Ayush266/textbook-qa-llama2

Textbook Q&A with LLaMA 2

This repo contains a combined FastAPI backend and Streamlit frontend for an interactive textbook Q&A system using LLaMA 2.

How to run locally

1. 1.Install dependencies:

pip install -r requirements.txt

1. 1.Set your Hugging Face token as environment variable:

export HF_TOKEN="your_hf_token_here"

1. 1.Run:

python app.py

1. 1.Open the Streamlit interface at http://localhost:8501

How to deploy on Hugging Face Spaces

1. 1.Upload all files in this repo.
2. 2.Add your Hugging Face token as secret HF_TOKEN in the Space settings.
3. 3.The Space will auto-install requirements and start.
4. 4.Use the web UI to upload PDFs/images and ask questions!

Notes

• —Backend (FastAPI) runs at port 8000 inside the Space.
• —Streamlit frontend communicates with backend locally.
• —Uses GPU if available for LLaMA 2 model.