Illustrious LoRA trained on Akipeko style. Tested on this merge.
Works without trigger.

v2 is trained on more recent images.