From 949a71bf27ed4f18591727923cb241c2c93e39db Mon Sep 17 00:00:00 2001
From: Lucas Wilkinson <LucasWilkinson@users.noreply.github.com>
Date: Fri, 24 Jan 2025 15:20:59 -0500
Subject: [PATCH] [Bugfix][Kernel] FA3 Fix - RuntimeError: This flash attention
 build only supports pack_gqa (for build size reasons). (#12405)

Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>
---
 CMakeLists.txt | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/CMakeLists.txt b/CMakeLists.txt
index 2f9da6fa3e1d3..c954731bf94ef 100755
--- 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 0aff05f577e8a10086066a00618609199b25231d
+          GIT_TAG 9732b0ce005d1e6216864788502d5570004678f5
           GIT_PROGRESS TRUE
           # Don't share the vllm-flash-attn build between build types
           BINARY_DIR ${CMAKE_BINARY_DIR}/vllm-flash-attn