PyTorch Implementation of the Maximum a Posteriori Policy Optimisation (paper1, paper2) Reinforcement Learning Algorithms for OpenAI gym environments.

I tested on the below environment.

• Windows 10
• Python 3.7
• PyTorch 1.8.1

Install PyTorch https://pytorch.org/

pip install gym Box2D IPython tqdm scipy tensorboard tensorboardx

python train.py \
  --device cuda:0 \
  --env LunarLanderContinuous-v2 \
  --log log_continuous \
  --render

python train.py \
  --device cuda:0 \
  --env LunarLander-v2 \
  --log log_discrete \
  --render

This repository is a clone of theogruner/rl_pro_telu, which is licensed under the GNU GPL3 License - see the LICENSE file for details