Floating point error (core dumped) in pytorch

When doing torchvision.transforms.Normalize(mean=[0.5], std=[0.5]),

get the error: 浮點數例外 (核心已轉儲) or Floating point error (core dumped) 

please check if the import image is integer type