app.py

from huggingface_hub import InferenceClient

import gradio as gr
import random

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


def respond(message, history):
    # responses = ["Yes", "No"]
    # return random.choice(responses)

    messages = [
        {"role":"system",
        "content": "You are a friendly chatbot!"
      }  
    ]
    
    if history:
        messages.extend(history)
    
    response = client.chat_completion(
        messages, max_tokens=100
    )
    
    return response['choices'][0]['message']['content'].strip()

chatbot = gr.ChatInterface(respond, type="messages")
chatbot.launch()