ComfyUI Extension: ComfyUI-SmolLM3

Authored by XmYx

Created

Updated

3 stars

Welcome to ComfyUI-SmolLM3, where we bring the magic of Hugging Face's SmolLM3 language models into your ComfyUI workflow! Whether you're crafting stories, generating ideas, or building AI-powered creativity tools, this node pack makes it delightfully simple.

Custom Nodes (0)

    README

    🪄 ComfyUI-SmolLM3 ✨

    SmolLM3 Banner License Stars

    "Small models, big dreams, infinite possibilities" 🌟

    Welcome to ComfyUI-SmolLM3, where we bring the magic of Hugging Face's SmolLM3 language models into your ComfyUI workflow! Whether you're crafting stories, generating ideas, or building AI-powered creativity tools, this node pack makes it delightfully simple.

    Created with 💜 by XmYx - bringing a touch of magic to the AI animation and VFX community!

    ✨ Features

    • 🎯 Simple & Intuitive - Just load and generate! No complex setups required
    • 🚀 Dual Model Support - Choose between base and instruct models
    • 🎨 Flexible Generation - From simple completions to full chat conversations
    • ⚡ Optimized Performance - Support for fp16, fp32, and bf16 precision
    • 🎲 Reproducible Results - Seed support for consistent outputs
    • 🌈 ComfyUI Native - Seamlessly integrates with your existing workflows

    🎬 What's Inside

    🔮 Three Magical Nodes

    1. SmolLM3 Model Loader - Your gateway to loading the models
    2. SmolLM3 Sampler (Chat) - Full-featured chat generation with all the bells and whistles
    3. SmolLM3 Simple Generate - Quick and easy text completion

    🚀 Installation

    1. Navigate to your ComfyUI custom nodes folder:

      cd ComfyUI/custom_nodes/

    2. Clone this repository:

      git clone https://github.com/XmYx/ComfyUI-SmolLM3.git

    3. Install the required dependencies:

      cd ComfyUI-SmolLM3
pip install -r requirements.txt

    4. Restart ComfyUI and let the magic begin! ✨

    📖 Usage

    Basic Workflow

    1. Add a SmolLM3 Model Loader node
    2. Select your preferred model:
      • HuggingFaceTB/SmolLM3-3B-Base - For general text generation
      • HuggingFaceTB/SmolLM3-3B - For chat and instruction following
    3. Connect to either:
      • SmolLM3 Sampler for chat-style interactions
      • SmolLM3 Simple Generate for basic completions
    4. Set your prompt and parameters
    5. Generate! 🎉

    🎯 Pro Tips

    • 🌡️ Use temperature = 0.6 and top_p = 0.95 for best results (as recommended by the SmolLM3 team)
    • 🎲 Set a seed for reproducible outputs
    • 💬 Enable chat template for conversational AI
    • 🖥️ Choose CPU mode if you're running on limited GPU memory

    🌟 Example Prompts

    "Give me a brief explanation of gravity in simple terms."
"Write a haiku about artificial intelligence."
"What are three creative uses for a paperclip?"
"Continue this story: Once upon a time in a digital realm..."

    🤝 Contributing

    Found a bug? Have an idea? Want to add more sparkle? ✨

    Contributions are warmly welcomed! Feel free to:

    • 🐛 Report issues
    • 💡 Suggest features
    • 🔧 Submit pull requests
    • ⭐ Star the repo if you find it useful!

    🎨 About the Creator

    This node pack is crafted with love by XmYx (Miklos Nagy), a human, artist and ML enthusiast who believes in making AI tools accessible and fun for creators. Check out my other projects:

    📜 License

    This project is licensed under the MIT License - because sharing is caring! 💝

    🙏 Acknowledgments

    • The amazing Hugging Face team for SmolLM3
    • The ComfyUI community for the incredible platform
    • You, for being part of this creative journey! 🌟
    <p align="center"> Made with 💜 and a sprinkle of ✨ by the AI creative community </p> <p align="center"> <i>"In a world of large models, sometimes the smallest ones cast the biggest spells"</i> 🪄 </p>