Instructions to use ostris/sdxl-sd1-vae-lora with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use ostris/sdxl-sd1-vae-lora with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0", dtype=torch.bfloat16, device_map="cuda") pipe.load_lora_weights("ostris/sdxl-sd1-vae-lora") prompt = "photo in the sd1 latent space" image = pipe(prompt).images[0] - Inference
- Notebooks
- Google Colab
- Kaggle
- Local Apps
- Draw Things
- DiffusionBee
I didn't quite understand.
#1
by razvanab - opened
Hi.
So it does do from SDXL Chekpont to SD15 Loras or vice versa?
Thank you.
Hi.
So it does do from SDXL Chekpont to SD15 Loras or vice versa?
Thank you.
No, this is, I'm guessing, adjusting the output (last layers I suppose?) of SDXL so that it can be turned into an image by the SD1.5 VAE rather than the horribly bloated and memory inefficient SDXL VAE.
I see. Thanks
Oh well, I am still waiting for a method to use SD1.5 Loras with SDXL.