| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 299.94117647058823, | |
| "global_step": 1200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 24.94, | |
| "learning_rate": 3.4649999999999995e-05, | |
| "loss": 10.2127, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 49.94, | |
| "learning_rate": 6.965e-05, | |
| "loss": 3.5438, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 74.94, | |
| "learning_rate": 6.306999999999999e-05, | |
| "loss": 3.0874, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 99.94, | |
| "learning_rate": 5.607e-05, | |
| "loss": 2.9853, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 124.94, | |
| "learning_rate": 4.906999999999999e-05, | |
| "loss": 2.9022, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 124.94, | |
| "eval_loss": 2.776294469833374, | |
| "eval_runtime": 20.018, | |
| "eval_samples_per_second": 25.427, | |
| "eval_steps_per_second": 3.197, | |
| "eval_wer": 0.9824253075571178, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 149.94, | |
| "learning_rate": 4.2069999999999995e-05, | |
| "loss": 2.7184, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 174.94, | |
| "learning_rate": 3.5069999999999995e-05, | |
| "loss": 2.2828, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 199.94, | |
| "learning_rate": 2.807e-05, | |
| "loss": 1.8793, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 224.94, | |
| "learning_rate": 2.1069999999999996e-05, | |
| "loss": 1.6491, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 249.94, | |
| "learning_rate": 1.4069999999999999e-05, | |
| "loss": 1.5112, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 249.94, | |
| "eval_loss": 1.173567533493042, | |
| "eval_runtime": 20.0655, | |
| "eval_samples_per_second": 25.367, | |
| "eval_steps_per_second": 3.19, | |
| "eval_wer": 0.7404803749267721, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 274.94, | |
| "learning_rate": 7.07e-06, | |
| "loss": 1.4311, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 299.94, | |
| "learning_rate": 6.999999999999999e-08, | |
| "loss": 1.3867, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 299.94, | |
| "step": 1200, | |
| "total_flos": 4.700653235698678e+19, | |
| "train_loss": 2.9658275985717775, | |
| "train_runtime": 15272.387, | |
| "train_samples_per_second": 20.331, | |
| "train_steps_per_second": 0.079 | |
| } | |
| ], | |
| "max_steps": 1200, | |
| "num_train_epochs": 300, | |
| "total_flos": 4.700653235698678e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |