Struct ContrastiveConfig 

pub struct ContrastiveConfig {
    pub temperature: f64,
    pub margin: f64,
    pub num_negatives: usize,
    pub mining_strategy: NegativeMiningStrategy,
    pub use_cosine: bool,
    pub label_smoothing: f64,
}

Configuration for contrastive learning.

Fields

temperature: f64

Temperature parameter for softmax scaling (default: 0.07).

margin: f64

Margin for triplet loss (default: 1.0).

num_negatives: usize

Number of negative samples per positive (default: 128).

mining_strategy: NegativeMiningStrategy

Hard negative mining strategy (default: SemiHard).

use_cosine: bool

Whether to use cosine similarity (true) or dot product (false).

label_smoothing: f64

Label smoothing factor (0.0 = no smoothing, default: 0.0).

Trait Implementations

impl Clone for ContrastiveConfig

fn clone(&self) -> ContrastiveConfig

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for ContrastiveConfig

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

Formats the value using the given formatter.

impl Default for ContrastiveConfig

fn default() -> Self

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for ContrastiveConfig

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

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for ContrastiveConfig

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

where __S: Serializer,

Serialize this value into the given Serde serializer.