Skip to content

README.md

Latest commit

 

History

History
50 lines (42 loc) · 1.89 KB

README.md

File metadata and controls

50 lines (42 loc) · 1.89 KB
Raw

RAG-MultiFile-QA

GitHub Repo stars GitHub forks

📚 Multi-File Retrieval-Augmented Generation (RAG) Q&A System

This project is a Streamlit-based Q&A application that allows users to upload multiple document types (PDF, DOCX, TXT, CSV) and ask questions about their content using retrieval-augmented generation (RAG).

🔹 Features

  • Upload and process multiple files at once.
  • Supports PDF, DOCX, TXT, and CSV formats.
  • Uses Hugging Face Embeddings and FAISS vector search for document retrieval.
  • Integrates Hugging Face Inference API for generating responses.
  • Maintains chat history for seamless user experience.
  • Clear all button to reset uploaded files and chat history.

🛠️ Tech Stack

  • Python
  • Streamlit (Frontend UI)
  • Langchain (Document Processing & Retrieval)
  • Hugging Face Inference API (LLM-based Answer Generation)
  • FAISS (Vector Store for Efficient Retrieval)
  • PyPDFLoader, TextLoader, CSVLoader (File Parsing)

🚀 How to Run

  1. Clone the repository: 
    git clone https://github.com/your-username/RAG-MultiFile-QA.git
cd RAG-MultiFile-QA
  2. Install dependencies: 
    pip install -r requirements.txt
  3. Set your Hugging Face API Key as an environment variable: 
    export HUGGINGFACE_API_KEY="your_api_key"
  4. Run the app: 
    streamlit run main.py

📌 Notes

  • Ensure your Hugging Face API Key is correctly set.
  • The system works best with structured documents containing well-defined sections and tables.
  • FAISS indexing helps in faster search and retrieval from large documents.

📜 License

This project is open-source and available under the MIT License.