Michael Müller
Sep 1, 2018 · 1 min read

Woah, this is great, how is this not getting more attention! Thank you! I was able to reproduce this till 02_play_steps, which gave already a nice speed improvement (maybe I have to learn Pytorch after all, combined with keras!), but sadly for anything higher (≥03) I get an error “line=253 error=63 : OS call failed or operation not supported on this OS”, it appears that pytorch doesn’t support cuda on windows for asynchronos methods, any idea how to solve this? Thank you very much for your great work so far, FG Mike