@tduncan / Nvidia-Optimization-Stack-VLLM Mirror · read-only / commits / ef533d25fb
Fork 0

A high-throughput and memory-efficient inference and serving engine for LLMs

Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git

Synced Nov 16, 00:45 UTC

[Bugfix] FA2 illegal memory access (#12848)

Lucas Wilkinson committed Feb 6, 2025 at 22:54 UTC ef533d25fba4b5ef8b4da9369de718c0773b9bce
1 file changed +1 -1
CMakeLists.txt
+1 -1
@@ -581,7 +581,7 @@ else()
581 FetchContent_Declare(
582 vllm-flash-attn
583 GIT_REPOSITORY https://github.com/vllm-project/flash-attention.git
584 - GIT_TAG d4e09037abf588af1ec47d0e966b237ee376876c
584 + GIT_TAG 720c94869cf2e0ff5a706e9c7f1dce0939686ade
585 GIT_PROGRESS TRUE
586 # Don't share the vllm-flash-attn build between build types
587 BINARY_DIR ${CMAKE_BINARY_DIR}/vllm-flash-attn