| { | |
| "best_metric": 0.4975221305697031, | |
| "best_model_checkpoint": "distilbert-base-uncased-finetuned-cola/run-0/checkpoint-4276", | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 4276, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 9.967147053469058e-06, | |
| "loss": 0.5657, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 8.647344742363515e-06, | |
| "loss": 0.5621, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 7.327542431257973e-06, | |
| "loss": 0.5429, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 6.0077401201524294e-06, | |
| "loss": 0.5214, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.5339541435241699, | |
| "eval_matthews_correlation": 0.40454640840116707, | |
| "eval_runtime": 0.7883, | |
| "eval_samples_per_second": 1323.162, | |
| "eval_steps_per_second": 83.728, | |
| "step": 2138 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 4.687937809046888e-06, | |
| "loss": 0.4709, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 3.368135497941345e-06, | |
| "loss": 0.4733, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "learning_rate": 2.048333186835802e-06, | |
| "loss": 0.4707, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 7.285308757302596e-07, | |
| "loss": 0.5123, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.7701239585876465, | |
| "eval_matthews_correlation": 0.4975221305697031, | |
| "eval_runtime": 0.7314, | |
| "eval_samples_per_second": 1426.085, | |
| "eval_steps_per_second": 90.241, | |
| "step": 4276 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4276, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "total_flos": 65123767413480.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 1.1286949364574601e-05, | |
| "num_train_epochs": 2, | |
| "per_device_train_batch_size": 4, | |
| "seed": 27 | |
| } | |
| } | |