English | 中文

AwesomeChat

Feature

🚀Quick Reply 🤖Streaming Reply 😎Conversation Memory 👏Chat Interruption 😃Rapid Deployment

Install

npm install awesomechat
or
yarn add awesomechat
or
pnpm install awesomechat

Usage

Simple

import AwesomeChatBot from "awesomechat";
const key = "sk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx"; // replace with your key
const systemDescription = "You are a good robot";
const chatBot = new AwesomeChatBot(key,systemDescription);

chatBot.chat({
  chatContext: "hello?",
  onReplying(word, reply, id) {
    console.log("onReplying", reply);
  },
  onError(error) {
    console.log("onError", error);
  },
});

With Context

import AwesomeChatBot from "awesomechat";
const key = "sk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx"; // replace with your key
const systemDescription =
  "You're a robot with emotions, and you need to add an emoji to the front of each sentence to express your emotions";

const chatBot = new AwesomeChatBot(key, systemDescription);

chatBot.chat({
  chatContext: [
    { role: "user", content: "100" },
    { role: "user", content: "200" },
    { role: "user", content: "300" },
    {
      role: "user",
      content: "What is the maximum number in the dialog above",
    },
  ],

  onReplying(word, reply, id) {
    console.log("onReplying", reply);
  },
  onReplyEnd(completion, id, usage) {
    if (completion.includes("300")) {
      console.log("😄You are right!");
    }
  },
  onError(error) {
    console.log("onError", error);
  },
});

Abort

import AwesomeChatBot from "awesomechat";
const key = "sk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx"; // replace with your key
const chatBot = new AwesomeChatBot(key);

chatBot.chat({
  chatContext: "Write a 500-word essay about AI",
  onReplyStart() {
    console.log("🚀copy that!");
  },
  onReplying(word, reply) {
    console.log("onReplying", reply);
  },
  onCancel(id, usage) {
    console.log("onCancel", id, usage);
  },
});

setTimeout(() => {
  chatBot.cancel();
}, 6000);

API

bot.chat({options})

Key | Type | Description --- | --- | --- |chatContext | ChatContext [] / string | An array of chat context objects or a string representing the current conversation context. |beforeReplyStart | (promptTokens: number) => boolean; | A function called before the conversation starts, returning a boolean value. If the value is true, the conversation continues; if the value is false, the following content will be interrupted. Typically used in scenarios such as conditional verification and authentication. |onReplyStart| (id: string) => void; | Function called when the model starts generating a response. |onReplying| (word: string, reply: string, id: string) => void; | Function called when the model is generating a response. The word parameter is the current character being generated, and the reply parameter is the concatenation result of the model's current generation content. |onReplyEnd| (completion: string, id: string, usage: ChatUsage) => void; | Function called when the model finishes generating a response. The completion parameter is the final generated response. |onError | (error: Error) => void; | Callback function called when an error occurs. The error parameter is an error object. |onCancel | (id: string, usage: ChatUsage)

bot.cancel()

Cancel the current chat in progress.

types

ChatContext

type ChatContext = {
    role: "assistant" | "user" | "system";
    content: string;
};

ChatConfig

type ChatConfig = {
    /**  What sampling temperature to use, between 0 and 2 */
    temperature?: number;
    /**  An alternative to sampling with temperature, where the model considers the results of the tokens with top_p probability mass */
    top_p?: number;
    /**  How many chat completion choices to generate for each input message */
    n?: number;
    /**  Up to 4 sequences where the API will stop generating further tokens */
    stop?: string | string[];
    /**  The maximum number of tokens to generate in the chat completion */
    max_tokens?: number;
    /**  Number between -2.0 and 2.0 */
    presence_penalty?: number;
    /**  Number between -2.0 and 2.0 */
    frequency_penalty?: number;
    /**  A json object that maps tokens to an associated bias value from -100 to 100 */
    logit_bias?: {
        [key: string]: number;
    };
    /**  A unique identifier representing your end-user */
    user?: string;
};

ChatUsage

type ChatUsage = {
    /**  The number of total tokens used*/
    totalTokens: number;
    /**  The number of requests made to the API */
    promptTokens: number;
    /**  The number of tokens generated by the model */
    completionTokens: number;
};