app.py

import os
import gradio as gr
from openai import OpenAI
from openai.error import BadRequestError

# Retrieve the Hugging Face API token from environment variables
TOKEN = os.getenv("HF_TOKEN")
if not TOKEN:
    raise ValueError("Hugging Face API token (HF_TOKEN) not set in environment variables.")

client = OpenAI(
    base_url="https://api-inference.huggingface.co/v1/",
    api_key=TOKEN,
)

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

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

    messages.append({"role": "user", "content": message})

    try:
        response = ""
        for msg in client.chat.completions.create(
            model="meta-llama/Meta-Llama-3.1-405B-Instruct-FP8",
            max_tokens=max_tokens,
            stream=True,
            temperature=temperature,
            top_p=top_p,
            messages=messages,
        ):
            token = msg.choices[0].delta.content
            response += token
            yield response
    except BadRequestError as e:
        error_message = f"Error: {e}. Please ensure your Hugging Face token is valid and you have a Pro subscription."
        yield error_message

# Define the Gradio interface
demo = gr.ChatInterface(
    fn=respond,
    additional_inputs=[
        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
        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)"),
    ],
)

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