All notable changes to this project will be documented in this file.
The format is based on Keep a Changelog and this project adheres to Semantic Versioning.
Initial release of a PyTorch Dataset for SOMA Experiments. This is a port and enhancement of code contributed by the Chan Zuckerberg Initiative Foundation CELLxGENE project.
This is not a one-for-one migration of the contributed code. Substantial changes have been made to the package utility (e.g., multi-GPU support), improved API UX, performance improvements, and more.
- Initial project organization and other scaffolding PR #4
- Simple, non-shuffling Dataset/DataPipe implementation PR #6
- Add CI workflows PR #7
- Add a DataLoader creation wrapper function PR #8
- Add shuffling support PR #9
- Add first draft of tutorial notebooks PR #10
- Archive script used to populate the repo commit history PR #11
- I/O buffer performance optimization PR #13