smids_3x_deit_tiny_rms_0001_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: 0.9250
- Accuracy: 0.8933
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 |
|---|---|---|---|---|
| 0.4046 | 1.0 | 225 | 0.3353 | 0.855 |
| 0.2238 | 2.0 | 450 | 0.2977 | 0.8967 |
| 0.2029 | 3.0 | 675 | 0.3292 | 0.8717 |
| 0.1861 | 4.0 | 900 | 0.3918 | 0.8633 |
| 0.097 | 5.0 | 1125 | 0.5728 | 0.87 |
| 0.0763 | 6.0 | 1350 | 0.3602 | 0.8867 |
| 0.1211 | 7.0 | 1575 | 0.3953 | 0.9067 |
| 0.0628 | 8.0 | 1800 | 0.5619 | 0.8917 |
| 0.1484 | 9.0 | 2025 | 0.5750 | 0.88 |
| 0.0452 | 10.0 | 2250 | 0.6659 | 0.89 |
| 0.0229 | 11.0 | 2475 | 0.6256 | 0.8933 |
| 0.0617 | 12.0 | 2700 | 0.7075 | 0.87 |
| 0.0553 | 13.0 | 2925 | 0.6972 | 0.8983 |
| 0.0308 | 14.0 | 3150 | 0.6494 | 0.8983 |
| 0.0312 | 15.0 | 3375 | 0.6866 | 0.9 |
| 0.011 | 16.0 | 3600 | 0.7253 | 0.895 |
| 0.0983 | 17.0 | 3825 | 0.7035 | 0.8933 |
| 0.0451 | 18.0 | 4050 | 0.8265 | 0.8933 |
| 0.0418 | 19.0 | 4275 | 0.8696 | 0.8767 |
| 0.0469 | 20.0 | 4500 | 0.6273 | 0.9133 |
| 0.0203 | 21.0 | 4725 | 0.7939 | 0.895 |
| 0.0102 | 22.0 | 4950 | 0.7226 | 0.8967 |
| 0.0005 | 23.0 | 5175 | 0.8708 | 0.8933 |
| 0.0229 | 24.0 | 5400 | 0.9025 | 0.89 |
| 0.0344 | 25.0 | 5625 | 0.7685 | 0.9033 |
| 0.0016 | 26.0 | 5850 | 0.7805 | 0.9067 |
| 0.0048 | 27.0 | 6075 | 0.7684 | 0.9033 |
| 0.0028 | 28.0 | 6300 | 0.8595 | 0.8933 |
| 0.0098 | 29.0 | 6525 | 0.8847 | 0.8983 |
| 0.0002 | 30.0 | 6750 | 0.8488 | 0.8917 |
| 0.0 | 31.0 | 6975 | 0.9022 | 0.8883 |
| 0.0 | 32.0 | 7200 | 0.8024 | 0.895 |
| 0.0047 | 33.0 | 7425 | 0.8208 | 0.8933 |
| 0.0001 | 34.0 | 7650 | 0.9019 | 0.9017 |
| 0.0033 | 35.0 | 7875 | 0.8774 | 0.8883 |
| 0.0 | 36.0 | 8100 | 0.8642 | 0.885 |
| 0.0189 | 37.0 | 8325 | 0.8309 | 0.8983 |
| 0.0 | 38.0 | 8550 | 0.9322 | 0.89 |
| 0.0 | 39.0 | 8775 | 0.9453 | 0.8933 |
| 0.0 | 40.0 | 9000 | 0.9411 | 0.89 |
| 0.0 | 41.0 | 9225 | 0.9468 | 0.8917 |
| 0.0 | 42.0 | 9450 | 0.9584 | 0.8967 |
| 0.003 | 43.0 | 9675 | 0.9469 | 0.8917 |
| 0.0 | 44.0 | 9900 | 0.9339 | 0.8917 |
| 0.0 | 45.0 | 10125 | 0.9259 | 0.89 |
| 0.0 | 46.0 | 10350 | 0.9294 | 0.8917 |
| 0.0 | 47.0 | 10575 | 0.9214 | 0.8917 |
| 0.0 | 48.0 | 10800 | 0.9235 | 0.8917 |
| 0.0 | 49.0 | 11025 | 0.9243 | 0.8933 |
| 0.0 | 50.0 | 11250 | 0.9250 | 0.8933 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2