xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-07-10 05:07:13 +08:00
Code Issues Packages Projects Releases Wiki Activity

[rocm] Fix wrong attention log (#18764)

Browse Source 
Signed-off-by: Felix Marty <felmarty@amd.com>
This commit is contained in:
fxmarty-amd 2025-05-28 04:45:41 +02:00 committed by GitHub
parent d73a9457a5
commit 794ae1f551
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 3 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
vllm/platforms/rocm.py
View File

@ -194,8 +194,9 @@ class RocmPlatform(Platform):
f" The selected backend, {selected_backend.name}," f" The selected backend, {selected_backend.name},"
f"is not MLA type while requested for MLA backend.") f"is not MLA type while requested for MLA backend.")
selected_backend = (_Backend.ROCM_FLASH if selected_backend if selected_backend is None or selected_backend == _Backend.FLASH_ATTN:
== _Backend.FLASH_ATTN else selected_backend) selected_backend = _Backend.ROCM_FLASH
if envs.VLLM_USE_V1: if envs.VLLM_USE_V1:
logger.info("Using Triton Attention backend on V1 engine.") logger.info("Using Triton Attention backend on V1 engine.")
return ("vllm.v1.attention.backends." return ("vllm.v1.attention.backends."