Struct SimpleEmbedder 

pub struct SimpleEmbedder { /* private fields */ }

Simple hash-based embedder.

Creates embeddings by hashing words into a fixed-dimension space. Uses multiple hash functions for better distribution.

Example

use phago_embeddings::SimpleEmbedder;
use phago_embeddings::Embedder;

let embedder = SimpleEmbedder::new(128);
let vec = embedder.embed("hello world").unwrap();
assert_eq!(vec.len(), 128);

Implementations

impl SimpleEmbedder

pub fn new(dimension: usize) -> Self

Create a new simple embedder with specified dimension.

pub fn default_dimension() -> Self

Create with default dimension (256).

Trait Implementations

impl Default for SimpleEmbedder

fn default() -> Self

Returns the “default value” for a type.

impl Embedder for SimpleEmbedder

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

Embed a single text string.

fn dimension(&self) -> usize

Get the embedding dimension.

fn model_name(&self) -> &str

Get the model name/identifier.

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

Embed multiple texts in a batch (more efficient).

fn similarity(&self, a: &[f32], b: &[f32]) -> EmbeddingResult<f32>

Compute cosine similarity between two vectors.

fn most_similar<'a>( &self, query: &str, candidates: &[&'a str], ) -> EmbeddingResult<Option<(&'a str, f32)>>

Find the most similar text from a list.

fn top_k_similar<'a>( &self, query: &str, candidates: &[&'a str], k: usize, ) -> EmbeddingResult<Vec<(&'a str, f32)>>

Find top-k most similar texts.