| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 122, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01639344262295082, | |
| "grad_norm": 178.06546020507812, | |
| "learning_rate": 4.959016393442623e-05, | |
| "loss": 8.943, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.47540983606557374, | |
| "grad_norm": 54.69188690185547, | |
| "learning_rate": 3.8114754098360655e-05, | |
| "loss": 2.2897, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.9508196721311475, | |
| "grad_norm": 16.769203186035156, | |
| "learning_rate": 2.6229508196721314e-05, | |
| "loss": 2.0108, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_cer": 0.1530448717948718, | |
| "eval_loss": 1.128391981124878, | |
| "eval_runtime": 125.6559, | |
| "eval_samples_per_second": 0.485, | |
| "eval_steps_per_second": 0.064, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.4262295081967213, | |
| "grad_norm": 30.61115837097168, | |
| "learning_rate": 1.4344262295081968e-05, | |
| "loss": 1.1471, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.901639344262295, | |
| "grad_norm": 10.20889663696289, | |
| "learning_rate": 2.459016393442623e-06, | |
| "loss": 0.5665, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_cer": 0.03125, | |
| "eval_loss": 0.5153284072875977, | |
| "eval_runtime": 107.7401, | |
| "eval_samples_per_second": 0.566, | |
| "eval_steps_per_second": 0.074, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 122, | |
| "total_flos": 7.193998576047882e+17, | |
| "train_loss": 1.5048968147058956, | |
| "train_runtime": 791.7321, | |
| "train_samples_per_second": 0.614, | |
| "train_steps_per_second": 0.154 | |
| } | |
| ], | |
| "logging_steps": 29, | |
| "max_steps": 122, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "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": 7.193998576047882e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |