import torch
from diffusers import DiffusionPipeline
# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained("stablediffusionapi/pony-diffusion-v6-xl", dtype=torch.bfloat16, device_map="cuda")
pipe.load_lora_weights("Horsebird/klxl")
prompt = "score_9, score_8_up, woman, fox ears, teal hair, ponytail, black sclera, golden eyes, slit pupils, lipstick, tan, chest tuft, qipao, lamia, smoke, shadow, glow, water, night"
image = pipe(prompt).images[0]KLXL
- Prompt
- score_9, score_8_up, woman, fox ears, teal hair, ponytail, black sclera, golden eyes, slit pupils, lipstick, tan, chest tuft, qipao, lamia, smoke, shadow, glow, water, night
- Negative Prompt
- chibi, 3d
Model description
A style lora without any specific trigger tokens, trained on PonyXL v6. Usable weight range appears to be 0.6-1.0 for the moment. Responds strongly to water stuff. Just prompt normally and see if you can have fun with it. It's still a WIP with more than a few clear bias issues affecting clothing, hairstyles, skin tones, architecture, and probably a lot of stuff I just haven't found yet.
Download model
Weights for this model are available in Safetensors format.
Download them in the Files & versions tab.
- Downloads last month
- -