AI Image Generator & Stylizer

A powerful Tkinter-based GUI application that harnesses Stable Diffusion models for image generation and stylization. Supports image‑to‑image (img2img) and text‑to‑image (text2img) workflows, integrated upscaling, and VRAM‑saving options—all in a user‑friendly interface.

---

✨ Features

• Two Generation Modes

  • Image to Image: Transform any input image using a text prompt and strength control.
  • Text to Image: Create entirely new images from scratch using a prompt.

• Flexible Model Management

  • Load local Stable Diffusion models (folders containing model_index.json).
  • Download models directly from Hugging Face Hub (e.g., Lykon/dreamshaper-8, stabilityai/stable-diffusion-2-1).
  • Saved model paths persist across restarts.

• Fine‑Grained Controls

  • Prompt with separate style and content fields (content overrides style in text2img mode).
  • Negative prompts to exclude unwanted elements.
  • Adjustable inference steps, guidance scale, and resolution.
  • Seed control for reproducible generations.

• Advanced Upscaling & Enhancement

  • Choose from: None (simple resize), Real‑ESRGAN (fast, realistic), or Stable Diffusion img2img (artistic, slower).
  • Final resolution presets: 1080p, 1440p, 4K, or custom dimensions.
  • Adjustable sharpness for final output.

• VRAM Saver Mode

  • Enable CPU offloading to run models on limited VRAM (slower but functional).

• Real‑Time Preview

  • See intermediate results during generation in the Preview tab.

• Memory Optimised

  • Automatic GPU/DirectML cache clearing after each generation.
  • Attention and VAE slicing to reduce memory footprint.

---

🖥️ System Requirements

• Operating System: Windows, Linux, macOS (tested primarily on Windows).
• Python: 3.8 or higher.
• RAM: 8 GB+ recommended.
• GPU (optional):
  • NVIDIA with CUDA (supports PyTorch CUDA)
  • AMD with DirectML (Windows only, see setup.bat)
  • CPU fallback (slower but works).

---

🚀 Installation

Windows (Quick Setup)

Double‑click setup.bat. This script will:

1. Create a Python virtual environment (venv).
2. Install PyTorch with DirectML support (for AMD GPUs) or CPU fallback.
3. Install all required libraries (diffusers, transformers, Real‑ESRGAN, etc.).

After completion, activate the environment and run the app:

venv\Scripts\activate
python img2img_ui.py