app.py

import gradio as gr
from diffusers import DDPMPipeline

model = 'alexktrs/CumulusCloudsGenerator'
device='cpu'

generator = DDPMPipeline.from_pretrained(model)
generator.to(device)

def generate(num_inference_steps):

    image1 = generator(num_inference_steps=num_inference_steps).images[0]
    image2 = generator(num_inference_steps=num_inference_steps).images[0]
    
    return [image1, image2]

with gr.Blocks() as demo:
    gr.Markdown("""
                
                # Generate Cumulus Clouds

                """)
    
    gallery=gr.Gallery(type="pil")
    slider=gr.Slider(label='Inference Steps', minimum=0, maximum=100)
    
    btn = gr.Button("Generate Clouds")
    btn.click(fn=generate, inputs=slider, outputs=gallery)

demo.launch()