463 Commits

Author SHA1 Message Date
Matthijs Douze
1a332509a3
Update 5-Multiple-GPUs.cpp
Thanks to gameofdimension
https://github.com/facebookresearch/faiss/pull/1060/files
2020-01-16 16:46:11 +01:00
glutamatt
d278355b9c Fix typo in Clustering_c.h (#1084) 2020-01-14 15:41:07 +01:00
Matthijs Douze
281c9f7113
fixing typedef
see #1064
2020-01-03 07:19:07 +01:00
Matthijs Douze
c250587c77
Update README.md 2019-12-09 13:32:36 +00:00
Lucas Hosseini
2ba6985f81 Facebook sync 20191129 (#1048)
Looks good!
v1.6.1
2019-12-04 07:21:02 +01:00
Kota Yamaguchi
83a341f300 Replace print with logging (#964) 2019-11-21 11:42:03 +01:00
Lucas Hosseini
da24fcc56e
Use intrinsics for float16 in GPU scalar quantizer. (#987) 2019-11-19 16:15:33 +01:00
mattg-sp
2ab4b4859c Fixes to enable building with -mavx, for Sandybridge CPUs. (#992)
The -mf16c (i.e. packed-half conversion instructions) weren't added until the
Ivy Bridge CPU generation (https://en.wikipedia.org/wiki/F16C#CPUs_with_F16C).
Therefore, code dependent on those intrinsics must be conditioned on __F16C__ -
not __AVX__.
2019-11-19 16:15:00 +01:00
Jose Truyol
c47da8a948 [C_API] Pass compilation in current state (#989)
* Pass compilation in current state

* Fix formatting and add missing parts

* Define DistanceComputer

* Add documentation to name changed on DistanceComputer::operator()()

* Apply suggestions from code review

Changes in docs

Co-Authored-By: Eduardo Pinho <enet4mikeenet@gmail.com>

* Updated MetricType
2019-11-19 11:36:44 +01:00
Matthijs Douze
4c32fa4d14
Update demo_sift1M.cpp 2019-11-19 11:18:47 +01:00
Jin Hai
e325c50e97 fix bugs with setNumProbes in GPUIndexIVF (#1011) 2019-11-10 07:37:06 +01:00
Jeff Johnson
2cce2e5f59
Merge pull request #1006 from milvus-io/1.6.0
Fix GPUSQ copyto bug
2019-10-31 23:57:57 -04:00
JinHai-CN
45639f62fa Fix GPUSQ copyto bug 2019-10-29 16:55:27 +08:00
Matthijs Douze
8bc09261f4
Update README.md 2019-10-20 09:59:18 +02:00
Matthijs Douze
10b5a1333e
Update README.md 2019-10-20 09:58:54 +02:00
Lucas Hosseini
31393764cc Prepare for v1.6.0 release. v1.6.0 2019-09-24 05:28:20 +02:00
Lucas Hosseini
d9220806a2
Fix GPU build. (#959) 2019-09-23 00:01:49 +02:00
Matthijs Douze
cdf9d914b8
Update AutoTune.h 2019-09-21 11:21:45 +02:00
Lucas Hosseini
3d0ece0cb7
Run time detection of avx2 for conda packages. (#957)
* Build conda CPU packages with run-time detection of AVX2.

* Add cudatoolkit 10.1 to conda list.

* Default to sse4 on x86_64.
2019-09-20 20:17:01 +02:00
Lucas Hosseini
36ddba9196
Facebook sync (2019-09-10) (#943)
* Facebook sync (2019-09-10)

* Fix depends Makefile target.

* Add faiss symlink for new include directives.

* Fix missing header.

* Fix tests.

* Fix Makefile.

* Update depend.

* Fix include directives spacing.
2019-09-20 18:59:10 +02:00
Scaled Dragon
8b682606e1 Fix spelling error (#919) 2019-09-02 13:51:07 +02:00
Matthijs Douze
f61e6228ca
Update README.md 2019-08-30 13:49:28 +02:00
Matthijs Douze
c364c2b91c
Update search_server.py 2019-08-29 15:36:20 +02:00
Matthijs Douze
983a1f6b8b
Update run_on_cluster.bash 2019-08-29 15:36:07 +02:00
Matthijs Douze
d9a01b2d5c
Update rpc.py 2019-08-29 15:35:53 +02:00
Matthijs Douze
64828d2851
Update merge_to_ondisk.py 2019-08-29 15:35:39 +02:00
Matthijs Douze
0fdd38055b
Update make_trained_index.py 2019-08-29 15:35:21 +02:00
Matthijs Douze
29e0514128
Update make_index_vslice.py 2019-08-29 15:35:07 +02:00
Matthijs Douze
87c83b9d97
Update distributed_query_demo.py 2019-08-29 15:34:50 +02:00
Matthijs Douze
a9a475b003
Update distributed_kmeans.py 2019-08-29 15:34:14 +02:00
Matthijs Douze
10ca6e20d4
Update combined_index.py 2019-08-29 15:33:50 +02:00
Matthijs Douze
8d08912453
Ondisk distributed index implementation (#930)
Adds the code for the distributed on-disk index
2019-08-29 13:44:08 +02:00
Lucas Hosseini
656368b5ed
Dynamically detect avx2 capabilities (#877)
* Install scipy via pip rather than system package manager

