CDoc: Chat with Your Document

CDoc empowers you to have a conversation with your documents using local large language models (LLMs) and the power of Ollama, ChromaDB, and LangChain.

Key Features:

• Chat with Documents: Ask questions and get answers directly from your documents.
• Local LLM Support: Leverage the capabilities of local LLMs for offline document interaction.
• ChromaDB Support: Store and manage document metadata efficiently with ChromaDB.
• LangChain Integration: Streamline information extraction from documents through LangChain.

Target Users:

• Researchers and students seeking an efficient way to interact with research papers.
• Developers and programmers looking to analyze code documentation.
• Professionals wanting to extract key information from contracts and legal documents (Optional with OpenAI API).

Installation

Prerequisites:

• Python >=3.9 (https://www.python.org/downloads/)
• pip (usually comes pre-installed with Python)
• Ollama (https://ollama.com/)

Installation Steps:

1. Clone the repository:

   git clone https://github.com/ChatDocDev/CDoc
   

2. Navigate to the project directory:

   cd CDoc
   

3. Open project directory in VSCode

   code .
   

or any other code editor

4. Install dependencies from requirements.txt

   pip install -r requirements.txt
   

5. Pull the required models from Ollama

   • Download & install Ollama if not installed

   • Open terminal & run these command to pull the required models into local machine

     For llama3

     ollama pull llama3:latest
     

     For nomic-embed-text

     ollama pull nomic-embed-text:latest
     

   • Insure both models are downloaded

     ollama ls
     

   • Serve Ollama

     ollama serve
     

     goto localhost:11434 & you should get Ollama is running

6. BACKEND

   go to backend directory

   cd backend
   

   create db folder for storing Chromadb files

   mkdir db
   

   Start Chromadb server:

   chroma run --path db --port 8001
   

   Open new terminal and go into backend folder(hint: cd backend) & Run backend server:

   python backend.py
   

7. FRONTEND

   Open new terminal and go to frontend folder

   cd frontend
   

   Run frontend.py

   streamlit run frontend.py