Struct HuggingFaceTextVectorizer 

pub struct HuggingFaceTextVectorizer { /* private fields */ }

HuggingFace local embedding adapter backed by ONNX Runtime.

Uses the fastembed crate to run embedding models locally. Models are automatically downloaded from the HuggingFace Hub on first use and cached on disk.

This vectorizer implements Vectorizer for synchronous embedding generation. For async use cases, wrap it with tokio::task::spawn_blocking or use it with the synchronous semantic extension APIs.

Implementations

impl HuggingFaceTextVectorizer

pub fn new(config: HuggingFaceConfig) -> Result<Self>

Creates a new HuggingFace local vectorizer.

This may download the model from HuggingFace Hub on first invocation.

Errors

Returns an error if the model cannot be loaded.

Trait Implementations

impl Debug for HuggingFaceTextVectorizer

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

Formats the value using the given formatter.

impl Vectorizer for HuggingFaceTextVectorizer

fn embed(&self, text: &str) -> Result<Vec<f32>>

Embeds a single string.

fn embed_many(&self, texts: &[&str]) -> Result<Vec<Vec<f32>>>

Embeds many strings.

impl Send for HuggingFaceTextVectorizer

impl Sync for HuggingFaceTextVectorizer