Activate Stochastic Muzero Policy

bwfbowen / muax

A project that provides help for using DeepMind's mctx on gym-style environments.

MIT License

45 stars 9 forks source link

Activate Stochastic Muzero Policy #8

Open Karlheinzniebuhr opened 8 months ago

Karlheinzniebuhr commented 8 months ago

I need to usue the stochastic muzero policy.

I inspected the muax class and can see that currently only muzero and gumbel policies are supported. Any guidance on how to implement stochastic policy would be appreciated

bwfbowen commented 8 months ago

Sure, thanks for asking. When I was implementing muax, stochastic muzero policy hasn't been released. I will be working on it!