hushem_5x_deit_tiny_rms_001_fold3
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: 0.9822
- Accuracy: 0.6977
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 |
|---|---|---|---|---|
| 1.9654 | 1.0 | 28 | 2.6283 | 0.2558 |
| 1.5373 | 2.0 | 56 | 2.1986 | 0.2558 |
| 1.4995 | 3.0 | 84 | 1.6042 | 0.2558 |
| 1.4355 | 4.0 | 112 | 1.6838 | 0.2558 |
| 1.4504 | 5.0 | 140 | 1.5958 | 0.2326 |
| 1.4097 | 6.0 | 168 | 1.4017 | 0.2791 |
| 1.4297 | 7.0 | 196 | 1.5664 | 0.2326 |
| 1.4137 | 8.0 | 224 | 1.4485 | 0.2558 |
| 1.3745 | 9.0 | 252 | 1.2801 | 0.4186 |
| 1.2697 | 10.0 | 280 | 1.2764 | 0.3256 |
| 1.1321 | 11.0 | 308 | 1.5227 | 0.3256 |
| 1.1096 | 12.0 | 336 | 1.2384 | 0.3953 |
| 1.0727 | 13.0 | 364 | 1.1395 | 0.4884 |
| 1.0037 | 14.0 | 392 | 1.3856 | 0.3953 |
| 1.0402 | 15.0 | 420 | 1.1134 | 0.5116 |
| 1.0378 | 16.0 | 448 | 1.2506 | 0.4419 |
| 0.958 | 17.0 | 476 | 1.1080 | 0.4651 |
| 0.9953 | 18.0 | 504 | 1.2467 | 0.4884 |
| 0.9958 | 19.0 | 532 | 1.0807 | 0.5814 |
| 0.9467 | 20.0 | 560 | 1.1055 | 0.4186 |
| 0.9535 | 21.0 | 588 | 1.1974 | 0.5116 |
| 0.9184 | 22.0 | 616 | 1.1307 | 0.4186 |
| 0.9252 | 23.0 | 644 | 1.0833 | 0.5116 |
| 0.8662 | 24.0 | 672 | 1.1623 | 0.5349 |
| 0.8421 | 25.0 | 700 | 0.9575 | 0.5814 |
| 0.8602 | 26.0 | 728 | 1.1189 | 0.5581 |
| 0.923 | 27.0 | 756 | 1.3369 | 0.5116 |
| 0.8226 | 28.0 | 784 | 1.0806 | 0.6279 |
| 0.8183 | 29.0 | 812 | 1.2385 | 0.4186 |
| 0.801 | 30.0 | 840 | 0.8599 | 0.6744 |
| 0.7516 | 31.0 | 868 | 1.3471 | 0.4884 |
| 0.7555 | 32.0 | 896 | 1.0726 | 0.5814 |
| 0.7219 | 33.0 | 924 | 0.8253 | 0.6977 |
| 0.7341 | 34.0 | 952 | 0.9501 | 0.6744 |
| 0.7645 | 35.0 | 980 | 0.9024 | 0.6512 |
| 0.6775 | 36.0 | 1008 | 0.6982 | 0.6977 |
| 0.6942 | 37.0 | 1036 | 0.8138 | 0.6744 |
| 0.6421 | 38.0 | 1064 | 1.1443 | 0.6279 |
| 0.6108 | 39.0 | 1092 | 0.7170 | 0.6977 |
| 0.7595 | 40.0 | 1120 | 0.7538 | 0.6744 |
| 0.6409 | 41.0 | 1148 | 1.2761 | 0.5581 |
| 0.6168 | 42.0 | 1176 | 1.0481 | 0.6279 |
| 0.5155 | 43.0 | 1204 | 0.7647 | 0.7209 |
| 0.5674 | 44.0 | 1232 | 0.9942 | 0.6744 |
| 0.5763 | 45.0 | 1260 | 0.8142 | 0.6977 |
| 0.4817 | 46.0 | 1288 | 0.8614 | 0.6977 |
| 0.4723 | 47.0 | 1316 | 1.0386 | 0.6512 |
| 0.4863 | 48.0 | 1344 | 0.9689 | 0.7209 |
| 0.4909 | 49.0 | 1372 | 0.9822 | 0.6977 |
| 0.4786 | 50.0 | 1400 | 0.9822 | 0.6977 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0