|
{ |
|
"best_metric": 0.3920317590236664, |
|
"best_model_checkpoint": "distilbert_sa_GLUE_Experiment_logit_kd_qnli_256/checkpoint-820", |
|
"epoch": 7.0, |
|
"global_step": 2870, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 1.0, |
|
"learning_rate": 4.9e-05, |
|
"loss": 0.4063, |
|
"step": 410 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"eval_accuracy": 0.5751418634449936, |
|
"eval_loss": 0.39294081926345825, |
|
"eval_runtime": 7.3441, |
|
"eval_samples_per_second": 743.864, |
|
"eval_steps_per_second": 2.996, |
|
"step": 410 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"learning_rate": 4.8e-05, |
|
"loss": 0.3924, |
|
"step": 820 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_accuracy": 0.5872231374702545, |
|
"eval_loss": 0.3920317590236664, |
|
"eval_runtime": 7.3386, |
|
"eval_samples_per_second": 744.415, |
|
"eval_steps_per_second": 2.998, |
|
"step": 820 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"learning_rate": 4.7e-05, |
|
"loss": 0.3823, |
|
"step": 1230 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"eval_accuracy": 0.5912502288120081, |
|
"eval_loss": 0.39368072152137756, |
|
"eval_runtime": 7.3567, |
|
"eval_samples_per_second": 742.587, |
|
"eval_steps_per_second": 2.99, |
|
"step": 1230 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"learning_rate": 4.600000000000001e-05, |
|
"loss": 0.3739, |
|
"step": 1640 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_accuracy": 0.586490939044481, |
|
"eval_loss": 0.3972700238227844, |
|
"eval_runtime": 7.3703, |
|
"eval_samples_per_second": 741.216, |
|
"eval_steps_per_second": 2.985, |
|
"step": 1640 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"learning_rate": 4.5e-05, |
|
"loss": 0.3668, |
|
"step": 2050 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"eval_accuracy": 0.5778876075416438, |
|
"eval_loss": 0.3996599316596985, |
|
"eval_runtime": 7.3492, |
|
"eval_samples_per_second": 743.348, |
|
"eval_steps_per_second": 2.994, |
|
"step": 2050 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"learning_rate": 4.4000000000000006e-05, |
|
"loss": 0.3591, |
|
"step": 2460 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"eval_accuracy": 0.5718469705290133, |
|
"eval_loss": 0.41333967447280884, |
|
"eval_runtime": 7.4841, |
|
"eval_samples_per_second": 729.951, |
|
"eval_steps_per_second": 2.94, |
|
"step": 2460 |
|
}, |
|
{ |
|
"epoch": 7.0, |
|
"learning_rate": 4.3e-05, |
|
"loss": 0.3518, |
|
"step": 2870 |
|
}, |
|
{ |
|
"epoch": 7.0, |
|
"eval_accuracy": 0.5648910854841662, |
|
"eval_loss": 0.4250338077545166, |
|
"eval_runtime": 7.3446, |
|
"eval_samples_per_second": 743.816, |
|
"eval_steps_per_second": 2.995, |
|
"step": 2870 |
|
}, |
|
{ |
|
"epoch": 7.0, |
|
"step": 2870, |
|
"total_flos": 1.2511057062920192e+16, |
|
"train_loss": 0.37608287425822085, |
|
"train_runtime": 1391.6849, |
|
"train_samples_per_second": 3763.172, |
|
"train_steps_per_second": 14.73 |
|
} |
|
], |
|
"max_steps": 20500, |
|
"num_train_epochs": 50, |
|
"total_flos": 1.2511057062920192e+16, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|