Add AVX512 support to k-nn FAISS#2069

Closed

akashsha1 wants to merge 76 commits intoopensearch-project:mainfrom

akashsha1:intel/avx512faiss

Commits on Aug 8, 2024

changes for avx512
root
committed

Commits on Aug 20, 2024

minor edits to avx512 check
root
committed

Commits on Aug 23, 2024

some more changes
root
committed

Commits on Aug 28, 2024

Commits on Sep 7, 2024

Commits on Sep 10, 2024

Commits on Sep 13, 2024

Commits on Sep 16, 2024

Update scripts/build.sh

akashsha1
and
naveentatikonda
authored
changes for avx512

root
authored and
akashsha1
committed
minor edits to avx512 check

root
authored and
akashsha1
committed
some more changes

root
authored and
akashsha1
committed
fixes
akashsha1
committed
more fixes
akashsha1
committed
Modify faiss submodule to point to local repo.

mulugetam
authored and
akashsha1
committed
Fix avx-512 flag check.

mulugetam
authored and
akashsha1
committed
Fix formatting issue.

mulugetam
authored and
akashsha1
committed
Refactor Around Mapper and Mapping (#1939)

jmazanec15
authored and
akashsha1
committed
Integrate Lucene Vector field with native engines to use KNNVectorFormat during segment creation (#1945)

navneet1v
authored and
akashsha1
committed
Disallow invalid characters for physical file name to be included within vector field name. (#1936)

0ctopus13prime
authored and
akashsha1
committed
Quantization Framework Implementation with 1bit and MultiBit Binary Quantizer (#1929)

Vikasht34
authored and
akashsha1
committed
Integrate KNNVectorValues with vector ANN Search flow (#1952)

navneet1v
authored and
akashsha1
committed
BackPort Java Doc Fix with Code Improvements (#1959)

Vikasht34
authored and
akashsha1
committed
Quantization Framework Code Structure Improvement (#1967)

Vikasht34
authored and
akashsha1
committed
Adds rescore parameter to KNNQuery (#1969)

jmazanec15
authored and
akashsha1
committed
Encapsulate dimension, vector data type validation/processing inside Library (#1957)

jmazanec15
authored and
akashsha1
committed
Fix bwc version for rescore (#1975)

jmazanec15
authored and
akashsha1
committed
Fixing the dimension for the vector when using Lucene field in ModelFieldMapper (#1980)

navneet1v
authored and
akashsha1
committed
Disables rewrite code path to debug latency issues (#1987)

shatejas
authored and
akashsha1
committed
Add quantization state cache (#1960)

ryanbogan
authored and
akashsha1
committed
Integrates FAISS iterative builds with NativeEngines990KnnVectorsFormat (#1950)

authored and
akashsha1
committed
Add support for qframework from interface (#1994)

jmazanec15
authored and
akashsha1
committed
Fix string for qframe parser key (#1999)

jmazanec15
authored and
akashsha1
committed
k-NN query rescore support for native engines (#1984)

jmazanec15
authored and
akashsha1
committed
Add HNSW changes to support Faiss byte vector (#1823)

naveentatikonda
authored and
akashsha1
committed
Integration of Quantization Framework for Binary Quantization with Indexing Flow (#1996)

Vikasht34
authored and
akashsha1
committed
Parallelize make to reduce build time (#2006)

naveentatikonda
authored and
akashsha1
committed
Align dimensions to the nearest multiple of 8 in QuantizationState (#2010)

authored and
akashsha1
committed
Changes NativeEngineKNNQuery to execute search and rescore in (#2014)

shatejas
authored and
akashsha1
committed
Add IVF changes to support Faiss byte vector (#2002)

naveentatikonda
authored and
akashsha1
committed
Makes NativeEngines990KnnVectorFormat as default for index version (#2020)

shatejas
authored and
akashsha1
committed
Adds graph build time metrics in NativeEngines990KnnVectorsWriter (#2018)

shatejas
authored and
akashsha1
committed
Fixing the format name for NativeEngines990KnnVectorsFormat (#2025)

navneet1v
authored and
akashsha1
committed
Introduce compression and mode mapping parms (#2019)

jmazanec15
authored and
akashsha1
committed
Fix memory overflow caused by cache behavior (#2015)

kotwanikunal
authored and
akashsha1
committed
Add release notes for 2.17 (#2037)

naveentatikonda
authored and
akashsha1
committed
Add quantization state reader and writer (#1997)

ryanbogan
authored and
akashsha1
committed
Introduce mode and compression param resolution (#2034)

jmazanec15
authored and
akashsha1
committed
Fix tests related to quantization state (#2045)

ryanbogan
authored and
akashsha1
committed
Add spaceType as a top level parameter while creating vector field. (#2044)

authored and
akashsha1
committed
Fix the force merge with Quantization failures when a segment has deleted docs in it (#2046)

navneet1v
authored and
akashsha1
committed
Add model version to model metadata and change model metadata reads to be from cluster metadata (#2005)

authored and
akashsha1
committed
Add tests for top level spaceType parameter with different combinations (#2058)

navneet1v
authored and
akashsha1
committed
Fix expected string for get model response tests (#2062)

ryanbogan
authored and
akashsha1
committed
Throws and exception for radial search when mapping is for on-disk mode (#2055)

shatejas
authored and
akashsha1
committed
few more fixes
akashsha1
committed
few more fixes
akashsha1
committed
address CR feedback
akashsha1
committed
fix spotless check
akashsha1
committed
Update CHANGELOG.md

akashsha1
and
naveentatikonda
committed
update build script per CR feedback
akashsha1
committed
update CI script
akashsha1
committed
update dev guide and build script
akashsha1
committed
DCO check Signed-off by Akash Shankaran <[email protected]>
akashsha1
committed
Update scripts/build.sh

akashsha1
and
naveentatikonda
committed
Merge branch 'intel/avx512faiss' of https://github.com/akashsha1/k-NN into intel/avx512faiss
akashsha1
committed