Struct FastTextClassifier 

pub struct FastTextClassifier { /* private fields */ }

FastText-inspired averaged word-vector text classifier.

Each document is represented as the average of its per-word embeddings. A linear layer maps the averaged vector to class logits. Training uses SGD with Hogwild-style updates.

Implementations

impl FastTextClassifier

pub fn new(n_classes: usize, dim: usize, classes: Vec<String>) -> Self

Create a new, unfitted classifier.

• n_classes – number of output classes.
• dim – word-embedding dimension.
• classes – ordered class name list.

pub fn fit( &mut self, corpus: &[(Vec<String>, usize)], n_epochs: usize, lr: f32, ) -> Result<()>

Train the classifier.

corpus is (tokens, class_id) pairs. lr is the learning rate.

pub fn predict(&self, tokens: &[String]) -> usize

Predict the class ID for a token sequence.

pub fn predict_proba(&self, tokens: &[String]) -> Vec<f32>

Predict class probability distribution.

pub fn class_names(&self) -> &[String]

Return the ordered class names.

pub fn is_fitted(&self) -> bool

true if the model has been trained.

Trait Implementations

impl Clone for FastTextClassifier

fn clone(&self) -> FastTextClassifier

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for FastTextClassifier

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

Formats the value using the given formatter.