#80 · Primary category: Image Generation

img2img-turbo

computer-vision deep-learning generative-adversarial-network generative-art stable-diffusion

One-step image-to-image with Stable Diffusion turbo: sketch2image, day2night, and more

Project last updated:08/01/25

GitHub Stars

2.5K

Forks

291

Contributors

8

License

MIT

Why we included this project

Turning a sketch or edge map into a finished image usually means running a long diffusion loop. This project adapts a single-step model like SD-Turbo to do it in one pass, so the same input gets rewritten rather than regenerated from scratch. The included recipes cover both paired tasks, such as sketch to photo, and unpaired edits like day-to-night, and the speed makes it practical for interactive tools or batch work where you want near-real-time results. There's also documentation for training on your own data and local Gradio demos to test the workflows first.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category