This is a lora made from NIji sweet spot v4 + 748cm style output.
I'm still learning how to create a proper dataset for trainning, so this will be experimental version.

BEST use with other than ANIma base01