, | |
| "learning_rate": 3.2346241457858773e-06, | |
| "loss": 24.6349, | |
| "step": 205, | |
| "true_loss": 2.9529 | |
| }, | |
| { | |
| "epoch": 3.442622950819672, | |
| "grad_norm": 239.830671087524, | |
| "learning_rate": 3.1776765375854217e-06, | |
| "loss": 24.5938, | |
| "step": 210, | |
| "true_loss": 3.3097 | |
| }, | |
| { | |
| "epoch": 3.442622950819672, | |
| "eval_accuracy": 0.1175115207373272, | |
| "eval_loss": 3.9209704399108887, | |
| "eval_runtime": 14.9774, | |
| "eval_samples_per_second": 28.977, | |
| "eval_steps_per_second": 3.672, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.5245901639344264, | |
| "grad_norm": 188.8971324623384, | |
| "learning_rate": 3.120728929384966e-06, | |
| "loss": 24.3125, | |
| "step": 215, | |
| "true_loss": 2.8892 | |
| }, | |
| { | |
| "epoch": 3.6065573770491803, | |
| "grad_norm": 211.46305164676468, | |
| "learning_rate": 3.0637813211845106e-06, | |
| "loss": 24.9786, | |
| "step": 220, | |
| "true_loss": 3.4745 | |
| }, | |
| { | |
| "epoch": 3.6065573770491803, | |
| "eval_accuracy": 0.11290322580645161, | |
| "eval_loss": 3.892094373703003, | |
| "eval_runtime": 15.0091, | |
| "eval_samples_per_second": 28.916, | |
| "eval_steps_per_second": 3.664, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.6885245901639343, | |
| "grad_norm": 209.57415125751984, | |
| "learning_rate": 3.0068337129840546e-06, | |
| "loss": 24.4776, | |
| "step": 225, | |
| "true_loss": 3.0013 | |
| }, | |
| { | |
| "epoch": 3.7704918032786887, | |
| "grad_norm": 219.2212947656317, | |
| "learning_rate": 2.949886104783599e-06, | |
| "loss": 25.0714, | |
| "step": 230, | |
| "true_loss": 3.3653 | |
| }, | |
| { | |
| "epoch": 3.7704918032786887, | |
| "eval_accuracy": 0.12211981566820276, | |
| "eval_loss": 3.88911509513855, | |
| "eval_runtime": 15.0058, | |
| "eval_samples_per_second": 28.922, | |
| "eval_steps_per_second": 3.665, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 3.8524590163934427, | |
| "grad_norm": 215.0136859889706, | |
| "learning_rate": 2.892938496583144e-06, | |
| "loss": 25.0681, | |
| "step": 235, | |
| "true_loss": 2.7688 | |
| }, | |
| { | |
| "epoch": 3.9344262295081966, | |
| "grad_norm": 212.6143565856701, | |
| "learning_rate": 2.8359908883826882e-06, | |
| "loss": 24.524, | |
| "step": 240, | |
| "true_loss": 3.0097 | |
| }, | |
| { | |
| "epoch": 3.9344262295081966, | |
| "eval_accuracy": 0.10599078341013825, | |
| "eval_loss": 3.901205062866211, | |
| "eval_runtime": 14.9742, | |
| "eval_samples_per_second": 28.983, | |
| "eval_steps_per_second": 3.673, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 4.016393442622951, | |
| "grad_norm": 203.1242362392191, | |
| "learning_rate": 2.7790432801822326e-06, | |
| "loss": 22.3977, | |
| "step": 245, | |
| "true_loss": 2.4811 | |
| }, | |
| { | |
| "epoch": 4.098360655737705, | |
| "grad_norm": 219.93955042504686, | |
| "learning_rate": 2.722095671981777e-06, | |
| "loss": 20.6891, | |
| "step": 250, | |
| "true_loss": 2.6231 | |
| }, | |
| { | |
| "epoch": 4.098360655737705, | |
| "eval_accuracy": 0.12903225806451613, | |
| "eval_loss": 3.918625831604004, | |
| "eval_runtime": 15.0247, | |
| "eval_samples_per_second": 28.886, | |
