smids_5x_deit_tiny_adamax_00001_fold2
This model is a fine-tuned version of facebook/deit-small-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 1.0539
- Accuracy: 0.8752
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 1e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 50
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.3245 | 1.0 | 375 | 0.3357 | 0.8586 |
| 0.2435 | 2.0 | 750 | 0.3012 | 0.8802 |
| 0.1837 | 3.0 | 1125 | 0.3092 | 0.8802 |
| 0.0922 | 4.0 | 1500 | 0.3362 | 0.8719 |
| 0.064 | 5.0 | 1875 | 0.4063 | 0.8619 |
| 0.0948 | 6.0 | 2250 | 0.4674 | 0.8619 |
| 0.0452 | 7.0 | 2625 | 0.5334 | 0.8602 |
| 0.0373 | 8.0 | 3000 | 0.6077 | 0.8619 |
| 0.0111 | 9.0 | 3375 | 0.6364 | 0.8769 |
| 0.0018 | 10.0 | 3750 | 0.7083 | 0.8636 |
| 0.0038 | 11.0 | 4125 | 0.7404 | 0.8752 |
| 0.0175 | 12.0 | 4500 | 0.8300 | 0.8719 |
| 0.0012 | 13.0 | 4875 | 0.8986 | 0.8652 |
| 0.0087 | 14.0 | 5250 | 0.8825 | 0.8686 |
| 0.004 | 15.0 | 5625 | 0.8822 | 0.8785 |
| 0.0001 | 16.0 | 6000 | 0.9237 | 0.8735 |
| 0.0162 | 17.0 | 6375 | 0.9830 | 0.8619 |
| 0.0 | 18.0 | 6750 | 1.0120 | 0.8702 |
| 0.0 | 19.0 | 7125 | 1.0192 | 0.8719 |
| 0.0001 | 20.0 | 7500 | 0.9781 | 0.8735 |
| 0.0 | 21.0 | 7875 | 1.0188 | 0.8702 |
| 0.0 | 22.0 | 8250 | 0.9776 | 0.8735 |
| 0.0 | 23.0 | 8625 | 1.0494 | 0.8702 |
| 0.0 | 24.0 | 9000 | 0.9531 | 0.8752 |
| 0.0 | 25.0 | 9375 | 1.0293 | 0.8719 |
| 0.0 | 26.0 | 9750 | 1.0427 | 0.8652 |
| 0.0 | 27.0 | 10125 | 1.0483 | 0.8719 |
| 0.0 | 28.0 | 10500 | 1.0202 | 0.8735 |
| 0.0 | 29.0 | 10875 | 1.0779 | 0.8686 |
| 0.0 | 30.0 | 11250 | 1.0065 | 0.8719 |
| 0.0018 | 31.0 | 11625 | 1.0762 | 0.8702 |
| 0.0202 | 32.0 | 12000 | 1.0874 | 0.8669 |
| 0.0024 | 33.0 | 12375 | 1.0366 | 0.8735 |
| 0.0 | 34.0 | 12750 | 1.1165 | 0.8686 |
| 0.0 | 35.0 | 13125 | 1.0244 | 0.8752 |
| 0.0 | 36.0 | 13500 | 1.1014 | 0.8719 |
| 0.0 | 37.0 | 13875 | 1.0995 | 0.8702 |
| 0.0 | 38.0 | 14250 | 1.1070 | 0.8719 |
| 0.0 | 39.0 | 14625 | 1.0209 | 0.8769 |
| 0.0048 | 40.0 | 15000 | 1.0540 | 0.8752 |
| 0.0 | 41.0 | 15375 | 1.0624 | 0.8752 |
| 0.0015 | 42.0 | 15750 | 1.0637 | 0.8752 |
| 0.0013 | 43.0 | 16125 | 1.0536 | 0.8752 |
| 0.0013 | 44.0 | 16500 | 1.0479 | 0.8752 |
| 0.0013 | 45.0 | 16875 | 1.0540 | 0.8752 |
| 0.0 | 46.0 | 17250 | 1.0694 | 0.8752 |
| 0.0016 | 47.0 | 17625 | 1.0601 | 0.8752 |
| 0.0 | 48.0 | 18000 | 1.0596 | 0.8752 |
| 0.0013 | 49.0 | 18375 | 1.0574 | 0.8752 |
| 0.0012 | 50.0 | 18750 | 1.0539 | 0.8752 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2