RaBitQ-Library v0.6.0

An official lightweight library for the RaBitQ algorithm and its applications in vector search.

Release notes

<!-- Release notes generated using configuration in .github/release.yml at v0.6.0 -->

## What's Changed
### Dependencies
* Upgrade vendored Eigen to 5.0.1 by @gouyt13 in https://github.com/VectorDB-NTU/RaBitQ-Library/pull/135
### Other changes
* docs: add benchmarks by @gouyt13 in https://github.com/VectorDB-NTU/RaBitQ-Library/pull/130
* Fix HNSW entry-point races and construction failure handling by @gouyt13 in https://github.com/VectorDB-NTU/RaBitQ-Library/pull/131
* fix: respect CPU affinity and reduce clustering parallel overhead by @gouyt13 in https://github.com/VectorDB-NTU/RaBitQ-Library/pull/132
* Relax dimension padding requirements from 64 to 32 by @gouyt13 in https://github.com/VectorDB-NTU/RaBitQ-Library/pull/133
* Add configurable IVF centroid routing with persisted initializer types by @gouyt13 in https://github.com/VectorDB-NTU/RaBitQ-Library/pull/134
* fix: isolate CMake optimization flags from consumers by @gouyt13 in https://github.com/VectorDB-NTU/RaBitQ-Library/pull/136
* fix: release Python GIL and isolate concurrent index searches by @gouyt13 in https://github.com/VectorDB-NTU/RaBitQ-Library/pull/137
* test: add historical index persistence compatibility coverage by @gouyt13 in https://github.com/VectorDB-NTU/RaBitQ-Library/pull/138


**Full Changelog**: https://github.com/VectorDB-NTU/RaBitQ-Library/compare/v0.5.2...v0.6.0

Share this resource