import torch
from diffusers import DiffusionPipeline
# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained("SDXL", dtype=torch.bfloat16, device_map="cuda")
pipe.load_lora_weights("jarod2212/WILDLIFE_MASTER_PRIME_XL")
prompt = "-"
image = pipe(prompt).images[0]WILDLIFE‑MASTER PRIME‑XL
- Prompt
- -
- Prompt
- -
- Prompt
- -
- Prompt
- -
- Prompt
- -
- Prompt
- -
- Prompt
- -
- Prompt
- -
- Prompt
- -
- Prompt
- -
- Prompt
- -
- Prompt
- -
Model description
WILDLIFE‑MASTER PRIME‑XL Universal Animal Morphology LoRA — SDXL
A modularly‑trained LoRA built with an anatomical and morphological approach, not a photographic one. The dataset includes an extremely large number of species — far too many to list. The model does not memorize individual photos: it learns the structure of animals, their materials, textures, proportions, and natural lighting.
It is designed to generalize, not imitate.
It can reproduce common animals, rare animals, exotic species, marine life, terrestrial wildlife, birds, reptiles, mammals… and even species you wouldn’t expect. Part of the experience is discovering what it can generate.
Trigger word: wildprime Recommended weight: 0.8 – 1.1 Recommended checkpoints: realistic SDXL models (photorealistic, neutral, non‑illustrative)
⭐ Recommended Prompts:
wildprime, close-up portrait of a wild animal, natural colors, realistic lighting, detailed texture, clean background
wildprime, full-body wildlife shot, natural environment, realistic anatomy, accurate proportions, detailed fur/feathers/skin, soft natural light
wildprime, dramatic wildlife close-up, sharp focus on eyes, natural colors, high-detail morphology, neutral background
Trigger words
You should use wildprime to trigger the image generation.
Download model
Download them in the Files & versions tab.
- Downloads last month
- 22