vit-finetuned-chest-xray-pneumonia
This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the chest-xray-pneumonia dataset. It achieves the following results on the evaluation set:
- Loss: 0.1271
- Accuracy: 0.9551
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: 2e-05
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| No log | 1.0 | 326 | 0.2739 | 0.9167 |
| 0.2238 | 2.0 | 652 | 0.2892 | 0.9071 |
| 0.2238 | 3.0 | 978 | 0.2077 | 0.9407 |
| 0.1385 | 4.0 | 1304 | 0.1349 | 0.9535 |
| 0.1347 | 5.0 | 1630 | 0.1271 | 0.9551 |
| 0.1347 | 6.0 | 1956 | 0.1458 | 0.9535 |
| 0.1112 | 7.0 | 2282 | 0.2040 | 0.9375 |
| 0.1063 | 8.0 | 2608 | 0.1423 | 0.9567 |
| 0.1063 | 9.0 | 2934 | 0.1473 | 0.9535 |
| 0.0944 | 10.0 | 3260 | 0.1385 | 0.9583 |
Example Images
Pneumonia Chest X-Ray
Normal Chest X-Ray
Framework versions
- Transformers 4.17.0
- Pytorch 1.10.0+cu111
- Datasets 1.18.4
- Tokenizers 0.11.6