Artur Niederfahrenhorst

> I am actually not a fan of learning_starts -> min_size. min_size seems to imply the size of the RB, not how many samples there are in the buffer. This...

@gjoliver Yes, that's the one linked in the description!

Absolutely. I simply copied the steps recommended by our "_DCO failed Details_" page without thinking much about what would be rebased here. Seems that a lot of changes from master...

I'll leave this as it is for a couple of hours so that we can decide if we want to change the steps recommended by the DCO page.

This PR includes both places that a change would be effective in. 1) The docker ml requirements affect all GPU tests and 2) the app config only affects RLLib.

@HJasperson I've discussed this with @kouroshHakha and we have come to the conclusion that I've [presented above](https://github.com/ray-project/ray/pull/27527#discussion_r968194058). Let's treat entropy and KL changes as the same "issue" and the logps...

Hey @HJasperson I'd do the changes myself but git is giving a cryptic error **for this PR** that I'll not resolve for now. > To github.com:ray-project/ray.git ! [remote rejected] HEAD...

@mgerstgrasser this does not look related to me. If the error persists, feel free to open a separate issue. Closing because this has not been an issue for a while...

@krfricke I'm on-call this week and will focus on #31293

Thought I had to fix stuff related to https://github.com/ray-project/ray/issues/31321 and https://github.com/ray-project/ray/issues/31314 . That's already fixed though so I ended up with these minor changes in comments @sven1977 .