This is a prototyping and learning workspace for Reinforcement Learning in Rust.
Goals:
- Reinforcement Algorithm possible to run on a typical developer laptop
- Use Rust as much as possible
- Direct GPU support is mandatory in production - it MUST NOT be rules out by the used components
The goal is to learn an ML model for playing interactive visual games like Breakout.
This is a learning project. So most things are implemented from scratch.