Download
1 variant available
fp16 SafeTensor
yt_haigure zimage_260813.safetensors
Half precision, best balance • 44 MB
Verified: 15 days ago
850 1 2 3 4 5 6 7 8 90 1 2 3 4 5 6 7 8 9
20 1 2 3 4 5 6 7 8 9
60 1 2 3 4 5 6 7 8 9
(4)
Aug 13, 2026
License:
Apache 2.0About this LoRA
This LoRA allows you to generate the "Haigure Pose" (or Komaneci Pose)
It was trained on a highly diverse dataset (men, women, elderly, costumes, etc.), so it has ZERO character corruption and can be applied to almost anyone—even mascots, wedding dresses, or superheroes!
This model works best with a few helper prompts to get the perfect squatting form and hand position.
* Trigger Word (トリガーワード): yt_haigure
* Recommended Weight (推奨強度): 0.8 - 1.0
* Prompt Helper (推奨プロンプト):
yt_haigure, full body shot, squatting, wide stance, bent knees, wide stance, hands on lower abdomen,
> PRO TIP:
> - If the pose doesn't trigger fully, add full body shot and squatting to force the framing!
> - If you get unwanted text on clothing, put text, logo, writing in the Negative Prompt.
- ✅ Works on Men, Women, Elderly, and even Mascots! (男性、女性、お年寄り、着ぐるみにも対応)
日本語による補足:loraの自由度を優先選したため、トリガーワードによるポーズの定着が弱いです。そのため、プロンプトによる補強が必要になっています。
下記タグを組み合わせて使用してみてください。
衣装により姿勢が固い場合、出にくい場合もあるのでその他タグなどでも工夫してみてください。



