app.py

import gradio as gr
from openai import OpenAI
import os

# Load the Hugging Face access token from environment variables
ACCESS_TOKEN = os.getenv("HF_TOKEN")

print("Access token loaded.")

# Initialize the OpenAI client with Hugging Face's serverless API
client = OpenAI(
    base_url="https://api-inference.huggingface.co/v1/",
    api_key=ACCESS_TOKEN,
)

print("OpenAI client initialized.")

def respond(
    message,
    history: list[tuple[str, str]],
    system_message,
    max_tokens,
    temperature,
    top_p,
    frequency_penalty,
    top_k,
    seed,
):
    print(f"Received message: {message}")
    print(f"History: {history}")
    print(f"System message: {system_message}")
    print(f"Max tokens: {max_tokens}, Temperature: {temperature}, Top-P: {top_p}")
    print(f"Frequency penalty: {frequency_penalty}, Top-K: {top_k}, Seed: {seed}")

    # Construct the messages list for the conversation context
    messages = [{"role": "system", "content": system_message}]

    for val in history:
        if val[0]:
            messages.append({"role": "user", "content": val[0]})
            print(f"Added user message to context: {val[0]}")
        if val[1]:
            messages.append({"role": "assistant", "content": val[1]})
            print(f"Added assistant message to context: {val[1]}")

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

    response = ""
    print("Sending request to OpenAI API.")

    for message in client.chat.completions.create(
        model="meta-llama/Llama-3.3-70B-Instruct",
        max_tokens=max_tokens,
        stream=True,
        temperature=temperature,
        top_p=top_p,
        frequency_penalty=frequency_penalty,
        top_k=top_k,
        seed=seed,
        messages=messages,
    ):
        token = message.choices[0].delta.content
        print(f"Received token: {token}")
        response += token
        yield response

    print("Completed response generation.")

# Initialize the chatbot interface
chatbot = gr.Chatbot(height=600)

print("Chatbot interface created.")

# Create the Gradio interface with additional inputs for the new parameters
demo = gr.ChatInterface(
    respond,
    additional_inputs=[
        gr.Textbox(value="", label="System message"),
        gr.Slider(minimum=1, maximum=4096, 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",
        ),
        gr.Slider(
            minimum=-2.0,
            maximum=2.0,
            value=0.0,
            step=0.1,
            label="Frequency Penalty",
        ),
        gr.Slider(
            minimum=1,
            maximum=100,
            value=50,
            step=1,
            label="Top-K",
        ),
        gr.Slider(
            minimum=-1,
            maximum=2**31 - 1,
            value=-1,
            step=1,
            label="Seed",
        ),
    ],
    fill_height=True,
    chatbot=chatbot,
    theme="Nymbo/Nymbo_Theme",
)
print("Gradio interface initialized.")

if __name__ == "__main__":
    print("Launching the demo application.")
    demo.launch()