Skip to main content

RopeScaling

Enum RopeScaling

pub enum RopeScaling {
    None,
    Linear {
        factor: f64,
    },
    Llama3 {
        factor: f64,
        low_freq_factor: f64,
        high_freq_factor: f64,
        original_max_position_embeddings: usize,
    },
    Yarn {
        factor: f64,
        original_max_position_embeddings: usize,
        beta_fast: f64,
        beta_slow: f64,
        attention_factor: Option<f64>,
    },
    LongRope {
        short_factor: Vec<f64>,
        long_factor: Vec<f64>,
        original_max_position_embeddings: usize,
        factor: f64,
        attention_factor: Option<f64>,
    },
}

Expand description

RoPE frequency scaling, parsed from HF rope_scaling (accepts both the modern rope_type and the legacy type key). Table math lives in combs-models::rope; this is parse-only.

Variants§

None

Linear

Fields

§factor: f64

Llama3

Llama-3.1+ piecewise frequency scaling.

Fields

§factor: f64

§low_freq_factor: f64

§high_freq_factor: f64

§original_max_position_embeddings: usize

Yarn

YaRN NTK-by-parts interpolation (+ attention temperature).

Fields

§factor: f64

§original_max_position_embeddings: usize

§beta_fast: f64

§beta_slow: f64

§attention_factor: Option<f64>

Explicit attention scaling; None = the YaRN default 0.1·ln(factor) + 1.

LongRope

Phi-3 LongRoPE: per-dimension frequency divisors, one set for prompts within the pretraining context and one beyond it.

Fields

§short_factor: Vec<f64>

§long_factor: Vec<f64>

§original_max_position_embeddings: usize

§factor: f64

Context extension ratio max_position / original_max (phi derives the attention temperature from it, not from a config factor key).

§attention_factor: Option<f64>

Explicit attention scaling; None = the LongRoPE default sqrt(1 + ln(factor)/ln(original_max)) (1.0 when factor ≤ 1).

Trait Implementations§

impl Clone for RopeScaling

fn clone(&self) -> RopeScaling

Returns a duplicate of the value. Read more

1.0.0 (const: unstable) · Source§

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

Performs copy-assignment from source. Read more

impl Debug for RopeScaling

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

Formats the value using the given formatter. Read more

impl Default for RopeScaling

fn default() -> RopeScaling

Returns the “default value” for a type. Read more

impl PartialEq for RopeScaling

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

Equality operator ==. Read more

1.0.0 (const: unstable) · Source§

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

Inequality operator !=. Read more

impl StructuralPartialEq for RopeScaling

Auto Trait Implementations§

impl Freeze for RopeScaling

impl RefUnwindSafe for RopeScaling

impl Send for RopeScaling

impl Sync for RopeScaling

impl Unpin for RopeScaling

impl UnsafeUnpin for RopeScaling

impl UnwindSafe for RopeScaling

Blanket Implementations§

impl<T> Any for T
where T: 'static + ?Sized,

fn type_id(&self) -> TypeId

Gets the TypeId of self. Read more

impl<T> Borrow<T> for T
where T: ?Sized,

fn borrow(&self) -> &T

Immutably borrows from an owned value. Read more

impl<T> BorrowMut<T> for T
where T: ?Sized,

fn borrow_mut(&mut self) -> &mut T

Mutably borrows from an owned value. Read more

impl<ST, DT> CastableFrom<ST, Initialized, Initialized> for DT
where ST: ?Sized, DT: ?Sized,

impl<ST, DT> CastableFrom<ST, Uninit, Uninit> for DT
where ST: ?Sized, DT: ?Sized,

impl<C> CloneExpand for C
where C: Clone,

fn __expand_clone_method(&self, _scope: &mut Scope) -> C

impl<T> CloneToUninit for T
where T: Clone,

unsafe fn clone_to_uninit(&self, dest: *mut u8)

🔬This is a nightly-only experimental API. (clone_to_uninit)

Performs copy-assignment from self to dest. Read more

impl<T> Downcast<T> for T

fn downcast(&self) -> &T

impl<T> From<T> for T

fn from(t: T) -> T

Returns the argument unchanged.

impl<T> Instrument for T

fn instrument(self, span: Span) -> Instrumented<Self> ⓘ

Instruments this type with the provided Span, returning an Instrumented wrapper. Read more

fn in_current_span(self) -> Instrumented<Self> ⓘ

Instruments this type with the current Span, returning an Instrumented wrapper. Read more

impl<T, U> Into<U> for T
where U: From<T>,

fn into(self) -> U

Calls U::from(self).

That is, this conversion is whatever the implementation of From<T> for U chooses to do.

impl<T> IntoComptime for T

fn comptime(self) -> Self

impl<T> Read<Exclusive, BecauseExclusive> for T
where T: ?Sized,

impl<T> ToOwned for T
where T: Clone,

type Owned = T

The resulting type after obtaining ownership.

fn to_owned(&self) -> T

Creates owned data from borrowed data, usually by cloning. Read more

fn clone_into(&self, target: &mut T)

Uses borrowed data to replace owned data, usually by cloning. Read more

impl<T, U> TryFrom<U> for T
where U: Into<T>,

type Error = Infallible

The type returned in the event of a conversion error.

fn try_from(value: U) -> Result<T, <T as TryFrom<U>>::Error>

Performs the conversion.

impl<T, U> TryInto<U> for T
where U: TryFrom<T>,

type Error = <U as TryFrom<T>>::Error

The type returned in the event of a conversion error.

fn try_into(self) -> Result<U, <U as TryFrom<T>>::Error>

Performs the conversion.

impl<T> TuneInputs for T
where T: Clone + Send + Sync + 'static,

type At<'a> = T

The concrete input type at lifetime 'a.

impl<T> Upcast<T> for T

fn upcast(&self) -> Option<&T>

impl<T> WasmNotSend for T
where T: Send,

impl<T> WasmNotSendSync for T
where T: WasmNotSend + WasmNotSync,

impl<T> WasmNotSync for T
where T: Sync,

impl<T> WithSubscriber for T

fn with_subscriber<S>(self, subscriber: S) -> WithDispatch<Self> ⓘ
where S: Into<Dispatch>,

Attaches the provided Subscriber to this type, returning a WithDispatch wrapper. Read more

fn with_current_subscriber(self) -> WithDispatch<Self> ⓘ

Attaches the current default Subscriber to this type, returning a WithDispatch wrapper. Read more