Skip to main content

MyScale

兼容性

仅支持 Node.js。

MyScale 是一个新兴的 AI 数据库，它将向量搜索和 SQL 分析的能力结合在一起，提供一种托管的、高效的、响应迅速的体验。

安装配置​

  1. 通过 MyScale 的 Web 控制台 启动一个集群。更多信息请参见 MyScale 官方文档。
  2. 启动集群后，从集群的 Actions 菜单中查看您的 连接详情（Connection Details）。您将需要主机地址、端口、用户名和密码。
  3. 在您的工作区中安装所需的 Node.js 对等依赖。

:::提示 请参阅安装集成包的一般说明部分。 :::

npm install -S @langchain/openai @clickhouse/client @langchain/community @langchain/core

索引和查询文档​

import { MyScaleStore } from "@langchain/community/vectorstores/myscale";
import { OpenAIEmbeddings } from "@langchain/openai";

const vectorStore = await MyScaleStore.fromTexts(
  ["Hello world", "Bye bye", "hello nice world"],
  [
    { id: 2, name: "2" },
    { id: 1, name: "1" },
    { id: 3, name: "3" },
  ],
  new OpenAIEmbeddings(),
  {
    host: process.env.MYSCALE_HOST || "localhost",
    port: process.env.MYSCALE_PORT || "8443",
    username: process.env.MYSCALE_USERNAME || "username",
    password: process.env.MYSCALE_PASSWORD || "password",
    database: "default", // defaults to "default"
    table: "your_table", // defaults to "vector_table"
  }
);

const results = await vectorStore.similaritySearch("hello world", 1);
console.log(results);

const filteredResults = await vectorStore.similaritySearch("hello world", 1, {
  whereStr: "metadata.name = '1'",
});
console.log(filteredResults);

API Reference:

从现有集合中查询文档​

import { MyScaleStore } from "@langchain/community/vectorstores/myscale";
import { OpenAIEmbeddings } from "@langchain/openai";

const vectorStore = await MyScaleStore.fromExistingIndex(
  new OpenAIEmbeddings(),
  {
    host: process.env.MYSCALE_HOST || "localhost",
    port: process.env.MYSCALE_PORT || "8443",
    username: process.env.MYSCALE_USERNAME || "username",
    password: process.env.MYSCALE_PASSWORD || "password",
    database: "default", // defaults to "default"
    table: "your_table", // defaults to "vector_table"
  }
);

const results = await vectorStore.similaritySearch("hello world", 1);
console.log(results);

const filteredResults = await vectorStore.similaritySearch("hello world", 1, {
  whereStr: "metadata.name = '1'",
});
console.log(filteredResults);

API Reference:

相关内容​

Was this page helpful?


You can also leave detailed feedback on GitHub.