Skip to content

Add CuTe DSL implementation of the swiglu liger kernel#1277

Open
Celaena24 wants to merge 3 commits into
linkedin:mainfrom
Celaena24:swiglu-cutedsl
Open

Add CuTe DSL implementation of the swiglu liger kernel#1277
Celaena24 wants to merge 3 commits into
linkedin:mainfrom
Celaena24:swiglu-cutedsl

Optimize using packed-f32x2 math, tvm ffi, and 128-bit vectorized loads

ef933d2
Select commit
Loading
Failed to load commit list.