Enum QuantizationStrategy 

pub enum QuantizationStrategy {
    PerTensorSymmetric(SymmetricQuantization<f32>),
    PerBlockSymmetric(Vec<SymmetricQuantization<f32>>, BlockSize),
}

Quantization strategy.

Variants

PerTensorSymmetric(SymmetricQuantization<f32>)

Per-tensor symmetric quantization.

PerBlockSymmetric(Vec<SymmetricQuantization<f32>>, BlockSize)

Per-block symmetric quantization.

Implementations

impl QuantizationStrategy

pub fn quantize(&self, values: &[f32]) -> Vec<i8>

Quantize the values to a lower precision data type.

pub fn dequantize(&self, values: &[i8]) -> Vec<f32>

Dequantize the values to a higher precision data type.

Trait Implementations

impl Clone for QuantizationStrategy

fn clone(&self) -> QuantizationStrategy

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for QuantizationStrategy

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for QuantizationStrategy

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl PartialEq for QuantizationStrategy

fn eq(&self, other: &QuantizationStrategy) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl Serialize for QuantizationStrategy

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl Eq for QuantizationStrategy

impl StructuralPartialEq for QuantizationStrategy