| "eval_steps_per_second": 3.661, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.180327868852459, | |
| "grad_norm": 269.83943945736723, | |
| "learning_rate": 2.6651480637813215e-06, | |
| "loss": 19.9576, | |
| "step": 255, | |
| "true_loss": 2.4069 | |
| }, | |
| { | |
| "epoch": 4.262295081967213, | |
| "grad_norm": 285.3088807670382, | |
| "learning_rate": 2.608200455580866e-06, | |
| "loss": 19.9079, | |
| "step": 260, | |
| "true_loss": 2.5739 | |
| }, | |
| { | |
| "epoch": 4.262295081967213, | |
| "eval_accuracy": 0.12672811059907835, | |
| "eval_loss": 3.955899477005005, | |
| "eval_runtime": 14.9722, | |
| "eval_samples_per_second": 28.987, | |
| "eval_steps_per_second": 3.673, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.344262295081967, | |
| "grad_norm": 235.65294642592144, | |
| "learning_rate": 2.5512528473804103e-06, | |
| "loss": 19.9913, | |
| "step": 265, | |
| "true_loss": 2.6661 | |
| }, | |
| { | |
| "epoch": 4.426229508196721, | |
| "grad_norm": 243.25746669728042, | |
| "learning_rate": 2.4943052391799547e-06, | |
| "loss": 21.5919, | |
| "step": 270, | |
| "true_loss": 2.931 | |
| }, | |
| { | |
| "epoch": 4.426229508196721, | |
| "eval_accuracy": 0.12672811059907835, | |
| "eval_loss": 3.8867075443267822, | |
| "eval_runtime": 14.9423, | |
| "eval_samples_per_second": 29.045, | |
| "eval_steps_per_second": 3.681, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 4.508196721311475, | |
| "grad_norm": 268.3143937900749, | |
| "learning_rate": 2.437357630979499e-06, | |
| "loss": 21.3721, | |
| "step": 275, | |
| "true_loss": 2.6628 | |
| }, | |
| { | |
| "epoch": 4.590163934426229, | |
| "grad_norm": 233.6702831549346, | |
| "learning_rate": 2.3804100227790436e-06, | |
| "loss": 21.651, | |
| "step": 280, | |
| "true_loss": 2.3591 | |
| }, | |
| { | |
| "epoch": 4.590163934426229, | |
| "eval_accuracy": 0.1359447004608295, | |
| "eval_loss": 3.855499029159546, | |
| "eval_runtime": 15.1421, | |
| "eval_samples_per_second": 28.662, | |
| "eval_steps_per_second": 3.632, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 4.672131147540983, | |
| "grad_norm": 219.0257023920148, | |
| "learning_rate": 2.323462414578588e-06, | |
| "loss": 21.5054, | |
| "step": 285, | |
| "true_loss": 2.5704 | |
| }, | |
| { | |
| "epoch": 4.754098360655737, | |
| "grad_norm": 287.08072005658994, | |
| "learning_rate": 2.2665148063781324e-06, | |
| "loss": 20.7173, | |
| "step": 290, | |
| "true_loss": 3.0756 | |
| }, | |
| { | |
| "epoch": 4.754098360655737, | |
| "eval_accuracy": 0.14055299539170507, | |
| "eval_loss": 3.8604798316955566, | |
| "eval_runtime": 15.1432, | |
| "eval_samples_per_second": 28.66, | |
| "eval_steps_per_second": 3.632, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 4.836065573770492, | |
| "grad_norm": 264.8569129204395, | |
| "learning_rate": 2.209567198177677e-06, | |
| "loss": 19.9181, | |
| "step": 295, | |
| "true_loss": 2.2299 | |
| }, | |
| { | |
| "epoch": 4.918032786885246, | |
| "grad_norm": 239.4181479000276, | |
| "learning_rate": 2.1526195899772212e-06, | |
| "loss": 22.0119, | |
| "step": 300, | |
| "true_loss": 3.0984 | |
| }, | |
| { | |
| "epoch": 4.918032786885246, | |
| "eval_accuracy": 0.1313364055299539, | |
