Description
Riaku Anima is a model specifically designed to generate high-quality images without the need for complex prompts.
Versions
Version 0.4 Lunaris (Latest) - New captionless training method. Acts as a pure visual filter, preserving base model anatomy and geometry while delivering a rich, atmospheric art style.
Version 0.3 Sophy - This version focuses on improving stability and maintaining the art style.
Version 0.1 Scarlet - An experimental training method with a modified dataset.
Version 0.21 Lily - Anatomy fixes (experimental).
Version 0.2 Lily - Built on the new anima-aesthetic-v1.1 base. Training methods have been improved.
Version 0.1A - Built on the new Anima-Aesthetic base. The image dataset has been significantly improved.
Test Fix - Based on Anima Base 1.0. This early update was intended to help resolve broken anatomy issues.
Test Version - Based on Anima Base 1.0. An initial experimental build. It may be unstable, but it was designed to push for a higher level of detail.
Description
New captionless training method. Acts as a pure visual filter, preserving base model anatomy and geometry while delivering a rich, atmospheric art style.