Usage (0.5)
Set the Lora strength anywhere between 0.8 and 1.8 on average. I find it worked best at 1.6 personally.
Use the DKT prompt at the beginning of your prompts before everything else, or alternatively after the quality tags.
If you are using a model or other lora's that heavily weight character styles/features you can force this model to adapt them by adjusting the tag's weight. Specific adjustments will vary.
Example for a weight of 1.2:(DKT:1.2),good anatomy,jagged lines,flat colors,This is based off of and is intended to also generate NSFW art. Be Responsible with what you post.
Notes:
Works wonderfully with Wai-Illustrious-SDXL
As of 0.5 it heavily favors revealing clothing and other lewd acts if your prompt is not descriptive enough, you can easily get around this with negative tags as well as detailing your prompt more.
I'm fairly certain the lora can create much less flawed images than my own examples, i'm just not that experienced with generation and all the additional refining during generation.
To do:
Reduce bias towards revealing clothes and extreme content when not prompted to
Either reinforce or create an additional lora for DKT eye style.
Fine tuning
Description
Looks like we don't have an active mirror for this file right now.
CivArchive is a community-maintained index — we catalog mirrors that volunteers upload to HuggingFace, torrents, and other public hosts. Looks like no one has uploaded a copy of this file yet.
Some files do get recovered over time through contributions. If you're looking for this one, feel free to ask in Discord, or help preserve it if you have a copy.


