Skip to content

[muon] Per-head Newton-Schulz for attention projections - #8384

Open
alanhuangyoo wants to merge 12 commits into
deepspeedai:masterfrom
alanhuangyoo:feat/per-head-muon
Open

[muon] Per-head Newton-Schulz for attention projections#8384
alanhuangyoo wants to merge 12 commits into
deepspeedai:masterfrom
alanhuangyoo:feat/per-head-muon

Consolidate the per-head Muon tests into two files

789a1e7
Select commit
Loading
Failed to load commit list.
DCO / DCO succeeded Sep 8, 2026 in 1s

DCO

All commits are signed off!

View more details on DCO