本LoRA已有一更好版本 https://civarchive.com/models/3004112/sinomale-ultraor-krea2?modelVersionId=3408191 但并非在同一训练集完成 因此不作为另一版本发布 恳请各位使用新版本权重

A better version of this LoRA is now available: https://civarchive.com/models/3004112/sinomale-ultraor-krea2?modelVersionId=3408191. Since it was not trained on the same dataset, it is not being released as an additional version of this specific entry. I kindly ask that you use the weights from the new version instead.

It is designed to generate natural-looking East Asian men aged 18–25, with a clean, mildly handsome appearance rather than an overly polished influencer-style face. The training set focuses mainly on athletic characters, university students, delivery riders, firefighters, police officers, and other occupational looks. It includes both close-up portraits and full-body images.
Trigger word：asianathman
模型主要用于生成 18–25 岁、自然耐看的年轻东亚男性。人物以清爽、小帅但不过度精致为主，尽量避免千篇一律的网红脸。训练集偏运动风，包含体育生、大学生、外卖骑手、消防员等形象，同时覆盖大头人像和全身照。
使用 asianathman 以激活