A number of agents (PPO, MuZero) with a Perceiver-based NN architecture that can be trained to achieve goals in nethack/minihack environments.
Why do you think that https://github.com/opendilab/LightZero is a good alternative to omega
A number of agents (PPO, MuZero) with a Perceiver-based NN architecture that can be trained to achieve goals in nethack/minihack environments.
Why do you think that https://github.com/opendilab/LightZero is a good alternative to omega