smids_1x_deit_tiny_adamax_0001_fold2
This model is a fine-tuned version of facebook/deit-tiny-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.8884
- Accuracy: 0.8735
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 |
|---|---|---|---|---|
| 0.4433 | 1.0 | 75 | 0.4087 | 0.8136 |
| 0.2718 | 2.0 | 150 | 0.3393 | 0.8636 |
| 0.1523 | 3.0 | 225 | 0.3549 | 0.8602 |
| 0.1126 | 4.0 | 300 | 0.4203 | 0.8785 |
| 0.1857 | 5.0 | 375 | 0.5016 | 0.8702 |
| 0.1676 | 6.0 | 450 | 0.6812 | 0.8403 |
| 0.0685 | 7.0 | 525 | 0.6011 | 0.8719 |
| 0.0238 | 8.0 | 600 | 0.6670 | 0.8686 |
| 0.0519 | 9.0 | 675 | 0.6013 | 0.8686 |
| 0.0386 | 10.0 | 750 | 0.7008 | 0.8719 |
| 0.0148 | 11.0 | 825 | 0.7193 | 0.8619 |
| 0.0007 | 12.0 | 900 | 0.7563 | 0.8752 |
| 0.0241 | 13.0 | 975 | 0.7693 | 0.8636 |
| 0.0307 | 14.0 | 1050 | 0.8760 | 0.8636 |
| 0.0115 | 15.0 | 1125 | 0.7808 | 0.8719 |
| 0.0035 | 16.0 | 1200 | 0.7588 | 0.8669 |
| 0.0001 | 17.0 | 1275 | 0.8971 | 0.8619 |
| 0.0102 | 18.0 | 1350 | 0.7909 | 0.8719 |
| 0.0001 | 19.0 | 1425 | 0.7984 | 0.8636 |
| 0.004 | 20.0 | 1500 | 0.8206 | 0.8669 |
| 0.0001 | 21.0 | 1575 | 0.8515 | 0.8752 |
| 0.0 | 22.0 | 1650 | 0.7887 | 0.8752 |
| 0.0 | 23.0 | 1725 | 0.9036 | 0.8719 |
| 0.0001 | 24.0 | 1800 | 0.8151 | 0.8735 |
| 0.0 | 25.0 | 1875 | 0.8674 | 0.8669 |
| 0.0 | 26.0 | 1950 | 0.8463 | 0.8702 |
| 0.0044 | 27.0 | 2025 | 0.8541 | 0.8669 |
| 0.0043 | 28.0 | 2100 | 0.8322 | 0.8669 |
| 0.002 | 29.0 | 2175 | 0.8405 | 0.8686 |
| 0.0043 | 30.0 | 2250 | 0.8433 | 0.8686 |
| 0.0034 | 31.0 | 2325 | 0.8353 | 0.8752 |
| 0.0 | 32.0 | 2400 | 0.8744 | 0.8702 |
| 0.0 | 33.0 | 2475 | 0.8688 | 0.8669 |
| 0.0 | 34.0 | 2550 | 0.8557 | 0.8669 |
| 0.0091 | 35.0 | 2625 | 0.8746 | 0.8669 |
| 0.0 | 36.0 | 2700 | 0.8586 | 0.8686 |
| 0.0 | 37.0 | 2775 | 0.8715 | 0.8702 |
| 0.0 | 38.0 | 2850 | 0.8844 | 0.8719 |
| 0.0019 | 39.0 | 2925 | 0.8957 | 0.8735 |
| 0.0 | 40.0 | 3000 | 0.8803 | 0.8752 |
| 0.0031 | 41.0 | 3075 | 0.8802 | 0.8752 |
| 0.0 | 42.0 | 3150 | 0.8828 | 0.8752 |
| 0.0026 | 43.0 | 3225 | 0.8803 | 0.8735 |
| 0.003 | 44.0 | 3300 | 0.8878 | 0.8752 |
| 0.0 | 45.0 | 3375 | 0.8894 | 0.8702 |
| 0.0 | 46.0 | 3450 | 0.8874 | 0.8735 |
| 0.0024 | 47.0 | 3525 | 0.8958 | 0.8752 |
| 0.0 | 48.0 | 3600 | 0.8893 | 0.8735 |
| 0.0023 | 49.0 | 3675 | 0.8880 | 0.8735 |
| 0.0021 | 50.0 | 3750 | 0.8884 | 0.8735 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0