How to deploy a model using S2I

Learning path | 1 resource | 20 Minutes | Published on September 9, 2021
Learn how to deploy a model using S2I.

  1. Overview: How to deploy a model using S2I

Overview: How to deploy a model using S2I

Data Science AI/ML
RHODS learning path

When you are using the Source-to-Image (S2I) deployment method, you need to log in to your Red Hat OpenShift Data Science platform and enable applications.

Log in to  Red Hat OpenShift Data Science

  1. Log in to the Red Hat OpenShift Dedicated (OSD) platform.
  2. Gain access to Red Hat OpenShift Data Science by selecting the OpenShift Data Science icon in the upper right corner.
  3. Select the Red Hat OpenShift Data Science menu option (Figure 1).
Launch OpenShift Data Science from the OpenShift Managed Services menu.
Figure 1. Launch OpenShift Data Science from the Red Hat OpenShift Managed Services menu.

Enable applications

  1. Locate the Enabled option in the left panel.
  2. Select Enabled.
  3. Select the JupyterHub managed service, then Launch (Figure 3).
The OpenShift Data Science dashboard.
Figure 3. The OpenShift Data Science dashboard.

A Start a notebook server page will appear (Figure 5).

    Select TensorFlow on the JupyterHub screen and start the server.
    Figure 5. Select TensorFlow on the JupyterHub screen and start the server.

    You are now authenticated and can create your Jupyter notebook. You can also start creating your notebook image. You'll be able to customize your:

    • Notebook image type
    • Deployment size
    • Environment variables

    If this procedure is successful, you have executed the first steps in deploying a model using S2I and Red Hat OpenShift Data Science.