Preview images were not upscaled, ADetailer was used on some of them, I recommend using both, just wanted to show the "bare" model.

This is a ILXL2.0 + NAIEPS1.1 merge spiced up with my LoRAs, trained on the former.

I encourage posting images of any kind! Ofc, check TOS.

Main parameters:

• Res: 1194x1536 (supports 512x512~1536x1536 resolutions)

• Steps: 28

• CFG: 5

• Sampler-scheduler: euler a-normal

• CFGZeroStar and sometimes Self-Attention Guidance: default for difficult backgrounds

Try to BREAK prompts into parts by 75 tokens max, this goes for most models..

Positive prompt:

number of characters (1girl/1boy), character name, artist prompt, scene/environment/camera angle, action, expression, items, quality prompts (masterpiece, best quality)

Negative prompt:

lowres, worst quality

More prompt recommendations in about version.

Enjoy..