[GSoC] Add e2e test for `tune` api with LLM hyperparameter optimization #1093

Job	Run time
e2e (v1.27.11, long-running-resume,from-volume-resume,median-stop)	15m 57s
e2e (v1.27.11, grid,bayesian-optimization,tpe,multivariate-tpe,cma-es,hyperband)	21m 59s
e2e (v1.27.11, file-metrics-collector,pytorchjob-mnist)	12m 54s
e2e (v1.27.11, median-stop-with-json-format,file-metrics-collector-with-json-format)	13m 9s
e2e (v1.28.7, long-running-resume,from-volume-resume,median-stop)	14m 11s
e2e (v1.28.7, grid,bayesian-optimization,tpe,multivariate-tpe,cma-es,hyperband)	19m 23s
e2e (v1.28.7, file-metrics-collector,pytorchjob-mnist)	13m 14s
e2e (v1.28.7, median-stop-with-json-format,file-metrics-collector-with-json-format)	11m 56s
e2e (v1.29.2, long-running-resume,from-volume-resume,median-stop)	14m 50s
e2e (v1.29.2, grid,bayesian-optimization,tpe,multivariate-tpe,cma-es,hyperband)	20m 56s
e2e (v1.29.2, file-metrics-collector,pytorchjob-mnist)	12m 41s
e2e (v1.29.2, median-stop-with-json-format,file-metrics-collector-with-json-format)	12m 16s
	3h 3m 26s

Provide feedback