@tduncan / Nvidia-Optimization-Stack-VLLM Mirror · read-only / commits / 8c1e77fb58
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, 05:45 UTC

[Kernel] Update vllm-flash-attn version to reduce CPU overheads (#10742)

Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>

Woosuk Kwon committed Nov 28, 2024 at 08:31 UTC 8c1e77fb585c4f42783a3d88c1efc7c9e15fd89f
1 file changed +1 -1
CMakeLists.txt
+1 -1
@@ -522,7 +522,7 @@ else()
522 FetchContent_Declare(
523 vllm-flash-attn
524 GIT_REPOSITORY https://github.com/vllm-project/flash-attention.git
525 - GIT_TAG d886f88165702b3c7e7744502772cd98b06be9e1
525 + GIT_TAG fdf6d72b48aea41f4ae6a89139a453dae554abc8
526 GIT_PROGRESS TRUE
527 # Don't share the vllm-flash-attn build between build types
528 BINARY_DIR ${CMAKE_BINARY_DIR}/vllm-flash-attn