用戶LoRA

XL
姿勢

見下ろし式半透明だいしゅきホールド / POV high-angle invisible leg lock

#第一人稱視角

#鎖腿

見下ろし式半透明だいしゅきホールド / POV high-angle invisible leg lock - AI Model cover image
上傳時間
2025年8月7日 凌晨2:06
使用數
493
評論
暫無評論
觸發詞

dsk-invi, 1girl, invisible boy, leg lock, nsfw

許可權
  • 允許圖片生成與分享
  • 允許使用者下載此模型
  • 允許生成圖片用於商業用途

描述

Recommended strength (v2.0): Around 1.3–1.5. It feels unstable at 1.0.Features of Old Version v1.0Major Issues in v1.0During training, the semi-transparent male's face overlapped the girl's face, making the girl's face prone to deformation.This issue becomes especially noticeable when using external LoRAs.While ADetailer would usually be the go-to solution, applying it with person can destroy the core concept (semi-transparent male).Applying it to face sometimes causes only the man's head to become fully opaque.Due to the above, compatibility with ADetailer is very poor.Character swaps are not recommended.Recommended strength (v1.0): 0.8–1.0

評論