How to utilize IndexLSH for bucketization feature and return the hash code?

[chenyuanTKCY]

Summary

After indexing the data with IndexLSH, how can we determine the exact location of each data point, specifically the hash bucket number, to facilitate subsequent pruning? This is crucial for optimizing the search process, as identifying the hash bucket in which a data point resides allows us to efficiently narrow down candidate points during query time, reducing unnecessary computations and improving the overall performance of the algorithm.

Platform

OS:

• Operating System: Ubuntu 20.04
• Kernel Version: 5.15.0-101-generic
• System Architecture: x86_64 (64-bit)

Faiss version: faiss-gpu 1.7.3

Installed from: anaconda

Faiss compilation options:

Running on:

• CPU
• GPU

Interface:

• C++
• Python

Reproduction instructions