Update: added 2 gig fp16 version in dropdown

My aim was to get away from the anime, cartoon look and get closer to realism but with a fantasy rendered cgi style.

Recipe was: 50/50 AOM2 and anything + 20 DreamShaper + 20 Experience

My prompts are in the example images.
Useful prompt keywords:

digital art, illustrated, realistic face, 3d rendered face, shaded face, best quality

This goes in \stable-diffusion-webui\models\Stable-diffusion

VAE used (goes in \stable-diffusion-webui\models\VAE): vae-ft-mse-840000-ema-pruned.ckpt

Example images settings:

This Lora is used (goes in \stable-diffusion-webui\models\Lora):
Sam Yang I keep it down at 0.3 as it can be overpowering.

These textual inversions are used (they go in \stable-diffusion-webui\embeddings):
bad_prompt_version2, bad-hands-5, EasyNegative, ng_deepnegative_v1_75t

Clip skip: 2
Eta noise seed delta: 31337

See more of my gens on https://www.deviantart.com/realitybyte5