import gradio as gr import torch import numpy as np from PIL import Image from diffusers import DiffusionPipeline # Initialize the DiffusionPipeline model with LoRA weights pipeline = DiffusionPipeline.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0") pipeline.load_lora_weights("ostris/super-cereal-sdxl-lora") def text_to_image(prompt): # Generate image using the DiffusionPipeline output = pipeline(prompt) generated_img_tensor = output.images[0] # Convert torch tensor to numpy array generated_img_array = generated_img_tensor.cpu().numpy().transpose((1, 2, 0)) return generated_img_array def create_cereal_box(input_image): # Convert the input numpy array to PIL Image cover_img = Image.fromarray((input_image.astype(np.uint8))) # Load the template image template_img = Image.open('CerealBoxMaker/template.jpeg') # Simplified cereal box creation logic scaling_factor = 1.5 rect_height = int(template_img.height * 0.32) new_width = int(rect_height * 0.70) cover_resized = cover_img.resize((new_width, rect_height), Image.LANCZOS) new_width_scaled = int(new_width * scaling_factor) new_height_scaled = int(rect_height * scaling_factor) cover_resized_scaled = cover_resized.resize((new_width_scaled, new_height_scaled), Image.LANCZOS) left_x = int(template_img.width * 0.085) left_y = int((template_img.height - new_height_scaled) // 2 + template_img.height * 0.012) left_position = (left_x, left_y) right_x = int(template_img.width * 0.82) - new_width_scaled right_y = left_y right_position = (right_x, right_y) template_copy = template_img.copy() template_copy.paste(cover_resized_scaled, left_position) template_copy.paste(cover_resized_scaled, right_position) # Convert the PIL Image back to a numpy array template_copy_array = np.array(template_copy) return template_copy_array def combined_function(prompt): generated_img_array = text_to_image(prompt) final_img = create_cereal_box(generated_img_array) return final_img # Create a Gradio Interface gr.Interface(fn=combined_function, inputs="text", outputs="image").launch()