xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2025-12-26 06:49:38 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix amd triton mla path (#17871)

Browse Source
This commit is contained in:
Shiyan Deng 2025-05-11 00:53:31 -07:00 committed by GitHub
parent 9112155283
commit eea22a56ab
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
vllm/attention/backends/mla/common.py
View File

@ -1063,7 +1063,7 @@ class MLACommonImpl(MLAAttentionImpl[T], Generic[T]):
softmax_scale,
None, # bias
)
if is_vllm_fa:
elif is_vllm_fa:
attn_out = self.flash_attn_varlen_func(
q=q,
k=k,