| { | |
| "best_metric": 0.8325688073394495, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-2/checkpoint-2108", | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 2108, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 0.00013462097079767568, | |
| "loss": 1.19, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.819954128440367, | |
| "eval_loss": 1.0689420700073242, | |
| "eval_runtime": 10.3665, | |
| "eval_samples_per_second": 84.117, | |
| "eval_steps_per_second": 0.675, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 0.00011218414233139642, | |
| "loss": 0.5832, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8279816513761468, | |
| "eval_loss": 1.1704012155532837, | |
| "eval_runtime": 6.2773, | |
| "eval_samples_per_second": 138.912, | |
| "eval_steps_per_second": 1.115, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 8.974731386511713e-05, | |
| "loss": 0.4391, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.823394495412844, | |
| "eval_loss": 1.1875587701797485, | |
| "eval_runtime": 6.2138, | |
| "eval_samples_per_second": 140.333, | |
| "eval_steps_per_second": 1.127, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 6.731048539883784e-05, | |
| "loss": 0.3642, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8325688073394495, | |
| "eval_loss": 1.1609015464782715, | |
| "eval_runtime": 8.1221, | |
| "eval_samples_per_second": 107.361, | |
| "eval_steps_per_second": 0.862, | |
| "step": 2108 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 3689, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 7, | |
| "save_steps": 500, | |
| "total_flos": 32344049138640.0, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.6204520709084631, | |
| "learning_rate": 0.00015705779926395499, | |
| "num_train_epochs": 7, | |
| "temperature": 16 | |
| } | |
| } | |