|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 0.14705882352941177, |
|
"eval_steps": 500, |
|
"global_step": 5000, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.014705882352941176, |
|
"grad_norm": 0.35731062293052673, |
|
"learning_rate": 4.975490196078432e-05, |
|
"loss": 0.5354, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.029411764705882353, |
|
"grad_norm": 2.1595866680145264, |
|
"learning_rate": 4.9509803921568634e-05, |
|
"loss": 0.1627, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 0.04411764705882353, |
|
"grad_norm": 0.14825384318828583, |
|
"learning_rate": 4.9264705882352944e-05, |
|
"loss": 0.101, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 0.058823529411764705, |
|
"grad_norm": 0.07121703773736954, |
|
"learning_rate": 4.901960784313725e-05, |
|
"loss": 0.0952, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 0.07352941176470588, |
|
"grad_norm": 0.06089532747864723, |
|
"learning_rate": 4.877450980392157e-05, |
|
"loss": 0.0881, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 0.08823529411764706, |
|
"grad_norm": 0.037034619599580765, |
|
"learning_rate": 4.8529411764705885e-05, |
|
"loss": 0.0717, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 0.10294117647058823, |
|
"grad_norm": 0.01999847963452339, |
|
"learning_rate": 4.82843137254902e-05, |
|
"loss": 0.0848, |
|
"step": 3500 |
|
}, |
|
{ |
|
"epoch": 0.11764705882352941, |
|
"grad_norm": 0.018650399520993233, |
|
"learning_rate": 4.803921568627452e-05, |
|
"loss": 0.068, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 0.1323529411764706, |
|
"grad_norm": 0.023023229092359543, |
|
"learning_rate": 4.7794117647058826e-05, |
|
"loss": 0.0626, |
|
"step": 4500 |
|
}, |
|
{ |
|
"epoch": 0.14705882352941177, |
|
"grad_norm": 0.018715515732765198, |
|
"learning_rate": 4.7549019607843135e-05, |
|
"loss": 0.0555, |
|
"step": 5000 |
|
} |
|
], |
|
"logging_steps": 500, |
|
"max_steps": 102000, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 3, |
|
"save_steps": 500, |
|
"total_flos": 3.09998518050816e+18, |
|
"train_batch_size": 8, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|