| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 8.0, | |
| "global_step": 11448, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 87.2627, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 52.4123, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 6e-06, | |
| "loss": 41.3932, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 36.7185, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 1e-05, | |
| "loss": 33.4597, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 1.2e-05, | |
| "loss": 31.1323, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 28.9204, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 27.0128, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 1.8e-05, | |
| "loss": 25.1703, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 2e-05, | |
| "loss": 23.5486, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 2.2000000000000003e-05, | |
| "loss": 21.8419, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 2.4e-05, | |
| "loss": 20.3387, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "loss": 18.7216, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "loss": 16.7862, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "learning_rate": 3e-05, | |
| "loss": 15.034, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 13.3747, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "loss": 11.7581, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 3.6e-05, | |
| "loss": 9.9702, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 3.8e-05, | |
| "loss": 8.3363, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 4e-05, | |
| "loss": 6.9199, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "eval_bleu": 3.4801, | |
| "eval_em": 0.0, | |
| "eval_gen_len": 220.966, | |
| "eval_loss": 6.3751444816589355, | |
| "eval_runtime": 1604.7901, | |
| "eval_samples_per_second": 0.751, | |
| "eval_steps_per_second": 0.094, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "learning_rate": 4.2e-05, | |
| "loss": 6.2524, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 5.8029, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 5.606, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "learning_rate": 4.8e-05, | |
| "loss": 5.3757, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "learning_rate": 5e-05, | |
| "loss": 5.1043, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "learning_rate": 4.944121591417077e-05, | |
| "loss": 4.9414, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 4.888243182834153e-05, | |
| "loss": 4.7381, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "learning_rate": 4.8323647742512295e-05, | |
| "loss": 4.6214, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "learning_rate": 4.776486365668306e-05, | |
| "loss": 4.2971, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "learning_rate": 4.720607957085382e-05, | |
| "loss": 4.1602, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "learning_rate": 4.664729548502459e-05, | |
| "loss": 4.0391, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "learning_rate": 4.6088511399195353e-05, | |
| "loss": 3.9211, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "learning_rate": 4.552972731336611e-05, | |
| "loss": 3.7642, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "learning_rate": 4.497094322753688e-05, | |
| "loss": 3.6698, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "learning_rate": 4.4412159141707646e-05, | |
| "loss": 3.5409, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "learning_rate": 4.385337505587841e-05, | |
| "loss": 3.4016, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "learning_rate": 4.329459097004918e-05, | |
| "loss": 3.2761, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "learning_rate": 4.2735806884219945e-05, | |
| "loss": 3.1708, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "learning_rate": 4.2177022798390704e-05, | |
| "loss": 3.0849, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 4.161823871256147e-05, | |
| "loss": 3.0222, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "eval_bleu": 27.8543, | |
| "eval_em": 0.0, | |
| "eval_gen_len": 36.8, | |
| "eval_loss": 2.8796441555023193, | |
| "eval_runtime": 501.7382, | |
| "eval_samples_per_second": 2.402, | |
| "eval_steps_per_second": 0.301, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "learning_rate": 4.105945462673223e-05, | |
| "loss": 2.9079, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "learning_rate": 4.0500670540903e-05, | |
| "loss": 2.8688, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 3.994188645507376e-05, | |
| "loss": 2.7398, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "learning_rate": 3.938310236924452e-05, | |
| "loss": 2.5503, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "learning_rate": 3.882431828341529e-05, | |
| "loss": 2.5121, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "learning_rate": 3.8265534197586055e-05, | |
