app.py

import gradio as gr
from huggingface_hub import InferenceClient

client = InferenceClient("facebook/opt-350m")

def respond(
    message,
    history: list[tuple[str, str]],
    system_message,
    max_tokens,
    temperature,
    top_p,
):
    # Construct the prompt from history and current message
    prompt = system_message + "\n\n"
    for user_msg, bot_msg in history:
        prompt += f"Human: {user_msg}\nAI: {bot_msg}\n"
    prompt += f"Human: {message}\nAI:"

    # Generate response
    response = client.text_generation(
        prompt,
        max_new_tokens=max_tokens,
        temperature=temperature,
        top_p=top_p,
        do_sample=True,
    )

    # Extract only the AI's response
    ai_response = response.split("AI:")[-1].strip()
    
    return ai_response

demo = gr.ChatInterface(
    respond,
    additional_inputs=[
        gr.Textbox(value="You are a helpful AI assistant.", label="System message"),
        gr.Slider(minimum=1, maximum=256, value=128, step=1, label="Max new tokens"),
        gr.Slider(minimum=0.1, maximum=2.0, 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()