|
{ |
|
"best_metric": 1.374252200126648, |
|
"best_model_checkpoint": "outputs/checkpoint-104", |
|
"epoch": 2.982078853046595, |
|
"eval_steps": 500, |
|
"global_step": 104, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.5734767025089605, |
|
"grad_norm": 1.346787452697754, |
|
"learning_rate": 2e-05, |
|
"loss": 2.8792, |
|
"step": 20 |
|
}, |
|
{ |
|
"epoch": 0.974910394265233, |
|
"eval_loss": 2.1643929481506348, |
|
"eval_runtime": 34.9442, |
|
"eval_samples_per_second": 31.937, |
|
"eval_steps_per_second": 4.006, |
|
"step": 34 |
|
}, |
|
{ |
|
"epoch": 1.146953405017921, |
|
"grad_norm": 0.6896965503692627, |
|
"learning_rate": 4e-05, |
|
"loss": 2.3839, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 1.7204301075268817, |
|
"grad_norm": 0.6072823405265808, |
|
"learning_rate": 6e-05, |
|
"loss": 1.8972, |
|
"step": 60 |
|
}, |
|
{ |
|
"epoch": 1.978494623655914, |
|
"eval_loss": 1.7010903358459473, |
|
"eval_runtime": 34.9827, |
|
"eval_samples_per_second": 31.902, |
|
"eval_steps_per_second": 4.002, |
|
"step": 69 |
|
}, |
|
{ |
|
"epoch": 2.293906810035842, |
|
"grad_norm": 0.6996594071388245, |
|
"learning_rate": 8e-05, |
|
"loss": 1.7112, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 2.867383512544803, |
|
"grad_norm": 1.0495513677597046, |
|
"learning_rate": 0.0001, |
|
"loss": 1.5167, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 2.982078853046595, |
|
"eval_loss": 1.374252200126648, |
|
"eval_runtime": 34.9766, |
|
"eval_samples_per_second": 31.907, |
|
"eval_steps_per_second": 4.003, |
|
"step": 104 |
|
} |
|
], |
|
"logging_steps": 20, |
|
"max_steps": 170, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 5, |
|
"save_steps": 500, |
|
"total_flos": 5518564418224128.0, |
|
"train_batch_size": 2, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|