hushem_5x_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: 0.2927
- Accuracy: 0.9048
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 |
|---|---|---|---|---|
| 0.7792 | 1.0 | 28 | 0.8167 | 0.6667 |
| 0.1236 | 2.0 | 56 | 0.4141 | 0.8333 |
| 0.0368 | 3.0 | 84 | 0.3101 | 0.8810 |
| 0.0033 | 4.0 | 112 | 0.3632 | 0.8571 |
| 0.0009 | 5.0 | 140 | 0.3885 | 0.8571 |
| 0.0009 | 6.0 | 168 | 0.2838 | 0.9286 |
| 0.0003 | 7.0 | 196 | 0.2908 | 0.9048 |
| 0.0003 | 8.0 | 224 | 0.2764 | 0.9286 |
| 0.0002 | 9.0 | 252 | 0.2768 | 0.9048 |
| 0.0002 | 10.0 | 280 | 0.2786 | 0.9048 |
| 0.0002 | 11.0 | 308 | 0.2796 | 0.9048 |
| 0.0001 | 12.0 | 336 | 0.2810 | 0.9048 |
| 0.0001 | 13.0 | 364 | 0.2830 | 0.9048 |
| 0.0001 | 14.0 | 392 | 0.2832 | 0.9048 |
| 0.0001 | 15.0 | 420 | 0.2837 | 0.9048 |
| 0.0001 | 16.0 | 448 | 0.2819 | 0.9048 |
| 0.0001 | 17.0 | 476 | 0.2839 | 0.9048 |
| 0.0001 | 18.0 | 504 | 0.2843 | 0.9048 |
| 0.0001 | 19.0 | 532 | 0.2842 | 0.9048 |
| 0.0001 | 20.0 | 560 | 0.2846 | 0.9048 |
| 0.0001 | 21.0 | 588 | 0.2846 | 0.9048 |
| 0.0001 | 22.0 | 616 | 0.2856 | 0.9048 |
| 0.0001 | 23.0 | 644 | 0.2866 | 0.9048 |
| 0.0001 | 24.0 | 672 | 0.2880 | 0.9048 |
| 0.0001 | 25.0 | 700 | 0.2878 | 0.9048 |
| 0.0001 | 26.0 | 728 | 0.2882 | 0.9048 |
| 0.0001 | 27.0 | 756 | 0.2888 | 0.9048 |
| 0.0001 | 28.0 | 784 | 0.2897 | 0.9048 |
| 0.0001 | 29.0 | 812 | 0.2897 | 0.9048 |
| 0.0001 | 30.0 | 840 | 0.2901 | 0.9048 |
| 0.0001 | 31.0 | 868 | 0.2903 | 0.9048 |
| 0.0001 | 32.0 | 896 | 0.2907 | 0.9048 |
| 0.0001 | 33.0 | 924 | 0.2909 | 0.9048 |
| 0.0 | 34.0 | 952 | 0.2903 | 0.9048 |
| 0.0001 | 35.0 | 980 | 0.2907 | 0.9048 |
| 0.0001 | 36.0 | 1008 | 0.2909 | 0.9048 |
| 0.0 | 37.0 | 1036 | 0.2912 | 0.9048 |
| 0.0 | 38.0 | 1064 | 0.2916 | 0.9048 |
| 0.0 | 39.0 | 1092 | 0.2918 | 0.9048 |
| 0.0 | 40.0 | 1120 | 0.2919 | 0.9048 |
| 0.0 | 41.0 | 1148 | 0.2921 | 0.9048 |
| 0.0 | 42.0 | 1176 | 0.2924 | 0.9048 |
| 0.0 | 43.0 | 1204 | 0.2926 | 0.9048 |
| 0.0 | 44.0 | 1232 | 0.2925 | 0.9048 |
| 0.0 | 45.0 | 1260 | 0.2927 | 0.9048 |
| 0.0 | 46.0 | 1288 | 0.2927 | 0.9048 |
| 0.0 | 47.0 | 1316 | 0.2927 | 0.9048 |
| 0.0 | 48.0 | 1344 | 0.2927 | 0.9048 |
| 0.0 | 49.0 | 1372 | 0.2927 | 0.9048 |
| 0.0 | 50.0 | 1400 | 0.2927 | 0.9048 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0