apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2026-04-23 00:17:25 +08:00
Code Issues Actions 19 Packages Projects Releases Wiki Activity

check init_flash_attn_version log (#7399)

Browse Source
This commit is contained in:
chen
2026-04-15 11:05:10 +08:00
committed by GitHub
parent 13b9fe7299
commit 616b29ce08
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
fastdeploy/model_executor/layers/attention/flash_attn_backend.py
+1 -1
View File
@@ -95,7 +95,7 @@ def init_flash_attn_version():
logger.info(f"The current platform[sm{get_sm_version()}] can't import Flash Attention V4.")
if FLASH_ATTN_VERSION is None:
if sm_version >= 89 and any(num >= 89 for num in paddle.version.cuda_archs()):
if sm_version == 90 and 90 in paddle.version.cuda_archs():
FLASH_ATTN_VERSION = 3
logger.info("The current platform supports Flash Attention V3.")
else: