Function quantized_matmul 

pub fn quantized_matmul(x: &Array, w: &Quantized, transpose: bool) -> Array

x · dequantize(w)ᵀ (or without the transpose); see Array::quantized_matmul.