new trigger word : olmn
recommended checkpoint : Realistic Vision 3 No VAE
if you like my stuff pls consider supporting me on Kofi
accepting Commissions as well...
Show More
Description
should work just fine but lemme know
FAQ
Comments (10)
Don't you mean "that hot girl from G4TV"?
That could be so many people.
hehe, I know her now, I was a baby boy when that show used to air
@dogu_cat She was best on HBO's "The Newsroom".
Amazing work ! can your share your training settings and details?
This LoRa looks so good and works on many models. How did you do that? Can you share some tips? I tried making some LoRas but mine look bad. I followed some guides but that didn't help much.
use good photos, use gfp gan and gigapixel to increase detail
@dogu_cat Your stuff is phenomenal, I have photos of myself I'm using that are very high quality, but the result always comes out cartoonish / anime like. Your Lora's work super well for hyper realistic scenarios, I just dont get how.
Is it possible to use too many photos? Could that be my problem?
@Redhawk117 too much photos + too many steps/image can cause overburn or cartoonish artifacts, if you have many images try low step count
@dogu_cat Is 100 steps a good number?