Stable-Diffusion

Overview

Welcome to the Stable Diffusion PyTorch repository! This project leverages the power of PyTorch to implement the Stable Diffusion model, enabling high-quality image generation from textual descriptions. Built with efficiency and scalability in mind, our implementation supports various customizations and advanced features, making it suitable for both research and practical applications.

Features

• High-quality image generation with Stable Diffusion
• Support for custom prompts and image editing
• Efficient PyTorch implementation optimized for GPU acceleration
• Extensible architecture for further customization and research

Prerequisites

Before you begin, ensure you have the following installed:

• Python 3.8 or newer
• PyTorch 1.8 or newer
• CUDA Toolkit (for GPU acceleration)
• Other dependencies listed in requirements.txt

Installation

To set up the project, follow these steps:

Clone the repository:

git clone https://github.com/Ionio-io/Stable-diffusion-PyTorch.git
cd Stable-diffusion-PyTorch

Install the required Python packages:

pip install -r requirements.txt

Download weights and tokenizer files:

• Download vocab.json and merges.txt from https://huggingface.co/runwayml/stable-diffusion-v1-5/tree/main/tokenizer and save them in the data folder
• Download v1-5-pruned-emaonly.ckpt from https://huggingface.co/runwayml/stable-diffusion-v1-5/tree/main and save it in the data folder

Usage

To generate images with Stable Diffusion, use the following command:

python generate.py --prompt "A text description of the image you want to generate"

Advanced Options
--uncond_prompt: Specify an unconditional prompt for more abstract generations.
--strength: Control the influence of the conditional prompt on the generated image.
--n_inference_steps: Specify the number of diffusion steps.
Refer to pipeline.py for more details and options.

Customization

This implementation allows for easy customization and extension. You can modify the generation parameters, integrate new models, or experiment with different diffusion techniques by editing the corresponding Python scripts.

License

This project is released under the MIT License. See LICENSE for more details.

Acknowledgments

Special thanks to the original creators of the Stable Diffusion model and the PyTorch community for their invaluable resources and support. Additional thanks to the YouTube channel by Umar Jamil. For a detailed explanation and a list of resources, please visit https://www.ionio.ai/blog/harnessing-stable-diffusion-a-revolutionary-guide-for-creators-and-business-innovators