Enum AlignmentMethod 

pub enum AlignmentMethod {
    DPO {
        beta: f64,
    },
    ORPO {
        lambda: f64,
    },
    None,
}

Alignment training method.

Variants

DPO

Direct Preference Optimization.

Fields

beta: f64

DPO beta parameter.

ORPO

Odds Ratio Preference Optimization (single-pass).

Fields

lambda: f64

ORPO lambda parameter.

None

No alignment, standard SFT only.

Implementations

impl AlignmentMethod

pub fn dpo() -> Self

Create DPO alignment with default beta.

pub fn orpo() -> Self

Create ORPO alignment with default lambda.

Trait Implementations

impl Clone for AlignmentMethod

fn clone(&self) -> AlignmentMethod

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for AlignmentMethod

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

Formats the value using the given formatter.

impl Default for AlignmentMethod

fn default() -> AlignmentMethod

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for AlignmentMethod

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

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl PartialEq for AlignmentMethod

fn eq(&self, other: &AlignmentMethod) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl Serialize for AlignmentMethod

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

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl Copy for AlignmentMethod

impl StructuralPartialEq for AlignmentMethod