ComfyUI-OnDemand-Lora-Loader

This is a custom node for ComfyUI that allows you to download and apply LoRA models directly from Civitai/HuggingFace. It simplifies your workflow by managing LoRA downloads automatically, so you don't have to manually download files and place them in the correct folder.

Features

• Automatic Downloading: Downloads LoRA models from Civitai/HuggingFace URLs specified in a configuration file.
• No Re-downloads: Checks if a LoRA file already exists before attempting to download it.
• Progress Bar: Displays a progress bar in the console during download.
• Private Model Support: Access private or early-access models using your Civitai/HuggingFace API key.
• Centralized Configuration: Manage your list of LoRAs from a simple config.json file.
• Seamless Integration: Functions as a standard LoRA loader node within the ComfyUI interface.

Installation

1. Navigate to your ComfyUI custom_nodes directory:

   cd ComfyUI/custom_nodes/
   

2. Clone this repository:

   git clone https://github.com/francarl/ComfyUI-OnDemand-Lora-Loader.git
   

3. Install the required dependencies:

   cd ComfyUI-OnDemand-Lora-Loader
   pip install -r requirements.txt
   

4. Restart ComfyUI.

Usage

1. Configure Your LoRAs

After installation, you need to configure the LoRAs you want to use.

1. Navigate to the ComfyUI/custom_nodes/ComfyUI-OnDemand-Lora-Loader/ directory.
2. Create a file named config.json. If the file does not exist, the node will fall back to a default example configuration.
3. Add your LoRAs to the config.json file. Each entry requires a name (which will appear in the node's dropdown menu) and a url (the Civitai/HuggingFace download link).

You can get the download link from a model's page on Civitai/HuggingFace by right-clicking the download button and copying the link address. It should look like https://civitai.com/api/download/models/MODEL_ID or https://huggingface.co/user/folder/resolve/main/loras.safetensors.

Example config.json:

{
    "loras": [
        {
            "name": "Epi Noise Offset",
            "url": "https://civitai.com/api/download/models/16576"
        },
        {
            "name": "High Lora",
            "url": "https://huggingface.co/lightx2v/Wan2.2-Lightning/resolve/main/Wan2.2-T2V-A14B-4steps-lora-rank64-Seko-V1.1/high_noise_model.safetensors"
        },
        {
            "name": "My Private Lora",
            "url": "https://civitai.com/api/download/models/123456?type=Model&format=SafeTensor"
        }
    ]
}

2. Add the Node in ComfyUI

1. In ComfyUI, double-click on the canvas to open the search menu.
2. Search for and add the OnDemand Lora Loader node.
3. Connect it in your workflow just as you would with a standard LoraLoader node.

3. Node Inputs

• model: The input model from a checkpoint loader.
• clip (optional): The input CLIP from a checkpoint loader.
• lora_name: A dropdown menu to select the LoRA you configured in config.json.
• strength_model: The strength of the LoRA applied to the model.
• strength_clip: The strength of the LoRA applied to the CLIP model.
• api_key (optional): Your API key. Use this to download private or early-access models. Alternatively, you can set the CIVITAI_TOKEN and HUGGINGFACE_TOKEN environment variable.
• download_chunks (optional): The chunk size (in KB) for downloading files. Adjust if you experience slow downloads.

The first time you select a LoRA, it will be downloaded to your ComfyUI/models/loras/ directory. Subsequent uses will load the existing file.

License

This project is licensed under the MIT License. See the LICENSE.txt file for details.