Pytorch Implementation of MuZero Unplugged for gym environment. This algorithm is capable of supporting a wide range of action and observation spaces, including both discrete and continuous variations.
Why do you think that https://github.com/openai/retro is a good alternative to Muzero-unplugged