hushem_1x_deit_small_adamax_00001_fold1
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: 1.1270
- 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: 1e-05
- 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 | 1.3199 | 0.3333 |
| 1.3414 | 2.0 | 12 | 1.2923 | 0.4667 |
| 1.3414 | 3.0 | 18 | 1.2886 | 0.4667 |
| 1.0791 | 4.0 | 24 | 1.2761 | 0.4667 |
| 0.9244 | 5.0 | 30 | 1.2453 | 0.4889 |
| 0.9244 | 6.0 | 36 | 1.2252 | 0.4667 |
| 0.7694 | 7.0 | 42 | 1.2158 | 0.5111 |
| 0.7694 | 8.0 | 48 | 1.2163 | 0.4667 |
| 0.6552 | 9.0 | 54 | 1.2081 | 0.5111 |
| 0.5314 | 10.0 | 60 | 1.1883 | 0.5556 |
| 0.5314 | 11.0 | 66 | 1.1802 | 0.5556 |
| 0.4407 | 12.0 | 72 | 1.1737 | 0.5778 |
| 0.4407 | 13.0 | 78 | 1.1623 | 0.6222 |
| 0.3864 | 14.0 | 84 | 1.1625 | 0.6222 |
| 0.3093 | 15.0 | 90 | 1.1653 | 0.6222 |
| 0.3093 | 16.0 | 96 | 1.1658 | 0.6222 |
| 0.2597 | 17.0 | 102 | 1.1519 | 0.6444 |
| 0.2597 | 18.0 | 108 | 1.1466 | 0.6222 |
| 0.2099 | 19.0 | 114 | 1.1591 | 0.6 |
| 0.1766 | 20.0 | 120 | 1.1509 | 0.5778 |
| 0.1766 | 21.0 | 126 | 1.1488 | 0.5778 |
| 0.1537 | 22.0 | 132 | 1.1482 | 0.5778 |
| 0.1537 | 23.0 | 138 | 1.1427 | 0.6222 |
| 0.1244 | 24.0 | 144 | 1.1370 | 0.6 |
| 0.103 | 25.0 | 150 | 1.1285 | 0.6 |
| 0.103 | 26.0 | 156 | 1.1323 | 0.6 |
| 0.089 | 27.0 | 162 | 1.1268 | 0.6 |
| 0.089 | 28.0 | 168 | 1.1377 | 0.6 |
| 0.0777 | 29.0 | 174 | 1.1346 | 0.6 |
| 0.068 | 30.0 | 180 | 1.1274 | 0.6 |
| 0.068 | 31.0 | 186 | 1.1199 | 0.6 |
| 0.0597 | 32.0 | 192 | 1.1245 | 0.6 |
| 0.0597 | 33.0 | 198 | 1.1296 | 0.6 |
| 0.0547 | 34.0 | 204 | 1.1270 | 0.6 |
| 0.0493 | 35.0 | 210 | 1.1241 | 0.6 |
| 0.0493 | 36.0 | 216 | 1.1250 | 0.6 |
| 0.0441 | 37.0 | 222 | 1.1253 | 0.6 |
| 0.0441 | 38.0 | 228 | 1.1296 | 0.6 |
| 0.0409 | 39.0 | 234 | 1.1287 | 0.6 |
| 0.0405 | 40.0 | 240 | 1.1275 | 0.6 |
| 0.0405 | 41.0 | 246 | 1.1272 | 0.6 |
| 0.0391 | 42.0 | 252 | 1.1270 | 0.6 |
| 0.0391 | 43.0 | 258 | 1.1270 | 0.6 |
| 0.0395 | 44.0 | 264 | 1.1270 | 0.6 |
| 0.0377 | 45.0 | 270 | 1.1270 | 0.6 |
| 0.0377 | 46.0 | 276 | 1.1270 | 0.6 |
| 0.0388 | 47.0 | 282 | 1.1270 | 0.6 |
| 0.0388 | 48.0 | 288 | 1.1270 | 0.6 |
| 0.0366 | 49.0 | 294 | 1.1270 | 0.6 |
| 0.0396 | 50.0 | 300 | 1.1270 | 0.6 |
Framework versions
- Transformers 4.35.0
- Pytorch 2.1.0+cu118
- Datasets 2.14.6
- Tokenizers 0.14.1