* WIP Initial commit of refactored notebooks section.
* Minor additions.
* Finished draft of notebooks quickstart.
* Renamed some files and clarified the notebooks setup doc.
* Updated the guide to custom notebook images.
* A couple more screenshots plus text refinements.
* Fixed references to JupyterHub throughout the docs.
* Updated in response to review comments.
* One more link fix.
* Clarified auth in custom notebooks doc.
* WIP Initial commit of component dev tutorial.
* Further textual improvements.
* Further improvements.
* Further improvements.
* Added the best practices.
* Fixed a link.
* Tidied up the code comments.
* More code comment fixes.
* WIP Started an intro to the Pipelines DSL.
* Added more content.
* Responded to review comments.
* Changed function formatting to be parallel with previous example.
Most recent version of miniconda installs pip, not pip3 with Python 3.7. When using GCP Cloud Shell, pip3 is in /usr/bin and will cause kfp package to end up in the users site packages directory and not in the mlpipelines env.
As I understand it, pip3 is the Python3 version of pip. On systems with both Python 2 and 3, pip will use Python 2. This affects hosted notebook environments that have both Python 2 and 3 installed, such as Kubeflow notebooks or Google Cloud Notebooks. Since the SDK requires Python 3, I'd recommend changing this to be pip3.
* Sets apart component output for each component.
- Pulls out the component output from each component to make it a tiny bit easier to find with a quick scan.
- Makes a guess at what the component output for the BigQuery query might be (needs verification).
- Removes the **Component:** preceding each component -- this seems redundant, unless other non-component items will be added to this list. Otherwise, the bolded URL / name gives each component sufficient visual distinction on the page.
* Updated component output description
Updated component output description for BigQuery query, per its README:
https://github.com/kubeflow/pipelines/tree/master/components/gcp/bigquery/query#output
* Restructured the pipelines docs.
* Added redirects for the deleted files.
* Standardized use of full stop in descriptions.
* More full stops.
* Fixed internal links to the moved pages.
* First commit of pipelines tutorial.
* Still in progress: Pipelines end-to-end guide.
* Finished first run of tutorial.
* A few more tidy-ups.
* Set back all the page weights to values in head.
* Reverted change to gpu/tpu line containing page weight.
* Added the finishing touches to the tutorial.
* Updated in response to review comments.
* One more change resulting from review comments.
* WIP: Initial commit of refactored component guide.
* Continuing work on the guide to building a pipeline.
* Further updates to the pipeline component guides.
* Further improvements to the component refactoring.
* Updated in response to review comments.
* In progress: Upgrade guide for Kubeflow Pipelines.
* Added more content to the upgrade guide.
* Added a note in the main upgrade guide.
* Tweaked the upgrade guide after testing.
* Better images.
* Fixed a typo in the GKE getting-started guide.
* Updated in response to review comments.
* Great guide on how to use GPU and TPU in KFP
* Correct gpu type in example
* Break lines in code example to make it look nicer
* Add python imports in code examples.
* In progress: Refactoring the left-hand navigation panel.
* Added .description to section index pages.
* Removed the redundant Components Overview as the new index page of the components section does a better job.
* Added redirects for the files I've moved/deleted.
* Fixed the URLs in all doc links.
Sarah Maddox
Jorge
Roy Lenferink
IronPan
chiefkarlin
Ning
Hung-Ting Wen
Joe Liedtke
Crystal Gomes
Yasser Elsayed
Ryan Williams
Yuan (Terry) Tang
Jin Chi He
hongye-sun
Guang Ya Liu
CraigSterrett