app.py

import os
import json
from datetime import datetime

import gradio as gr
from openai import OpenAI


def print_now(msg):
    now = datetime.now()
    formatted_time = now.strftime("%Y-%m-%d %H:%M:%S.%f")
    print(f"{msg}:{formatted_time}")
    return formatted_time

def respond(
    message,
    history: list[tuple[str, str]],
    system_message,
    max_tokens,
    temperature,
    top_p,
):
    try:
        weekdays = ["周一", "周二", "周三", "周四", "周五", "周六", "周日"]
        now = datetime.now()
        weekday_num = now.weekday()
        weekday_chinese = weekdays[weekday_num]
        formatted_time = now.strftime("%Y-%m-%d %H:%M:%S") + " " + weekday_chinese
        default_system = f"你是一个由腾讯开发的有用的人工智能助手，你的名字是“腾讯元宝”，简称“元宝”，你的英文名是“Tencent Yuanbao”，你乐于帮助大家解答问题。\n现在的时间是{formatted_time}"

        messages = [{"Role": "system", "Content": default_system}]
        client = OpenAI(
            api_key=os.getenv('HUNYUAN_API_KEY'),
            base_url="https://api.hunyuan.cloud.tencent.com/v1",
        )
        for val in history:
            if val[0] and val[1]:
                messages.append({"Role": "user", "Content": val[0]})
                pure_response = val[1].split("**End thinking**")[-1].strip()
                messages.append({"Role": "assistant", "Content": pure_response})
                
        
        messages.append({"Role": "user", "Content": message})
        completion = client.chat.completions.create(
            model="hunyuan-t1-latest",
            messages=messages,
            stream=True,
            extra_body={
            "stream_moderation": True,
            "enable_enhancement": False,
            }
        )
        response = ""
        is_reasoning_start = True
        is_reasoning_end = True
        

        for event in completion:
            if message == "深圳哪里好玩？":
                print(event)
                print(event.choices[0].delta)
            if hasattr(event.choices[0].delta, 'reasoning_content'):
                if is_reasoning_start:
                    response += '> **Start thinking**\n\n'
                    is_reasoning_start = False
                token = event.choices[0].delta.reasoning_content
                response += token
            else:
                if is_reasoning_end:
                    response += '> **End thinking**\n\n'
                    is_reasoning_end = False
                token = event.choices[0].delta.answer_content
                response += token
            yield response
    except Exception as e:
        raise gr.Error(f"发生错误: {str(e)}")

example_prompts = [
    ["Write a short papragraph where the 1st letter of each sentence spells out the word 'CODE'. The message should appear natural and not obviously hide this pattern."],
    ["Compose an engaging travel blog post about a recent trip to Hawaii, highlighting cultural experiences and must-see attractions."],
    ["Why has online learning been able to spread rapidly in recent years?"],
    ["How many 'e' in Deeplearning?"],
    ["Write a 3-line poem"]
]
latex_delimiters = [
    {"left": "$$", "right": "$$", "display": True},
    {"left": "\\[", "right": "\\]", "display": True},{"left": "$", "right": "$", "display": False},
    {"left": "\\(", "right": "\\)", "display": False}
]


chatbot = gr.Chatbot(latex_delimiters=latex_delimiters, scale=9)


demo = gr.ChatInterface(respond,
    title="Hunyuan-T1",
    examples=example_prompts,
    chatbot=chatbot,
    description="当前体验demo为非联网Hunyuan-T1 最新推理模型，完整版联网/非联网能力即将在元宝上线，敬请期待!</br>The current  demo is the latest reasoning  model of Hunyuan-T1. The full version with browsing will be launched on Tencent Yuanbao soon. Stay tuned."                    

)

if __name__ == "__main__":
    demo.queue(default_concurrency_limit=100)
    demo.launch(max_threads=100)