| { |
| "best_metric": 1.0, |
| "best_model_checkpoint": "./swin-soiral/checkpoint-32", |
| "epoch": 10.0, |
| "eval_steps": 500, |
| "global_step": 80, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.62, |
| "learning_rate": 7.5e-06, |
| "loss": 1.3592, |
| "step": 5 |
| }, |
| { |
| "epoch": 1.0, |
| "eval_accuracy": 0.3, |
| "eval_loss": 1.3065917491912842, |
| "eval_runtime": 1.1424, |
| "eval_samples_per_second": 17.507, |
| "eval_steps_per_second": 2.626, |
| "step": 8 |
| }, |
| { |
| "epoch": 1.25, |
| "learning_rate": 1.4791666666666668e-05, |
| "loss": 1.2383, |
| "step": 10 |
| }, |
| { |
| "epoch": 1.88, |
| "learning_rate": 1.375e-05, |
| "loss": 0.9142, |
| "step": 15 |
| }, |
| { |
| "epoch": 2.0, |
| "eval_accuracy": 0.8, |
| "eval_loss": 0.751232922077179, |
| "eval_runtime": 1.0508, |
| "eval_samples_per_second": 19.034, |
| "eval_steps_per_second": 2.855, |
| "step": 16 |
| }, |
| { |
| "epoch": 2.5, |
| "learning_rate": 1.2916666666666668e-05, |
| "loss": 0.6865, |
| "step": 20 |
| }, |
| { |
| "epoch": 3.0, |
| "eval_accuracy": 0.75, |
| "eval_loss": 0.5134674310684204, |
| "eval_runtime": 1.5982, |
| "eval_samples_per_second": 12.514, |
| "eval_steps_per_second": 1.877, |
| "step": 24 |
| }, |
| { |
| "epoch": 3.12, |
| "learning_rate": 1.1874999999999999e-05, |
| "loss": 0.5304, |
| "step": 25 |
| }, |
| { |
| "epoch": 3.75, |
| "learning_rate": 1.0833333333333334e-05, |
| "loss": 0.3784, |
| "step": 30 |
| }, |
| { |
| "epoch": 4.0, |
| "eval_accuracy": 1.0, |
| "eval_loss": 0.25737839937210083, |
| "eval_runtime": 1.0292, |
| "eval_samples_per_second": 19.433, |
| "eval_steps_per_second": 2.915, |
| "step": 32 |
| }, |
| { |
| "epoch": 4.38, |
| "learning_rate": 9.791666666666666e-06, |
| "loss": 0.2918, |
| "step": 35 |
| }, |
| { |
| "epoch": 5.0, |
| "learning_rate": 8.750000000000001e-06, |
| "loss": 0.2105, |
| "step": 40 |
| }, |
| { |
| "epoch": 5.0, |
| "eval_accuracy": 1.0, |
| "eval_loss": 0.13916663825511932, |
| "eval_runtime": 1.0609, |
| "eval_samples_per_second": 18.851, |
| "eval_steps_per_second": 2.828, |
| "step": 40 |
| }, |
| { |
| "epoch": 5.62, |
| "learning_rate": 7.708333333333332e-06, |
| "loss": 0.1395, |
| "step": 45 |
| }, |
| { |
| "epoch": 6.0, |
| "eval_accuracy": 1.0, |
| "eval_loss": 0.07838249206542969, |
| "eval_runtime": 1.0235, |
| "eval_samples_per_second": 19.542, |
| "eval_steps_per_second": 2.931, |
| "step": 48 |
| }, |
| { |
| "epoch": 6.25, |
| "learning_rate": 6.666666666666667e-06, |
| "loss": 0.0924, |
| "step": 50 |
| }, |
| { |
| "epoch": 6.88, |
| "learning_rate": 5.625e-06, |
| "loss": 0.0822, |
| "step": 55 |
| }, |
| { |
| "epoch": 7.0, |
| "eval_accuracy": 1.0, |
| "eval_loss": 0.043302156031131744, |
| "eval_runtime": 1.0416, |
| "eval_samples_per_second": 19.202, |
| "eval_steps_per_second": 2.88, |
| "step": 56 |
| }, |
| { |
| "epoch": 7.5, |
| "learning_rate": 4.583333333333334e-06, |
| "loss": 0.0772, |
| "step": 60 |
| }, |
| { |
| "epoch": 8.0, |
| "eval_accuracy": 1.0, |
| "eval_loss": 0.028308678418397903, |
| "eval_runtime": 1.0513, |
| "eval_samples_per_second": 19.024, |
| "eval_steps_per_second": 2.854, |
| "step": 64 |
| }, |
| { |
| "epoch": 8.12, |
| "learning_rate": 3.5416666666666665e-06, |
| "loss": 0.0307, |
| "step": 65 |
| }, |
| { |
| "epoch": 8.75, |
| "learning_rate": 2.4999999999999998e-06, |
| "loss": 0.0363, |
| "step": 70 |
| }, |
| { |
| "epoch": 9.0, |
| "eval_accuracy": 1.0, |
| "eval_loss": 0.023053836077451706, |
| "eval_runtime": 1.0793, |
| "eval_samples_per_second": 18.531, |
| "eval_steps_per_second": 2.78, |
| "step": 72 |
| }, |
| { |
| "epoch": 9.38, |
| "learning_rate": 1.4583333333333335e-06, |
| "loss": 0.0319, |
| "step": 75 |
| }, |
| { |
| "epoch": 10.0, |
| "learning_rate": 4.1666666666666667e-07, |
| "loss": 0.0466, |
| "step": 80 |
| }, |
| { |
| "epoch": 10.0, |
| "eval_accuracy": 1.0, |
| "eval_loss": 0.020184706896543503, |
| "eval_runtime": 1.0352, |
| "eval_samples_per_second": 19.32, |
| "eval_steps_per_second": 2.898, |
| "step": 80 |
| } |
| ], |
| "logging_steps": 5, |
| "max_steps": 80, |
| "num_train_epochs": 10, |
| "save_steps": 500, |
| "total_flos": 2.1075452515934208e+17, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