| "loss": 2.4625, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 3.28, | |
| "learning_rate": 3.7706750111756815e-05, | |
| "loss": 2.3833, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "learning_rate": 3.714796602592758e-05, | |
| "loss": 2.3133, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 3.42, | |
| "learning_rate": 3.658918194009835e-05, | |
| "loss": 2.2152, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "learning_rate": 3.603039785426911e-05, | |
| "loss": 2.2304, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "learning_rate": 3.5471613768439874e-05, | |
| "loss": 2.1563, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 3.63, | |
| "learning_rate": 3.491282968261064e-05, | |
| "loss": 2.0991, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "learning_rate": 3.4354045596781406e-05, | |
| "loss": 2.0702, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "learning_rate": 3.379526151095217e-05, | |
| "loss": 1.9914, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "learning_rate": 3.323647742512294e-05, | |
| "loss": 1.9235, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 3.91, | |
| "learning_rate": 3.26776933392937e-05, | |
| "loss": 1.8922, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 3.98, | |
| "learning_rate": 3.2118909253464465e-05, | |
| "loss": 1.8512, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 4.05, | |
| "learning_rate": 3.156012516763523e-05, | |
| "loss": 1.6966, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 4.12, | |
| "learning_rate": 3.100134108180599e-05, | |
| "loss": 1.6399, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 4.19, | |
| "learning_rate": 3.0442556995976757e-05, | |
| "loss": 1.5982, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 4.19, | |
| "eval_bleu": 56.0747, | |
| "eval_em": 0.0017, | |
| "eval_gen_len": 43.9021, | |
| "eval_loss": 1.7495189905166626, | |
| "eval_runtime": 423.2589, | |
| "eval_samples_per_second": 2.847, | |
| "eval_steps_per_second": 0.357, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 4.26, | |
| "learning_rate": 2.9883772910147524e-05, | |
| "loss": 1.5968, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 4.33, | |
| "learning_rate": 2.9324988824318283e-05, | |
| "loss": 1.5813, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "learning_rate": 2.876620473848905e-05, | |
| "loss": 1.5421, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 4.47, | |
| "learning_rate": 2.8207420652659816e-05, | |
| "loss": 1.4852, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 4.54, | |
| "learning_rate": 2.7648636566830576e-05, | |
| "loss": 1.5027, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 4.61, | |
| "learning_rate": 2.7089852481001342e-05, | |
| "loss": 1.4951, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 4.68, | |
| "learning_rate": 2.653106839517211e-05, | |
| "loss": 1.4803, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 4.75, | |
| "learning_rate": 2.597228430934287e-05, | |
| "loss": 1.4127, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 4.82, | |
| "learning_rate": 2.5413500223513638e-05, | |
| "loss": 1.3896, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 4.89, | |
| "learning_rate": 2.48547161376844e-05, | |
| "loss": 1.3656, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 4.96, | |
| "learning_rate": 2.4295932051855164e-05, | |
| "loss": 1.3432, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 5.03, | |
| "learning_rate": 2.373714796602593e-05, | |
| "loss": 1.2224, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "learning_rate": 2.3178363880196693e-05, | |
| "loss": 1.1396, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 5.17, | |
| "learning_rate": 2.2619579794367456e-05, | |
| "loss": 1.1475, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 5.24, | |
| "learning_rate": 2.206079570853822e-05, | |
| "loss": 1.0669, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 5.31, | |
| "learning_rate": 2.1502011622708985e-05, | |
| "loss": 1.1356, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 5.38, | |
| "learning_rate": 2.0943227536879752e-05, | |
| "loss": 1.0965, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 5.45, | |
| "learning_rate": 2.0384443451050515e-05, | |
| "loss": 1.1086, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 5.52, | |
| "learning_rate": 1.982565936522128e-05, | |
| "loss": 1.0642, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 5.59, | |
| "learning_rate": 1.9266875279392044e-05, | |
| "loss": 1.0717, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 5.59, | |
| "eval_bleu": 69.9606, | |
| "eval_em": 0.0199, | |
| "eval_gen_len": 46.0722, | |
| "eval_loss": 1.26251220703125, | |
| "eval_runtime": 355.6397, | |
| "eval_samples_per_second": 3.388, | |
| "eval_steps_per_second": 0.425, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 5.66, | |
| "learning_rate": 1.8708091193562807e-05, | |
