NoduleNet CUDA out of memory when set net.use

CUDA out of memory when set net.use_mask = True at testing time

Open LXYTSOS opened this issue 4 years ago • 5 comments

no matter how many GPUs(11GB each) I use, I always get out of memory error when set net.use_mask = True at testing time, how can I solve this?

Jun 30 '20 09:06 LXYTSOS

Hi, I also have the same question, Do you have solved it?

Oct 29 '20 07:10 fengz63

no, I just give up.

Oct 29 '20 07:10 LXYTSOS

OK, thank you.

Oct 29 '20 07:10 fengz63

no matter how many GPUs(11GB each) I use, I always get out of memory error when set net.use_mask = True at testing time, how can I solve this?

@LXYTSOS What is the GPU usage rate when train.py is running? How about the GPU memory usage?

Oct 19 '21 09:10 dreamwish1998

@dreamwish1998 Hi, I also have the same question, Do you have solved it？

Mar 27 '23 07:03 chovvvvvy