Intelli Chess Lab

Project Overview

Intelli Chess Lab is an interactive tool that allows users to set chess positions on a virtual board and request a strategic analysis of the position through a chatbot. Users can choose between receiving a general report on the position or focusing on specific aspects, such as the pawn structure, piece activity, and others. The project leverages advanced natural language processing technologies via LLMs to provide understandable and useful natural language analyses, going beyond traditional chess engines that only provide the best move.

Project Components

The project consists of two main services that work together to provide the complete functionality:

• api-server
  This is the backend service that receives chess positions and generates strategic reports requested by the user using ChatGPT for analysis. It is developed in Python and Flask.

• web-client
  This is the frontend service, a web interface that allows users to set chess positions on a virtual board and request reports on the positions they configure. This component is developed with Node.js and React, providing an intuitive and user-friendly interface.

Use of Stockfish

The backend uses a modified version of Stockfish 15.1, which includes text traces with relevant strategic information about the chess position. These traces provide additional details that are used by the language model (ChatGPT) to generate more comprehensive and accurate analyses. The modified version of Stockfish can be found in the repository StockfishTraces.

To run the service manually, you need to download this project, compile it, and place the resulting stockfish executable in the api-server/stockfish directory. This allows the backend to access Stockfish for position analysis before passing the information to the language model.

Running the Project

Option 1: Running with Docker

1. Configuring the ChatGPT Token:
   You need to configure the ChatGPT token in the environment variables for the api-server service, within the docker-compose.yml file. Other environment variables are optional, but the default configuration should work fine.

2. Starting the Services with Docker:
   Once the environment variables are set, you can run the following command in the project’s root directory:

   docker compose up --build

   This command will build and start both the backend (api-server) and frontend (web-client) services, using the configurations defined in the docker-compose.yml file.

Option 2: Running Manually

1. Configuring the ChatGPT Token

You need to edit the .env file located in api-server/.env to set the ChatGPT token. Other configuration properties in the .env file can also be modified, but the default configuration is valid.

2. Creating a Python Virtual Environment (Optional but Recommended)

It is recommended to create a virtual environment to isolate the project's dependencies. Run the following commands in the api-server directory:

cd api-server
python -m venv intelli-chess-lab
source intelli-chess-lab/bin/activate  # On Windows: intelli-chess-lab\Scripts\activate

3. Installing Backend Dependencies

Once the virtual environment is activated, install the required dependencies for the backend by running:

pip install -r requirements.txt

4. Download and Compile Modified Stockfish

Download the modified version of Stockfish from the following repository:
StockfishTraces.
Compile it and copy the stockfish executable to the api-server/stockfish folder.

5. Running the Backend Manually

With the virtual environment activated, dependencies installed, and Stockfish configured, you can start the backend service by running:

python src/app.py

6. Running the Frontend Manually

Navigate to the web-client directory, install the Node.js dependencies, and start the frontend server with the following commands:

cd web-client
npm install
npm start

With both services running, the application will be ready to interact and generate strategic chess analyses.