smids_1x_deit_tiny_sgd_0001_fold4
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: 0.8842
- Accuracy: 0.5767
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 |
|---|---|---|---|---|
| 1.2663 | 1.0 | 75 | 1.2852 | 0.3417 |
| 1.1957 | 2.0 | 150 | 1.2221 | 0.36 |
| 1.1637 | 3.0 | 225 | 1.1784 | 0.3767 |
| 1.1755 | 4.0 | 300 | 1.1489 | 0.365 |
| 1.123 | 5.0 | 375 | 1.1280 | 0.3733 |
| 1.103 | 6.0 | 450 | 1.1122 | 0.3833 |
| 1.0888 | 7.0 | 525 | 1.0988 | 0.3917 |
| 1.0793 | 8.0 | 600 | 1.0870 | 0.4 |
| 1.0578 | 9.0 | 675 | 1.0759 | 0.4133 |
| 1.0777 | 10.0 | 750 | 1.0654 | 0.4183 |
| 1.017 | 11.0 | 825 | 1.0553 | 0.4183 |
| 1.0652 | 12.0 | 900 | 1.0460 | 0.43 |
| 1.0325 | 13.0 | 975 | 1.0371 | 0.4417 |
| 1.0323 | 14.0 | 1050 | 1.0281 | 0.45 |
| 1.0256 | 15.0 | 1125 | 1.0196 | 0.4567 |
| 1.0233 | 16.0 | 1200 | 1.0112 | 0.465 |
| 0.9433 | 17.0 | 1275 | 1.0027 | 0.4667 |
| 0.989 | 18.0 | 1350 | 0.9944 | 0.4733 |
| 0.9912 | 19.0 | 1425 | 0.9866 | 0.4783 |
| 0.9822 | 20.0 | 1500 | 0.9787 | 0.4817 |
| 0.9897 | 21.0 | 1575 | 0.9718 | 0.495 |
| 0.9177 | 22.0 | 1650 | 0.9648 | 0.5067 |
| 0.95 | 23.0 | 1725 | 0.9583 | 0.5133 |
| 0.9863 | 24.0 | 1800 | 0.9520 | 0.515 |
| 0.9643 | 25.0 | 1875 | 0.9462 | 0.5217 |
| 0.9026 | 26.0 | 1950 | 0.9408 | 0.5267 |
| 0.9173 | 27.0 | 2025 | 0.9357 | 0.5283 |
| 0.927 | 28.0 | 2100 | 0.9309 | 0.5333 |
| 0.9287 | 29.0 | 2175 | 0.9265 | 0.5417 |
| 0.9309 | 30.0 | 2250 | 0.9222 | 0.5467 |
| 0.9103 | 31.0 | 2325 | 0.9183 | 0.5483 |
| 0.9355 | 32.0 | 2400 | 0.9146 | 0.5583 |
| 0.9176 | 33.0 | 2475 | 0.9110 | 0.56 |
| 0.883 | 34.0 | 2550 | 0.9078 | 0.5617 |
| 0.9163 | 35.0 | 2625 | 0.9049 | 0.5633 |
| 0.8785 | 36.0 | 2700 | 0.9021 | 0.5667 |
| 0.8597 | 37.0 | 2775 | 0.8996 | 0.565 |
| 0.8694 | 38.0 | 2850 | 0.8972 | 0.5717 |
| 0.8945 | 39.0 | 2925 | 0.8952 | 0.5733 |
| 0.8725 | 40.0 | 3000 | 0.8932 | 0.575 |
| 0.875 | 41.0 | 3075 | 0.8915 | 0.575 |
| 0.8552 | 42.0 | 3150 | 0.8899 | 0.575 |
| 0.8903 | 43.0 | 3225 | 0.8886 | 0.575 |
| 0.8534 | 44.0 | 3300 | 0.8873 | 0.575 |
| 0.8746 | 45.0 | 3375 | 0.8863 | 0.5767 |
| 0.8976 | 46.0 | 3450 | 0.8855 | 0.5767 |
| 0.8522 | 47.0 | 3525 | 0.8849 | 0.5767 |
| 0.8759 | 48.0 | 3600 | 0.8845 | 0.5767 |
| 0.9028 | 49.0 | 3675 | 0.8842 | 0.5767 |
| 0.8753 | 50.0 | 3750 | 0.8842 | 0.5767 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0