|
{ |
|
"best_metric": 0.9894, |
|
"best_model_checkpoint": "AML_A2_Q4/checkpoint-6750", |
|
"epoch": 3.0, |
|
"global_step": 6750, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.22, |
|
"learning_rate": 1.851851851851852e-05, |
|
"loss": 0.2407, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.44, |
|
"learning_rate": 1.7037037037037038e-05, |
|
"loss": 0.1887, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 0.67, |
|
"learning_rate": 1.555555555555556e-05, |
|
"loss": 0.1843, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 0.89, |
|
"learning_rate": 1.4074074074074075e-05, |
|
"loss": 0.1642, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"eval_accuracy": 0.9862, |
|
"eval_loss": 0.057202473282814026, |
|
"eval_runtime": 71.1804, |
|
"eval_samples_per_second": 70.244, |
|
"eval_steps_per_second": 17.561, |
|
"step": 2250 |
|
}, |
|
{ |
|
"epoch": 1.11, |
|
"learning_rate": 1.2592592592592593e-05, |
|
"loss": 0.1586, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 1.33, |
|
"learning_rate": 1.1111111111111113e-05, |
|
"loss": 0.1594, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 1.56, |
|
"learning_rate": 9.62962962962963e-06, |
|
"loss": 0.1632, |
|
"step": 3500 |
|
}, |
|
{ |
|
"epoch": 1.78, |
|
"learning_rate": 8.148148148148148e-06, |
|
"loss": 0.1734, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"learning_rate": 6.666666666666667e-06, |
|
"loss": 0.1503, |
|
"step": 4500 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_accuracy": 0.9854, |
|
"eval_loss": 0.05908837914466858, |
|
"eval_runtime": 70.902, |
|
"eval_samples_per_second": 70.52, |
|
"eval_steps_per_second": 17.63, |
|
"step": 4500 |
|
}, |
|
{ |
|
"epoch": 2.22, |
|
"learning_rate": 5.185185185185185e-06, |
|
"loss": 0.1564, |
|
"step": 5000 |
|
}, |
|
{ |
|
"epoch": 2.44, |
|
"learning_rate": 3.7037037037037037e-06, |
|
"loss": 0.1595, |
|
"step": 5500 |
|
}, |
|
{ |
|
"epoch": 2.67, |
|
"learning_rate": 2.222222222222222e-06, |
|
"loss": 0.1665, |
|
"step": 6000 |
|
}, |
|
{ |
|
"epoch": 2.89, |
|
"learning_rate": 7.407407407407407e-07, |
|
"loss": 0.1818, |
|
"step": 6500 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"eval_accuracy": 0.9894, |
|
"eval_loss": 0.04315419867634773, |
|
"eval_runtime": 70.3971, |
|
"eval_samples_per_second": 71.026, |
|
"eval_steps_per_second": 17.756, |
|
"step": 6750 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"step": 6750, |
|
"total_flos": 1.046216869705728e+19, |
|
"train_loss": 0.17347003343370226, |
|
"train_runtime": 5139.576, |
|
"train_samples_per_second": 26.267, |
|
"train_steps_per_second": 1.313 |
|
} |
|
], |
|
"max_steps": 6750, |
|
"num_train_epochs": 3, |
|
"total_flos": 1.046216869705728e+19, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|