munchausen_addon and action

[quyouyuan]

Hello!Excuse me！I used your code in another environment, but I encountered difficulties. Action is a decimal array,, so how to rewrite the " munchausen_addon = log_pi.gather(1, actions)" line of code! And the action space is very large! By the way, these lines of code are:adding the scald log-policy to the immediate reward
looking forward to your reply！

)