Momento 支持的聊天记忆

对于跨聊天会话的分布式、无服务器持久性，您可以使用 Momento 支持的聊天消息历史记录。由于 Momento 缓存可立即使用且无需任何基础设施维护，因此无论是在本地构建还是在生产中，它都是开始使用聊天历史记录的绝佳方法。

设置

您需要在您的项目中安装 Momento 客户端库。鉴于 Momento 与 Node.js、浏览器和边缘环境兼容，请确保安装了相关软件包。

要为 Node.js 安装

npm

npm install @gomomento/sdk

Yarn

yarn add @gomomento/sdk

pnpm

pnpm add @gomomento/sdk

要为 浏览器/边缘工作者 安装

npm

npm install @gomomento/sdk-web

Yarn

yarn add @gomomento/sdk-web

pnpm

pnpm add @gomomento/sdk-web

提示

请参阅此部分以获取有关安装集成软件包的通用说明.

npm

npm install @langchain/openai @langchain/community

Yarn

yarn add @langchain/openai @langchain/community

pnpm

pnpm add @langchain/openai @langchain/community

您还需要从 Momento 获取 API 密钥。您可以在这里注册免费帐户。

用法

为了区分一个聊天历史会话与另一个，我们需要一个唯一的 sessionId。您也可以提供一个可选的 sessionTtl 使会话在给定秒数后过期。

import {
  CacheClient,
  Configurations,
  CredentialProvider,
} from "@gomomento/sdk"; // `from "gomomento/sdk-web";` for browser/edge
import { BufferMemory } from "langchain/memory";
import { ChatOpenAI } from "@langchain/openai";
import { ConversationChain } from "langchain/chains";
import { MomentoChatMessageHistory } from "@langchain/community/stores/message/momento";

// See https://github.com/momentohq/client-sdk-javascript for connection options
const client = new CacheClient({
  configuration: Configurations.Laptop.v1(),
  credentialProvider: CredentialProvider.fromEnvironmentVariable({
    environmentVariableName: "MOMENTO_API_KEY",
  }),
  defaultTtlSeconds: 60 * 60 * 24,
});

// Create a unique session ID
const sessionId = new Date().toISOString();
const cacheName = "langchain";

const memory = new BufferMemory({
  chatHistory: await MomentoChatMessageHistory.fromProps({
    client,
    cacheName,
    sessionId,
    sessionTtl: 300,
  }),
});
console.log(
  `cacheName=${cacheName} and sessionId=${sessionId} . This will be used to store the chat history. You can inspect the values at your Momento console at https://console.gomomento.com.`
);

const model = new ChatOpenAI({
  model: "gpt-3.5-turbo",
  temperature: 0,
});

const chain = new ConversationChain({ llm: model, memory });

const res1 = await chain.invoke({ input: "Hi! I'm Jim." });
console.log({ res1 });
/*
{
  res1: {
    text: "Hello Jim! It's nice to meet you. My name is AI. How may I assist you today?"
  }
}
*/

const res2 = await chain.invoke({ input: "What did I just say my name was?" });
console.log({ res2 });

/*
{
  res1: {
    text: "You said your name was Jim."
  }
}
*/

// See the chat history in the Momento
console.log(await memory.chatHistory.getMessages());

API 参考

• BufferMemory 来自 langchain/memory
• ChatOpenAI 来自 @langchain/openai
• ConversationChain 来自 langchain/chains
• MomentoChatMessageHistory 来自 @langchain/community/stores/message/momento