| { | |
| "best_metric": 0.8256880733944955, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-11/checkpoint-527", | |
| "epoch": 8.0, | |
| "eval_steps": 500, | |
| "global_step": 4216, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 6.5331573486328125, | |
| "learning_rate": 0.00014562243502675204, | |
| "loss": 0.547, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8256880733944955, | |
| "eval_loss": 0.5717546343803406, | |
| "eval_runtime": 2.8071, | |
| "eval_samples_per_second": 310.639, | |
| "eval_steps_per_second": 2.494, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 4.755884647369385, | |
| "learning_rate": 0.0001248192300229303, | |
| "loss": 0.3165, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.823394495412844, | |
| "eval_loss": 0.6189394593238831, | |
| "eval_runtime": 2.8231, | |
| "eval_samples_per_second": 308.882, | |
| "eval_steps_per_second": 2.48, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 13.318252563476562, | |
| "learning_rate": 0.0001040160250191086, | |
| "loss": 0.2438, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8256880733944955, | |
| "eval_loss": 0.6849867701530457, | |
| "eval_runtime": 2.8211, | |
| "eval_samples_per_second": 309.105, | |
| "eval_steps_per_second": 2.481, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 6.025650501251221, | |
| "learning_rate": 8.321282001528688e-05, | |
| "loss": 0.2015, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8256880733944955, | |
| "eval_loss": 0.7157025337219238, | |
| "eval_runtime": 2.8159, | |
| "eval_samples_per_second": 309.671, | |
| "eval_steps_per_second": 2.486, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 6.3858113288879395, | |
| "learning_rate": 6.240961501146515e-05, | |
| "loss": 0.1725, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8153669724770642, | |
| "eval_loss": 0.7607192397117615, | |
| "eval_runtime": 2.8278, | |
| "eval_samples_per_second": 308.365, | |
| "eval_steps_per_second": 2.475, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 12.167182922363281, | |
| "learning_rate": 4.160641000764344e-05, | |
| "loss": 0.1517, | |
| "step": 3162 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8119266055045872, | |
| "eval_loss": 0.8524329662322998, | |
| "eval_runtime": 2.8071, | |
| "eval_samples_per_second": 310.643, | |
| "eval_steps_per_second": 2.494, | |
| "step": 3162 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 6.675292015075684, | |
| "learning_rate": 2.080320500382172e-05, | |
| "loss": 0.1413, | |
| "step": 3689 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8107798165137615, | |
| "eval_loss": 0.862307608127594, | |
| "eval_runtime": 2.8162, | |
| "eval_samples_per_second": 309.641, | |
| "eval_steps_per_second": 2.486, | |
| "step": 3689 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 2.5445616245269775, | |
| "learning_rate": 0.0, | |
| "loss": 0.1299, | |
| "step": 4216 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8107798165137615, | |
| "eval_loss": 0.8725943565368652, | |
| "eval_runtime": 2.8212, | |
| "eval_samples_per_second": 309.086, | |
| "eval_steps_per_second": 2.481, | |
| "step": 4216 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4216, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 8, | |
| "save_steps": 500, | |
| "total_flos": 64687167744780.0, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.9318516048675823, | |
| "learning_rate": 0.00016642564003057375, | |
| "num_train_epochs": 8, | |
| "temperature": 23 | |
| } | |
| } | |