This LoRA was trained by me.

Base model : Anima-base-v1.0
For the dataset: I used images generated from the model "neta lumina".
For usage: my settings are "Anima aesthetic V1.1 int8 and Turbo lora (strength=1), step=6~12, cfg=1, dpmpp_2m_sde_gpu, beta (α=0.5, β=0.7)".
For strength: 0.9~1.2. Excessive strength may cause the error of "eye loss".