HuiZeng/Grid-Anchor-based-Image-Cropping

Overfitting

andriyrizhiy opened this issue · 3 comments

I trying to implement your paper in tensorflow and run experiments in less models, but i get overfitting in training set. You encountered this problem in your experiment (I did not find this in the paper)?

Can you provide more information about overfitting? You mean the training loss is smaller than the validatation/testing loss?

yes, my training loss smaller and decrease and val loss big and increase, and ACC4/5 in training increases but val is staying in place

Please refer to issue #14