Trained based on animagineXLV3_v30, generating images in the style of artists.

Version update:

v3.0:

New additions: Alphonse (whitedatura) and Muk (monsieur) art styles.Musouzuki art style optimization.

Recommended weight: Recommend 0.9 or 1.

Changes: All trigger words have been changed to Danbooru tags:

as109

alphonse\(white datura\)

muk\(monsieur\)

musouzuki

v2.0:

New: Added the style of artist musouduki.

Recommended weight: Recommend 0.9.

The dataset was trained with a resolution of 1024*1024, consisting of 865 images (as109: 549, musouduki: 316).

Trigger words: musoudukistyle, as109style

There's a certain probability of limb fragmentation, currently, elves discovered will carry the style of as109.

Looking forward to seeing everyone's feedback (lol).

v1.0:

Recommended weight: Recommend 1.

The dataset was trained with a resolution of 1152*1152, consisting of 549 images.

Trigger words: No trigger words, just use directly (perhaps it can't be written out?

If the image is too gray, it's recommended to add [saturation] to the prompt.

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

基于animagineXLV3_v30训练，生成画师的画风图像。

版本更新：

v3.0:

新增：alphonse(whitedatura)，muk(monsieur)画风。musouzuki画风优化。

推荐权重：推荐0.9或1。

改动：触发词全部改用danbooru标签：

as109

alphonse\(white datura\)

muk\(monsieur\)

musouzuki

v2.0:

新增：musouduki画师画风。

推荐权重：推荐0.9。

数据集以1024*1024分辨率进行训练，有865张(as109:549,musouduki:316),。

触发词：musoudukistyle,as109style

一定概率的肢体破碎，目前发现的elf会附带as109的画风。

非常期待大家的返图（笑

v1.0:

推荐权重：推荐1。

数据集以1152*1152分辨率进行训练，有549张。

触发词：没有触发词，直接用就好(有没有可能是因为不能写出来？

如果画面偏灰建议在提示词加上：[saturation]。