backend.ai-client-js

Backend.AI Client Library for Javascript

  • 所有者: lablup/backend.ai-client-js
  • 平台:
  • 许可证: MIT License
  • 分类:
  • 主题:
  • 喜欢:
    0
      比较:

Github星跟踪图

Backend.AI Client for Javascript (node.js / ES6+)

Requirements

This client SDK runs on CommonJs (with node.js) / ES6-compatible Javascript runtimes with async/await supports
such as NodeJS 7+ and modern web browsers released since 2017.

This client library supports Backend.AI API v3 and above.

Install

$ npm install backend.ai-client

You can also use yarn.

$ yarn install backend.ai-client

Build

Package preparation

NPM:

$ npm i

Yarn:

$ yarn install

ES6+ library

$ make es6

Node.js library

$ make node 

Usage

TypeScript:

import * as ai from 'backend.ai-client-node';

let config = ai.backend.ClientConfig.createFromEnv();
let client = new ai.backend.Client(config);

CommonJS-style:

const ai = require('backend.ai-client-node');

let config = ai.backend.ClientConfig.createFromEnv();
let client = new ai.backend.Client(config);

ES6+:

import './backend.ai-client-es6.js';

let config = new ai.backend.ClientConfig(
  '[ADD_ACCESS_KEY_HERE]',
  '[ADD_SECRET_KEY_HERE]',
  '[ENDPOINT_HERE]'
);
let client = new ai.backend.Client(
  config,
  `Backend.AI ES6 App.`,
);

When creating ClientConfig object, you can manually pass accessKey,
secretKey, and optional endpoint arguments.
The environment variables are:

  • BACKEND_ACCESS_KEY
  • BACKEND_SECRET_KEY
  • BACKEND_ENDPOINT (optional, defaults to https://api.backend.ai)

All API functions return a promise that resolves into a parsed object
when success according to server-provided Content-Type and rejects with an
object with type and message attributes if failed.

client.createIfNotExists('python:latest', 'my-session-id')
.then(response => {
  console.log(`my session is created: ${response.kernelId}`);
}).catch(err => {
  switch (err.type) {
  case ai.backend.Client.ERR_SERVER:
    console.log(`session creation failed: ${err.message}`);
    break;
  default:
    console.log(`request/response failed: ${err.message}`);
  }
});

The result objects returned with success has different formats API by API.
Please check out our official documentation.

err.type is one of the following values:

  • ai.backend.Client.ERR_SERVER: The server responded with failure.
    In this case, err.message includes HTTP status and additional error information
    returned by the API server.
  • ai.backend.Client.ERR_RESPONSE: An error occurred while reading the response.
    err.message includes an exception value passed from your Javascript runtime.
  • ai.backend.Client.ERR_REQUEST: An error occurred while sending the request.
    err.message includes an exception value passed from your Javascript runtime.

主要指标

概览
名称与所有者lablup/backend.ai-client-js
主编程语言JavaScript
编程语言JavaScript (语言数: 3)
平台
许可证MIT License
所有者活动
创建于2017-09-01 00:49:40
推送于2023-04-19 02:19:07
最后一次提交2020-04-27 13:48:27
发布数0
用户参与
星数5
关注者数18
派生数4
提交数61
已启用问题?
问题数10
打开的问题数4
拉请求数6
打开的拉请求数12
关闭的拉请求数6
项目设置
已启用Wiki?
已存档?
是复刻?
已锁定?
是镜像?
是私有?