Hugging Face Text Generation WebUI (Gradio Edition)

A modern user interface for running Hugging Face text-generation models locally using Gradio.

Features

• Chat tab — conversational UI with history, persona/impersonation, advanced controls
• Notebook tab — longform notes/prompt engineering scratchpad
• Model selection — choose from several Hugging Face (local) models
• Advanced parameters — temperature, top-k
• Responsive UI — responsive ui
• Runs fully local — all generation happens on your machine

Quickstart

1. Install Python 3.8+ and run:

   pip install gradio transformers torch huggingface_hub

2. To add a model (by link or ID):

   python add_model.py mistralai/Mistral-7B-Instruct-v0.2
   # or
   python add_model.py https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2

   This will:

   • Download the model to models/mistralai-Mistral-7B-Instruct-v0.2
   • Auto-update AVAILABLE_MODELS in webui_gradio.py to include it

3. Run:

   python webui_gradio.py

4. Open http://localhost:7860 in your browser.

Customizing

• Add your favorite Hugging Face models to the models/ directory using add_model.py.
• Tweak the CSS in webui_gradio.py for a different look.
• For more options (max tokens, top-p, etc.), see the Gradio and Transformers docs.

Credits

• Powered by Gradio, Hugging Face Transformers, huggingface_hub