smids_10x_deit_tiny_adamax_00001_fold3
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: 0.8454
- Accuracy: 0.9117
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 |
|---|---|---|---|---|
| 0.3334 | 1.0 | 750 | 0.3137 | 0.8867 |
| 0.2212 | 2.0 | 1500 | 0.2847 | 0.8933 |
| 0.1557 | 3.0 | 2250 | 0.2655 | 0.9017 |
| 0.1618 | 4.0 | 3000 | 0.2547 | 0.9083 |
| 0.1458 | 5.0 | 3750 | 0.2827 | 0.9117 |
| 0.0827 | 6.0 | 4500 | 0.3021 | 0.9067 |
| 0.0585 | 7.0 | 5250 | 0.3302 | 0.915 |
| 0.038 | 8.0 | 6000 | 0.3826 | 0.9117 |
| 0.0072 | 9.0 | 6750 | 0.4421 | 0.9033 |
| 0.0278 | 10.0 | 7500 | 0.4604 | 0.9117 |
| 0.024 | 11.0 | 8250 | 0.5200 | 0.9083 |
| 0.0385 | 12.0 | 9000 | 0.5950 | 0.905 |
| 0.0036 | 13.0 | 9750 | 0.6062 | 0.9167 |
| 0.0266 | 14.0 | 10500 | 0.6478 | 0.9083 |
| 0.0002 | 15.0 | 11250 | 0.6832 | 0.905 |
| 0.0008 | 16.0 | 12000 | 0.6882 | 0.9083 |
| 0.0001 | 17.0 | 12750 | 0.7251 | 0.9033 |
| 0.0001 | 18.0 | 13500 | 0.7412 | 0.9 |
| 0.0 | 19.0 | 14250 | 0.7303 | 0.905 |
| 0.0 | 20.0 | 15000 | 0.7516 | 0.9083 |
| 0.0 | 21.0 | 15750 | 0.7572 | 0.905 |
| 0.0 | 22.0 | 16500 | 0.7738 | 0.9033 |
| 0.0 | 23.0 | 17250 | 0.7783 | 0.905 |
| 0.0 | 24.0 | 18000 | 0.7887 | 0.9083 |
| 0.0 | 25.0 | 18750 | 0.8047 | 0.9067 |
| 0.0 | 26.0 | 19500 | 0.7947 | 0.905 |
| 0.0 | 27.0 | 20250 | 0.8074 | 0.9017 |
| 0.0 | 28.0 | 21000 | 0.8188 | 0.9067 |
| 0.0 | 29.0 | 21750 | 0.8171 | 0.9083 |
| 0.0 | 30.0 | 22500 | 0.8254 | 0.91 |
| 0.0 | 31.0 | 23250 | 0.8136 | 0.91 |
| 0.0 | 32.0 | 24000 | 0.8161 | 0.9117 |
| 0.0 | 33.0 | 24750 | 0.8510 | 0.9083 |
| 0.0 | 34.0 | 25500 | 0.8503 | 0.905 |
| 0.0 | 35.0 | 26250 | 0.8125 | 0.9083 |
| 0.0 | 36.0 | 27000 | 0.8406 | 0.9083 |
| 0.0 | 37.0 | 27750 | 0.8651 | 0.9033 |
| 0.0 | 38.0 | 28500 | 0.8378 | 0.9083 |
| 0.0 | 39.0 | 29250 | 0.8214 | 0.905 |
| 0.0 | 40.0 | 30000 | 0.8342 | 0.91 |
| 0.0 | 41.0 | 30750 | 0.8303 | 0.905 |
| 0.0 | 42.0 | 31500 | 0.8436 | 0.9117 |
| 0.0 | 43.0 | 32250 | 0.8482 | 0.9083 |
| 0.0 | 44.0 | 33000 | 0.8466 | 0.9117 |
| 0.0 | 45.0 | 33750 | 0.8396 | 0.9083 |
| 0.0 | 46.0 | 34500 | 0.8440 | 0.9117 |
| 0.0 | 47.0 | 35250 | 0.8439 | 0.9117 |
| 0.0 | 48.0 | 36000 | 0.8441 | 0.91 |
| 0.0 | 49.0 | 36750 | 0.8457 | 0.9117 |
| 0.0 | 50.0 | 37500 | 0.8454 | 0.9117 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2