| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 100.0, | |
| "global_step": 341600, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 9.474208061136448e-05, | |
| "loss": 14.0127, | |
| "step": 34160 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_cer": 0.9619047619047619, | |
| "eval_loss": 3.3398807048797607, | |
| "eval_runtime": 1.0489, | |
| "eval_samples_per_second": 200.219, | |
| "eval_steps_per_second": 25.742, | |
| "eval_wer": 1.0, | |
| "step": 34160 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "learning_rate": 8.422007888573894e-05, | |
| "loss": 2.7792, | |
| "step": 68320 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_cer": 0.9402597402597402, | |
| "eval_loss": 3.2982892990112305, | |
| "eval_runtime": 1.0271, | |
| "eval_samples_per_second": 204.45, | |
| "eval_steps_per_second": 26.286, | |
| "eval_wer": 1.0, | |
| "step": 68320 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "learning_rate": 7.36980771601134e-05, | |
| "loss": 2.7144, | |
| "step": 102480 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "eval_cer": 0.935064935064935, | |
| "eval_loss": 3.2650129795074463, | |
| "eval_runtime": 1.0213, | |
| "eval_samples_per_second": 205.615, | |
| "eval_steps_per_second": 26.436, | |
| "eval_wer": 1.0, | |
| "step": 102480 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "learning_rate": 6.317607543448786e-05, | |
| "loss": 2.6884, | |
| "step": 136640 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "eval_cer": 0.9341991341991343, | |
| "eval_loss": 3.2275328636169434, | |
| "eval_runtime": 1.0304, | |
| "eval_samples_per_second": 203.803, | |
| "eval_steps_per_second": 26.203, | |
| "eval_wer": 1.0, | |
| "step": 136640 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "learning_rate": 5.26537655614446e-05, | |
| "loss": 2.6742, | |
| "step": 170800 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "eval_cer": 0.935064935064935, | |
| "eval_loss": 3.228600025177002, | |
| "eval_runtime": 1.0287, | |
| "eval_samples_per_second": 204.145, | |
| "eval_steps_per_second": 26.247, | |
| "eval_wer": 1.0, | |
| "step": 170800 | |
| }, | |
| { | |
| "epoch": 60.0, | |
| "learning_rate": 4.213176383581906e-05, | |
| "loss": 2.6644, | |
| "step": 204960 | |
| }, | |
| { | |
| "epoch": 60.0, | |
| "eval_cer": 0.935064935064935, | |
| "eval_loss": 3.192678213119507, | |
| "eval_runtime": 1.0316, | |
| "eval_samples_per_second": 203.564, | |
| "eval_steps_per_second": 26.172, | |
| "eval_wer": 1.0, | |
| "step": 204960 | |
| }, | |
| { | |
| "epoch": 70.0, | |
| "learning_rate": 3.161007025761124e-05, | |
| "loss": 2.659, | |
| "step": 239120 | |
| }, | |
| { | |
| "epoch": 70.0, | |
| "eval_cer": 0.9333333333333333, | |
| "eval_loss": 3.189307689666748, | |
| "eval_runtime": 1.0408, | |
| "eval_samples_per_second": 201.77, | |
| "eval_steps_per_second": 25.942, | |
| "eval_wer": 1.0, | |
| "step": 239120 | |
| }, | |
| { | |
| "epoch": 80.0, | |
| "learning_rate": 2.108776038456798e-05, | |
| "loss": 2.6548, | |
| "step": 273280 | |
| }, | |
| { | |
| "epoch": 80.0, | |
| "eval_cer": 0.9324675324675324, | |
| "eval_loss": 3.1870908737182617, | |
| "eval_runtime": 1.0301, | |
| "eval_samples_per_second": 203.86, | |
| "eval_steps_per_second": 26.211, | |
| "eval_wer": 1.0, | |
| "step": 273280 | |
| }, | |
| { | |
| "epoch": 90.0, | |
| "learning_rate": 1.0566066806360162e-05, | |
| "loss": 2.6524, | |
| "step": 307440 | |
| }, | |
| { | |
| "epoch": 90.0, | |
| "eval_cer": 0.9324675324675324, | |
| "eval_loss": 3.1949210166931152, | |
| "eval_runtime": 1.0305, | |
| "eval_samples_per_second": 203.786, | |
| "eval_steps_per_second": 26.201, | |
| "eval_wer": 1.0, | |
| "step": 307440 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "learning_rate": 4.4373228152348083e-08, | |
| "loss": 2.651, | |
| "step": 341600 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "eval_cer": 0.9324675324675324, | |
| "eval_loss": 3.189643144607544, | |
| "eval_runtime": 1.0382, | |
| "eval_samples_per_second": 202.281, | |
| "eval_steps_per_second": 26.008, | |
| "eval_wer": 1.0, | |
| "step": 341600 | |
| } | |
| ], | |
| "max_steps": 341600, | |
| "num_train_epochs": 100, | |
| "total_flos": 3.24550895516716e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |