python - 我想为 twitch 创建我的 Neuro-sama 模拟，但是

Question

几天前我开始学习 Python，我做的第一件事是制作一个带有查询表单的 Python 应用程序，我们可以在其中提出问题并获得有声的文本到语音的答案，而前一天我发现发现有神经大人这样的表演，我想用技术术语来复制它 神经大人的工作原理是这样的:分析 Twitch Chat 并按常规向 ChatGPT 发送文本请求，然后当它收到来自 ChatGPT 的响应时，它会播放这种格式的响应“用户名，回复”

我发现这个任务很简单，并开始编写自己的代码，但我遇到了一些问题，我正在请求感兴趣的人帮助我，我准备支付 10 美元以获得合理的回应并指出我的错误，我在下面附上我失败的代码:

import tkinter as tk
import requests
import pyttsx3
import twitchio
import os

# Создание окна
root = tk.Tk()
root.title("Пример приложения с ChatGPT и озвучкой")

# Создание текстового поля и кнопки
text_field = tk.Entry(root, width=50)
text_field.pack()
button = tk.Button(root, text="Отправить")

# Создание объекта для озвучивания речи
engine = pyttsx3.init()

# Инициализация Twitch-клиента
bot_nickname = "pasha_tech"
bot_oauth = "YOUR_BOT_OAUTH_TOKEN" # замените "YOUR_BOT_OAUTH_TOKEN" на OAuth-токен вашего бота
bot = twitchio.Client(bot_nickname, oauth=bot_oauth)

# Функция для отправки запроса к API ChatGPT и получения ответа
def get_response(username, text):
    url = "https://api.openai.com/v1/engines/davinci-codex/completions"
    headers = {
        "Content-Type": "application/json",
        "Authorization": "Bearer API_KEY" # замените "API_KEY" на ваш API ключ от ChatGPT
    }
    data = {
        "prompt": f"Пользователь {username} написал в чате: {text}\nAI ответит:",
        "max_tokens": 50,
        "temperature": 0.7
    }
    response = requests.post(url, headers=headers, json=data)
    if response.status_code == 200:
        result = response.json()["choices"][0]["text"]
    else:
        result = "Ошибка при обработке запроса"
    # Отображение ответа в окне
    response_label.config(text=result)
    # Озвучивание ответа
    engine.say(result)
    engine.runAndWait()

# Функция-обработчик сообщений в чате Twitch
async def on_message(channel, user, message):
    if user.name != bot_nickname:
        # Отправка сообщения на обработку в ChatGPT
        get_response(user.name, message)
        
# Функция для подключения к IRC-чату Twitch
async def connect_to_twitch_chat():
    await bot.connect()
    await bot.join(os.environ['CHANNEL_NAME']) # замените "CHANNEL_NAME" на имя канала, к которому подключается бот
    print(f"Бот {bot_nickname} подключился к чату {os.environ['CHANNEL_NAME']}")

# Привязка функции к кнопке
button.config(command=get_response)
button.pack()

# Создание метки для отображения ответа
response_label = tk.Label(root, text="")
response_label.pack()

# Запуск бесконечного цикла обработки сообщений в чате Twitch
bot.loop.create_task(connect_to_twitch_chat())
bot.loop.run_forever()

root.mainloop()

使用这段代码我得到了错误:bot = twitchio.Client(bot_nickname, token=bot_token)TypeError:init() 为参数“token”获取了多个值，帮助我理解，这是我坐在这里想知道我的问题是什么的第 3 天

代码最初是通过ChatGPT生成的，请指出此服务可能存在的错误，以便我知道将来如何修复它们，谢谢!

Answer 1

我接近 AI Vtuber 项目的稳定版本，请改用新的 gpt-3.5-turbo 模型。此外，您还需要提供初始提示，以便我承担角色和反馈循环以提供短期内存(与机器人内存是自动的 chatgpt 不同)

https://www.youtube.com/watch?v=dkgJBcTitpE

尝试遵循文档中相同的代码结构:https://twitchio.dev/en/latest/quickstart.html

class Bot(commands.Bot):

    def __init__(self):
        # Initialise our Bot with our access token, prefix and a list of channels to join on boot...
        # prefix can be a callable, which returns a list of strings or a string...
        # initial_channels can also be a callable which returns a list of strings...

        # --- GET YOUR TWITCH ACCEESS_TOKEN USING: https://twitchtokengenerator.com/ ----

        super().__init__(token='ACCESS_TOKEN', prefix='?', initial_channels=['...'])

    async def event_ready(self):
        # Notify us when everything is ready!
        # We are logged in and ready to chat and use commands...
        print(f'Logged in as | {self.nick}')
        print(f'User id is | {self.user_id}')

    async def event_message(self, message):
        # Messages with echo set to True are messages sent by the bot...
        # For now we just want to ignore them...
        if message.echo:
            return

        # ---------------PUT YOUR OPEN AI  CODE HERE----------
        # SOME VARIABLE
        # user unique Id (never changes per account): message.author.id 
        # user chat name: message.author.name
        # user's message you include in your prompt: message.content

        # Since we have commands and are overriding the default `event_message`
        # We must let the bot know we want to handle and invoke our commands...
        await self.handle_commands(message)

        # WHEN YOU SEND ?hello THIS STILL GOES THROUGH event_message() THEN CAUGHT BY THIS METHOD
    @commands.command()
    async def hello(self, ctx: commands.Context):
        # Here we have a command hello, we can invoke our command with our prefix and command name
        # e.g ?hello
        # We can also give our commands aliases (different names) to invoke with.

        # Send a hello back!
        # Sending a reply back to the channel is easy... Below is an example.
        await ctx.send(f'Hello {ctx.author.name}!')


bot = Bot()
bot.run()    # bot.run() is blocking and will stop execution of any below code here until stopped or closed.

在开放的 Ai 代码块上集成此技术以赋予机器人内存和角色:https://github.com/daveshap/LongtermChatExternalSources