smids_3x_deit_tiny_rms_001_fold2
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: 1.9828
- Accuracy: 0.8037
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 |
|---|---|---|---|---|
| 0.8903 | 1.0 | 225 | 0.8221 | 0.5707 |
| 0.8376 | 2.0 | 450 | 0.7917 | 0.6023 |
| 0.7073 | 3.0 | 675 | 0.7643 | 0.5790 |
| 0.7158 | 4.0 | 900 | 0.7590 | 0.6506 |
| 0.6892 | 5.0 | 1125 | 0.7293 | 0.6506 |
| 0.6685 | 6.0 | 1350 | 0.6489 | 0.6755 |
| 0.606 | 7.0 | 1575 | 0.5908 | 0.7471 |
| 0.6652 | 8.0 | 1800 | 0.5586 | 0.7537 |
| 0.6311 | 9.0 | 2025 | 0.5971 | 0.7504 |
| 0.5645 | 10.0 | 2250 | 0.5700 | 0.7571 |
| 0.6575 | 11.0 | 2475 | 0.6329 | 0.7338 |
| 0.5727 | 12.0 | 2700 | 0.5598 | 0.7521 |
| 0.5704 | 13.0 | 2925 | 0.5359 | 0.7754 |
| 0.5215 | 14.0 | 3150 | 0.5587 | 0.7571 |
| 0.583 | 15.0 | 3375 | 0.5430 | 0.7787 |
| 0.5848 | 16.0 | 3600 | 0.5789 | 0.7255 |
| 0.5726 | 17.0 | 3825 | 0.5236 | 0.7854 |
| 0.489 | 18.0 | 4050 | 0.4992 | 0.8003 |
| 0.4753 | 19.0 | 4275 | 0.5863 | 0.7621 |
| 0.5263 | 20.0 | 4500 | 0.5219 | 0.7704 |
| 0.4735 | 21.0 | 4725 | 0.4900 | 0.7970 |
| 0.5712 | 22.0 | 4950 | 0.5240 | 0.8070 |
| 0.5071 | 23.0 | 5175 | 0.5142 | 0.7854 |
| 0.5079 | 24.0 | 5400 | 0.5416 | 0.7887 |
| 0.4923 | 25.0 | 5625 | 0.5063 | 0.8053 |
| 0.3927 | 26.0 | 5850 | 0.5633 | 0.7870 |
| 0.3745 | 27.0 | 6075 | 0.6375 | 0.7621 |
| 0.3879 | 28.0 | 6300 | 0.5787 | 0.7787 |
| 0.3817 | 29.0 | 6525 | 0.5796 | 0.7887 |
| 0.3478 | 30.0 | 6750 | 0.5474 | 0.8070 |
| 0.3866 | 31.0 | 6975 | 0.5776 | 0.7920 |
| 0.3993 | 32.0 | 7200 | 0.5703 | 0.8103 |
| 0.2926 | 33.0 | 7425 | 0.6089 | 0.7970 |
| 0.2772 | 34.0 | 7650 | 0.6185 | 0.8020 |
| 0.2754 | 35.0 | 7875 | 0.6777 | 0.7854 |
| 0.2402 | 36.0 | 8100 | 0.7422 | 0.7870 |
| 0.2412 | 37.0 | 8325 | 0.7873 | 0.7754 |
| 0.2165 | 38.0 | 8550 | 0.8318 | 0.8070 |
| 0.1829 | 39.0 | 8775 | 0.8522 | 0.7920 |
| 0.1669 | 40.0 | 9000 | 0.9913 | 0.7903 |
| 0.135 | 41.0 | 9225 | 1.0029 | 0.8003 |
| 0.1443 | 42.0 | 9450 | 1.0999 | 0.7987 |
| 0.0893 | 43.0 | 9675 | 1.2541 | 0.8020 |
| 0.0533 | 44.0 | 9900 | 1.3854 | 0.8003 |
| 0.065 | 45.0 | 10125 | 1.6713 | 0.7754 |
| 0.0332 | 46.0 | 10350 | 1.6089 | 0.8003 |
| 0.0177 | 47.0 | 10575 | 1.8304 | 0.7920 |
| 0.0017 | 48.0 | 10800 | 1.9277 | 0.7937 |
| 0.0154 | 49.0 | 11025 | 1.9904 | 0.8053 |
| 0.0079 | 50.0 | 11250 | 1.9828 | 0.8037 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2