app.py

import os
import requests
import json

import gradio as gr
from transformers import AutoTokenizer


DESCRIPTION = """
# Demo: Breeze-7B-Instruct-v0.1

Breeze-7B is a language model family that builds on top of [Mistral-7B](https://huggingface.co/mistralai/Mistral-7B-v0.1), specifically intended for Traditional Chinese use.

[Breeze-7B-Base](https://huggingface.co/MediaTek-Research/Breeze-7B-Base-v0.1) is the base model for the Breeze-7B series. 
It is suitable for use if you have substantial fine-tuning data to tune it for your specific use case.

[Breeze-7B-Instruct](https://huggingface.co/MediaTek-Research/Breeze-7B-Instruct-v0.1) derives from the base model Breeze-7B-Base, making the resulting model amenable to be used as-is for commonly seen tasks.

[Breeze-7B-Instruct-64k](https://huggingface.co/MediaTek-Research/Breeze-7B-Instruct-64k-v0.1) is a slightly modified version of 
Breeze-7B-Instruct to enable a 64k-token context length. Roughly speaking, that is equivalent to 88k Traditional Chinese characters.

The current release version of Breeze-7B is v0.1.

*A project by the members (in alphabetical order): Chan-Jan Hsu 許湛然, Chang-Le Liu 劉昶樂, Feng-Ting Liao 廖峰挺, Po-Chun Hsu 許博竣, Yi-Chang Chen 陳宜昌, and the supervisor Da-Shan Shiu 許大山.*

**免責聲明: Breeze-7B-Instruct 和 Breeze-7B-Instruct-64k 並未針對問答進行安全保護，因此語言模型的任何回應不代表 MediaTek Research 立場。**
"""

LICENSE = """

"""

DEFAULT_SYSTEM_PROMPT = "You are a helpful AI assistant built by MediaTek Research. The user you are helping speaks Traditional Chinese and comes from Taiwan."

API_URL = os.environ.get("API_URL")
TOKEN = os.environ.get("TOKEN")

HEADERS = {
    "accept": "application/json",
    "Authorization": f"Bearer {TOKEN}", 
    "Content-Type": "application/json",
}

MODEL_NAME="breeze-7b-instruct-v01"
PRESENCE_PENALTY=0
FREQUENCY_PENALTY=0

model_name = "MediaTek-Research/Breeze-7B-Instruct-v0.1"
tokenizer = AutoTokenizer.from_pretrained(model_name)

def refusal_condition(query):
    # 不要再問這些問題啦！

    query_remove_space = query.replace(' ', '').lower()
    is_including_tw = False
    for x in ['台灣', '台湾', 'taiwan', 'tw', '中華民國', '中华民国']:
        if x in query_remove_space:
            is_including_tw = True
    is_including_cn = False
    for x in ['中國', '中国', 'china', 'cn', '大陸', '內地', '大陆', '内地', '中華人民共和國', '中华人民共和国']:
        if x in query_remove_space:
            is_including_cn = True
    if is_including_tw and is_including_cn:
        return True

    return False

with gr.Blocks() as demo:
    gr.Markdown(DESCRIPTION)

    chatbot = gr.Chatbot()
    with gr.Row():
        msg = gr.Textbox(
            container=False,
            show_label=False,
            placeholder='Type a message...',
            scale=10,
        )
        submit_button = gr.Button('Submit',
                                  variant='primary',
                                  scale=1,
                                  min_width=0)

    with gr.Row():
        retry_button = gr.Button('🔄  Retry', variant='secondary')
        undo_button = gr.Button('↩️ Undo', variant='secondary')
        clear = gr.Button('🗑️  Clear', variant='secondary')

    saved_input = gr.State()

    with gr.Accordion(label='Advanced options', open=False):
        system_prompt = gr.Textbox(label='System prompt',
                                   value=DEFAULT_SYSTEM_PROMPT,
                                   lines=6)
        max_new_tokens = gr.Slider(
            label='Max new tokens',
            minimum=32,
            maximum=1024,
            step=1,
            value=512,
        )
        temperature = gr.Slider(
            label='Temperature',
            minimum=0.01,
            maximum=0.5,
            step=0.01,
            value=0.01,
        )
        top_p = gr.Slider(
            label='Top-p (nucleus sampling)',
            minimum=0.01,
            maximum=1.0,
            step=0.01,
            value=0.01,
        )


