| { | |
| "best_metric": 0.4260311424732208, | |
| "best_model_checkpoint": "hBERTv1_new_pretrain_w_init_48_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": 0.5336, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8176605504587156, | |
| "eval_loss": 0.4260311424732208, | |
| "eval_runtime": 1.6096, | |
| "eval_samples_per_second": 541.764, | |
| "eval_steps_per_second": 4.349, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 3.8400000000000005e-05, | |
| "loss": 0.2532, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8211009174311926, | |
| "eval_loss": 0.5553512573242188, | |
| "eval_runtime": 1.6268, | |
| "eval_samples_per_second": 536.036, | |
| "eval_steps_per_second": 4.303, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 3.76e-05, | |
| "loss": 0.1845, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8107798165137615, | |
| "eval_loss": 0.5317797660827637, | |
| "eval_runtime": 1.6118, | |
| "eval_samples_per_second": 541.01, | |
| "eval_steps_per_second": 4.343, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 3.680000000000001e-05, | |
| "loss": 0.1484, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8107798165137615, | |
| "eval_loss": 0.6237322688102722, | |
| "eval_runtime": 1.6067, | |
| "eval_samples_per_second": 542.726, | |
| "eval_steps_per_second": 4.357, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 3.6e-05, | |
| "loss": 0.1195, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8096330275229358, | |
| "eval_loss": 0.530974805355072, | |
| "eval_runtime": 1.6016, | |
| "eval_samples_per_second": 544.465, | |
| "eval_steps_per_second": 4.371, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 3.52e-05, | |
| "loss": 0.1, | |
| "step": 3162 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8027522935779816, | |
| "eval_loss": 0.7588031888008118, | |
| "eval_runtime": 1.6195, | |
| "eval_samples_per_second": 538.438, | |
| "eval_steps_per_second": 4.322, | |
| "step": 3162 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "step": 3162, | |
| "total_flos": 5.984501559302554e+16, | |
| "train_loss": 0.22320289298147433, | |
| "train_runtime": 2235.7769, | |
| "train_samples_per_second": 1506.165, | |
| "train_steps_per_second": 11.786 | |
| } | |
| ], | |
| "max_steps": 26350, | |
| "num_train_epochs": 50, | |
| "total_flos": 5.984501559302554e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |