Raise on failed inference endpoint in InferenceEndpoint.wait()

[Wauplin]

Originally from @philschmid on slack (internal):

i am working on a IE example with the new version of huggingface_hub which supports custom container image configs. I noticed that .wait will not break if the endpoint failed.
I made a mistake an the endpoint is in the "failed" status through the UI but the endpoint.wait() is still "blocking".

We now raise an error if the inference endpoint gets into a failed state instead of waiting forever.

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.

[philschmid]

LGTM!

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (7c2aac6) 82.44% compared to head (456159f) 82.42%.
Report is 1 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1935      +/-   ##
==========================================
- Coverage   82.44%   82.42%   -0.03%     
==========================================
  Files          66       66              
  Lines        8138     8140       +2     
==========================================
  Hits         6709     6709              
- Misses       1429     1431       +2     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[Wauplin]

Thanks for reviewing!