该模型使用1.2k张图片训练，可生成Logo，徽章，图标，与其他lora结合等。

欢迎使用、评论和返图。

This model is trained with 1.2k images and can generate logos, badges, icons, and be combined with other Loras.

Welcome to use, comment, and return pictures.

• 召唤词：logo,

• Summoning words: logo,

用于生成logo很不错，其它可以尝试。

It's great for generating logos, you can try other options.

• 你仅需要输入logo，即可召唤，一个提示词大多是随机抽卡（可能随机带有english text）。

• You only need to enter the logo to summon, and a prompt word is mostly a random draw (possibly with English text randomly included).

你可以使用任何1.5的底膜出图，二次元人物对象用二次元的较佳，其它可用现实风的。

• 二次元较容易，现实风不容易生成人物。

You can use any 1.5 base film for mapping. anime character objects are better to use anime ones, and others can use realistic ones.

• Anime is easier, but realistic wind is not easy to generate characters.

不同底膜出图会有不一样的表现效果，你可以尝试。

Different bottom films will have different performance effects, you can try it out.

其中若你不想要english text,

If you don't want English text,

请在负向提示词加上例如：(english text:2), (artist name:2),

Please add an example to the negative prompt word: (English text: 2), (artist name: 2),

但也不一定不带，你可能需要进行抽卡。

But it doesn't necessarily mean you don't bring it, you may need to draw cards.

大多情况下，也行带text的可能效果更好。

In most cases, using text may also yield better results.

lora合用：两个lora权重之和建议不超过1.6，我不确定是否这好用。

Lora combined: It is recommended that the sum of two Lora weights should not exceed 1.6. I am not sure if this is useful.