Struct OptimizerMetrics 

pub struct OptimizerMetrics {
    pub name: String,
    pub step_count: u64,
    pub total_step_time: Duration,
    pub avg_step_time: Duration,
    pub current_learning_rate: f64,
    pub gradient_stats: GradientStatistics,
    pub parameter_stats: ParameterStatistics,
    pub convergence: ConvergenceMetrics,
    pub memory_usage: usize,
}

Optimizer performance metrics

Tracks key performance indicators for optimizer operations including step timing, gradient statistics, parameter updates, and convergence.

Fields

name: String

Optimizer name

step_count: u64

Total number of optimization steps

total_step_time: Duration

Total time spent in optimization steps

avg_step_time: Duration

Average time per step

current_learning_rate: f64

Current learning rate

gradient_stats: GradientStatistics

Gradient statistics

parameter_stats: ParameterStatistics

Parameter statistics

convergence: ConvergenceMetrics

Convergence metrics

memory_usage: usize

Memory usage (bytes)

Implementations

impl OptimizerMetrics

pub fn new(name: impl Into<String>) -> Self

Create new metrics for an optimizer

pub fn update_step<A: Float>( &mut self, step_duration: Duration, learning_rate: f64, gradients: &ArrayView1<'_, A>, params_before: &ArrayView1<'_, A>, params_after: &ArrayView1<'_, A>, )

Update metrics after an optimization step

pub fn throughput(&self) -> f64

Get throughput (steps per second)

pub fn reset(&mut self)

Reset all metrics

Trait Implementations

impl Clone for OptimizerMetrics

fn clone(&self) -> OptimizerMetrics

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for OptimizerMetrics

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

Formats the value using the given formatter.