Welcome to DocTalk! This project leverages advanced language models to create a powerful chatbot that can interact with PDF documents. The application is built using Streamlit, LangChain, and other modern technologies to provide an intuitive and interactive user experience.
- Upload and Chat with PDFs: Easily upload your PDF documents and ask questions to extract relevant information.
- Advanced Language Model: Powered by the Llama3-8b-8192 model from Groq.
- Efficient Text Processing: Utilizes FAISS for similarity search and vector indexing.
- Google Generative AI Embeddings: Enhanced document understanding with Google's state-of-the-art embeddings.
- Streamlit Integration: Simple and interactive user interface built with Streamlit.
- WhatsApp Integration: Seamlessly integrate the chatbot with WhatsApp for enhanced accessibility (future work).
Ensure you have Python installed. Recommended version: Python 3.8 or higher.
-
Clone the repository:
git clone https://github.com/your-username/doctalk.git cd doctalk -
Create virtual environment:
conda create -p venv python==3.12 conda activate venv/
-
Install the required dependencies:
pip install -r requirements.txt
-
Set up environment variables:
Create a
.envfile in the root directory and add your API keys:GROQ_API_KEY=your_groq_api_key GOOGLE_API_KEY=your_google_api_key -
Run the application:
streamlit run Main.py
-
Upload your PDF and start chatting!
Main.py: Main application file containing the Streamlit code and chatbot logic.requirements.txt: List of dependencies required for the project..env.example: Example of the environment variables file.data/: Directory to store uploaded PDF files.
- Upload a PDF: Use the file uploader to select and upload a PDF document.
- Embed the Document: The PDF is processed, and text is extracted and split into chunks for efficient querying.
- Ask Questions: Enter your question related to the document.
- Get Answers: The chatbot processes your question using the LLM and retrieves relevant information from the PDF.
- Streamlit: For building the web interface.
- PyPDF2: For PDF processing.
- LangChain: For handling language model interactions.
- FAISS: For efficient similarity search and vector indexing.
- Google Generative AI: For advanced text embeddings.
- dotenv: For managing environment variables.
- WhatsApp Integration: Extend the chatbot to work with WhatsApp for more accessible interactions.
- Enhanced PDF Processing: Improve the text extraction and processing capabilities for better accuracy.
We welcome contributions! Please fork the repository and submit a pull request to contribute to the project.
For any inquiries, please contact [email protected].
The chatbot in DocTalk utilizes a Retrieval-Augmented Generation (RAG) approach, which combines retrieval-based and generation-based techniques to enhance performance, especially for tasks involving large-scale or complex documents like PDFs.
- Document Retrieval: Utilizes FAISS for efficient similarity search and vector indexing to retrieve relevant text passages from PDF documents.
- Generation with Context: Employs the Groq-powered language model to generate contextually relevant answers based on the retrieved text.
- Enhanced Understanding: Integrates Google Generative AI Embeddings for semantic understanding and processing of document text, improving the accuracy and relevance of responses.
- Improved Accuracy: By combining retrieval with generation, the chatbot provides more accurate and contextually relevant responses.
- Efficient Processing: The RAG approach allows for efficient handling of large documents without requiring the entire document to be processed at once.
- Enhanced User Experience: Users can interact with PDF documents seamlessly, extracting information quickly and effectively.
DocTalk effectively leverages RAG principles to provide an efficient and powerful chatbot experience for interacting with PDF documents.