Skip to content

Get Dataset and Code

To work through this module you will need the code and data we have provided. If you are not familiar with Kale we strongly recommend you complete Kale 101. If you do not have the completed notebook available you may download and unzip the handout.

Upload the handout files

Once you’ve unzipped the handout, you should see the following files.

1. Review the handout files


car_prices.csv is our data file.

data_dictionary-carprices.xlsx contains examples images for the notebook and will need to be treated as a separate folder.

predict_car_price_katib.ipynb is a notebook containing Python code that builds and evaluates three models for predicting car prices based on our dataset. We will build on the modified and annoted code from Kale 101 to define and run Katib Experiments in Kubeflow Pipelines!

requirements.txt lists the Python modules required for our notebook. We'll use this file to install those requirements in a later step.

2. Open the learn-katib-pipelines-vol-1 folder

Double-click on the directory, learn-katib-pipelines-vol-1.

vol 1 directory

3. Click the file upload button

upload files

4. Upload handout files

In the file dialog that pops up, select the three handout files you unzipped and upload them to your Jupyter notebook environment.

select handout files

You will see them appear in the learn-katib-pipelines-vol-1 directory.

vol 1 directory

5. Create a new folder

Click the button to create a new folder.

create folder

6. Name the folder "data"

name the folder

7. Move data files

Drag and drop car_prices.csv and data_dictionary-carprices.xlsx into the data folder.

move data

8. Open our notebook

Double-click predict_car_price.ipynb in the file browser pane.

open our notebook

9. Enable Kale

Click the Enable toggle in the Kale Deployment panel to enable Kale.

enable Kale