micppo: Another Proximal Policy Optimization (PPO) implementation

This is my implementation of Proximal Policy Optimization.

I wrote this code following a tutorial from Costa Huang.

Installation

Clone the repository, cd into it and then run

pip install .

The command above will install the micppo script into your PATH.

Usage

The micppo script allows to make experiments with the gymnasium environments. By default it runs on "CartPole", but you can configure it. It logs results for tensorboard. Run

micppo --help

for having a complete command-line help.

Author

Michele De Stefano