Unofficial Docker image for FAISS, a library for efficient similarity search
997
FAISS is a library for efficient similarity search and clustering of dense vectors, developed by Facebook Research and MIT licensed. This repo contains inofficial Dockerfiles for building an installing the C++ library as well as the Python bindings.
The container is based on the official installation instructions and uses the the official CUDA docker container nvidia/cuda:11.4.0-cudnn8-devel-ubuntu20.04 as the base image.
Sources available on https://github.com/stes/faiss-docker.
Containers can be pulled using
docker pull stffsc/faiss:latest
docker pull stffsc/faiss:1.7.1-cuda11.4.0-cudnn8-devel-ubuntu20.04
All containers come with python 3.8 installed and FAISS python bindings:
docker run -u $(id -u) --ipc host -it stffsc/faiss:latest python
Python 3.8.10 (default, Jun 2 2021, 10:49:15)
[GCC 9.4.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import faiss
>>> print(faiss.__version__)
1.7.1
Content type
Image
Digest
Size
5.5 GB
Last updated
about 5 years ago
docker pull stffsc/faiss