hushem_40x_deit_small_rms_001_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: 7.2597
- Accuracy: 0.4222
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 |
|---|---|---|---|---|
| 1.3825 | 1.0 | 215 | 1.4688 | 0.2667 |
| 1.3534 | 2.0 | 430 | 1.4553 | 0.3778 |
| 0.9498 | 3.0 | 645 | 1.8460 | 0.3333 |
| 0.7874 | 4.0 | 860 | 1.0992 | 0.4444 |
| 0.6519 | 5.0 | 1075 | 1.5864 | 0.4222 |
| 0.6238 | 6.0 | 1290 | 1.5678 | 0.4444 |
| 0.6712 | 7.0 | 1505 | 1.5837 | 0.3778 |
| 0.6234 | 8.0 | 1720 | 1.4844 | 0.3778 |
| 0.6842 | 9.0 | 1935 | 1.4360 | 0.4 |
| 0.5244 | 10.0 | 2150 | 1.9225 | 0.3778 |
| 0.5422 | 11.0 | 2365 | 1.4512 | 0.4667 |
| 0.4482 | 12.0 | 2580 | 2.2789 | 0.3556 |
| 0.5899 | 13.0 | 2795 | 1.6124 | 0.4222 |
| 0.4227 | 14.0 | 3010 | 1.8210 | 0.4444 |
| 0.4862 | 15.0 | 3225 | 1.4215 | 0.4667 |
| 0.4615 | 16.0 | 3440 | 2.1496 | 0.3778 |
| 0.6895 | 17.0 | 3655 | 1.7698 | 0.4667 |
| 0.3741 | 18.0 | 3870 | 2.6905 | 0.3556 |
| 0.3762 | 19.0 | 4085 | 2.4546 | 0.4222 |
| 0.3383 | 20.0 | 4300 | 2.0176 | 0.3778 |
| 0.3622 | 21.0 | 4515 | 2.9706 | 0.4 |
| 0.3284 | 22.0 | 4730 | 2.9396 | 0.4 |
| 0.2403 | 23.0 | 4945 | 2.3459 | 0.4889 |
| 0.345 | 24.0 | 5160 | 3.1195 | 0.4222 |
| 0.3045 | 25.0 | 5375 | 2.4187 | 0.4667 |
| 0.2936 | 26.0 | 5590 | 2.9167 | 0.3556 |
| 0.249 | 27.0 | 5805 | 2.5521 | 0.4667 |
| 0.2161 | 28.0 | 6020 | 3.7842 | 0.3778 |
| 0.2382 | 29.0 | 6235 | 3.0584 | 0.4 |
| 0.1225 | 30.0 | 6450 | 4.4557 | 0.4 |
| 0.2075 | 31.0 | 6665 | 4.7131 | 0.3111 |
| 0.1575 | 32.0 | 6880 | 3.8714 | 0.3556 |
| 0.1516 | 33.0 | 7095 | 4.5510 | 0.4 |
| 0.1231 | 34.0 | 7310 | 5.0636 | 0.3778 |
| 0.0943 | 35.0 | 7525 | 4.2212 | 0.4 |
| 0.0741 | 36.0 | 7740 | 4.4947 | 0.4 |
| 0.0582 | 37.0 | 7955 | 4.8808 | 0.4222 |
| 0.0412 | 38.0 | 8170 | 5.2254 | 0.3778 |
| 0.0508 | 39.0 | 8385 | 5.2558 | 0.3556 |
| 0.0566 | 40.0 | 8600 | 5.9529 | 0.3556 |
| 0.0397 | 41.0 | 8815 | 5.9087 | 0.3333 |
| 0.0462 | 42.0 | 9030 | 6.2634 | 0.4444 |
| 0.0245 | 43.0 | 9245 | 6.0294 | 0.4222 |
| 0.0398 | 44.0 | 9460 | 6.9015 | 0.4222 |
| 0.0182 | 45.0 | 9675 | 5.5112 | 0.4667 |
| 0.0162 | 46.0 | 9890 | 6.0476 | 0.4889 |
| 0.0028 | 47.0 | 10105 | 6.5416 | 0.4667 |
| 0.0087 | 48.0 | 10320 | 6.8964 | 0.4444 |
| 0.0011 | 49.0 | 10535 | 7.0908 | 0.4222 |
| 0.0007 | 50.0 | 10750 | 7.2597 | 0.4222 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2