Instructions to use ckpt/anything-v3-vae-swapped with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use ckpt/anything-v3-vae-swapped with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("ckpt/anything-v3-vae-swapped", 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
- Draw Things
- DiffusionBee
Source of vae?
#1
by godly-devotion - opened
Thanks for this repo, the generated images are definitely much better than the plain anything v3. I was wondering which vae did you use because I've tried using SD's official diffusers based vaes but the colors come out too strong, this one was just right.
Hi @godly-devotion π this is the vae https://huggingface.co/Linaqruf/anything-v3.0/blob/main/Anything-V3.0.vae.pt
Oh wow, how were you able to convert it to diffusers format?
I will add converter here https://huggingface.co/spaces/camenduru/converter