Antonin RAFFIN

Duplicate of https://github.com/DLR-RM/stable-baselines3/issues/720, it is addressed in https://github.com/DLR-RM/stable-baselines3/pull/704 but I think I would welcome a separate PR as it is not clear if #720 will be merged and when ;)

Hello, > The documentation says that MultiBinary is supported (at least for some of the RL agents), however it looks like the only supported case is MultiBinary(n) where n is...

@Rocamonde in case you have time in the coming weeks, could you do a similar PR to our contrib repo? Otherwise, I will open an issue in that repo not...

> I suppose you're referring to the return types of specific algorithms? yes, and I mean this repo: https://github.com/Stable-Baselines-Team/stable-baselines3-contrib It follows the exact same structure as SB3. > What is...

Hello, Layer normalization was originally made for using parameter noise, so they were removed in SB3 (cf. docs). However, recently, there have some works that do use LN and dropout,...

Hello, > tests are not being type checked, yes, this is intentional, as @AdamGleave mentioned, we are doing something funky things to test the lib and I didn't to spend...

Hello, I understand your motivation but this is a special problem that would require too much changes internally (replacing fixed-size numpy rollout buffer by variable size list, this will have...

> Right now, if number of steps is less than length of episode then a rollout may not have the final reward, so it won't be discounted for any transition...

Hello, `logger.close` seems be call nowhere indeed... but that's weird we didn't get the problem earlier (I never experienced that, even when using early stopping), could you share your configuration...

Hello, thanks for the detailed report =) Looks like a legitimate bug indeed. Will leave it to @Miffyli or @qgallouedec if they have time, otherwise, I will have a closer...