Add kubeflow-katib and kserve packages to default notebook images

[thesuperzapper]

/kind feature

Currently we don't include the kubeflow-katib and kserve pip packages in the default notebooks. However, we do include the kfp package in the -full suffixed images, for example jupyter-pytorch-cuda-full.

Currently we are blocked by the following upstream issues because kfp has conflicting version dependencies with these packages:

• kubeflow-katib: Python SDK for Katib and Kubeflow Pipelines can not be installed at the same time katib#2346
• kserve: Python SDK for KServe and Kubeflow Pipelines can not be installed at the same time kserve/kserve#3730

[andreyvelich]

Thanks for creating this @thesuperzapper .
kubeflow-katib should be unblocked now since @tenzen-y updated the protobuf version in this PR: kubeflow/katib#2358
The latest version is kubeflow-katib==0.17.0rc1

Also, we might want to include kubeflow-training SDK to the Kubeflow Notebooks as well: https://pypi.org/project/kubeflow-training/

[andreyvelich]

/transfer notebooks

[google-oss-prow]

@thesuperzapper: The label(s) kind/feature cannot be applied, because the repository doesn't have them.

In response to this:

/kind feature

Currently we don't include the kubeflow-katib and kserve pip packages in the default notebooks. However, we do include the kfp package in the -full suffixed images, for example jupyter-pytorch-cuda-full.

Currently we are blocked by the following upstream issues because kfp has conflicting version dependencies with these packages:

• kubeflow-katib: Python SDK for Katib and Kubeflow Pipelines can not be installed at the same time katib#2346
• kserve: Python SDK for KServe and Kubeflow Pipelines can not be installed at the same time kserve/kserve#3730

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.