Struct QuantumLearnedOptimizer 

pub struct QuantumLearnedOptimizer<T>
where
    T: Float + ScalarOperand + Debug + Send + Sync + 'static,
{ /* private fields */ }

A learned-interface adapter over the quantum-inspired core optimizers.

QuantumLearnedOptimizer wraps one of the QuantumBackend variants and implements AdvancedOptimizer, tracking the step count and an exponential moving average of the gradient L2 norm so the learned stack can query optimizer health via AdvancedOptimizer::get_state.

Implementations

impl<T> QuantumLearnedOptimizer<T>

where T: Float + ScalarOperand + Debug + Send + Sync + 'static,

pub fn annealing(learning_rate: T) -> Self

Create a quantum-annealing-backed learned optimizer.

The learning_rate scales the perturbation magnitude of the underlying QuantumAnnealing optimizer.

pub fn variational(learning_rate: T) -> Self

Create a variational-quantum (SPSA) backed learned optimizer.

The learning_rate is the SPSA gain numerator a of the underlying VariationalQuantumOptimizer.

pub fn hybrid(learning_rate: T, switch_step: usize) -> Self

Create a hybrid quantum-classical backed learned optimizer.

The first switch_step updates are produced by quantum annealing; later updates are produced by the classical Adam refinement phase of the underlying HybridQuantumClassical optimizer.

pub fn norm_ema_decay(&self) -> T

Returns the decay coefficient used for the gradient-norm EMA.

pub fn set_norm_ema_decay(&mut self, decay: T) -> Result<()>

Override the gradient-norm EMA decay coefficient.

Errors

Returns OptimError::InvalidConfig unless decay lies in the open interval (0, 1); a decay outside that range would make the EMA either ignore new observations entirely or diverge.

pub fn backend(&self) -> &QuantumBackend<T>

Immutable access to the underlying backend (useful for diagnostics).

pub fn step_count(&self) -> usize

Current step count.

Trait Implementations

impl<T> AdvancedOptimizer<T> for QuantumLearnedOptimizer<T>

where T: Float + ScalarOperand + Debug + Send + Sync + 'static,

fn step( &mut self, params: &Array1<T>, gradients: &Array1<T>, ) -> Result<Array1<T>>

Perform one optimization step, returning the updated parameters.

fn get_learning_rate(&self) -> T

Get the current effective learning rate.

fn set_learning_rate(&mut self, lr: T)

Set the base learning rate.

fn name(&self) -> &str

Get the optimizer name.

fn get_state(&self) -> OptimizerStateInfo<T>

Get a snapshot of the optimizer’s internal state.

impl<T> Debug for QuantumLearnedOptimizer<T>

where T: Float + ScalarOperand + Debug + Send + Sync + 'static + Debug,

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

Formats the value using the given formatter.