Instructions to use madebyollin/sdxl-vae-fp16-fix with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use madebyollin/sdxl-vae-fp16-fix with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("madebyollin/sdxl-vae-fp16-fix", 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
can you fix the encoder?
#9
by BoyuanJiang - opened
encoder still overflow when fp16
Hmm, are you using the latest version (including this commit)? The latest version of the sdxl-vae-fp16-fix encoder is not supposed to overflow:
If you've found an image that reliably overflows the latest version of the sdxl-vae-fp16-fix encoder, please share and I can look into it.

