search

Jannoshh / simple-sam

Sharpness-Aware Minimization for Efficiently Improving Generalization
MIT License
41 stars 9 forks source link

Change `grad_norm` to use `gradients` #3

Closed SimonBiggs closed 3 years ago

Jannoshh commented 3 years ago

Thank you. Surprised my version still increased the accuracy.

SimonBiggs commented 3 years ago

Thank you.

My pleasure :slightly_smiling_face:.