Right now making a stack of models to compare them side-by-side and then release ones with most quality. A large amount of work with datasets and generations. Still, all my work is long waited but free!