erykbanatt/stable-diffusion-systems repository overview

⁠Stable Diffusion for Comic Generation

This project aims to implement Stable Diffusion for image generation, in order to assist artists in iterating upon concepts more quickly.

This project was built leveraging the Huggingface Diffusers library, which provides a flexible and easy-to-use implementation of stable diffusion. This allowed us to quickly and efficiently implement the algorithm, and focus on developing features and improvements specifically tailored to the needs of artists.

⁠Using the Code

⁠Conditional Image Generation

To use this docker image for text-to-image generation, use the following command

docker run --gpus all --rm stable-diffusion python diffusion.py --prompt "A superhero flying over the city"

⁠Image to Image Generation

To use this docker image for image-to-image generation, use the following command

docker run --gpus all --rm stable-diffusion python diffusion.py --prompt "A superhero flying over the city" --img path/to/input_image.png --saveas path/to/output_image.png