26.05.16
Retrained using a new base model.
This is a Blue Archive–style LoRA trained using Anima Base 1.0.
The trigger word is “@BlueArchStyle” .
If you are using ComfyUI and connecting the LoRA directly through nodes, you can use it without a trigger word.
It works well even with a LoRA weight of 1.0, so feel free to adjust it to your preference.
The style applies especially well to Blue Archive characters, but it also works quite nicely on non–Blue Archive characters.
Anima Base 1.0으로 블루아카이브 스타일 lora 입니다.
트리거 워드는 '@BlueArchStyle' 입니다. ComfyUI로 lora 직접 연동하여 사용하는 경우 트리거 워드 없이 사용 가능합니다.
LoRA weight는 1.0으로 돌려도 무난하게 잘 뽑혔습니다. 취향대로 조절하여 사용하면 될 것 같습니다.
블루아카이브 캐릭터의 경우 특히나 스타일 적용도가 높으며 아닌 경우에도 꽤 잘 먹습니다.
Description
FAQ
Comments (2)
Just wanted to note that using the trigger word (at least on forge neo) makes the image unstable, but using the lora without it lets it work as intended.
Since I’ve only used it in ComfyUI and applied it directly through the LoRA stacker without a trigger word, I’m not too familiar with how it behaves in other environments. Thank you for sharing your experience!