📚 PDF RAG Chatbot

A powerful PDF question-answering chatbot built with LangChain, Groq, and Gradio. Upload any PDF document and ask questions about its content using advanced RAG (Retrieval-Augmented Generation) technology.

Live APP HF Link

HuggingFace

✨ Features

• 📄 Upload and process PDF documents
• 🤖 Ask questions about PDF content using natural language
• 🚀 Fast responses powered by Groq's Mixtral model
• 🎯 Accurate answers using RAG with vector similarity search
• 🖥️ Clean and intuitive web interface with Gradio
• 🔍 MMR (Maximal Marginal Relevance) search for diverse results

🛠️ Technologies Used

• LangChain: Framework for building LLM applications
• Groq: Ultra-fast LLM inference
• HuggingFace Embeddings: Sentence transformers for text embeddings
• ChromaDB: Vector database for efficient similarity search
• Gradio: Web UI framework
• PyPDF: PDF processing

📋 Prerequisites

• Python 3.8 or higher
• Groq API key (get it from console.groq.com)

🚀 Installation

1. Clone the repository

git clone <your-repo-url>
cd pdf-chatbot

2. Create a virtual environment

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

3. Install dependencies

pip install gradio langchain-groq langchain-community langchain-core langchain-text-splitters chromadb pypdf sentence-transformers python-dotenv torch

4. Set up environment variables

Create a .env file in the project root:

touch .env

Add your Groq API key to the .env file:

GROQ_API_KEY=gsk_your_actual_api_key_here

5. Add .env to .gitignore (to keep your API key secure)

echo ".env" >> .gitignore

🎮 Usage

1. Run the application

python app.py

2. Open your browser

Navigate to http://localhost:7860

3. Upload and Ask
   • Click on the file upload area and select a PDF
   • Type your question in the text box
   • Click Submit to get your answer

📁 Project Structure

pdf-chatbot/
├── .venv/              # Virtual environment
├── app.py              # Main application file
├── .env                # Environment variables (API keys)
├── .gitignore          # Git ignore file
└── README.md           # This file

🔑 Getting a Groq API Key

1. Visit console.groq.com
2. Sign up or log in
3. Navigate to the API Keys section
4. Create a new API key
5. Copy the key and add it to your .env file

💡 How It Works

1. Document Processing: The PDF is loaded and split into chunks
2. Embeddings: Text chunks are converted to vector embeddings
3. Vector Storage: Embeddings are stored in ChromaDB
4. Query Processing: User questions are converted to embeddings
5. Retrieval: Similar chunks are retrieved using MMR search
6. Generation: Groq's Mixtral model generates answers based on retrieved context

⚙️ Configuration

You can modify these parameters in app.py:

• Model: Change mixtral-8x7b-32768 to other Groq models
• Temperature: Adjust creativity (0.0-1.0)
• Max Tokens: Control response length
• Chunk Size: Modify document splitting (default: 1000)
• Chunk Overlap: Adjust context continuity (default: 200)
• K value: Number of documents to retrieve (default: 5)

🐛 Troubleshooting

Issue: ModuleNotFoundError

pip install --upgrade pip
pip install -r requirements.txt

Issue: API Key Error

• Ensure your .env file exists in the project root
• Verify the API key is correct
• Check that python-dotenv is installed

Issue: PDF Not Loading

• Ensure the PDF is not password-protected
• Check that the file is a valid PDF format
• Try with a smaller PDF first

📝 Example Questions

• "What is the main topic of this document?"
• "Can you summarize the key points?"
• "What does the document say about [specific topic]?"
• "Who are the authors mentioned?"
• "What are the conclusions?"

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

📄 License

This project is licensed under the MIT License.

🙏 Acknowledgments

• LangChain for the RAG framework
• Groq for lightning-fast LLM inference
• Gradio for the web interface
• HuggingFace for embeddings models

📞 Support

If you encounter any issues or have questions, please open an issue on GitHub.

---

Made with ❤️ using LangChain and Groq