These instructions will get you a copy of the project up and running on your local machine for development and testing purposes. See deployment for notes on how to deploy the project on a live system.
To build, deploy and run the image you´ll need to have:
- docker with nvidia-docker
- zip (unzip)
To build the docker container, simply run the supplied build script
For local testing, install the following dependencies manually:
pip3 install scikit-learn sklearn torch requests pip3 install --upgrade https://github.com/Theano/Theano/archive/master.zip pip3 install --upgrade https://github.com/Lasagne/Lasagne/archive/master.zip
To execute the test for the core functionalities run
python3 -m unittest tests/test_evaluaterestencoder.py
This image is meant to be run as a Kubernetes job. The configuration (encoder url, parameter, …) has to be done by environment variables. See
run_k8sjob.sh for details. Upon completion, the results are saved as JSON in the corresponding results folder mount.
An example of how to deploy the image onto the datexis registry can be found in
deploy.sh. Don´t forget to login into the registry before deployment and change the namespace according to your configuration.
To start a job on the K8s cluster, change the namespaces within
k8s-senteval-jobs.yaml and run
kubectl create -f k8s-senteval-job.yaml
run_local.sh for an example.