| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 2440, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.196406364440918, | |
| "learning_rate": 4.75e-05, | |
| "loss": 0.5417, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7543859649122807, | |
| "eval_f1": 0.6730602006688964, | |
| "eval_loss": 0.47316503524780273, | |
| "eval_precision": 0.7027985359158151, | |
| "eval_recall": 0.6612111292962357, | |
| "eval_runtime": 1.7797, | |
| "eval_samples_per_second": 224.194, | |
| "eval_steps_per_second": 28.094, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 3.380220651626587, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.4395, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7919799498746867, | |
| "eval_f1": 0.7705259948585406, | |
| "eval_loss": 0.4128379225730896, | |
| "eval_precision": 0.7612839958158996, | |
| "eval_recall": 0.8028277868703401, | |
| "eval_runtime": 1.7767, | |
| "eval_samples_per_second": 224.574, | |
| "eval_steps_per_second": 28.142, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 2.2583518028259277, | |
| "learning_rate": 4.25e-05, | |
| "loss": 0.3319, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8646616541353384, | |
| "eval_f1": 0.8315033783783784, | |
| "eval_loss": 0.32298392057418823, | |
| "eval_precision": 0.8438775510204082, | |
| "eval_recall": 0.8217403164211674, | |
| "eval_runtime": 1.7811, | |
| "eval_samples_per_second": 224.023, | |
| "eval_steps_per_second": 28.073, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.8704787492752075, | |
| "learning_rate": 4e-05, | |
| "loss": 0.2873, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8521303258145363, | |
| "eval_f1": 0.8237962290701417, | |
| "eval_loss": 0.322201669216156, | |
| "eval_precision": 0.8201159969225307, | |
| "eval_recall": 0.8278777959629023, | |
| "eval_runtime": 1.7743, | |
| "eval_samples_per_second": 224.873, | |
| "eval_steps_per_second": 28.18, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.8079779148101807, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.2571, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8721804511278195, | |
| "eval_f1": 0.8484099018899409, | |
| "eval_loss": 0.29681602120399475, | |
| "eval_precision": 0.8430645161290322, | |
| "eval_recall": 0.8545644662665939, | |
| "eval_runtime": 1.7796, | |
| "eval_samples_per_second": 224.208, | |
| "eval_steps_per_second": 28.096, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 2.8642373085021973, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.2443, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8671679197994987, | |
| "eval_f1": 0.8466330637850383, | |
| "eval_loss": 0.29177311062812805, | |
| "eval_precision": 0.8353276671885485, | |
| "eval_recall": 0.8635206401163849, | |
| "eval_runtime": 1.7765, | |
| "eval_samples_per_second": 224.594, | |
| "eval_steps_per_second": 28.145, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.3272399306297302, | |
| "learning_rate": 3.2500000000000004e-05, | |
| "loss": 0.2256, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8646616541353384, | |
| "eval_f1": 0.8447157518450185, | |
| "eval_loss": 0.2981509566307068, | |
| "eval_precision": 0.8325401217487549, | |
| "eval_recall": 0.864248045099109, | |
| "eval_runtime": 1.7799, | |
| "eval_samples_per_second": 224.168, | |
| "eval_steps_per_second": 28.091, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 2.9290239810943604, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2172, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8922305764411027, | |
| "eval_f1": 0.8646934961080748, | |
| "eval_loss": 0.2722471356391907, | |
| "eval_precision": 0.882551000198059, | |
| "eval_recall": 0.8512456810329151, | |
| "eval_runtime": 1.7818, | |
| "eval_samples_per_second": 223.935, | |
| "eval_steps_per_second": 28.062, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 6.677186012268066, | |
| "learning_rate": 2.7500000000000004e-05, | |
| "loss": 0.2049, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.8947368421052632, | |
| "eval_f1": 0.8748655913978494, | |
| "eval_loss": 0.264840692281723, | |
| "eval_precision": 0.86983032873807, | |
| "eval_recall": 0.8805237315875614, | |
| "eval_runtime": 1.7819, | |
| "eval_samples_per_second": 223.922, | |
| "eval_steps_per_second": 28.06, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 2.770735502243042, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.1914, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.9072681704260651, | |
| "eval_f1": 0.8848664457009163, | |
| "eval_loss": 0.2680298984050751, | |
| "eval_precision": 0.8977236138837015, | |
| "eval_recall": 0.8743862520458265, | |
| "eval_runtime": 1.7909, | |
| "eval_samples_per_second": 222.787, | |
| "eval_steps_per_second": 27.918, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.3618270754814148, | |
| "learning_rate": 2.25e-05, | |
| "loss": 0.1724, | |
| "step": 1342 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.899749373433584, | |
| "eval_f1": 0.8808243727598566, | |
| "eval_loss": 0.264539897441864, | |
| "eval_precision": 0.875706963591375, | |
