README.md

metadata

title: PaperChat - Research Paper Q&A
emoji: 📄
colorFrom: blue
colorTo: purple
sdk: streamlit
sdk_version: 1.52.2
app_file: app.py
pinned: false

📄 PaperChat - Research Paper Q&A Assistant

Ask questions about any research paper using AI-powered retrieval and reasoning.

🎯 Features

• Upload Any Paper: Support for PDF research papers
• Natural Questions: Ask questions in plain English
• Accurate Answers: Powered by Agentic RAG with source citations
• Chat Interface: Interactive conversation with context
• Fast Processing: Efficient document chunking and retrieval

🚀 How to Use

1. Upload a research paper (PDF) using the sidebar
2. Wait for the paper to be processed
3. Ask any question about the paper
4. Get detailed answers with citations

🛠️ Technology Stack

• Framework: Streamlit
• Agent System: smolagents (Hugging Face)
• Retrieval: BM25 with LangChain
• LLM: Qwen/Qwen3-Next-80B (via HF Inference API)
• PDF Processing: PyPDF + LangChain text splitters

💡 Example Questions

• "What is the main contribution of this paper?"
• "What methodology was used in the experiments?"
• "What datasets were used for evaluation?"
• "What are the key results and findings?"
• "What are the limitations mentioned by the authors?"

🔐 API Configuration

This app uses the Hugging Face Inference API. Make sure to set your HF_TOKEN environment variable or add it in the Hugging Face Spaces secrets.

📝 Local Development

# Install dependencies
pip install -r requirements.txt

# Set your HF token
export HF_TOKEN=your_token_here

# Run the app
streamlit run app.py

📄 License

MIT License

🙏 Acknowledgments

Built with smolagents, LangChain, and Streamlit.