Instructions to use yijunwang2/krea2-reid with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use yijunwang2/krea2-reid with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline from diffusers.utils import load_image # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("krea/Krea-2-Turbo", dtype=torch.bfloat16, device_map="cuda") pipe.load_lora_weights("yijunwang2/krea2-reid") prompt = "Turn this cat into a dog" input_image = load_image("https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/cat.png") image = pipe(image=input_image, prompt=prompt).images[0] - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Draw Things
Workflow example
#1
by Kimboooo - opened
Hey! Excited to try this out. Do you have ComfyUI workflow example? Thanks
I just updated a workflow, feel free to have a play and let me know.
Warm reminder for ComfyUI testing: please use the official workflow included in the model repository rather than the default generic Krea 2 Edit graph. This ReID adapter was trained with isolated reference attention, so kv_cache must remain enabled. The tested settings are the Comfy-Org Krea 2 Turbo INT8 ConvRot checkpoint, LoRA strength 1.0, 8 steps, CFG 1.0, and Euler/simple sampling. The workflow also prepares the reference at the released 384Γ384 pixel budget