|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 0.014414414414414415, |
|
"eval_steps": 2, |
|
"global_step": 4, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.0036036036036036037, |
|
"grad_norm": 4.331277847290039, |
|
"learning_rate": 1e-05, |
|
"loss": 3.2096, |
|
"step": 1 |
|
}, |
|
{ |
|
"epoch": 0.0036036036036036037, |
|
"eval_loss": 4.149227619171143, |
|
"eval_runtime": 109.8188, |
|
"eval_samples_per_second": 4.262, |
|
"eval_steps_per_second": 2.131, |
|
"step": 1 |
|
}, |
|
{ |
|
"epoch": 0.007207207207207207, |
|
"grad_norm": 3.7934200763702393, |
|
"learning_rate": 2e-05, |
|
"loss": 3.032, |
|
"step": 2 |
|
}, |
|
{ |
|
"epoch": 0.007207207207207207, |
|
"eval_loss": 4.095294952392578, |
|
"eval_runtime": 5.0906, |
|
"eval_samples_per_second": 91.935, |
|
"eval_steps_per_second": 45.967, |
|
"step": 2 |
|
}, |
|
{ |
|
"epoch": 0.010810810810810811, |
|
"grad_norm": 4.124750137329102, |
|
"learning_rate": 3e-05, |
|
"loss": 3.235, |
|
"step": 3 |
|
}, |
|
{ |
|
"epoch": 0.014414414414414415, |
|
"grad_norm": 4.565180778503418, |
|
"learning_rate": 4e-05, |
|
"loss": 3.4669, |
|
"step": 4 |
|
}, |
|
{ |
|
"epoch": 0.014414414414414415, |
|
"eval_loss": 3.972473621368408, |
|
"eval_runtime": 5.1083, |
|
"eval_samples_per_second": 91.616, |
|
"eval_steps_per_second": 45.808, |
|
"step": 4 |
|
} |
|
], |
|
"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": false |
|
}, |
|
"attributes": {} |
|
} |
|
}, |
|
"total_flos": 282696170078208.0, |
|
"train_batch_size": 2, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|