turbovec
A vector index built on TurboQuant, written in Rust with Python bindings
Integrate Turbovec with LangChain for persistent, quantized vector storage. Replace in-memory stores easily with TurboQuantVectorStore without code changes. Boost performance and scalability.
How to Use Turbovec for In-Memory Serialization and DeserializationLearn to use turbovec for fast in-memory serialization and deserialization of quantized vector indexes. Achieve deterministic, disk-free round-trip data handling.
How turbovec Handles Large Batch Operations and Ctrl-C Interruption: Architecture Deep DiveExplore how turbovec manages large batch operations and Ctrl-C interruption. Discover its architecture for efficient parallel query processing and graceful aborts.
How to Perform Hybrid Retrieval with Filtered Search in Turbovec Using an AllowlistLearn to emulate hybrid retrieval in Turbovec by filtering metadata to an allowlist and performing vector search on the constrained set. Enhance your search accuracy now.
SIMD Kernel Optimizations in Turbovec: x86 AVX2, AVX‑512, and ARM NEON ImplementationsExplore Turbovec's SIMD kernel optimizations using x86 AVX2, AVX-512, and ARM NEON for accelerated 4-bit vector search. Achieve bit-identical scoring across architectures.
Effective TQ+ Calibration Rules in Turbovec: A Complete GuideMaster effective TQ+ calibration rules in Turbovec. Learn the two-step process, sampling needs, and why early calibration is crucial for accurate results. Avoid common pitfalls.
When and How to Apply TQ+ Calibration in Turbovec for Maximum RecallBoost TurboVect recall by 2.5 points on average with TQ+ calibration. Learn when and how to apply this per-coordinate shift and scale transformation for optimal results.
How to Configure Threading for Parallel Search in Turbovec Using `RAYON_NUM_THREADS`Optimize Turbovec search performance by configuring threading with RAYON_NUM_THREADS. Learn how to set the environment variable for efficient parallel processing and maximize your hardware utilization. Control thread pool size ...
Turbovec Bit Widths: How 2, 3, and 4-Bit Quantization Impact Compression and PerformanceExplore Turbovec's 2, 3, and 4-bit quantization options to understand their impact on compression ratios and performance. Optimize your vector search with precise bit width selection.
How to Perform Incremental Saves with Turbovec Using `sync()`Learn how to perform incremental saves with Turbovec's sync() method. Efficiently update your index by writing only changed data, avoiding full rewrites.
Turbovec Index File Formats Explained: `.tv` and `.tvim` Binary StructureExplore the Turbovec index file formats .tv and .tvim. Understand the binary structure for compact snapshots and persistent indexes. Learn how Turbovec stores vector data.
How Turbovec Handles Deletions in the TurboQuantIndex and IdMapIndexLearn how turbovec efficiently handles deletions in TurboQuantIndex and IdMapIndex using swap_remove and dual-swap for O(1) synchronized updates without element shifting.
Have a question about this repo?
These articles cover the highlights, but your codebase questions are specific. Give your agent direct access to the source. Share this with your agent to get started:
curl -s "https://instagit.com/install.md" 
Maintain an open-source project? Get it listed too →