airkjw/RETFound
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

update readme

Browse Source
This commit is contained in:
rmaphoh
2025-09-02 16:33:57 +01:00
parent bda7a6c69f
commit 17768be893
1 changed files with 23 additions and 21 deletions
Download Patch File Download Diff File Expand all files Collapse all files
README.md
+23 -21
View File
@@ -1,11 +1,11 @@
## RETFound - A foundation model for retinal imaging
## RETFound - A foundation model for retinal images
Official repo including a series of foundation models and applications in retinal imaging.<br>
Official repo including a series of foundation models and applications for retinal images.<br>
`[RETFound-MAE]`:[RETFound: a foundation model for generalizable disease detection from retinal images](https://www.nature.com/articles/s41586-023-06555-x).<br>
`[RETFound-DINOv2]`:[Revealing the Impact of Pre-training Data on Medical Foundation Models](https://www.researchsquare.com/article/rs-6080254/v1).<br>
`[DINOv2]`:[General-purpose vision foundation models DINOv2](https://github.com/facebookresearch/dinov2).<br>
`[DINOv3]`:[General-purpose vision foundation models DINOv3](https://github.com/facebookresearch/dinov3).<br>
`[DINOv2]`:[General-purpose vision foundation models DINOv2 by Meta](https://github.com/facebookresearch/dinov2).<br>
`[DINOv3]`:[General-purpose vision foundation models DINOv3 by Meta](https://github.com/facebookresearch/dinov3).<br>
Please contact **ykzhoua@gmail.com** or **yukun.zhou.19@ucl.ac.uk** if you have questions.
@@ -51,8 +51,6 @@ pip install -r requirements.txt
### 🌱Fine-tuning with RETFound weights
To fine tune RETFound on your own data, follow these steps:
1. Get access to the pre-trained models on HuggingFace (register an account and fill in the form) and go to step 2:
<table><tbody>
<!-- START TABLE -->
@@ -102,7 +100,9 @@ huggingface-cli login --token YOUR_HUGGINGFACE_TOKEN
export HF_ENDPOINT=https://hf-mirror.com
```
3. Organise your data into this directory structure (Public datasets used in this study can be [downloaded here](BENCHMARK.md))
3. If you would like to fine-tune [DINOv2]((https://github.com/facebookresearch/dinov2)) and [DINOv3](https://github.com/facebookresearch/dinov3), please visit their GitHub repositories to download the model weights and put them in the RETFound folder.
4. Organise your data into this directory structure (Public datasets used in this study can be [downloaded here](BENCHMARK.md))
```
├── data folder
@@ -120,12 +120,12 @@ export HF_ENDPOINT=https://hf-mirror.com
├──class_c
```
4. If you would like to use DINOv2 and DINOv3, please visit their GitHub repositories to download the model weights and put them in the RETFound folder.
4. Start fine-tuning by running `sh train.sh`.
The model can be selected by changing the hyperparameters `MODEL`, `MODEL_ARCH`, `FINETUNE` in `train.sh`:
5. Start fine-tuning by running `sh train.sh`.
In `train.sh`, the model can be selected by changing the hyperparameters `MODEL`, `MODEL_ARCH`, `FINETUNE`:
**RETFound**:
@@ -161,6 +161,8 @@ The model can be selected by changing the hyperparameters `MODEL`, `MODEL_ARCH`,
| Dinov2 | dinov2_vitg14 | dinov2_vitg14_pretrain.pth | ~1.1B |
Change the DATA_PATH to your dataset directory.
```
# ==== Model settings ====
# adaptation {finetune,lp}
@@ -173,8 +175,10 @@ FINETUNE="RETFound_dinov2_meh"
# change the dataset name and corresponding class number
DATASET="MESSIDOR2"
NUM_CLASS=5
data_path="./${DATASET}"
task="${MODEL_ARCH}_${DATASET}_${ADAPTATION}"
# =======================
DATA_PATH="PATH TO THE DATASET"
TASK="${MODEL_ARCH}_${DATASET}_${ADAPTATION}"
torchrun --nproc_per_node=1 --master_port=48766 main_finetune.py \
--model "${MODEL}" \
@@ -186,16 +190,16 @@ torchrun --nproc_per_node=1 --master_port=48766 main_finetune.py \
--world_size 1 \
--epochs 50 \
--nb_classes "${NUM_CLASS}" \
--data_path "${data_path}" \
--data_path "${DATA_PATH}" \
--input_size 224 \
--task "${task}" \
--task "${TASK}" \
--adaptation "${ADAPTATION}"
```
4. For evaluation only (download data and model checkpoints [here](BENCHMARK.md); change the path below)
6. For evaluation only (download data and model checkpoints [here](BENCHMARK.md); change the DATA_PATH below)
```
@@ -208,7 +212,9 @@ FINETUNE="RETFound_dinov2_meh"
# ==== Data/settings (match training) ====
DATASET="MESSIDOR2"
NUM_CLASS=5
DATA_PATH="./${DATASET}"
# =======================
DATA_PATH="PATH TO THE DATASET"
TASK="${MODEL_ARCH}_${DATASET}_${ADAPTATION}"
# Path to the trained checkpoint (adjust if you saved elsewhere)
@@ -237,10 +243,6 @@ torchrun --nproc_per_node=1 --master_port=48766 main_finetune.py \
If you find this repository useful, please consider citing this paper:
```
TBD
```
```
@article{zhou2023foundation,
title={A foundation model for generalizable disease detection from retinal images},