app.py

import os

import gradio as gr
import requests

API_TOKEN = os.environ['API_TOKEN']

state = []

headers = {'Authorization': f'Bearer {API_TOKEN}'}

def query(payload, model_id):
    API_URL = f'https://api-inference.huggingface.co/models/{model_id}'
    response = requests.post(API_URL, headers=headers, json=payload)
    return response.json()

def chat(message):
    response = query({
        'inputs': message,
        'parameters': {
            'max_length': 500,
        }
    }, 'IssakaAI/health-chatbot')
    reply = ''
    if isinstance(response, list):
        reply = response[0]['generated_text'][len(message) + 1:]
    elif isinstance(response, dict):
        reply = f'{response["error"]}. Please wait about {int(response["estimated_time"])} seconds.'
    state.append((message, reply))
    return gr.Textbox.update(value=''), state

def clear_message():
    state.clear()
    return gr.Chatbot.update(value=[])

with gr.Blocks() as blk:
    gr.Markdown('# Interact with IssakaAI NLP models')
    with gr.Row():
        chatbot = gr.Chatbot()
        with gr.Box():
            message = gr.Textbox(value='What is the menstrual cycle?', lines=10)
            send = gr.Button('Send', variant='primary')
            clear = gr.Button('Clear history', variant='secondary')
    send.click(fn=chat, inputs=message, outputs=[message, chatbot])
    clear.click(fn=clear_message, inputs=[], outputs=chatbot)
blk.launch(debug=True)