Some questions for your code

[sudalvxin]

Hi, thanks for your codes. I have some small questions.
Q1. train.py: the 161th row: cls_loss = F.cross_entropy(outputs, target, reduction="none", ignore_index=-1).mean(). The outputs contains labled data and unabled data.
I think it may be inappropriate.
Q2. which way you use to calculate the loss in pseudo_label.py

[perrying]

The outputs contain labeled data and unlabeled data.

yes, but the targets of unlabeled data are given -1 and it's ignored by ignore_index option. So, cls_loss is calculated with only labeled data.

which way you use to calculate the loss in pseudo_label.py

cross entropy between pseudo labels and outputs

[sudalvxin]

Thanks for your replay, but the .mean() will calculate the mean values of 100 rather than 50

[sudalvxin]

Besides, in pseudo_label.py, I think only the unlabeled samples are needed for calculating the loss.

[perrying]

but the .mean() will calculate the mean values of 100 rather than 50

Original implementation also calculate the loss with the same way.

I think only the unlabeled samples are needed for calculating the loss.

You are right, my implementation is wrong.
I'll fix it. Thanks.