Instructions to use stabilityai/stable-diffusion-3.5-large-controlnet-blur with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use stabilityai/stable-diffusion-3.5-large-controlnet-blur with Diffusers:
pip install -U diffusers transformers accelerate
from diffusers import ControlNetModel, StableDiffusionControlNetPipeline controlnet = ControlNetModel.from_pretrained("stabilityai/stable-diffusion-3.5-large-controlnet-blur") pipe = StableDiffusionControlNetPipeline.from_pretrained( "fill-in-base-model", controlnet=controlnet ) - Notebooks
- Google Colab
- Kaggle
- Local Apps
- Draw Things
- DiffusionBee
How to get good results in ComfyUI?
#4
by Little-God - opened
Hi, first i love the output SD 3.5 delivers. Unfortunately all the workflows i found online using this controlNet model did almost nothing. So i build my own as you suggested in your Dokumentation using tiled diffusion. This makes slightly better images, but its nothing compared to a "normal" ultimate SD Upscale workflow. No one seems to know how to uses this properly.
Can you please provide a default workflow and also which settings to use.
I tried every setting i could think of while using Tiled diffusion but nothing.