perf(GDN-MOE-BF16-OUT): the GDN recurrence output and the z gate are bf16 on MoE too, as both references keep them - #1223

perf(GDN-MOE-BF16-OUT): the GDN recurrence output and the z gate are bf16 on MoE too, as both references keep them#1223

localai-bot merged 5 commits into

mainmudler/vllm.cpp:mainfrom

row/GDN-MOE-BF16-OUT-IMPLmudler/vllm.cpp:row/GDN-MOE-BF16-OUT-IMPLCopy head branch name to clipboard