This codebase is the official implementation of Test-Time Classifier Adjustment Module for Model-Agnostic Domain Generalization (NeurIPS2021, Spotlight)
MIT License
93
stars
10
forks
source link
How can you inference on the weights of the classifier? #4
How are you able to inference on the weights of the classifier using the classifier? That is like saying f( f.weights() ). That doesn't make sense. Can you please explain it to me?
In the constructor for T3A, you write
warmup_supports = self.classifier.weight.dataself.warmup_supports = warmup_supportswarmup_prob = self.classifier(self.warmup_supports)How are you able to inference on the weights of the classifier using the classifier? That is like saying f( f.weights() ). That doesn't make sense. Can you please explain it to me?