Crate phago_embeddings 

Phago Embeddings

Embedding backends for Phago semantic intelligence.

This crate provides vector embeddings for semantic understanding:

• Text → vector conversion
• Similarity computation
• Chunking and normalization

Features

• local: ONNX-based local embeddings (no API needed)
• api: API-based embeddings (OpenAI, Voyage, etc.)
• full: Both local and API support

Usage

use phago_embeddings::{Embedder, SimpleEmbedder};

let embedder = SimpleEmbedder::new();
let vector = embedder.embed("cell membrane transport");
let similarity = embedder.cosine_similarity(&v1, &v2);

Modules

prelude

Prelude for convenient imports.

Structs

ChunkConfig

Chunking configuration.

Chunker

Text chunker.

SimpleEmbedder

Simple hash-based embedder.

Enums

EmbeddingError

Embedding error types.

Traits

Embedder

Core trait for embedding providers.

Functions

cosine_similarity

Compute cosine similarity between two vectors.

dot_product

Compute dot product between two vectors.

euclidean_distance

Compute Euclidean distance between two vectors.

normalize_l1

L1 normalize a vector (sum to 1).

normalize_l2

L2 normalize a vector (unit length).

normalize_minmax

Min-max normalize to [0, 1] range.

normalize_zscore

Z-score normalize (mean=0, std=1).

Type Aliases

EmbeddingResult

Result type for embedding operations.