Instructions to use Linaqruf/animagine-xl-2.0 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use Linaqruf/animagine-xl-2.0 with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("Linaqruf/animagine-xl-2.0", dtype=torch.bfloat16, device_map="cuda") prompt = "face focus, cute, masterpiece, best quality, 1girl, green hair, sweater, looking at viewer, upper body, beanie, outdoors, night, turtleneck" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Draw Things
- DiffusionBee
Do we need to include VAE?
#4
by MicahB - opened
For some SDXL Models, the VAE is needed, but it appears that yours operates without it. However, I'm getting pretty blurry images in Automatic1111's Stable Diffusion UI.
Running the latest version of Automatic1111, No XFormers, on a 4070 12G, with 32GB of system ram, alongside a Ryzen 9 5900X.
Generation settings:
- 512*768 Image
- CFG: 12
- Sampler: Euler a
- Clip Skip: 1
- Prompt
Also, when I try to load the VAE that you've uploaded in the VAE folder, I get this error.
Please let me know if I have anything setup wrong and I'll try to get it fixed.
The VAE is already baked inside, I use https://huggingface.co/madebyollin/sdxl-vae-fp16-fix
And this is SDXL, you need to use resolution at 1024x1024 to make it work