Trained on images by a user on pixiv called Matcha_. I used Anything v3 as a base and trained it with dreambooth. To get the style add (in the artstyle of Matchapix) to your prompts. Follow standard danbooru tags when making prompts