A LORA for Merida from Brave. This LORA was trained to work with PonyXL based models, so use the standard prompts: score_9, score_8_up, score_7_up, score_6_up, score_5_up, score_4_up