class Contextual Attention F.conv_transpose2d problem

[kangjh0226]

Thanks for posting this code!
I'm trying to change original generative-inpainting code to pytorch too, your version help me a lot!
I have some problem on here

generative-inpainting-pytorch/model/networks.py

Line 323 in c6cdaea

yi = F.conv_transpose2d(yi, wi_center, stride=self.rate, padding=1) / 4. # (B=1, C=128, H=64, W=64)

the output error is:
RuntimeError: Given transposed=1, weight of size [128, 1024, 4, 4], expected input[1, 1024, 32, 32] to have 128 channels, but got 1024 channels instead

Do you have any experience before?
Thank you!!