Enum ImportanceMethod 

pub enum ImportanceMethod {
    L1Norm,
    L2Norm,
    GradientWeighted,
    TaylorExpansion,
    Random {
        seed: u64,
    },
    Movement,
    Fisher,
}

Importance score computation method for unstructured pruning

Variants

L1Norm

L1 norm (absolute value): |w|

L2Norm

L2 norm (squared magnitude): w^2

GradientWeighted

Gradient-weighted importance: |w * g|

TaylorExpansion

Taylor expansion: |w * g * a| or first-order Taylor approximation

Random

Random scores (baseline for comparison)

Fields

seed: u64

Seed for reproducibility

Movement

Movement-based: tracks weight changes during training

Fisher

Fisher information approximation: g^2 (gradient squared)

Trait Implementations

impl Clone for ImportanceMethod

fn clone(&self) -> ImportanceMethod

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for ImportanceMethod

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

Formats the value using the given formatter.

impl Default for ImportanceMethod

fn default() -> ImportanceMethod

Returns the “default value” for a type.

impl PartialEq for ImportanceMethod

fn eq(&self, other: &ImportanceMethod) -> 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 Copy for ImportanceMethod

impl StructuralPartialEq for ImportanceMethod