app.py

import numpy as np
import gradio as gr

def flip_text(x):
    return x[::-1]

def flip_image(x):
    return np.fliplr(x)

interface_title = "TSAI-ERAV1 Capstone - Multimodal GPT"

with gr.Interface(
    fn=None,
    live=True,
    title=interface_title,
    layout="vertical",
    description="Choose text mode/image mode/audio mode for generation",
) as demo:

    with demo.Tab("Text mode"):
        text_input = demo.Textbox(placeholder="Enter a prompt", label="Input")
        text_input_count = demo.Textbox(placeholder="Enter number of characters you want to generate", label="Count")
        text_button = demo.Button("Generate Text")
        text_output = demo.Textbox(label="Chat GPT like text")        

        text_button.click(flip_text, inputs=text_input, outputs=text_output)

    with demo.Tab("Image mode"):
        image_input = demo.Image()
        image_text_input = demo.Textbox(placeholder="Enter a question/prompt around the image", label="Question/Prompt")
        image_button = demo.Button("Generate Text")   
        image_text_output = demo.Textbox(label="Answer")

        image_button.click(flip_text, inputs=image_text_input, outputs=image_text_output)

    with demo.Tab("Audio mode"):
        audio_text_input = demo.Textbox()
        audio_text_output = demo.Textbox()
        audio_button = demo.Button("Generate Text")

        audio_button.click(flip_text, inputs=audio_text_input, outputs=audio_text_output)

demo.launch()