English translation is being prepared.
View original postEnglish translation is being prepared.
View original
In Oct last year, Representation Autoencoders provided an elegant solution to unified tokenization for understanding and generation. Today we make them a bit more simple. a bit more general. Result: >10x faster convergence, better reconstruction, better generation. And yes we test them on T2I and world models :) Introducing RAEv2
Body
The English body is being prepared. You can open the original source.