hushem_1x_deit_small_adamax_001_fold3
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: 3.7699
- Accuracy: 0.4651
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: 0.001
- 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 |
|---|---|---|---|---|
| No log | 1.0 | 6 | 1.4218 | 0.2558 |
| 1.7221 | 2.0 | 12 | 1.4061 | 0.3953 |
| 1.7221 | 3.0 | 18 | 1.4801 | 0.3256 |
| 1.2972 | 4.0 | 24 | 1.5453 | 0.3023 |
| 1.2115 | 5.0 | 30 | 1.2993 | 0.3953 |
| 1.2115 | 6.0 | 36 | 1.4486 | 0.3721 |
| 1.1196 | 7.0 | 42 | 1.4881 | 0.3721 |
| 1.1196 | 8.0 | 48 | 1.2031 | 0.4419 |
| 1.0394 | 9.0 | 54 | 1.1825 | 0.4651 |
| 0.9076 | 10.0 | 60 | 1.3831 | 0.3953 |
| 0.9076 | 11.0 | 66 | 1.5606 | 0.3953 |
| 0.8351 | 12.0 | 72 | 1.6879 | 0.3721 |
| 0.8351 | 13.0 | 78 | 1.5744 | 0.5581 |
| 0.7325 | 14.0 | 84 | 2.1220 | 0.5116 |
| 0.5767 | 15.0 | 90 | 2.2458 | 0.4884 |
| 0.5767 | 16.0 | 96 | 2.4745 | 0.3953 |
| 0.487 | 17.0 | 102 | 2.9255 | 0.3953 |
| 0.487 | 18.0 | 108 | 2.8169 | 0.4186 |
| 0.265 | 19.0 | 114 | 2.9600 | 0.4419 |
| 0.2739 | 20.0 | 120 | 3.0131 | 0.3953 |
| 0.2739 | 21.0 | 126 | 3.2413 | 0.4186 |
| 0.1684 | 22.0 | 132 | 4.9920 | 0.3953 |
| 0.1684 | 23.0 | 138 | 3.1514 | 0.5116 |
| 0.3265 | 24.0 | 144 | 4.1598 | 0.3953 |
| 0.2652 | 25.0 | 150 | 3.3248 | 0.4651 |
| 0.2652 | 26.0 | 156 | 3.1898 | 0.4884 |
| 0.1992 | 27.0 | 162 | 3.7937 | 0.3953 |
| 0.1992 | 28.0 | 168 | 3.9838 | 0.4884 |
| 0.1826 | 29.0 | 174 | 3.5764 | 0.3721 |
| 0.124 | 30.0 | 180 | 4.1231 | 0.4419 |
| 0.124 | 31.0 | 186 | 4.1455 | 0.4186 |
| 0.1353 | 32.0 | 192 | 3.9925 | 0.4186 |
| 0.1353 | 33.0 | 198 | 3.7016 | 0.5581 |
| 0.0743 | 34.0 | 204 | 3.7997 | 0.5349 |
| 0.0362 | 35.0 | 210 | 3.6073 | 0.4884 |
| 0.0362 | 36.0 | 216 | 3.6198 | 0.4651 |
| 0.0082 | 37.0 | 222 | 3.6509 | 0.4651 |
| 0.0082 | 38.0 | 228 | 3.7081 | 0.4651 |
| 0.003 | 39.0 | 234 | 3.7432 | 0.4651 |
| 0.002 | 40.0 | 240 | 3.7616 | 0.4651 |
| 0.002 | 41.0 | 246 | 3.7690 | 0.4651 |
| 0.0018 | 42.0 | 252 | 3.7699 | 0.4651 |
| 0.0018 | 43.0 | 258 | 3.7699 | 0.4651 |
| 0.0016 | 44.0 | 264 | 3.7699 | 0.4651 |
| 0.0017 | 45.0 | 270 | 3.7699 | 0.4651 |
| 0.0017 | 46.0 | 276 | 3.7699 | 0.4651 |
| 0.0017 | 47.0 | 282 | 3.7699 | 0.4651 |
| 0.0017 | 48.0 | 288 | 3.7699 | 0.4651 |
| 0.0018 | 49.0 | 294 | 3.7699 | 0.4651 |
| 0.0017 | 50.0 | 300 | 3.7699 | 0.4651 |
Framework versions
- Transformers 4.35.0
- Pytorch 2.1.0+cu118
- Datasets 2.14.6
- Tokenizers 0.14.1