MLflow server. Key features
- MLflow UI
- MLflow tracking server
- MLflow model registry
- PostgreSQL backend store
- Running on Kubernetes
- One click deployment on staroid.
- How Open Data Studio Jupyter notebook connects MLflow server - step by step instruction how to use with Jupyter notebook and how it works underneath.
Run locally with skaffold command.
$ skaffold dev --port-forward -p minikube
and browse http://localhost:5000
It creates secret mlflow-ssh-key dynamically. Skaffold does not remove this one. Clean up resource with
$ kubectl delete secret mlflow-ssh-key