/PPO_implementation_v4.0

PPO algorithm implemetation for TF 2.8.0

Primary LanguagePythonMIT LicenseMIT

PPO_implementation_v4.0

PPO algorithm implemetation for continuous action spaces (TF 2.8.0)

The algorithm was tested on OpenAI Gym 'Pendulum-v1' environment

This implementation is heavily inspired by @mandrakedrink's one for PyTorch (https://github.com/mandrakedrink/PPO-pytorch)

To take a look at the parameters you can set, run: python main.py -h