import torch
from diffusers import DiffusionPipeline
# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained("John6666/wai-nsfw-illustrious-v110-sdxl", dtype=torch.bfloat16, device_map="cuda")
pipe.load_lora_weights("Speedso/Saezuru_Usagi_style")
prompt = "masterpiece, best quality, amazing quality, absurdres, 1girl, portrait, solo, standing, looking at viewer, red eyes, blue hair, twintails, :<, red hakama, rabbit ears, tail, rabbit girl, ahoge"
image = pipe(prompt).images[0]Saezuru Usagi style
- Prompt
- masterpiece, best quality, amazing quality, absurdres, 1girl, portrait, solo, standing, looking at viewer, red eyes, blue hair, twintails, :<, red hakama, rabbit ears, tail, rabbit girl, ahoge
- Prompt
- masterpiece, best quality, amazing quality, absurdres, 1girl, fox ears, traditional media, fox tail, solo, cowboy shot, medium breasts, looking at viewer, medium hair, smile, one eye closed, thigh gap, thick thighs, double fox shadow puppet,
- Prompt
- masterpiece, best quality, amazing quality, absurdres, 1girl, solo, animal ears, blue hair, simple background, smile, ahoge, :d, white kimono, open mouth, fox ears, animal ear fluff, fang, thick eyebrows, kimono, upper body, short eyebrows, half updo, red eyes, short hair, french braid
Model description
Saezuru Usagi/Freezeusagi style.
Trained on 3326 pieces of artwork, used WAI-NSFW-illustrious-SDXL as a base.
No trigger words.
Check out "traditional media(+shikishi)" tags for hand drawn+painted images, I think they look pretty good.
Download model
Download them in the Files & versions tab.
- Downloads last month
- 16
Model tree for Speedso/Saezuru_Usagi_style
Base model
KBlueLeaf/kohaku-xl-beta5 Finetuned
John6666/wai-nsfw-illustrious-v110-sdxl