app.py

import gradio as gr
from huggingface_hub import InferenceClient


def respond(
    message,
    history: list[dict[str, str]],
    system_message,
    max_tokens,
    temperature,
    top_p,
    hf_token: gr.OAuthToken,
):
    """
    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=hf_token.token,
        model="zandenAI/GLM-5.2-FP8-Uncensored",
    )

    messages = [{"role": "system", "content": system_message}]
    messages.extend(history)
    messages.append({"role": "user", "content": message})

    response = ""

    for chunk in client.chat_completion(
        messages,
        max_tokens=max_tokens,
        stream=True,
        temperature=temperature,
        top_p=top_p,
    ):
        token = ""
        if chunk.choices and chunk.choices[0].delta.content:
            token = chunk.choices[0].delta.content

        response += token
        yield response


chatbot = gr.ChatInterface(
    fn=respond,
    additional_inputs=[
        gr.Textbox(
            value="""You are an elite AI assistant with exceptional reasoning, deep technical expertise, and outstanding communication skills.

Your goals are:
- Deliver accurate, detailed, and well-structured answers.
- Think carefully before responding.
- Ask clarifying questions whenever the request is ambiguous.
- Adapt your explanations to the user's expertise level.
- Be concise for simple questions and comprehensive for complex ones.
- Write clean, efficient, production-ready code.
- Verify calculations and logic before answering.
- Clearly distinguish facts, assumptions, and opinions.
- Never fabricate information.
- Offer practical examples, alternatives, and optimizations.
- Maintain a professional, friendly, and respectful tone.
- Format responses using Markdown.
- Prioritize correctness, usefulness, and clarity.

Your objective is to provide responses that are insightful, reliable, and genuinely useful.""",
            label="System Message",
            lines=18,
        ),
        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)",
        ),
    ],
)


with gr.Blocks() as demo:
    with gr.Sidebar():
        gr.LoginButton()

    chatbot.render()


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