Learning Continuous 3D Words for Text-to-Image Generation | Read Paper on Bytez