-
Notifications
You must be signed in to change notification settings - Fork 3.7k
Related projects
This page references other projects and repositories that use Faiss or that are related to it. They are unaffiliated the main Faiss, so any issue with those should be redirected to the respective packages. If you did something interesting with Faiss and you want to add a link to your project, just ask us (via an issue).
There are projects to interface Faiss with other languages, see
-
the Rust bindings at https://github.com/Enet4/faiss-rs
-
the .NET Core 3.0 bindings at https://gitlab.com/josetruyol/faisssharp
-
the Java interface at https://github.com/gameofdimension/jni-faiss
Other platforms:
-
Faiss is reported to work on Raspberry Pi (see https://github.com/facebookresearch/faiss/issues/1071)
Other packages:
-
Faiss C++ on CPU is available via a Brew package: https://github.com/Homebrew/homebrew-core/pull/49235
-
Faiss is reported to work on Google collab: https://stackoverflow.com/questions/47967252/installing-faiss-on-google-colaboratory
-
QuickerADC, an implementation of IVFPQ variants using SIMD with better precision/speed operating points: https://github.com/technicolor-research/faiss-quickeradc See also the paper Quicker ADC : Unlocking the hidden potential of Product Quantization with SIMD, André et al.
-
pgANN: Approximate Nearest Neighbor (ANN) searches with a PostgreSQL database. https://github.com/netrasys/pgANN
-
milvus offers a uniform interface to Faiss + other packages and manages it like a database engine (client/server and the like): https://github.com/milvus-io/milvus
-
vearch: a search engine based on Faiss with good scalability and real-time updates: https://github.com/vearch/vearch
-
port of (an old version of) Faiss to Windows: https://github.com/bitsun/faiss-windows
There are many approximate nearest neighbor search libraries. You may be interested in:
-
NMSLib (An efficient similarity search library and a toolkit for evaluation of k-NN methods for generic non-metric spaces): https://github.com/nmslib/nmslib
-
an implementation of IVF-HNSW https://github.com/dbaranchuk/ivf-hnsw
-
Annoy (Approximate Nearest Neighbors in C++/Python): https://github.com/spotify/annoy
-
Approximate nearest neighbor benchmark page: https://github.com/erikbern/ann-benchmarks
-
A promising GPU index here (vaporware as of 2020-01-20): https://github.com/cgtuebingen/ggnn
Faiss building blocks: clustering, PCA, quantization
Index IO, cloning and hyper parameter tuning
Threads and asynchronous calls
Inverted list objects and scanners
Indexes that do not fit in RAM
Brute force search without an index
Fast accumulation of PQ and AQ codes (FastScan)
Setting search parameters for one query
Binary hashing index benchmark