smids_3x_deit_small_adamax_001_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: 1.0171
- Accuracy: 0.8819
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: 0.001
- 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 |
|---|---|---|---|---|
| 0.4401 | 1.0 | 225 | 0.3048 | 0.8735 |
| 0.2924 | 2.0 | 450 | 0.3208 | 0.8702 |
| 0.2562 | 3.0 | 675 | 0.3212 | 0.8702 |
| 0.2244 | 4.0 | 900 | 0.3324 | 0.8735 |
| 0.2012 | 5.0 | 1125 | 0.4267 | 0.8602 |
| 0.1372 | 6.0 | 1350 | 0.5512 | 0.8353 |
| 0.1267 | 7.0 | 1575 | 0.3624 | 0.8869 |
| 0.0899 | 8.0 | 1800 | 0.5724 | 0.8436 |
| 0.0839 | 9.0 | 2025 | 0.5571 | 0.8702 |
| 0.0466 | 10.0 | 2250 | 0.5349 | 0.8569 |
| 0.0801 | 11.0 | 2475 | 0.5913 | 0.8502 |
| 0.0671 | 12.0 | 2700 | 0.5854 | 0.8752 |
| 0.0348 | 13.0 | 2925 | 0.6837 | 0.8552 |
| 0.0379 | 14.0 | 3150 | 0.5712 | 0.8752 |
| 0.0439 | 15.0 | 3375 | 0.6348 | 0.8702 |
| 0.003 | 16.0 | 3600 | 0.7977 | 0.8686 |
| 0.0261 | 17.0 | 3825 | 0.6729 | 0.8735 |
| 0.002 | 18.0 | 4050 | 0.7795 | 0.8636 |
| 0.001 | 19.0 | 4275 | 0.6782 | 0.8852 |
| 0.0058 | 20.0 | 4500 | 0.6727 | 0.8918 |
| 0.0166 | 21.0 | 4725 | 0.6389 | 0.8835 |
| 0.0009 | 22.0 | 4950 | 0.7419 | 0.8752 |
| 0.0138 | 23.0 | 5175 | 0.7956 | 0.8769 |
| 0.0224 | 24.0 | 5400 | 0.7981 | 0.8735 |
| 0.0076 | 25.0 | 5625 | 0.7548 | 0.8802 |
| 0.0182 | 26.0 | 5850 | 0.8476 | 0.8586 |
| 0.0001 | 27.0 | 6075 | 0.8394 | 0.8669 |
| 0.0 | 28.0 | 6300 | 0.7756 | 0.8802 |
| 0.0001 | 29.0 | 6525 | 0.8666 | 0.8752 |
| 0.0033 | 30.0 | 6750 | 0.8769 | 0.8935 |
| 0.0046 | 31.0 | 6975 | 0.8571 | 0.8835 |
| 0.0 | 32.0 | 7200 | 0.8804 | 0.8802 |
| 0.0022 | 33.0 | 7425 | 0.8723 | 0.8902 |
| 0.0 | 34.0 | 7650 | 0.8812 | 0.8869 |
| 0.0 | 35.0 | 7875 | 0.9182 | 0.8819 |
| 0.0032 | 36.0 | 8100 | 0.9012 | 0.8869 |
| 0.0 | 37.0 | 8325 | 0.9270 | 0.8885 |
| 0.0 | 38.0 | 8550 | 0.8810 | 0.8902 |
| 0.0032 | 39.0 | 8775 | 0.9223 | 0.8918 |
| 0.0 | 40.0 | 9000 | 0.9755 | 0.8918 |
| 0.0003 | 41.0 | 9225 | 0.9917 | 0.8885 |
| 0.0 | 42.0 | 9450 | 0.9880 | 0.8869 |
| 0.0 | 43.0 | 9675 | 0.9953 | 0.8869 |
| 0.0 | 44.0 | 9900 | 0.9968 | 0.8869 |
| 0.0 | 45.0 | 10125 | 1.0016 | 0.8869 |
| 0.0 | 46.0 | 10350 | 1.0040 | 0.8852 |
| 0.0 | 47.0 | 10575 | 1.0063 | 0.8835 |
| 0.0 | 48.0 | 10800 | 1.0129 | 0.8835 |
| 0.0023 | 49.0 | 11025 | 1.0163 | 0.8819 |
| 0.0023 | 50.0 | 11250 | 1.0171 | 0.8819 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2