smids_1x_deit_small_adamax_00001_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: 0.6794
- Accuracy: 0.8667
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: 1e-05
- 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 |
|---|---|---|---|---|
| 0.6229 | 1.0 | 75 | 0.5343 | 0.7933 |
| 0.4114 | 2.0 | 150 | 0.3999 | 0.8333 |
| 0.3246 | 3.0 | 225 | 0.3573 | 0.835 |
| 0.2962 | 4.0 | 300 | 0.3318 | 0.8633 |
| 0.2498 | 5.0 | 375 | 0.3315 | 0.86 |
| 0.1818 | 6.0 | 450 | 0.3035 | 0.87 |
| 0.1932 | 7.0 | 525 | 0.3061 | 0.875 |
| 0.1474 | 8.0 | 600 | 0.3041 | 0.87 |
| 0.0826 | 9.0 | 675 | 0.3169 | 0.87 |
| 0.081 | 10.0 | 750 | 0.3163 | 0.8667 |
| 0.0813 | 11.0 | 825 | 0.3296 | 0.87 |
| 0.0261 | 12.0 | 900 | 0.3349 | 0.87 |
| 0.0464 | 13.0 | 975 | 0.3659 | 0.8667 |
| 0.0215 | 14.0 | 1050 | 0.4099 | 0.87 |
| 0.027 | 15.0 | 1125 | 0.4201 | 0.87 |
| 0.0112 | 16.0 | 1200 | 0.4420 | 0.8717 |
| 0.0162 | 17.0 | 1275 | 0.4669 | 0.8733 |
| 0.016 | 18.0 | 1350 | 0.5089 | 0.87 |
| 0.0103 | 19.0 | 1425 | 0.4963 | 0.8717 |
| 0.0014 | 20.0 | 1500 | 0.5125 | 0.8733 |
| 0.0096 | 21.0 | 1575 | 0.5220 | 0.8733 |
| 0.0069 | 22.0 | 1650 | 0.5718 | 0.8617 |
| 0.0159 | 23.0 | 1725 | 0.5556 | 0.8717 |
| 0.0008 | 24.0 | 1800 | 0.5732 | 0.8667 |
| 0.0352 | 25.0 | 1875 | 0.5727 | 0.8683 |
| 0.0005 | 26.0 | 1950 | 0.5893 | 0.8683 |
| 0.0095 | 27.0 | 2025 | 0.6176 | 0.8667 |
| 0.0202 | 28.0 | 2100 | 0.5996 | 0.865 |
| 0.0236 | 29.0 | 2175 | 0.6069 | 0.8633 |
| 0.0103 | 30.0 | 2250 | 0.6179 | 0.865 |
| 0.0003 | 31.0 | 2325 | 0.6857 | 0.8633 |
| 0.0003 | 32.0 | 2400 | 0.6471 | 0.8667 |
| 0.0077 | 33.0 | 2475 | 0.6466 | 0.8667 |
| 0.0003 | 34.0 | 2550 | 0.6723 | 0.8633 |
| 0.0076 | 35.0 | 2625 | 0.6448 | 0.865 |
| 0.0002 | 36.0 | 2700 | 0.6372 | 0.87 |
| 0.0037 | 37.0 | 2775 | 0.6601 | 0.87 |
| 0.0002 | 38.0 | 2850 | 0.6572 | 0.8683 |
| 0.0002 | 39.0 | 2925 | 0.6720 | 0.8683 |
| 0.0002 | 40.0 | 3000 | 0.6642 | 0.8683 |
| 0.0072 | 41.0 | 3075 | 0.6568 | 0.8683 |
| 0.0002 | 42.0 | 3150 | 0.6668 | 0.8633 |
| 0.0069 | 43.0 | 3225 | 0.6577 | 0.8683 |
| 0.0002 | 44.0 | 3300 | 0.6738 | 0.8667 |
| 0.0042 | 45.0 | 3375 | 0.6742 | 0.865 |
| 0.0002 | 46.0 | 3450 | 0.6815 | 0.8667 |
| 0.0111 | 47.0 | 3525 | 0.6812 | 0.8667 |
| 0.0112 | 48.0 | 3600 | 0.6837 | 0.8683 |
| 0.0002 | 49.0 | 3675 | 0.6809 | 0.8667 |
| 0.0038 | 50.0 | 3750 | 0.6794 | 0.8667 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0