| { | |
| "best_metric": 0.8245412844036697, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-29/checkpoint-527", | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 1581, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 12.870553970336914, | |
| "learning_rate": 0.00018060344891044337, | |
| "loss": 1.0875, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8245412844036697, | |
| "eval_loss": 1.0481752157211304, | |
| "eval_runtime": 2.8193, | |
| "eval_samples_per_second": 309.298, | |
| "eval_steps_per_second": 2.483, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 6.4301323890686035, | |
| "learning_rate": 0.00015802801779663797, | |
| "loss": 0.5298, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8211009174311926, | |
| "eval_loss": 1.1450493335723877, | |
| "eval_runtime": 2.8147, | |
| "eval_samples_per_second": 309.8, | |
| "eval_steps_per_second": 2.487, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.0001354954243130295, | |
| "loss": 0.3909, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.819954128440367, | |
| "eval_loss": 1.176020622253418, | |
| "eval_runtime": 2.8147, | |
| "eval_samples_per_second": 309.805, | |
| "eval_steps_per_second": 2.487, | |
| "step": 1581 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4743, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9, | |
| "save_steps": 500, | |
| "total_flos": 24228458327580.0, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.6300420149444647, | |
| "learning_rate": 0.0002031788800242488, | |
| "num_train_epochs": 9, | |
| "temperature": 25 | |
| } | |
| } | |