A large-scale text-to-image generation model based on latent diffusion, developed by the Kuaishou Kolors team