| "eval_recall": 0.8865702855064557, | |
| "eval_runtime": 1.7802, | |
| "eval_samples_per_second": 224.138, | |
| "eval_steps_per_second": 28.087, | |
| "step": 1342 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 2.054783582687378, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1689, | |
| "step": 1464 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.899749373433584, | |
| "eval_f1": 0.881931703852755, | |
| "eval_loss": 0.27462852001190186, | |
| "eval_precision": 0.8740012737378415, | |
| "eval_recall": 0.8915711947626841, | |
| "eval_runtime": 1.778, | |
| "eval_samples_per_second": 224.405, | |
| "eval_steps_per_second": 28.121, | |
| "step": 1464 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 0.6106524467468262, | |
| "learning_rate": 1.75e-05, | |
| "loss": 0.1473, | |
| "step": 1586 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.9047619047619048, | |
| "eval_f1": 0.8800854160075926, | |
| "eval_loss": 0.28370755910873413, | |
| "eval_precision": 0.9002425410326267, | |
| "eval_recall": 0.8651118385160939, | |
| "eval_runtime": 1.785, | |
| "eval_samples_per_second": 223.535, | |
| "eval_steps_per_second": 28.012, | |
| "step": 1586 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 4.3967108726501465, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.1577, | |
| "step": 1708 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.9022556390977443, | |
| "eval_f1": 0.884617951284618, | |
| "eval_loss": 0.2892190217971802, | |
| "eval_precision": 0.8772893772893773, | |
| "eval_recall": 0.8933442444080741, | |
| "eval_runtime": 1.7818, | |
| "eval_samples_per_second": 223.936, | |
| "eval_steps_per_second": 28.062, | |
| "step": 1708 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 0.48143357038497925, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.1468, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.9022556390977443, | |
| "eval_f1": 0.8829621606985718, | |
| "eval_loss": 0.27894169092178345, | |
| "eval_precision": 0.8802419354838709, | |
| "eval_recall": 0.8858428805237315, | |
| "eval_runtime": 1.7862, | |
| "eval_samples_per_second": 223.376, | |
| "eval_steps_per_second": 27.992, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.6658376455307007, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1473, | |
| "step": 1952 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.8972431077694235, | |
| "eval_f1": 0.8775533117267087, | |
| "eval_loss": 0.28521186113357544, | |
| "eval_precision": 0.873246730188791, | |
| "eval_recall": 0.8822967812329514, | |
| "eval_runtime": 1.778, | |
| "eval_samples_per_second": 224.41, | |
| "eval_steps_per_second": 28.121, | |
| "step": 1952 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 1.4181182384490967, | |
| "learning_rate": 7.5e-06, | |
| "loss": 0.1274, | |
| "step": 2074 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.9047619047619048, | |
| "eval_f1": 0.8856836962422341, | |
| "eval_loss": 0.28584179282188416, | |
| "eval_precision": 0.8838235294117647, | |
| "eval_recall": 0.8876159301691217, | |
| "eval_runtime": 1.7792, | |
| "eval_samples_per_second": 224.253, | |
| "eval_steps_per_second": 28.102, | |
| "step": 2074 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 5.91610050201416, | |
| "learning_rate": 5e-06, | |
| "loss": 0.1318, | |
| "step": 2196 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.899749373433584, | |
| "eval_f1": 0.8802521008403361, | |
| "eval_loss": 0.29269006848335266, | |
| "eval_precision": 0.8767168083714847, | |
| "eval_recall": 0.8840698308783415, | |
| "eval_runtime": 1.7862, | |
| "eval_samples_per_second": 223.383, | |
| "eval_steps_per_second": 27.993, | |
| "step": 2196 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 6.210901737213135, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.1355, | |
| "step": 2318 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.9072681704260651, | |
| "eval_f1": 0.8884169154604891, | |
| "eval_loss": 0.2884277403354645, | |
| "eval_precision": 0.8874630556728391, | |
| "eval_recall": 0.8893889798145117, | |
| "eval_runtime": 1.7806, | |
| "eval_samples_per_second": 224.08, | |
| "eval_steps_per_second": 28.08, | |
| "step": 2318 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 1.5748217105865479, | |
| "learning_rate": 0.0, | |
| "loss": 0.1367, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.9072681704260651, | |
| "eval_f1": 0.8884169154604891, | |
| "eval_loss": 0.2881532311439514, | |
| "eval_precision": 0.8874630556728391, | |
| "eval_recall": 0.8893889798145117, | |
| "eval_runtime": 1.7811, | |
| "eval_samples_per_second": 224.016, | |
| "eval_steps_per_second": 28.072, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 2440, | |
| "total_flos": 8444128359504000.0, | |
| "train_loss": 0.2206378909408069, | |
| "train_runtime": 624.8021, | |
| "train_samples_per_second": 116.453, | |
| "train_steps_per_second": 3.905 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2440, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 8444128359504000.0, | |
| "train_batch_size": 30, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |