app.py

import gradio as gr
import torch
from diffusers import StableDiffusion3Pipeline, ControlNetModel, UniPCMultistepScheduler
from huggingface_hub import login
import os

# Log in to Hugging Face with token from environment variables
token = os.getenv("HF_TOKEN")
login(token=token)

# Model IDs for the base Stable Diffusion model and ControlNet variant
model_id = "stabilityai/stable-diffusion-3.5-large-turbo"
controlnet_id = "lllyasviel/control_v11p_sd15_inpaint"  # Adjust based on ControlNet needs

# Load ControlNet and Stable Diffusion models
controlnet = ControlNetModel.from_pretrained(controlnet_id, torch_dtype=torch.bfloat16)
pipe = StableDiffusion3Pipeline.from_pretrained(model_id, controlnet=controlnet, torch_dtype=torch.bfloat16)
pipe.scheduler = UniPCMultistepScheduler.from_config(pipe.scheduler.config)
pipe = pipe.to("cuda") if torch.cuda.is_available() else pipe

# Gradio interface function
def generate_image(prompt, reference_image):
    # Prepare the reference image
    reference_image = reference_image.convert("RGB").resize((512, 512))

    # Generate the image using the pipeline with ControlNet
    generated_image = pipe(
        prompt=prompt,
        image=reference_image,
        controlnet_conditioning_scale=1.0,
        guidance_scale=7.5,
        num_inference_steps=50
    ).images[0]
    return generated_image

# Set up Gradio interface
interface = gr.Interface(
    fn=generate_image,
    inputs=[
        gr.Textbox(label="Prompt"),
        gr.Image(type="pil", label="Reference Image (Style)")
    ],
    outputs="image",
    title="Image Generation with Stable Diffusion 3.5 and ControlNet",
    description="Generates an image based on a text prompt and style reference image using Stable Diffusion 3.5 and ControlNet."
)

# Launch the Gradio interface
interface.launch()