Post #300 2.5K Jun 15, 2021, 20:07 UTC This is the architecture. Content encoder encodes text, Style encoder extracts style and Generator generates stylized text conditioned on a style vector.