search

aidudezzz / deepbots

A wrapper framework for Reinforcement Learning in the Webots robot simulator using Python 3.
https://deepbots.readthedocs.io/
GNU General Public License v3.0
236 stars 50 forks source link

StableBaselines Intergration #42

Closed MentalGear closed 2 years ago

MentalGear commented 4 years ago

Hey guys, Pretty amazing work here. Actually found this only a day before having completed my own gym & stableBaselines integration with Webots. But yours is a nice package format, so I actually prefer this I guess. Alright, here's my question: Do you guys are already working on adding stableBaselines to deepbots, or should I give it a shot ?

eakirtas commented 4 years ago

Hello @MentalGear,

Thanking for opening this issue. Indeed, integrating stableBaseline is one of our future plans. However, this isn't something that we are currently working on. It is more than welcome to give it a shot and we will be happy to help you. I feel that it will be an awesome enchantment.

It seems, with a quick look, that the compatibility is broken because that the SupervisorEnv doesn't have some extra attributes OpenGym uses (ex. action_space, observation_space) which they are necessary to the stableBaselines (for example in DDPG). I feel that this is a good staring point.

You can use the deepworlds repository which has already some implemented examples to test your code and feel free to contacting us, open PRs, ask for help or request a code review.

cheers

eakirtas commented 3 years ago

Hello @MentalGear, stable baselines can be integrated using the latest-dev version of deepbots. We will happy to hear if that works for you.

Any feedback will be appreciated!

eakirtas commented 3 years ago

@all-contributors please add @MentalGear for ideas

allcontributors[bot] commented 3 years ago

@ManosMagnus

I've put up a pull request to add @MentalGear! :tada:

tsampazk commented 2 years ago

Issue is a duplicate of #85