Ben

@DerThorsten on point 2, I think num_output > 1 is buggy and needs some tests added.

@matthiasplappert I need to rebase the tests if we end up using this parameterization. First need to confirm whether it fixes NaN issues or not. It seems to be learning...

@matthiasplappert I was thinking about making it an alternative. Need to test this versus the version with the moved epsilon. Seems to be fixing the NaNs but I want to...

Check sumo as to how to install traci module. There is also a branch which should support within process sumo, but that is a more complicated install. On Oct 29,...

Hi guys. I have a branch of gym I was using when I wrote this. I haven't gotten around to cleaning this up for current gym. You can checkout the...

I just checked it was gym 0.7.3 at the time. I'm sure a lot has changed since the beginning of the year. https://github.com/bstriner/gym/blob/bstriner/gym/version.py The `bstriner` branch combines some changes from...

@phausamann Interesting catch. Should be easy enough to fix if `None` means `stderr`. Just curious, why are you trying to deepcopy the callback? It is part of training, not part...

Definitely never seen that before. Can you give some details on your system (python, browser, etc.)? When are you getting the error? Maybe a screenshot could help. I'd be curious...

IIRC you can click save widgets or some similarly-labeled button. There is some combination of menu buttons that will get it to save. The normal autosave doesn't do the trick....

Text mode looks weird in IPython but it's really meant to be run on the command line. Uploaded an example to the README of what TQDMCallback does on the command...