ComfyUI_Personalize_Anything

ComfyUI_Personalize_Anything
★ 44

个性化训练Diffusion TransformerComfyUI 集成Wrapper 模式
在ComfyUI中通过Diffusion Transformer实现免费个性化训练，支持wrapper模式，便于快速创建自定义人物与风格。
💡 在ComfyUI内快速为模型添加自定义人物或风格
🍴 1 Forks💻 Python🔄 2025-03-26
📦
网盘下载
复制链接后前往夸克网盘下载
https://pan.quark.cn/s/c1eafc754fbb
📦 requirements.txt
accelerate
diffusers>=0.32.2
#ipykernel
#gradio
#matplotlib
opencv-python
protobuf
sentencepiece
transformers>=4.49.0
📄 README

ComfyUI_Personalize_Anything

Personalize Anything for Free with Diffusion Transformer,use it in comfyUI with wrapper mode

Update

  • fix inpainting bug, inpainting only use prompt line one /修复内绘的bug，注意内绘只用第一行的prompt

    • 1. Installation

    In the ./ComfyUI/custom_node directory, run the following:

    git clone https://github.com/smthemex/ComfyUI_Personalize_Anything

    2. Requirements

    pip install -r requirements.txt

    3.Model

    3.1 flux dev repo （preference/推荐）

  • 3.1 use flux dev diffusser repo

    • if OOM chocie nf4 or fp8 type 如果OOM，开启nf4或者fp8量化

    3.2 single model

  • 3.1.1 download kijai ‘flux1-dev.safetensors’ fp8 from here 下载kijai的flux单体模型模型,文件结构如下图
  • 3.1.2 download ‘clip_l.safetensors’ and ‘t5xxl_fp8_e4m3fn.safetensors’ from here下载T5和clip l单体模型模型,文件结构如下图

    • --   ComfyUI/models/diffusion_models  # or unet
    ├── flux1-dev-fp8-e4m3fn.safetensors  #11G
--   ComfyUI/models/clip
    ├── clip_l.safetensors
    ├── t5xxl_fp8_e4m3fn.safetensorsvae
--   ComfyUI/models/vae
    ├── ae.safetensor

    4.Example

  • Inpainting , use up prompt only
  • Outpainting
  • Reconstruction
  • Single_personalize
  • comfy single model

    • 5.Tips

  • 使用方法请参考示例图片/Please refer to the example image for usage instructions

    • 6. Citation

    @article{feng2025personalize,
  title={Personalize Anything for Free with Diffusion Transformer},
  author={Feng, Haoran and Huang, Zehuan and Li, Lin and Lv, Hairong and Sheng, Lu},
  journal={arXiv preprint arXiv:2503.12590},
  year={2025}
}