Share your images here and leave a like if you enjoyed it! Get exclusive NSFW models at DeviantArt or Booth.pm.
Tag: princess carry
Recommended quality tags: score_9, score_8_up, score_7_up, score_6_up. Optional: source_anime
Recommended negative tags: score_4, score_5. Optional: 3d, lips
Recommended checkpoint: AutismMix_confetti
Description
FAQ
Comments (26)
Do you recommend using Euler a for your LoRAs in general? I always use DPM++ 2M Karras and I just noticed that it works better for this Lora to use Euler a.
It depends. This is my limited opinion on the subject, but from my experience that one works best in other base models, and especially in semi-realistic and realistic ones. For anime I tend to use Euler a. But I wouldn't be able to recommend you one or the other.
Greatly done and wholesome AF, loving it!
thanks!
so, I'm guessing you gen with the lora to get the pose and use inpaint to make the faces after?
no inpainting. these are all regular gens with hires fix
@psoft So it never mixes the boy's and girl's faces together? you never get 2boys or 2girls, when you actually wanted 1 of each?
If i may make a additional concept idea.
How about making a concept of a person riding on a woman's back? Like on all fours and piggyback rides?
You doing great training and uploading these, btw.
Nice one!
Wow! Senior....... You made me feel like a man should be ;-)
Can you reverse the genders with this?
it's very unlikely with this lora, since the dataset was limited to male+female or female+female
@psoft why
Its possible if you use img2img and have another picture as reference
好像可以很刑的样子
I have many pictures of carrying where the roles are reversed. if you're interested, I would like to share these pictures to you for feeding the AI. Can this be done?
男角色要是有点成熟的味道，霸道总裁的感觉就好了
Challenge to everyone: a girl carrying a boy
You could try to just change the faces with AfterDetailer lmao
This lora is very well done (at least for yuri carry, which is the only thing I tested).
Unlike other concept loras for couple/two people that I tried in this site, this one combines perfectly with other loras and has a big success rate.
Thanks!
It is indeed very tricky to train LoRAs with more than one person, and often doesn't work well, but this one came out nicely :)
@psoft I love your work! And I'm curious about how to make it work. How many imgs are needed or how to tag the imgs or any other tricks? Can you give us a simple introduction? Many thanks!
This is a wonderful LoRA. Would be great if you could make an XL version of this.
I might look into it at some point. I only recently started training SDXL LoRAs
@psoft Looking forward for SDXL version too