Skip to content

[Bugfix][Metrics][MFU] Bill linear-attention layers of hybrid models separately - #58836

Open
jayzuccarelli wants to merge 2 commits into
vllm-project:mainfrom
jayzuccarelli:mfu-linear-attention
Open

jayzuccarelli wants to merge 2 commits into
vllm-project:mainfrom
jayzuccarelli:mfu-linear-attention