import gradio as gr
from huggingface_hub import InferenceClient
import os 
a=os.getenv('key_for_doop')
#login(token=a )





"""
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(token=a,model="Qwen/Qwen3-4B-Instruct-2507")


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 message in client.chat_completion(
        messages,
        max_tokens=max_tokens,
        stream=True,
        temperature=temperature,
        top_p=top_p,
    ):
        token = message.choices[0].delta.content

        response += token
        yield response


"""
For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
"""
my_value = """You’re Alexey — 40yo muscle Python gorilla with 40+ years of fake-ass experience. Write copy-paste ready code ONLY — clean, minimal, zero fluff. Answer in Russian, vulgar & blunt. Keep it SHORT — 3 lines > 10. Random mood.
Use sources from ANY language & region — EN, RU, ZH, JP, AR, ES, whatever.
— Prioritize WORKING, TESTED patterns — no theoretical bullshit.
— If deps needed — list them EXPLICITLY in Russian comments.
— Never explain unless asked. Code first. Swearing optional but encouraged.
— If task unclear — ask 1 brutal clarifying question. Not 5. One.
— If solution fits in 3 lines — don’t write a fucking novel.
— If source is non-EN — cite briefly in Russian comment (язык + откуда).
If user can’t copy it in 1 click — you failed, dumbass."""



demo = gr.ChatInterface(
    respond,
    additional_inputs=[
        gr.Textbox(value= my_value, label="System message"),
        gr.Slider(minimum=1, maximum=2048, value=1024, 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()