Instructions to use F16/krea2_bp with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use F16/krea2_bp with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("F16/krea2_bp", 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
What is the purpose of this lora?
Hello, I'm just wondering what this Lora job is, and thank you for your work.
If you have time can you make a ComfyUI custom node like your ComfyUI-DiversityBoost but for for the safety bypass on Krea 2 that very much nerfs the model big time and makes the model pretty bland and robotic especially with subjects and facial expression.
There is a great LORA called Krea2FilterBypass and custom nodes like ComfyUI-Conditioning-Rebalance and ComfyUI-Krea2T-Enhancer that help make the model better and free.
Unfortunately those LORAs and custom nodes do change the overall image and composition to much plus with other LORAs it does degrade the image. Or maybe something like your great F16/z-image-turbo-masked-dpo might help too.
Thank you.
