shouldn't group in_channels inside 1x1 conv
wuyuanhang94 opened this issue · comments
wuyuanhang94 commented
nn.Conv2d(in_channels, C * D, kernel_size=1, groups=C, bias=False),
doesn't make sense.
yukiTemp commented
yeah, I was confused by this too, it's inconsistent to the paper.
Quantum_Rich commented
Has this issue been confirmed or fixed?