{ "best_metric": 0.6540625925560273, "best_model_checkpoint": "whisper-base-bn/checkpoint-200", "epoch": 20.0, "eval_steps": 50, "global_step": 200, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 1.0, "grad_norm": 79.03315734863281, "learning_rate": 1.4000000000000001e-06, "loss": 1.6251, "step": 10 }, { "epoch": 2.0, "grad_norm": 23.89545249938965, "learning_rate": 3.4000000000000005e-06, "loss": 1.1182, "step": 20 }, { "epoch": 3.0, "grad_norm": 13.786792755126953, "learning_rate": 5.400000000000001e-06, "loss": 0.7697, "step": 30 }, { "epoch": 4.0, "grad_norm": 8.665822982788086, "learning_rate": 7.4e-06, "loss": 0.5954, "step": 40 }, { "epoch": 5.0, "grad_norm": 7.227974891662598, "learning_rate": 9.4e-06, "loss": 0.4909, "step": 50 }, { "epoch": 5.0, "eval_loss": 0.45442524552345276, "eval_runtime": 164.709, "eval_samples_per_second": 1.919, "eval_steps_per_second": 0.061, "eval_wer": 1.0225519287833829, "step": 50 }, { "epoch": 6.0, "grad_norm": 8.914044380187988, "learning_rate": 9.533333333333334e-06, "loss": 0.4142, "step": 60 }, { "epoch": 7.0, "grad_norm": 7.179055213928223, "learning_rate": 8.866666666666668e-06, "loss": 0.3577, "step": 70 }, { "epoch": 8.0, "grad_norm": 7.231939792633057, "learning_rate": 8.2e-06, "loss": 0.3156, "step": 80 }, { "epoch": 9.0, "grad_norm": 5.616255283355713, "learning_rate": 7.533333333333334e-06, "loss": 0.2797, "step": 90 }, { "epoch": 10.0, "grad_norm": 7.509573936462402, "learning_rate": 6.866666666666667e-06, "loss": 0.2505, "step": 100 }, { "epoch": 10.0, "eval_loss": 0.295759916305542, "eval_runtime": 163.8777, "eval_samples_per_second": 1.928, "eval_steps_per_second": 0.061, "eval_wer": 0.8072217837411207, "step": 100 }, { "epoch": 11.0, "grad_norm": 7.341796398162842, "learning_rate": 6.200000000000001e-06, "loss": 0.2277, "step": 110 }, { "epoch": 12.0, "grad_norm": 6.398680686950684, "learning_rate": 5.533333333333334e-06, "loss": 0.2071, "step": 120 }, { "epoch": 13.0, "grad_norm": 5.016919136047363, "learning_rate": 4.866666666666667e-06, "loss": 0.1882, "step": 130 }, { "epoch": 14.0, "grad_norm": 4.969520568847656, "learning_rate": 4.2000000000000004e-06, "loss": 0.1717, "step": 140 }, { "epoch": 15.0, "grad_norm": 5.127407073974609, "learning_rate": 3.5333333333333335e-06, "loss": 0.1582, "step": 150 }, { "epoch": 15.0, "eval_loss": 0.2277686595916748, "eval_runtime": 164.435, "eval_samples_per_second": 1.922, "eval_steps_per_second": 0.061, "eval_wer": 0.7243742550655542, "step": 150 }, { "epoch": 16.0, "grad_norm": 4.846920013427734, "learning_rate": 2.866666666666667e-06, "loss": 0.1477, "step": 160 }, { "epoch": 17.0, "grad_norm": 4.428435325622559, "learning_rate": 2.2e-06, "loss": 0.1398, "step": 170 }, { "epoch": 18.0, "grad_norm": 4.65584135055542, "learning_rate": 1.5333333333333334e-06, "loss": 0.132, "step": 180 }, { "epoch": 19.0, "grad_norm": 4.605101585388184, "learning_rate": 8.666666666666668e-07, "loss": 0.1282, "step": 190 }, { "epoch": 20.0, "grad_norm": 4.232695579528809, "learning_rate": 2.0000000000000002e-07, "loss": 0.124, "step": 200 }, { "epoch": 20.0, "eval_loss": 0.1989278793334961, "eval_runtime": 164.8836, "eval_samples_per_second": 1.917, "eval_steps_per_second": 0.061, "eval_wer": 0.6540625925560273, "step": 200 } ], "logging_steps": 10, "max_steps": 200, "num_input_tokens_seen": 0, "num_train_epochs": 20, "save_steps": 50, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 7.38778630255018e+18, "train_batch_size": 16, "trial_name": null, "trial_params": null }