softlearning icon indicating copy to clipboard operation
softlearning copied to clipboard

Published 20 hours ago • verified publisher icon rail-berkeley

using deterministic policy in enviroment like lunarlander?

Open Elktrn opened this issue 2 years ago • 0 comments

Hi and thank you for such a genius algorithm. I wonder how by using mu of gaussian policy in sac in enviroments like lunar lander is it guranteed to converge cuz i see some trials fails to converges. specialy on lunar lander and humnoid v3

Elktrn avatar Jun 07 '22 10:06 Elktrn