State-of-the-art diffusion and generative models that turn text prompts into detailed images. Includes lightweight CPU-friendly and photorealistic mdl