hushem_5x_deit_small_rms_0001_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: 1.0120
- Accuracy: 0.8444
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 |
|---|---|---|---|---|
| 0.7286 | 1.0 | 27 | 0.7253 | 0.7333 |
| 0.1535 | 2.0 | 54 | 0.7112 | 0.7333 |
| 0.0294 | 3.0 | 81 | 0.7268 | 0.7333 |
| 0.0088 | 4.0 | 108 | 0.6613 | 0.7778 |
| 0.002 | 5.0 | 135 | 0.6035 | 0.8 |
| 0.0011 | 6.0 | 162 | 0.5742 | 0.8444 |
| 0.0009 | 7.0 | 189 | 0.5650 | 0.8222 |
| 0.0006 | 8.0 | 216 | 0.5900 | 0.8444 |
| 0.0005 | 9.0 | 243 | 0.6288 | 0.8444 |
| 0.0004 | 10.0 | 270 | 0.6359 | 0.8444 |
| 0.0003 | 11.0 | 297 | 0.6555 | 0.8444 |
| 0.0003 | 12.0 | 324 | 0.6530 | 0.8444 |
| 0.0002 | 13.0 | 351 | 0.6624 | 0.8444 |
| 0.0002 | 14.0 | 378 | 0.6755 | 0.8444 |
| 0.0002 | 15.0 | 405 | 0.7042 | 0.8444 |
| 0.0002 | 16.0 | 432 | 0.7026 | 0.8444 |
| 0.0001 | 17.0 | 459 | 0.7218 | 0.8444 |
| 0.0001 | 18.0 | 486 | 0.7361 | 0.8444 |
| 0.0001 | 19.0 | 513 | 0.7477 | 0.8444 |
| 0.0001 | 20.0 | 540 | 0.7601 | 0.8444 |
| 0.0001 | 21.0 | 567 | 0.7574 | 0.8444 |
| 0.0001 | 22.0 | 594 | 0.7945 | 0.8444 |
| 0.0001 | 23.0 | 621 | 0.8035 | 0.8444 |
| 0.0001 | 24.0 | 648 | 0.8071 | 0.8444 |
| 0.0001 | 25.0 | 675 | 0.8201 | 0.8444 |
| 0.0 | 26.0 | 702 | 0.8263 | 0.8444 |
| 0.0 | 27.0 | 729 | 0.8414 | 0.8444 |
| 0.0 | 28.0 | 756 | 0.8501 | 0.8444 |
| 0.0 | 29.0 | 783 | 0.8628 | 0.8444 |
| 0.0 | 30.0 | 810 | 0.8618 | 0.8444 |
| 0.0 | 31.0 | 837 | 0.8808 | 0.8444 |
| 0.0 | 32.0 | 864 | 0.8998 | 0.8444 |
| 0.0 | 33.0 | 891 | 0.9078 | 0.8444 |
| 0.0 | 34.0 | 918 | 0.9219 | 0.8444 |
| 0.0 | 35.0 | 945 | 0.9258 | 0.8444 |
| 0.0 | 36.0 | 972 | 0.9286 | 0.8444 |
| 0.0 | 37.0 | 999 | 0.9396 | 0.8444 |
| 0.0 | 38.0 | 1026 | 0.9562 | 0.8444 |
| 0.0 | 39.0 | 1053 | 0.9622 | 0.8444 |
| 0.0 | 40.0 | 1080 | 0.9723 | 0.8444 |
| 0.0 | 41.0 | 1107 | 0.9820 | 0.8444 |
| 0.0 | 42.0 | 1134 | 0.9901 | 0.8444 |
| 0.0 | 43.0 | 1161 | 0.9932 | 0.8444 |
| 0.0 | 44.0 | 1188 | 0.9995 | 0.8444 |
| 0.0 | 45.0 | 1215 | 1.0030 | 0.8444 |
| 0.0 | 46.0 | 1242 | 1.0074 | 0.8444 |
| 0.0 | 47.0 | 1269 | 1.0103 | 0.8444 |
| 0.0 | 48.0 | 1296 | 1.0119 | 0.8444 |
| 0.0 | 49.0 | 1323 | 1.0120 | 0.8444 |
| 0.0 | 50.0 | 1350 | 1.0120 | 0.8444 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0