[Bugfix][Kernel] Fix perf regression caused by PR #12405 (#12434)

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>

[Bugfix][Kernel] Fix perf regression caused by PR #12405 (#12434)
Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>
68f11149 · Lucas Wilkinson · GitHub · 72f48804 · 68f11149
Unverified Commit 68f11149 authored Jan 26, 2025 by Lucas Wilkinson Committed by GitHub Jan 26, 2025
Hide whitespace changes
Inline Side-by-side

Showing with 1 addition and 1 deletion

CMakeLists.txt CMakeLists.txt +1 -1

No files found.
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -576,7 +576,7 @@ else()
  FetchContent_Declare(
          vllm-flash-attn
          GIT_REPOSITORY https://github.com/vllm-project/flash-attention.git
-          GIT_TAG 9732b0ce005d1e6216864788502d5570004678f5
+          GIT_TAG d4e09037abf588af1ec47d0e966b237ee376876c
          GIT_PROGRESS TRUE
          # Don't share the vllm-flash-attn build between build types
          BINARY_DIR ${CMAKE_BINARY_DIR}/vllm-flash-attn