Skip to content

Optimize rmsnorm/layernorm to get better performance than aiter/triton#610

Open
cschenjunlin wants to merge 3 commits into
mainfrom
cjl/norm_optimization
Open

Optimize rmsnorm/layernorm to get better performance than aiter/triton#610
cschenjunlin wants to merge 3 commits into
mainfrom
cjl/norm_optimization