POV騎乗位と断面図二枚（膣内断面図と性行為中女性器アップ） 要素の組み合わせLoRA 学習量が足りない面があり、現状だとあまり描写成功率が良くない また要素盛りすぎてる面がある。 ■推奨適用強度 1.0- 下限が0.8。断面図二枚描写に特に1.0が必要と認識している。 　　二断面（cross-section）の組み合わせがうまく出る確率は４割程度かと思っている。 溶けやすいためAdetailer推奨 holding handsやnipple tweak は元モデル時点で難しいため、 どうしても溶けやすい面がある。 ■元々目指してた構図、要素は以下 ①cross-section二枚を画面左上と右上に挿入 ②holding hands &amp; nipple tweak （片手動作＆片手動作の学習） ③clothing cutout &amp; nipples （乳首だけ切り抜き服） ②は学習させてはいるものの描写品質があまり良くないため、 好みによって別タグ使う形でいいかと思う。 POV riding position and two cross-sectional views (vaginal cross-section and close-up of female genitalia during intercourse) LoRA element combinationThere are aspects where the training amount is insufficient, and currently the rendering success rate is not very good.Also, there are aspects where too many elements are included. ■Recommended application strength1.0 - Lower limit is 0.8. We recognize that 1.0 is especially necessary for rendering two cross-sectional views.We estimate the probability of successfully rendering the combination of two cross-sections to be around 40%.Adetailer recommended due to its tendency to meltHolding hands and nipple tweaks are difficult in the original model,and therefore tend to melt easily. ■The original composition and elements I was aiming for are as follows:① Insert two cross-section images in the upper left and upper right corners of the screen.② Holding hands &amp; nipple tweak (learning one-handed movements).③ Clothing cutout &amp; nipples (clothing with only nipples cut out).Although I've trained the model for ②, the rendering quality isn't very good, so I think it's fine to use a different tag depending on your preference.

Files with SHA256: 002b507da23955745e19f69aec06f3ef99dd1ba6abe2c261ee39a54868fe77fe