hushem_1x_deit_tiny_sgd_lr0001_fold5
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: 1.5789
- Accuracy: 0.1463
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 | 1.6455 | 0.1220 |
| 1.6035 | 2.0 | 12 | 1.6420 | 0.1220 |
| 1.6035 | 3.0 | 18 | 1.6386 | 0.1463 |
| 1.6142 | 4.0 | 24 | 1.6353 | 0.1463 |
| 1.5857 | 5.0 | 30 | 1.6321 | 0.1463 |
| 1.5857 | 6.0 | 36 | 1.6289 | 0.1463 |
| 1.5718 | 7.0 | 42 | 1.6259 | 0.1463 |
| 1.5718 | 8.0 | 48 | 1.6232 | 0.1463 |
| 1.5833 | 9.0 | 54 | 1.6206 | 0.1463 |
| 1.5737 | 10.0 | 60 | 1.6178 | 0.1463 |
| 1.5737 | 11.0 | 66 | 1.6153 | 0.1463 |
| 1.5614 | 12.0 | 72 | 1.6128 | 0.1220 |
| 1.5614 | 13.0 | 78 | 1.6104 | 0.1220 |
| 1.5648 | 14.0 | 84 | 1.6081 | 0.1220 |
| 1.5575 | 15.0 | 90 | 1.6060 | 0.1220 |
| 1.5575 | 16.0 | 96 | 1.6040 | 0.1220 |
| 1.5452 | 17.0 | 102 | 1.6020 | 0.1220 |
| 1.5452 | 18.0 | 108 | 1.6002 | 0.1220 |
| 1.5768 | 19.0 | 114 | 1.5984 | 0.1220 |
| 1.5464 | 20.0 | 120 | 1.5966 | 0.1220 |
| 1.5464 | 21.0 | 126 | 1.5950 | 0.1220 |
| 1.5149 | 22.0 | 132 | 1.5934 | 0.1220 |
| 1.5149 | 23.0 | 138 | 1.5920 | 0.1220 |
| 1.6056 | 24.0 | 144 | 1.5905 | 0.1220 |
| 1.5161 | 25.0 | 150 | 1.5892 | 0.1220 |
| 1.5161 | 26.0 | 156 | 1.5879 | 0.1220 |
| 1.519 | 27.0 | 162 | 1.5868 | 0.1220 |
| 1.519 | 28.0 | 168 | 1.5857 | 0.1220 |
| 1.5531 | 29.0 | 174 | 1.5848 | 0.1220 |
| 1.5347 | 30.0 | 180 | 1.5839 | 0.1220 |
| 1.5347 | 31.0 | 186 | 1.5831 | 0.1220 |
| 1.5238 | 32.0 | 192 | 1.5824 | 0.1220 |
| 1.5238 | 33.0 | 198 | 1.5817 | 0.1463 |
| 1.5463 | 34.0 | 204 | 1.5811 | 0.1463 |
| 1.5219 | 35.0 | 210 | 1.5805 | 0.1463 |
| 1.5219 | 36.0 | 216 | 1.5800 | 0.1463 |
| 1.5056 | 37.0 | 222 | 1.5797 | 0.1463 |
| 1.5056 | 38.0 | 228 | 1.5794 | 0.1463 |
| 1.5505 | 39.0 | 234 | 1.5791 | 0.1463 |
| 1.5261 | 40.0 | 240 | 1.5790 | 0.1463 |
| 1.5261 | 41.0 | 246 | 1.5789 | 0.1463 |
| 1.5175 | 42.0 | 252 | 1.5789 | 0.1463 |
| 1.5175 | 43.0 | 258 | 1.5789 | 0.1463 |
| 1.5317 | 44.0 | 264 | 1.5789 | 0.1463 |
| 1.5241 | 45.0 | 270 | 1.5789 | 0.1463 |
| 1.5241 | 46.0 | 276 | 1.5789 | 0.1463 |
| 1.5533 | 47.0 | 282 | 1.5789 | 0.1463 |
| 1.5533 | 48.0 | 288 | 1.5789 | 0.1463 |
| 1.4945 | 49.0 | 294 | 1.5789 | 0.1463 |
| 1.5379 | 50.0 | 300 | 1.5789 | 0.1463 |
Framework versions
- Transformers 4.35.0
- Pytorch 2.1.0+cu118
- Datasets 2.14.6
- Tokenizers 0.14.1