questions of code details

[dengwanxia1991]

Thanks for sharing the code. I have a question. In the TarDisClusterLoss, why is the true label of the target used in the 141 line?

[huitangtang]

Hello, glad to hear from you. For your problem, note that the target domain label was pseudo-labeled by K-means clustering in line 145 in main.py. Please check the corresponding function in trainer.py. We do this mainly for initialization at the first epoch and we set all target labels to -1 in line 115 to make sure no use of ground truth target labels.

[dengwanxia1991]

Hello, glad to hear from you. For your problem, note that the target domain label was pseudo-labeled by K-means clustering in line 145 in main.py. Please check the corresponding function in trainer.py. We do this mainly for initialization at the first epoch and we set all target labels to -1 in line 115 to make sure no use of ground truth target labels.

Thanks for the reply and solving my question.