



















I didn't have a lot of training data for this girl, so I experimented some with her training configuration.
I used flip aug, color aug, and shuffle caption, which required me to turn off latent caching. It didn't really seem to impact the training time.
I felt like things were too tight at dim 24, so I moved to 32 for a couple of models. For this one, I cut that in half to 16, and I had plenty of good epochs to choose from. I ran 31 epochs, and ended up choosing the 17th.
None of those displayed any artifacts, and I had epochs to spare on both sides.
22+ showed artifacts around her nose, though none in the skin or hair.
You can find the training data and settings attached.
1. Modèle partagé uniquement à l'apprentissage et au partage. Droits d'auteur et interprétation finale réservés à l'auteur original.
2. Auteur souhaitant revendiquer le modèle : Contactez officiellement SeaArt AI pour l'authentification. Nous protégeons les droits de chaque auteur. Cliquer pour revendiquer
