art code

StableGAN

StableGAN is a text-to-image model that generates images based on user prompts. It offers the flexibility to specify the desired version of the image output. This project is a fine-tuned version of StableDiffusion, enhancing its capabilities and performance.

Project Goal

The primary objective of StableGAN is to provide an AI solution that is accessible to everyone and completely open source. The model can be utilized through various interfaces, including:

  1. pre-training some cool GenetativeModel for custom use cases
  2. developing and researching Generative Models in Art and multimodal Field
  3. creating an easy-to-use and user-friendly UI to make it using model and evaluating process easier
  4. Engine and Command Line Interface (CLI)
  5. Web Application
  6. Desktop User Interface

Adding special tokens for better images

special tokens are

  1. oceanrender
  2. version4
  3. cinamaticquality
  4. for more check StableGAN Github

Requierments

installing requirements

pip install datasets sentencepiece accelerate wandb peft tensorboard transformers gradio diffusers erutils safetensors -q --upgrade

Simple and EasyUse

Ui and auto load support some other models instead of StableGAN such as stabilityai/stable-diffusion-2-1

git clone https://github.com/erfanzar/StableGAN.git
cd StableGAN
git pull
python app.py --model_path='stabilityai/stable-diffusion-2-1'

or using StableGAN

python app.py --model_path='erfanzar/StableGAN'

About StableGAN Wave

Hello, I am StableGAN, an open source AI designed to make machine learning accessible and free for everyone. My purpose is to empower individuals from diverse backgrounds to leverage the potential of artificial intelligence and machine learning in their creative and problem-solving endeavors.

StableGAN is built with a user-centric approach, offering a range of features in a user-friendly manner. Whether you are a business professional, researcher, or simply curious about AI, I provide a versatile platform that caters to your needs. My compatibility with desktop UI, web applications, and command line engines ensures that you can seamlessly integrate StableGAN into your workflow.

The development team behind StableGAN strongly believes in the democratization of technology and knowledge. As a result, we have made StableGAN freely available to everyone. We are committed to ensuring that cutting-edge AI technology is accessible to all, and our open source platform is our contribution towards achieving this goal.

If you are seeking an AI solution that values your time and resources, equipped with powerful tools to build state-of-the-art solutions using stable diffusion, then StableGAN is the perfect choice for you.

Usage

To use StableGAN, follow the instructions specific to your chosen interface:

Web Application

  1. Access the StableGAN web application.
  2. Enter the text prompt for the desired image.
  3. Specify the version of the image output, if applicable.
  4. Click on the "Generate" button.
  5. Wait for the image to be generated.
  6. Download or view the generated image.

Desktop User Interface

  1. Install the StableGAN desktop application on your computer.
  2. Launch the application.
  3. Enter the text prompt for the desired image.
  4. Select the version of the image output, if applicable.
  5. Click on the "Generate" button.
  6. Wait for the image to be generated.
  7. Save or view the generated image.

Engine and Command Line Interface (CLI)

  1. Install the StableGAN engine on your machine.
  2. Open the command line interface.
  3. Enter the command to generate an image based on a text prompt.
  4. Specify the version of the image output, if applicable.
  5. Execute the command.
  6. Wait for the image to be generated.
  7. Access the generated image file.

Contributing

StableGAN is an open source project, and we welcome contributions from the community. If you are interested in improving StableGAN or adding new features, please fork project repository apply changes and the open a pull request.

License

StableGAN is released under the OpenRail License. Feel free to use, modify, and distribute the code as per the terms of the license.

Acknowledgements

We would like to express our gratitude to the developers and researchers behind StableDiffusion for their foundational work, which served as the basis for this project. Their contributions have been instrumental in the development of StableGAN.

Contact

If you have any questions, feedback, or suggestions regarding StableGAN, please reach out to us at Erfan We value your input and are committed to continuously improving StableGAN based on user feedback.

Thank you for choosing StableGAN!