smids_1x_deit_small_adamax_0001_fold2
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: 0.8040
- Accuracy: 0.8769
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 |
|---|---|---|---|---|
| 0.3172 | 1.0 | 75 | 0.3354 | 0.8636 |
| 0.1909 | 2.0 | 150 | 0.2950 | 0.8785 |
| 0.1696 | 3.0 | 225 | 0.3030 | 0.8902 |
| 0.1088 | 4.0 | 300 | 0.3512 | 0.8735 |
| 0.0828 | 5.0 | 375 | 0.3819 | 0.8719 |
| 0.0816 | 6.0 | 450 | 0.5235 | 0.8735 |
| 0.017 | 7.0 | 525 | 0.4625 | 0.8802 |
| 0.0096 | 8.0 | 600 | 0.6732 | 0.8536 |
| 0.0297 | 9.0 | 675 | 0.5099 | 0.8852 |
| 0.0129 | 10.0 | 750 | 0.6168 | 0.8819 |
| 0.0004 | 11.0 | 825 | 0.6434 | 0.8769 |
| 0.0003 | 12.0 | 900 | 0.6532 | 0.8752 |
| 0.0115 | 13.0 | 975 | 0.7781 | 0.8669 |
| 0.0025 | 14.0 | 1050 | 0.6839 | 0.8735 |
| 0.0031 | 15.0 | 1125 | 0.6481 | 0.8802 |
| 0.0037 | 16.0 | 1200 | 0.7018 | 0.8719 |
| 0.0001 | 17.0 | 1275 | 0.6843 | 0.8752 |
| 0.0057 | 18.0 | 1350 | 0.6963 | 0.8819 |
| 0.0001 | 19.0 | 1425 | 0.6873 | 0.8802 |
| 0.0036 | 20.0 | 1500 | 0.7059 | 0.8785 |
| 0.0001 | 21.0 | 1575 | 0.7123 | 0.8819 |
| 0.0 | 22.0 | 1650 | 0.7298 | 0.8785 |
| 0.0 | 23.0 | 1725 | 0.7182 | 0.8785 |
| 0.0 | 24.0 | 1800 | 0.7389 | 0.8752 |
| 0.0 | 25.0 | 1875 | 0.7283 | 0.8785 |
| 0.0 | 26.0 | 1950 | 0.7283 | 0.8802 |
| 0.0038 | 27.0 | 2025 | 0.7334 | 0.8819 |
| 0.0034 | 28.0 | 2100 | 0.7554 | 0.8735 |
| 0.0022 | 29.0 | 2175 | 0.7526 | 0.8752 |
| 0.0035 | 30.0 | 2250 | 0.7536 | 0.8769 |
| 0.0026 | 31.0 | 2325 | 0.7690 | 0.8719 |
| 0.0 | 32.0 | 2400 | 0.7598 | 0.8769 |
| 0.0 | 33.0 | 2475 | 0.7644 | 0.8752 |
| 0.0 | 34.0 | 2550 | 0.7770 | 0.8769 |
| 0.0081 | 35.0 | 2625 | 0.7696 | 0.8735 |
| 0.0 | 36.0 | 2700 | 0.7747 | 0.8735 |
| 0.0 | 37.0 | 2775 | 0.7776 | 0.8735 |
| 0.0 | 38.0 | 2850 | 0.7800 | 0.8735 |
| 0.0022 | 39.0 | 2925 | 0.7797 | 0.8735 |
| 0.0 | 40.0 | 3000 | 0.7884 | 0.8752 |
| 0.0028 | 41.0 | 3075 | 0.7926 | 0.8785 |
| 0.0 | 42.0 | 3150 | 0.7941 | 0.8769 |
| 0.0025 | 43.0 | 3225 | 0.7995 | 0.8752 |
| 0.0026 | 44.0 | 3300 | 0.7969 | 0.8752 |
| 0.0 | 45.0 | 3375 | 0.7932 | 0.8785 |
| 0.0 | 46.0 | 3450 | 0.8020 | 0.8752 |
| 0.0023 | 47.0 | 3525 | 0.8011 | 0.8702 |
| 0.0 | 48.0 | 3600 | 0.8043 | 0.8769 |
| 0.0022 | 49.0 | 3675 | 0.8040 | 0.8769 |
| 0.0022 | 50.0 | 3750 | 0.8040 | 0.8769 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0