This is a tutorial notebook on using Stable Diffusion img2img to generate images, and then upload those images to your Roboflow project. You can run the img2img.ipynb anywhere you have a GPU.
Stability AI GitHub here
Hugging Face Stability v2 model here