Chess AI Project

This project demonstrates a simple reinforcement learning agent that learns to play chess through self‑play.

Setup

1. Create a Python virtual environment.
2. Install dependencies:

   pip install -r requirements.txt

Training

Run the training script to let the agent learn by self-play:

python -m chess_ai.train --episodes 1000 --checkpoint agent.pth

This will store a model checkpoint at agent.pth.

Play against the AI

After training, you can play against the agent either in the terminal or using the graphical interface.

Terminal UI

python -m chess_ai.play_ui --model agent.pth

Moves are entered in UCI notation (e.g., e2e4).

Pygame UI

python -m chess_ai.play_pygame --model agent.pth

When a piece is selected, the square is highlighted and all legal target squares are marked. Click a highlighted square to move the piece. The AI will then make its move in response.

Repository Structure

• chess_ai/ – Package containing the environment, agent, and scripts
  • assets/svg/ – SVG icons for chess pieces
  • play_pygame.py – interactive Pygame interface
  • play_ui.py – terminal interface
  • train.py – training entry point
• requirements.txt – Project dependencies
• README.md – Project overview and instructions