hushem_5x_deit_tiny_rms_001_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: 1.8130
- Accuracy: 0.6098
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 |
|---|---|---|---|---|
| 1.7861 | 1.0 | 28 | 1.6178 | 0.2439 |
| 1.5274 | 2.0 | 56 | 1.5079 | 0.2683 |
| 1.4754 | 3.0 | 84 | 1.5602 | 0.2439 |
| 1.4772 | 4.0 | 112 | 1.4942 | 0.2683 |
| 1.3612 | 5.0 | 140 | 1.2123 | 0.4146 |
| 1.3875 | 6.0 | 168 | 1.3792 | 0.3415 |
| 1.2469 | 7.0 | 196 | 1.2932 | 0.3171 |
| 1.2171 | 8.0 | 224 | 1.2748 | 0.3902 |
| 1.2175 | 9.0 | 252 | 1.1348 | 0.4390 |
| 1.166 | 10.0 | 280 | 1.0946 | 0.4878 |
| 1.164 | 11.0 | 308 | 0.8922 | 0.6829 |
| 1.0132 | 12.0 | 336 | 0.8702 | 0.7317 |
| 1.0418 | 13.0 | 364 | 1.2510 | 0.5366 |
| 0.9219 | 14.0 | 392 | 0.9079 | 0.5854 |
| 0.9159 | 15.0 | 420 | 1.0015 | 0.5366 |
| 0.9295 | 16.0 | 448 | 0.9465 | 0.6829 |
| 0.8951 | 17.0 | 476 | 0.9426 | 0.6585 |
| 0.8137 | 18.0 | 504 | 0.9329 | 0.6098 |
| 0.8716 | 19.0 | 532 | 0.8386 | 0.7073 |
| 1.0116 | 20.0 | 560 | 0.7967 | 0.7073 |
| 0.8813 | 21.0 | 588 | 0.8366 | 0.7561 |
| 0.8835 | 22.0 | 616 | 1.0236 | 0.7073 |
| 0.8442 | 23.0 | 644 | 0.8949 | 0.5366 |
| 0.7804 | 24.0 | 672 | 0.9501 | 0.6829 |
| 0.7587 | 25.0 | 700 | 0.8384 | 0.6829 |
| 0.6929 | 26.0 | 728 | 0.9426 | 0.6098 |
| 0.7631 | 27.0 | 756 | 0.9144 | 0.6829 |
| 0.7309 | 28.0 | 784 | 0.8469 | 0.7317 |
| 0.6928 | 29.0 | 812 | 0.8593 | 0.6341 |
| 0.7758 | 30.0 | 840 | 0.8452 | 0.7561 |
| 0.6741 | 31.0 | 868 | 0.6468 | 0.7317 |
| 0.6387 | 32.0 | 896 | 0.8346 | 0.7317 |
| 0.6393 | 33.0 | 924 | 0.9678 | 0.7073 |
| 0.6388 | 34.0 | 952 | 0.9404 | 0.7317 |
| 0.5553 | 35.0 | 980 | 0.9751 | 0.6829 |
| 0.5488 | 36.0 | 1008 | 0.8789 | 0.6829 |
| 0.5749 | 37.0 | 1036 | 0.9082 | 0.6829 |
| 0.5631 | 38.0 | 1064 | 1.0681 | 0.6829 |
| 0.5744 | 39.0 | 1092 | 1.0203 | 0.6341 |
| 0.4507 | 40.0 | 1120 | 1.0247 | 0.6341 |
| 0.4706 | 41.0 | 1148 | 1.0726 | 0.6585 |
| 0.2984 | 42.0 | 1176 | 1.3959 | 0.6585 |
| 0.3096 | 43.0 | 1204 | 1.3220 | 0.6098 |
| 0.3345 | 44.0 | 1232 | 1.8003 | 0.6341 |
| 0.2779 | 45.0 | 1260 | 1.6742 | 0.6098 |
| 0.2186 | 46.0 | 1288 | 1.6974 | 0.6098 |
| 0.2223 | 47.0 | 1316 | 1.8144 | 0.6098 |
| 0.1904 | 48.0 | 1344 | 1.8092 | 0.6098 |
| 0.1904 | 49.0 | 1372 | 1.8130 | 0.6098 |
| 0.1964 | 50.0 | 1400 | 1.8130 | 0.6098 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0