The Lora model is specifically trained to render a Sexy Sheer Lace Inserts Cheongsam, with satin textile and See-Through Inserts.

A native 1024x1024 dataset was selected to train a performing Lora for Illustrious, with reliable layout for anime, manga and cartoon images.

So first, if you like this Lora, please review it, use it and share many pictures.

Best of them will be reviewed with tips.

For a great output please read the following instructions.

Triggering prompt keyword:

EmbrShCheong

Lora weight:

A value very close to 1.0 is highly recommended. Higher values can trigger artifacts.

So a good activation keyword for lora is:

<lora:EmbrShCheong_IL:1>

Sampling steps:

About 20-40 sampling steps.

Additional istructions

-          The triggering keyword should be followed by another token, that define the colour of satin parts of dress:

[colour chosen] EmbrShCheong,

 So the two token for rendering the Cheongsam are :

EmbrShCheong, [colour chosen] EmbrShCheong,

Example for yellow Cheongsam:

 EmbrShCheong, yellow EmbrShCheong,

Unlike the satin parts, lace inserts are always black.

For entire body images an inpainting step focused on the bodysuit could be needed to improve the output details.

Image dimensions

Better result are obtained with a base image of about 1200 px dimension on the long image side with an image width/length ratio of  1.5.

USE DEMO IMAGES PROMPTS AS REFERENCE, ALSO TO TRIGGER ADDITIONAL SHAPES OF THIS LORA!!!

Final statements:

1. You are solely responsible for any creative work using this model.

2. You cannot use this model to intentionally produce or share illegal images or content.

3. Furthermore can't use this model for malicious, hurtful, defamatory, fraudulent or political purposes.