PDF Assistant Using Lamma2 (RAG - Project)

Overview

This project is a Retrieval-Augmented Generation (RAG) application that enables users to ask questions about PDF documents using open-source LLMs. Built with Llama2, the application retrieves relevant information from uploaded PDFs and generates accurate, context-aware responses. It's an ideal proof-of-concept for integrating NLP with document querying for educational, research, or knowledge management purposes.

Features

• PDF Upload & Processing: Upload and extract text from PDF files for efficient indexing and retrieval
• Natural Language Q&A: Ask questions about PDF content and receive AI-generated answers
• Open-Source Stack: Uses Llama2 for LLM capabilities without reliance on proprietary APIs
• Vector Search: Embeddings-based retrieval for semantic matching and accuracy
• Streamlit UI: Intuitive web interface for easy interaction
• Error Handling: Robust file processing with automatic cleanup

Tech Stack

• Language: Python 3.8+
• LLM: Llama2 (via Ollama)
• Libraries:
  • LangChain - RAG pipeline orchestration
  • PyPDF2/PyPDFLoader - PDF text extraction
  • Ollama Embeddings - Vector embeddings
  • DocArrayInMemorySearch - Vector store
  • Streamlit - Web interface

Installation

Prerequisites

• Python 3.8+
• Ollama installed with llama2 model
  • Download: ollama.ai
  • Run: ollama pull llama2

Setup Steps

1. Clone the repository:

   git clone https://github.com/harshsharma/Ragapplication.git
   cd Ragapplication

2. Create a virtual environment:

   python -m venv .venv
   source .venv/bin/activate  # On Windows: .venv\Scripts\activate

3. Install dependencies:

   pip install -r requirements.txt

4. Run the application:

   streamlit run app.py

5. Open your browser to http://localhost:8501

Usage

1. Upload PDF: Click the upload area and select a PDF file
2. Ask Question: Enter your question in the text input (e.g., "What are the main topics?")
3. Get Answer: The app retrieves relevant sections and generates a response

Project Structure

Ragapplication/
├── app.py              # Main Streamlit application
├── requirements.txt    # Python dependencies
├── README.md           # This file
└── .gitignore          # Git ignore rules

Troubleshooting

Issue	Solution
ModuleNotFoundError	Ensure virtual environment is activated: source .venv/bin/activate
Ollama connection error	Verify Ollama is running and llama2 model is installed
Slow responses	Check system resources; Llama2 requires significant memory
PDF processing fails	Ensure PDF is not corrupted or password-protected

Future Enhancements

• Support for multiple document formats (DOCX, TXT, HTML)
• Chat history and conversation memory
• Custom model support and fine-tuning
• Performance optimization with caching
• Deployment via Docker

Contributing

Contributions are welcome! Please fork the repository and submit a pull request.

License

MIT License - feel free to use this project as a reference or starting point.

Contact

For questions or collaboration opportunities, reach out via GitHub Issues.