tutu's HiSilk (30D black mesh pantyhose stockings)/图图的嗨丝（30D黑色网格纹连裤袜丝袜）

zhaotutuzhaotutu
255
1.8K
セクシーfemaleclothespantyhose衣装
v1.0
model-image-0
model-image-1
model-image-2
model-image-3
model-image-4
model-image-5
model-image-6

詳細

ファイルをダウンロード (1)

モデル説明

Created with Tutu AI Workflow

Trained with Tutu Trainer. Tagged with Tutu Annotator.

This LoRA is part of my Tutu AI workflow. Tutu Trainer is used for LoRA training, model resource management, parameter presets, and training checks. Tutu Annotator is used for dataset annotation, prompt reverse engineering, English-friendly tags, and cleaner searchable captions.

For a mesh pantyhose LoRA, clean training is important because the model must learn both the black hosiery texture and the mesh pattern. If the captions are messy, the pattern may disappear or spread into clothing and background.

Tutu AI Workflow - Tutu Trainer and Tutu Annotator

Tutu HiSilk: 30D Black Mesh Pantyhose

This is a HiSilk LoRA for 30D black mesh pantyhose / patterned black pantyhose. It improves the stocking texture in the image while trying to keep the rest of the image stable.

The important thing about this model is the mesh pattern. Because of resolution and camera distance, the pattern is easier to see when the image resolution is higher and the camera is closer. If the camera is too far away, the mesh texture may become unclear. This needs some testing with your base model and image size.

  • Type: 30D black mesh pantyhose
  • Color: black
  • Base model: SD 1.5
  • Trigger words: black pantyhose, tutuls
  • Recommended start: (black pantyhose), (tutuls:1.5)
  • Example LoRA weight: around 0.85
  • Recommended CLIP skip: 2
  • CFG Scale: start around 7

What this model is good at

  • Improving black pantyhose texture while keeping a 30D mesh / patterned look.
  • Making mesh texture visible in close enough, high enough resolution images.
  • Reducing pollution to clothing, shoes, body shape, background, and composition.
  • Working especially well with realistic-style base models after parameter adjustment.
  • Separating this mesh pantyhose direction from ordinary black pantyhose and glossy stockings.

Why another black pantyhose LoRA?

There are already many black stocking LoRAs, but many of them mix thickness, style, shine, pattern, and clothing together. Some affect the clothing, some affect shoes, some affect body shape. I wanted a LoRA that focuses on the black pantyhose texture and mesh pattern while affecting other image content as little as possible.

This model was made with real high-definition photo material and a differential extraction workflow. It is part of my HiSilk series, where different thicknesses and textures are separated instead of being mixed into one vague "black stocking" effect.

Recommended settings

  • Make sure the LoRA is loaded correctly.
  • Recommended CLIP skip / CLIP stop layers: 2.
  • Recommended prompt start: (black pantyhose), (tutuls:1.5).
  • LoRA weight can start around 0.85, then adjust according to the base model.
  • Start CFG Scale around 7.
  • Do not add too many negative prompts, or the LoRA effect may be weakened.
  • I did not use LoRA Block Weight in the original test, but you can try it if you need local control.

Prompt keywords

Main trigger group:

black pantyhose, tutuls

Useful prompt phrases:

  • black pantyhose
  • mesh pantyhose
  • patterned pantyhose
  • 30D pantyhose
  • black mesh stockings
  • sheer black pantyhose
  • visible mesh pattern
  • close-up legs
  • high resolution

If the mesh pattern is weak, use a closer camera, higher resolution, and clearer words such as mesh pantyhose and visible mesh pattern. If the LoRA affects clothes or background, lower the weight and describe those parts more clearly.

Model compatibility

In my tests, Moyou-style and Maiju-style base models worked relatively well, and the model seems more friendly to realistic base models. Different base models still need parameter adjustment.

FAQ

Why can I not see the mesh pattern?

The camera may be too far away or the image resolution may be too low. Move the camera closer, increase resolution, and describe the mesh pattern clearly.

Why does the pattern affect clothing or background?

Lower the LoRA weight first. Also describe the clothing and environment so the model does not use hosiery texture to fill unrelated areas.

What should I do if a mosaic or broken texture appears after increasing weight?

Lower this LoRA weight first. If you use many LoRAs together, lower other LoRA weights too. You can also try LoRA Block Weight.

Responsible use

  • You are fully responsible for any creative work made with this model.
  • Do not use this model to create or share illegal, harmful, malicious, defamatory, fraudulent, or political content.
  • If you use this model commercially, please inform me. Thank you.

