A small test to see how few images are really needed for a character LoRA. Surprised by how well it came out considering the art available for Nathan doesn't even reach double digits.
Forever in test mode.
Description
FAQ
Comments (2)
Amazing, it must have been really hard to train this one, since there aren´t many reference pictures. Thank you so much!
The hard part was just finding good quality artwork, it wouldn't let me make a LoRA without 10 pictures so I had to pad it out with a couple of in-game sprites. Now that I know it can turn out like this even with only a few, I'll see if I can remake it at some point with even better quality images. I noticed the face messes up a lot depending on the checkpoint but I'm pretty sure that can be easily fixed.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.










