LLM Hackathon Template

A simple template for building LLM applications quickly, for hackathons and rapid prototyping. This template contains basic LLM client setup, environment variable management, a simple Gradio web interface, and is ready to extend with your features!

Quick Start

1. Install dependencies: Download uv here to install and manage packages (recommended). Change the folder my_project to the name of your project, and modify the pyproject.toml file to include the name and a short description of your project. Then, install the package and its dependencies:

   uv sync
   uv pip install -e .

   (Alternative: create a conda environment and run pip install -e .)

2. Add your API key:

   cp .env.example .env
   # Edit .env with your API key

3. Run the app:

   uv run main.py

   (Alternative: run with python main.py)

4. Open your browser to http://localhost:7860 to use the web interface!

5. Customize! Modify the Gradio app, package structure, and add your own features! Make sure to add a "graphical abstract" of your work as well.

Adding New Features

# Add a new package
uv add package-name

Project Structure

├── main.py              # Your main application with Gradio
├── notebooks/           # Notebooks for prototyping
├── src/
│   └── my_project/      # Your code goes here
├── .env.example         # Template for API keys
└── pyproject.toml       # Dependencies