autofaiss reference documentation

Creating and tuning quantized KNN indices for efficient similarity search with low memory usage.

Introduction and installation

• Introduction
• Installing the requirements

Getting started with the CLI

• Creating an index
  • The use-case
  • The build_index command
  • Time required
• Tuning an existing index
  • The use-case
  • The tune_index command
  • Parameters
  • Returns
  • Time required
  • What it does behind
• Getting scores on an index
  • The use-case
  • The score command
  • Parameters
  • Time required
• Creating partitioned indexes
  • The use-case
  • The build_partitioned_indexes command
  • What it does behind

Use autofaiss from python

• Python documentation
  • Quantize
• autofaiss
  • autofaiss package

Getting started with notebooks

• Autofaiss getting started
  • Information
  • Parameters
  • Embeddings creation (add your own embeddings here)
  • Save your embeddings on the disk
  • Build the KNN index with Autofaiss
  • Load the index and play with it
  • (Bonus) Python version of the CLI
• Autofaiss Index selection demo slider

Indices and tables

• Index

• Module Index

• Search Page