QA Chatbot

⚠️ Disclaimer

This project is in an experimental phase and remains under active development. Please use it with caution and understand that you do so at your own risk.

Architecture

Docker setup

Set environment variables

• Prepare the docker environment:

• Open the file named .env-example.

  • Replace the placeholder configurations with your actual settings:

    # For OpenAI configuration
    OPENAI_KEY="your-openai-key"
    
    # Default model settings
    MODEL_NAME="gpt-4o-mini"
    EMBEDDING_MODEL_NAME="text-embedding-3-small"
    TEMPERATURE=0.0

  • If using OpenAI models, make sure OPENAI_KEY is set, and you can select MODEL_NAME and EMBEDDING_MODEL_NAME from https://platform.openai.com/docs/models.

  • If using Ollama, remove OPENAI_KEY from the environment to avoid conflicts, and you can select MODEL_NAME and EMBEDDING_MODEL_NAME from https://www.ollama.com/search.

  • Save and close the file.

  • Rename the file to .env.

• Prepare the ollama environment:

1. Start the Ollama Docker image using docker-compose

  docker-compose start ollama

2. Pull a model supported by Ollama. You can find Ollama models here and replace mistral with another model, and avr/sfr-embedding-mistral with another embedding.

    docker exec -it ollama ollama pull mistral
    docker exec -it ollama ollama pull avr/sfr-embedding-mistral

Documents Setup

Since this project is based on context (RAG), all files in the *.txt format should be added to the backend/docs folder.

Please note that if you modify the models, you should ensure the embedding dimensions are correct. Otherwise, you can delete the existing data in chromaDB by calling the following function from your browser after running the server, and then restart the server to re-embed the docs automatically:

http://localhost:5000/delete-collection

Build & RUN

     docker compose up --build -d

RUN

     docker compose up -d

Development setup

Prerequisites

You can skip this part if your system meets all the requirements listed below 👇

• Backend requirements
• Frontend requirements

Install dependencies & set environment variables

• Clone the project in your preferred IDE (e.g. VScode)

• Prepare the backend environment:

  1. Navigate to the backend using backend folder: cd backend
  2. Create a virtual environment for dependencies called env using the following command:

       python -m venv env

  3. Activate the created env by running:
     • Windows:

       env\Scripts\activate.bat

     • Unix/MacOS:

       env/bin/activate

  4. Install the required libraries listed in requirements.txt by running:

       pip install -r requirements.txt

• Prepare the frontend environment:

  1. Navigate to the frontend folder: cd ../frontend
  2. follow the same steps of the backend environment

• Set Up Environment Variables:

  • Open the file named .env-example.

  • Replace the placeholder configurations with your actual settings:

    # For OpenAI configuration
    OPENAI_KEY="your-openai-key"
    
    # For Ollama configuration
    LLM_BASE_URL="http://localhost:11434"
    
    # Default model settings
    MODEL_NAME="gpt-4o-mini"
    EMBEDDING_MODEL_NAME="text-embedding-3-small"
    TEMPERATURE=0.0
    
     #for backend configuration
     API_BASE_URL="http://localhost:5000"
    
    # Database configuration
    CHROMA_DB_SERVER_HOST="localhost"
    CHROMA_DB_SERVER_PORT=8000
    CHROMA_DB_COLLECTION_NAME="docs_collection"

  • If using OpenAI models, make sure OPENAI_KEY is set, and you can select MODEL_NAME and EMBEDDING_MODEL_NAME from https://platform.openai.com/docs/models.

  • If using Ollama, remove OPENAI_KEY from the environment to avoid conflicts, and you can select MODEL_NAME and EMBEDDING_MODEL_NAME from https://www.ollama.com/search.

  • Save and close the file.

  • Rename the file to .env.

• Prepare the ollama environment:

  1. Start the Ollama Docker image using docker-compose

    docker-compose start ollama

  2. Pull a model supported by Ollama. You can find Ollama models here and replace mistral with another model, and avr/sfr-embedding-mistral with another embedding.

      docker exec -it ollama ollama pull mistral
      docker exec -it ollama ollama pull avr/sfr-embedding-mistral

• Prepare the ChromaDB environment:

  1. Start the ChromaDB Docker image using docker-compose

    docker-compose start chromadb

Documents Setup

Since this project is based on context (RAG), all files in the *.txt format should be added to the backend/docs folder.

Please note that if you modify the models, you should ensure the embedding dimensions are correct. Otherwise, you can delete the existing data in chromaDB by calling the following function from your browser after running the server, and then restart the server to re-embed the docs automatically:

http://localhost:5000/delete-collection

Run

1. Navigate to the backend using backend folder: cd backend
2. Run the application:

   uvicorn app:app --host 0.0.0.0 --port 5000 --reload

3. Open a new terminal and navigate to the frontendfolder: cd frontend
4. Run the development web server:

   streamlit run app.py --server.port=8080 --server.address=0.0.0.0

5. Access the web application on the displayed localhost: http://localhost:8080