LORA of Cantabile in Arknights, Trained base on AOM3 checkpoint
明日方舟晓歌的LORA,基于AOM3模型训练
Someone in the QQ chat group said that there are too few Cantabile fanarts, therefore I trained this LORA.
有群友说小鸟的色图和二创太少,所以我就进行了一个模型的炼
This LORA performs well from 0.7 to 0.9, higher than 1 is not recommended.
该LORA在0.7—0.9的权重下表现良好,不建议高于1
Supports the generation of original costume and New Year costume, keywords are (Origin_skin) and (chinese_dress)
可以进行原皮肤和灯如昼皮肤的生成,关键词分别为(Origin_skin)和(chinese_dress)
Would like to have a reliable LORA training guide BTW:)
顺便求一个靠谱的LORA训练教程:)
有模型的训练需求可以B站私信我 id:不吃青蛙的阿龙
有空就会做
Description
Details
Downloads
500
Platform
CivitAI
Platform Status
Deleted
Created
3/9/2023
Updated
8/9/2025
Deleted
8/7/2025
Trigger Words:
cantabile
origin_skin
chinese_dress
Files
arknightsCantabile_v10.safetensors
Mirrors
CivitAI (1 mirrors)