Model trained on Anima base-v1.0 to imitate the style of Zankuro=====Support the artist: <a target="_blank" rel="ugc" href="https://www.pixiv.net/en/users/7964">https://www.pixiv.net/en/users/7964</a>=====Third time's a charm, right? I think I narrowed down the issue with my dataset and training, and hopefully this should no longer be something that would be considered to breach ToS. May want to put "white outline" in the negatives sometimes, as I've noticed this version tends to add those.NOTE: All the images I generate for the showcase are done with base Anima without the use of any quality tags, Adetailer, or Inpainting, only Hires. fix to do very basic upscaling. Anima is a base checkpoint and so you can easily get nicer looking results using finetunes or more precise generating methods, but the results speak for themselves on the base.

Zankuro Style (Anima) - Retrain

Tighter, more selective dataset. Cleaned up images/aspects that could well have been the reason my first two attempts were considered to breach ToS.

Files with SHA256: e1065adca495e71ed43eed6ea3c44ae0b1f2dea09344e31b64a488c4bd302880