mhubii / ppo_libtorch

C++ implementation of Proximal Policy Optimization
71 stars 16 forks source link

this->is_training() not working #2

Closed N1ckfm closed 2 years ago

N1ckfm commented 2 years ago

Hi, when I test in models.h if(this->is_training()){} ever is true, but execute the "else", I was thinking maybe is for the version of libtorch, I am using 1.2.0 I will appreciate you indicate which version are you using and if you have one idea why not change the is_training() thanks

mhubii commented 2 years ago

hm thats strange, just tested and seems to work fine for the latest release, see https://pytorch.org/cppdocs/installing.html