4/15 update:

new version for AnimagineXLV3.1, it's stabler than 1.0, and you can even use it as a 2.5D or 3D stylizer. It's also compatible with other pose or character loras.

(Due to the data, this more is not good at drawing bare feet, you might need gacha)

Recommended weight: 0.9 - 1.1,

Trigger: asanagi-figure, photorealistic, photo \(medium\),

Trained specifically for figure works Illustrated by Asanagi, trained on AnimagineXLV3, but also works on v3.1 (though it seems that V3 performs a little bit better). Most data was collected on Twitter. Because of the lack of the data, it might need a little bit gacha to work well.

Recommended weight: 0.9 - 1.1,

lower the weight if faces are twisted, and specific character tags are helpful for making faces look good.

Tigger: asanagi-figure,

Recommended tags (Try to remove them if faces are weird): photorealistic, photo \(medium\),

Support nsfw contents, but you may need to add extra tags if some parts are missing.

专门为朝凪的手办化作品训练的画风，底模AnimagineXLV3，3.1也能用不过貌似3.0效果稍微好一点。数据来源主要是推特返图，因为能用的图很少所以这个模型用的时候可能要稍微抽下卡

建议权重 0.9 - 1.1, 如果人物的脸扭曲了或者很怪就降低些权重，使用特定的角色tag可以比较有效的避免脸部抽风

触发词：asanagi-figure,

建议同时使用（如果脸部很怪的话也可以移除这些tag）： photorealistic, photo \(medium\),

也可以画nsfw内容，不过需要额外的tag来补充丢失的部分