This is my first style that I decided to publish. Its main feature is the emphasis on anime-style eyes and enhanced elements of realism.
V1.0
Test version for the Zimage Turbo model. The style works perfectly without a trigger and produces excellent results that combine anime and semi-realism. The trigger enhances the main feature of the style — the characters' large and expressive eyes. The style focuses exclusively on female characters. This version is based on the original dataset. Standard danbooru tags were used as captions, which work perfectly with the style.
Description
LoRA works without a trigger, but the trigger enhances the main element of style — expressive and larger eyes.