include stop string in completions output

[saiatmakuri]

Pull Request Summary

Bring over flag added in vllm-project/vllm#1976 to llm engine

Test Plan and Usage Guide

test locally
without the flag:

> curl -H "Content-Type: application/json" -H "Authorization: Basic <AUTH>" -d '{ "prompt": "Whats 2+2? ", "stop_sequences": ["\n"], "max_new_tokens": 10, "temperature": 0.1 }' http://localhost:5000/v1/llm/completions-sync?model_endpoint_name=llama-7b-vllm-test-temp
{"request_id":"907c759f-e704-4d32-98a5-4dbe388ec04a","output":{"text":"4.","num_prompt_tokens":9,"num_completion_tokens":3,"tokens":null}}

note: "4." above

with the flag:

> curl -H "Content-Type: application/json" -H "Authorization: Basic <AUTH>" -d '{ "prompt": "Whats 2+2? ", "stop_sequences": ["\n"], "max_new_tokens": 10, "temperature": 0.1, "include_stop_str_in_output": true }' http://localhost:5000/v1/llm/completions-sync?model_endpoint_name=llama-7b-vllm-test-temp
{"request_id":"7f465cd7-8d42-4470-96ab-6b8cdfb2b5b0","output":{"text":"4.\n","num_prompt_tokens":9,"num_completion_tokens":3,"tokens":null}}

note: "4.\n" above