VRChat Chiffon

Trained on animagineXLV3_v30, the generation of Chiffon images of 3D models sold on the booth.

Recommended weights: I recommend 1 or 0.9. At 1, there might be minor flaws, but the difference is negligible until 0.8.

The dataset is trained at a resolution of 1152*1152, comprising 53 images.

Trigger words: For triggering the original clothing, use "vrchat chiffon clothe"; when stronger guidance is needed, use "vrchat chiffon." (Note: There is a typo in the word "vrchat" during labeling.)

The tagging method involves removing character features.

=========================================================

基于animagineXLV3_v30训练，生成Booth上售卖的3D模型Chiffon图像。

推荐权重：推荐1或0.9,1可能会有小瑕疵，直到0.8区别都不大。

数据集以1152*1152分辨率进行训练，有53张。

触发词：触发原始衣服用 vrcaht chiffon clothe，当需要强化引导时使用 vrcaht chiffon。(标注时，单词写错了)

打标方式是删除人物形象特征。