strkyyy/anime-cyberpunk

SDXL LoRA for Dark Futuristic Style: Cyberpunk Anime Aesthetic

What it does

This LoRA model applies a dark cyberpunk anime aesthetic to your prompts, inspired by iconic works like Ghost in the Shell and Akira. It adds moody neon-lit megacities, rain-slicked streets, holographic ads, and gritty industrial details with deep shadows and neon accents.

How to use

To apply this style, include the trigger word 'cyb3rn1me' in your prompt. A suggested strength of 0.85 works well for most scenes, but adjust as needed depending on how dominant you want the cyberpunk elements to be. Use EMA sampler with 25 steps for best results.

Best for

• Scenes featuring hackers and detectives in neon-lit environments
• Futuristic megacities with flying cars and towering skyscrapers
• Cybernetic characters with advanced technology and samurai-inspired aesthetics
• Rainy alleyways filled with noodle stalls and hidden secrets
• Industrial settings with a gritty, dystopian feel

Tips

• Experiment with different strengths to find your preferred level of cyberpunk influence.
• Combine 'cyb3rn1me' with other style triggers for unique visual blends.

Example Prompts

cyb3rn1me cybernetic catgirl portrait, rainy night neon reflections, moody lighting

cyb3rn1me futuristic skyline at dawn, rain-slick streets, soft glowing neon signs, architectural details

cyb3rn1me robotic arm with circuitry exposed, metallic sheen, intricate wiring, sharp focus

cyb3rn1me biker gang on rainy street, speeding motorcycles, wet asphalt, dramatic shadows and highlights

cyb3rn1me gloomy alleyway with neon signs, rain drops hitting puddles, dark mood, low angle shot

cyb3rn1me abstract digital pattern, glitch effect, neon colors, sharp contrast

Trigger word

Activate with cyb3rn1me. Recommended strength: 0.9.

Showcase

Usage with diffusers

from diffusers import AutoPipelineForText2Image
import torch

pipe = AutoPipelineForText2Image.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0", torch_dtype=torch.float16).to("cuda")
pipe.load_lora_weights("strkyyy/anime-cyberpunk")
image = pipe("cyb3rn1me, your prompt", num_inference_steps=30).images[0]