| "loss": 1.0546, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 5.73, | |
| "learning_rate": 1.8149307107733573e-05, | |
| "loss": 1.0579, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "learning_rate": 1.7590523021904336e-05, | |
| "loss": 1.0204, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 5.87, | |
| "learning_rate": 1.70317389360751e-05, | |
| "loss": 1.0398, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 5.94, | |
| "learning_rate": 1.6472954850245866e-05, | |
| "loss": 0.9992, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 6.01, | |
| "learning_rate": 1.5914170764416632e-05, | |
| "loss": 0.9756, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 6.08, | |
| "learning_rate": 1.5355386678587395e-05, | |
| "loss": 0.8385, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 6.15, | |
| "learning_rate": 1.479660259275816e-05, | |
| "loss": 0.8815, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 6.22, | |
| "learning_rate": 1.4237818506928924e-05, | |
| "loss": 0.8447, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 6.29, | |
| "learning_rate": 1.3679034421099687e-05, | |
| "loss": 0.8553, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 6.36, | |
| "learning_rate": 1.312025033527045e-05, | |
| "loss": 0.8188, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 6.43, | |
| "learning_rate": 1.2561466249441217e-05, | |
| "loss": 0.8241, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "learning_rate": 1.2002682163611981e-05, | |
| "loss": 0.8118, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 6.57, | |
| "learning_rate": 1.1443898077782746e-05, | |
| "loss": 0.8357, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 6.64, | |
| "learning_rate": 1.0885113991953509e-05, | |
| "loss": 0.8063, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 6.71, | |
| "learning_rate": 1.0326329906124274e-05, | |
| "loss": 0.8263, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 6.78, | |
| "learning_rate": 9.767545820295038e-06, | |
| "loss": 0.8064, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 6.85, | |
| "learning_rate": 9.208761734465803e-06, | |
| "loss": 0.7858, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 6.92, | |
| "learning_rate": 8.649977648636568e-06, | |
| "loss": 0.7854, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 6.99, | |
| "learning_rate": 8.09119356280733e-06, | |
| "loss": 0.7765, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 6.99, | |
| "eval_bleu": 74.7723, | |
| "eval_em": 0.0349, | |
| "eval_gen_len": 46.1685, | |
| "eval_loss": 1.0809996128082275, | |
| "eval_runtime": 352.8566, | |
| "eval_samples_per_second": 3.415, | |
| "eval_steps_per_second": 0.428, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 7.06, | |
| "learning_rate": 7.532409476978096e-06, | |
| "loss": 0.696, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 7.13, | |
| "learning_rate": 6.973625391148861e-06, | |
| "loss": 0.6991, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "learning_rate": 6.414841305319625e-06, | |
| "loss": 0.6795, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 7.27, | |
| "learning_rate": 5.856057219490389e-06, | |
| "loss": 0.6953, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 7.34, | |
| "learning_rate": 5.297273133661153e-06, | |
| "loss": 0.6854, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 7.41, | |
| "learning_rate": 4.738489047831918e-06, | |
| "loss": 0.6798, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 7.48, | |
| "learning_rate": 4.1797049620026825e-06, | |
| "loss": 0.6701, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 7.55, | |
| "learning_rate": 3.6209208761734468e-06, | |
| "loss": 0.684, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 7.62, | |
| "learning_rate": 3.062136790344211e-06, | |
| "loss": 0.686, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 7.69, | |
| "learning_rate": 2.5033527045149757e-06, | |
| "loss": 0.6764, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 7.76, | |
| "learning_rate": 1.94456861868574e-06, | |
| "loss": 0.6461, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 7.83, | |
| "learning_rate": 1.3857845328565042e-06, | |
| "loss": 0.6723, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 7.9, | |
| "learning_rate": 8.270004470272687e-07, | |
| "loss": 0.6453, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 7.97, | |
| "learning_rate": 2.682163611980331e-07, | |
| "loss": 0.6544, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "step": 11448, | |
| "total_flos": 7528078235838336.0, | |
| "train_loss": 6.2339570505647375, | |
| "train_runtime": 11756.9791, | |
| "train_samples_per_second": 15.575, | |
| "train_steps_per_second": 0.974 | |
| } | |
| ], | |
| "max_steps": 11448, | |
| "num_train_epochs": 8, | |
| "total_flos": 7528078235838336.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |