Deploy Using HPE Machine Learning Development Environment Software (MLDES) Managed Service#

Attention

MLDES only supports Determined Enterprise Edition.

Overview#

You can deploy a Determined cluster using MLDES. MLDES is a managed service that simplifies the infrastructure management complexities. It offers an intuitive interface for deploying, managing, and utilizing Determined clusters.

Prerequisites#

MLDES supports both AWS and GCP as backends for Determined cluster resources. To start, ensure you have an AWS account with the required IAM roles or a GCP project ID. For more information, visit the documentation.

Deployment Guide#

Access the Console#

  1. Visit the MLDES console.

  2. Sign in using your HPE Passport account.

Create an Organization#

  1. Select Organization to open the Organizations panel and create a new organization.

  2. Enter a unique name and select Save.

Deploy the Cluster#

  1. Select Clusters to open the Clusters panel.

  2. Select New MLDE Cluster to start creating a new cluster.

  3. Enter a unique name.

  4. Review the default cluster and resource pool configuration and make any necessary changes.

  5. Select Create Cluster.

MLDE provisions the cluster and lets you know when it is ready.

Create and Manage Experiments#

  1. Once the cluster is ready, select Open.

  2. To open a Jupyter notebook, Launch JupyterLab.