smids_10x_deit_small_adamax_001_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: 1.6552
- Accuracy: 0.8733
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 |
|---|---|---|---|---|
| 0.2966 | 1.0 | 750 | 0.3694 | 0.8683 |
| 0.2339 | 2.0 | 1500 | 0.4363 | 0.8417 |
| 0.2355 | 3.0 | 2250 | 0.4010 | 0.8633 |
| 0.1515 | 4.0 | 3000 | 0.4419 | 0.8767 |
| 0.2106 | 5.0 | 3750 | 0.4830 | 0.8683 |
| 0.1439 | 6.0 | 4500 | 0.4916 | 0.8683 |
| 0.0791 | 7.0 | 5250 | 0.5811 | 0.875 |
| 0.077 | 8.0 | 6000 | 0.7477 | 0.8617 |
| 0.0484 | 9.0 | 6750 | 0.9268 | 0.8383 |
| 0.0478 | 10.0 | 7500 | 0.8560 | 0.855 |
| 0.0454 | 11.0 | 8250 | 0.7616 | 0.8733 |
| 0.0414 | 12.0 | 9000 | 0.8591 | 0.8617 |
| 0.0224 | 13.0 | 9750 | 0.8231 | 0.8833 |
| 0.0062 | 14.0 | 10500 | 0.9264 | 0.8717 |
| 0.004 | 15.0 | 11250 | 0.8932 | 0.8783 |
| 0.0299 | 16.0 | 12000 | 0.8030 | 0.8733 |
| 0.0268 | 17.0 | 12750 | 0.8616 | 0.88 |
| 0.0071 | 18.0 | 13500 | 0.9511 | 0.8767 |
| 0.0023 | 19.0 | 14250 | 0.9282 | 0.8783 |
| 0.008 | 20.0 | 15000 | 1.1898 | 0.855 |
| 0.0085 | 21.0 | 15750 | 1.0698 | 0.8683 |
| 0.0003 | 22.0 | 16500 | 1.1571 | 0.8633 |
| 0.0004 | 23.0 | 17250 | 1.1256 | 0.8783 |
| 0.0035 | 24.0 | 18000 | 1.2671 | 0.8633 |
| 0.0 | 25.0 | 18750 | 1.1579 | 0.8683 |
| 0.002 | 26.0 | 19500 | 1.2159 | 0.87 |
| 0.0001 | 27.0 | 20250 | 1.2282 | 0.8717 |
| 0.0 | 28.0 | 21000 | 1.2713 | 0.8683 |
| 0.0 | 29.0 | 21750 | 1.3150 | 0.8683 |
| 0.0 | 30.0 | 22500 | 1.2639 | 0.8733 |
| 0.0 | 31.0 | 23250 | 1.4238 | 0.865 |
| 0.0 | 32.0 | 24000 | 1.3138 | 0.8717 |
| 0.0 | 33.0 | 24750 | 1.4236 | 0.8733 |
| 0.0 | 34.0 | 25500 | 1.4930 | 0.865 |
| 0.0 | 35.0 | 26250 | 1.4369 | 0.87 |
| 0.0 | 36.0 | 27000 | 1.4573 | 0.8667 |
| 0.0 | 37.0 | 27750 | 1.4567 | 0.8717 |
| 0.0 | 38.0 | 28500 | 1.4973 | 0.8767 |
| 0.0 | 39.0 | 29250 | 1.5427 | 0.8667 |
| 0.0 | 40.0 | 30000 | 1.5656 | 0.8717 |
| 0.0 | 41.0 | 30750 | 1.5787 | 0.8717 |
| 0.0 | 42.0 | 31500 | 1.5845 | 0.87 |
| 0.0 | 43.0 | 32250 | 1.5904 | 0.8717 |
| 0.0 | 44.0 | 33000 | 1.5995 | 0.8717 |
| 0.0 | 45.0 | 33750 | 1.6192 | 0.8717 |
| 0.0 | 46.0 | 34500 | 1.6307 | 0.8717 |
| 0.0 | 47.0 | 35250 | 1.6406 | 0.8733 |
| 0.0 | 48.0 | 36000 | 1.6477 | 0.8733 |
| 0.0 | 49.0 | 36750 | 1.6529 | 0.8733 |
| 0.0 | 50.0 | 37500 | 1.6552 | 0.8733 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2