Imagen - Pytorch

Implementation of Imagen, Google's Text-to-Image Neural Network that beats DALL-E2, in Pytorch. It is the new SOTA for text-to-image synthesis. Architecturally, it is actually much simpler than DALL-E2. It consists of a cascading DDPM conditioned on text embeddings from a large pre-trained T5 model (attention network). It also contains dynamic clipping for improved classifier-free guidance, noise level conditioning, and a memory-efficient unit design. It appears neither CLIP nor prior network is needed after all. And so research continues. For simpler training, you can directly supply text strings instead of precomputing text encodings. (Although for scaling purposes, you will definitely want to precompute the textual embeddings + mask)