Unity Reinforcement Learning - Cart Pole

A demo of a cart pole agent that was trained using a custom framework for RL in Unity. This agent was trained in Python with PyTorch using a build of a Unity project as the training environment. The agent was then exported as an onnx file and integrated into the Unity project using the Barracuda plugin.

To learn more about the RL training framework, click here.

Note: When the agent is disabled, a new agent that makes random decisions takes over.