| "eval_loss": 3.848721742630005, | |
| "eval_runtime": 15.1331, | |
| "eval_samples_per_second": 28.679, | |
| "eval_steps_per_second": 3.634, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 216.31247846676033, | |
| "learning_rate": 2.0956719817767656e-06, | |
| "loss": 20.7322, | |
| "step": 305, | |
| "true_loss": 2.2975 | |
| }, | |
| { | |
| "epoch": 5.081967213114754, | |
| "grad_norm": 266.0474875386491, | |
| "learning_rate": 2.03872437357631e-06, | |
| "loss": 17.2354, | |
| "step": 310, | |
| "true_loss": 2.0884 | |
| }, | |
| { | |
| "epoch": 5.081967213114754, | |
| "eval_accuracy": 0.1382488479262673, | |
| "eval_loss": 3.835157871246338, | |
| "eval_runtime": 15.1115, | |
| "eval_samples_per_second": 28.72, | |
| "eval_steps_per_second": 3.64, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 5.163934426229508, | |
| "grad_norm": 258.5339989822872, | |
| "learning_rate": 1.9817767653758545e-06, | |
| "loss": 17.442, | |
| "step": 315, | |
| "true_loss": 2.1051 | |
| }, | |
| { | |
| "epoch": 5.245901639344262, | |
| "grad_norm": 308.2469254205616, | |
| "learning_rate": 1.924829157175399e-06, | |
| "loss": 17.0414, | |
| "step": 320, | |
| "true_loss": 2.5328 | |
| }, | |
| { | |
| "epoch": 5.245901639344262, | |
| "eval_accuracy": 0.15207373271889402, | |
| "eval_loss": 3.899200201034546, | |
| "eval_runtime": 15.7217, | |
| "eval_samples_per_second": 27.605, | |
| "eval_steps_per_second": 3.498, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 5.327868852459017, | |
| "grad_norm": 297.7046229142417, | |
| "learning_rate": 1.8678815489749433e-06, | |
| "loss": 16.4162, | |
| "step": 325, | |
| "true_loss": 2.2139 | |
| }, | |
| { | |
| "epoch": 5.409836065573771, | |
| "grad_norm": 262.17760383527263, | |
| "learning_rate": 1.8109339407744877e-06, | |
| "loss": 17.9956, | |
| "step": 330, | |
| "true_loss": 2.3699 | |
| }, | |
| { | |
| "epoch": 5.409836065573771, | |
| "eval_accuracy": 0.15668202764976957, | |
| "eval_loss": 3.869380235671997, | |
| "eval_runtime": 14.9258, | |
| "eval_samples_per_second": 29.077, | |
| "eval_steps_per_second": 3.685, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 5.491803278688525, | |
| "grad_norm": 267.855545172113, | |
| "learning_rate": 1.753986332574032e-06, | |
| "loss": 18.2262, | |
| "step": 335, | |
| "true_loss": 2.1083 | |
| }, | |
| { | |
| "epoch": 5.573770491803279, | |
| "grad_norm": 263.032871409478, | |
| "learning_rate": 1.6970387243735763e-06, | |
| "loss": 17.7507, | |
| "step": 340, | |
| "true_loss": 2.0572 | |
| }, | |
| { | |
| "epoch": 5.573770491803279, | |
| "eval_accuracy": 0.14516129032258066, | |
| "eval_loss": 3.860694646835327, | |
| "eval_runtime": 14.9606, | |
| "eval_samples_per_second": 29.01, | |
| "eval_steps_per_second": 3.676, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 5.655737704918033, | |
| "grad_norm": 251.13017134627248, | |
| "learning_rate": 1.640091116173121e-06, | |
| "loss": 18.9093, | |
| "step": 345, | |
| "true_loss": 2.552 | |
| }, | |
| { | |
| "epoch": 5.737704918032787, | |
| "grad_norm": 283.50724134889464, | |
| "learning_rate": 1.5831435079726654e-06, | |
| "loss": 17.0755, | |
| "step": 350, | |
| "true_loss": 2.621 | |
| }, | |
| { | |
