Struct GnnOptimizer 

pub struct GnnOptimizer<T: Float + Debug + Send + Sync + 'static> { /* private fields */ }

A learned optimizer driven by a message-passing GNN with a GRU node update.

Implements AdvancedOptimizer. The learned weights are fixed; the per-node EMAs and GRU hidden vectors are persistent state that evolves across steps.

Implementations

impl<T: Float + Debug + Send + Sync + 'static> GnnOptimizer<T>

pub fn weight_vector(&self) -> Vec<f64>

Every learned weight, flattened in a stable order.

The layout is an implementation detail, but it is stable across calls on the same architecture, which is what a meta-training loop needs: set_weight_vector(&weight_vector()) is the identity.

pub fn weight_count(&self) -> usize

Number of learned weights.

pub fn set_weight_vector(&mut self, flat: &[f64]) -> Result<()>

Overwrite every learned weight from a vector laid out like Self::weight_vector’s output.

Errors

Returns crate::error::OptimError::InvalidConfig if flat is not exactly the right length for this architecture — a silent partial load would leave the network in a state no caller could reason about.

pub fn reset_state(&mut self)

Clear the persistent optimization state (per-node EMAs, GRU hidden vectors, step count, graph layout) while keeping the learned weights.

Meta-training evaluates the same weights on several tasks, and a rollout must not inherit momentum from the previous one. The graph layout is discarded too because it is derived from the parameter length, which differs between tasks; it is rebuilt lazily on the next step.

impl<T: Float + Debug + Send + Sync + 'static> GnnOptimizer<T>

pub fn new(config: GnnOptimizerConfig) -> Result<Self>

Construct a new optimizer from a configuration.

The learned weights are initialised deterministically from config.seed. Graph-dependent state is created lazily on the first AdvancedOptimizer::step once the parameter length is known.

Errors

Returns OptimError::InvalidConfig if a hyper-parameter is out of range.

pub fn num_nodes(&self) -> usize

The actual number of graph nodes (after clamping to the parameter count).

Returns 0 before the first step, when the parameter length is unknown.

pub fn hidden_dim(&self) -> usize

The GRU hidden / message dimension.

pub fn seed(&self) -> u64

The seed used for deterministic weight initialisation.

Trait Implementations

impl<T: Float + Debug + Send + Sync + 'static> AdvancedOptimizer<T> for GnnOptimizer<T>

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: Clone + Float + Debug + Send + Sync + 'static> Clone for GnnOptimizer<T>

fn clone(&self) -> Self

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl<T: Debug + Float + Debug + Send + Sync + 'static> Debug for GnnOptimizer<T>

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

Formats the value using the given formatter.

impl<T: Float + Debug + Send + Sync + 'static> MetaTrainable<T> for GnnOptimizer<T>

fn weight_vector(&self) -> Vec<f64>

Every learned weight, flattened in an order that is stable across calls for a given architecture, so that set_weight_vector(&weight_vector()) is the identity.

fn set_weight_vector(&mut self, flat: &[f64]) -> Result<()>

Overwrite every learned weight from a vector laid out like Self::weight_vector’s output.

fn reset_state(&mut self)

Clear the persistent optimization state (momentum, hidden vectors, memory, step count) while keeping the learned weights.

fn weight_count(&self) -> usize

Number of learned weights.