smids_3x_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: 1.6588
- Accuracy: 0.8167
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 |
|---|---|---|---|---|
| 0.972 | 1.0 | 225 | 1.0247 | 0.3783 |
| 0.8568 | 2.0 | 450 | 0.8274 | 0.5533 |
| 0.8081 | 3.0 | 675 | 0.8040 | 0.5483 |
| 0.7625 | 4.0 | 900 | 0.7546 | 0.5933 |
| 0.7576 | 5.0 | 1125 | 0.7605 | 0.64 |
| 0.6884 | 6.0 | 1350 | 0.7873 | 0.5983 |
| 0.766 | 7.0 | 1575 | 0.7270 | 0.6583 |
| 0.6978 | 8.0 | 1800 | 0.7176 | 0.64 |
| 0.6732 | 9.0 | 2025 | 0.7347 | 0.645 |
| 0.6839 | 10.0 | 2250 | 0.7289 | 0.6267 |
| 0.6569 | 11.0 | 2475 | 0.6542 | 0.7183 |
| 0.6318 | 12.0 | 2700 | 0.6186 | 0.7283 |
| 0.6796 | 13.0 | 2925 | 0.6663 | 0.71 |
| 0.6092 | 14.0 | 3150 | 0.6155 | 0.7117 |
| 0.6242 | 15.0 | 3375 | 0.6625 | 0.6967 |
| 0.5314 | 16.0 | 3600 | 0.5775 | 0.7533 |
| 0.5564 | 17.0 | 3825 | 0.5848 | 0.7533 |
| 0.5755 | 18.0 | 4050 | 0.5751 | 0.7583 |
| 0.5677 | 19.0 | 4275 | 0.5731 | 0.7617 |
| 0.5761 | 20.0 | 4500 | 0.5204 | 0.785 |
| 0.4524 | 21.0 | 4725 | 0.5722 | 0.75 |
| 0.4782 | 22.0 | 4950 | 0.5385 | 0.7733 |
| 0.4908 | 23.0 | 5175 | 0.5176 | 0.7933 |
| 0.5195 | 24.0 | 5400 | 0.5242 | 0.7917 |
| 0.4871 | 25.0 | 5625 | 0.5298 | 0.7983 |
| 0.5293 | 26.0 | 5850 | 0.5066 | 0.8 |
| 0.504 | 27.0 | 6075 | 0.4969 | 0.81 |
| 0.4467 | 28.0 | 6300 | 0.5630 | 0.79 |
| 0.4177 | 29.0 | 6525 | 0.5247 | 0.8067 |
| 0.3722 | 30.0 | 6750 | 0.5359 | 0.8117 |
| 0.3286 | 31.0 | 6975 | 0.5623 | 0.795 |
| 0.3205 | 32.0 | 7200 | 0.5594 | 0.8017 |
| 0.3627 | 33.0 | 7425 | 0.5968 | 0.815 |
| 0.2799 | 34.0 | 7650 | 0.5562 | 0.825 |
| 0.2664 | 35.0 | 7875 | 0.6268 | 0.81 |
| 0.2603 | 36.0 | 8100 | 0.6102 | 0.82 |
| 0.2382 | 37.0 | 8325 | 0.6448 | 0.8083 |
| 0.1999 | 38.0 | 8550 | 0.7396 | 0.825 |
| 0.1413 | 39.0 | 8775 | 0.7329 | 0.8167 |
| 0.1906 | 40.0 | 9000 | 0.8804 | 0.81 |
| 0.1179 | 41.0 | 9225 | 0.7998 | 0.84 |
| 0.0965 | 42.0 | 9450 | 0.9317 | 0.8217 |
| 0.0987 | 43.0 | 9675 | 0.9015 | 0.825 |
| 0.1035 | 44.0 | 9900 | 1.1023 | 0.8083 |
| 0.0347 | 45.0 | 10125 | 1.2315 | 0.82 |
| 0.054 | 46.0 | 10350 | 1.2317 | 0.8083 |
| 0.014 | 47.0 | 10575 | 1.4229 | 0.82 |
| 0.0044 | 48.0 | 10800 | 1.5732 | 0.8217 |
| 0.0012 | 49.0 | 11025 | 1.6140 | 0.8183 |
| 0.0003 | 50.0 | 11250 | 1.6588 | 0.8167 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2