Instructions to use Vaelico/Wulver with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use Vaelico/Wulver with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("Vaelico/Wulver", dtype=torch.bfloat16, device_map="cuda") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Draw Things
- DiffusionBee
Thank you.
Legit good. I really do not care about furry, but I do care about anime, and this checkpoint is amazing for it, with great aesthetics and follows the prompt just as well as Krea 2. Have not tested it with loras yet, but they might not be needed. I did use 14 steps like the readme recommends, but I see that your workflow uses 8, like Krea 2. Not sure which to follow, but thank you for the checkpoint; love it.
It does have 1 issue: it's not that good with hands and feet. It has basic issues we have not had for a while, like extra fingers, feet not rendering toes, etc. Visually really strong, so I hope these get ironed out.
thanks! funny enough anime came out stronger than i expected. the workflow was my bad, 8 steps was left from an old test, readme is right and good catch on hands/feet, you're the first to report it. i think i know where it comes from (my training undersampled the timesteps where fine detail forms) and the next version goes right at it. Hopefully that fixes that.