Announcing Flower Datasets 0.6.0

The Flower Team is excited to announce the release of Flower Datasets 0.6.0!

Flower Datasets (flwr-datasets) is a library designed to quickly and easily create datasets for federated learning, federated evaluation, and federated analytics.

Thanks to our contributors

We would like to give our special thanks to all the contributors who made this new version of Flower Datasets possible (in git shortlog order):

Adam Tupper, Alireza Ghasemi, Daniel Hinjos García, Haoran Jie, Heng Pan, Iason Ofeidis, Javier, Yan Gao

What's new?

• Introduce Flower Datasets CLI (#6514, #6520)

  Run flwr-datasets in your terminal to access a new CLI. The first command, create, lets you quickly generate federated datasets from Hugging Face datasets and save them to disk. This is perfect for creating demo data to test Flower SuperNodes. Learn more.

• Add ContinuousPartitioner (#5235)

  ContinuousPartitioner enables non-IID partitioning based on real-valued dataset properties with adjustable strictness. It interpolates between IID and non-IID partitioning using a strictness parameter that blends standardized property vectors with Gaussian noise.

• New Recommended Datasets (#6483, #4966, #4853)

  • Add FedJam multimodal federated dataset: A 36k sample dataset featuring spectrogram images and time-series KPIs for wireless jamming classification
  • Add CareQA medical benchmark: A dataset of 5,621 QA pairs from official Spanish healthcare exams (2020–2024) for medical LLM evaluation
  • Add two ChEMBL molecular datasets to the recommended FL datasets list

• Bug Fixes (#5340)

  • Fix NaturalIdPartitioner and DirichletPartitioner producing inconsistent partition assignments across dataset splits by sorting unique IDs/classes before mapping. This issue occurred when partitioning different splits (e.g., train/test) or datasets with different orderings.

• Bump datasets dependency (#6126)

  The datasets dependency now supports versions 4.x (previously only up to 3.1.0).

• Improved Documentation (#5948,#6127)

• General Improvements (#6126, #6132, #6131, #4854, #5201, #5989, #5248, #6543)