Skip to main content

compute_q_params

burn::backend::quantization

Function compute_q_params 

pub fn compute_q_params<B>(
    scheme: &QuantScheme,
    min: <B as BackendTypes>::FloatTensorPrimitive,
    max: <B as BackendTypes>::FloatTensorPrimitive,
) -> QuantizationParametersPrimitive<B>
where
    B: Backend,
Expand description

Compute the quantization parameters.