app.py

import gradio as gr
from huggingface_hub import InferenceClient, get_inference_endpoint
from openai import OpenAI
import os

MODEL_URL_MAP = {"EuroLLM-9B-Instruct": os.getenv('ENDPOINT_URL_9B'),
}

def respond(
    message,
    history: list[tuple[str, str]],
    system_message,
    max_tokens,
    temperature,
    model
):
    client = OpenAI(
        base_url=MODEL_URL_MAP.get(model), 
    	api_key=os.getenv('AUTH_TOKEN')
    )
    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 = ""
        
    try:
        chat_completion = client.chat.completions.create(
        	model=model,
        	messages=messages,
        	temperature=temperature,
        	max_tokens=max_tokens,
        	stream=True,
        )  
        for message in chat_completion:
            token = message.choices[0].delta.content
            response += token
            yield response
    except Exception:
        yield response

default_system_prompt = """You are EuroLLM, a Large Language Model (LLM) optimized for European languages.
You power an AI assistant.
Your knowledge base was last updated on 2023-10-01.
The current date is 2024-12-04.

When you're not sure about some information, you say that you don't have the information and don't make up anything."""

demo = gr.ChatInterface(
    respond,
    additional_inputs=[
        gr.Textbox(value=default_system_prompt, label="System Prompt"),
        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
        gr.Slider(minimum=0.0, maximum=4.0, value=0.0, step=0.1, label="Temperature"),
        gr.Dropdown(["EuroLLM-9B-Instruct"], label="Model Name", value="EuroLLM-9B-Instruct")
    ],
)


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