| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 13.0, | |
| "eval_steps": 2179, | |
| "global_step": 28327, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 3e-05, | |
| "loss": 2.2464, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3e-05, | |
| "loss": 2.0544, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9968, | |
| "step": 1308 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9537, | |
| "step": 1744 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6693020521036767, | |
| "eval_loss": 1.5623117685317993, | |
| "eval_runtime": 220.2436, | |
| "eval_samples_per_second": 33.622, | |
| "eval_steps_per_second": 2.102, | |
| "step": 2179 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_exact_match": 10.24983119513842, | |
| "eval_f1": 15.257097040945366, | |
| "eval_qa_bleu": 5.8382901471491575, | |
| "eval_qa_exact_match": 0.09736664415935178, | |
| "eval_recite_bleu": 17.0930417063702, | |
| "eval_recite_exact_match": 0.0, | |
| "step": 2179 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9029, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 3e-05, | |
| "loss": 1.5215, | |
| "step": 2616 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 3e-05, | |
| "loss": 1.5077, | |
| "step": 3052 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 3e-05, | |
| "loss": 1.5016, | |
| "step": 3488 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "learning_rate": 3e-05, | |
| "loss": 1.4734, | |
| "step": 3924 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6923618412861704, | |
| "eval_loss": 1.2099343538284302, | |
| "eval_runtime": 222.1542, | |
| "eval_samples_per_second": 33.333, | |
| "eval_steps_per_second": 2.084, | |
| "step": 4358 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_exact_match": 10.61444969615125, | |
| "eval_f1": 15.935672519396675, | |
| "eval_qa_bleu": 6.524750946372503, | |
| "eval_qa_exact_match": 0.10209318028359217, | |
| "eval_recite_bleu": 18.618913529891735, | |
| "eval_recite_exact_match": 0.0, | |
| "step": 4358 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 3e-05, | |
| "loss": 1.4425, | |
| "step": 4360 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0827, | |
| "step": 4796 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0876, | |
| "step": 5232 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0887, | |
| "step": 5668 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0665, | |
| "step": 6104 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7147465927772421, | |
| "eval_loss": 0.9177509546279907, | |
| "eval_runtime": 226.6441, | |
| "eval_samples_per_second": 32.672, | |
| "eval_steps_per_second": 2.043, | |
| "step": 6537 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_exact_match": 12.397029034436192, | |
| "eval_f1": 17.721572805797255, | |
| "eval_qa_bleu": 8.139751517788653, | |
| "eval_qa_exact_match": 0.11978392977717758, | |
| "eval_recite_bleu": 21.769859861022653, | |
| "eval_recite_exact_match": 0.0, | |
| "step": 6537 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7772, | |
| "step": 6540 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7688, | |
| "step": 6976 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7598, | |
| "step": 7412 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "learning_rate": 3e-05, | |
| "loss": 0.77, | |
| "step": 7848 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7684, | |
| "step": 8284 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7331474586575765, | |
| "eval_loss": 0.6987847089767456, | |
| "eval_runtime": 221.7249, | |
| "eval_samples_per_second": 33.397, | |
| "eval_steps_per_second": 2.088, | |
| "step": 8716 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_exact_match": 13.409858203916272, | |
| "eval_f1": 19.326947132558665, | |
| "eval_qa_bleu": 8.79837393539171, | |
| "eval_qa_exact_match": 0.12869682646860228, | |
| "eval_recite_bleu": 23.848732661358156, | |
| "eval_recite_exact_match": 0.0, | |
| "step": 8716 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 3e-05, | |
| "loss": 0.7559, | |
| "step": 8720 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5354, | |
| "step": 9156 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5511, | |
| "step": 9592 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5496, | |
| "step": 10028 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "learning_rate": 3e-05, | |
| "loss": 0.548, | |
| "step": 10464 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7465929589970037, | |
| "eval_loss": 0.5567022562026978, | |
| "eval_runtime": 226.4909, | |
| "eval_samples_per_second": 32.694, | |
