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)