| { | |
| "best_metric": 58.31498729889923, | |
| "best_model_checkpoint": "./whisper-tiny-hi2_test/checkpoint-80", | |
| "epoch": 6.28125, | |
| "eval_steps": 40, | |
| "global_step": 201, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 8.544885635375977, | |
| "learning_rate": 3.75e-05, | |
| "loss": 1.8228, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 6.32118558883667, | |
| "learning_rate": 3.75e-05, | |
| "loss": 0.6803, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "eval_loss": 0.6102083325386047, | |
| "eval_runtime": 542.5987, | |
| "eval_samples_per_second": 1.843, | |
| "eval_steps_per_second": 1.843, | |
| "eval_wer": 65.94411515664692, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.875, | |
| "grad_norm": 6.4600510597229, | |
| "learning_rate": 3.75e-05, | |
| "loss": 0.4966, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 5.194849491119385, | |
| "learning_rate": 3.75e-05, | |
| "loss": 0.3649, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "eval_loss": 0.5036975741386414, | |
| "eval_runtime": 540.6408, | |
| "eval_samples_per_second": 1.85, | |
| "eval_steps_per_second": 1.85, | |
| "eval_wer": 58.31498729889923, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.125, | |
| "grad_norm": 4.111001014709473, | |
| "learning_rate": 3.75e-05, | |
| "loss": 0.3075, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 4.358007431030273, | |
| "learning_rate": 3.75e-05, | |
| "loss": 0.2296, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "eval_loss": 0.48966631293296814, | |
| "eval_runtime": 538.04, | |
| "eval_samples_per_second": 1.859, | |
| "eval_steps_per_second": 1.859, | |
| "eval_wer": 66.71464860287891, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.375, | |
| "grad_norm": 4.111566066741943, | |
| "learning_rate": 3.75e-05, | |
| "loss": 0.1811, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 6.702174186706543, | |
| "learning_rate": 3.75e-05, | |
| "loss": 0.1564, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.5130774974822998, | |
| "eval_runtime": 529.7879, | |
| "eval_samples_per_second": 1.888, | |
| "eval_steps_per_second": 1.888, | |
| "eval_wer": 65.34292972057578, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 5.625, | |
| "grad_norm": 3.0882856845855713, | |
| "learning_rate": 3.75e-05, | |
| "loss": 0.101, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "grad_norm": 4.524318695068359, | |
| "learning_rate": 3.75e-05, | |
| "loss": 0.0905, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "eval_loss": 0.5503659248352051, | |
| "eval_runtime": 587.5545, | |
| "eval_samples_per_second": 1.702, | |
| "eval_steps_per_second": 1.702, | |
| "eval_wer": 63.903471634208294, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "step": 200, | |
| "total_flos": 1.5401574531072e+17, | |
| "train_loss": 0.44307032763957976, | |
| "train_runtime": 3561.8571, | |
| "train_samples_per_second": 1.797, | |
| "train_steps_per_second": 0.056 | |
| }, | |
| { | |
| "epoch": 6.28125, | |
| "step": 201, | |
| "total_flos": 1.5480354963456e+17, | |
| "train_loss": 0.04041733433358112, | |
| "train_runtime": 8.5506, | |
| "train_samples_per_second": 748.49, | |
| "train_steps_per_second": 23.39 | |
| }, | |
| { | |
| "epoch": 6.28125, | |
| "eval_loss": 0.5036975741386414, | |
| "eval_runtime": 613.1737, | |
| "eval_samples_per_second": 1.631, | |
| "eval_steps_per_second": 1.631, | |
| "eval_wer": 58.31498729889923, | |
| "step": 201 | |
| } | |
| ], | |
| "logging_steps": 20, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 7, | |
| "save_steps": 40, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.5480354963456e+17, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |