Instructions to use stabilityai/stable-cascade with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use stabilityai/stable-cascade 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-cascade", 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
How to create Comfy Checkpoints from BF16 fine-tuned B/C/Text Encoder Safetensors
#56
by c0ur4ge - opened
I'd like to take the fine tuned versions of Cascade that I've created and adopt them for release for use in the official comfy workflow rather than inferencing them as unet/te/etc. How does the comfy team (or can I) output the model in the combined format preferred by Comfy UI from what OneTrainer outputs for BF16?