跳至主要内容

Pinecone

本指南将帮助您开始使用由Pinecone 向量存储支持的检索器。有关所有功能和配置的详细文档，请前往API 参考。

概述​

自查询检索器 通过基于某些输入查询动态生成元数据过滤器来检索文档。这使得检索器除了纯语义相似性之外，还可以考虑潜在文档元数据来获取结果。

它使用一个名为 Translator 的模块，该模块根据有关元数据字段和给定向量存储支持的查询语言的信息生成过滤器。

集成详细信息​

支持向量存储自托管云产品包Py 支持
PineconeStore❌✅@langchain/pinecone✅

设置​

按照此处的说明设置 Pinecone 实例。设置以下环境变量

process.env.PINECONE_API_KEY = "YOUR_API_KEY";
process.env.PINECONE_ENVIRONMENT = "YOUR_ENVIRONMENT";
process.env.PINECONE_INDEX = "YOUR_INDEX";

如果您希望从单个查询中获得自动跟踪，还可以通过取消注释以下内容来设置您的LangSmith API 密钥

// process.env.LANGSMITH_API_KEY = "<YOUR API KEY HERE>";
// process.env.LANGSMITH_TRACING = "true";

安装​

向量存储位于 @langchain/pinecone 包中。您还需要安装 langchain 包才能导入主要的 SelfQueryRetriever 类。

官方 Pinecone SDK (@pinecone-database/pinecone) 会自动安装为 @langchain/pinecone 的依赖项，但您也可以独立安装它。

在此示例中，我们还将使用 OpenAI 嵌入，因此您需要安装 @langchain/openai 包并获取 API 密钥

提示

查看本节以获取有关安装集成包的常规说明。

yarn add @langchain/pinecone langchain @langchain/openai @pinecone-database/pinecone

实例化​

首先，使用包含元数据的某些文档初始化您的 Pinecone 向量存储

import { Pinecone } from "@pinecone-database/pinecone";

import { OpenAIEmbeddings } from "@langchain/openai";
import { PineconeStore } from "@langchain/pinecone";
import { Document } from "@langchain/core/documents";
import type { AttributeInfo } from "langchain/chains/query_constructor";

/**
 * First, we create a bunch of documents. You can load your own documents here instead.
 * Each document has a pageContent and a metadata field. Make sure your metadata matches the AttributeInfo below.
 */
const docs = [
  new Document({
    pageContent:
      "A bunch of scientists bring back dinosaurs and mayhem breaks loose",
    metadata: { year: 1993, rating: 7.7, genre: "science fiction" },
  }),
  new Document({
    pageContent:
      "Leo DiCaprio gets lost in a dream within a dream within a dream within a ...",
    metadata: { year: 2010, director: "Christopher Nolan", rating: 8.2 },
  }),
  new Document({
    pageContent:
      "A psychologist / detective gets lost in a series of dreams within dreams within dreams and Inception reused the idea",
    metadata: { year: 2006, director: "Satoshi Kon", rating: 8.6 },
  }),
  new Document({
    pageContent:
      "A bunch of normal-sized women are supremely wholesome and some men pine after them",
    metadata: { year: 2019, director: "Greta Gerwig", rating: 8.3 },
  }),
  new Document({
    pageContent: "Toys come alive and have a blast doing so",
    metadata: { year: 1995, genre: "animated" },
  }),
  new Document({
    pageContent: "Three men walk into the Zone, three men walk out of the Zone",
    metadata: {
      year: 1979,
      director: "Andrei Tarkovsky",
      genre: "science fiction",
      rating: 9.9,
    },
  }),
];

/**
 * Next, we define the attributes we want to be able to query on.
 * in this case, we want to be able to query on the genre, year, director, rating, and length of the movie.
 * We also provide a description of each attribute and the type of the attribute.
 * This is used to generate the query prompts.
 */
const attributeInfo: AttributeInfo[] = [
  {
    name: "genre",
    description: "The genre of the movie",
    type: "string or array of strings",
  },
  {
    name: "year",
    description: "The year the movie was released",
    type: "number",
  },
  {
    name: "director",
    description: "The director of the movie",
    type: "string",
  },
  {
    name: "rating",
    description: "The rating of the movie (1-10)",
    type: "number",
  },
  {
    name: "length",
    description: "The length of the movie in minutes",
    type: "number",
  },
];

