Struct MetaLearningConfig 

pub struct MetaLearningConfig {

    pub algorithm: MetaAlgorithm,
    pub inner_lr: f64,
    pub meta_lr: f64,
    pub inner_steps: usize,
    pub support_size: usize,
    pub query_size: usize,
    pub num_ways: usize,
    pub num_shots: usize,
    pub first_order: bool,
    pub temperature: f64,
    pub embedding_dim: usize,
    pub normalize_embeddings: bool,
    pub memory_size: usize,
    pub memory_key_dim: usize,
    pub memory_value_dim: usize,
    pub meta_batch_size: usize,
    pub task_specific_params: bool,
    pub inner_l2_reg: f64,
    pub grad_clip_norm: f64,
}

Configuration for meta-learning

Fields

algorithm: MetaAlgorithm

Meta-learning algorithm to use

inner_lr: f64

Inner loop learning rate (task-specific adaptation)

meta_lr: f64

Meta learning rate (across-task learning)

inner_steps: usize

Number of inner loop gradient steps

support_size: usize

Number of support examples per task

query_size: usize

Number of query examples per task

num_ways: usize

Number of ways (classes) per task

num_shots: usize

Number of shots (examples per class) per task

first_order: bool

Whether to use first-order approximation

temperature: f64

Temperature for softmax in prototypical networks

embedding_dim: usize

Dimension of learned embeddings

normalize_embeddings: bool

Whether to normalize embeddings

memory_size: usize

Memory size for memory-augmented networks

memory_key_dim: usize

Memory key dimension

memory_value_dim: usize

Memory value dimension

meta_batch_size: usize

Number of meta-training tasks per batch

task_specific_params: bool

Whether to use task-specific parameters

inner_l2_reg: f64

L2 regularization for inner loop

grad_clip_norm: f64

Gradient clipping threshold

Trait Implementations

impl Clone for MetaLearningConfig

fn clone(&self) -> MetaLearningConfig

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for MetaLearningConfig

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

Formats the value using the given formatter.

impl Default for MetaLearningConfig

fn default() -> Self

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for MetaLearningConfig

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for MetaLearningConfig

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.