| "eval_steps_per_second": 2.044, | |
| "step": 10895 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_exact_match": 14.220121539500338, | |
| "eval_f1": 20.736856637999697, | |
| "eval_qa_bleu": 9.542509177303282, | |
| "eval_qa_exact_match": 0.13652937204591492, | |
| "eval_recite_bleu": 27.315073242003514, | |
| "eval_recite_exact_match": 0.0009453072248480756, | |
| "step": 10895 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 3e-05, | |
| "loss": 0.5518, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "learning_rate": 3e-05, | |
| "loss": 0.3837, | |
| "step": 11336 | |
| }, | |
| { | |
| "epoch": 5.4, | |
| "learning_rate": 3e-05, | |
| "loss": 0.3963, | |
| "step": 11772 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "learning_rate": 3e-05, | |
| "loss": 0.404, | |
| "step": 12208 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "learning_rate": 3e-05, | |
| "loss": 0.4039, | |
| "step": 12644 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7551407100982113, | |
| "eval_loss": 0.4728190004825592, | |
| "eval_runtime": 223.8973, | |
| "eval_samples_per_second": 33.073, | |
| "eval_steps_per_second": 2.068, | |
| "step": 13074 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_exact_match": 16.12424037812289, | |
| "eval_f1": 23.0965386152669, | |
| "eval_qa_bleu": 10.706681466541065, | |
| "eval_qa_exact_match": 0.1524645509790682, | |
| "eval_recite_bleu": 31.825156559426375, | |
| "eval_recite_exact_match": 0.0036461850101282916, | |
| "step": 13074 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 3e-05, | |
| "loss": 0.405, | |
| "step": 13080 | |
| }, | |
| { | |
| "epoch": 6.2, | |
| "learning_rate": 3e-05, | |
| "loss": 0.285, | |
| "step": 13516 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2984, | |
| "step": 13952 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "learning_rate": 3e-05, | |
| "loss": 0.3051, | |
| "step": 14388 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "learning_rate": 3e-05, | |
| "loss": 0.3044, | |
| "step": 14824 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.7600033806411041, | |
| "eval_loss": 0.4375591278076172, | |
| "eval_runtime": 225.3993, | |
| "eval_samples_per_second": 32.853, | |
| "eval_steps_per_second": 2.054, | |
| "step": 15253 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_exact_match": 16.948008102633356, | |
| "eval_f1": 23.931422337180113, | |
| "eval_qa_bleu": 11.914541362683222, | |
| "eval_qa_exact_match": 0.16029709655638083, | |
| "eval_recite_bleu": 35.91072342079755, | |
| "eval_recite_exact_match": 0.010938555030384874, | |
| "step": 15253 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2046, | |
| "step": 15260 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2265, | |
| "step": 15696 | |
| }, | |
| { | |
| "epoch": 7.4, | |
| "learning_rate": 3e-05, | |
| "loss": 0.236, | |
| "step": 16132 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2399, | |
| "step": 16568 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2446, | |
| "step": 17004 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7628389388058522, | |
| "eval_loss": 0.42204150557518005, | |
| "eval_runtime": 223.5169, | |
| "eval_samples_per_second": 33.129, | |
| "eval_steps_per_second": 2.071, | |
| "step": 17432 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_exact_match": 18.433490884537473, | |
| "eval_f1": 25.859990277646197, | |
| "eval_qa_bleu": 13.293068057422849, | |
| "eval_qa_exact_match": 0.17609723160027008, | |
| "eval_recite_bleu": 41.25950959223269, | |
| "eval_recite_exact_match": 0.015530047265361242, | |
| "step": 17432 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2466, | |
| "step": 17440 | |
| }, | |
| { | |
| "epoch": 8.2, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1887, | |
| "step": 17876 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1967, | |
| "step": 18312 | |
| }, | |
| { | |
| "epoch": 8.6, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2026, | |
| "step": 18748 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2039, | |
| "step": 19184 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.7641864696201232, | |
| "eval_loss": 0.41897761821746826, | |
| "eval_runtime": 222.9387, | |
| "eval_samples_per_second": 33.215, | |
| "eval_steps_per_second": 2.077, | |
| "step": 19611 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_exact_match": 18.933153274814316, | |
| "eval_f1": 26.522905186169627, | |
| "eval_qa_bleu": 12.546175820192783, | |
| "eval_qa_exact_match": 0.1797434166103984, | |
| "eval_recite_bleu": 44.27793915167843, | |
| "eval_recite_exact_match": 0.02228224172856178, | |
