Struct FastEmbedEmbedder 

pub struct FastEmbedEmbedder { /* private fields */ }

fastembed-rs–backed Embedder.

fastembed 5.x requires &mut TextEmbedding for embed; the trait surface gives us &self. Interior Mutex serialises concurrent embed calls — acceptable because ONNX inference is already CPU- saturating.

Implementations

impl FastEmbedEmbedder

pub fn new() -> Result<FastEmbedEmbedder, MemoryError>

Build with DEFAULT_MODEL. Triggers model download on first call if not already cached.

pub fn with_model( model: EmbeddingModel, dim: usize, ) -> Result<FastEmbedEmbedder, MemoryError>

Build with the supplied fastembed model. The caller asserts the model’s output dimension; mismatch is reported at first embed call against a dimension-checked store.

Trait Implementations

impl Embedder for FastEmbedEmbedder

fn dimension(&self) -> usize

Embedding dimensionality. Must be constant for a given Embedder instance — long-term memory backends reject vectors of the wrong length at runtime.

fn embed<'life0, 'life1, 'async_trait>( &'life0 self, texts: &'life1 [String], ) -> Pin<Box<dyn Future<Output = Result<Vec<Vec<f32>>, MemoryError>> + Send + 'async_trait>>

where 'life0: 'async_trait, 'life1: 'async_trait, FastEmbedEmbedder: 'async_trait,

Compute one embedding per input text.

fn is_ranking(&self) -> bool

Whether distinct inputs produce distinct directions, i.e. whether a vector store wired with this embedder actually ranks by relevance.