app.py

import gradio as gr
from huggingface_hub import InferenceClient

# Initialize the InferenceClient
client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")

def respond(
    message,
    history: list[dict],  # Use a list of dictionaries instead of tuples
    system_message,
    max_tokens,
    temperature,
    top_p,
):
    messages = [{"role": "system", "content": system_message}]

    for val in history:
        messages.append({"role": val['role'], "content": val['content']})

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

    response = ""
    
    # Use chat_completion to get responses
    for message in client.chat_completion(
        messages,
        max_tokens=max_tokens,
        stream=True,
        temperature=temperature,
        top_p=top_p,
    ):
        token = message.choices[0].delta.content
        response += token
        yield response

# Create the Gradio Interface for API
api_interface = gr.Interface(
    fn=respond,
    inputs=[
        gr.Textbox(label="Message"),
        gr.JSON(label="History"),
        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)",
        ),
    ],
    outputs=gr.Textbox(label="Response"),
)

# Launch the API
if __name__ == "__main__":
    api_interface.launch(server_name="0.0.0.0", server_port=7860, share=False)  # Set share=False to avoid Hugging Face Spaces