Cytus II style [animagineXLV3_v30]

Trained on animagineXLV3_v30, it generates images in the art style of the game Cytus.

Version update:

v2.0

Change：Slightly optimized the "mud" lines from version 1.0, and added the functionality to generate game titles.

Recommended weight: 1.

The dataset is trained at a resolution of 1024*1024, with 98 images.

Trigger words: "cytus ii style" for triggering the art style, "cytus title" for triggering the game title (due to limited data, quality may not be high, or may not even appear).

The recommended resolution is 960x1344. This Lora might not be compatible with other Loras.

v1.0

Recommended weight: 1.

The dataset is trained at a resolution of 1152*1152, with 81 images.

There are no trigger words; you can use it directly.

The recommended resolution is 960x1344, which is the same ratio as most datasets. This LORA may not work well with other LORAs.

There may be unexpected images when used on A3.1 (not rigorously tested).

==========================================================

基于animagineXLV3_v30训练，生成游戏Cytus美术风格的图像。

版本更新：

v2.0

变化：稍微优化了1.0版本的“泥浆”线条，加入生成游戏标题的功能。

推荐权重：推荐1。

数据集以1024*1024分辨率进行训练，有98张。

触发词：cytus ii sytle用于触发画风，cytus title用于触发游戏标题(由于数据很少质量不会很高，甚至不会出现)

推荐分辨率是960x1344，这个lora可能无法和其他lora一起使用。

v1.0

推荐权重：推荐1。

数据集以1152*1152分辨率进行训练，有81张。

触发词：没有触发词，直接用就好(

推荐分辨率是960x1344，大多数据集都是这个比例。这个lora可能无法和其他lora一起使用。

也许在A3.1上使用会有意想不到的图像(不严谨测试)