More PCA related fixes

This commit is contained in:
Dominik Moritz Roth 2023-08-22 00:00:24 +02:00
parent 81e223183a
commit f3683afb86

View File

@ -677,7 +677,7 @@ class ActorCriticPolicy(BasePolicy):
elif isinstance(self.action_dist, StateDependentNoiseDistribution):
return self.action_dist.proba_distribution(mean_actions, self.log_std, latent_pi)
elif isinstance(self.action_dist, PCA_Distribution):
return self.action_dist.proba_distribution(mean_actions, self.log_std)
return self.action_dist.proba_distribution(mean_actions, th.ones_like(mean_actions) * self.log_std.exp())
else:
raise ValueError("Invalid action distribution")