DP-μP transfers learning rates under differential privacy
Scaling rules cover DP-SGD, DP-Adam, and DP-Muon on MLP, ViT, and GPT-2.
1 stories, newest first · one card per story
Scaling rules cover DP-SGD, DP-Adam, and DP-Muon on MLP, ViT, and GPT-2.
Today 1.6K readers · 2.4K views · 774 reads · all time 189K views