Checkpoint: Pipeline Design
Having completed the lessons in this module, you should now be able to do the following comfortably:
- Identify code in a notebook that implements a discrete step in a machine learning workflow and annotate that cell as a Pipeline Step.
- Identify the data that one step produces as output and the step or steps that depend on that data as input.
- Specify single-step and multi-step dependency relationships between Kubeflow pipeline steps using the Depends on parameter of the Pipeline Step annotation.
- Create pipeline branches that can run in parallel using the Depends on parameter of the Pipeline Step annotation.
- Organize the Python statements that import modules your pipeline steps need into a small number of cells and mark those cells using the Imports annotation.
- Identify cells in a notebook that should be excluded from pipeline runs and annotate them as Skip Cells.
- Organize and annotate Functions cells.