This repository is a work in progress implementation of: Learning to Drive a Bicycle Using Reinforcement Learning and Shaping I am not affiliated with the authors of the paper.
The file main.py contains all the code. The Bicycle class describes the environment. The agent sends the action to the environment through the method Bicycle.take_action and the method returns next state, reward and is_end.
- Implement Bicycle Environment
- Implement Model in PyTorch
- Run episodes
- Obtain Convergence
The work is in progress to obtain convergence.
Learning to ride bicycle using reinforcement learning and shaping Bicycle environment