Maintain a scoreboard or list of notable agents/solutions/implementations

[cirosantilli]

I know this might entail a lot of work or difficult decisions. But it would be really cool if it were easy to see cool selected solutions to the problems.

The ultimate solution I guess is to have an automated online judge system. The other option is for maintainers to try and run things locally. Or maybe only published materials. Not sure exactly.

But it would be really cool and valuable if there was something.

https://github.com/DLR-RM/rl-baselines3-zoo seems to contain some implementations and pre-trained models.

[pseudo-rnd-thoughts]

Hi, historically, there has been a leaderboard on Gym (https://github.com/openai/gym/wiki/Leaderboard), however, due to the difficulty of maintaining it and of verifying results, I haven't pursed it