Deepseek AI Document Q&A

An advanced document analysis and question-answering application powered by the LLAMA RAG (Retrieval-Augmented Generation) model. The system enables users to upload documents or paste text content and receive AI-powered insights and contextual answers through an intuitive web interface.

Overview

Deepseek AI Document Q&A leverages cutting-edge natural language processing to transform how users interact with their documents. By combining document parsing, intelligent chunking, and retrieval-augmented generation, the application provides accurate, context-aware responses to user queries based on uploaded content.

Technology Stack

Core Technologies

• Python 3.8+ - Primary development language
• LLAMA 3.2 - Large language model for response generation
• LangChain - Framework for LLM application development
• Streamlit - Web application framework
• Chroma/FAISS - Vector database for document embedding storage

Document Processing

• PyPDF2/pdfplumber - PDF document parsing
• python-docx - DOCX file processing
• tiktoken - Text tokenization and chunking
• sentence-transformers - Text embedding generation

AI/ML Libraries

• transformers - Hugging Face model integration
• torch - PyTorch for deep learning operations
• numpy - Numerical computing
• pandas - Data manipulation and analysis

Features

• Multi-format Support: Seamlessly processes PDF, DOCX, and TXT documents
• Intelligent Document Indexing: Advanced text chunking with metadata preservation
• Semantic Search: Vector-based retrieval for relevant document sections
• Context-aware Responses: RAG-powered answers with source attribution
• Interactive Web Interface: User-friendly Streamlit dashboard
• Configurable Parameters: Adjustable model settings and processing options
• Real-time Processing: Live document analysis and query processing

Project Structure

├── app.py                          # Main Streamlit application
├── src/
│   ├── document_processor.py       # Document parsing and chunking
│   ├── embeddings.py              # Text embedding generation
│   ├── retriever.py               # Document retrieval logic
│   ├── llm_handler.py             # LLAMA model integration
│   └── utils.py                   # Utility functions
├── config/
│   ├── model_config.yaml          # Model configuration settings
│   └── app_config.yaml            # Application parameters
├── data/
│   ├── uploads/                   # Uploaded document storage
│   └── vector_store/              # Vector database storage
├── requirements.txt               # Python dependencies
├── Dockerfile                     # Container configuration
├── docker-compose.yml             # Multi-service orchestration
└── tests/
    ├── test_processor.py          # Document processing tests
    └── test_retriever.py          # Retrieval system tests

Installation

Prerequisites

• Python 3.8 or higher
• 8GB+ RAM (recommended for optimal model performance)
• GPU support (optional, for faster inference)

Setup Instructions

1. Clone the repository

   git clone https://github.com/yourusername/LLM-Powered-Document-Q-A-with-Llama-3.2.git
   cd LLM-Powered-Document-Q-A-with-Llama-3.2

2. Create and activate virtual environment

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

3. Install dependencies

   pip install -r requirements.txt

4. Configure environment variables (optional)

   cp .env.example .env
   # Edit .env with your API keys and model paths

Dependencies

Core Requirements

streamlit>=1.28.0
langchain>=0.0.300
transformers>=4.35.0
torch>=2.0.0
sentence-transformers>=2.2.2
chromadb>=0.4.0

Document Processing

PyPDF2>=3.0.1
python-docx>=0.8.11
pdfplumber>=0.9.0
tiktoken>=0.5.1

Additional Libraries

numpy>=1.24.0
pandas>=2.0.0
regex>=2023.8.8
tqdm>=4.65.0

Usage

Starting the Application

1. Launch the Streamlit interface

   streamlit run app.py

2. Access the web interface

   • Open your browser and navigate to http://localhost:8501
   • The application will be ready for document upload and queries

Using the Interface

1. Document Upload: Use the file uploader to select PDF, DOCX, or TXT files
2. Text Input: Alternatively, paste text directly into the text area
3. Query Processing: Enter your questions in the query input field
4. Configuration: Adjust settings in the sidebar for optimal results

Configuration

Model Parameters

Configure the application through the Streamlit sidebar:

• Model Selection: Choose from available LLAMA model variants
• Temperature (0.0-1.0): Controls response creativity and randomness
• Chunk Size (100-2000): Sets document segmentation size
• Overlap Size (0-200): Defines chunk overlap for context preservation
• Max Tokens: Limits response length

Advanced Settings

Edit config/model_config.yaml for advanced configurations:

model:
  name: "llama-3.2-8b"
  temperature: 0.7
  max_tokens: 512

processing:
  chunk_size: 1000
  chunk_overlap: 100
  embedding_model: "all-MiniLM-L6-v2"

retrieval:
  similarity_threshold: 0.7
  max_results: 5

Docker Deployment

For containerized deployment:

docker-compose up -d

This will start the application with all necessary services configured.

Performance Optimization

• GPU Acceleration: Enable CUDA for faster model inference
• Model Caching: Pre-load models to reduce startup time
• Batch Processing: Process multiple documents simultaneously
• Vector Store Optimization: Use persistent storage for large document collections

Contributing

We welcome contributions to improve the project! Please follow these steps:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/enhancement)
3. Commit your changes (git commit -am 'Add new feature')
4. Push to the branch (git push origin feature/enhancement)
5. Create a Pull Request

Development Guidelines

• Follow PEP 8 style guidelines
• Add unit tests for new functionality
• Update documentation for API changes
• Ensure compatibility with Python 3.8+

License

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

Acknowledgments

• Meta AI for the LLAMA 3.2 model
• LangChain community for the RAG framework
• Streamlit team for the web application framework
• Hugging Face for model hosting and transformers library

Support

For questions, issues, or feature requests:

• Open an issue on GitHub
• Check the documentation for detailed guides
• Join our community discussions