mlpack
blog
|
Proximal Policy Optimization - Week 8
Proximal Policy Optimization - Week 8
This week, I finished all functionality of PPO, such as update the actor and critic model. Synchronizing the oldActorNetwork. I made the code consistent with the latest code which modifies the Pendulum environment. I fixed some bugs in the Step
and Update
function. But there are some bugs like memory access violation
need to fix. Next week, I will focus on those bugs and tuning the parameter to make it work.
Thanks for reading :).
Generated by 1.8.13