强度：1

已知问题：

1.在训练lora时是单独训练了帽子的，但帽子的效果很不理想，特别是和人物lora共同使用或者视角偏低时，这顶小礼帽会被完全改变形状，和衣服很不搭调，所以我没有把帽子的提示词放入右侧栏中：plaid hat,pink hat bow,lace trim,

2.糟糕的素材，往往只能得到糟糕的lora，这个lora仅仅能大体维持住衣服的外观，细节就不存在一致性了，完全看大模型自己发挥了。

---

Strength: 1

Known Issues:

• While the hat was trained separately, its results are often poor. When used with character LoRAs or at lower camera angles, the hat becomes heavily distorted and poorly matches the outfit. So these hat prompts are not included in the side panel: "plaid hat, pink hat bow, lace trim".

• Poor training materials often lead to unreliable results. This LoRA only roughly maintains the outfit's appearance. Details lack consistency, and it's up to the base model to fill in the details.