fix: correct evaluation and logging output

[binaryaaron]

Summary

• keep JSONL events canonical while rendering Rich tables only on the plain console
• report exact membership-inference attack counts and percentages without Plotly renormalization
• correct PII progress timing and step labels, fraction formatting, and user-facing log severities
• store final evaluation, generation, timing, failure, and vLLM completion values in W&B summary while retaining Hugging Face training curves in history
• add the evaluation scorecard panel and an explicitly opted-in HTML/report artifact; report upload is best-effort, retry-safe, and may disclose distribution labels

Test plan

• mise run check
• mise run typecheck
• focused Engineer Arc suite: 196 passed
• W&B/results/CLI/vLLM focused suite: 145 passed
• genuine offline W&B smoke: no scalar history keys; media history limited to evaluation/scorecard and opted-in evaluation/report
• mixed console/JSONL and reset-timing regressions: 2 passed
• systemic, style/type-safety, and general council reviews accepted the final product patch

[coderabbitai]

Important

Review skipped

Draft detected.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: 5ba51ed4-52e7-43b2-b0e0-94d08d5fe5ee

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch binaryaaron/log-fixes

---

_{Comment @coderabbitai help to get the list of available commands.}

[codecov]

Codecov Report

❌ Patch coverage is 97.84946% with 8 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
src/nemo_safe_synthesizer/cli/wandb_setup.py	92.75%	5 Missing ⚠️
.../evaluation/components/text_semantic_similarity.py	0.00%	1 Missing ⚠️
src/nemo_safe_synthesizer/observability.py	96.96%	1 Missing ⚠️
..._safe_synthesizer/pii_replacer/data_editor/edit.py	75.00%	1 Missing ⚠️

📢 Thoughts on this report? Let us know!