|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 14.4, |
|
"eval_steps": 500, |
|
"global_step": 180, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 1.2, |
|
"grad_norm": 0.8679821491241455, |
|
"learning_rate": 0.00019868265225415265, |
|
"loss": 4.9095, |
|
"step": 15 |
|
}, |
|
{ |
|
"epoch": 2.4, |
|
"grad_norm": 1.064207911491394, |
|
"learning_rate": 0.00019075754196709572, |
|
"loss": 3.2019, |
|
"step": 30 |
|
}, |
|
{ |
|
"epoch": 3.6, |
|
"grad_norm": 0.6014457941055298, |
|
"learning_rate": 0.00017621620551276366, |
|
"loss": 1.7706, |
|
"step": 45 |
|
}, |
|
{ |
|
"epoch": 4.8, |
|
"grad_norm": 0.4954649806022644, |
|
"learning_rate": 0.00015611870653623825, |
|
"loss": 1.2936, |
|
"step": 60 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"grad_norm": 0.418344110250473, |
|
"learning_rate": 0.000131930153013598, |
|
"loss": 1.0465, |
|
"step": 75 |
|
}, |
|
{ |
|
"epoch": 7.2, |
|
"grad_norm": 0.40778854489326477, |
|
"learning_rate": 0.00010541389085854176, |
|
"loss": 0.9186, |
|
"step": 90 |
|
}, |
|
{ |
|
"epoch": 8.4, |
|
"grad_norm": 0.3741260766983032, |
|
"learning_rate": 7.85029559788976e-05, |
|
"loss": 0.8364, |
|
"step": 105 |
|
}, |
|
{ |
|
"epoch": 9.6, |
|
"grad_norm": 0.4041515290737152, |
|
"learning_rate": 5.3159155930021e-05, |
|
"loss": 0.7861, |
|
"step": 120 |
|
}, |
|
{ |
|
"epoch": 10.8, |
|
"grad_norm": 0.3697131872177124, |
|
"learning_rate": 3.123005411465766e-05, |
|
"loss": 0.7488, |
|
"step": 135 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"grad_norm": 0.4183999001979828, |
|
"learning_rate": 1.4314282383241096e-05, |
|
"loss": 0.7274, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 13.2, |
|
"grad_norm": 0.4263518154621124, |
|
"learning_rate": 3.6450007480777093e-06, |
|
"loss": 0.7245, |
|
"step": 165 |
|
}, |
|
{ |
|
"epoch": 14.4, |
|
"grad_norm": 0.3251570165157318, |
|
"learning_rate": 0.0, |
|
"loss": 0.7288, |
|
"step": 180 |
|
}, |
|
{ |
|
"epoch": 14.4, |
|
"step": 180, |
|
"total_flos": 116292530995200.0, |
|
"train_loss": 1.4744000752766928, |
|
"train_runtime": 81.674, |
|
"train_samples_per_second": 18.366, |
|
"train_steps_per_second": 2.204 |
|
} |
|
], |
|
"logging_steps": 15, |
|
"max_steps": 180, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 15, |
|
"save_steps": 500, |
|
"total_flos": 116292530995200.0, |
|
"train_batch_size": 4, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|