中文说明

这是图图嗨丝系列里的 30D 黑色网格纹连裤袜 LoRA。它的重点是改善黑色连裤袜质感，同时尽量保留 30D 网格纹 / 花纹效果，并减少对画面其他内容的影响。

这个模型最重要的地方是网格纹。由于分辨率和镜头距离的原因，分辨率高、镜头拉近时，纹路和花纹会更明显；如果镜头拉远，纹路就可能看不清。这个需要结合底模和出图尺寸反复调一下。

  • 类型：30D 黑色网格纹连裤袜
  • 颜色：黑色
  • 基础模型：SD 1.5
  • 触发词：black pantyhose, tutuls
  • 推荐起手：(black pantyhose), (tutuls:1.5)
  • LoRA 权重示例：0.85 左右
  • CLIP 终止层：2
  • CFG Scale：从 7 左右开始

这个模型适合什么

  • 改善黑色连裤袜质感，同时保留 30D 网格纹效果。
  • 在较近镜头和较高分辨率下表现丝袜纹路。
  • 尽量减少对衣服、鞋子、身材、背景和构图的污染。
  • 对写实类底模更友好，但不同底模都需要调参。
  • 把网格纹连裤袜和普通黑丝、油亮黑丝区分开。

为什么还要做一个黑丝 LoRA

因为很多黑丝 LoRA 会把厚度、样式、光泽、花纹、衣服、鞋子都混在一起。有的会污染衣服，有的会影响鞋子，有的会影响身材。我想做的是尽量只影响黑色连裤袜质感和网格纹，不影响其他画面内容。

这个模型使用真人高清实拍素材，并用差异提取流程制作。它属于嗨丝系列的一部分：不同厚度、不同质感、不同花纹应该拆开做，而不是混成一个模糊的“黑丝效果”。

推荐参数

  • 确保正确加载 LoRA。
  • CLIP 终止层数建议设置为 2。
  • 推荐起手：(black pantyhose), (tutuls:1.5)。
  • LoRA 权重可以从 0.85 左右开始，再根据底模调整。
  • CFG Scale 可以从 7 开始。
  • 负面提示词不要堆太多，否则会影响 LoRA 效果。
  • 原始测试中我没有使用 LoRA Block Weight，但如果需要局部控制，可以尝试。

出图关键词

主触发词组：

black pantyhose, tutuls

  • black pantyhose
  • mesh pantyhose
  • patterned pantyhose
  • 30D pantyhose
  • black mesh stockings
  • sheer black pantyhose
  • visible mesh pattern
  • close-up legs
  • high resolution

如果网格纹不明显，请拉近镜头、提高分辨率，并补充 mesh pantyhose、visible mesh pattern。如果影响衣服或背景，先降低权重，并把衣服和环境描述清楚。

模型适配

目前测试里，墨幽系列、麦橘系列效果相对不错，似乎对真人模型更加友好。不同大模型需要根据实际效果调参数。

常见问题

为什么看不清网格纹？

可能镜头太远或者分辨率不够。请拉近镜头、提高分辨率，并明确描述网格纹。

为什么花纹影响到衣服或背景？

先降低 LoRA 权重，同时把衣服、环境描述清楚，避免模型把丝袜纹理扩散到其他区域。

权重调高后出现马赛克或破碎纹理怎么办？

先降低本 LoRA 权重。如果同时使用很多 LoRA，也降低其他 LoRA 权重，或者尝试 LoRA Block Weight。

声明

  • 你需要对使用本模型生成的任何作品承担全部责任。
  • 不要使用本模型制作或传播非法、有害、恶意、诽谤、诈骗或政治用途内容。
  • 如果将本模型用于商业用途或定制合作，请提前联系我，谢谢。

日本語説明

これは Tutu HiSilk シリーズの 30D black mesh pantyhose LoRA です。黒いパンストの質感と、30D のメッシュ / パターンを出すためのモデルです。

  • Trigger: black pantyhose, tutuls
  • Recommended start: (black pantyhose), (tutuls:1.5)
  • LoRA weight: around 0.85, then adjust
  • CLIP skip: 2
  • CFG Scale: start around 7

メッシュ模様は解像度とカメラ距離の影響を受けます。遠景では見えにくいため、近い構図、高い解像度、mesh pantyhose, visible mesh pattern などの語を使うと効果が出やすくなります。

このモデルで生成された画像