| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.7862407862407862, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 2.3000000000000004e-06, | |
| "loss": 2.1385, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 0.5702, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "eval_loss": 0.25573939085006714, | |
| "eval_runtime": 564.0777, | |
| "eval_samples_per_second": 3.469, | |
| "eval_steps_per_second": 0.869, | |
| "eval_wer": 0.10067539046010975, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 7.2000000000000005e-06, | |
| "loss": 0.182, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 9.7e-06, | |
| "loss": 0.1181, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "eval_loss": 0.11441328376531601, | |
| "eval_runtime": 575.4946, | |
| "eval_samples_per_second": 3.401, | |
| "eval_steps_per_second": 0.851, | |
| "eval_wer": 0.07750105529759392, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 7.800000000000002e-06, | |
| "loss": 0.1088, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 5.300000000000001e-06, | |
| "loss": 0.1073, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "eval_loss": 0.07400055974721909, | |
| "eval_runtime": 571.9053, | |
| "eval_samples_per_second": 3.422, | |
| "eval_steps_per_second": 0.857, | |
| "eval_wer": 0.05289151540734487, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 0.088, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 3.0000000000000004e-07, | |
| "loss": 0.0747, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "eval_loss": 0.06278568506240845, | |
| "eval_runtime": 564.3091, | |
| "eval_samples_per_second": 3.468, | |
| "eval_steps_per_second": 0.868, | |
| "eval_wer": 0.036935415787252006, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "step": 200, | |
| "total_flos": 6.531871475907953e+18, | |
| "train_loss": 0.4234592032432556, | |
| "train_runtime": 4041.7087, | |
| "train_samples_per_second": 1.583, | |
| "train_steps_per_second": 0.049 | |
| } | |
| ], | |
| "max_steps": 200, | |
| "num_train_epochs": 1, | |
| "total_flos": 6.531871475907953e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |