hushem_5x_deit_small_sgd_001_fold3
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: 1.1692
- Accuracy: 0.4186
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 |
|---|---|---|---|---|
| 1.4612 | 1.0 | 28 | 1.5118 | 0.2558 |
| 1.3805 | 2.0 | 56 | 1.4497 | 0.2093 |
| 1.3041 | 3.0 | 84 | 1.4123 | 0.2093 |
| 1.3096 | 4.0 | 112 | 1.3866 | 0.2558 |
| 1.3306 | 5.0 | 140 | 1.3677 | 0.2558 |
| 1.2628 | 6.0 | 168 | 1.3531 | 0.2791 |
| 1.2467 | 7.0 | 196 | 1.3404 | 0.2791 |
| 1.2266 | 8.0 | 224 | 1.3306 | 0.3023 |
| 1.194 | 9.0 | 252 | 1.3214 | 0.3023 |
| 1.1663 | 10.0 | 280 | 1.3122 | 0.2791 |
| 1.1821 | 11.0 | 308 | 1.3053 | 0.3023 |
| 1.1428 | 12.0 | 336 | 1.2975 | 0.3023 |
| 1.1221 | 13.0 | 364 | 1.2914 | 0.3023 |
| 1.0981 | 14.0 | 392 | 1.2852 | 0.3023 |
| 1.1081 | 15.0 | 420 | 1.2792 | 0.3488 |
| 1.0434 | 16.0 | 448 | 1.2755 | 0.3488 |
| 1.0661 | 17.0 | 476 | 1.2676 | 0.3721 |
| 1.0349 | 18.0 | 504 | 1.2624 | 0.3721 |
| 1.0251 | 19.0 | 532 | 1.2563 | 0.3721 |
| 0.9878 | 20.0 | 560 | 1.2508 | 0.3721 |
| 1.0276 | 21.0 | 588 | 1.2459 | 0.3721 |
| 0.9848 | 22.0 | 616 | 1.2419 | 0.3721 |
| 0.9825 | 23.0 | 644 | 1.2362 | 0.3721 |
| 0.9289 | 24.0 | 672 | 1.2312 | 0.3721 |
| 0.9221 | 25.0 | 700 | 1.2274 | 0.3721 |
| 0.9187 | 26.0 | 728 | 1.2222 | 0.3721 |
| 0.9248 | 27.0 | 756 | 1.2177 | 0.3953 |
| 0.9505 | 28.0 | 784 | 1.2135 | 0.3721 |
| 0.9022 | 29.0 | 812 | 1.2094 | 0.3721 |
| 0.8445 | 30.0 | 840 | 1.2060 | 0.3721 |
| 0.861 | 31.0 | 868 | 1.2023 | 0.3953 |
| 0.9005 | 32.0 | 896 | 1.1985 | 0.3953 |
| 0.8936 | 33.0 | 924 | 1.1956 | 0.3953 |
| 0.8469 | 34.0 | 952 | 1.1923 | 0.3953 |
| 0.9675 | 35.0 | 980 | 1.1892 | 0.3953 |
| 0.8615 | 36.0 | 1008 | 1.1869 | 0.3953 |
| 0.8944 | 37.0 | 1036 | 1.1838 | 0.3953 |
| 0.8374 | 38.0 | 1064 | 1.1820 | 0.3953 |
| 0.8431 | 39.0 | 1092 | 1.1797 | 0.3953 |
| 0.8451 | 40.0 | 1120 | 1.1776 | 0.3953 |
| 0.8614 | 41.0 | 1148 | 1.1756 | 0.3953 |
| 0.8347 | 42.0 | 1176 | 1.1741 | 0.3953 |
| 0.8159 | 43.0 | 1204 | 1.1724 | 0.4419 |
| 0.8229 | 44.0 | 1232 | 1.1713 | 0.4419 |
| 0.8046 | 45.0 | 1260 | 1.1703 | 0.4186 |
| 0.8129 | 46.0 | 1288 | 1.1698 | 0.4186 |
| 0.8156 | 47.0 | 1316 | 1.1693 | 0.4186 |
| 0.8259 | 48.0 | 1344 | 1.1692 | 0.4186 |
| 0.7943 | 49.0 | 1372 | 1.1692 | 0.4186 |
| 0.8288 | 50.0 | 1400 | 1.1692 | 0.4186 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0