c34c426441
|
||
|---|---|---|
| .. | ||
| README.md | ||
| artifact_location.py | ||
| condition.py | ||
| exit_handler.py | ||
| immediate_value.py | ||
| parallel_join.py | ||
| recursion.py | ||
| retry.py | ||
| sequential.py | ||
| sidecar.py | ||
README.md
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 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.txtandgs://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.