Instructions to use Lakonik/pi-FLUX.2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use Lakonik/pi-FLUX.2 with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("Lakonik/pi-FLUX.2", 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 Settings
- Draw Things
- DiffusionBee
Can we get this distillation for Kandinsky 5.0 Pro?
#1
by natalie5 - opened
Hi there, I really like your distillation of Qwen Image, though a video model distillation should be much better since video models require exponentially more compute per step than image models. I would recommend Kandinsky 5.0 Pro for distillation since it's currently the #1 open model in arena https://lmarena.ai/leaderboard/text-to-video/overall and it doesn't have other distillations like lighhtning/self forcing like Wan has. Your distillation will make it possible to run a model that's comparable to Luma Ray 3 on consumer GPUs. 🥰🥰
This comment has been hidden (marked as Resolved)
Thanks for the suggestion. Unfortunately, I don’t have access to any GPU resources over the next few weeks to do the distillation… (sponsors welcome!)