def lf_mulb
import bendtorch-gpu@1.3.0.0/main.bend as BendtorchGpuimport Base
import bend-ml-tensor-gpu@0.1.5.0/main.bend as TG
128def lf_mulb(+t: U32, +l: U32, +e: EN, x1: Array<F32>, x2: Array<F32>, x3: Array<F32>, x4: Array<F32>, o: TG.O2) -> TG.O2:
129 match e o:
130 case EN{+n} TG.O2{a, b}: mulb_loop(TG.share(t, n, l), t, l, Array.get(F32, x1, t), Array.get(F32, a, t), Array.get(F32, b, t))