Struct MultiFidelityConfig 

pub struct MultiFidelityConfig {
    pub max_budget: f64,
    pub min_budget: f64,
    pub eta: usize,
    pub n_initial: usize,
}

Configuration for multi-fidelity optimization algorithms.

The budget parameters control how resources (e.g. training epochs) are allocated. eta is the reduction factor: at each successive halving round the number of surviving configurations is divided by eta while the per-configuration budget is multiplied by eta.

Fields

max_budget: f64

Maximum resource budget per configuration (e.g. epochs).

min_budget: f64

Minimum resource budget per configuration.

eta: usize

Reduction factor (default 3).

n_initial: usize

Number of initial configurations. When zero the algorithm computes a sensible default from the budget ratio and eta.

Implementations

impl MultiFidelityConfig

pub fn validate(&self) -> OptimizeResult<()>

Validate configuration parameters.

Trait Implementations

impl Clone for MultiFidelityConfig

fn clone(&self) -> MultiFidelityConfig

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for MultiFidelityConfig

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

Formats the value using the given formatter.

impl Default for MultiFidelityConfig

fn default() -> Self

Returns the “default value” for a type.