pipelines/samples

The sample pipelines give you a quick start to building and deploying machine learning pipelines with Kubeflow.

• Follow the guide to deploy the Kubeflow pipelines service.
• Build and deploy your pipeline using the provided samples.

This page tells you how to use the basic sample pipelines contained in the repo.

Compile the pipeline specification

Follow the guide to building a pipeline to install the Kubeflow Pipelines SDK and compile the sample Python into a workflow specification. The specification takes the form of a YAML file compressed into a .tar.gz file.

For convenience, you can download a pre-compiled, compressed YAML file containing the specification of the core/sequential.py pipeline. This saves you the steps required to compile and compress the pipeline specification: sequential.tar.gz

Deploy

Open the Kubeflow pipelines UI, and follow the prompts to create a new pipeline and upload the generated workflow specification, my-pipeline.tar.gz (example: sequential.tar.gz).

Run

Follow the pipeline UI to create pipeline runs.

Useful parameter values:

• For the "exit_handler" and "sequential" samples: gs://ml-pipeline-playground/shakespeare1.txt
• For the "parallel_join" sample: gs://ml-pipeline-playground/shakespeare1.txt and gs://ml-pipeline-playground/shakespeare2.txt

Components source

All samples use pre-built components. The command to run for each container is built into the pipeline file.