| { | |
| "best_metric": 0.4606306850910187, | |
| "best_model_checkpoint": "hBERTv1_new_pretrain_w_init__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": 0.363, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8038990825688074, | |
| "eval_loss": 0.4606306850910187, | |
| "eval_runtime": 1.5758, | |
| "eval_samples_per_second": 553.365, | |
| "eval_steps_per_second": 4.442, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 3.8400000000000005e-05, | |
| "loss": 0.2256, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8119266055045872, | |
| "eval_loss": 0.6466029286384583, | |
| "eval_runtime": 1.581, | |
| "eval_samples_per_second": 551.543, | |
| "eval_steps_per_second": 4.428, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 3.76e-05, | |
| "loss": 0.1754, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8176605504587156, | |
| "eval_loss": 0.5100513696670532, | |
| "eval_runtime": 1.5786, | |
| "eval_samples_per_second": 552.387, | |
| "eval_steps_per_second": 4.434, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 3.680000000000001e-05, | |
| "loss": 0.1394, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8176605504587156, | |
| "eval_loss": 0.49214717745780945, | |
| "eval_runtime": 1.5822, | |
| "eval_samples_per_second": 551.116, | |
| "eval_steps_per_second": 4.424, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 3.6e-05, | |
| "loss": 0.1111, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.819954128440367, | |
| "eval_loss": 0.5109940767288208, | |
| "eval_runtime": 1.5752, | |
| "eval_samples_per_second": 553.567, | |
| "eval_steps_per_second": 4.444, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 3.52e-05, | |
| "loss": 0.0937, | |
| "step": 3162 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8211009174311926, | |
| "eval_loss": 0.6468406319618225, | |
| "eval_runtime": 1.5781, | |
| "eval_samples_per_second": 552.564, | |
| "eval_steps_per_second": 4.436, | |
| "step": 3162 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "step": 3162, | |
| "total_flos": 5.984501559302554e+16, | |
| "train_loss": 0.18471441202567845, | |
| "train_runtime": 2261.4914, | |
| "train_samples_per_second": 1489.039, | |
| "train_steps_per_second": 11.652 | |
| } | |
| ], | |
| "max_steps": 26350, | |
| "num_train_epochs": 50, | |
| "total_flos": 5.984501559302554e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |