[WIP] Add e2e test for `tune` api with LLM hyperparameter optimization #1018

Job	Run time
e2e (v1.27.11, long-running-resume,from-volume-resume,median-stop)	11m 44s
e2e (v1.27.11, grid,bayesian-optimization,tpe,multivariate-tpe,cma-es,hyperband)	11m 44s
e2e (v1.27.11, file-metrics-collector,pytorchjob-mnist)	11m 44s
e2e (v1.27.11, median-stop-with-json-format,file-metrics-collector-with-json-format)	11m 43s
e2e (v1.28.7, long-running-resume,from-volume-resume,median-stop)	11m 35s
e2e (v1.28.7, grid,bayesian-optimization,tpe,multivariate-tpe,cma-es,hyperband)	11m 42s
e2e (v1.28.7, file-metrics-collector,pytorchjob-mnist)	11m 43s
e2e (v1.28.7, median-stop-with-json-format,file-metrics-collector-with-json-format)	11m 37s
e2e (v1.29.2, long-running-resume,from-volume-resume,median-stop)	11m 34s
e2e (v1.29.2, grid,bayesian-optimization,tpe,multivariate-tpe,cma-es,hyperband)	11m 36s
e2e (v1.29.2, file-metrics-collector,pytorchjob-mnist)	11m 38s
e2e (v1.29.2, median-stop-with-json-format,file-metrics-collector-with-json-format)	11m 34s
	2h 19m 54s

Provide feedback