The Changes Listed Above May be useful

[lionyouko]

While I still didn't upload the entire example I created and trained using omnet++, I would like to give the following suggestions:

1. To make known that one can use Stable baselines 3 easily with veinsgym framework, hence using state-of-art agents.
2. To alert about the use for multi-agent or federated learning (while I am working on it and I may be able to help regarding it in future). I could not link Issue 5 or the .py, that can be done.
3. To change the code just a bit, being the float(reward[0]) an easy fix for an issue raised from check_env() from stable_baselines3 regarding reward not being float even though it was declared as one in StepResult class.
4. To provide small example agent with some annotations to help who is going to use.

Hope it helps a bit. Dr Dbuse may want to tell me if pull request is the right way to contribute.

[dbuse]

Hi @lionyouko

Thanks for sharing. I might cherry-pick some of the contents later. But I do not want to bind VeinsGym too tightly to a certain RL library such as StableBaselines. So I'll close this for now and potentially come back later.