v2.0:
强度：0.7
新版本使用了更多更高清的素材图片并替换了提示词，这让lora能生成更稳定的衣服结构，以及可以接受近距离特写的考验。但在别的方面并没有什么改善，甚至在生成中远距离视角的图片时，衣服比旧版更模糊，因此新版本适合近距离特写镜头或需要放大精修处理的情况。
---
Strength: 0.7
The new version uses more high-quality reference images and updated prompts, which helps the Lora generate more stable clothing structures and pass close-up tests. However, it doesn't improve other aspects — in fact, clothes may appear blurrier than the old version in distant and middle views. So, this version works better for close-up shots or situations requiring upscale and detailed editing.
v1.0:
强度：0.8（根据需要调整）
如果要生成全身图像，推荐使用二次采样或高清放大。
碎碎念：这条连衣裙，本来肩部应该是覆盖着薄纱的，使用“sheer shoulders”或者“see through shoulders”可以还原这一点，但会连带着胸口附近也变成凌乱的半透明状，挺难看的，所以我没有使用这两个提示词。
---
Strength: 0.8 (adjust as needed)
For full-body images, secondary sampling or high-resolution upscaling is recommended.
A little side note: Regarding the dress, the shoulders should be lightly veiled. While using the Prompts “sheer shoulders” or ”see through shoulders” can recreate this detail, they often extend the effect to the chest area, creating a messy, semi-transparent appearance that is less aesthetically appealing. So I removed these Prompts.