santa hat
deerdeer nosedeer glow
Sign In

翠玉白菜SDXL | Jadeite Cabbage | Chinese treasure

775
4.1k
101
Updated: Sep 20, 2023
style
Verified:
SafeTensor
Type
LoRA
Stats
4,139
Reviews
Published
Sep 20, 2023
Base Model
SDXL 1.0
Hash
AutoV2
4120CD6D7D
default creator card background decoration
XiongSan's Avatar
XiongSan

翠玉白菜SDXL|玉雕风格

简介:

此lora为翠玉白菜的SDXL版本

【翠玉白菜】(The Green Jade Cabbage),取自清代文物,是由翠玉所琢碾出白菜形状的玉器,现藏台北故宫博物院。它的色彩质感既有一般白玉的凝重,又有翡翠绿色的鲜活。此lora生成的角色正是有相似的特点,肌肤嫩如白玉、服装则温润透亮如翡翠一般,鲜活,精致、有流动感的,因此借用了此名。

使用说明:

推荐权重:0.7~0.9

推荐关键词:

【起手式】:jade,jade sculpture (用于激活大模型本身的玉雕元素)

套词:masterpiece,best quality,jade,jade sculpture,1 girl,looking at viewer,Strapless dress,bare shoulders,short hair,dancing, butterfly, butterfly wings, front light,cinematic light

衣服元素:cheongsam,hanfu,strapless dress, bikini

补充:

大家也可以试试生成动物或者建筑之类的,由于训练集中并没有此类素材,所以更多的依靠的是大模型的能力

English:Jadeite Cabbage

introduction:

This lora is a style model of jade material and is good at generating female characters with Chinese style.

The name Jadeite Cabbage, taken from the cultural relics of the Qing Dynasty, is a piece of jade in the shape of cabbage carved by Cuiyu and is now hidden in the Palace Museum in Taipei.

Its color texture is not only the pure of white jade, but also the freshness of emerald green.

The characters generated by this lora have similar characteristics, the skin is as tender as white jade, and the clothing is as warm and bright as green jade, fresh, delicate and mobile, so the name is borrowed.

Instructions for use:

Recommended weight: 0.7~0.9

Recommended keywords:

Start with the following tag: jade,jade sculpture (used to activate the jade carving elements of the large model itself).

Idioms: masterpiece,best quality,jade,jade sculpture,1 girl,looking at viewer,Strapless dress,bare shoulders,short hair,dancing, butterfly, butterfly wings, front light,cinematic light

Clothing elements: cheongsam,hanfu,strapless dress, bikini.

You can also try to generate animals or buildings, because there is no such material in the training set, so we rely more on the ability of large models.