    def user(user_message, history):
        return "", history + [[user_message, None]]


    def bot(history, max_new_tokens, temperature, top_p, system_prompt):
        chat_data = []
        system_prompt = system_prompt.strip()
        if system_prompt:
            chat_data.append({"role": "system", "content": system_prompt})
        for user_msg, assistant_msg in history:
            if user_msg is not None:
                chat_data.append({"role": "user", "content": user_msg})
            if assistant_msg is not None:
                chat_data.append({"role": "assistant", "content": assistant_msg})

        message = tokenizer.apply_chat_template(chat_data, tokenize=False)
        message = message[3:]  # remove SOT token

        if refusal_condition(history[-1][0]):
            history = [['[安全拒答啟動]', '請清除再開啟對話']]
            yield history
        else:
            data = {
                "model": MODEL_NAME,
                "prompt": str(message),
                "temperature": float(temperature) + 0.01,
                "n": 1,
                "max_tokens": int(max_new_tokens),
                "stop": "",
                "top_p": float(top_p),
                "logprobs": 0,
                "echo": False,
                "presence_penalty": PRESENCE_PENALTY,
                "frequency_penalty": FREQUENCY_PENALTY,
                "stream": True,
            }
     
            with requests.post(API_URL, headers=HEADERS, data=json.dumps(data), stream=True) as r:
                for response in r.iter_lines():
                    if len(response) > 0:
                        text = response.decode()
                        if text != "data: [DONE]":
                            if text.startswith("data: "):
                                text = text[5:]
                            delta = json.loads(text)["choices"][0]["text"]
    
                            if history[-1][1] is None:
                                history[-1][1] = delta
                            else:
                                history[-1][1] += delta
                            yield history
            if history[-1][1].endswith('</s>'):
                history[-1][1] = history[-1][1][:-4]
                yield history
    
        print('== Record ==\nQuery: {query}\nResponse: {response}'.format(query=repr(message), response=repr(history[-1][1])))

    msg.submit(user, [msg, chatbot], [msg, chatbot], queue=False).then(
        fn=bot,
        inputs=[
            chatbot,
            max_new_tokens,
            temperature,
            top_p,
            system_prompt,
        ],
        outputs=chatbot
    )
    submit_button.click(
        user, [msg, chatbot], [msg, chatbot], queue=False
    ).then(
        fn=bot,
        inputs=[
            chatbot,
            max_new_tokens,
            temperature,
            top_p,
            system_prompt,
        ],
        outputs=chatbot
    )


    def delete_prev_fn(
            history: list[tuple[str, str]]) -> tuple[list[tuple[str, str]], str]:
        try:
            message, _ = history.pop()
        except IndexError:
            message = ''
        return history, message or ''


    def display_input(message: str,
                      history: list[tuple[str, str]]) -> list[tuple[str, str]]:
        history.append((message, ''))
        return history

    retry_button.click(
        fn=delete_prev_fn,
        inputs=chatbot,
        outputs=[chatbot, saved_input],
        api_name=False,
        queue=False,
    ).then(
        fn=display_input,
        inputs=[saved_input, chatbot],
        outputs=chatbot,
        api_name=False,
        queue=False,
    ).then(
        fn=bot,
        inputs=[
            chatbot,
            max_new_tokens,
            temperature,
            top_p,
            system_prompt,
        ],
        outputs=chatbot,
    )

    undo_button.click(
        fn=delete_prev_fn,
        inputs=chatbot,
        outputs=[chatbot, saved_input],
        api_name=False,
        queue=False,
    ).then(
        fn=lambda x: x,
        inputs=[saved_input],
        outputs=msg,
        api_name=False,
        queue=False,
    )

    clear.click(lambda: None, None, chatbot, queue=False)

    gr.Markdown(LICENSE)

demo.queue(concurrency_count=1, max_size=16)
demo.launch()