Adding usage info in llm.py

[Vidit-Ostwal]

Fixes #4170

---

Note

Introduces consistent token usage accounting across sync/async and streaming/non-streaming LiteLLM paths, with callback logging and event emission.

• In llm.py, extract usage from LiteLLM responses and stream chunks, call _track_token_usage_internal(usage_info), and invoke _handle_streaming_callbacks(...) in streaming flow; mirror usage tracking for _handle_non_streaming_response and _ahandle_non_streaming_response after completion calls
• Async streaming handler _ahandle_streaming_response now captures usage_info, logs success callbacks, and updates internal counters before emitting completion events
• Adds tests in tests/test_llm.py to validate _token_usage updates for call/acall in both streaming and non-streaming modes
• Adds VCR cassettes for the new tests: test_usage_info_* (streaming and non-streaming, sync and async)

^{Written by Cursor Bugbot for commit 104075a. This will update automatically on new commits. Configure here.}

[cursor]

This PR is being reviewed by Cursor Bugbot

Details

Your team is on the Bugbot Free tier. On this plan, Bugbot will review limited PRs each billing cycle for each member of your team.

To receive Bugbot reviews on all of your PRs, visit the Cursor dashboard to activate Pro and start your 14-day free trial.

[lorenzejay]

Lovely ! Can you add a couple tests showing this with

llm.call
llm.acall?

[lorenzejay]

tested and LGTM