app.py

import gradio as gr
from huggingface_hub import InferenceClient

client = InferenceClient("unsloth/Meta-Llama-3.1-8B-Instruct-bnb-4bit")

def respond(
    message,
    history: list[tuple[str, str]],
    system_message,
    max_tokens,
    temperature,
    top_p,
):
    messages = [{"role": "system", "content": system_message}]

    for user_text, assistant_text in history:
        if user_text:
            messages.append({"role": "user", "content": user_text})
        if assistant_text:
            messages.append({"role": "assistant", "content": assistant_text})

    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

# The additional_inputs are now hidden by setting visible=False.
demo = gr.ChatInterface(
    respond,
    additional_inputs=[
        gr.Textbox(
            value="You are a friendly Chatbot. Respond only in bisaya language. No english translation.",
            label="System message",
            visible=False,
        ),
        gr.Slider(
            minimum=1,
            maximum=2048,
            value=512,
            step=1,
            label="Max new tokens",
            visible=False,
        ),
        gr.Slider(
            minimum=0.1,
            maximum=4.0,
            value=0.7,
            step=0.1,
            label="Temperature",
            visible=False,
        ),
        gr.Slider(
            minimum=0.1,
            maximum=1.0,
            value=0.95,
            step=0.05,
            label="Top-p (nucleus sampling)",
            visible=False,
        ),
    ],
)

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