smids_3x_deit_small_adamax_0001_fold4
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: 1.1878
- Accuracy: 0.89
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: 0.0001
- 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 |
|---|---|---|---|---|
| 0.358 | 1.0 | 225 | 0.3442 | 0.86 |
| 0.0876 | 2.0 | 450 | 0.3933 | 0.8767 |
| 0.0775 | 3.0 | 675 | 0.6722 | 0.8367 |
| 0.035 | 4.0 | 900 | 0.5573 | 0.875 |
| 0.0884 | 5.0 | 1125 | 0.6478 | 0.8783 |
| 0.0046 | 6.0 | 1350 | 0.6948 | 0.8817 |
| 0.0172 | 7.0 | 1575 | 0.8233 | 0.8867 |
| 0.0024 | 8.0 | 1800 | 0.7176 | 0.8833 |
| 0.0123 | 9.0 | 2025 | 0.8547 | 0.875 |
| 0.0046 | 10.0 | 2250 | 0.9722 | 0.87 |
| 0.0001 | 11.0 | 2475 | 0.8743 | 0.89 |
| 0.0097 | 12.0 | 2700 | 1.0010 | 0.8783 |
| 0.0001 | 13.0 | 2925 | 0.9311 | 0.88 |
| 0.0002 | 14.0 | 3150 | 0.9762 | 0.8833 |
| 0.0029 | 15.0 | 3375 | 0.9797 | 0.8867 |
| 0.0 | 16.0 | 3600 | 1.0447 | 0.8867 |
| 0.0 | 17.0 | 3825 | 1.0609 | 0.8817 |
| 0.0 | 18.0 | 4050 | 1.0141 | 0.8867 |
| 0.0 | 19.0 | 4275 | 1.0239 | 0.8883 |
| 0.0 | 20.0 | 4500 | 1.0236 | 0.8867 |
| 0.0 | 21.0 | 4725 | 1.0764 | 0.8883 |
| 0.0 | 22.0 | 4950 | 1.0464 | 0.8883 |
| 0.0 | 23.0 | 5175 | 1.0429 | 0.8917 |
| 0.0 | 24.0 | 5400 | 1.0507 | 0.8883 |
| 0.0 | 25.0 | 5625 | 1.0648 | 0.8883 |
| 0.0 | 26.0 | 5850 | 1.0647 | 0.8867 |
| 0.0 | 27.0 | 6075 | 1.0952 | 0.8883 |
| 0.0 | 28.0 | 6300 | 1.1057 | 0.8867 |
| 0.0 | 29.0 | 6525 | 1.0765 | 0.8917 |
| 0.0 | 30.0 | 6750 | 1.0978 | 0.89 |
| 0.0 | 31.0 | 6975 | 1.1017 | 0.8917 |
| 0.0 | 32.0 | 7200 | 1.1042 | 0.8917 |
| 0.0 | 33.0 | 7425 | 1.1208 | 0.8917 |
| 0.0 | 34.0 | 7650 | 1.1187 | 0.89 |
| 0.0 | 35.0 | 7875 | 1.1334 | 0.89 |
| 0.0 | 36.0 | 8100 | 1.1314 | 0.8883 |
| 0.0 | 37.0 | 8325 | 1.1340 | 0.89 |
| 0.0 | 38.0 | 8550 | 1.1418 | 0.8883 |
| 0.0 | 39.0 | 8775 | 1.1558 | 0.8883 |
| 0.0 | 40.0 | 9000 | 1.1604 | 0.89 |
| 0.0 | 41.0 | 9225 | 1.1596 | 0.8867 |
| 0.0028 | 42.0 | 9450 | 1.1661 | 0.8867 |
| 0.0 | 43.0 | 9675 | 1.1764 | 0.8883 |
| 0.0026 | 44.0 | 9900 | 1.1730 | 0.8867 |
| 0.0 | 45.0 | 10125 | 1.1807 | 0.89 |
| 0.0 | 46.0 | 10350 | 1.1827 | 0.8883 |
| 0.0 | 47.0 | 10575 | 1.1851 | 0.89 |
| 0.0 | 48.0 | 10800 | 1.1866 | 0.89 |
| 0.0 | 49.0 | 11025 | 1.1856 | 0.8883 |
| 0.0 | 50.0 | 11250 | 1.1878 | 0.89 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2