|
{ |
|
"best_metric": 0.21577584743499756, |
|
"best_model_checkpoint": "mobilebert_sa_GLUE_Experiment_logit_kd_pretrain_qnli/checkpoint-8190", |
|
"epoch": 15.0, |
|
"global_step": 12285, |
|
"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": 1.1418, |
|
"step": 819 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"eval_accuracy": 0.5053999633900788, |
|
"eval_loss": 1.0623146295547485, |
|
"eval_runtime": 10.2085, |
|
"eval_samples_per_second": 535.141, |
|
"eval_steps_per_second": 4.212, |
|
"step": 819 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"learning_rate": 4.8e-05, |
|
"loss": 1.1397, |
|
"step": 1638 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_accuracy": 0.5053999633900788, |
|
"eval_loss": 1.0616631507873535, |
|
"eval_runtime": 10.1836, |
|
"eval_samples_per_second": 536.452, |
|
"eval_steps_per_second": 4.222, |
|
"step": 1638 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"learning_rate": 4.7e-05, |
|
"loss": 1.1439, |
|
"step": 2457 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"eval_accuracy": 0.5053999633900788, |
|
"eval_loss": 1.063395380973816, |
|
"eval_runtime": 10.2401, |
|
"eval_samples_per_second": 533.492, |
|
"eval_steps_per_second": 4.199, |
|
"step": 2457 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"learning_rate": 4.600000000000001e-05, |
|
"loss": 1.1397, |
|
"step": 3276 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_accuracy": 0.5053999633900788, |
|
"eval_loss": 1.0635251998901367, |
|
"eval_runtime": 10.2624, |
|
"eval_samples_per_second": 532.329, |
|
"eval_steps_per_second": 4.19, |
|
"step": 3276 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"learning_rate": 4.5e-05, |
|
"loss": 1.14, |
|
"step": 4095 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"eval_accuracy": 0.5053999633900788, |
|
"eval_loss": 1.0643025636672974, |
|
"eval_runtime": 10.1907, |
|
"eval_samples_per_second": 536.078, |
|
"eval_steps_per_second": 4.22, |
|
"step": 4095 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"learning_rate": 4.4000000000000006e-05, |
|
"loss": 1.1399, |
|
"step": 4914 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"eval_accuracy": 0.5053999633900788, |
|
"eval_loss": 1.0611162185668945, |
|
"eval_runtime": 10.1973, |
|
"eval_samples_per_second": 535.731, |
|
"eval_steps_per_second": 4.217, |
|
"step": 4914 |
|
}, |
|
{ |
|
"epoch": 7.0, |
|
"learning_rate": 4.3e-05, |
|
"loss": 1.14, |
|
"step": 5733 |
|
}, |
|
{ |
|
"epoch": 7.0, |
|
"eval_accuracy": 0.5053999633900788, |
|
"eval_loss": 1.06246018409729, |
|
"eval_runtime": 10.1876, |
|
"eval_samples_per_second": 536.238, |
|
"eval_steps_per_second": 4.221, |
|
"step": 5733 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"learning_rate": 4.2e-05, |
|
"loss": 1.0013, |
|
"step": 6552 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"eval_accuracy": 0.8420281896393923, |
|
"eval_loss": 0.380109041929245, |
|
"eval_runtime": 10.1848, |
|
"eval_samples_per_second": 536.385, |
|
"eval_steps_per_second": 4.222, |
|
"step": 6552 |
|
}, |
|
{ |
|
"epoch": 9.0, |
|
"learning_rate": 4.1e-05, |
|
"loss": 0.3353, |
|
"step": 7371 |
|
}, |
|
{ |
|
"epoch": 9.0, |
|
"eval_accuracy": 0.9029837085850265, |
|
"eval_loss": 0.21630515158176422, |
|
"eval_runtime": 10.2676, |
|
"eval_samples_per_second": 532.064, |
|
"eval_steps_per_second": 4.188, |
|
"step": 7371 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"learning_rate": 4e-05, |
|
"loss": 0.2165, |
|
"step": 8190 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"eval_accuracy": 0.8984074684239429, |
|
"eval_loss": 0.21577584743499756, |
|
"eval_runtime": 10.195, |
|
"eval_samples_per_second": 535.853, |
|
"eval_steps_per_second": 4.218, |
|
"step": 8190 |
|
}, |
|
{ |
|
"epoch": 11.0, |
|
"learning_rate": 3.9000000000000006e-05, |
|
"loss": 0.1593, |
|
"step": 9009 |
|
}, |
|
{ |
|
"epoch": 11.0, |
|
"eval_accuracy": 0.9057294526816767, |
|
"eval_loss": 0.22050638496875763, |
|
"eval_runtime": 10.1795, |
|
"eval_samples_per_second": 536.669, |
|
"eval_steps_per_second": 4.224, |
|
"step": 9009 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"learning_rate": 3.8e-05, |
|
"loss": 0.126, |
|
"step": 9828 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"eval_accuracy": 0.9077429983525536, |
|
"eval_loss": 0.22907224297523499, |
|
"eval_runtime": 10.1911, |
|
"eval_samples_per_second": 536.053, |
|
"eval_steps_per_second": 4.219, |
|
"step": 9828 |
|
}, |
|
{ |
|
"epoch": 13.0, |
|
"learning_rate": 3.7e-05, |
|
"loss": 0.1049, |
|
"step": 10647 |
|
}, |
|
{ |
|
"epoch": 13.0, |
|
"eval_accuracy": 0.9071938495332235, |
|
"eval_loss": 0.23231425881385803, |
|
"eval_runtime": 10.2915, |
|
"eval_samples_per_second": 530.824, |
|
"eval_steps_per_second": 4.178, |
|
"step": 10647 |
|
}, |
|
{ |
|
"epoch": 14.0, |
|
"learning_rate": 3.6e-05, |
|
"loss": 0.0903, |
|
"step": 11466 |
|
}, |
|
{ |
|
"epoch": 14.0, |
|
"eval_accuracy": 0.8984074684239429, |
|
"eval_loss": 0.2675889730453491, |
|
"eval_runtime": 10.2001, |
|
"eval_samples_per_second": 535.585, |
|
"eval_steps_per_second": 4.216, |
|
"step": 11466 |
|
}, |
|
{ |
|
"epoch": 15.0, |
|
"learning_rate": 3.5e-05, |
|
"loss": 0.0819, |
|
"step": 12285 |
|
}, |
|
{ |
|
"epoch": 15.0, |
|
"eval_accuracy": 0.900604063701263, |
|
"eval_loss": 0.23767022788524628, |
|
"eval_runtime": 10.1798, |
|
"eval_samples_per_second": 536.652, |
|
"eval_steps_per_second": 4.224, |
|
"step": 12285 |
|
}, |
|
{ |
|
"epoch": 15.0, |
|
"step": 12285, |
|
"total_flos": 4.989595309572096e+16, |
|
"train_loss": 0.6733751687471781, |
|
"train_runtime": 7169.216, |
|
"train_samples_per_second": 730.505, |
|
"train_steps_per_second": 5.712 |
|
} |
|
], |
|
"max_steps": 40950, |
|
"num_train_epochs": 50, |
|
"total_flos": 4.989595309572096e+16, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|