Struct GnnOptimizerConfig 

pub struct GnnOptimizerConfig {
    pub num_nodes: usize,
    pub hidden_dim: usize,
    pub num_message_rounds: usize,
    pub topology: GraphTopology,
    pub aggregation: MessageAggregation,
    pub activation: MessageActivation,
    pub base_lr: f64,
    pub momentum_decay: f64,
    pub rms_decay: f64,
    pub history_decay: f64,
    pub max_grad_norm: f64,
    pub seed: u64,
}

Configuration for GnnOptimizer.

The numeric hyper-parameters are stored as f64 and converted to the optimizer’s scalar type T when the optimizer is constructed.

Fields

num_nodes: usize

Requested number of graph nodes (clamped to the parameter count at run time).

hidden_dim: usize

Dimension of the GRU hidden state and message space.

num_message_rounds: usize

Number of synchronous message-passing rounds per optimization step.

topology: GraphTopology

Inter-node graph connectivity.

aggregation: MessageAggregation

Neighbour message aggregation rule.

activation: MessageActivation

Message nonlinearity.

base_lr: f64

Base learning rate applied before the learned per-node step scale.

momentum_decay: f64

EMA decay for the gradient (momentum), often called beta1.

rms_decay: f64

EMA decay for the squared gradient (RMS), often called beta2.

history_decay: f64

EMA decay for the per-node update-history feature.

max_grad_norm: f64

Maximum gradient L2 norm; larger gradients are rescaled before use.

seed: u64

Seed for deterministic weight initialisation.

Trait Implementations

impl Clone for GnnOptimizerConfig

fn clone(&self) -> Self

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for GnnOptimizerConfig

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

Formats the value using the given formatter.

impl Default for GnnOptimizerConfig

fn default() -> Self

Returns the “default value” for a type.