Instructions to use bitfount/RETFound_DR_IDRID with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- timm
How to use bitfount/RETFound_DR_IDRID with timm:
import timm model = timm.create_model("hf_hub:bitfount/RETFound_DR_IDRID", pretrained=True) - Notebooks
- Google Colab
- Kaggle
Model card for RETFound_DR_IDRID
A model fine-tuned on the IDRID dataset of retinal fundus images, initialised from the base open-eye/RETFound_MAE model for fundus images.
This model achieves good accuracy in the task of classifying level of diabetic retinopathy in fundus (0 = no DR, 1 = mild DR, 2 = moderate DR, 3 = severe DR, 4 = progressed DR).
NOTE this model was trained on greyscale images, so please ensure your images are converted to greyscale before running inference using this model.
- Downloads last month
- 15