It's Skeletor, you bubbling nincompoops!
-LTX2 Version 1-
This one is fairly stable and I rather like it. One thing to note is that I have noticed depending upon the seed number results can vary greatly. Most of the time it will nail the voice and character consistency - but in some generations with the same prompt with a different seed you'll get a raspy bro voice vibe and it looking like a modernized version of him. If this happens, I suggest that you change the seed and try again (unless you like that version, no judgement here).
I've made this one where it can use very simple prompting so you can try to vary the prompt a bit but generally just changing the seed number will do the trick if you find that its happening w/ your generation attempt.
-Z-Image Turbo Version 1-
I created this one on a whim after watching the newest He-Man movie trailer and getting memory flashbacks watching the TV series from the 80s.
This version is made from a dataset using the original He-Man series, with some selective enhancements.. heh.
It does a fairly decent job of prompt adherence, but since I used detailed captioning you'll need to be fairly specific in prompting (i.e. don't expect much with short prompts).
Description
Created on approx 100 HQ images from the original TV series
FAQ
Comments (6)
I thought this model can't render penises properly. Did you train the lora with nsfw images?
The example images in the gallery are txt2img examples, using the base Z-Image Turbo bf16 checkpoint. - so what you see there is what kind of penis quality you can expect from it.
what is with the ***** and shit this site you cant cuss or something but post xxx pics im confused
@jrewingwannabe947 so is this a image to text model or does it do both cause all i get is some shirtless guy
@brianzivkovich02575 it's a text 2 image model. Try some of the samples in the gallery and look at my notes in the model page about seed #s.
@brianzivkovich02575 Are you seeing censors instead of the words? I don't see a single censor on this page. Maybe it's your profile settings?





