README.md

PuLID

📖 PuLID: Pure and Lightning ID Customization via Contrastive Alignment

Zinan Guo*, Yanze Wu*✝, Zhuowei Chen, Lang Chen, Qian He
(*Equal Contribution, ✝Corresponding Author)
ByteDance Inc

🚩 Updates

• 2024.09.12: 💥 We're thrilled to announce the release of the PuLID-FLUX-v0.9.0 model. Enjoy exploring its capabilities! 😊 Learn more about this model
• 2024.05.23: share the preview of our upcoming v1.1 model, please stay tuned
• 2024.05.01: release v1 codes&models, also the 🤗HuggingFace Demo
• 2024.04.25: release arXiv paper.

PuLID for FLUX

Please check the doc and demo of PuLID-FLUX here.

We will actively update and maintain this repository in the near future, so please stay tuned.

updates

• Local gradio demo is ready now
• Online HuggingFace demo is ready now
• We have optimized the codes to support consumer-grade GPUS, and now PuLID-FLUX can run on a 16GB graphic card. Check the details here

Below results are generated with PuLID-FLUX.

Examples

Images generated with our PuLID Applications

pulid_demo.mp4

🔧 Dependencies and Installation

• Python >= 3.9 (Recommend to use Anaconda or Miniconda)
• PyTorch >= 2.0 if you don't need flux-dev-fp8, otherwise PyTorch >= 2.4.1

# clone PuLID repo
git clone https://github.com/ToTheBeginning/PuLID.git
cd PuLID
# create conda env
conda create --name pulid python=3.10
# activate env
conda activate pulid
# Install dependent packages
# 1. if you don't need flux-fp8, e.g., you are using xl or flux-bf16, install the following requirements.txt
pip install -r requirements.txt
# 2. if you need flux-fp8 (to put flux on consumer-grade gpu), install the following requirements_fp8.txt
pip install -r requirements_fp8.txt

⚡ Quick Inference

Local Gradio Demo

python app.py

Online HuggingFace Demo

Thanks for the GPU grant from HuggingFace team, you can try PuLID HF demo in https://huggingface.co/spaces/yanze/PuLID

📎 Related Resources

Following are some third-party implementations of PuLID we have found in the Internet. We appreciate the efforts of the respective developers for making PuLID accessible to a wider audience. If there are any PuLID based resources and applications that we have not mentioned here, please let us know, and we will include them in this list.

Online Demo

• Colab: https://github.com/camenduru/PuLID-jupyter provided by camenduru
• Replicate: https://replicate.com/zsxkib/pulid provided by zsxkib

ComfyUI

• https://github.com/cubiq/PuLID_ComfyUI provided by cubiq, native ComfyUI implementation
• https://github.com/ZHO-ZHO-ZHO/ComfyUI-PuLID-ZHO provided by ZHO, diffusers-based implementation

WebUI

• Mikubill/sd-webui-controlnet#2838 provided by huchenlei

Disclaimer

This project strives to impact the domain of AI-driven image generation positively. Users are granted the freedom to create images using this tool, but they are expected to comply with local laws and utilize it responsibly. The developers do not assume any responsibility for potential misuse by users.

Citation

If PuLID is helpful, please help to ⭐ the repo.

If you find this project useful for your research, please consider citing our paper:

@article{guo2024pulid,
  title={PuLID: Pure and Lightning ID Customization via Contrastive Alignment},
  author={Guo, Zinan and Wu, Yanze and Chen, Zhuowei and Chen, Lang and He, Qian},
  journal={arXiv preprint arXiv:2404.16022},
  year={2024}
}

📧 Contact

If you have any comments or questions, please open a new issue or feel free to contact Yanze Wu and Zinan Guo.