fix type error in gpu mode

The-Learning-And-Vision-Atelier-LAVA / SOF-VSR

[ACCV 2018 & TIP 2020] Deep Video Super-resolution using HR Optical Flow Estimation

169 stars 43 forks source link

Open FeiZhu opened 5 years ago

FeiZhu commented 5 years ago

fix "TypeError: can't convert CUDA tensor to numpy. Use Tensor.cpu to copy the tensor to host memory first" error in gpu mode