Lora trained in recreate the Redlight style.

Illustrious v1

I just cannot live without this style, it is included in the illustrious base style, but is really sketchy. Maybe this one is a little burned, try weight values under 1. I'll see if is necessary retrain it.

Weight: .65 - .9

Pony v2

This model had several problems with eyes and fingers, so I pruned a lot of scketchy images from the dataset, keeping the most interestings features of this style.

The resource tag was buggy and it didn't tag the model used for the images. It was KNK PonyMerge v2

Weight: .8

Pony v1

After some test, I can create goods Loras based in Doujins

Pruned sketches and other low quality images, better prune of tags (doesn't need "lips" any more), but sometimes still appears "speech bubbles" you can just add them to negative. So that helps to get more defined generations and better eyes.

Weight: .8-1

Pony Beta1.0

This version was an experiment, using all the artist's doujins without cut or prone something. I just removed the repeated images and I keep some pages in colorized and non colorized. I don't know how, but It works so well. Just I didn't prune so well the tags and is necessary add "lips" to the prompt to get a more like style.

I'll try to improvee it, but is enoguht good to be used.

Recommended weight: .8

Prompt: lips, uncensored

By the way, If you ask for the model used in the lasts images, is just a mix of Hassaku XL (Hentai) and MomoiroPony1.2 at 50-50. Actually I released it as KNK PonyMerge