[FIX] Use the new entry point so the /metrics endpoint works.

[xiangshen-dk]

Description

This PR change the entrypoint command to use vllm.entrypoints.openai.api_server. The old one vllm.entrypoints.api_server doesn't expose the /metrics route.

Once it's merged, the example in the doc needs to be updated to the following:

USER_PROMPT="I'm new to coding. If you could only recommend one programming language to start with, what would it be and why?"

curl -X POST http://localhost:8000/v1/completions \
  -H "Content-Type: application/json" \
  -d @- <<EOF
{
    "model": "google/gemma-2-2b-it",
    "prompt": "<start_of_turn>user\n${USER_PROMPT}<end_of_turn>\n",
    "temperature": 0.90,
    "top_p": 1.0,
    "max_tokens": 128
}
EOF

The gradio.yaml file needs to be updated as well.

Tasks

• The contributing guide has been read and followed.
• The samples added / modified have been fully tested.
• Workflow files have been added / modified, if applicable.
• Region tags have been properly added, if new samples.
• All dependencies are set to up-to-date versions, as applicable.
• Merge this pull-request for me once it is approved.