jasonz-lanceandClaude Sonnet 5
1b16057d23
_TwoPhaseSplitReader retains _plan_block_pos/_plan_local_offset for its whole lifetime, one entry per row in the entire split (not bounded by window_blocks -- see the open whole-split-memory review thread), so their dtype is a real per-row memory cost, not a rounding choice. Both were int64 despite holding small values: block_pos is a block index (bounded by this split's block count) and local_offset is an in-block row offset (bounded by block_size), both far smaller than the row-rank range sigma/order need int64 for. Switch to int32, falling back to int64 only if a single split's block count or block size itself exceeds what int32 can represent (practically never). Halves this bookkeeping's memory cost for any realistic config, as a stopgap short of the full lazy/windowed redesign that would bound it by window_blocks instead of split size -- deliberately deferred, since that touches the same hot-path indexing logic as the prefetch-loop regression fixed earlier and deserves its own dedicated pass. Added a regression test pinning the dtype (so a future change can't silently regress it back to int64) and checking values are still correct, not just narrower. Full suite: 211 passed. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
The Multimodal AI Lakehouse
How to Install ✦ Detailed Documentation ✦ Tutorials and Recipes ✦ Contributors
The ultimate multimodal data platform for AI/ML applications.
LanceDB is designed for fast, scalable, and production-ready vector search. It is built on top of the Lance columnar format. You can store, index, and search over petabytes of multimodal data and vectors with ease. LanceDB is a central location where developers can build, train and analyze their AI workloads.
Demo: Multimodal Search by Keyword, Vector or with SQL
Star LanceDB to get updates!
⭐ Click here ⭐ to see how fast we're growing!
Key Features:
- Fast Vector Search: Search billions of vectors in milliseconds with state-of-the-art indexing.
- Comprehensive Search: Support for vector similarity search, full-text search and SQL.
- Multimodal Support: Store, query and filter vectors, metadata and multimodal data (text, images, videos, point clouds, and more).
- Advanced Features: Zero-copy, automatic versioning, manage versions of your data without needing extra infrastructure. GPU support in building vector index.
Products:
- Open Source & Local: 100% open source, runs locally or in your cloud. No vendor lock-in.
- Cloud and Enterprise: Production-scale vector search with no servers to manage. Complete data sovereignty and security.
Ecosystem:
- Columnar Storage: Built on the Lance columnar format for efficient storage and analytics.
- Seamless Integration: Python, Node.js, Rust, and REST APIs for easy integration. Native Python and Javascript/Typescript support.
- Rich Ecosystem: Integrations with LangChain 🦜️🔗, LlamaIndex 🦙, Apache-Arrow, Pandas, Polars, DuckDB and more on the way.
How to Install:
Follow the Quickstart doc to set up LanceDB locally.
API & SDK: We also support Python, Typescript and Rust SDKs
| Interface | Documentation |
|---|---|
| Python SDK | https://lancedb.github.io/lancedb/python/python/ |
| Typescript SDK | https://lancedb.github.io/lancedb/js/globals/ |
| Rust SDK | https://docs.rs/lancedb/latest/lancedb/index.html |
| REST API | https://docs.lancedb.com/api-reference/rest |
Join Us and Contribute
We welcome contributions from everyone! Whether you're a developer, researcher, or just someone who wants to help out.
If you have any suggestions or feature requests, please feel free to open an issue on GitHub or discuss it on our Discord server.
Check out the GitHub Issues if you would like to work on the features that are planned for the future. If you have any suggestions or feature requests, please feel free to open an issue on GitHub.
Contributors
Stay in Touch With Us