Harry Potter Quiz App

A Streamlit application that uses RAG (Retrieval-Augmented Generation) to answer questions about the Harry Potter universe.

Features

• Ask questions about Harry Potter
• Retrieves relevant passages from the books
• Uses Groq's LLM to generate accurate answers based on the retrieved content

Local Setup

1. Clone this repository
2. Install dependencies:

   pip install -r requirements.txt
   

3. Create a .env file with your Groq API key:

   GROQ_API_KEY=your-api-key-here
   

4. Run the database formation script to create the vector database:

   python database_formation.py
   

5. Run the Streamlit app:

   streamlit run app.py
   

Deploying to Streamlit Cloud

1. Fork this repository to your GitHub account
2. Sign up for Streamlit Cloud
3. Create a new app and connect it to your GitHub repository
4. In the app settings, add your Groq API key as a secret:
   • Go to "Advanced settings" > "Secrets"
   • Add a new secret with the key GROQ_API_KEY and your API key as the value
5. Deploy the app

Project Structure

• app.py: The main Streamlit application
• database_formation.py: Script to create the vector database from text files
• documents/: Directory containing Harry Potter text files
• db/: Directory where the vector database is stored
• requirements.txt: List of Python dependencies

Notes

• The app requires SQLite 3.35.0 or higher, which is available on Streamlit Cloud
• Make sure to upload your Harry Potter text files to the documents/ directory before running database_formation.py