Updating class_loss and consistent_loss at the same time?

LiJunnan1992 / MLNT

Meta-Learning based Noise-Tolerant Training

123 stars 29 forks source link

Updating class_loss and consistent_loss at the same time? #6

Open htmap opened 4 years ago

htmap commented 4 years ago

Dear author. Why are you updating class_loss and consistent_loss at the same time? In Algorithm1, it seems that its processing is decoupled. I'm sorry if I have misunderstood.