Open luoww1992 opened 3 years ago
Looks like this is more a libtorch issue rather than our model issue. Please try run other models to make sure it is not your environment problem.
@PeterL1n i run it in gpu with c++ sucessfully,but it is slow. i am try to do it. how many fps you test in c++? what should we do to get high fps in c++ gpu with 1920*1080 ?
my env: win10 vs2019 libtorch 1.7.1 -cuda101-debug cuda101
i can run it by change the device in torch::Device("cpu"); but how to use it in gpu? How did you do it in gpu with torchscript model?