chore: update to latest openinference embedding schema

[codefromthecrypt]

Description

This updates to the latest openinference semantic conventions for CreateEmbeddings requests, and re-records all spans. This also removes the ModelO3Mini as OpenAI themselves recently removed it.

Related Issues/PRs (if applicable)

#1085

[codefromthecrypt]

wait.. there's a huge data here.. lemme look which

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 79.68%. Comparing base (8c767bb) to head (9efd3de).

❌ Your project status has failed because the head coverage (79.68%) is below the target coverage (86.00%). You can increase the head coverage or adjust the target coverage.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1232      +/-   ##
==========================================
- Coverage   79.72%   79.68%   -0.04%     
==========================================
  Files          90       90              
  Lines       10254    10254              
==========================================
- Hits         8175     8171       -4     
- Misses       1715     1720       +5     
+ Partials      364      363       -1     

☔ 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.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[codefromthecrypt]

ah right because the span data is decoded vectors, which ends up really big in json due to otel json encoding not being efficient:

      "key": "embedding.embeddings.4.embedding.vector",
      "value": {
        "arrayValue": {
          "values": [
            {
              "doubleValue": 0.030946535989642143
            },
            {
              "doubleValue": 0.007837203331291676
            },

[mathetake]

LGTM-- I confirmed that this does not have any functional change, which is aligned with the PR description. Pure tests updates and hardening sounds good to me