vllm
1fd6337e - [Bugfix] Keep legacy WNA16 MoE dispatch on Triton under batch invariance (#61038)

Commit
24 days ago
[Bugfix] Keep legacy WNA16 MoE dispatch on Triton under batch invariance (#61038) Signed-off-by: Yifan Feng <evanfengff@gmail.com> Signed-off-by: yifanFengg <yifanfen@amazon.com> Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com> Co-authored-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com> Co-authored-by: vllm-agent <claw@inferact.ai>
Author
Parents
Loading