add link to github

README.md

@@ -7,7 +7,7 @@ pipeline_tag: image-classification
 
 # Unlocking the Hidden Potential of CLIP in Generalizable Deepfake Detection
 
-[![arXiv Badge](https://img.shields.io/badge/arXiv-B31B1B?logo=arxiv&logoColor=FFF&style=for-the-badge)](https://arxiv.org/abs/2503.19683)
+[![arXiv Badge](https://img.shields.io/badge/arXiv-B31B1B?logo=arxiv&logoColor=FFF)](https://arxiv.org/abs/2503.19683) [![GitHub Badge](https://img.shields.io/badge/GitHub-181717?logo=github&logoColor=fff)](https://github.com/yermandy/deepfake-detection)
 
 This repository contains the model for the paper:
 
@@ -42,14 +42,7 @@ Generalization of models trained on the FF++ dataset to unseen datasets and forg
 
 ## Example
 
-Find the code in our [github](https://github.com/yermandy/deepfake-detection) project. Read `inference.py`, it automatically downloads the model from [huggingface](https://huggingface.co/yermandy/deepfake-detection/tree/main) and runs inference on sample images. Make sure to have the required dependencies installed before running the script.
-
-``` bash
-python inference.py
-```
-
-**❗ Important note**: sample images are already preprocessed. To get the same results as in the paper, you need to preprocess images using DeepfakeBench [preprocessing](https://github.com/SCLBD/DeepfakeBench/blob/fb6171a8e1db2ae0f017d9f3a12be31fd9e0a3fb/preprocessing/preprocess.py) pipeline.
-
+See usage examples in our [github](https://github.com/yermandy/deepfake-detection) project
 
 ## Cite