RAG Demo Application

This Streamlit application demonstrates the power of Retrieval-Augmented Generation (RAG) by comparing standard LLM responses with RAG-enhanced responses. It clearly illustrates why RAG matters when dealing with recent information.

Features

• Side-by-side comparison of standard LLM vs. RAG-enhanced responses
• Built-in sample documents with recent information
• Support for custom document uploads
• Knowledge base explorer to view document contents
• Educational content explaining how RAG works

Installation

1. Clone this repository:

git clone https://github.com/yourusername/rag-demo.git
cd rag-demo

2. Create a virtual environment and activate it:

# For Windows
python -m venv venv
venv\Scripts\activate

# For macOS/Linux
python -m venv venv
source venv/bin/activate

3. Install the required packages:

pip install -r requirements.txt

4. Set up your OpenAI API key:
   • Create a .env file in the project root directory
   • Add your OpenAI API key to the .env file:

     OPENAI_API_KEY=your_openai_api_key_here
     

Usage

1. Run the Streamlit application:

streamlit run app.py

2. Open your web browser and go to http://localhost:8501

3. Use the application:

   • Ask questions about the sample documents or your own uploaded documents
   • Compare the standard LLM response with the RAG-enhanced response
   • Explore the knowledge base to view document contents
   • Learn about how RAG works

Application Structure

• app.py: Main Streamlit application
• requirements.txt: Required Python packages
• .env: Environment variables (add your OpenAI API key here)
• knowledge_base/: Directory for storing uploaded documents
• sample_docs/: Directory for sample documents
• chroma_db/: Directory for the Chroma vector database

How It Works

This application demonstrates Retrieval-Augmented Generation (RAG), which enhances language model outputs by:

1. Creating a knowledge base of documents
2. Converting documents into embeddings (vector representations)
3. Storing these embeddings in a vector database
4. When a question is asked:
   • Converting the question to an embedding
   • Finding the most similar document chunks
   • Providing these chunks as context to the language model
   • Generating a response based on this specific context

This approach addresses two major limitations of large language models:

• Knowledge cutoff: LLMs only know information up to their training date
• Hallucinations: LLMs can generate plausible but incorrect information

Sample Documents

The application includes sample documents about:

• A breakthrough in quantum computing
• A new climate agreement
• Revolutionary battery technology for electric vehicles
• Advances in Alzheimer's treatment
• Economic forecasts from a global forum

These documents contain "recent" information that would be beyond a typical LLM's knowledge cutoff date, demonstrating the value of RAG for accessing up-to-date information.

Custom Documents

You can upload your own text documents to the knowledge base. The application will process these documents and make them available for RAG queries.

Requirements

• Python 3.8+
• OpenAI API key
• Required packages listed in requirements.txt

License

This project is licensed under the MIT License - see the LICENSE file for details.