tessl/npm-llamaindex

Data framework for your LLM application

—

Pending

Quality

Pending

Does it follow best practices?

Impact

Pending

No eval scenarios have been run

Securityby

Pending

The risk profile of this skill

Overview

Eval results

Files

Embeddings

Name: tessl/npm-llamaindex
Author: tessl

Text embedding generation and similarity operations for semantic search and retrieval in LlamaIndex.TS.

Import

import { OpenAIEmbedding, Settings } from "llamaindex";

Overview

Embeddings convert text into numerical vector representations that capture semantic meaning, enabling similarity search and retrieval operations in LlamaIndex.TS.

Base Embedding Interface

interface BaseEmbedding {
  getTextEmbedding(text: string): Promise<number[]>;
  getQueryEmbedding(query: string): Promise<number[]>;
  
  embedBatchSize?: number;
  dimensions?: number;
}

OpenAI Embeddings

class OpenAIEmbedding implements BaseEmbedding {
  constructor(options?: {
    model?: string;
    dimensions?: number;
    apiKey?: string;
  });
  
  getTextEmbedding(text: string): Promise<number[]>;
  getQueryEmbedding(query: string): Promise<number[]>;
}

Basic Usage

import { OpenAIEmbedding, Settings } from "llamaindex";

// Configure global embedding model
Settings.embedModel = new OpenAIEmbedding({
  model: "text-embedding-3-large",
  dimensions: 1536,
});

// Generate embeddings
const text = "LlamaIndex is a data framework";
const embedding = await Settings.embedModel.getTextEmbedding(text);
console.log("Embedding dimensions:", embedding.length);

Similarity Functions

function similarity(embedding1: number[], embedding2: number[]): number;

function getTopKEmbeddings(
  queryEmbedding: number[],
  embeddings: number[][],
  k: number
): number[];

Usage with Indices

// Embeddings are automatically used by VectorStoreIndex
const index = await VectorStoreIndex.fromDocuments(documents);
// Uses Settings.embedModel to generate embeddings for documents