Refactor: reintroduce `LLM_MODEL` Env Variable for Flexible Model Implementation in `chatqna-xeon-ui-server` #2319

Job	Run time
get-test-matrix / Get-test-matrix	10s
example-test (ChatQnA, gaudi) / get-test-case	10s
example-test (ChatQnA, xeon) / get-test-case	4s
example-test (ChatQnA, xeon) / run-test (test_compose_on_xeon.sh)	16m 29s
example-test (ChatQnA, xeon) / run-test (test_compose_pinecone_on_xeon.sh)	31m 37s
example-test (ChatQnA, xeon) / run-test (test_compose_qdrant_on_xeon.sh)	13m 21s
example-test (ChatQnA, xeon) / run-test (test_compose_vllm_on_xeon.sh)	17m 7s
example-test (ChatQnA, xeon) / run-test (test_compose_without_rerank_on_xeon.sh)	15m 27s
example-test (ChatQnA, gaudi) / run-test (test_compose_guardrails_on_gaudi.sh)	14m 13s
example-test (ChatQnA, gaudi) / run-test (test_compose_on_gaudi.sh)	14m 32s
example-test (ChatQnA, gaudi) / run-test (test_compose_vllm_on_gaudi.sh)	15m 25s
example-test (ChatQnA, gaudi) / run-test (test_compose_without_rerank_on_gaudi.sh)	13m 39s
	2h 32m 14s

Provide feedback