Add response token count logic to LLM instrumentations

[umaannamalai]

This PR adds new logic to capture token counts directly from LLM response objects in OpenAI, Bedrock, and Gemini.

[github-actions]

✅MegaLinter analysis: Success

Descriptor	Linter	Files	Fixed	Errors	Warnings	Elapsed time
✅ ACTION	actionlint	7		0	0	0.88s
✅ MARKDOWN	markdownlint	7	0	0	0	1.32s
✅ PYTHON	ruff	948	0	0	0	0.98s
✅ PYTHON	ruff-format	948	0	0	0	0.39s
✅ YAML	prettier	15	0	0	0	1.51s
✅ YAML	v8r	15		0	0	5.32s
✅ YAML	yamllint	15		0	0	0.68s

See detailed reports in MegaLinter artifacts

[codecov-commenter]

Codecov Report

❌ Patch coverage is 90.21739% with 18 lines in your changes missing coverage. Please review.
✅ Project coverage is 81.89%. Comparing base (8d18c0a) to head (1245264).

Files with missing lines	Patch %	Lines
newrelic/hooks/mlmodel_gemini.py	67.39%	7 Missing and 8 partials ⚠️
newrelic/hooks/external_botocore.py	97.27%	0 Missing and 3 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1583      +/-   ##
==========================================
+ Coverage   81.82%   81.89%   +0.06%     
==========================================
  Files         207      207              
  Lines       23953    24081     +128     
  Branches     3799     3812      +13     
==========================================
+ Hits        19599    19720     +121     
- Misses       3086     3088       +2     
- Partials     1268     1273       +5     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.