Antonin RAFFIN
Antonin RAFFIN
Good to hear =) then i would be happy to receive a PR that solves this issue ;)
Hello, > Could you elaborate why do you think, that removing the probs is not a good idea? The idea behind it is to use a feature that is in...
Hello, are you proposing to implement it or it is a request? If it is a request, please add a bit more motivation of why this algorithm is needed, otherwise...
Hello, this is a duplicate of https://github.com/Stable-Baselines-Team/stable-baselines3-contrib/issues/49#issuecomment-957629253 we would apppreciate a PR that solves this issue ;)
> At the very least, there should be a warning about this as it is a very different value from what you usually use yes, I would be for a...
Hello, I've got mixed feeling about that one. It would be interesting to have it, at least in a separate repo. However, the current focus of SB3 is model free...
closing as out of scope of SB3 (we focus on model-free RL), however, if you want a model-free RL algorithm that is as sample efficient as MBPO, you can take...
Hello, > model we obtain valid actions but during evaluation we obtain actions of 0. Both the evaluation and train environment are the same except for using different but similar...
Hello, Can you elaborate a bit more, what do you want to implement exactly? Where should it be included? Which algorithms do you plan to support?
>The reason for applying the augmentation prior to the network instead of as a wrapper is to make best use of multiple pass-throughs of the data. e.g. instead of storing...