Run Jupyter Notebooks on Red Hat OpenShift Data Science

Run Jupyter Notebooks on Red Hat OpenShift Data Science - WIP

Prerequisites

• Configure RHODS workspace has been completed.

Run Jupyter notebooks

1. prepare_env.ipynb
2. Train model.ipynb
3. Test_Model.ipynb
4. Store_Model.ipynb
   Change the minio vaules and run the notebook

Create pipelines with Tekton integration

Create a file called demo.pipeline

Click on Pipeline Editor Rename pipeline to demo pipeline

Configure Pipeline Runtime

• Data Science Pipelines API Endpoint*
  • https://ds-pipeline-pipelines-definition-mltesting.apps.mltesting.example.com
• Public Data Science Pipelines API Endpoint
  • https://rhods-dashboard-redhat-ods-applications.apps.mltesting.sandbox2831.opentlc.com/pipelineRuns/mltesting
• Data Science Pipelines API Endpoint Password Or Token
  • OpenShift API token
• Cloud Object Storage Endpoint*
  • https://minio-api-mltesting.apps.mltesting.sandbox2831.opentlc.com
• Cloud Object Storage Bucket Name*
  • mybucket
• Cloud Object Storage Credentials Secret
  • ` aws-connection-mybucket`
• Cloud Object Storage Username
  • minio
• Cloud Object Storage Password
  • minio123

The data volume for each job is below

• Mount Path* /opt/app-root/src/
• Persistent Volume Claim Name* pins-workspace

Optional: Add GPU information

• GPU: 1
• GPU Vendor: nvidia.com/gpu

1. prepare_env.ipynb
2. Train model.ipynb
3. Test_Model.ipynb
4. Store_Model.ipynb
   

Final View of the pipeline

Start pipeline run

You should see this screen on scuessful run