| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 30, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "eval_cer": 0.5559814332934806, | |
| "eval_loss": 3.5223801136016846, | |
| "eval_runtime": 54.3029, | |
| "eval_samples_per_second": 11.51, | |
| "eval_steps_per_second": 0.055, | |
| "eval_wer": 1.131116687578419, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 23.496896743774414, | |
| "learning_rate": 6.000000000000001e-07, | |
| "loss": 2.4889, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_cer": 0.5018285392784303, | |
| "eval_loss": 3.4807357788085938, | |
| "eval_runtime": 52.1391, | |
| "eval_samples_per_second": 11.987, | |
| "eval_steps_per_second": 0.058, | |
| "eval_wer": 1.0566185696361354, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_cer": 0.4985934313242844, | |
| "eval_loss": 3.2107956409454346, | |
| "eval_runtime": 53.2554, | |
| "eval_samples_per_second": 11.736, | |
| "eval_steps_per_second": 0.056, | |
| "eval_wer": 1.0561480552070264, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 17.40036964416504, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 2.3707, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_cer": 0.5154722554328716, | |
| "eval_loss": 2.944542169570923, | |
| "eval_runtime": 53.7146, | |
| "eval_samples_per_second": 11.636, | |
| "eval_steps_per_second": 0.056, | |
| "eval_wer": 1.0583437892095358, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 12.635431289672852, | |
| "learning_rate": 2.6e-06, | |
| "loss": 2.0528, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_cer": 0.5467684084675434, | |
| "eval_loss": 2.6806018352508545, | |
| "eval_runtime": 53.3587, | |
| "eval_samples_per_second": 11.713, | |
| "eval_steps_per_second": 0.056, | |
| "eval_wer": 1.0472082810539523, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 30, | |
| "total_flos": 9.962495329645363e+17, | |
| "train_loss": 2.304109636942546, | |
| "train_runtime": 1167.3913, | |
| "train_samples_per_second": 12.678, | |
| "train_steps_per_second": 0.026 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 30, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9.962495329645363e+17, | |
| "train_batch_size": 256, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |