| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.01774622892635315, | |
| "eval_steps": 2, | |
| "global_step": 5, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00354924578527063, | |
| "grad_norm": NaN, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00354924578527063, | |
| "eval_loss": NaN, | |
| "eval_runtime": 103.1348, | |
| "eval_samples_per_second": 4.606, | |
| "eval_steps_per_second": 2.308, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00709849157054126, | |
| "grad_norm": NaN, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.00709849157054126, | |
| "eval_loss": NaN, | |
| "eval_runtime": 4.2297, | |
| "eval_samples_per_second": 112.3, | |
| "eval_steps_per_second": 56.268, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.01064773735581189, | |
| "grad_norm": NaN, | |
| "learning_rate": 3e-05, | |
| "loss": 0.0, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.01419698314108252, | |
| "grad_norm": NaN, | |
| "learning_rate": 4e-05, | |
| "loss": 0.0, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.01419698314108252, | |
| "eval_loss": NaN, | |
| "eval_runtime": 3.8846, | |
| "eval_samples_per_second": 122.278, | |
| "eval_steps_per_second": 61.268, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.01774622892635315, | |
| "grad_norm": NaN, | |
| "learning_rate": 5e-05, | |
| "loss": 0.0, | |
| "step": 5 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 5, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 2, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 581926750519296.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |