Skip to content

README.md

Latest commit

 

History

History
16 lines (11 loc) · 880 Bytes

README.md

File metadata and controls

16 lines (11 loc) · 880 Bytes

🧠 Simple RAG System - Document Query Assistant

This project implements a lightweight Retrieval-Augmented Generation (RAG) system using Streamlit, FAISS, and LLMs like OpenAI and Ollama. It allows users to upload PDF, TXT, or DOCX documents, embed them into a searchable vector database, and query them using natural language. The model then returns context-aware answers based on retrieved document chunks.

🚀 Features

  • 📄 Upload PDF, TXT, or DOCX files
  • 🔍 Split content into semantic chunks with overlapping context
  • 📦 Vector store backed by FAISS for efficient similarity search
  • 🔗 Embeddings from OpenAI, Nomic, or Chroma
  • 🧠 LLM-based question answering via OpenAI (GPT-4) or Ollama (LLaMA3)
  • 🧵 Streamlit UI with model selection, caching, and source highlighting