| { | |
| "best_metric": 0.5141951979542654, | |
| "best_model_checkpoint": "distilbert-base-uncased-finetuned-cola\\run-2\\checkpoint-1605", | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 2140, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 7.330426643717586e-06, | |
| "loss": 0.5309, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.4970090985298157, | |
| "eval_matthews_correlation": 0.43624361512112764, | |
| "eval_runtime": 6.9324, | |
| "eval_samples_per_second": 150.453, | |
| "eval_steps_per_second": 9.521, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 5.095540471852468e-06, | |
| "loss": 0.4001, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.5170313715934753, | |
| "eval_matthews_correlation": 0.4414558823469975, | |
| "eval_runtime": 6.257, | |
| "eval_samples_per_second": 166.694, | |
| "eval_steps_per_second": 10.548, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 2.8606542999873506e-06, | |
| "loss": 0.3288, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.512870192527771, | |
| "eval_matthews_correlation": 0.5141951979542654, | |
| "eval_runtime": 6.366, | |
| "eval_samples_per_second": 163.839, | |
| "eval_steps_per_second": 10.368, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "learning_rate": 6.257681281222329e-07, | |
| "loss": 0.2647, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.5683671236038208, | |
| "eval_matthews_correlation": 0.4982331824617808, | |
| "eval_runtime": 6.6371, | |
| "eval_samples_per_second": 157.146, | |
| "eval_steps_per_second": 9.944, | |
| "step": 2140 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2140, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "total_flos": 171385647695124.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 9.565312815582705e-06, | |
| "num_train_epochs": 4, | |
| "per_device_train_batch_size": 16, | |
| "seed": 23 | |
| } | |
| } | |