FVTs don't run due to "No space left on device" error

[ckadner]

Describe the bug

https://github.com/kserve/modelmesh-serving/actions/runs/4831653258/jobs/8609373544#step:11:367

failed to register layer: ApplyLayer exit status 1 stdout:  stderr: write /usr/local/lib64/python3.8/site-packages/tensorflow/libtensorflow_cc.so.2: no space left on device
Error: Process completed with exit code 1.

We should investigate what is using up the additional disk space and follow some best practices to reduces disk usage for GitHub action runners: https://github.com/marketplace/actions/maximize-build-disk-space

Or here: actions/runner-images#2840 (comment)

sudo rm -rf /usr/share/dotnet
sudo rm -rf /opt/ghc
sudo rm -rf "/usr/local/share/boost"
sudo rm -rf "$AGENT_TOOLSDIRECTORY"

To Reproduce
Steps to reproduce the behavior:

1. Go to the FVT jobs list
2. Click on any of the most recent failures, e.g. run #768
3. Scroll down to see the build failure
4. See error no space left on device

Expected behavior

FVT run should complete successfully, or at the very least run some of the actuall FVT tests

Additional context

These errors first appeared during the HPA PR checks after the minikube setup was changed

@Jooho @rafvasq @tjohnson31415 @njhill