feat: add latency and token_usage info in ai-proxy access log

[Revolyssup]

Description

Fixes # (issue)
After the traffic passes through the AI Gateway (also known as APISIX), users hope to: be able to record more variables required for AI proxy scenarios in the access log of AI requests and responses.
Some metadata explicitly mentioned by the user, and corresponding explanations:

• Token: The token usage for each request and response phase needs to be recorded separately.
• Latency: The waiting time for the first response after the request is sent to the AI Instance through the AI Gateway proxy. - Time to first byte.

127.0.0.1 - - [12/Mar/2025:16:24:57 +0530] 127.0.0.1:9080 "POST /anything HTTP/1.1" 429 349 1.088 "-" "curl/8.9.1" - - - "https://somerandom.com "ai_token_usage={\x22prompt_tokens\x22:0,\x22completion_tokens\x22:0,\x22total_tokens\x22:0}" "ai_time_to_first_byte(in seconds)=1.0880000591278""

Checklist

• I have explained the need for this PR and the problem it solves
• I have explained the changes or the new features added to this PR
• I have added tests corresponding to this change
• I have updated the documentation to reflect this change
• I have verified that this change is backward compatible (If not, please discuss on the APISIX mailing list first)

[Baoyuantop]

Hi @Revolyssup, is this PR still being worked on?

[Revolyssup]

Hi @Revolyssup, is this PR still being worked on?

I will take a look at it this weekend.

[github-actions]

This pull request has been marked as stale due to 60 days of inactivity. It will be closed in 4 weeks if no further activity occurs. If you think that's incorrect or this pull request should instead be reviewed, please simply write any comment. Even if closed, you can still revive the PR at any time or discuss it on the [email protected] list. Thank you for your contributions.

[Revolyssup]

superceded by #12518