Function quantize_tensor 

pub fn quantize_tensor<S: Scale, const BITS: u32>(
    values: &[f32],
) -> Result<Quantized<S>>

Quantize the entire tensor with a single scale.