mirror of
https://github.com/facebookresearch/faiss.git
synced 2025-06-03 21:54:02 +08:00
6d0bc58db6
Summary: IndexPQ and IndexIVFPQ implementations with AVX shuffle instructions. The training and computing of the codes does not change wrt. the original PQ versions but the code layout is "packed" so that it can be used efficiently by the SIMD computation kernels. The main changes are: - new IndexPQFastScan and IndexIVFPQFastScan objects - simdib.h for an abstraction above the AVX2 intrinsics - BlockInvertedLists for invlists that are 32-byte aligned and where codes are not sequential - pq4_fast_scan.h/.cpp: for packing codes and look-up tables + optmized distance comptuation kernels - simd_result_hander.h: SIMD version of result collection in heaps / reservoirs Misc changes: - added contrib.inspect_tools to access fields in C++ objects - moved .h and .cpp code for inverted lists to an invlists/ subdirectory, and made a .h/.cpp for InvertedListsIOHook - added a new inverted lists type with 32-byte aligned codes (for consumption by SIMD) - moved Windows-specific intrinsics to platfrom_macros.h Pull Request resolved: https://github.com/facebookresearch/faiss/pull/1542 Test Plan: ``` buck test mode/opt -j 4 //faiss/tests/:test_fast_scan_ivf //faiss/tests/:test_fast_scan buck test mode/opt //faiss/manifold/... ``` Reviewed By: wickedfoo Differential Revision: D25175439 Pulled By: mdouze fbshipit-source-id: ad1a40c0df8c10f4b364bdec7172e43d71b56c34
92 lines
1.9 KiB
Python
92 lines
1.9 KiB
Python
# Copyright (c) Facebook, Inc. and its affiliates.
|
|
#
|
|
# This source code is licensed under the MIT license found in the
|
|
# LICENSE file in the root directory of this source tree.
|
|
|
|
import os
|
|
import time
|
|
import numpy as np
|
|
import pdb
|
|
|
|
import faiss
|
|
from datasets import load_sift1M, evaluate
|
|
|
|
|
|
print("load data")
|
|
|
|
xb, xq, xt, gt = load_sift1M()
|
|
nq, d = xq.shape
|
|
|
|
# we need only a StandardGpuResources per GPU
|
|
res = faiss.StandardGpuResources()
|
|
|
|
|
|
#################################################################
|
|
# Exact search experiment
|
|
#################################################################
|
|
|
|
print("============ Exact search")
|
|
|
|
flat_config = faiss.GpuIndexFlatConfig()
|
|
flat_config.device = 0
|
|
|
|
index = faiss.GpuIndexFlatL2(res, d, flat_config)
|
|
|
|
print("add vectors to index")
|
|
|
|
index.add(xb)
|
|
|
|
print("warmup")
|
|
|
|
index.search(xq, 123)
|
|
|
|
print("benchmark")
|
|
|
|
for lk in range(11):
|
|
k = 1 << lk
|
|
t, r = evaluate(index, xq, gt, k)
|
|
|
|
# the recall should be 1 at all times
|
|
print("k=%d %.3f ms, R@1 %.4f" % (k, t, r[1]))
|
|
|
|
|
|
#################################################################
|
|
# Approximate search experiment
|
|
#################################################################
|
|
|
|
print("============ Approximate search")
|
|
|
|
index = faiss.index_factory(d, "IVF4096,PQ64")
|
|
|
|
# faster, uses more memory
|
|
# index = faiss.index_factory(d, "IVF16384,Flat")
|
|
|
|
co = faiss.GpuClonerOptions()
|
|
|
|
# here we are using a 64-byte PQ, so we must set the lookup tables to
|
|
# 16 bit float (this is due to the limited temporary memory).
|
|
co.useFloat16 = True
|
|
|
|
index = faiss.index_cpu_to_gpu(res, 0, index, co)
|
|
|
|
print("train")
|
|
|
|
index.train(xt)
|
|
|
|
print("add vectors to index")
|
|
|
|
index.add(xb)
|
|
|
|
print("warmup")
|
|
|
|
index.search(xq, 123)
|
|
|
|
print("benchmark")
|
|
|
|
for lnprobe in range(10):
|
|
nprobe = 1 << lnprobe
|
|
index.setNumProbes(nprobe)
|
|
t, r = evaluate(index, xq, gt, 100)
|
|
|
|
print("nprobe=%4d %.3f ms recalls= %.4f %.4f %.4f" % (nprobe, t, r[1], r[10], r[100]))
|