* Dynamically detect AVX2 capability

* Simplify build on TravisCI
2019-06-28 13:50:46 +02:00
Matthijs Douze
70f5b04206
Update CONTRIBUTING.md 2019-06-25 17:27:23 +02:00
Lucas Hosseini
0ca46ee474
Update README.md 2019-06-24 18:51:00 +02:00
Lucas Hosseini
efa81eec76 Prepare for v1.5.3 release. v1.5.3 2019-06-24 17:41:35 +02:00
Eduardo Pinho
828dcea400 [C API] June 2019 update (#869)
* [c_api] use all relevant flags in compilation

* [c_api] Remove redundant IndexIVFFlat declarations

- From IndexIVF_c.h, already declared in IndexIVFFlat_c.h

* [c_api] type changes

- replace `long` with a more suitable type
- provide definitions for `faiss_component_t` and `faiss_distance_t`

* [c_api] Define CFLAGS and CUDACFLAGS
2019-06-24 13:29:38 +02:00
Lucas Hosseini
1cbc999610
Add -mavx2 -mf16c to CPUFLAGS. (#871) 2019-06-24 13:28:58 +02:00
chck
fc9856a25e Fix gpu compilation with cuda 10.1 related to #751 (#865) 2019-06-21 19:43:53 +02:00
Lucas Hosseini
d224d11c9d
Revert swigfaiss import for python 3 compatibility. (#868) 2019-06-20 06:13:38 +02:00
Lucas Hosseini
3896b12c65
Facebook sync (Jun 2019) (#862)
Bugfixes:
- slow scanning of inverted lists (#836).

Features:
- add basic support for 6 new metrics in CPU `IndexFlat` and `IndexHNSW` (#848);
- add support for `IndexIDMap`/`IndexIDMap2` with binary indexes (#780).

Misc:
- throw python exception for OOM (#758);
- make `DistanceComputer` available for all random access indexes;
- gradually moving from `long` to `int64_t` for portability.
2019-06-19 15:59:06 +02:00
Eduardo Pinho
2bfbead8f1 [C API] Update for compatibility with v1.5 (#858)
* [c_api] Update impl and interface for v1.5

- move IndexShards to dedicated module IndexShards_c.{h|cpp}
- remove getter/setters to unreachable fields
- reimplement faiss_IndexIVF_imbalance_factor (to use invlists)
- minor IndexIVF documentation tweaks
- Remove QueryResult, provide RangeQueryResult

* [c_api] Document FaissErrorCode

* [c_api] Update GPU impl and interface for v1.5

- Remove unavailable method setTempMemoryFraction

* [c_api] Relicense to MIT

In accordance to the rest of the project
2019-06-12 15:46:08 +02:00
Gao
c4c5d6f73c Update README.md (#849)
fix typo
2019-05-31 21:02:17 +02:00
Lucas Hosseini
ae09cabd6a
Fix Makefile in gpu/test (#843) 2019-05-29 14:32:05 +02:00
Lucas Hosseini
dacd5a9f51 Prepare for v1.5.2 release v1.5.2 2019-05-28 16:22:38 +02:00
Lucas Hosseini
a8118acbc5
Facebook sync (May 2019) + relicense (#838)
Changelog:

- changed license: BSD+Patents -> MIT
- propagates exceptions raised in sub-indexes of IndexShards and IndexReplicas
- support for searching several inverted lists in parallel (parallel_mode != 0)
- better support for PQ codes where nbit != 8 or 16
- IVFSpectralHash implementation: spectral hash codes inside an IVF
- 6-bit per component scalar quantizer (4 and 8 bit were already supported)
- combinations of inverted lists: HStackInvertedLists and VStackInvertedLists
- configurable number of threads for OnDiskInvertedLists prefetching (including 0=no prefetch)
- more test and demo code compatible with Python 3 (print with parentheses)
- refactored benchmark code: data loading is now in a single file
2019-05-28 16:17:22 +02:00
gensmusic
712edb043a tutorial build add CPPFLAGS which contains cuda include pathi when ./configure --with-cuda (#833) 2019-05-24 14:46:53 +02:00
Pierre Letessier
5d1ed5b6fb Fix unable to read untrained index (#817) 2019-05-07 18:14:15 +02:00
Lucas Hosseini
012954ebbf
Fix HNSW.fill_with_random_links. (#801) 2019-04-23 16:52:39 +02:00
Xuhui Sun
0a8f92d6d6 Update cpp tutorial makefile for GPU part (#774) 2019-04-11 14:20:38 +02:00