Embeddings

Embeddings convert text into vector representations, enabling semantic search and similarity comparison.

Usage

from yosrai.engine.embeddings import OpenAIEmbedder, OllamaEmbedder

# OpenAI
embedder = OpenAIEmbedder(model="text-embedding-3-small")
vector = embedder.embed_query("Hello world")

# Ollama (Local)
local_embedder = OllamaEmbedder(model="nomic-embed-text")
vector = local_embedder.embed_query("Hello world")

API Reference

yosrai.engine.embeddings.base.Embedder

Bases: ABC

Abstract base class for Embedding providers.

aembed_documents(texts) abstractmethod async

Async embed a batch of documents.

aembed_query(text) abstractmethod async

Async embed a single query string.

embed_documents(texts) abstractmethod

Embed a batch of documents.

embed_query(text) abstractmethod

Embed a single query string.

yosrai.engine.embeddings.openai_embedder.OpenAIEmbedder

Bases: Embedder

yosrai.engine.embeddings.ollama_embedder.OllamaEmbedder

Bases: Embedder