[WIP] Add e2e test for `tune` api with LLM hyperparameter optimization #1131

Job	Run time
e2e (v1.27.11, long-running-resume,from-volume-resume,median-stop)	16m 25s
e2e (v1.27.11, grid,bayesian-optimization,tpe,multivariate-tpe,cma-es,hyperband)	19m 10s
e2e (v1.27.11, file-metrics-collector,pytorchjob-mnist)	13m 56s
e2e (v1.27.11, median-stop-with-json-format,file-metrics-collector-with-json-format)	12m 22s
e2e (v1.28.7, long-running-resume,from-volume-resume,median-stop)	15m 47s
e2e (v1.28.7, grid,bayesian-optimization,tpe,multivariate-tpe,cma-es,hyperband)	19m 18s
e2e (v1.28.7, file-metrics-collector,pytorchjob-mnist)	12m 10s
e2e (v1.28.7, median-stop-with-json-format,file-metrics-collector-with-json-format)	12m 37s
e2e (v1.29.2, long-running-resume,from-volume-resume,median-stop)	16m 4s
e2e (v1.29.2, grid,bayesian-optimization,tpe,multivariate-tpe,cma-es,hyperband)	18m 30s
e2e (v1.29.2, file-metrics-collector,pytorchjob-mnist)	12m 31s
e2e (v1.29.2, median-stop-with-json-format,file-metrics-collector-with-json-format)	12m 53s
	3h 1m 43s

Provide feedback