biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

Separate MLAAttention class from Attention (#25103)

Browse Source 
Signed-off-by: Naveenraj Kamalakannan <therealnaveenkamal@gmail.com>
Signed-off-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
This commit is contained in:
Naveenraj Kamalakannan
2025-10-08 20:11:11 -04:00
committed by GitHub
parent 2a03f93de9
commit e614ab7806
10 changed files with 502 additions and 163 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
vllm/config/compilation.py
View File

@@ -380,6 +380,8 @@ class CompilationConfig:
_attention_ops: ClassVar[list[str]] = [
"vllm.unified_attention",
"vllm.unified_attention_with_output",
"vllm.unified_mla_attention",
"vllm.unified_mla_attention_with_output",
"vllm.mamba_mixer2",
"vllm.mamba_mixer",
"vllm.short_conv",