Assetto Corsa RL

A reinforcement learning project for driving in Assetto Corsa.

A reinforcement learning agent (using the SAC algorithm) learns to drive a Formula 1 car around the Monaco GP circuit in Assetto Corsa. The agent controls steering, acceleration, and braking by interacting with the track, receiving feedback, and improving over time.

• 1. Train a Soft Actor-Critic (SAC) agent on a simplified 2D environment: OpenAI’s CarRacing-v3.
• 2. Adapt and port the trained agent to the Assetto Corsa racing simulator.
• 3. Implement full race simulations using AI-driven race strategies.

Contributions are welcome! Suggested workflow:

1. Fork the repo and create a feature branch
2. Add tests for new behavior
3. Open a pull request and describe the change

If you have questions or want to collaborate, open an issue or reach out via the project's issue tracker.