| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 15.0, | |
| "global_step": 11085, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 9.831924793466908e-05, | |
| "loss": 18.2448, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_cer": 1.0, | |
| "eval_loss": 3.3216891288757324, | |
| "eval_new_wer": 1.0, | |
| "eval_old_wer": 1.0, | |
| "eval_runtime": 7.2541, | |
| "eval_samples_per_second": 24.814, | |
| "eval_steps_per_second": 3.171, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 9.130187066755295e-05, | |
| "loss": 3.4894, | |
| "step": 1478 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_cer": 1.0, | |
| "eval_loss": 3.335986375808716, | |
| "eval_new_wer": 1.0, | |
| "eval_old_wer": 1.0, | |
| "eval_runtime": 7.2646, | |
| "eval_samples_per_second": 24.778, | |
| "eval_steps_per_second": 3.166, | |
| "step": 1478 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 8.428449340043682e-05, | |
| "loss": 3.4869, | |
| "step": 2217 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_cer": 1.0, | |
| "eval_loss": 3.368795394897461, | |
| "eval_new_wer": 1.0, | |
| "eval_old_wer": 1.0, | |
| "eval_runtime": 7.3486, | |
| "eval_samples_per_second": 24.494, | |
| "eval_steps_per_second": 3.13, | |
| "step": 2217 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 7.726711613332067e-05, | |
| "loss": 3.4836, | |
| "step": 2956 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_cer": 1.0, | |
| "eval_loss": 3.357470750808716, | |
| "eval_new_wer": 1.0, | |
| "eval_old_wer": 1.0, | |
| "eval_runtime": 7.3992, | |
| "eval_samples_per_second": 24.327, | |
| "eval_steps_per_second": 3.108, | |
| "step": 2956 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 7.024973886620455e-05, | |
| "loss": 3.4864, | |
| "step": 3695 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_cer": 1.0, | |
| "eval_loss": 3.36539888381958, | |
| "eval_new_wer": 1.0, | |
| "eval_old_wer": 1.0, | |
| "eval_runtime": 7.5815, | |
| "eval_samples_per_second": 23.742, | |
| "eval_steps_per_second": 3.034, | |
| "step": 3695 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 6.323236159908841e-05, | |
| "loss": 3.4849, | |
| "step": 4434 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_cer": 1.0, | |
| "eval_loss": 3.3517842292785645, | |
| "eval_new_wer": 1.0, | |
| "eval_old_wer": 1.0, | |
| "eval_runtime": 7.7896, | |
| "eval_samples_per_second": 23.108, | |
| "eval_steps_per_second": 2.953, | |
| "step": 4434 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 5.6214984331972276e-05, | |
| "loss": 3.484, | |
| "step": 5173 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_cer": 1.0, | |
| "eval_loss": 3.3279595375061035, | |
| "eval_new_wer": 1.0, | |
| "eval_old_wer": 1.0, | |
| "eval_runtime": 7.3916, | |
| "eval_samples_per_second": 24.352, | |
| "eval_steps_per_second": 3.112, | |
| "step": 5173 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 4.9197607064856144e-05, | |
| "loss": 3.4854, | |
| "step": 5912 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_cer": 1.0, | |
| "eval_loss": 3.361226797103882, | |
| "eval_new_wer": 1.0, | |
| "eval_old_wer": 1.0, | |
| "eval_runtime": 7.4066, | |
| "eval_samples_per_second": 24.303, | |
| "eval_steps_per_second": 3.105, | |
| "step": 5912 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "learning_rate": 4.2180229797740006e-05, | |
| "loss": 3.4807, | |
| "step": 6651 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_cer": 1.0, | |
| "eval_loss": 3.3284316062927246, | |
| "eval_new_wer": 1.0, | |
| "eval_old_wer": 1.0, | |
| "eval_runtime": 7.4013, | |
| "eval_samples_per_second": 24.32, | |
| "eval_steps_per_second": 3.108, | |
| "step": 6651 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 3.5162852530623875e-05, | |
| "loss": 3.4854, | |
| "step": 7390 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_cer": 1.0, | |
| "eval_loss": 3.360283851623535, | |
| "eval_new_wer": 1.0, | |
| "eval_old_wer": 1.0, | |
| "eval_runtime": 7.389, | |
| "eval_samples_per_second": 24.361, | |
| "eval_steps_per_second": 3.113, | |
| "step": 7390 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "learning_rate": 2.8145475263507744e-05, | |
| "loss": 3.4854, | |
| "step": 8129 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_cer": 1.0, | |
| "eval_loss": 3.3530423641204834, | |
| "eval_new_wer": 1.0, | |
| "eval_old_wer": 1.0, | |
| "eval_runtime": 7.4543, | |
| "eval_samples_per_second": 24.147, | |
| "eval_steps_per_second": 3.085, | |
| "step": 8129 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "learning_rate": 2.112809799639161e-05, | |
| "loss": 3.4855, | |
| "step": 8868 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_cer": 1.0, | |
| "eval_loss": 3.3396615982055664, | |
| "eval_new_wer": 1.0, | |
| "eval_old_wer": 1.0, | |
| "eval_runtime": 7.3505, | |
| "eval_samples_per_second": 24.488, | |
| "eval_steps_per_second": 3.129, | |
| "step": 8868 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "learning_rate": 1.4110720729275472e-05, | |
| "loss": 3.4813, | |
| "step": 9607 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_cer": 1.0, | |
| "eval_loss": 3.352962017059326, | |
| "eval_new_wer": 1.0, | |
| "eval_old_wer": 1.0, | |
| "eval_runtime": 7.5315, | |
| "eval_samples_per_second": 23.9, | |
| "eval_steps_per_second": 3.054, | |
| "step": 9607 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "learning_rate": 7.093343462159339e-06, | |
| "loss": 3.4785, | |
| "step": 10346 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_cer": 1.0, | |
| "eval_loss": 3.345838785171509, | |
| "eval_new_wer": 1.0, | |
| "eval_old_wer": 1.0, | |
| "eval_runtime": 7.0543, | |
| "eval_samples_per_second": 25.516, | |
| "eval_steps_per_second": 3.26, | |
| "step": 10346 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "learning_rate": 7.596619504320577e-08, | |
| "loss": 3.4798, | |
| "step": 11085 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_cer": 1.0, | |
| "eval_loss": 3.345510482788086, | |
| "eval_new_wer": 1.0, | |
| "eval_old_wer": 1.0, | |
| "eval_runtime": 7.0297, | |
| "eval_samples_per_second": 25.606, | |
| "eval_steps_per_second": 3.272, | |
| "step": 11085 | |
| } | |
| ], | |
| "max_steps": 11085, | |
| "num_train_epochs": 15, | |
| "total_flos": 8.649191629868698e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |