| { | |
| "best_metric": 17.79463834092059, | |
| "best_model_checkpoint": "large_config/checkpoint-8000", | |
| "epoch": 5.089, | |
| "global_step": 10000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 9.960000000000001e-06, | |
| "loss": 1.1815, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 9.475789473684212e-06, | |
| "loss": 0.5155, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "eval_cer": 12.13082003827182, | |
| "eval_loss": 0.404460608959198, | |
| "eval_runtime": 2929.9503, | |
| "eval_samples_per_second": 1.571, | |
| "eval_steps_per_second": 0.098, | |
| "eval_wer": 27.7845220030349, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 8.950526315789474e-06, | |
| "loss": 0.3945, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 8.42421052631579e-06, | |
| "loss": 0.5436, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "eval_cer": 11.918136465412264, | |
| "eval_loss": 0.37842079997062683, | |
| "eval_runtime": 2388.6432, | |
| "eval_samples_per_second": 1.927, | |
| "eval_steps_per_second": 0.121, | |
| "eval_wer": 27.098297083122574, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 7.897894736842106e-06, | |
| "loss": 0.4885, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 7.371578947368422e-06, | |
| "loss": 0.265, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "eval_cer": 10.023625272869095, | |
| "eval_loss": 0.3226892054080963, | |
| "eval_runtime": 2404.3282, | |
| "eval_samples_per_second": 1.915, | |
| "eval_steps_per_second": 0.12, | |
| "eval_wer": 22.881470241106054, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 6.845263157894737e-06, | |
| "loss": 0.3096, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 6.318947368421054e-06, | |
| "loss": 0.4765, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "eval_cer": 9.600502808657737, | |
| "eval_loss": 0.30207160115242004, | |
| "eval_runtime": 2483.346, | |
| "eval_samples_per_second": 1.854, | |
| "eval_steps_per_second": 0.116, | |
| "eval_wer": 21.97437194402293, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 5.7926315789473685e-06, | |
| "loss": 0.2808, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "learning_rate": 5.266315789473685e-06, | |
| "loss": 0.1818, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "eval_cer": 9.182992048215759, | |
| "eval_loss": 0.3002430200576782, | |
| "eval_runtime": 2415.8377, | |
| "eval_samples_per_second": 1.906, | |
| "eval_steps_per_second": 0.119, | |
| "eval_wer": 21.03355252065419, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 4.74e-06, | |
| "loss": 0.3291, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "learning_rate": 4.214736842105263e-06, | |
| "loss": 0.4031, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "eval_cer": 7.918114018597186, | |
| "eval_loss": 0.24964982271194458, | |
| "eval_runtime": 2159.9122, | |
| "eval_samples_per_second": 2.132, | |
| "eval_steps_per_second": 0.133, | |
| "eval_wer": 18.091384252234025, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "learning_rate": 3.6884210526315794e-06, | |
| "loss": 0.2113, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.15, | |
| "learning_rate": 3.1621052631578953e-06, | |
| "loss": 0.1991, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.15, | |
| "eval_cer": 9.99837260590688, | |
| "eval_loss": 0.2970982789993286, | |
| "eval_runtime": 2859.4268, | |
| "eval_samples_per_second": 1.61, | |
| "eval_steps_per_second": 0.101, | |
| "eval_wer": 21.702916877423707, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "learning_rate": 2.635789473684211e-06, | |
| "loss": 0.3346, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 4.07, | |
| "learning_rate": 2.110526315789474e-06, | |
| "loss": 0.3023, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 4.07, | |
| "eval_cer": 7.924848063120445, | |
| "eval_loss": 0.24453558027744293, | |
| "eval_runtime": 2309.4868, | |
| "eval_samples_per_second": 1.994, | |
| "eval_steps_per_second": 0.125, | |
| "eval_wer": 17.79463834092059, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 4.12, | |
| "learning_rate": 1.5842105263157894e-06, | |
| "loss": 0.1504, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "learning_rate": 1.0578947368421054e-06, | |
| "loss": 0.2185, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "eval_cer": 9.289053249457067, | |
| "eval_loss": 0.2841958999633789, | |
| "eval_runtime": 1995.3868, | |
| "eval_samples_per_second": 2.307, | |
| "eval_steps_per_second": 0.144, | |
| "eval_wer": 20.375990558084638, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 5.04, | |
| "learning_rate": 5.315789473684211e-07, | |
| "loss": 0.3632, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 5.09, | |
| "learning_rate": 5.263157894736842e-09, | |
| "loss": 0.2114, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 5.09, | |
| "eval_cer": 8.00734010853035, | |
| "eval_loss": 0.24409358203411102, | |
| "eval_runtime": 1991.9589, | |
| "eval_samples_per_second": 2.311, | |
| "eval_steps_per_second": 0.145, | |
| "eval_wer": 17.85702242454898, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 5.09, | |
| "step": 10000, | |
| "total_flos": 1.037552386572288e+19, | |
| "train_loss": 0.3680129928588867, | |
| "train_runtime": 124053.1844, | |
| "train_samples_per_second": 2.58, | |
| "train_steps_per_second": 0.081 | |
| } | |
| ], | |
| "max_steps": 10000, | |
| "num_train_epochs": 9223372036854775807, | |
| "total_flos": 1.037552386572288e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |