Try size: 768 x 1024 or 1024 x 1024
Weights: 1.0
Clip skip: 2
Note: I tried my best to retain a little bit of the stiffness in the robot's eyes, sacrificing some qualities that could have made it more human. I borrowed materials from several Douyin AI creators out of respect and admiration. / 我盡力地保留了一點點機器人眼神的生硬感，犧牲了一些可以更像人的特質。我借用了幾位抖音AI創作者的素材，是以致敬、欽佩的心態借用的。(Anima10)
Note: In Asian realistic base models, "asian" can be added; otherwise, the facial features presented by the LoRA will be closer to those of Caucasian women. And reducing the LoRA weight might be better. / 在亞洲寫實模型裏可以添加"asian"，否則外表更接近高加索女性。並且降低LoRA權重，可能會比較好一些。 (ILXL)
Reference Video: