paper cut ,background,pastel,剪纸风格,背景,kk

这款Lora是我是训练过的,最喜欢的一款,V1版本修复了人脸的问题,

使用方法:

触发词:parper cut

CFG:7-10

Lora权重:0.6~0.8

Steps:30-50

模型建议:写实类的,或者2.5D风格的,在这款经典的ReV Animated模型下表现已经非常不错,

如果你使用其它大模型效果不佳,就换个底模.

底模推荐:

ReV Animated

https://civitai.com/models/7371

GhostMixReV

https://civitai.com/models/36520

描述词:

Softlight,(warm color:1.15),best quality exquisite details,3d rendering,Octane render,pastel,

(paper_cut:1.15), (white background:1.35), lora:KK_paperCut_v1:0.8

如果想要干净的白色背景,可以添加 white background

或者配合这款立绘Lora使用

https://civitai.com/models/16997

如果想要画面炫丽

可以搭配一些增炫丽的描述或者Lora

例如:(fractal art:1.3),

zentangle, mandala, tangle, entangle,

surrounded by water and fire and wings as behind,

Lora:https://civitai.com/models/17824

VAE:推荐使用NyanMix.vae

剪纸风格,花花草草,江河湖海的自然画面表现会不错

人物也是可以,描述人物的时候,脸如果有崩坏,就打开面部修复,

tag建议添加一些炫丽的效果描述词,可以去翻一翻元素法典,把里面的一些炫丽魔法添加进去,

动物表现也还可以,有问题请留言.

更新状态:如果没有大的问题,就不更新啦,目前测试下,表现还不错.

如果底模表现不佳,请换底模.

欢迎自行测试并分享图片

paper cut ,background,pastel,paper cut style,background,kk This Lora is the one I have trained and like the most, V1 version fixes the face problem. How to use.

Trigger word: parper cut CFG:7-10 Lora weight:0.6~0.8 Steps:30-50

Model recommendation: realistic, or 2.5D style, in this classic ReV Animated model has performed very well. If you don't get good results with other large models, change the base model. Base model recommendations. ReV Animated https://civitai.com/models/7371 GhostMixReV https://civitai.com/models/36520 Descriptors. Softlight, (warm color:1.15), best quality exquisite details, 3d rendering, Octane render, pastel. (paper_cut:1.15), (white background:1.35), For a clean white background, you can add white background Or use it with this Lora print https://civitai.com/models/16997 For a dazzling image you can use it with a description or Lora that adds a bit of flair For example: (fractal art: 1.3). zentangle, mandala, tangle, entangle. surrounded by water and fire and wings as behind. Lora:https://civitai.com/models/17824 Recommended use VAE: NyanMix.vae

Papercut style, flowers and plants, rivers, lakes and seas would be good for natural images Characters are also possible, when describing the characters, if the face is broken, open the face repair. tag suggests adding some dazzling effect descriptors, you can go through the elemental codex and add some dazzling magic to it. The animals are doing ok, please leave a comment if you have any questions. Update status: If there are no major problems, I will not update it, it's not bad at all. If the base is not performing well, please change it. welcome to test and share your pictures

上传的图片不能自动识别参数信息,我把图片参数信息全部添加到正面描述中了,

就这样将就看吧.

(复制全部参数然后点小箭头就可以把参数添加进SD webui中了)

The uploaded image does not automatically recognise the parameters, so I have added them all to the front description.

(Copy all the parameters and click the little arrow to add them to the SD webui)