RL From Scratch

This repository accompanies a reinforcement learning course focused on implementing algorithms from first principles.

Rules

• All algorithms must be implemented in .py files
• Notebooks are for visualization and analysis only
• External RL libraries are not permitted

Setup

pip install -r requirements.txt

Instructor Notes

• This repo is intentionally incomplete: students fill in algorithms
• Designed for pytest-based autograding
• Easily extensible to robotics (ROS2, real sensors)

This repository was designed for use with Github classroom.