[2/N][Attention] Enable masked MHA for sparse MLA prefills (#48770)

Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
Co-authored-by: OpenAI Codex <noreply@openai.com>
This commit is contained in:
Matthew Bonanni
2026-07-31 06:01:24 -07:00
committed by GitHub
co-authored by OpenAI Codex
parent 864a87febb
commit 82ae4164ee
11 changed files with 855 additions and 26 deletions
@@ -39,7 +39,7 @@ else()
FetchContent_Declare(
vllm-flash-attn
GIT_REPOSITORY https://github.com/vllm-project/flash-attention.git
GIT_TAG ed4b7342bc8f0489dd9b649d5288867e35fc6a32
GIT_TAG ec4a882d70b19235c824848db944d84a6cb8f9ee
GIT_PROGRESS TRUE
# Don't share the vllm-flash-attn build between build types
BINARY_DIR ${CMAKE_BINARY_DIR}/vllm-flash-attn