Mini-service for retrieving embedding vectors and similarity computation. An example can be found in the Embedding service API notebook.
On its starting the service requires Nexus authentication, in order to fetch available models. Therefore, before launching the service, get your token from Nexus staging and export it to an environment variable as follows:
export NEXUS_TOKEN=<your_token>
Simply execute
flask run
First, build the docker image:
docker build . -f services/embedder/Dockerfile
Then, run the container by passing your nexus token env variable using the --env parameter.
docker run --env NEXUS_TOKEN=$NEXUS_TOKEN -p 5000:5000 <image>