the coefficient scheduler for gradient reversal module

thuml / Universal-Domain-Adaptation

Code release for Universal Domain Adaptation(CVPR 2019)

272 stars 52 forks source link

Closed cmhungsteve closed 5 years ago

cmhungsteve commented 5 years ago

self.grl = GradientReverseModule(lambda step: aToBSheduler(step, 0.0, 1.0, gamma=10, max_iter=10000))

It seems that max_iter is set as 10000.

From the DANN paper, I thought it depends on the total iteration number during training. Is there any reason to use a fixed number?

Thank you.

youkaichao commented 5 years ago

Because the total iterations in our experiments are basically the same.