Skip to content

Draft: Linear attention support for KVCacheManager #12185

Draft
VALLIS-NERIA wants to merge 10 commits intoNVIDIA:mainfrom
VALLIS-NERIA:user/xiweny/linear_reuse_new
Draft

Draft: Linear attention support for KVCacheManager #12185
VALLIS-NERIA wants to merge 10 commits intoNVIDIA:mainfrom
VALLIS-NERIA:user/xiweny/linear_reuse_new

Commits

Commits on Mar 6, 2026

Commits on Mar 9, 2026

Commits on Mar 13, 2026