search

YU1ut / MixMatch-pytorch

Code for "MixMatch - A Holistic Approach to Semi-Supervised Learning"
MIT License
636 stars 129 forks source link

train_loss #40

Open Wanghd-MVP opened 3 years ago

Wanghd-MVP commented 3 years ago

why train_loss increase？is this normal？ image

khatrimann commented 1 year ago

I am also facing the same @hturki @jason718 @ustcjerry @dddzg