1.5 trained with around 350 images, mix of non AI and AI
2.0 trained with NO AI
2.5 trained with 200+ ai images
Needs more testing but it seems 2.5 works better around 0.8
Where as older versions seem to work better higher, around 1.3
Description
Needs testing, I'm not sure how the anima model differs
FAQ
Comments (8)
Bruh I just can't make it work. I've tried a number of models, weights and parameters but either I get a small blur where I want the trapped person to be or the tiny person is straight up outside of clothing. If anyone has found a combination of model/settings that works every time, please hit me up.
I think I need to scrap this version, I'm not sure how anima works exactly, so I'm going to have to rethink how to make it work. Maybe it works better with captions
@toon667 Thanks for the reply, I'm on WAI Illustrious and I just can't get the guy inside of clothes. I can only manage to get a tiny guy in different positions sort of around where I want him to be. I appreciate your efforts and it might just be a me problem. 😅
@mrkatzbalger632 Remix one of the existing images and change the loras/models and set the seed to random and then you should be able to change words to get different poses/images
you should update this to Anima 👀
I released a model but it didn't do the clothing entrapment properly so I have deleted it.
@toon667 Is that so? That's a shame. I hope you don't give up though, this lora of yours is way too good not to exist for Anima. And with the way anima works... you might even be able to improve it.
Best of luck! 💪
@oyaoyaoya58326874 I can re-upload the anima trained version, it's harder for me to figure out how it works when I'm only generating online.