Run Our Pipeline
Now it’s time to run our pipeline. To do this:
- Open the Kale Deployment panel, by clicking on the Kale icon in your Jupyter notebook environment.
- Click the pull-down menu for Select experiment and create a new experiment called car-price.
- Enter “predict-car-price” as the pipeline name.
- Click the COMPILE AND RUN button.
Once the pipeline is running, view the run by clicking the View link.
This will open a panel to enable you to view the complete pipeline graph as the pipeline executes.
Note that, as expected, training and evaluation for our models run in parallel.
The steps in the pipeline are clickable and provide detailed information about that step. Most of the detail view for a step is outside the scope of this module, but let’s click on the output step and view the Logs tab.
If we zoom in, we can see the output produced by the output step reporting on the prediction performance of all three of our models.
Feel free to explore other output tabs and other aspects of the pipeline run. We’ll address everything you see here in upcoming training modules.