| "epoch": 5.737704918032787, | |
| "eval_accuracy": 0.15668202764976957, | |
| "eval_loss": 3.821519613265991, | |
| "eval_runtime": 15.0984, | |
| "eval_samples_per_second": 28.745, | |
| "eval_steps_per_second": 3.643, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 5.8196721311475414, | |
| "grad_norm": 751.7678054126261, | |
| "learning_rate": 1.5261958997722096e-06, | |
| "loss": 18.1668, | |
| "step": 355, | |
| "true_loss": 1.9259 | |
| }, | |
| { | |
| "epoch": 5.901639344262295, | |
| "grad_norm": 318.53273858625954, | |
| "learning_rate": 1.469248291571754e-06, | |
| "loss": 17.0525, | |
| "step": 360, | |
| "true_loss": 2.2937 | |
| }, | |
| { | |
| "epoch": 5.901639344262295, | |
| "eval_accuracy": 0.1497695852534562, | |
| "eval_loss": 3.812807083129883, | |
| "eval_runtime": 15.0494, | |
| "eval_samples_per_second": 28.838, | |
| "eval_steps_per_second": 3.655, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 5.983606557377049, | |
| "grad_norm": 259.86569786349736, | |
| "learning_rate": 1.4123006833712984e-06, | |
| "loss": 16.5851, | |
| "step": 365, | |
| "true_loss": 1.926 | |
| }, | |
| { | |
| "epoch": 6.065573770491803, | |
| "grad_norm": 236.16491616521617, | |
| "learning_rate": 1.355353075170843e-06, | |
| "loss": 14.0549, | |
| "step": 370, | |
| "true_loss": 1.6638 | |
| }, | |
| { | |
| "epoch": 6.065573770491803, | |
| "eval_accuracy": 0.14516129032258066, | |
| "eval_loss": 3.8400564193725586, | |
| "eval_runtime": 15.0501, | |
| "eval_samples_per_second": 28.837, | |
| "eval_steps_per_second": 3.654, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 6.147540983606557, | |
| "grad_norm": 267.4679343061016, | |
| "learning_rate": 1.2984054669703875e-06, | |
| "loss": 13.1733, | |
| "step": 375, | |
| "true_loss": 1.4947 | |
| }, | |
| { | |
| "epoch": 6.229508196721311, | |
| "grad_norm": 303.4138473047156, | |
| "learning_rate": 1.2414578587699317e-06, | |
| "loss": 14.2548, | |
| "step": 380, | |
| "true_loss": 1.8772 | |
| }, | |
| { | |
| "epoch": 6.229508196721311, | |
| "eval_accuracy": 0.14516129032258066, | |
| "eval_loss": 3.897352933883667, | |
| "eval_runtime": 15.0678, | |
| "eval_samples_per_second": 28.803, | |
| "eval_steps_per_second": 3.65, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 6.311475409836065, | |
| "grad_norm": 304.09080880251156, | |
| "learning_rate": 1.1845102505694763e-06, | |
| "loss": 14.1155, | |
| "step": 385, | |
| "true_loss": 1.4 | |
| }, | |
| { | |
| "epoch": 6.39344262295082, | |
| "grad_norm": 328.52917098962524, | |
| "learning_rate": 1.1275626423690205e-06, | |
| "loss": 14.0041, | |
| "step": 390, | |
| "true_loss": 1.5224 | |
| }, | |
| { | |
| "epoch": 6.39344262295082, | |
| "eval_accuracy": 0.1313364055299539, | |
| "eval_loss": 3.8768956661224365, | |
| "eval_runtime": 14.9485, | |
| "eval_samples_per_second": 29.033, | |
| "eval_steps_per_second": 3.679, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 6.475409836065574, | |
| "grad_norm": 374.87828963261063, | |
| "learning_rate": 1.0706150341685651e-06, | |
| "loss": 13.4204, | |
| "step": 395, | |
| "true_loss": 1.7657 | |
| }, | |
| { | |
| "epoch": 6.557377049180328, | |
| "grad_norm": 290.7307487199588, | |
