Updated version. Much better.




















This is the "real" branch of my merged models. Please Read:
19-11-23 - V1.6 Includes VAE burned in.
As I stated before, I started training my own LORAs and merging them into my older model mixes. I am quite pleased with this newest version.
This model is a bit different than the previous versions. Excellent results can be had with little to no negatives. I do still recommend a negative chant of "worst quality, low quality, normal quality" with variance on the strength (i.e. (worst quality, low quality, normal quality:1.2) ). Having those tags too high will create an overdeveloped skin texture. See the examples I posted. Also having just "Best Quality" in your "positive" prompt will almost achieve the same effect.
See my samples for recommended settings. I use the CFG scale from 5.5 - 7.0. I use DPM++ 2M Karras for first pass and DPM++ 2M SDE Heun Karras for second pass. 512x768 is my usual resolution and it can go higher.
This model seems to excels at ?????? and ??? scenes.
Thanks for downloading and let me know what you think. And please post some examples!
I started training my own LORAs and merging them into my older merges. The output seems ok and shareable.
Enjoy.
Recommended NEG: "(worst quality, low quality, normal quality:1.8), (blurry, semi-realistic, cgi, 3d, render, sketch, cartoon, drawing, anime:1.4),
I will upload a non-VAE and a VAE burned in for each model.
Updated version. Much better.
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
