| { | |
| "best_metric": 0.13462364673614502, | |
| "best_model_checkpoint": "fast_food_image_detection/checkpoint-2500", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 2500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.28725844621658325, | |
| "learning_rate": 1.6326530612244897e-06, | |
| "loss": 0.225, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9757074881041823, | |
| "eval_loss": 0.14208008348941803, | |
| "eval_runtime": 39.2448, | |
| "eval_samples_per_second": 101.746, | |
| "eval_steps_per_second": 12.741, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.3262328803539276, | |
| "learning_rate": 1.2244897959183673e-06, | |
| "loss": 0.2101, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9769596794390183, | |
| "eval_loss": 0.13922728598117828, | |
| "eval_runtime": 38.0806, | |
| "eval_samples_per_second": 104.856, | |
| "eval_steps_per_second": 13.13, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.27534928917884827, | |
| "learning_rate": 8.163265306122449e-07, | |
| "loss": 0.1932, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9774605559729527, | |
| "eval_loss": 0.1367102861404419, | |
| "eval_runtime": 37.9368, | |
| "eval_samples_per_second": 105.254, | |
| "eval_steps_per_second": 13.18, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 3.9497718811035156, | |
| "learning_rate": 4.0816326530612243e-07, | |
| "loss": 0.1909, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9769596794390183, | |
| "eval_loss": 0.13517306745052338, | |
| "eval_runtime": 37.8037, | |
| "eval_samples_per_second": 105.625, | |
| "eval_steps_per_second": 13.226, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.22323879599571228, | |
| "learning_rate": 0.0, | |
| "loss": 0.1859, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9774605559729527, | |
| "eval_loss": 0.13462364673614502, | |
| "eval_runtime": 38.526, | |
| "eval_samples_per_second": 103.644, | |
| "eval_steps_per_second": 12.978, | |
| "step": 2500 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 6.188179040444621e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |