This LoRA is trained on a dataset of 9,936 images, but due to the poor quality of the dataset, some characters are not possible.
Here are some samples of captions from this dataset.
kita ikuyo, bocchi the rock!, sensitive
ijichi nijika, yamada ryo, bocchi the rock!, generalgotoh hitori is possible without LoRA on Pony Diffusion V6 XL models. Of course, it is also possible with this LoRA.

