smids_1x_deit_small_adamax_0001_fold5
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: 0.6669
- Accuracy: 0.8983
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 |
|---|---|---|---|---|
| 0.4029 | 1.0 | 75 | 0.3535 | 0.845 |
| 0.3043 | 2.0 | 150 | 0.3006 | 0.88 |
| 0.16 | 3.0 | 225 | 0.2739 | 0.8917 |
| 0.1191 | 4.0 | 300 | 0.3321 | 0.8867 |
| 0.1124 | 5.0 | 375 | 0.4055 | 0.885 |
| 0.0475 | 6.0 | 450 | 0.4499 | 0.89 |
| 0.0249 | 7.0 | 525 | 0.4528 | 0.89 |
| 0.0478 | 8.0 | 600 | 0.5581 | 0.8883 |
| 0.0013 | 9.0 | 675 | 0.5854 | 0.8867 |
| 0.0299 | 10.0 | 750 | 0.6219 | 0.8833 |
| 0.0306 | 11.0 | 825 | 0.6492 | 0.8833 |
| 0.0013 | 12.0 | 900 | 0.5930 | 0.89 |
| 0.006 | 13.0 | 975 | 0.6079 | 0.8933 |
| 0.018 | 14.0 | 1050 | 0.6720 | 0.885 |
| 0.009 | 15.0 | 1125 | 0.7215 | 0.8917 |
| 0.0001 | 16.0 | 1200 | 0.5817 | 0.8967 |
| 0.0052 | 17.0 | 1275 | 0.5765 | 0.8983 |
| 0.0097 | 18.0 | 1350 | 0.5975 | 0.8967 |
| 0.0033 | 19.0 | 1425 | 0.6180 | 0.8983 |
| 0.0 | 20.0 | 1500 | 0.6163 | 0.9 |
| 0.0043 | 21.0 | 1575 | 0.5951 | 0.9 |
| 0.0041 | 22.0 | 1650 | 0.6227 | 0.8983 |
| 0.0044 | 23.0 | 1725 | 0.6229 | 0.895 |
| 0.0 | 24.0 | 1800 | 0.6143 | 0.9 |
| 0.0068 | 25.0 | 1875 | 0.6320 | 0.9017 |
| 0.0 | 26.0 | 1950 | 0.6233 | 0.8983 |
| 0.0062 | 27.0 | 2025 | 0.6313 | 0.8967 |
| 0.0083 | 28.0 | 2100 | 0.6337 | 0.8983 |
| 0.0062 | 29.0 | 2175 | 0.6268 | 0.9033 |
| 0.0034 | 30.0 | 2250 | 0.6363 | 0.9 |
| 0.0 | 31.0 | 2325 | 0.6524 | 0.8983 |
| 0.0 | 32.0 | 2400 | 0.6386 | 0.9 |
| 0.0057 | 33.0 | 2475 | 0.6534 | 0.9 |
| 0.0 | 34.0 | 2550 | 0.6542 | 0.8967 |
| 0.0048 | 35.0 | 2625 | 0.6498 | 0.9017 |
| 0.0 | 36.0 | 2700 | 0.6512 | 0.9033 |
| 0.003 | 37.0 | 2775 | 0.6532 | 0.9017 |
| 0.0 | 38.0 | 2850 | 0.6543 | 0.8967 |
| 0.0 | 39.0 | 2925 | 0.6569 | 0.8967 |
| 0.0 | 40.0 | 3000 | 0.6561 | 0.9 |
| 0.0028 | 41.0 | 3075 | 0.6588 | 0.8967 |
| 0.0 | 42.0 | 3150 | 0.6601 | 0.9017 |
| 0.007 | 43.0 | 3225 | 0.6640 | 0.9017 |
| 0.0 | 44.0 | 3300 | 0.6634 | 0.8967 |
| 0.0024 | 45.0 | 3375 | 0.6647 | 0.9033 |
| 0.0 | 46.0 | 3450 | 0.6647 | 0.8983 |
| 0.0051 | 47.0 | 3525 | 0.6655 | 0.8983 |
| 0.0049 | 48.0 | 3600 | 0.6656 | 0.8983 |
| 0.0 | 49.0 | 3675 | 0.6665 | 0.8967 |
| 0.0043 | 50.0 | 3750 | 0.6669 | 0.8983 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0