MLFlow Server dockerised setup using NGINX as basic authentication and PostgreSQL for entity storage (artifacts are stored locally)
Based on: https://github.com/ymym3412/mlflow-docker-compose
echo "{USER_NAME}:$(openssl passwd -apr1 {PASSWORD})" >> ${HOST}
${HOST}
as defined in your .env
file
docker-compose build
docker-compose up -d
docker-compose down
- Include DVC for data versioning
- Include Kubeflow for
- Improve artifact storage
- Make ports more flexible (currently fixed due to
acme-companion
, includingMLFLOW_PORT
in env) - Add GoogleCloud and AWS setups