| { | |
| "best_global_step": 100, | |
| "best_metric": 74.07821923770442, | |
| "best_model_checkpoint": "./JUDICM/checkpoint-100", | |
| "epoch": 1.1764705882352942, | |
| "eval_steps": 100, | |
| "global_step": 100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.23529411764705882, | |
| "grad_norm": 9.61864185333252, | |
| "learning_rate": 1.7000000000000002e-06, | |
| "loss": 2.9117, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 8.609731674194336, | |
| "learning_rate": 3.7e-06, | |
| "loss": 2.0419, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7058823529411765, | |
| "grad_norm": 7.390444278717041, | |
| "learning_rate": 5.7e-06, | |
| "loss": 1.5565, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 5.834730625152588, | |
| "learning_rate": 7.7e-06, | |
| "loss": 0.9792, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.1764705882352942, | |
| "grad_norm": 5.353567123413086, | |
| "learning_rate": 9.7e-06, | |
| "loss": 0.8465, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.1764705882352942, | |
| "eval_loss": 0.681628942489624, | |
| "eval_runtime": 544.1503, | |
| "eval_samples_per_second": 1.251, | |
| "eval_steps_per_second": 0.158, | |
| "eval_wer": 74.07821923770442, | |
| "step": 100 | |
| } | |
| ], | |
| "logging_steps": 20, | |
| "max_steps": 850, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.265935704064e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |