







My first attempt at a concept LoRA, The attempt over-night went better than expected.
I've tried in may models, and I've found that it's a LoRA to work enough to steer the concept for the lady to play with at least one toy. - Some models are more stubborn than others.
Focuses on ????????? and view from the front
Using an additional keyword "????? in ?????", "????? in ???" and "spread leg", helps to nudge the LoRA to give higher success. (result tested with and without these additional keywords)
Recommendation
Model - Most models seem to work well (sample provided). It works ok in both realistic and animation
Sample - Restart
Sampling steps - 30 (I find Restart can use a little more than 30)
Upscale - You don't need it.
As with my other work, feedback is welcome, as I learn/get the hang of what is needed for the modesl.
As I'm starting to get the hang of it, I can take some requests. Please feel free to reach out.
Any support will be welcomed https://thelaughterlab.etsy.com/
~~~~~~~~
女性のモデルではなく、コンセプトの初挑戦です。実験が思ったよりうまくいったので、V1をアップロードすることにしました。
モデルによって成功の度合いが違いますが、アニメでも現実的なえでも両方オッケーです。
正面目線で、足を開いたポーズです。
トリガープラス+"????? in ?????", "????? in ???", "spread leg" を使う事で成功の確率がかなり上がります。
基本設定
Modelー大体お好きなやつでいいです
SamplerーRestart
Steps-30位がベストです。
Upscalerー無しでも行けます。
コツを掴んできたので、T-シャツ買ってくれればリクエスト取ります。
1. 轉载模型僅供學習與交流分享,其版權及最終解释權归原作者。
2. 模型原作者如需認領模型,請通過官方渠道联系海藝AI工作人員進行認證。我們致力於保護每一位創作者的權益。 點擊去認領
