| { | |
| "best_metric": 0.4098166227340698, | |
| "best_model_checkpoint": "hBERTv1_new_pretrain_48_KD_sst2/checkpoint-527", | |
| "epoch": 6.0, | |
| "global_step": 3162, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 3.9200000000000004e-05, | |
| "loss": 0.3947, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8165137614678899, | |
| "eval_loss": 0.4098166227340698, | |
| "eval_runtime": 1.8008, | |
| "eval_samples_per_second": 484.227, | |
| "eval_steps_per_second": 3.887, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 3.8400000000000005e-05, | |
| "loss": 0.2426, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8256880733944955, | |
| "eval_loss": 0.47963330149650574, | |
| "eval_runtime": 1.8005, | |
| "eval_samples_per_second": 484.31, | |
| "eval_steps_per_second": 3.888, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 3.76e-05, | |
| "loss": 0.1948, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8188073394495413, | |
| "eval_loss": 0.4834706783294678, | |
| "eval_runtime": 1.8087, | |
| "eval_samples_per_second": 482.113, | |
| "eval_steps_per_second": 3.87, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 3.680000000000001e-05, | |
| "loss": 0.1702, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8027522935779816, | |
| "eval_loss": 0.5115558505058289, | |
| "eval_runtime": 1.8015, | |
| "eval_samples_per_second": 484.037, | |
| "eval_steps_per_second": 3.886, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 3.6e-05, | |
| "loss": 0.1484, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8084862385321101, | |
| "eval_loss": 0.5547378063201904, | |
| "eval_runtime": 1.8032, | |
| "eval_samples_per_second": 483.586, | |
| "eval_steps_per_second": 3.882, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 3.52e-05, | |
| "loss": 0.1355, | |
| "step": 3162 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7993119266055045, | |
| "eval_loss": 0.6597509980201721, | |
| "eval_runtime": 1.8009, | |
| "eval_samples_per_second": 484.198, | |
| "eval_steps_per_second": 3.887, | |
| "step": 3162 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "step": 3162, | |
| "total_flos": 5.984501559302554e+16, | |
| "train_loss": 0.21436870558061305, | |
| "train_runtime": 2799.5615, | |
| "train_samples_per_second": 1202.849, | |
| "train_steps_per_second": 9.412 | |
| } | |
| ], | |
| "max_steps": 26350, | |
| "num_train_epochs": 50, | |
| "total_flos": 5.984501559302554e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |