Saumya-Shah
Saumya-Shah
I had the same problem. Just replace AT_CHECK and THCState_getCurrentStream(state) with TORCH_CHECK and at::cuda::getCurrentCUDAStream(), respectively and it works. https://github.com/hyangwinter/flownet3d_pytorch/issues/8#issuecomment-663925414
> > I had the same problem. > > Just replace AT_CHECK and THCState_getCurrentStream(state) with TORCH_CHECK and at::cuda::getCurrentCUDAStream(), respectively and it works. > > [hyangwinter/flownet3d_pytorch#8 (comment)](https://github.com/hyangwinter/flownet3d_pytorch/issues/8#issuecomment-663925414) > > Thank you,...
That's probably a common cuda error. See if [this](https://www.programmersought.com/article/8191430209/) solves your problem.