



A comment below asked me to do it, and while I've been struggling to do 2 or more characters image generations, figured I'd give it a try.
... Still can't do it. Maybe it's a "me" problem, but Rachel usually looks a bit off, zack's heterochromia mixes things up, clothes are a mess... Not too happy about the result, but since I'd put the effort, might as well share it, as someone was asking for it.
If anyone wants to try the LoRA and gets succesful generations on them together, please do share and upload them over here! I'd love to finally be able to (somehow) make decent multiple characters image generations and LoRAs.
OLD VERSION:
Anyway, here's Zack from the Angels of Death media.
Lora made of 20 images consisting of anime screen caps, official and fan arts.
This Lora was trained on the Hollostrawberry's google colab training generator. It was trained on the Pony XL model, so it won't probably work well with other XL models.
Works fine with a 0.9 or even 1~ strenght. The bangades can get quite tricky with lower strenght like 0.8 or 0.7, but lowering it can help with using with other LoRAs, so it's your choice.
1. 재게시된 모델의 권리는 원 제작자에게 있습니다.
2. 모델 원작자가 모델을 인증받으려면 공식 채널을 통해 SeaArt.AI 직원에게 문의하세요. 저희는 모든 창작자의 권리를 보호하기 위해 노력합니다. 인증하러 이동
