爻光_LOKR
第一次练lokr,可能会有些奇妙的地方过拟合/不拟合,质量还行。
本模型使用v预测来训练,请使用同样支持v预测的底膜使用
训练使用NoobAI-XL-v-pred-1.0
跑图使用WAI-SHUFFLE-NOOB-2.0(unet-fp8)
跑图工具comfyui
推荐强度0.9(0.8-1)
触发词:yaoguang, 1girl, solo,blue eyes,jewelry, hair ornament
默认服装:yao dress
其他:holding envelope,hongbao(可能有效)
如果要换装,请把yao dress添加到负面提示词中。
This is my first attempt at training a LoKR (Locon), so there might be some peculiarities with overfitting/underfitting, but the overall quality is decent.
The model was trained using v-prediction. Please ensure you use a base model that also supports v-prediction.
Trained on base model: NoobAI-XL-v-pred-1.0
Used for generating examples: WAI-SHUFFLE-NOOB-2.0 (unet-fp8)
Generation tool: ComfyUI
Recommended strength: 0.9 (range 0.8 - 1.0)
Trigger words:yaoguang, 1girl, solo, blue eyes, jewelry, hair ornament
Default outfit:yao dress
Other elements:holding envelope, hongbao (may work)
If you want to change the outfit, please add yao dress to the negative prompt.
Description
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.


