Check out my Twitter where I post daily and you can check my commission status there.
I have a Patreon where I post loads of NSFW work every month and on my feed!
V1 doesn't do too well with her original outfits, but should give some good results.
V2 I added additional triggers, the shirt logo gives the Incredibles logo and red bodysuit gives the general outfit. At the time of making the Lora I didn't think about black underpants or a yellow belt, so its hit-or-miss if those trigger. Give it a shot!
Description
Added the incredibles logo as well as some suit data
FAQ
Comments (18)
+1 uncanny valley
imperfect resemblance :^ )
Did you really just axe my picture?
I loved your image! I'm not sure what happened :( Feel free to repost it!
I also updated a new model, and civitai creates a new gallery for each version. If it got taken down, then someone may have reported it, especially if it wasn't flagged NSFW. I thought it was really funny and creative :)
@NotEnoughVRAM Fair. Dont' know why they would have, but will do.
im goin to steel your image too <3
@ELZEN6 just be sure to credit where you see fit! :3
I have one image posted, it got flagged and came back after 2 weeks, so it happens
So the image was removed for TOS. Not going to repost it. The people who helped me were nice, but can't put it back up despite its positive reception.
@Bombalurina not even when flagged as mature content?
@Bombalurina That sucks, I am curious to see the result. Maybe post it on someplace like baraag.net or maybe pixiv were they're not so stingy on more questionable artwork.
@Bombalurina looks like someone axed one of my pictures too. And they were SFW! How did you manage to get in contact with someone to get some help?
I sort of skipped version 1 and tried this(version2) earlier today just to see how it fares with the more common 2.5D semi-realistic animated/anime style checkpoints which there are a ton of in this site. Overall after testing around 10 of them the results were pretty good and quite consistent. It's a very well built LORA overall.
No offense to the creator though, I did try it with PsuedoRP3D (their own checkpoint) first and comparing it with the other 2.5D models I found the 2.5D models more clear and less flawed while still maintaining the same visual style.
Out of curiosity, which checkpoint was used to train this LORA?
That's completely valid! I've actually sacrificed some of the "perfection" aspects to allow for a wider range of concepts. You'll be able to see some of the variety I put into the images I post daily on my civitai. I'd love to see what your generations look like, if you would like to share!
Additionally, I trained on a ckpt merge of CyberRealistic v1.3, DisneyPixarCartoon, and Kotosmix. You may find that Clip Skip 2 actually yields better results. I find that it helps sometimes given that Kotosmix is trained to work better with it.
@NotEnoughVRAM I ran Kotosmix as one of the ten checkpoints during my test run, it turned out great. And by default I have clip skip 2 on basically everything as well. Suffice to say, great LORA regardless.
Worth noting that you should not keep the LoRA weight more than 0.5. For me, I got very splotchy pictures.
It's usually how it is with the LoRAs I produce
hey, whenever you have time again, could you please create a fortnite nidek for headhunter pls
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.