All process update the network and then sync the grad?
nizhihao opened this issue · 1 comments
nizhihao commented
Hi, I have the doubt.
In this distributed RL, because of OS scheduling, every process will have the near state and do the same thing?
so all process will update the network and then sync the grad together? like sync algorithm A2C？
thanks very much.
TianhongDai commented
Yes, it can be refereed as using a very large batch size for the training.