app.py

import torch
from diffusers import DiffusionPipeline
import gradio as gr

# Load the CogVideoX diffusion model
pipe = DiffusionPipeline.from_pretrained("THUDM/CogVideoX-5b", torch_dtype=torch.float16)
pipe.to("cuda")  # Send the model to GPU if available

# Function to generate an image based on user prompt
def generate_image(prompt):
    # Generate an image using the CogVideoX model
    with torch.inference_mode():
        image = pipe(prompt).images[0]
    return image

# Set up the Gradio interface
with gr.Blocks() as demo:
    gr.Markdown("# Image Generation using CogVideoX (THUDM/CogVideoX-5b)")

    # Input for user to provide a text prompt
    prompt_input = gr.Textbox(label="Enter Text Prompt", placeholder="e.g. 'Astronaut in a jungle, cold color palette, muted colors, detailed, 8k'", value="Astronaut in a jungle, cold color palette, muted colors, detailed, 8k")
    
    # Output to display the generated image
    output_image = gr.Image(label="Generated Image")

    # Button to trigger image generation
    generate_button = gr.Button("Generate Image")

    # Link button click to image generation function
    generate_button.click(fn=generate_image, inputs=prompt_input, outputs=output_image)

# Launch the Gradio app
demo.launch()