[TRTLLM-15115][test] Prune Starcoder2-3B functional and unit tests

[xinhe-nv]

Remove out-of-scope Starcoder2-3B coverage per Test Scope Review 2026.7, including QA/CI list entries, encode/cli harnesses, and 3B-specific unit tests while leaving StarCoder2-7B/15B coverage intact.

Dev Engineer Review

• Removed StarCoder2-3B accuracy references, harness coverage, fixtures, performance configuration, and unit-test coverage.
• Preserved StarCoder2-7B and StarCoder2-15B coverage.
• Updated QA, A100, and test-duration entries consistently.
• No production-code or API changes were introduced.
• Configuration and test-list changes have valid scope and no apparent format or duplication issues.

QA Engineer Review

• Removed TestStarcoder2_3B CLI accuracy coverage.
• Removed test_starcoder2_multi_lora, STARCODER2_3B_CONFIG, and related StarCoder2-3B modeling scenarios.
• Removed StarCoder2-3B tokenizer parameterization from the incremental decoding test.
• Removed the related accuracy entries from:
  • tests/integration/test_lists/qa/llm_function_core.txt
  • tests/integration/test_lists/test-db/l0_a100.yml
• The integration accuracy test was covered by the QA and CI lists. The removed unit-test cases had no separate test-list entries.
• StarCoder2-7B and StarCoder2-15B coverage remains.
• Verdict: needs follow-up because two A100 CI runs failed. The subsequent A100X run passed with fail-fast disabled.

Description

Test Coverage

PR Checklist

Please review the following before submitting your PR:

• PR description clearly explains what and why. If using CodeRabbit's summary, please make sure it makes sense.

• PR Follows TRT-LLM CODING GUIDELINES to the best of your knowledge.

• Test cases are provided for new code paths (see test instructions)

• If PR introduces API changes, an appropriate PR label is added - either api-compatible or api-breaking. For api-breaking, include BREAKING in the PR title.

• Any new dependencies have been scanned for license and vulnerabilities

• CODEOWNERS updated if ownership changes

• Documentation updated as needed

• Update tava architecture diagram if there is a significant design change in PR.

• The reviewers assigned automatically/manually are appropriate for the PR.

• Please check this after reviewing the above items as appropriate for this PR.

GitHub Bot Help

To see a list of available CI bot commands, please comment /bot help.

[xinhe-nv]

/bot run --stage-list "a100"

[coderabbitai]

Note

Reviews paused

It looks like this branch is under active development. To avoid overwhelming you with review comments due to an influx of new commits, CodeRabbit has automatically paused this review. You can configure this behavior by changing the reviews.auto_review.auto_pause_after_reviewed_commits setting.

Use the following commands to manage reviews:

• @coderabbitai resume to resume automatic reviews.
• @coderabbitai review to trigger a single review.

Use the checkboxes below for quick actions:

• ▶️ Resume reviews
• 🔍 Trigger review

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: 1f91dc7f-e1c0-4023-b097-43bf9755bb4d

📥 Commits

Reviewing files that changed from the base of the PR and between 18dc0c7 and 2bdd5db.

📒 Files selected for processing (1)

• tests/integration/test_lists/test-db/l0_a100.yml

🚧 Files skipped from review as they are similar to previous changes (1)

• tests/integration/test_lists/test-db/l0_a100.yml

Included review availability: Your plan includes up to 12 reviews per rolling hour; 11 remain after this review.

---

Walkthrough

The PR removes StarCoder2-3B from accuracy references, integration test parameters, model fixtures, performance configurations, timing data, attention settings, modeling tests, and tokenizer test parameters.

Changes

StarCoder2-3B coverage removal

Layer / File(s)	Summary
Integration model coverage tests/integration/defs/accuracy/test_llm_api_pytorch_encode.py, tests/integration/defs/conftest.py, tests/integration/defs/perf/allowed_configs.py	The decoder model parameter, model fixtures, and performance allowlist entry no longer include starcoder2-3b.
Integration references and schedules tests/integration/defs/accuracy/references/gsm8k.yaml, tests/integration/defs/accuracy/test_cli_flow.py, tests/integration/test_lists/qa/llm_function_core.txt, tests/integration/defs/.test_durations	StarCoder2-3B accuracy references, CLI accuracy coverage, scheduled decoder encode coverage, and recorded test timing are removed.
Unit model coverage tests/unittest/_torch/attention/model_attn_config.py, tests/unittest/_torch/modeling/test_modeling_starcoder2.py, tests/unittest/llmapi/test_llm.py	The StarCoder2-3B attention configuration, modeling scenarios, related imports, and tokenizer parameter are removed. StarCoder2-7B and StarCoder2-15B coverage remains.

Estimated code review effort: 2 (Simple) | ~10 minutes

Merge Risk: ⚪ Minimal · up to 2bdd5

This change prunes Starcoder2-3B test coverage while retaining Starcoder2-7B/15B coverage; no actionable merge-blocking risk remains beyond normal checks and review.

Suggested reviewers: asfiyab-nvidia, brnguyen2, yechank-nvidia

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly identifies the StarCoder2-3B test coverage removal and uses the required ticket and type format.
Description check	✅ Passed	The description explains the scope and retained coverage, includes the required template and checklist, and marks the checklist complete.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

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

[tensorrt-cicd]

PR_Github #66167 [ run ] triggered by Bot. Commit: 3e3cce7 Link to invocation

[tensorrt-cicd]

PR_Github #66167 [ run ] completed with state FAILURE. Commit: 3e3cce7
/LLM/main/L0_MergeRequest_PR pipeline #53841 (Partly Tested) completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

CI Agent Failure Analysis

Link to invocation

[xinhe-nv]

/bot run --gpu-type "A100X"

[tensorrt-cicd]

PR_Github #66202 [ run ] triggered by Bot. Commit: 3e3cce7 Link to invocation

[tensorrt-cicd]

PR_Github #66202 [ run ] completed with state FAILURE. Commit: 3e3cce7
/LLM/main/L0_MergeRequest_PR pipeline #53871 (Partly Tested) completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

• Please check the failed tests and fix your PR
• If you cannot view the failures, ask the CI triggerer to share details
• Once fixed, request an NVIDIA team member to trigger CI again

CI Agent Failure Analysis

Link to invocation

[xinhe-nv]

/bot run --gpu-type "A100X" --disable-fail-fast

[tensorrt-cicd]

PR_Github #66296 [ run ] triggered by Bot. Commit: db9af45 Link to invocation

[tensorrt-cicd]

PR_Github #66296 [ run ] completed with state SUCCESS. Commit: db9af45
/LLM/main/L0_MergeRequest_PR pipeline #53955 (Partly Tested) completed with status: 'SUCCESS'

CI Report

Link to invocation

[xinhe-nv]

/bot reuse-pipeline

[QiJune]

LGTM

[xinhe-nv]

/bot reuse-pipeline

[tensorrt-cicd]

PR_Github #66642 [ reuse-pipeline ] triggered by Bot. Commit: 9c1e2ea Link to invocation

[tensorrt-cicd]

PR_Github #66642 [ reuse-pipeline ] completed with state SUCCESS. Commit: 9c1e2ea
Reusing PR_Github #66296 (Partly Tested) for commit 9c1e2ea

Link to invocation