Basic-RAG-Chatbot-FAISS-Phi3

An offline RAG (Retrieval-Augmented Generation) chatbot using FAISS for document retrieval and Microsoft Phi3 for response generation. Easily integrate via Flask API for private, fast Q&A from indexed documents. Includes a Jupyter notebook for setup, making it ideal for localized FAQ, customer support, and knowledge base applications.

Features

• FAISS-powered Document Retrieval: Quickly indexes and searches through documents.
• Local LLM Integration: Uses Microsoft Phi3 for offline, context-aware responses.
• Prompt Engineering: Tailored prompts ensure relevant, concise answers.
• Flask API: Simple integration with other applications.

Getting Started

1. Setup: Create a Conda environment and install dependencies (faiss-cpu, flask, torch, etc.).
2. Document Indexing: Load documents with FAISS to enable retrieval.
3. Run: Start the chatbot via the Flask server and begin querying.

Use Cases

Ideal for customer support, technical knowledge bases, and internal FAQs.

For more detailed instructions, refer to the included Jupyter notebook.