[VSSL-7701] Demo LLM API with VESSL Run and vLLM (#29)

* Add vllm-online-serving

* Add prom metrics

* Update monitoring

* remove logging

* Add labels

* Use vllm directly from upstream latest to pick up vllm-project/vllm#2316

* Roll back vllm to 0.3.0

* Get patch files for metrics in vllm-project/vllm#2316

* Update llm_engine.py

* Write documents

* Add vllm-online-serving/README-ko.md

* write README.md

Author: hippothewild