Struct CandleEmbeddingModel 

pub struct CandleEmbeddingModel { /* private fields */ }

A lazily-loaded embedding model backed by Candle.

On first embed call (or explicit warmup), the model weights and tokenizer are downloaded from HuggingFace Hub and loaded into memory.

Implementations

impl CandleEmbeddingModel

pub fn new( model_type: CandleTextModel, revision: Option<String>, cache_dir: PathBuf, ) -> Self

Trait Implementations

impl EmbeddingModel for CandleEmbeddingModel

fn embed<'life0, 'life1, 'life2, 'async_trait>( &'life0 self, texts: &'life1 [&'life2 str], ) -> Pin<Box<dyn Future<Output = Result<EmbedResult>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait, 'life2: 'async_trait,

Embed a batch of text strings into dense vectors.

fn dimensions(&self) -> u32

The dimensionality of the embedding vectors produced by this model.

fn warmup<'life0, 'async_trait>( &'life0 self, ) -> Pin<Box<dyn Future<Output = Result<()>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait,

Optional warmup hook (e.g. load weights into memory on first access). The default is a no-op.

impl ModelInfo for CandleEmbeddingModel

fn model_id(&self) -> &str

The underlying model identifier (e.g. a HuggingFace repo ID or API model name).

fn active_execution_providers(&self) -> Vec<String>

Names of the ONNX Runtime execution providers requested for the underlying session, in priority order. Empty for remote and non-ONNX models.