| "step": 19611 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2067, | |
| "step": 19620 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1667, | |
| "step": 20056 | |
| }, | |
| { | |
| "epoch": 9.4, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1709, | |
| "step": 20492 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1774, | |
| "step": 20928 | |
| }, | |
| { | |
| "epoch": 9.8, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1787, | |
| "step": 21364 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.7648945867029934, | |
| "eval_loss": 0.4250437319278717, | |
| "eval_runtime": 225.6304, | |
| "eval_samples_per_second": 32.819, | |
| "eval_steps_per_second": 2.052, | |
| "step": 21790 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_exact_match": 20.054017555705606, | |
| "eval_f1": 27.607870062960497, | |
| "eval_qa_bleu": 14.597537360428676, | |
| "eval_qa_exact_match": 0.1913571910871033, | |
| "eval_recite_bleu": 46.3428836838782, | |
| "eval_recite_exact_match": 0.029169480081026333, | |
| "step": 21790 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1829, | |
| "step": 21800 | |
| }, | |
| { | |
| "epoch": 10.2, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1498, | |
| "step": 22236 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1553, | |
| "step": 22672 | |
| }, | |
| { | |
| "epoch": 10.6, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1612, | |
| "step": 23108 | |
| }, | |
| { | |
| "epoch": 10.8, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1652, | |
| "step": 23544 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.7653962302216591, | |
| "eval_loss": 0.42946067452430725, | |
| "eval_runtime": 227.2445, | |
| "eval_samples_per_second": 32.586, | |
| "eval_steps_per_second": 2.037, | |
| "step": 23969 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_exact_match": 19.581363943281566, | |
| "eval_f1": 27.25909178622117, | |
| "eval_qa_bleu": 14.227446461856255, | |
| "eval_qa_exact_match": 0.18595543551654287, | |
| "eval_recite_bleu": 46.797968452871594, | |
| "eval_recite_exact_match": 0.03430114787305875, | |
| "step": 23969 | |
| }, | |
| { | |
| "epoch": 11.01, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1439, | |
| "step": 23980 | |
| }, | |
| { | |
| "epoch": 11.21, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1402, | |
| "step": 24416 | |
| }, | |
| { | |
| "epoch": 11.41, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1462, | |
| "step": 24852 | |
| }, | |
| { | |
| "epoch": 11.61, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1492, | |
| "step": 25288 | |
| }, | |
| { | |
| "epoch": 11.81, | |
| "learning_rate": 3e-05, | |
| "loss": 0.154, | |
| "step": 25724 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.7654645700633325, | |
| "eval_loss": 0.43657177686691284, | |
| "eval_runtime": 220.7713, | |
| "eval_samples_per_second": 33.541, | |
| "eval_steps_per_second": 2.097, | |
| "step": 26148 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_exact_match": 19.945982444294394, | |
| "eval_f1": 27.398959656999555, | |
| "eval_qa_bleu": 13.50737485018698, | |
| "eval_qa_exact_match": 0.1886563133018231, | |
| "eval_recite_bleu": 47.72099251148126, | |
| "eval_recite_exact_match": 0.036326806212018906, | |
| "step": 26148 | |
| }, | |
| { | |
| "epoch": 12.01, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1569, | |
| "step": 26160 | |
| }, | |
| { | |
| "epoch": 12.21, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1344, | |
| "step": 26596 | |
| }, | |
| { | |
| "epoch": 12.41, | |
| "learning_rate": 3e-05, | |
| "loss": 0.14, | |
| "step": 27032 | |
| }, | |
| { | |
| "epoch": 12.61, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1427, | |
| "step": 27468 | |
| }, | |
| { | |
| "epoch": 12.81, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1441, | |
| "step": 27904 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.7656634099218181, | |
| "eval_loss": 0.44285184144973755, | |
| "eval_runtime": 222.075, | |
| "eval_samples_per_second": 33.345, | |
| "eval_steps_per_second": 2.085, | |
| "step": 28327 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_exact_match": 20.405131667792034, | |
| "eval_f1": 28.457441663021587, | |
| "eval_qa_bleu": 14.596564409389117, | |
| "eval_qa_exact_match": 0.19446320054017555, | |
| "eval_recite_bleu": 48.14075990339527, | |
| "eval_recite_exact_match": 0.03902768399729912, | |
| "step": 28327 | |
| } | |
| ], | |
| "logging_steps": 436, | |
| "max_steps": 43580, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 1.3113812200077312e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |