Skip to content

AutoML

Please review the overview section to understand details about the end-to-end Katib Experiment you will implement in the steps below.


Step 1: Login

In this step, you will login to your MLOps Platform.

  • Navigate to the URL (This will be provided by your platform team)
  • Login using your local credentials or SSO credentials (Identity Provider such as Okta)

Login

Once logged in, you will see the home dashboard screen.

Dashboard


Step 2: Create a Notebook

In this step, you will create a Jupyter Notebook. The notebook will be used to create a Katib experiment.

  • Navigate to Notebooks
  • Click New Notebook
  • Enter a name for the notebook
  • Select JupyterLab
  • Set the minimum CPU to .5
  • Set the minimum memory to 1
  • Click Launch

Launch

It will take 1-2 minutes to create the notebook.

Launch


Step 3: Generate Experiment

In this step, you will use the notebook to create a Katib experiment which will evaluate function parameters to maximize the function value.

Download Notebook

  • Navigate to Notebooks
  • Click Connect on the previously created notebook
  • Download the following notebook file
  • In the left hand folder tree, click on the upload files icon
  • Upload the previously downloaded katib.ipynb file
  • Double click the katib.ipynb file in the folder tree to open the notebook
  • Select the first cell and click the run icon

Pipeline

  • Once the first cell has finished running, click the Restart Kernel and Run All Cells icon
  • After ~2 minutes, the experiment will be complete

Pipeline


Step 4: View AutoML Results

In this step, we will view the results of the Katib experiment within the Katib UI to view the trials and results.

  • Navigate back to the Kubeflow dashboard
  • Click Experiments (AutoML)
  • Click on the experiment name

You will see the parameters that delivered the best result.

Experiment

  • Click on Trials to see the parameters and results of each trials

Experiment


Recap

Congratulations! At this point, you have successfully created a Jupyter notebook to create a Katib experiment for hyperparameter optimization.