no trigger words
v1.1: more faithful to the dataset, lower contrastion when generation than v1.0 , more train epoches
因为我觉得别人训练的于我而言不好用,所以我自己训练了一个
I thought the style trained by others were not suitable for me, so I trained one myself.
need improve
落舟pile style for Anima.safetensors