Vector Quantized Diffusion Model for Text-to-Image Synthesis