/**
 * Next, we instantiate a vector store. This is where we store the embeddings of the documents.
 * We also need to provide an embeddings object. This is used to embed the documents.
 */

const pinecone = new Pinecone();

const pineconeIndex = pinecone.Index(process.env.PINECONE_INDEX!);

const embeddings = new OpenAIEmbeddings();
const vectorStore = await PineconeStore.fromDocuments(docs, embeddings, {
  pineconeIndex: pineconeIndex,
});

现在我们可以实例化我们的检索器

选择您的聊天模型

安装依赖项

提示

查看 本节以获取有关安装集成包的常规说明.

yarn add @langchain/openai

添加环境变量

OPENAI_API_KEY=your-api-key

实例化模型

import { ChatOpenAI } from "@langchain/openai";

const llm = new ChatOpenAI({
  model: "gpt-4o-mini",
  temperature: 0
});
import { SelfQueryRetriever } from "langchain/retrievers/self_query";
import { PineconeTranslator } from "@langchain/pinecone";

const selfQueryRetriever = SelfQueryRetriever.fromLLM({
  llm: llm,
  vectorStore: vectorStore,
  /** A short summary of what the document contents represent. */
  documentContents: "Brief summary of a movie",
  attributeInfo: attributeInfo,
  /**
   * We need to create a basic translator that translates the queries into a
   * filter format that the vector store can understand. We provide a basic translator
   * translator here, but you can create your own translator by extending BaseTranslator
   * abstract class. Note that the vector store needs to support filtering on the metadata
   * attributes you want to query on.
   */
  structuredQueryTranslator: new PineconeTranslator(),
});

用法​

现在，问一个需要了解文档元数据才能回答的问题。您可以看到检索器会生成正确的结果

await selfQueryRetriever.invoke("Which movies are rated higher than 8.5?");
[
  Document {
    pageContent: 'A psychologist / detective gets lost in a series of dreams within dreams within dreams and Inception reused the idea',
    metadata: { director: 'Satoshi Kon', rating: 8.6, year: 2006 },
    id: undefined
  },
  Document {
    pageContent: 'Three men walk into the Zone, three men walk out of the Zone',
    metadata: {
      director: 'Andrei Tarkovsky',
      genre: 'science fiction',
      rating: 9.9,
      year: 1979
    },
    id: undefined
  }
]

在链中使用​

与其他检索器一样，Pinecone 自查询检索器可以通过链集成到 LLM 应用程序中。

请注意，由于它们的返回答案可能在很大程度上取决于文档元数据，因此我们以不同的格式对检索到的文档进行格式化，以包括该信息。

import { ChatPromptTemplate } from "@langchain/core/prompts";
import {
  RunnablePassthrough,
  RunnableSequence,
} from "@langchain/core/runnables";
import { StringOutputParser } from "@langchain/core/output_parsers";

import type { Document } from "@langchain/core/documents";

const prompt = ChatPromptTemplate.fromTemplate(`
Answer the question based only on the context provided.

Context: {context}

Question: {question}`);

const formatDocs = (docs: Document[]) => {
  return docs.map((doc) => JSON.stringify(doc)).join("\n\n");
};

// See https://js.langchain.ac.cn/v0.2/docs/tutorials/rag
const ragChain = RunnableSequence.from([
  {
    context: selfQueryRetriever.pipe(formatDocs),
    question: new RunnablePassthrough(),
  },
  prompt,
  llm,
  new StringOutputParser(),
]);
await ragChain.invoke("Which movies are rated higher than 8.5?");
The movies rated higher than 8.5 are the ones directed by Satoshi Kon (rating: 8.6) and Andrei Tarkovsky (rating: 9.9).

默认搜索参数​

您也可以将 searchParams 字段传递给上述方法，该字段提供了除了任何生成的查询之外还应用的默认过滤器。过滤器语法与支持的 Pinecone 向量存储相同

const selfQueryRetrieverWithDefaultParams = SelfQueryRetriever.fromLLM({
  llm: llm,
  vectorStore: vectorStore,
  documentContents: "Brief summary of a movie",
  attributeInfo: attributeInfo,
  structuredQueryTranslator: new PineconeTranslator(),
  searchParams: {
    filter: {
      rating: {
        $gt: 8.5,
      },
    },
    mergeFiltersOperator: "and",
  },
});

API 参考​

有关所有 Pinecone 自查询检索器功能和配置的详细文档，请前往API 参考。

此页面是否有用？


您也可以在 GitHub 上留下详细的反馈 on GitHub.