Struct FixedDistTargetModel 

pub struct FixedDistTargetModel { /* private fields */ }

Target model returning the same categorical distribution for every position.

Implementations

impl FixedDistTargetModel

pub fn new(probs: Vec<f64>) -> SpeculativeDecodingResult<Self>

Build a mock target from a linear-space probability vector. See FixedDistDraftModel::new for invariants.

pub fn probs(&self) -> &[f64]

Access the (normalized) linear-space probabilities.

pub fn logprobs(&self) -> &[LogProb]

Access the log-probability row.

Trait Implementations

impl Clone for FixedDistTargetModel

fn clone(&self) -> FixedDistTargetModel

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for FixedDistTargetModel

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

Formats the value using the given formatter.

impl TargetModel for FixedDistTargetModel

fn vocab_size(&self) -> usize

Vocabulary cardinality the target emits log-probs over. Must match the draft’s vocab_size().

fn verify( &self, _prefix: &[TokenId], draft_tokens: &[TokenId], ) -> SpeculativeDecodingResult<TargetScores>

Score prefix concatenated with draft_tokens: return k + 1 distributions (the k draft-covered positions plus the bonus).