Layerwise Repacking: Variable-Length Hybrid-Attention Training without an Extra Redistribution Pass
Paper • 2610.35989 • Published • 1
zhangjiahao
wuqiao123
AI & ML interests
None yet
Recent Activity
upvoted a paper about 15 hours ago
Layerwise Repacking: Variable-Length Hybrid-Attention Training without an Extra Redistribution Pass upvoted a paper about 16 hours ago
KPage: Cross-Block Page Compaction for Hardware-Efficient Sparse AttentionOrganizations
None yet