Struct TernLayer 

pub struct TernLayer {
    pub name: String,
    pub scale: f32,
    pub storage: LayerStorage,
    pub sparsity: f64,
    pub original_dtype: String,
}

One compressed weight matrix (one transformer projection, etc.).

Fields

name: String

Layer name (e.g. “model.layers.0.self_attn.q_proj.weight”).

scale: f32

Per-layer scale α = mean(|W_original|).

storage: LayerStorage

Weight storage (dense or sparse).

sparsity: f64

Fraction of zero weights.

original_dtype: String

Original dtype (for metadata / roundtrip info).

Implementations

impl TernLayer

pub fn num_params(&self) -> usize

Total number of parameters in this layer.

pub fn memory_bytes(&self) -> usize

Memory used by ternary weights in bytes.

Trait Implementations

impl Clone for TernLayer

fn clone(&self) -> TernLayer

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for TernLayer

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

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for TernLayer

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

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for TernLayer

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

where __S: Serializer,

Serialize this value into the given Serde serializer.