Struct NaturalGradientConfig 

pub struct NaturalGradientConfig<T: Float> {
    pub learning_rate: T,
    pub max_iterations: usize,
    pub gradient_tolerance: T,
    pub parameter_tolerance: T,
    pub fisher_regularization: T,
    pub use_line_search: bool,
    pub line_search_beta: T,
    pub line_search_alpha: T,
}

Configuration for natural gradient optimization

Fields

learning_rate: T

Learning rate (step size)

max_iterations: usize

Maximum number of iterations

gradient_tolerance: T

Convergence tolerance for gradient norm

parameter_tolerance: T

Convergence tolerance for parameter changes

fisher_regularization: T

Regularization parameter for Fisher information matrix

use_line_search: bool

Use line search for adaptive step size

line_search_beta: T

Line search backtracking factor

line_search_alpha: T

Line search initial step scaling

Trait Implementations

impl<T: Clone + Float> Clone for NaturalGradientConfig<T>

fn clone(&self) -> NaturalGradientConfig<T>

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl<T: Debug + Float> Debug for NaturalGradientConfig<T>

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

Formats the value using the given formatter.

impl<T: Float> Default for NaturalGradientConfig<T>

fn default() -> Self

Returns the “default value” for a type.