smids_10x_deit_tiny_adamax_00001_fold5
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: 0.8518
- Accuracy: 0.8933
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: 1e-05
- 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 |
|---|---|---|---|---|
| 0.292 | 1.0 | 750 | 0.2930 | 0.885 |
| 0.2303 | 2.0 | 1500 | 0.2459 | 0.9033 |
| 0.1782 | 3.0 | 2250 | 0.2507 | 0.9017 |
| 0.1592 | 4.0 | 3000 | 0.2451 | 0.9 |
| 0.1917 | 5.0 | 3750 | 0.2847 | 0.8967 |
| 0.0682 | 6.0 | 4500 | 0.3017 | 0.895 |
| 0.0567 | 7.0 | 5250 | 0.3497 | 0.8983 |
| 0.0573 | 8.0 | 6000 | 0.3610 | 0.9 |
| 0.1186 | 9.0 | 6750 | 0.4077 | 0.9067 |
| 0.0339 | 10.0 | 7500 | 0.4570 | 0.9017 |
| 0.0496 | 11.0 | 8250 | 0.5695 | 0.8933 |
| 0.0132 | 12.0 | 9000 | 0.6133 | 0.8983 |
| 0.0021 | 13.0 | 9750 | 0.5988 | 0.8933 |
| 0.0118 | 14.0 | 10500 | 0.6072 | 0.8933 |
| 0.0074 | 15.0 | 11250 | 0.6482 | 0.8967 |
| 0.0005 | 16.0 | 12000 | 0.7207 | 0.8933 |
| 0.0021 | 17.0 | 12750 | 0.7385 | 0.885 |
| 0.0 | 18.0 | 13500 | 0.7467 | 0.895 |
| 0.0005 | 19.0 | 14250 | 0.7555 | 0.8967 |
| 0.0 | 20.0 | 15000 | 0.7776 | 0.9 |
| 0.0 | 21.0 | 15750 | 0.8136 | 0.8933 |
| 0.0 | 22.0 | 16500 | 0.7839 | 0.8983 |
| 0.0002 | 23.0 | 17250 | 0.7795 | 0.8983 |
| 0.0 | 24.0 | 18000 | 0.7219 | 0.8967 |
| 0.0 | 25.0 | 18750 | 0.7978 | 0.9 |
| 0.0 | 26.0 | 19500 | 0.7801 | 0.8983 |
| 0.0 | 27.0 | 20250 | 0.8617 | 0.9 |
| 0.0 | 28.0 | 21000 | 0.7908 | 0.8967 |
| 0.0 | 29.0 | 21750 | 0.7838 | 0.9 |
| 0.0154 | 30.0 | 22500 | 0.8352 | 0.895 |
| 0.0 | 31.0 | 23250 | 0.8276 | 0.8967 |
| 0.0 | 32.0 | 24000 | 0.8168 | 0.8983 |
| 0.0 | 33.0 | 24750 | 0.8030 | 0.8933 |
| 0.0 | 34.0 | 25500 | 0.8162 | 0.895 |
| 0.0 | 35.0 | 26250 | 0.8097 | 0.895 |
| 0.0 | 36.0 | 27000 | 0.8226 | 0.8967 |
| 0.0047 | 37.0 | 27750 | 0.8297 | 0.895 |
| 0.0 | 38.0 | 28500 | 0.8250 | 0.8933 |
| 0.0 | 39.0 | 29250 | 0.8392 | 0.8983 |
| 0.0 | 40.0 | 30000 | 0.8326 | 0.9 |
| 0.0 | 41.0 | 30750 | 0.8399 | 0.895 |
| 0.0 | 42.0 | 31500 | 0.8505 | 0.895 |
| 0.0 | 43.0 | 32250 | 0.8292 | 0.8967 |
| 0.0 | 44.0 | 33000 | 0.8390 | 0.895 |
| 0.0 | 45.0 | 33750 | 0.8458 | 0.895 |
| 0.0 | 46.0 | 34500 | 0.8491 | 0.8933 |
| 0.0 | 47.0 | 35250 | 0.8489 | 0.8933 |
| 0.0 | 48.0 | 36000 | 0.8502 | 0.8933 |
| 0.0 | 49.0 | 36750 | 0.8534 | 0.8933 |
| 0.0 | 50.0 | 37500 | 0.8518 | 0.8933 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2