hushem_1x_deit_tiny_sgd_001_fold4
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: 1.2335
- Accuracy: 0.4524
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 |
|---|---|---|---|---|
| No log | 1.0 | 6 | 1.5918 | 0.2857 |
| 1.6404 | 2.0 | 12 | 1.5188 | 0.2857 |
| 1.6404 | 3.0 | 18 | 1.4665 | 0.2857 |
| 1.5241 | 4.0 | 24 | 1.4299 | 0.3333 |
| 1.4755 | 5.0 | 30 | 1.4106 | 0.3571 |
| 1.4755 | 6.0 | 36 | 1.3938 | 0.3095 |
| 1.4186 | 7.0 | 42 | 1.3803 | 0.2857 |
| 1.4186 | 8.0 | 48 | 1.3677 | 0.3810 |
| 1.3819 | 9.0 | 54 | 1.3558 | 0.3810 |
| 1.3541 | 10.0 | 60 | 1.3456 | 0.3810 |
| 1.3541 | 11.0 | 66 | 1.3370 | 0.3810 |
| 1.3363 | 12.0 | 72 | 1.3284 | 0.3810 |
| 1.3363 | 13.0 | 78 | 1.3193 | 0.3571 |
| 1.3168 | 14.0 | 84 | 1.3103 | 0.4048 |
| 1.2875 | 15.0 | 90 | 1.3032 | 0.4048 |
| 1.2875 | 16.0 | 96 | 1.2966 | 0.4048 |
| 1.2638 | 17.0 | 102 | 1.2902 | 0.4048 |
| 1.2638 | 18.0 | 108 | 1.2846 | 0.4048 |
| 1.2758 | 19.0 | 114 | 1.2805 | 0.4048 |
| 1.2611 | 20.0 | 120 | 1.2763 | 0.4048 |
| 1.2611 | 21.0 | 126 | 1.2724 | 0.4048 |
| 1.2411 | 22.0 | 132 | 1.2693 | 0.4048 |
| 1.2411 | 23.0 | 138 | 1.2666 | 0.4048 |
| 1.2357 | 24.0 | 144 | 1.2628 | 0.4048 |
| 1.231 | 25.0 | 150 | 1.2590 | 0.4048 |
| 1.231 | 26.0 | 156 | 1.2555 | 0.4048 |
| 1.2026 | 27.0 | 162 | 1.2531 | 0.4048 |
| 1.2026 | 28.0 | 168 | 1.2508 | 0.4048 |
| 1.2253 | 29.0 | 174 | 1.2482 | 0.4048 |
| 1.1949 | 30.0 | 180 | 1.2457 | 0.4048 |
| 1.1949 | 31.0 | 186 | 1.2436 | 0.4286 |
| 1.2025 | 32.0 | 192 | 1.2420 | 0.4286 |
| 1.2025 | 33.0 | 198 | 1.2406 | 0.4524 |
| 1.1709 | 34.0 | 204 | 1.2390 | 0.4524 |
| 1.1908 | 35.0 | 210 | 1.2376 | 0.4524 |
| 1.1908 | 36.0 | 216 | 1.2365 | 0.4524 |
| 1.1663 | 37.0 | 222 | 1.2358 | 0.4524 |
| 1.1663 | 38.0 | 228 | 1.2349 | 0.4524 |
| 1.1875 | 39.0 | 234 | 1.2342 | 0.4524 |
| 1.1799 | 40.0 | 240 | 1.2338 | 0.4524 |
| 1.1799 | 41.0 | 246 | 1.2336 | 0.4524 |
| 1.1658 | 42.0 | 252 | 1.2335 | 0.4524 |
| 1.1658 | 43.0 | 258 | 1.2335 | 0.4524 |
| 1.1875 | 44.0 | 264 | 1.2335 | 0.4524 |
| 1.1627 | 45.0 | 270 | 1.2335 | 0.4524 |
| 1.1627 | 46.0 | 276 | 1.2335 | 0.4524 |
| 1.1689 | 47.0 | 282 | 1.2335 | 0.4524 |
| 1.1689 | 48.0 | 288 | 1.2335 | 0.4524 |
| 1.1911 | 49.0 | 294 | 1.2335 | 0.4524 |
| 1.1557 | 50.0 | 300 | 1.2335 | 0.4524 |
Framework versions
- Transformers 4.35.0
- Pytorch 2.1.0+cu118
- Datasets 2.14.6
- Tokenizers 0.14.1