It is often a trade-off between speed and accuracy. While the framework is the same, there are many different ways to carry out this denoising process. The sampler gradually produces cleaner and cleaner images. Read the article “How does Stable Diffusion work?” if you want to understand the whole model.īelow is a sampling process in action. Sampling is just one part of the Stable Diffusion model. The method used in sampling is called the sampler or sampling method. This denoising process is called sampling because Stable Diffusion generates a new sample image in each step. The predicted noise is subtracted from the image. The noise predictor then estimates the noise of the image. To produce an image, Stable Diffusion first generates a completely random image in the latent space.
What is Sampling? The sampler is responsible for carrying out the denoising steps.
