pub fn expert(
x: &[f32],
w1: &[f32],
w2: &[f32],
w3: &[f32],
rows: usize,
dim: usize,
inter: usize,
swiglu_limit: f32,
weights: Option<&[f32]>,
) -> Vec<f32>Expand description
One SwiGLU expert. w1, w3: [inter, dim]; w2: [dim, inter]; all bf16 values.
x: [rows, dim] bf16 values; weights: per-row routing weight (f32) or None for the
shared expert. Returns [rows, dim] bf16 values.