Enum WeightQuantization 

pub enum WeightQuantization {
    Affine(AffineQuantization),
    MxFp4,
    GgufIQuant {
        ggml_type: GgmlType,
        endian: Endian,
    },
}

Packed physical encoding of a model weight. Dense storage is represented by None at the use site.

Variants

Affine(AffineQuantization)

Per-group affine integer storage.

MxFp4

Microscaling FP4 with E2M1 values and E8M0 scales.

GgufIQuant

Checkpoint-native GGML blocks.

Fields

ggml_type: GgmlType

Native GGML tensor encoding.

endian: Endian

Byte order declared by the GGUF container.

Implementations

impl WeightQuantization

pub const MXFP4_GROUP_SIZE: i32 = 32

MXFP4 group size fixed by the format.

pub const MXFP4_BITS: i32 = 4

MXFP4 packed value width fixed by the format.

pub fn group_size(self) -> i32

Returns the grouping used by packed execution.

pub fn bits(self) -> i32

Returns the packed storage width.

pub const fn has_biases(self) -> bool

Returns whether the encoding stores affine bias companions.

pub const fn gguf_iquant(self) -> Option<(GgmlType, Endian)>

Returns checkpoint-native GGML metadata, when present.

pub fn validate(self) -> Result<(), Error>

Validates portable storage geometry.

Trait Implementations

impl Clone for WeightQuantization

fn clone(&self) -> WeightQuantization

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for WeightQuantization

impl Debug for WeightQuantization

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

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for WeightQuantization

fn deserialize<D>(deserializer: D) -> Result<Self, D::Error>

where D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Eq for WeightQuantization

impl From<AffineQuantization> for WeightQuantization

fn from(value: AffineQuantization) -> Self

Converts to this type from the input type.

impl From<WeightQuantization> for LinearFormat

fn from(value: WeightQuantization) -> Self

Converts to this type from the input type.

impl PartialEq for WeightQuantization

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

Equality operator ==.

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

Inequality operator !=.

impl Serialize for WeightQuantization

fn serialize<S>(&self, serializer: S) -> Result<S::Ok, S::Error>

where S: Serializer,

Serialize this value into the given Serde serializer.

impl StructuralPartialEq for WeightQuantization