Mia as a lora trained on a latina girl generated with SD1.5 and refined with SDXL. Wildcards can be used, but the lora needs testing with your SDXL checkpoint before using for long run, it's not stable with ever checkpoint.