prashantbhandari/rag-linux-kernel

1---2title: Chat with Any Book3emoji: 📚4colorFrom: blue5colorTo: green6sdk: streamlit7sdk_version: 1.41.08app_file: app.py9pinned: false10---11 12# 📚 Chat with Any Book (RAG)13 14Upload any PDF book or document and ask questions — get answers with exact page citations.15 16## How it works17 181. Upload a PDF192. The book is split into ~800-token chunks and embedded203. Your question is matched to the most relevant passages214. Groq (Llama3-70B) answers using only those passages22 23## Stack24 25- **LLM**: Groq API — `llama-3.3-70b-versatile`26- **Embeddings**: `sentence-transformers/all-MiniLM-L6-v2`27- **Vector DB**: ChromaDB (in-memory, per session)28- **Frontend**: Streamlit29 30## Run locally31 32```bash33git clone https://huggingface.co/spaces/YOUR_USERNAME/rag-linux-kernel34cd rag-linux-kernel35pip install -r requirements.txt36GROQ_API_KEY=your_key streamlit run app.py37```38