| { | |
| "best_metric": 0.3393769209893431, | |
| "best_model_checkpoint": "distilbert-base-uncased-finetuned-cola/run-7/checkpoint-642", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 1070, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.5881524085998535, | |
| "eval_matthews_correlation": 0.0, | |
| "eval_runtime": 0.5145, | |
| "eval_samples_per_second": 2027.266, | |
| "eval_steps_per_second": 128.283, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.6095048785209656, | |
| "eval_matthews_correlation": 0.2920277228925403, | |
| "eval_runtime": 0.517, | |
| "eval_samples_per_second": 2017.266, | |
| "eval_steps_per_second": 127.651, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "grad_norm": 7.986351013183594, | |
| "learning_rate": 6.259012645042397e-06, | |
| "loss": 0.5246, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.8854954838752747, | |
| "eval_matthews_correlation": 0.3393769209893431, | |
| "eval_runtime": 0.5407, | |
| "eval_samples_per_second": 1928.84, | |
| "eval_steps_per_second": 122.055, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 1.2050410509109497, | |
| "eval_matthews_correlation": 0.2903129121495111, | |
| "eval_runtime": 0.518, | |
| "eval_samples_per_second": 2013.528, | |
| "eval_steps_per_second": 127.414, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "grad_norm": 0.35196247696876526, | |
| "learning_rate": 7.68650675706961e-07, | |
| "loss": 0.1995, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 1.173661708831787, | |
| "eval_matthews_correlation": 0.3306791468948119, | |
| "eval_runtime": 0.5275, | |
| "eval_samples_per_second": 1977.377, | |
| "eval_steps_per_second": 125.126, | |
| "step": 1070 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 1070, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 26085728519712.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 1.1749374614377834e-05, | |
| "num_train_epochs": 5, | |
| "per_device_train_batch_size": 4, | |
| "seed": 24 | |
| } | |
| } | |