Instructions to use AuraDiffusion/16ch-vae with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use AuraDiffusion/16ch-vae with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("AuraDiffusion/16ch-vae", 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
Training, and why fp16?
#4
by ppbrown - opened
Hello,
I was wondering if you might answer me two questions:
what were the specific settings you used when you trained the vae?
why did you choose fp16 rather than bf16?
@cloneofsimo should know the answer to number 1 but for fp16 vs bf16, the increased precision was wanted rather than the greater range of bf16.