app.py

import streamlit as st
import torch
import os
from medigan import Generators
from torchvision.transforms.functional import to_pil_image

def install_and_patch_model(model_id):
    try:
        # Install model if not exists
        generators = Generators()
        if model_id not in generators.get_model_ids():
            generators.download_and_install_model(model_id)
        
        # Locate model files
        model_path = os.path.join(os.path.expanduser("~"), ".medigan", "models", model_id)
        
        # Patch base_GAN.py
        base_gan_path = os.path.join(model_path, "model19", "base_GAN.py")
        if os.path.exists(base_gan_path):
            with open(base_gan_path, "r") as f:
                content = f.read()
            
            # Replace problematic optimizer lines
            content = content.replace(
                "betas=(self.beta1, self.beta2)",
                "betas=(0.5, 0.999)"  # Force correct beta values
            )
            
            with open(base_gan_path, "w") as f:
                f.write(content)
            
            st.success("Model successfully patched!")
        return True
    except Exception as e:
        st.error(f"Patching failed: {str(e)}")
        return False

MODEL_IDS = [
    "00001_DCGAN_MMG_CALC_ROI",
    "00002_DCGAN_MMG_MASS_ROI",
    "00003_CYCLEGAN_MMG_DENSITY_FULL",
    "00004_PIX2PIX_MMG_MASSES_W_MASKS",
    "00019_PGGAN_CHEST_XRAY"
]

def main():
    st.set_page_config(page_title="MEDIGAN Generator", layout="wide")
    st.title("🧠 Medical Image Generator")
    
    with st.sidebar:
        st.header("⚙️ Settings")
        model_id = st.selectbox("Select Model", MODEL_IDS)
        num_images = st.slider("Number of Images", 1, 8, 4)
        
        if st.button("✨ Generate Images"):
            if "00019" in model_id and not install_and_patch_model(model_id):
                return
                
            with st.spinner("Generating images..."):
                generate_images(num_images, model_id)

def generate_images(num_images, model_id):
    try:
        generators = Generators()
        images = []
        
        for i in range(num_images):
            sample = generators.generate(
                model_id=model_id,
                num_samples=1,
                install_dependencies=False
            )
            img = to_pil_image(sample[0]).convert("RGB")
            images.append(img)
        
        cols = st.columns(4)
        for idx, img in enumerate(images):
            with cols[idx % 4]:
                st.image(img, caption=f"Image {idx+1}", use_column_width=True)
                st.markdown("---")
                
    except Exception as e:
        st.error(f"Generation failed: {str(e)}")

if __name__ == "__main__":
    main()