Create huggingface integration test example for v0.12.1

[misohu]

Context

With v0.12.1 huggingface ClusterServingRuntime was introduced. We should add integration test for this runtime.

What needs to get done

Create example yaml for huggingfaceserver serving runtime. Similar to other examples.

Definition of Done

1. huggingfaceserver example is part of integration tests.

[syncronize-issues-to-jira]

Thank you for reporting us your feedback!

The internal ticket has been created: https://warthogs.atlassian.net/browse/KF-5727.

This message was autogenerated

[misohu]

First I followed the docs from kserve documentation and I have tried to deploy the llama3 with huggingface inference service. This model tho requires 25GBs of space + a GPU.

After that I played around and found smaller model on hugging face which I successfully executed locally. Here is the PR #298. Problem with this test is that in order to run the test we need todeploy a pod with huggingface image to the runner which is almost 5 GB big (almost 8GB for the rock). This is problem for the runner as the image cant fit.