| { |
| "best_metric": 0.9784, |
| "best_model_checkpoint": "fine-tune-DeiT-s-cifar10\\checkpoint-3520", |
| "epoch": 5.0, |
| "global_step": 3520, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.71, |
| "learning_rate": 0.00017159090909090908, |
| "loss": 0.689, |
| "step": 500 |
| }, |
| { |
| "epoch": 1.0, |
| "eval_accuracy": 0.9356, |
| "eval_loss": 0.1929679960012436, |
| "eval_runtime": 23.4437, |
| "eval_samples_per_second": 213.277, |
| "eval_steps_per_second": 3.37, |
| "step": 704 |
| }, |
| { |
| "epoch": 1.42, |
| "learning_rate": 0.0001431818181818182, |
| "loss": 0.4759, |
| "step": 1000 |
| }, |
| { |
| "epoch": 2.0, |
| "eval_accuracy": 0.954, |
| "eval_loss": 0.1267678141593933, |
| "eval_runtime": 23.0301, |
| "eval_samples_per_second": 217.107, |
| "eval_steps_per_second": 3.43, |
| "step": 1408 |
| }, |
| { |
| "epoch": 2.13, |
| "learning_rate": 0.00011477272727272728, |
| "loss": 0.404, |
| "step": 1500 |
| }, |
| { |
| "epoch": 2.84, |
| "learning_rate": 8.636363636363637e-05, |
| "loss": 0.3335, |
| "step": 2000 |
| }, |
| { |
| "epoch": 3.0, |
| "eval_accuracy": 0.9632, |
| "eval_loss": 0.10697369277477264, |
| "eval_runtime": 22.2257, |
| "eval_samples_per_second": 224.965, |
| "eval_steps_per_second": 3.554, |
| "step": 2112 |
| }, |
| { |
| "epoch": 3.55, |
| "learning_rate": 5.7954545454545464e-05, |
| "loss": 0.2955, |
| "step": 2500 |
| }, |
| { |
| "epoch": 4.0, |
| "eval_accuracy": 0.9696, |
| "eval_loss": 0.07925179600715637, |
| "eval_runtime": 21.8511, |
| "eval_samples_per_second": 228.822, |
| "eval_steps_per_second": 3.615, |
| "step": 2816 |
| }, |
| { |
| "epoch": 4.26, |
| "learning_rate": 2.954545454545455e-05, |
| "loss": 0.2487, |
| "step": 3000 |
| }, |
| { |
| "epoch": 4.97, |
| "learning_rate": 1.1363636363636364e-06, |
| "loss": 0.2187, |
| "step": 3500 |
| }, |
| { |
| "epoch": 5.0, |
| "eval_accuracy": 0.9784, |
| "eval_loss": 0.06780806183815002, |
| "eval_runtime": 23.3979, |
| "eval_samples_per_second": 213.694, |
| "eval_steps_per_second": 3.376, |
| "step": 3520 |
| } |
| ], |
| "max_steps": 3520, |
| "num_train_epochs": 5, |
| "total_flos": 4.4035226145792e+18, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
