ComfyUI-IF_AI_HFDownloaderNode

Download HF repos from comfy.

Installation

• You can clone this repo to your comfy custom nodes folder.

  git clone https://github.com/if-ai/ComfyUI-IF_AI_HFDownloaderNode.git
  

1. Activate your ComfyUI environment and install the package:

   pip install requirements.txt
   

2. Export the environment variable:

   nano ~/.bashrc
   

   Add the following line:

   export HF_TOKEN=Your_Access_Token_from_your_HF_account
   

   Save and close the file, then reload your bash configuration:

   source ~/.bashrc
   

3. For Windows:

   • Press Win + R and type systempropertiesadvanced.

   • Click on the "Environment Variables" button.

- Create a new system variable with the name `HF_TOKEN` and the value `Your_Access_Token_from_your_HF_account`.

Prerequisites

HuggingFAce hub from the reqs

1-.Take a repo ID from HF it can be a HF Space too. example "Kwai-Kolors/Kolors"

2-. On the bottom you can select individual Mode here you can select a single or coma separate names from a repo "vae/diffusion_pytorch_model.bin,model_index.json" will download the vae model inside the vae folder and the model_index.json at the root of kolors

3-. You can specify a path or use the default directory at the root of comfy ComfyUI/models/IF_AI with the same name as the repo you are downloading

4-. Exclude the files when on whole repo or space mode works the same as number 2 but exclude the files instead

5-. HuggingFace token you need to either write an access token here form the settings on your HF profile or create an environment variable for your user named HF_TOKEN, optionally Create a .env file in the root directory of the node with HF_TOKEN=your_token_here

6-. You can specify a folder to download

7-. Mode individual or download the full repo/space

🚨Do not save or share your workflows with the TOKENS weritten in fact do not share WF using this Node⚠️

Tutorial

VIDEO USING THE NODE

Related Tools

• IF_prompt_MKR - A similar tool available for Stable Diffusion WebUI
• IF_AI_tools - LLM Nodes for chating and Creating SD prompts locally with Ollama or Via APIs