Struct MambaPredictor 

pub struct MambaPredictor { /* private fields */ }

A predictor using the Mamba neural network architecture.

Implementations

impl MambaPredictor

pub fn new(model: Arc<MambaModel>) -> Self

Creates a new MambaPredictor from an initialized Model.

pub fn from_method(method: &str) -> Result<Self, String>

Creates a new MambaPredictor from a method string.

Trait Implementations

impl Predictor for MambaPredictor

Available on crate feature backend-mamba only.

fn update(&mut self, sym: bool)

Incorporates a new symbol into the model’s training history.

fn revert(&mut self)

Reverts the model to its state before the last update.

fn predict_prob(&mut self, sym: bool) -> f64

Predicts the probability of the next symbol being sym.

fn model_name(&self) -> String

Returns a human-readable name of the predictive model.

fn boxed_clone(&self) -> Box<dyn Predictor>

Creates a boxed clone of this predictor.

fn commit_update(&mut self, sym: bool)

Incorporates a new symbol as committed training history without retaining rollback state when the predictor supports that optimization.

fn update_history(&mut self, sym: bool)

Appends a symbol to the model’s interaction history without necessarily updating the training counts immediately (backend dependent).

fn commit_update_history(&mut self, sym: bool)

Appends a symbol as committed interaction history without retaining rollback state when the predictor supports that optimization.

fn pop_history(&mut self)

Reverts the model to its state before the last update_history.

fn begin_rollback_scope(&mut self)

Begins an optional coarse rollback scope for simulation-heavy callers.

fn rollback_scope(&mut self) -> bool

Rolls back to the last scope opened with begin_rollback_scope.

fn predict_one(&mut self) -> f64

Shorthand for predict_prob(true).