smids_3x_deit_small_adamax_0001_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: 1.0636
- Accuracy: 0.8769
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: 0.0001
- 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 |
|---|---|---|---|---|
| 0.2927 | 1.0 | 225 | 0.3073 | 0.8686 |
| 0.1512 | 2.0 | 450 | 0.3158 | 0.8902 |
| 0.0739 | 3.0 | 675 | 0.4826 | 0.8602 |
| 0.0696 | 4.0 | 900 | 0.4520 | 0.8835 |
| 0.0095 | 5.0 | 1125 | 0.6442 | 0.8636 |
| 0.0162 | 6.0 | 1350 | 0.8340 | 0.8403 |
| 0.0336 | 7.0 | 1575 | 0.6936 | 0.8669 |
| 0.0025 | 8.0 | 1800 | 0.8353 | 0.8669 |
| 0.0105 | 9.0 | 2025 | 0.8036 | 0.8702 |
| 0.0025 | 10.0 | 2250 | 0.8245 | 0.8769 |
| 0.0004 | 11.0 | 2475 | 0.9095 | 0.8802 |
| 0.0003 | 12.0 | 2700 | 1.0840 | 0.8669 |
| 0.0001 | 13.0 | 2925 | 0.9475 | 0.8735 |
| 0.016 | 14.0 | 3150 | 0.9122 | 0.8835 |
| 0.0 | 15.0 | 3375 | 0.7790 | 0.8918 |
| 0.0 | 16.0 | 3600 | 0.7787 | 0.8785 |
| 0.0029 | 17.0 | 3825 | 0.8308 | 0.8918 |
| 0.0 | 18.0 | 4050 | 0.9812 | 0.8752 |
| 0.0006 | 19.0 | 4275 | 0.9540 | 0.8686 |
| 0.0 | 20.0 | 4500 | 0.9292 | 0.8785 |
| 0.0 | 21.0 | 4725 | 0.9652 | 0.8785 |
| 0.0 | 22.0 | 4950 | 0.9762 | 0.8802 |
| 0.0029 | 23.0 | 5175 | 0.9826 | 0.8702 |
| 0.0 | 24.0 | 5400 | 0.9763 | 0.8669 |
| 0.0047 | 25.0 | 5625 | 0.9733 | 0.8719 |
| 0.0 | 26.0 | 5850 | 0.9680 | 0.8686 |
| 0.0 | 27.0 | 6075 | 0.9775 | 0.8869 |
| 0.0 | 28.0 | 6300 | 0.9726 | 0.8769 |
| 0.0 | 29.0 | 6525 | 1.0293 | 0.8702 |
| 0.0036 | 30.0 | 6750 | 1.0029 | 0.8885 |
| 0.0 | 31.0 | 6975 | 0.9992 | 0.8686 |
| 0.0 | 32.0 | 7200 | 0.9937 | 0.8686 |
| 0.0 | 33.0 | 7425 | 0.9950 | 0.8802 |
| 0.0 | 34.0 | 7650 | 0.9997 | 0.8735 |
| 0.0 | 35.0 | 7875 | 1.0478 | 0.8769 |
| 0.0033 | 36.0 | 8100 | 1.0278 | 0.8686 |
| 0.0 | 37.0 | 8325 | 1.0247 | 0.8802 |
| 0.0 | 38.0 | 8550 | 1.0317 | 0.8785 |
| 0.0032 | 39.0 | 8775 | 1.0297 | 0.8819 |
| 0.0 | 40.0 | 9000 | 1.0396 | 0.8802 |
| 0.0 | 41.0 | 9225 | 1.0404 | 0.8752 |
| 0.0 | 42.0 | 9450 | 1.0476 | 0.8785 |
| 0.0 | 43.0 | 9675 | 1.0513 | 0.8802 |
| 0.0 | 44.0 | 9900 | 1.0502 | 0.8769 |
| 0.0 | 45.0 | 10125 | 1.0584 | 0.8752 |
| 0.0 | 46.0 | 10350 | 1.0568 | 0.8769 |
| 0.0 | 47.0 | 10575 | 1.0576 | 0.8785 |
| 0.0 | 48.0 | 10800 | 1.0616 | 0.8785 |
| 0.0023 | 49.0 | 11025 | 1.0626 | 0.8769 |
| 0.0023 | 50.0 | 11250 | 1.0636 | 0.8769 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2