search

weaviate / contextionary

Weaviate's own language vectorizer, which allows for semantic context-based searches in Weaviate
https://weaviate.io/developers/weaviate/modules/retriever-vectorizer-modules/text2vec-contextionary
BSD 3-Clause "New" or "Revised" License
14 stars 2 forks source link
machine-learning vectorizer weaviate

readme

Weaviate Contextionary Weaviate logo

The contextionary powers the semantic, context-based searches in Weaviate.

Not intended for stand-alone use. Used by Weaviate - the ML-first vector search engine.

Versioning

The version tag is <language-of-db><semver-of-db>-v<semver-of-app>. So for example the app version 0.1.0 deployed with the contextionary vector db version 0.6.0 of the English language will have the version en0.6.0-v0.1.0. This also corresponds to the Docker tag.

Languages

Currently available languages include:

Other languages coming soon.

Docker Requirements

The build pipeline makes use of Docker's buildx for multi-arch builds. Make sure you run a Docker version which supports buildx and have run docker buildx create --use at least once.

How to build and test project

  1. Regenerate schema:
./gen_proto_code.sh
  1. Build image:
LANGUAGE=en MODEL_VERSION=0.16.0 ./build.sh
  1. Run journey tests:
LANGUAGE=en MODEL_VERSION=0.16.0 ./build.sh && DIMENSIONS=300 ./test/journey.sh