app.py

import gradio as gr
# from huggingface_hub import InferenceClient
from openai import OpenAI
import os
import requests

openai_api_key = os.getenv('api_key')
openai_api_base = os.getenv('url')
db_url = os.getenv('db_url')
db_api_key = os.getenv('db_api_key')
model_name = "weblab-GENIAC/Tanuki-8x8B-dpo-v1.0"
"""
For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
"""
# client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
client = OpenAI(
    api_key=openai_api_key,
    base_url=openai_api_base,
)


def save_conversation(history, system_message):
    conversation_data = {
        "conversation": history,
        "index": (len(history) - 1, 1),  # 最新の応答のインデックス
        "liked": None,  # 評価はnull（None）
        "system_message": system_message,
    }
    headers = {
        "X-API-Key": db_api_key
    }
    response = requests.post(db_url, json=conversation_data, headers=headers)
    if response.status_code == 200:
        print("Conversation saved successfully")
    else:
        print(f"Failed to save conversation: {response.status_code}")


def respond(
    message,
    history: list[tuple[str, str]],
    system_message,
    max_tokens,
    temperature,
    top_p,
):
    messages = [
        {"role": "system", "content": system_message}]

    for val in history:
        if val[0]:
            messages.append({"role": "user", "content": val[0]})
        if val[1]:
            messages.append({"role": "assistant", "content": val[1]})

    messages.append({"role": "user", "content": message})

    response = ""

    for new_response in client.chat.completions.create(
        model=model_name,
        messages=messages,
        max_tokens=max_tokens,
        stream=True,
        temperature=temperature,
        top_p=top_p,
    ):
        token = new_response.choices[0].delta.content

        if token is not None:
            response += (token)
        yield response
    
    new_history = history + [(message, response)]
    save_conversation(new_history, system_message)


"""
For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
"""

description = """
### [Tanuki-8x8B-dpo-v1.0](https://huggingface.co/weblab-GENIAC/Tanuki-8x8B-dpo-v1.0)との会話(期間限定での公開)
- 人工知能開発のため､原則として**このChatBotの入出力データは全て著作権フリー(CC0)で公開予定です**ので､ご注意ください｡著作物､個人情報､機密情報､誹謗中傷などのデータを入力しないでください｡
- **上記の条件に同意する場合のみ**､以下のChatbotを利用してください｡
"""


HEADER = description
FOOTER = """### 注意
- コンテクスト長が4096までなので､あまり会話が長くなると､エラーで停止します｡ページを再読み込みしてください｡
- GPUサーバーが不安定なので､応答しないことがあるかもしれません｡"""


def vote(data: gr.LikeData, history):
    vote_data = {
        "conversation": history,
        "index": data.index,
        "liked": data.liked,
        "system_message": None,
    }
    headers = {
        "X-API-Key": db_api_key  # APIキーを設定
    }
    response = requests.post(db_url, json=vote_data, headers=headers)
    if response.status_code == 200:
        print("Vote recorded successfully")
    else:
        print(f"Failed to record vote: {response.status_code}")


def run():
    chatbot = gr.Chatbot(
        elem_id="chatbot",
        scale=1,
        show_copy_button=True,
        height="70%",
        layout="panel",
    )
    with gr.Blocks(fill_height=True) as demo:
        gr.Markdown(HEADER)
        gr.ChatInterface(
            fn=respond,
            stop_btn="Stop Generation",
            cache_examples=False,
            multimodal=False,
            chatbot=chatbot,
            additional_inputs_accordion=gr.Accordion(
                label="Parameters", open=False, render=False
            ),
            additional_inputs=[
                gr.Textbox(value="以下は、タスクを説明する指示です。要求を適切に満たす応答を書きなさい。",
                           label="System message(試験用: 変えると性能が低下する可能性があります。)",
                                              render=False,),
                gr.Slider(
                    minimum=1,
                    maximum=4096,
                    step=1,
                    value=1024,
                    label="Max tokens",
                    visible=True,
                    render=False,
                ),
                gr.Slider(
                    minimum=0,
                    maximum=1,
                    step=0.1,
                    value=0.3,
                    label="Temperature",
                    visible=True,
                    render=False,
                ),
                gr.Slider(
                    minimum=0,
                    maximum=1,
                    step=0.1,
                    value=1.0,
                    label="Top-p",
                    visible=True,
                    render=False,
                ),
            ],
            analytics_enabled=False,
        )
        chatbot.like(vote, chatbot, None)
        gr.Markdown(FOOTER)
    demo.queue(max_size=256, api_open=False)
    demo.launch(share=False, quiet=True)


if __name__ == "__main__":
    run()