Struct QuantizationConfig 

pub struct QuantizationConfig {
    pub block_size: usize,
    pub double_quant: bool,
    pub compute_dtype: ComputeDType,
    pub strategy: QuantizationStrategy,
    pub use_zero_point: bool,
}

Configuration for quantization.

Fields

block_size: usize

Block size for quantization (number of values sharing a scale).

double_quant: bool

Whether to use double quantization (quantize the scales).

compute_dtype: ComputeDType

Data type for computation (usually bf16 or f16).

strategy: QuantizationStrategy

Quantization strategy (per-tensor or per-channel).

use_zero_point: bool

Whether to use zero-point quantization (asymmetric).

Trait Implementations

impl Clone for QuantizationConfig

fn clone(&self) -> QuantizationConfig

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for QuantizationConfig

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

Formats the value using the given formatter.

impl Default for QuantizationConfig

fn default() -> QuantizationConfig

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for QuantizationConfig

fn deserialize<__D>( __deserializer: __D, ) -> Result<QuantizationConfig, <__D as Deserializer<'de>>::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for QuantizationConfig

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

where __S: Serializer,

Serialize this value into the given Serde serializer.