Instructions to use ostris/OpenFLUX.1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use ostris/OpenFLUX.1 with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("ostris/OpenFLUX.1", 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
Is someone working on controlnet for OpenFlux?
#11
by martintomov - opened
what the title says. it would be a game changer
Is there benfits to training openflux vs the regular model?
Also would it be better to have a controlnet trained for this schnell version or the dev version de-distilled
@tristan22mc This supports normal CFG(negative prompts as well) and might be a bit more stable in training.
For the 2nd question, I don't believe there will be a huge difference in OpenFLUX.1 and dev de-distilled in terms of performance in training but OpenFLUX.1 has a much more permissive license since its based of schnell.