Description
FAQ
Comments (8)
I'm curious about the training process for this? I'm thinking about instance prompt, class prompt and such parameters in dreambooth, and source image content.
Asking because I'm already doing loras for faces, but I'm thinking about making loras to modify various body elements without changing the likeness of the persons face etc.
source images? hmm, if i recall "total recall"? a classic and the first thing everyone has in mind when seeing women with 3 breasts. xD
Thank you for responding. :)
I know, the source images would depend on what concept is being trained. I'm more curious about is the dreambooth entries for "instance prompt" and "class prompt" and such. If I'm training a "person" class this would train the face as well, so the class prompt would need to be different? Is there a list of class prompts somewhere, or can it be anything?
@PixViz i think, you replied to the wrong person.
@stapfschuh yes, I see that now. If the creator of this Lora would like to chime in that'd be great. :)
The model we never wanted, but somehow exists.
3 Tits....awesome!
Is there any trigger word?