Library / SDK
Inference & ServingFine-tuning & Training
Diffusers
1Pretrained 2diffusion 3model 4pipelines 5for 6generation
the six Ws · specification
W1
Who
Hugging Face maintains Diffusers as an open source Python library.
W2
What
Diffusers provides pretrained diffusion model pipelines for image, audio and video generation.
W3
Where
Available on GitHub at huggingface/diffusers and distributed via PyPI.
W4
When
First released in 2022 and under continuous active development.
W5
Why
It gives developers ready to use, modular diffusion pipelines instead of reimplementing sampling and schedulers from scratch.
W6
With
Built on PyTorch and integrates with the Hugging Face Hub, Transformers and Accelerate.
W7
Watch
Apache 2.0 licensed, open source, maintained by Hugging Face, roughly 34k GitHub stars, actively maintained.
alternatives