SSD-1B is the diffusion-based text-to-image model and it's 50% smaller and 60% faster than SDXL 1.0.

ssd-1b

The width in pixels of the generated image.

The height in pixels of the generated image.

The prompt to guide what to not include in image generation.

A higher guidance scale makes the model focus more on matching text to images, but this can reduce image quality.

The number of denoising steps. More denoising steps usually lead to a higher quality image at the expense of slower inference.