Problem about WGAN

I run WGAN’s code


I get this bug:
in the line 184
input.resize_as_(real_cpu).copy_(real_cpu)
TypeError: resize_as_ received an invalid combination of arguments - got (!torch.FloatTensor!), but expected (torch.cuda.FloatTensor template)
Is any body have the same problem? How to solve this?

this has been fixed in the latest version of the code.

your meaning the fixed problem is WGAN code or pytorch?

I solve this problem!
Thanks very much!