cyberneticlibrary

Search billions of vectors fast

faissskillsetup L29,423
Orchestra-Research/AI-Research-SKILLs
What it does

Search billion-scale embeddings with GPU acceleration and multiple index types

Best for

High-throughput vector search where metadata filtering is not required and GPU acceleration helps.

Inputs
  • · dense vectors
  • · optional query vectors
  • · index type (Flat/IVF/HNSW/PQ)
Outputs
  • · indices
  • · distances
Requires
  • · faiss-cpu or faiss-gpu
  • · numpy
Preconditions

Vectors as float32 numpy arrays; dimensionality must be consistent

Failure modes

Wrong index type causes poor speed/accuracy tradeoff; training omitted crashes on untrained IVF

Trust signals
  • · 31.7k GitHub stars
  • · Meta/Facebook AI Research
  • · C++/Python bindings