这是一个英雄联盟游戏中佐伊的角色lora。

在这一次的lora训练中，我同时训练了4个概念，因此对于不同的形态的佐伊，会有不同的触发词。

总共有四个形态的佐伊可以选择：

• 原画佐伊：因为tag标记失误，需要用触发词“zoe”和“multicolored hair”来共同触发。

• 泳池派对佐伊：需要用“zoe”和“waterpool”作为触发词来触发。

• 纯白物语炫彩佐伊：需要用“zoe”和“whitezoe”来触发。

• 星之守护者佐伊：需要用“zoe”和“starguard”来共同触发。

实际使用中权重我推荐0.6-0.8，但是具体视使用者而定。尽管数据集不是很好，但是因为多概念的存在，仍然有很大的发挥空间。希望可以让您喜欢，也希望能一起多多创作关于佐伊的好作品。

因为数据集不够丰富，所以实际效果可能没有想象的那么好。但是由于Zoe是我的挚爱，我会再次训练她直到有一个更好的效果。

This is a character Lora from the game League of Legends, featuring Zoe.

In this Lora training, I trained four concepts simultaneously, so there are different trigger words for different forms of Zoe.

There are four forms of Zoe to choose from:

• Original Zoe: Due to tag marking errors, use the trigger words "zoe" and "multicolored hair" together to activate.

• Pool Party Zoe: Use "zoe" and "waterpool" as trigger words to activate.

• Pure White Story Dazzling Zoe: Use "zoe" and "whitezoe" to activate.

• Star Guardian Zoe: Use "zoe" and "starguard" together to activate.

For practical use, I recommend a weight of 0.6-0.8, but it depends on the user. Although the dataset is not great, the existence of multiple concepts still leaves room for creativity. I hope you enjoy it and look forward to creating more good works about Zoe together.

Since the dataset is not rich enough, the actual effect may not be as good as imagined. However, as Zoe is my favorite, I will train her again until there is a better result.