[SC-13352] Small fixes to TrainingTestDegradation test

[juanmleng]

Pull Request Description

What and why?

What

Removed incorrect references to "accuracy" metric from the docstring. The test only evaluates precision, recall, and f1-score per class, not accuracy.

Why

The test docstring is used as context for LLM-based test descriptions. The inaccuracy in the docstring causes the LLM to incorrectly state that accuracy is being computed, which results in low faithfulness scores when evaluating the generated descriptions against the actual test implementation.

How to test

Run the TrainingTestDegradation test—using the customer churn or the credit risk scorecard notebook—and check that the generated description does not reference accuracy as one of the computed metrics.

Before:

After:

What needs special review?

Dependencies, breaking changes, and deployment notes

Release notes

Checklist

• What and why
• Screenshots or videos (Frontend)
• How to test
• What needs special review
• Dependencies, breaking changes, and deployment notes
• Labels applied
• PR linked to Shortcut
• Unit tests added (Backend)
• Tested locally
• Documentation updated (if required)
• Environment variable additions/changes documented (if required)

[github-actions]

PR Summary

This PR updates the documentation within the TrainingTestDegradation test file by removing references to the accuracy metric. The changes clarify that the test now focuses solely on precision, recall, and f1 score to evaluate the degradation between the training and test datasets. Additionally, the explanation of the threshold for acceptable degradation was updated to explicitly include the default value (0.10). These modifications aim to improve clarity and consistency in the test's description without altering the underlying test logic.

Test Suggestions

• Run the tests to verify that model performance is only evaluated using precision, recall, and f1 score.
• Validate that the output table correctly lists the train score, test score, degradation percentage, and pass/fail status for the specified metrics.
• Check that documentation changes are reflected accurately in any auto-generated test reports or rendered documentation views.

[AnilSorathiya]

thanks!