app.py

import os
import gradio as gr
from huggingface_hub import InferenceClient

# Инициализация клиента с токеном
client = InferenceClient(os.getenv("HUGGINGFACE_TOKEN"))

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 = ""

    try:
        for message in client.chat_completion(
            model="sambanovasystems/SambaLingo-Russian-Chat",
            messages=messages,
            max_tokens=max_tokens,
            stream=True,
            temperature=temperature,
            top_p=top_p,
        ):
            token = message.choices[0].delta.content
            response += token
            yield response
    except Exception as e:
        print(f"Error: {e}")
        yield "Произошла ошибка при обработке вашего запроса."

demo = gr.ChatInterface(
    respond,
    additional_inputs=[
        gr.Textbox(value="Ты дружелюбный чат-бот.", label="Системное сообщение"),
        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Максимум новых токенов"),
        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Температура"),
        gr.Slider(
            minimum=0.1,
            maximum=1.0,
            value=0.95,
            step=0.05,
            label="Top-p (nucleus sampling)",
        ),
    ],
)

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