calc episodic infos for some fancy envs
This commit is contained in:
@@ -399,6 +399,7 @@ class PPO(GaussianRolloutCollectorAuxclass, OnPolicyAlgorithm):
|
||||
self.logger.record("train/clip_fraction", np.mean(clip_fractions))
|
||||
self.logger.record("train/loss", loss.item())
|
||||
self.logger.record("train/explained_variance", explained_var)
|
||||
self.logger.record("train/ssf", self.sde_sample_freq)
|
||||
if hasattr(self.policy, "log_std"):
|
||||
self.logger.record(
|
||||
"train/std", th.exp(self.policy.log_std).mean().item())
|
||||
|
||||
Reference in New Issue
Block a user