| { | |
| "best_metric": 0.8241134285926819, | |
| "best_model_checkpoint": "mobilebert_sa_GLUE_Experiment_logit_kd_sst2/checkpoint-1054", | |
| "epoch": 7.0, | |
| "global_step": 3689, | |
| "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": 1.0702, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7740825688073395, | |
| "eval_loss": 0.8403842449188232, | |
| "eval_runtime": 2.1461, | |
| "eval_samples_per_second": 406.316, | |
| "eval_steps_per_second": 3.262, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.5148, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7821100917431193, | |
| "eval_loss": 0.8241134285926819, | |
| "eval_runtime": 2.0443, | |
| "eval_samples_per_second": 426.558, | |
| "eval_steps_per_second": 3.424, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 4.7e-05, | |
| "loss": 0.3924, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7855504587155964, | |
| "eval_loss": 1.0814857482910156, | |
| "eval_runtime": 2.0476, | |
| "eval_samples_per_second": 425.864, | |
| "eval_steps_per_second": 3.419, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.3413, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7809633027522935, | |
| "eval_loss": 0.9158650040626526, | |
| "eval_runtime": 2.0515, | |
| "eval_samples_per_second": 425.049, | |
| "eval_steps_per_second": 3.412, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.3068, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7947247706422018, | |
| "eval_loss": 0.883631706237793, | |
| "eval_runtime": 2.0338, | |
| "eval_samples_per_second": 428.759, | |
| "eval_steps_per_second": 3.442, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.2837, | |
| "step": 3162 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.786697247706422, | |
| "eval_loss": 0.9832125306129456, | |
| "eval_runtime": 2.0645, | |
| "eval_samples_per_second": 422.377, | |
| "eval_steps_per_second": 3.391, | |
| "step": 3162 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 4.3e-05, | |
| "loss": 0.2648, | |
| "step": 3689 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.7981651376146789, | |
| "eval_loss": 0.963298499584198, | |
| "eval_runtime": 2.0555, | |
| "eval_samples_per_second": 424.233, | |
| "eval_steps_per_second": 3.406, | |
| "step": 3689 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "step": 3689, | |
| "total_flos": 1.4781746813861888e+16, | |
| "train_loss": 0.45344161263453214, | |
| "train_runtime": 2923.0123, | |
| "train_samples_per_second": 1152.048, | |
| "train_steps_per_second": 9.015 | |
| } | |
| ], | |
| "max_steps": 26350, | |
| "num_train_epochs": 50, | |
| "total_flos": 1.4781746813861888e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |