# Kpop Demon Hunters 2025

A style LoRA for Krea 2. Applies a consistent visual treatment without locking the subject matter.

## Usage

Include the trigger word *kdm** in your prompt.

Recommended weight: 1. Start there and adjust — styles often read well below 1.0, and lower weights blend more naturally with other LoRAs.

## Prompt Examples

Here are some prompt examples:

"kdm, A surprised animated character in a light blue blazer holds a thick folder, standing in a dimly lit, modern indoor space with tiled walls and distant figures. The camera frames them in a medium shot, emphasizing their wide-eyed expression under cool, ambient lighting."

"kdm, two animated women stand side by side in a brightly lit, teal-toned indoor space, one in a yellow hat and floral jacket looking surprised, the other in a cap and glasses with a calm expression, framed in a medium shot."

"kdm, five stylized animated figures with vibrant, colorful hair stand together in a softly lit studio setting, all dressed in pink-toned outfits; the central figure holds a microphone while the others pose around them, captured in a medium group shot with a muted background."

"kdm, three animated female characters stand side by side on a city street in front of a coffee shop, each with distinct outfits and expressions. The scene is lit with warm, ambient evening light, and the camera frames them in a medium shot that captures their upper bodies and the storefront behind them."

"kdm, A stylized animated character with long magenta pigtails stands confidently in a futuristic, neon-lit cabin, holding a small object while looking toward the viewer. The scene is framed with a shallow depth of field, emphasizing the character against soft, glowing purple and blue ambient lighting."

"kdm, A cheerful animated character with purple hair and a pink hoodie gives a thumbs-up while seated at a desk in an office, surrounded by books, framed certificates, and a calendar, under soft, even lighting with a shallow depth of field focusing on the subject."

"kdm, two stylized female performers in futuristic, metallic outfits strike dynamic dance poses on a brightly lit stage, framed by a colorful, abstract backdrop with warm stage lighting casting a glow across the scene."

"kdm, An aerial view captures a bustling urban sidewalk scene with pedestrians walking past a bank and food vendors, under bright daylight that casts sharp shadows, framed from above to emphasize the lively street activity and architectural details."

"kdm, A young person with vibrant purple hair sits outdoors in a grassy, sunlit field, looking directly at the camera with a calm, thoughtful expression. The background is softly blurred, emphasizing the subject, while warm, natural lighting highlights their features and the gentle folds of their clothing."

"kdm, A stylized female warrior with pink hair and spiked headgear grips a glowing sword, her intense gaze fixed forward in a tense, action-ready stance. The scene is bathed in vibrant green lighting with motion blur and glowing energy effects, framed tightly to emphasize her determined expression and dynamic posture."

"kdm, three stylized animated characters pose confidently at a table during a fan event, with bold purple lighting and a backdrop displaying “FAN EVENT” in large letters; the camera captures them from a frontal, eye-level angle."

"kdm, A woman and a young girl with purple hair sit together on the grass in a sunlit forest clearing, surrounded by rocks and wildflowers, with a vase of white blooms and apples nearby; the scene is framed with soft, natural lighting and a shallow depth of field that blurs the background."

"kdm, A diverse crowd of animated characters is gathered, many with raised hands and expressions of excitement or awe, suggesting they are watching an event. The scene is brightly lit with a vibrant, colorful palette, framed in a medium shot that captures the energy and collective engagement of the audience."

"kdm, A group of animated characters with expressive, surprised faces are gathered closely together, their arms outstretched as if reacting to something exciting or shocking. The scene is brightly lit with vibrant colors, set outdoors in a bustling, slightly blurred background, captured in a tight, eye-level shot that emphasizes their animated expressions and colorful outfits."

"kdm, three stylized animated women with distinct hair colors and ornate outfits stand in a row, hands clasped in front of them, against a vibrant purple and blue gradient background. The camera frames them in a medium shot with soft, glowing lighting that highlights their expressions and detailed clothing."

"kdm, A serene, stylized figure with a towel wrapped around their head rests in a steaming hot tub, eyes closed in relaxation, framed by soft, diffused lighting and a tiled, spa-like background."

"kdm, in a hazy, green-lit environment, a figure with a glowing staff stands poised near a hovering, mechanical object, while shadowy, elongated creatures lurk in the misty background. The scene is framed wide, emphasizing the eerie, atmospheric glow and the sense of mystery surrounding the central action."

"kdm, A stylized animated scene features a central figure with vibrant pink hair and a white robe, surrounded by several people in a dimly lit, industrial-looking space, as they appear to be applying makeup or accessories. The lighting casts dramatic shadows and highlights, with the camera framed in a medium shot that captures the group’s focused interaction."

"kdm, A stylized animated character with dark hair in buns and expressive eyes is enthusiastically pulling open a snack bag while seated on a purple couch, framed in a close-up shot with soft, ambient lighting."

"kdm, two figures stand in a futuristic, high-rise lounge overlooking a sprawling city at night, framed by expansive windows and illuminated by vibrant, colorful lighting beneath a curved sofa. The camera captures the scene from a wide, elevated angle, emphasizing the modern, minimalist interior and the glowing urban expanse beyond."

## Training

| | |

|---|---|

| Base model | Krea 2 (trained on Raw) |

| Type | Style LoRA |

| Dataset | 130 images |

| Steps | 2000 |

| Network rank / alpha | 32 / 32 |

| Learning rate | 0.0002 |

| Optimizer | adamw8bit |

| Resolutions | 768 / 1024 |

## Generation settings

Trained on Krea 2 Raw, and transfers to Krea 2 Turbo for faster

generation.

- Raw: ~30 steps, guidance 3.5

- Turbo: ~8 steps, guidance 1.0