Instructions to use gauravgandhi2411/aetherart-ukiyo-sd21 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use gauravgandhi2411/aetherart-ukiyo-sd21 with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("sd2-community/stable-diffusion-2-1", dtype=torch.bfloat16, device_map="cuda") pipe.load_lora_weights("gauravgandhi2411/aetherart-ukiyo-sd21") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Draw Things
- DiffusionBee
docs: warn that the auto-generated usage snippet is broken/untested
#1
by gauravgandhi2411 - opened
The auto-generated "Use this model" snippet omits the required madebyollin/sdxl-vae-fp16-fix VAE (SDXL) or uses an untested bfloat16/device_map config (SD 2.1), diverging from this repo own validated Usage section. Adds a prominent warning at the top of the card body pointing to the correct code below. For the SDXL card the failure mode is measured directly: ~280s/denoising-step VRAM-oversubscription on an 8GB GPU when the auto-snippet is copied verbatim.
Superseded β applied directly to main.
gauravgandhi2411 changed pull request status to closed