Controls eyebrow density/fullness, not shape or width. Push it negative and the eyebrows thin out to sparse, almost-translucent hairs, push it positive and they get thick, dense, and dark with no visible skin between the hairs.
Recommended range: -4 to +4, that's what every example above uses. It has also held up cleanly at more extreme weights (tested to ±10 on one subject with zero identity/age/ethnicity drift), but that hasn't been confirmed as reliably safe across many subjects yet, so treat anything past ±4 as experimental , your mileage may vary the further out you push it.
Description
initial release
FAQ
Comments (6)
Nice to finally see some high quality sliders in the feed. It looks like you have managed to keep ethnicity, exposure and contrast at bay aswell. Well done!
Loraholic himself!! 🔥 This honestly means a lot coming from you! Your sliders have been a huge inspiration, so seeing you say this about mine is pretty special. Thank you so much for the kind words and the Buzz! ❤️
Man who am I to argue with Loraholic? Really though, no notes. Tested with ±7 on a goofy prompt where I didn't define the faces at all and the prompt focused on everything else - still got a progression of brow hair without either the dog or the frog helmet growing weird eyebrows.
Thank you! Knowing it handled a those prompt with a dog and a frog helmet without bleeding effects is the ultimate stress test success. Thanks for testing it out and sharing! 🔥
@shakir_abdo Made my goofy prompt dumber by subbing in "Kratos" and "Aloy" to get some character consistency. Did base Krea2 capture their likeness? Lol no but that's not important. Did this LoRA still adjust their eyebrows even with helmet visors in the way (tried to avoid that in my first prompt)? Yep! Whatever your updated workflow is I would say it's working.
@boomhilde I'm very pleased to hear that ✨🤩