hushem_5x_deit_tiny_adamax_0001_fold3
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: 0.2866
- Accuracy: 0.9302
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 |
|---|---|---|---|---|
| 0.8773 | 1.0 | 28 | 0.7509 | 0.6512 |
| 0.2729 | 2.0 | 56 | 0.4204 | 0.8140 |
| 0.0799 | 3.0 | 84 | 0.2897 | 0.9070 |
| 0.0491 | 4.0 | 112 | 0.3313 | 0.8372 |
| 0.0108 | 5.0 | 140 | 0.2304 | 0.9535 |
| 0.0073 | 6.0 | 168 | 0.3224 | 0.9302 |
| 0.0019 | 7.0 | 196 | 0.1523 | 0.9302 |
| 0.0004 | 8.0 | 224 | 0.1884 | 0.9070 |
| 0.0002 | 9.0 | 252 | 0.2071 | 0.9535 |
| 0.0001 | 10.0 | 280 | 0.2200 | 0.9535 |
| 0.0001 | 11.0 | 308 | 0.2281 | 0.9302 |
| 0.0001 | 12.0 | 336 | 0.2328 | 0.9302 |
| 0.0001 | 13.0 | 364 | 0.2370 | 0.9302 |
| 0.0001 | 14.0 | 392 | 0.2423 | 0.9302 |
| 0.0001 | 15.0 | 420 | 0.2453 | 0.9302 |
| 0.0001 | 16.0 | 448 | 0.2484 | 0.9302 |
| 0.0001 | 17.0 | 476 | 0.2522 | 0.9302 |
| 0.0001 | 18.0 | 504 | 0.2526 | 0.9302 |
| 0.0001 | 19.0 | 532 | 0.2564 | 0.9302 |
| 0.0001 | 20.0 | 560 | 0.2573 | 0.9302 |
| 0.0001 | 21.0 | 588 | 0.2605 | 0.9302 |
| 0.0001 | 22.0 | 616 | 0.2600 | 0.9302 |
| 0.0001 | 23.0 | 644 | 0.2633 | 0.9302 |
| 0.0001 | 24.0 | 672 | 0.2654 | 0.9302 |
| 0.0001 | 25.0 | 700 | 0.2772 | 0.9302 |
| 0.0 | 26.0 | 728 | 0.2757 | 0.9302 |
| 0.0 | 27.0 | 756 | 0.2754 | 0.9302 |
| 0.0 | 28.0 | 784 | 0.2779 | 0.9302 |
| 0.0 | 29.0 | 812 | 0.2773 | 0.9302 |
| 0.0 | 30.0 | 840 | 0.2767 | 0.9302 |
| 0.0 | 31.0 | 868 | 0.2771 | 0.9302 |
| 0.0 | 32.0 | 896 | 0.2773 | 0.9302 |
| 0.0 | 33.0 | 924 | 0.2769 | 0.9302 |
| 0.0 | 34.0 | 952 | 0.2787 | 0.9302 |
| 0.0 | 35.0 | 980 | 0.2791 | 0.9302 |
| 0.0 | 36.0 | 1008 | 0.2792 | 0.9302 |
| 0.0 | 37.0 | 1036 | 0.2807 | 0.9302 |
| 0.0 | 38.0 | 1064 | 0.2808 | 0.9302 |
| 0.0 | 39.0 | 1092 | 0.2845 | 0.9302 |
| 0.0 | 40.0 | 1120 | 0.2848 | 0.9302 |
| 0.0 | 41.0 | 1148 | 0.2848 | 0.9302 |
| 0.0 | 42.0 | 1176 | 0.2855 | 0.9302 |
| 0.0 | 43.0 | 1204 | 0.2852 | 0.9302 |
| 0.0 | 44.0 | 1232 | 0.2853 | 0.9302 |
| 0.0 | 45.0 | 1260 | 0.2865 | 0.9302 |
| 0.0 | 46.0 | 1288 | 0.2864 | 0.9302 |
| 0.0 | 47.0 | 1316 | 0.2865 | 0.9302 |
| 0.0 | 48.0 | 1344 | 0.2866 | 0.9302 |
| 0.0 | 49.0 | 1372 | 0.2866 | 0.9302 |
| 0.0 | 50.0 | 1400 | 0.2866 | 0.9302 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0