Set up your MLOps platform on your Kubernetes cluster
This repository hold the base code to run your MLOps platform on your Kubernetes cluster with MLflow.
A quay.io repository exists to hold the container image of this project.
- Kubernetes cluster
- MinIO cluster (optional)
Ensure that you modified the manifests within the kubernetes folder to match your setup.
If you want a production-ready container through Docker run the following command :
#!/bin/bash
docker run -d\
-P\
-v "$(pwd)/scripts:/usr/src/app/scripts"\
quay.io/rafidini/k8s-mlflow:1.0.0-alpha.5Get into a Command prompt/Terminal and follow the instructions :
# (optional) Launch MinIO
kubectl apply -f kubernetes/minio-standalone.yaml
# Launch MLflow
kubectl apply -f kubernetes/k8s-mlflow.yamlA guideline will be produced later on.
Note
To update the changelog
git-changelog -o CHANGELOG.md -s angular .