tessl/npm-langchain--core

Core LangChain.js abstractions and schemas for building applications with Large Language Models

—

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-langchain--core
Author: tessl

Embedding model abstractions for converting text to vector representations. Embeddings enable semantic search and similarity comparisons by transforming text into high-dimensional numerical vectors.

Capabilities

Embeddings

Abstract base class for all embedding implementations.

/**
 * Abstract base class for embedding models
 */
abstract class Embeddings {
  /** Async caller for managing API requests */
  caller: AsyncCaller;
  
  constructor(params: EmbeddingsParams);
  
  /** Embed multiple documents */
  abstract embedDocuments(documents: string[]): Promise<number[][]>;
  
  /** Embed a single query */
  abstract embedQuery(document: string): Promise<number[]>;
}

Usage Examples:

import { Embeddings } from "@langchain/core/embeddings";

// Example embedding implementation
class MyEmbeddings extends Embeddings {
  async embedDocuments(texts: string[]): Promise<number[][]> {
    // Implementation would call embedding API
    return texts.map(text => this.generateVector(text));
  }
  
  async embedQuery(text: string): Promise<number[]> {
    // Single text embedding
    return this.generateVector(text);
  }
  
  private generateVector(text: string): number[] {
    // Placeholder - real implementation would use ML model
    return new Array(384).fill(0).map(() => Math.random());
  }
}

// Usage
const embeddings = new MyEmbeddings();

// Embed documents for vector store
const docVectors = await embeddings.embedDocuments([
  "LangChain is a framework for building LLM applications",
  "Vector stores enable semantic search over documents"
]);

// Embed query for similarity search
const queryVector = await embeddings.embedQuery("What is LangChain?");

Types

interface EmbeddingsInterface {
  embedDocuments(documents: string[]): Promise<number[][]>;
  embedQuery(document: string): Promise<number[]>;
}

interface EmbeddingsParams extends AsyncCallerParams {
  /** Maximum number of documents to embed at once */
  maxConcurrency?: number;
  /** Maximum number of retries for failed requests */
  maxRetries?: number;
}

interface AsyncCallerParams {
  /** Maximum number of concurrent requests */
  maxConcurrency?: number;
  /** Maximum number of retries */
  maxRetries?: number;
  /** Base delay between retries (ms) */
  onFailedAttempt?: (error: Error) => void;
}