| "learning_rate": 1.0136674259681093e-06, | |
| "loss": 14.282, | |
| "step": 400, | |
| "true_loss": 1.8787 | |
| }, | |
| { | |
| "epoch": 6.557377049180328, | |
| "eval_accuracy": 0.14285714285714285, | |
| "eval_loss": 3.83672833442688, | |
| "eval_runtime": 15.0933, | |
| "eval_samples_per_second": 28.755, | |
| "eval_steps_per_second": 3.644, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 6.639344262295082, | |
| "grad_norm": 293.9907784197036, | |
| "learning_rate": 9.567198177676538e-07, | |
| "loss": 13.0058, | |
| "step": 405, | |
| "true_loss": 1.9678 | |
| }, | |
| { | |
| "epoch": 6.721311475409836, | |
| "grad_norm": 329.0206706011657, | |
| "learning_rate": 8.997722095671982e-07, | |
| "loss": 15.0129, | |
| "step": 410, | |
| "true_loss": 1.8787 | |
| }, | |
| { | |
| "epoch": 6.721311475409836, | |
| "eval_accuracy": 0.15668202764976957, | |
| "eval_loss": 3.854370594024658, | |
| "eval_runtime": 15.0162, | |
| "eval_samples_per_second": 28.902, | |
| "eval_steps_per_second": 3.663, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 6.80327868852459, | |
| "grad_norm": 435.30888223467264, | |
| "learning_rate": 8.428246013667427e-07, | |
| "loss": 14.3213, | |
| "step": 415, | |
| "true_loss": 2.0839 | |
| }, | |
| { | |
| "epoch": 6.885245901639344, | |
| "grad_norm": 283.4882127418948, | |
| "learning_rate": 7.858769931662871e-07, | |
| "loss": 13.6361, | |
| "step": 420, | |
| "true_loss": 1.5802 | |
| }, | |
| { | |
| "epoch": 6.885245901639344, | |
| "eval_accuracy": 0.1497695852534562, | |
| "eval_loss": 3.839586019515991, | |
| "eval_runtime": 14.9583, | |
| "eval_samples_per_second": 29.014, | |
| "eval_steps_per_second": 3.677, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 6.967213114754099, | |
| "grad_norm": 353.6661268712156, | |
| "learning_rate": 7.289293849658314e-07, | |
| "loss": 15.2028, | |
| "step": 425, | |
| "true_loss": 1.7352 | |
| }, | |
| { | |
| "epoch": 7.049180327868853, | |
| "grad_norm": 273.2552710608739, | |
| "learning_rate": 6.71981776765376e-07, | |
| "loss": 12.2664, | |
| "step": 430, | |
| "true_loss": 1.4752 | |
| }, | |
| { | |
| "epoch": 7.049180327868853, | |
| "eval_accuracy": 0.14746543778801843, | |
| "eval_loss": 3.839120388031006, | |
| "eval_runtime": 15.1557, | |
| "eval_samples_per_second": 28.636, | |
| "eval_steps_per_second": 3.629, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 7.131147540983607, | |
| "grad_norm": 278.2724305429665, | |
| "learning_rate": 6.150341685649204e-07, | |
| "loss": 10.7664, | |
| "step": 435, | |
| "true_loss": 1.3734 | |
| }, | |
| { | |
| "epoch": 7.213114754098361, | |
| "grad_norm": 316.474377641746, | |
| "learning_rate": 5.580865603644648e-07, | |
| "loss": 10.5371, | |
| "step": 440, | |
| "true_loss": 1.6166 | |
| }, | |
| { | |
| "epoch": 7.213114754098361, | |
| "eval_accuracy": 0.16359447004608296, | |
| "eval_loss": 3.866461753845215, | |
| "eval_runtime": 14.9602, | |
| "eval_samples_per_second": 29.01, | |
| "eval_steps_per_second": 3.676, | |
| "step": 440 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 488, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 8, | |
| "save_steps": 10, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |