thuml/Universal-Domain-Adaptation

Question about classifier

tarun005 opened this issue · 1 comments

tarun005 commented

Shouldn't there be non-linearity between the bottleneck layers in the classifier?

Universal-Domain-Adaptation/net.py

Line 127 in 5d7caa9

self.main = nn.Sequential(self.bottleneck, self.fc, nn.Softmax(dim=-1))

youkaichao commented

That's a nice point. I suspect UAN can get better results with a non-linearity layer.