Struct LearnedOptimizationConfig 

pub struct LearnedOptimizationConfig {
    pub meta_training_episodes: usize,
    pub meta_learning_rate: f64,
    pub inner_steps: usize,
    pub inner_learning_rate: f64,
    pub batch_size: usize,
    pub max_parameters: usize,
    pub use_transformer: bool,
    pub hidden_size: usize,
    pub num_heads: usize,
    pub few_shot_adaptation: bool,
    pub exploration_temperature: f64,
}

Configuration for learned optimizers

Fields

meta_training_episodes: usize

Number of meta-training episodes

meta_learning_rate: f64

Learning rate for meta-optimizer

inner_steps: usize

Number of inner optimization steps

inner_learning_rate: f64

Inner learning rate

batch_size: usize

Batch size for meta-learning

max_parameters: usize

Maximum number of parameters to optimize

use_transformer: bool

Whether to use transformer architecture

hidden_size: usize

Hidden size for neural networks

num_heads: usize

Number of attention heads (for transformer)

few_shot_adaptation: bool

Whether to enable few-shot adaptation

exploration_temperature: f64

Temperature for exploration

Trait Implementations

impl Clone for LearnedOptimizationConfig

fn clone(&self) -> LearnedOptimizationConfig

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for LearnedOptimizationConfig

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

Formats the value using the given formatter.

impl Default for LearnedOptimizationConfig

fn default() -> Self

Returns the “default value” for a type.