hushem_1x_deit_tiny_adamax_001_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: 1.6766
- Accuracy: 0.6
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 |
|---|---|---|---|---|
| No log | 1.0 | 6 | 1.3194 | 0.2889 |
| 1.3705 | 2.0 | 12 | 1.2766 | 0.3778 |
| 1.3705 | 3.0 | 18 | 1.3268 | 0.5333 |
| 0.7361 | 4.0 | 24 | 1.2927 | 0.5556 |
| 0.3404 | 5.0 | 30 | 1.3610 | 0.5556 |
| 0.3404 | 6.0 | 36 | 1.1429 | 0.5778 |
| 0.1188 | 7.0 | 42 | 1.5833 | 0.5333 |
| 0.1188 | 8.0 | 48 | 1.2765 | 0.6667 |
| 0.0229 | 9.0 | 54 | 1.4099 | 0.6222 |
| 0.0046 | 10.0 | 60 | 1.4395 | 0.6 |
| 0.0046 | 11.0 | 66 | 1.6161 | 0.5556 |
| 0.0013 | 12.0 | 72 | 1.5774 | 0.5778 |
| 0.0013 | 13.0 | 78 | 1.5201 | 0.6 |
| 0.0007 | 14.0 | 84 | 1.5608 | 0.6 |
| 0.0005 | 15.0 | 90 | 1.6187 | 0.5778 |
| 0.0005 | 16.0 | 96 | 1.6424 | 0.5778 |
| 0.0004 | 17.0 | 102 | 1.6470 | 0.5778 |
| 0.0004 | 18.0 | 108 | 1.6480 | 0.6 |
| 0.0003 | 19.0 | 114 | 1.6471 | 0.6 |
| 0.0003 | 20.0 | 120 | 1.6450 | 0.6 |
| 0.0003 | 21.0 | 126 | 1.6532 | 0.6 |
| 0.0003 | 22.0 | 132 | 1.6559 | 0.6 |
| 0.0003 | 23.0 | 138 | 1.6612 | 0.6 |
| 0.0003 | 24.0 | 144 | 1.6668 | 0.6 |
| 0.0002 | 25.0 | 150 | 1.6718 | 0.6 |
| 0.0002 | 26.0 | 156 | 1.6748 | 0.6 |
| 0.0002 | 27.0 | 162 | 1.6728 | 0.6 |
| 0.0002 | 28.0 | 168 | 1.6726 | 0.6 |
| 0.0002 | 29.0 | 174 | 1.6718 | 0.6 |
| 0.0002 | 30.0 | 180 | 1.6716 | 0.6 |
| 0.0002 | 31.0 | 186 | 1.6738 | 0.6 |
| 0.0002 | 32.0 | 192 | 1.6734 | 0.6 |
| 0.0002 | 33.0 | 198 | 1.6748 | 0.6 |
| 0.0002 | 34.0 | 204 | 1.6753 | 0.6 |
| 0.0002 | 35.0 | 210 | 1.6740 | 0.6 |
| 0.0002 | 36.0 | 216 | 1.6735 | 0.6 |
| 0.0002 | 37.0 | 222 | 1.6732 | 0.6 |
| 0.0002 | 38.0 | 228 | 1.6740 | 0.6 |
| 0.0002 | 39.0 | 234 | 1.6751 | 0.6 |
| 0.0002 | 40.0 | 240 | 1.6758 | 0.6 |
| 0.0002 | 41.0 | 246 | 1.6766 | 0.6 |
| 0.0002 | 42.0 | 252 | 1.6766 | 0.6 |
| 0.0002 | 43.0 | 258 | 1.6766 | 0.6 |
| 0.0002 | 44.0 | 264 | 1.6766 | 0.6 |
| 0.0002 | 45.0 | 270 | 1.6766 | 0.6 |
| 0.0002 | 46.0 | 276 | 1.6766 | 0.6 |
| 0.0002 | 47.0 | 282 | 1.6766 | 0.6 |
| 0.0002 | 48.0 | 288 | 1.6766 | 0.6 |
| 0.0002 | 49.0 | 294 | 1.6766 | 0.6 |
| 0.0002 | 50.0 | 300 | 1.6766 | 0.6 |
Framework versions
- Transformers 4.35.0
- Pytorch 2.1.0+cu118
- Datasets 2.14.6
- Tokenizers 0.14.1