app.py

import gradio as gr
from huggingface_hub import InferenceClient

client = InferenceClient("microsoft/Phi-3.5-mini-instruct")

# Predefined prompts
prompts = [
    "Tell me a joke about programming",
    "Write a short story about a time-traveling robot",
    "Explain quantum computing to a 5-year-old",
    "Create a recipe for the most unusual pizza",
    "Describe an alien civilization's first contact with Earth"
]

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.text_generation(
        prompt="\n".join([m["content"] for m in messages]),
        max_new_tokens=max_tokens,
        stream=True,
        temperature=temperature,
        top_p=top_p,
    ):
        token = message.token.text

        response += token
        yield response

def update_textbox(prompt):
    return gr.Textbox.update(value=prompt)

with gr.Blocks() as demo:
    chatbot = gr.Chatbot()
    msg = gr.Textbox()
    clear = gr.ClearButton([msg, chatbot])

    with gr.Accordion("Advanced options", open=False):
        system = gr.Textbox(value="You are a friendly Chatbot.", label="System message")
        max_tokens = gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens")
        temperature = gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature")
        top_p = gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)")

    prompt_dropdown = gr.Dropdown(choices=prompts, label="Select a premade prompt")
    prompt_dropdown.change(update_textbox, inputs=[prompt_dropdown], outputs=[msg])

    msg.submit(respond, [msg, chatbot, system, max_tokens, temperature, top_p], chatbot)
    clear.click(lambda: None, None, chatbot, queue=False)

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