[None][refactor] Remove TensorRT performance baseline and update to PyTorch only by yufeiwu-nv · Pull Request #15256 · NVIDIA/TensorRT-LLM

yufeiwu-nv · 2026-06-11T09:14:07Z

Deleted base_perf.csv as TensorRT performance stages are no longer supported.
Updated scripts and documentation to reference base_perf_pytorch.csv for performance benchmarking.
Adjusted performance check logic to reflect the removal of TensorRT metrics.

Signed-off-by: yufeiwu-nv 230315618+yufeiwu-nv@users.noreply.github.com

Summary by CodeRabbit

Chores
- Consolidated performance baseline testing to use a single PyTorch reference configuration; removed TensorRT performance baseline stage.
Documentation
- Updated performance testing documentation to reflect the updated baseline file references.

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.

…yTorch only - Deleted `base_perf.csv` as TensorRT performance stages are no longer supported. - Updated scripts and documentation to reference `base_perf_pytorch.csv` for performance benchmarking. - Adjusted performance check logic to reflect the removal of TensorRT metrics. Signed-off-by: yufeiwu-nv <230315618+yufeiwu-nv@users.noreply.github.com>

coderabbitai · 2026-06-11T09:15:22Z

📝 Walkthrough

Walkthrough

This PR consolidates performance validation infrastructure to use only the PyTorch baseline. It removes TensorRT perf baseline selection logic from Jenkins CI configuration and the perf validation script, hardcodes PyTorch baseline usage, removes the corresponding TensorRT perf test stage, and updates documentation and user-facing messages to reflect the change.

Changes

TensorRT Perf Baseline Removal

Layer / File(s)	Summary
Jenkins CI baseline selection and stage removal `jenkins/L0_Test.groovy`	Removes conditional perf baseline selection to always use `base_perf_pytorch.csv` and removes the `H100_PCIe-TensorRT-Perf-1` perf stage from x86TestConfigs mapping.
Perf script baseline simplification `jenkins/scripts/slurm_run.sh`	Simplifies perfMode baseline selection from stage-dependent conditional logic to hardcoded `basePerfFilename="base_perf_pytorch.csv"`.
Documentation and output message updates `tests/integration/defs/perf/README.md`, `tests/integration/defs/perf/sanity_perf_check.py`	Updates README to define and reference `base_perf_pytorch.csv` as the baseline file, and changes sanity_perf_check.py output message to use dynamic filename via `self.base_perf_csv.name`.

🎯 1 (Trivial) | ⏱️ ~4 minutes

🚥 Pre-merge checks | ✅ 3 | ❌ 2

❌ Failed checks (1 warning, 1 inconclusive)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.
Description check	❓ Inconclusive	The PR description provides a clear summary of changes but lacks detailed explanations in the template sections.	Expand the Description section to explain why TensorRT support was removed and the impact of this change. Complete the Test Coverage section by listing specific tests that validate these changes.

✅ Passed checks (3 passed)

Check name	Status	Explanation
Title check	✅ Passed	Title clearly describes removing TensorRT baseline and transitioning to PyTorch-only metrics.
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.}

yufeiwu-nv · 2026-06-11T09:15:38Z

/bot run

coderabbitai

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@tests/integration/defs/perf/README.md`:
- Line 45: The markdown line starting with "** In the CI, `<base_perf_csv_path>`
is the `base_perf_pytorch.csv` file path mentioned above." has an opening bold
marker but no closing marker; fix it by closing the emphasis (e.g., add trailing
"**" or remove the opening "**") so the line becomes properly balanced around
the intended emphasized text (refer to the literal string "<base_perf_csv_path>"
in the README content).

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

Push a commit to this branch (recommended)
Create a new PR with the fixes

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: 8aa497cf-51bd-42ec-b321-513621f98b16

📥 Commits

Reviewing files that changed from the base of the PR and between 84b349f and 8ee3818.

⛔ Files ignored due to path filters (1)

tests/integration/defs/perf/base_perf.csv is excluded by !**/*.csv

📒 Files selected for processing (4)

jenkins/L0_Test.groovy
jenkins/scripts/slurm_run.sh
tests/integration/defs/perf/README.md
tests/integration/defs/perf/sanity_perf_check.py

tensorrt-cicd · 2026-06-11T09:22:42Z

PR_Github #53540 [ run ] triggered by Bot. Commit: 8ee3818 Link to invocation

Signed-off-by: yufeiwu-nv <230315618+yufeiwu-nv@users.noreply.github.com>

yufeiwu-nv · 2026-06-11T09:54:48Z

/bot run

tensorrt-cicd · 2026-06-11T10:00:38Z

PR_Github #53548 [ run ] triggered by Bot. Commit: 1c32c0b Link to invocation

tensorrt-cicd · 2026-06-11T10:00:44Z

PR_Github #53540 [ run ] completed with state ABORTED. Commit: 8ee3818

Link to invocation

tensorrt-cicd · 2026-06-11T14:32:02Z

PR_Github #53548 [ run ] completed with state SUCCESS. Commit: 1c32c0b
/LLM/main/L0_MergeRequest_PR pipeline #42699 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

jiaganc · 2026-06-12T02:31:21Z

/bot run

tensorrt-cicd · 2026-06-12T02:37:01Z

PR_Github #53753 [ run ] triggered by Bot. Commit: 1c32c0b Link to invocation

tensorrt-cicd · 2026-06-12T05:16:48Z

PR_Github #53753 [ run ] completed with state FAILURE. Commit: 1c32c0b
/LLM/main/L0_MergeRequest_PR pipeline #42876 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

yufeiwu-nv · 2026-06-12T06:26:11Z

/bot run --disable-fail-fast

tensorrt-cicd · 2026-06-12T06:32:22Z

PR_Github #53820 [ run ] triggered by Bot. Commit: 76d7739 Link to invocation

tensorrt-cicd · 2026-06-12T14:42:49Z

PR_Github #53820 [ run ] completed with state SUCCESS. Commit: 76d7739
/LLM/main/L0_MergeRequest_PR pipeline #42936 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

tburt-nv · 2026-06-12T17:24:30Z

You may need to reapply #15166 in this PR, since it has been temporarily reverted in #15310

…#15166) NVIDIA#15166 was temporarily reverted in NVIDIA#15310. Reapply it here so the TensorRT perf test removal in l0_perf.yml lands together with the TensorRT stage/baseline cleanup in this PR, keeping the test list and Jenkins perf stages consistent. Signed-off-by: yufeiwu-nv <230315618+yufeiwu-nv@users.noreply.github.com>

yufeiwu-nv · 2026-06-15T05:52:02Z

/bot run

yufeiwu-nv · 2026-06-15T05:57:40Z

/bot skip --comment "only test modify"

tensorrt-cicd · 2026-06-15T05:57:52Z

PR_Github #54228 [ run ] triggered by Bot. Commit: d5983b4 Link to invocation

tensorrt-cicd · 2026-06-15T06:03:13Z

PR_Github #54229 [ skip ] triggered by Bot. Commit: d5983b4 Link to invocation

tensorrt-cicd · 2026-06-15T06:17:50Z

PR_Github #54229 [ skip ] completed with state SUCCESS. Commit: d5983b4
Skipping testing for commit d5983b4

Link to invocation

yufeiwu-nv requested a review from a team June 11, 2026 09:14

yufeiwu-nv requested review from a team as code owners June 11, 2026 09:14

yufeiwu-nv requested review from dpitman-nvda and mlefeb01 June 11, 2026 09:14

github-actions Bot assigned yufeiwu-nv Jun 11, 2026

Merge branch 'main' into post_merge

8ee3818

coderabbitai Bot reviewed Jun 11, 2026

View reviewed changes

Comment thread tests/integration/defs/perf/README.md Outdated

[None][doc] Fix unclosed bold marker in perf README

1c32c0b

Signed-off-by: yufeiwu-nv <230315618+yufeiwu-nv@users.noreply.github.com>

mzweilz approved these changes Jun 11, 2026

View reviewed changes

Merge branch 'main' into post_merge

76d7739

yufeiwu-nv enabled auto-merge (squash) June 12, 2026 06:44

Merge branch 'main' into post_merge

c60aaf6

yufeiwu-nv added 2 commits June 15, 2026 13:34

Merge branch 'main' into post_merge

0c5d197

yufeiwu-nv requested a review from a team as a code owner June 15, 2026 05:50

ruodil approved these changes Jun 15, 2026

View reviewed changes

yufeiwu-nv merged commit 26ea499 into NVIDIA:main Jun 15, 2026
7 checks passed

Uh oh!

Conversation

yufeiwu-nv commented Jun 11, 2026 • edited by coderabbitai Bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary by CodeRabbit

Description

Test Coverage

PR Checklist

GitHub Bot Help

Uh oh!

coderabbitai Bot commented Jun 11, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Walkthrough

Changes

❌ Failed checks (1 warning, 1 inconclusive)

Uh oh!

yufeiwu-nv commented Jun 11, 2026

Uh oh!

coderabbitai Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

tensorrt-cicd commented Jun 11, 2026

Uh oh!

yufeiwu-nv commented Jun 11, 2026

Uh oh!

tensorrt-cicd commented Jun 11, 2026

Uh oh!

tensorrt-cicd commented Jun 11, 2026

Uh oh!

tensorrt-cicd commented Jun 11, 2026

Uh oh!

jiaganc commented Jun 12, 2026

Uh oh!

tensorrt-cicd commented Jun 12, 2026

Uh oh!

tensorrt-cicd commented Jun 12, 2026

Uh oh!

yufeiwu-nv commented Jun 12, 2026

Uh oh!

tensorrt-cicd commented Jun 12, 2026

Uh oh!

tensorrt-cicd commented Jun 12, 2026

Uh oh!

tburt-nv commented Jun 12, 2026

Uh oh!

yufeiwu-nv commented Jun 15, 2026

Uh oh!

yufeiwu-nv commented Jun 15, 2026

Uh oh!

tensorrt-cicd commented Jun 15, 2026

Uh oh!

tensorrt-cicd commented Jun 15, 2026

Uh oh!

tensorrt-cicd commented Jun 15, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

yufeiwu-nv commented Jun 11, 2026 •

edited by coderabbitai Bot

Loading

coderabbitai Bot commented Jun 11, 2026 •

edited

Loading