import torch
from diffusers import DiffusionPipeline
# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained("krea/Krea-2-Turbo", dtype=torch.bfloat16, device_map="cuda")
pipe.load_lora_weights("HuggingVince/w3dg1e_krea2")
prompt = "an amateur photo of woman wearing panty wedgie the panty elastic is pulled up over her shoulders, "
image = pipe(prompt).images[0]w3dg1e
- Prompt
- an amateur photo of woman wearing panty wedgie the panty elastic is pulled up over her shoulders,
- Prompt
- the cencor was added in photoshop cause new account havent read TOS on nudity
Model description
Over shoulders panty wedgie
Trigger words
Lora works best at 0.7 strength w3dg1e was the consistant trigger it was trained with although not really needed in generation and can add a watermark i suspect i need to understand Krea2 training more but this does the job. just use natural language a woman wearing panty the elastic stretched up over her shoulders
Download model
Download them in the Files & versions tab.
- Downloads last month
- -