Struct FallbackEmbedder 

pub struct FallbackEmbedder { /* private fields */ }

Hash-based fallback embedder.

Generates deterministic pseudo-embeddings using a combination of:

• Word-level hashing for vocabulary capture
• Character n-gram hashing for fuzzy matching
• Normalization to unit length for cosine similarity

This is NOT semantic similarity - it’s based on lexical overlap. Use FastEmbed for true semantic understanding.

Examples

use rlm_rs::embedding::{Embedder, FallbackEmbedder, DEFAULT_DIMENSIONS};

let embedder = FallbackEmbedder::new(DEFAULT_DIMENSIONS);
let emb1 = embedder.embed("hello world").unwrap();
let emb2 = embedder.embed("hello world").unwrap();
assert_eq!(emb1, emb2); // Deterministic

Implementations

impl FallbackEmbedder

pub const fn new(dimensions: usize) -> Self

Creates a new fallback embedder with the specified dimensions.

Trait Implementations

impl Embedder for FallbackEmbedder

fn dimensions(&self) -> usize

Returns the embedding dimensions.

fn model_name(&self) -> &'static str

Returns the model name/version identifier.

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

Generates an embedding for the given text.

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

Generates embeddings for multiple texts.