Antonin RAFFIN
Antonin RAFFIN
Looking at all the issues about LSTM: https://github.com/hill-a/stable-baselines/issues?utf8=%E2%9C%93&q=is%3Aissue+is%3Aopen+LSTM I think we need to gather them and close the duplicated ones.
>A PR to add this and update env_checker would be welcomed :) It would be a warning as it is specific to Stable Baselines. But yes, I forgot that one.
Hello, Two things: - please format your issue using code blocks for the equations (cf issue template) - as we told you in the email, we did not write that...
Hello, Please fill in the PR template completely.
Hello, I tested your code and ... it worked fine. See below for minimal code to reproduce (I got reward > 100) ```python import tensorflow as tf from stable_baselines import...
Maybe related to #351
Hello, > are features for the full observation-dictionary yes, this was a choice (to makes things simpler). Please tell us if it makes a difference.
Hello, I would be in favor of that minor change (unless another maintainer objects) and would appreciate a PR that solves this issue ;)
>I would expect the loss to become smaller as the DQN gets better at predicting Q-values (and therefore achieving better episode reward). Am I missing something :) ? Please open...
Thank you for the proposition. >When would be a good moment to attack this issue? I think you can almost already start. There is only two "big" PR currently: -...