| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.03076923076923077, | |
| "eval_steps": 3, | |
| "global_step": 3, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0, | |
| "eval_loss": 2.8341026306152344, | |
| "eval_runtime": 12.1694, | |
| "eval_samples_per_second": 3.451, | |
| "eval_steps_per_second": 1.726, | |
| "step": 0 | |
| }, | |
| { | |
| "epoch": 0.010256410256410256, | |
| "grad_norm": 1.5949193239212036, | |
| "learning_rate": 0.0, | |
| "loss": 1.9494, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.020512820512820513, | |
| "grad_norm": 1.186781644821167, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3728, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03076923076923077, | |
| "grad_norm": 1.1918362379074097, | |
| "learning_rate": 4e-05, | |
| "loss": 2.3531, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.03076923076923077, | |
| "eval_loss": 2.829076051712036, | |
| "eval_runtime": 9.3467, | |
| "eval_samples_per_second": 4.494, | |
| "eval_steps_per_second": 2.247, | |
| "step": 3 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 10, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 3, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 32846836137984.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |