Struct NeuralOptimizerConfig 

pub struct NeuralOptimizerConfig {
    pub meta_learning_rate: f32,
    pub hidden_size: usize,
    pub num_layers: usize,
    pub device: Arc<CpuDevice>,
    pub max_grad_norm: f32,
    pub coordinate_wise: bool,
    pub history_length: usize,
}

Configuration for neural optimizer

Fields

meta_learning_rate: f32

Learning rate for the meta-optimizer (optimizer of the optimizer)

hidden_size: usize

Hidden size for the LSTM optimizer network

num_layers: usize

Number of layers in the optimizer network

device: Arc<CpuDevice>

Device to run the neural optimizer on

max_grad_norm: f32

Maximum gradient norm for clipping

coordinate_wise: bool

Whether to use coordinate-wise optimization

history_length: usize

History length for the neural network

Trait Implementations

impl Clone for NeuralOptimizerConfig

fn clone(&self) -> NeuralOptimizerConfig

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for NeuralOptimizerConfig

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

Formats the value using the given formatter.

impl Default for NeuralOptimizerConfig

fn default() -> Self

Returns the “default value” for a type.