Add Maximal Update Parametrization (μP) #45848
Maximal Update Parametrization
6fde9cbd
trainer integration
9a8dea16
style
a71f5f6d
simplify
b3bf576f
Merge branch 'main' into mup
3289ae13
Merge branch 'main' into mup
99224d05
Merge branch 'main' into mup
da8150f9
Assignees
No one assigned
Login to write a write a comment.
Login via GitHub