smids_3x_deit_tiny_adamax_0001_fold1
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: 0.8020
- Accuracy: 0.8982
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 |
|---|---|---|---|---|
| 0.3976 | 1.0 | 226 | 0.3057 | 0.8848 |
| 0.2433 | 2.0 | 452 | 0.3108 | 0.8698 |
| 0.1348 | 3.0 | 678 | 0.3373 | 0.8982 |
| 0.1034 | 4.0 | 904 | 0.3559 | 0.9032 |
| 0.1083 | 5.0 | 1130 | 0.4835 | 0.8648 |
| 0.066 | 6.0 | 1356 | 0.5085 | 0.8881 |
| 0.0304 | 7.0 | 1582 | 0.5167 | 0.8982 |
| 0.0203 | 8.0 | 1808 | 0.4920 | 0.8915 |
| 0.0411 | 9.0 | 2034 | 0.6349 | 0.8948 |
| 0.0001 | 10.0 | 2260 | 0.5499 | 0.9015 |
| 0.0005 | 11.0 | 2486 | 0.8482 | 0.8715 |
| 0.0006 | 12.0 | 2712 | 0.6367 | 0.9082 |
| 0.0001 | 13.0 | 2938 | 0.6540 | 0.9082 |
| 0.0145 | 14.0 | 3164 | 0.7445 | 0.8848 |
| 0.0002 | 15.0 | 3390 | 0.7327 | 0.8948 |
| 0.0002 | 16.0 | 3616 | 0.7425 | 0.8898 |
| 0.0076 | 17.0 | 3842 | 0.8300 | 0.8881 |
| 0.0 | 18.0 | 4068 | 0.8281 | 0.8898 |
| 0.0117 | 19.0 | 4294 | 0.8348 | 0.8865 |
| 0.0 | 20.0 | 4520 | 0.7762 | 0.8932 |
| 0.0101 | 21.0 | 4746 | 0.8545 | 0.8781 |
| 0.0102 | 22.0 | 4972 | 0.6911 | 0.9048 |
| 0.0118 | 23.0 | 5198 | 0.7788 | 0.8965 |
| 0.0134 | 24.0 | 5424 | 0.8281 | 0.8898 |
| 0.0 | 25.0 | 5650 | 0.8580 | 0.8865 |
| 0.0002 | 26.0 | 5876 | 0.8155 | 0.8965 |
| 0.0 | 27.0 | 6102 | 0.7810 | 0.8982 |
| 0.0 | 28.0 | 6328 | 0.7738 | 0.9048 |
| 0.0 | 29.0 | 6554 | 0.7899 | 0.8915 |
| 0.0 | 30.0 | 6780 | 0.7800 | 0.9015 |
| 0.0 | 31.0 | 7006 | 0.7916 | 0.8982 |
| 0.0 | 32.0 | 7232 | 0.7756 | 0.8965 |
| 0.0037 | 33.0 | 7458 | 0.7754 | 0.9015 |
| 0.0 | 34.0 | 7684 | 0.7805 | 0.8982 |
| 0.0 | 35.0 | 7910 | 0.7846 | 0.8932 |
| 0.0 | 36.0 | 8136 | 0.8045 | 0.8932 |
| 0.0 | 37.0 | 8362 | 0.7958 | 0.8948 |
| 0.0 | 38.0 | 8588 | 0.7780 | 0.8965 |
| 0.0 | 39.0 | 8814 | 0.7869 | 0.8998 |
| 0.0 | 40.0 | 9040 | 0.7801 | 0.8965 |
| 0.0031 | 41.0 | 9266 | 0.7806 | 0.9032 |
| 0.0032 | 42.0 | 9492 | 0.8019 | 0.8982 |
| 0.0 | 43.0 | 9718 | 0.7990 | 0.8982 |
| 0.0 | 44.0 | 9944 | 0.8009 | 0.8965 |
| 0.0 | 45.0 | 10170 | 0.8036 | 0.8982 |
| 0.0 | 46.0 | 10396 | 0.8034 | 0.8982 |
| 0.0 | 47.0 | 10622 | 0.8034 | 0.8982 |
| 0.0 | 48.0 | 10848 | 0.8001 | 0.8982 |
| 0.0 | 49.0 | 11074 | 0.8018 | 0.8982 |
| 0.0 | 50.0 | 11300 | 0.8020 | 0.8982 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2