| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.5001611343860781, | |
| "global_step": 2327, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 8.01984817006747e-06, | |
| "loss": 0.2478, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 7.659729258075218e-06, | |
| "loss": 0.222, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 7.303211535202886e-06, | |
| "loss": 0.283, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 6.943092623210635e-06, | |
| "loss": 0.2102, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 6.5829737112183816e-06, | |
| "loss": 0.2127, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "eval_loss": 0.13347241282463074, | |
| "eval_runtime": 385.6175, | |
| "eval_samples_per_second": 14.099, | |
| "eval_steps_per_second": 0.882, | |
| "eval_wer": 0.1047094237500858, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 6.222854799226128e-06, | |
| "loss": 0.1923, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 5.862735887233875e-06, | |
| "loss": 0.2001, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 5.502616975241623e-06, | |
| "loss": 0.199, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 5.14249806324937e-06, | |
| "loss": 0.1944, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 4.782379151257117e-06, | |
| "loss": 0.1976, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "eval_loss": 0.1308828443288803, | |
| "eval_runtime": 385.5506, | |
| "eval_samples_per_second": 14.102, | |
| "eval_steps_per_second": 0.882, | |
| "eval_wer": 0.10391520487905322, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 4.422260239264865e-06, | |
| "loss": 0.1845, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 4.062141327272612e-06, | |
| "loss": 0.1885, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 3.7020224152803586e-06, | |
| "loss": 0.1799, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 3.3419035032881062e-06, | |
| "loss": 0.1749, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 2.981784591295853e-06, | |
| "loss": 0.1887, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "eval_loss": 0.13062627613544464, | |
| "eval_runtime": 383.0748, | |
| "eval_samples_per_second": 14.193, | |
| "eval_steps_per_second": 0.888, | |
| "eval_wer": 0.104032866934021, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "learning_rate": 2.6216656793036003e-06, | |
| "loss": 0.1917, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 2.261546767311348e-06, | |
| "loss": 0.1737, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 1.9014278553190947e-06, | |
| "loss": 0.1758, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 1.541308943326842e-06, | |
| "loss": 0.174, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 1.181190031334589e-06, | |
| "loss": 0.18, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "eval_loss": 0.13111014664173126, | |
| "eval_runtime": 382.6746, | |
| "eval_samples_per_second": 14.208, | |
| "eval_steps_per_second": 0.888, | |
| "eval_wer": 0.10379754282408542, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "learning_rate": 8.210711193423363e-07, | |
| "loss": 0.1667, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "learning_rate": 4.609522073500836e-07, | |
| "loss": 0.1774, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "learning_rate": 1.0083329535783077e-07, | |
| "loss": 0.1779, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "step": 2327, | |
| "total_flos": 1.6280526756920635e+19, | |
| "train_loss": 0.19505117766376623, | |
| "train_runtime": 7387.6715, | |
| "train_samples_per_second": 10.08, | |
| "train_steps_per_second": 0.315 | |
| } | |
| ], | |
| "max_steps": 2327, | |
| "num_train_epochs": 2, | |
| "total_flos": 1.6280526756920635e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |