[TRTLLM-13259][ci] Merge DGX_H100 DeepSeek and GptOss stages

[QiJune]

Both stages are 2-way splits of a single same-backend test suite. The test set is selected purely from the stage name via mako (backend=pytorch plus auto_trigger), and sharding is applied separately, so collapsing splits=2 -> 1 runs the identical set of tests on one node while paying docker pull + pip install only once.

• DGX_H100-4_GPUs-PyTorch-DeepSeek-1/2 -> DeepSeek-1 (1,1)
• DGX_H100-2_GPUs-PyTorch-GptOss-1/2 -> GptOss-1 (1,1)

Merged p99 stays well under the pre-merge critical path and the 4h stage cap, so there is no PR-turnaround impact while freeing one node per stage.

Summary by CodeRabbit

• Chores
  • Updated test infrastructure configurations for H100 GPU specializations to optimize resource allocation and testing workflows.

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.

[QiJune]

/bot run

[coderabbitai]

Ready to act? Review this PR in Change Stack to turn feedback into patch suggestions you can inspect and refine.

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: 5a3abed7-6ce8-427d-9ef1-0299e7df5075

📥 Commits

Reviewing files that changed from the base of the PR and between e47f26e and c50817c.

📒 Files selected for processing (1)

• jenkins/L0_Test.groovy

---

📝 Walkthrough

Walkthrough

SLURM test stage configurations in jenkins/L0_Test.groovy were adjusted for H100 PyTorch specializations: split counts were reduced from 2 to 1 for GptOss and DeepSeek stages, and one duplicate DeepSeek stage entry was removed.

Changes

H100 PyTorch Test Configuration

Layer / File(s)	Summary
H100 PyTorch SLURM stage split counts jenkins/L0_Test.groovy	Split count reduced from 2 to 1 for DGX_H100-2_GPUs-PyTorch-GptOss-1 and DGX_H100-4_GPUs-PyTorch-DeepSeek-1 test stages; DGX_H100-4_GPUs-PyTorch-DeepSeek-2 stage entry was removed.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~5 minutes

Possibly related PRs

• NVIDIA/TensorRT-LLM#14661: Both PRs modify SLURM stage split-count entries in jenkins/L0_Test.groovy for H100 PyTorch test stages.

Suggested reviewers

• dpitman-nvda
• tburt-nv

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly summarizes the main change: merging DGX_H100 DeepSeek and GptOss CI stages by reducing their split counts from 2 to 1.
Description check	✅ Passed	The PR description explains the motivation (avoiding duplicate docker pull and pip install), the specific stage changes, and performance impact assessment, but the template sections (Description, Test Coverage) are not explicitly filled and the PR checklist formatting is incomplete.
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.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

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

[tensorrt-cicd]

PR_Github #52485 [ run ] triggered by Bot. Commit: c50817c Link to invocation

[tensorrt-cicd]

PR_Github #52485 [ run ] completed with state SUCCESS. Commit: c50817c
/LLM/main/L0_MergeRequest_PR pipeline #41776 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

[QiJune]

/bot run

[tensorrt-cicd]

PR_Github #52511 [ run ] triggered by Bot. Commit: c50817c Link to invocation

[tensorrt-cicd]

PR_Github #52511 [ run ] completed with state FAILURE. Commit: c50817c
/LLM/main/L0_MergeRequest_PR pipeline #41801 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

[QiJune]

/bot skip --comment "DGX_H100 DeepSeek and GptOss stages passed"

[tensorrt-cicd]

PR_Github #52528 [ skip ] triggered by Bot. Commit: c50817c Link to invocation

[tensorrt-cicd]

PR_Github #52528 [ skip ] completed with state SUCCESS. Commit: c50817c
Release Check Pipeline #4118 failed

Link to invocation

[QiJune]

/bot skip --comment "DGX_H100 DeepSeek and GptOss stages passed"

[tensorrt-cicd]

PR_Github #52539 [ skip ] triggered by Bot. Commit: c50817c Link to invocation

[tensorrt-cicd]

PR_Github #52539 [ skip ] completed with state SUCCESS. Commit: c50817c
Skipping testing for commit c50817c

Link to invocation