类型
LORA
发布时间
2024-09-19
基础模型
SDXL 1.0
触发词
callahan
复制
版本介绍
V1 uses autotagged images, which will be improved in later versions
详情
推荐
V1
79
1
31
Experiment to explore ability of LoRA to capture the stark, off-symmetry Modernist compositions of Harry Callahan's photographs.
LoRA Weights
The sweet spot appears to be between 0.6-0.8 depending on Sampler, where the image becomes starker / off-symmetrical which gets distorted at higher weights.
查看译文
暂无数据
