app.py

import gradio as gr
from transformers import pipeline

# Загрузка модели Marco-o1 с квантизацией
pipe = pipeline("text-generation", model="AIDC-AI/Marco-o1", device_map="auto", torch_dtype="auto", trust_remote_code=True)

def respond(
    message,
    history: list[tuple[str, str]],
    system_message,
    max_tokens,
    temperature,
    top_p,
):
    messages = [system_message]
    
    for val in history:
        if val[0]:
            messages.append(val[0])
        if val[1]:
            messages.append(val[1])

    messages.append(message)
    
    # Объединяем все сообщения в одну строку для передачи в модель
    input_text = "\n".join(messages)
    
    response = pipe(
        input_text,
        max_length=max_tokens + len(input_text),
        temperature=temperature,
        top_p=top_p,
        num_return_sequences=1
    )[0]['generated_text']
    
    # Извлекаем новый ответ, исключая входные сообщения
    new_response = response[len(input_text):].strip()
    
    yield new_response

"""
For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
"""
demo = gr.ChatInterface(
    respond,
    additional_inputs=[
        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
        gr.Slider(
            minimum=0.1,
            maximum=1.0,
            value=0.95,
            step=0.05,
            label="Top-p (nucleus sampling)",
        ),
    ],
)

if __name__ == "__main__":
    demo.launch()