This LoRA works well for my gijinka project, but I have a small question. I’m a bit disappointed with the color flexibility, as the colors seem somewhat locked. I’m wondering if the dataset was trained only on the normal form and not on the shiny color palette?
