📚 RAG Based Local Document Chatter

A simple RAG system that enables you to have interactive conversations with your PDF documents using OpenAI's GPT models and FAISS vector storage.

🌟 Features

• PDF document processing and vectorization
• FAISS similarity search
• Interactive Streamlit chat interface
• Configurable RAG parameters
• Debug mode for retrieval analysis
• Chat history persistence

🚀 Setup

1. Clone the repository:

   git clone https://github.com/zumpious/rag-based-document-chatter
   cd rag-based-document-chatter

2. Create .env and set up your environment variables:

   cp .env.template .env   

3. Create and activate virtual environment (Optional):

   python -m venv venv
   source venv/bin/activate 

4. Install dependencies (Optional):

   pip install -r requirements.txt

📖 Usage

1. Process your document and create FAISS vector embedding:

   python scripts/process_document.py  

2. Verify setup (optional):

   python scripts/test_setup.py   

3. Run the chat interface:

   streamlit run main.py   

🐳 Docker Setup

1. Build the Docker image:

   docker build -t rag_app .

2. Run the container:

   docker run -d -p 8501:8501 --name rag_app rag_app

3. Acess the application:

   • Open http://localhost:8501 in your browser

🐋 Docker Compose Setup

1. Build and start the container:

   docker compose up -d

2. Access the application:

   • Open http://localhost:8501 in your browser