Skip to main content

Gradient AI

LangChain.js 支持与 Gradient AI 集成。查看 Gradient AI 获取可用模型列表。

¥LangChain.js supports integration with Gradient AI. Check out Gradient AI for a list of available models.

设置​

¥Setup

你需要安装官方 Gradient Node SDK 作为对等依赖：

¥You'll need to install the official Gradient Node SDK as a peer dependency:

npm i @gradientai/nodejs-sdk

你需要设置以下环境变量才能使用 Gradient AI API。

¥You will need to set the following environment variables for using the Gradient AI API.

  1. GRADIENT_ACCESS_TOKEN
  2. GRADIENT_WORKSPACE_ID

或者，可以在 GradientAI 类实例化期间分别将它们设置为 gradientAccessKey 和 workspaceId。例如：

¥Alternatively, these can be set during the GradientAI Class instantiation as gradientAccessKey and workspaceId respectively. For example:

const model = new GradientLLM({
  gradientAccessKey: "My secret Access Token"
  workspaceId: "My secret workspace id"
});

用法​

¥Usage

tip

请参阅 此部分包含有关安装集成包的一般说明。

¥See this section for general instructions on installing integration packages.

npm install @langchain/community @langchain/core

使用 Gradient 的基础模型​

¥Using Gradient's Base Models

import { GradientLLM } from "@langchain/community/llms/gradient_ai";

// Note that inferenceParameters are optional
const model = new GradientLLM({
  modelSlug: "llama2-7b-chat",
  inferenceParameters: {
    maxGeneratedTokenCount: 20,
    temperature: 0,
  },
});
const res = await model.invoke(
  "What would be a good company name for a company that makes colorful socks?"
);

console.log({ res });

API Reference:

使用自定义微调适配器​

¥Using your own fine-tuned Adapters

要使用你自己的自定义适配器，只需在设置期间设置 adapterId 即可。

¥The use your own custom adapter simply set adapterId during setup.

import { GradientLLM } from "@langchain/community/llms/gradient_ai";

// Note that inferenceParameters are optional
const model = new GradientLLM({
  adapterId: process.env.GRADIENT_ADAPTER_ID,
  inferenceParameters: {
    maxGeneratedTokenCount: 20,
    temperature: 0,
  },
});
const res = await model.invoke(
  "What would be a good company name for a company that makes colorful socks?"
);

console.log({ res });

API Reference:

¥Related