I need the community's help!
I'm currently working extensively with Z Image Turbo under Forge Neo WubUI. I'm impressed by the speed and quality.
I'm having a problem with the generation variations. I've created a prompt for an image. No matter which seed I set, the results are largely the same, without any significant variations. (I've noticed the same thing with other prompts, but not to this extreme.)
My question to the community:
Is it my settings or my prompt?
How can I achieve more image variations?
I hope you have some tips for me!!!!!
Prompt:
Fantasy Art, High-Resolution Photo,
A mystical warrior in a fantasy mountain landscape. half-body shot.
A woman, portrayed as a mystical warrior.
She has short black hair.
Her large blue eyes are accentuated by long eyelashes and striking makeup.
She has fair skin with a few freckles.
She has a slender figure with medium breasts.
She wears a brown leather corset with a deep neckline and open, loose lacing, (sideboob), and a miniskirt.
Around her neck, she wears several metal chains of varying lengths, each adorned with small, glowing Celtic symbols.
A wide silver metal bangle, also featuring glowing symbols, graces her upper arm.
In her hands, she holds a large, glowing rock crystal radiating magical energy that bathes her face and upper body in warm light, casting soft shadows.
The surroundings are a mystical, rugged mountain landscape with craggy rocks, large boulders, and scattered patches of grass.
Dark storm clouds with flashes of lightning are visible in the sky, while rain pours down.
The entire atmosphere is dark, creating a striking contrast with the glowing rock crystal.
breast focus, mystical, energy, fantasy, rain, raindrops, extremely detailed, dark, 8K,
Metadata:
Steps: 10
Sampler: Euler a
Schedule type: Simple
CFG scale: 1
Shift: 4
Seed: 335980777
Size: 832x1216
Model: zImageTurbo_turbo (https://civitai.red/models/2168935/z-image-turbo)
Module 1: zImageTurbo_vae
Module 2: zImageTurbo_textEncoder
The result: