V3.0优化了生成细节,能够生成更复杂的场景
关键词:vollymaker
欢迎大家使用
V3.0 optimizes the generation details and can generate more complex scenes.
Keywords: vollymaker
Welcome to use
__________________________________________________________________________________________________
v2.0增加了暗色图标的训练素材,强化材质细节,能生成更加具有真实感的场景。
暗色模式需添加关键词:black_safeicon, newsafeiconv2
亮色模式需添加关键词:white_safeicon, newsafeiconv2
欢迎大家使用
v2.0 adds training materials for dark icons, strengthens material details, and can generate more realistic scenes.
Dark mode prompt: black_safeicon, newsafeiconv2
Bright mode prompt: white_safeicon, newsafeiconv2
Welcome to use
__________________________________________________________________________________________________
应用SD_xl_1.0_base模型进行训练,建议强度在(0.8-1.2)区间内
可以配合一些增强细节的lora一起使用,能得到更好的效果。
V1.0版本对毛玻璃、塑料材质的prompt比较敏感,对金属类的prompt识别较弱。
Apply the SD_xl_1.0_base model for training, and the recommended intensity is within the range of (0.8-1.2)
It can be used together with some lora that enhance details to get better results.
The V1.0 version is more sensitive to prompts made of frosted glass and plastic materials, and is weak in recognizing prompts made of metals.