Isaac Gym Baselines with rl_games configurations implemented in SKRL

[noshaba]

Hi!

As the title suggests it would be super helpful to have at least one example of the runs and configurations of the Isaac Gym environments currently implemented with the rl_games library implemented in skrl. It doesn't need to be the whole thing, just the PPO, etc. baseline configurations, and which variable maps to which from rl_games to skrl, and having a tutorial about it on your website :)

Best,
Noshaba

[Toni-SM]

I will work on it :)

[Toni-SM]

Hi @noshaba

I added, in develop branch, examples for all Isaac Gym environments... Only the timesteps parameter needs to be adjusted for some environments (I am running all the original Isaac Gym examples to get this value).

You can see them, for now, in the skrl docs (develop branch): https://skrl.readthedocs.io/en/develop/intro/examples.html#learning-in-an-isaac-gym-environment-one-agent-multiple-environments

[noshaba]

Thank you so much!! Will check it out!! ^.^

[Toni-SM]

Now all parameters are set... in the skrl docs (develop branch)