[shengliangxu]

What does this PR do?

Type of change: CI (test runtime); the only non-test file is examples/torch_onnx/torch_quant_to_onnx.py

Three cuts to the torch_onnx example lane:

• No DataLoader workers in the calibration loaders. load_calibration_data builds its DataLoaders over lists of tensors that are already on the GPU, so the four workers load nothing, and since the script forced the spawn start method (CUDA cannot be used in forked workers), every loader pass spawned four processes that re-imported torch, timm and modelopt and received the dataset over CUDA IPC: about 14 s per pass and about 300 s per torch_onnx CI job. Both loaders use num_workers=0, and the module-level mp.set_start_method("spawn", force=True) is dropped, which also stops importing the script from changing the start method of the importing process (two tests import it inside the pytest process). Nothing else in the script uses multiprocessing. The --evaluate path keeps the CPU-only DataLoader workers of examples/onnx_ptq/evaluation.py, now under the default start method, exactly as in the onnx_ptq example that calls the same function.
• swinv2 int8, mxfp8 and auto are skipped on pull requests. swinv2_tiny runs all five formats on top of swin_tiny, which already covers every format through the same exporter and TensorRT path; swinv2 only adds its cosine attention and post-norm blocks. On PRs (GITHUB_REF under refs/heads/pull-request/) fp8 and nvfp4 still run and exercise those blocks; scheduled and local runs keep every combination, and test ids are unchanged. swinv2_tiny[auto] alone averages 121 s over 304 CI runs; the three skipped cases cost about 110 s of every PR run.
• build_trt_engine runs trtexec for one iteration. It let trtexec run its default benchmark after the build (200 ms warm-up, then at least 3 s and 10 iterations) and discarded the output. --warmUp=0 --duration=0 --iterations=1 still executes the engine once and removes about 3 s per engine, roughly 40 s per lane run. --trt_build users see no difference besides it returning sooner.

Time improvement (measured in CI)

torch_onnx lane; before = #2846, #2848 and #2850 (875-880 s; #2849 ran on a slower runner: 1,060 s), after = #2853, #2854 and #2855 (452-463 s).

	before	after	change
lane pytest	877 s	458 s	-419 s (-48%)
of which: the three swinv2 cases skipped on PRs	143 s	0	-143 s
of which: the cases that still run (no DataLoader workers, trtexec runs once)			-277 s
whole job	18.6 min	12.2 min	-6.4 min

This lane repeats to within 1% (875-880 s before, 452-463 s after), so it is the cleanest measurement of the series. For example swin_tiny with auto quantization goes from 66 s to 24 s and vit_tiny with auto from 60 s to 17 s.

Method: every PR of the series has its own CI run, so "before" is the run of the PR(s) below that lack the change and "after" the runs that carry it. Times come from the job logs (--durations, or the gap between consecutive result lines). Where runner speed matters, the tests that no PR of the series touches scale each run, and the raw and the scaled medians are both shown. One test varies by 15-30% from run to run on its own, so smaller differences are not evidence.

Usage

No API change. The skip is keyed on GITHUB_REF, so a local run keeps every combination.

Testing

py_compile, ruff 0.15.20 and pytest --collect-only on CPU (24 tests, ids unchanged). With the example command stubbed, GITHUB_REF=refs/heads/pull-request/123 skips exactly swinv2 int8, mxfp8 and auto, and any other ref skips none; the two in-process mapping tests that import the script pass on CPU, and build_trt_engine passes the three flags to a fake trtexec. Not run: the lane itself (it needs a GPU, TensorRT and model downloads) and the flags against a real trtexec, so this PR's torch_onnx job is the check. Pre-commit passes. #2685 (open) also edits this example; the changes touch different lines.

Before your PR is "Ready for review"

Make sure you read and follow Contributor guidelines and your commits are signed (git commit -s -S).

Make sure you read and follow the Security Best Practices (e.g. avoiding hardcoded trust_remote_code=True, torch.load(..., weights_only=False), pickle, etc.).

• Is this change backward compatible?: N/A (CI and tests)
• If you copied code from any other sources or added a new PIP dependency, did you follow guidance in CONTRIBUTING.md: no new dependency
• Did you write any new necessary tests?: N/A (existing tests; the PR-only skip keeps test ids)
• Did you update Changelog?: N/A (CI and tests)
• Did you get Claude approval on this PR?: N/A (optional, not run)

Additional Information

Slice 9 of 11 of the GPU-CI speed-up work, split out of the reference draft #2730 (not for merge; it holds the whole series with its measurements). Slices 1, 2, 5, 8 and 10 (#2845, #2846, #2849, #2852, #2854) have merged. The remaining slices (#2847, #2848, #2850, #2851, #2853, #2855) are independent PRs on main: each applies to main on its own, and merging all six onto main in different orders gives the same tree. The times above are from the series' CI runs before 2026-10-21, when all slices formed one linear stack; a PR number in a method note names that PR's run from then.

1. [OMNIML-6178] [1/11] Run the unit tests in parallel with pytest-xdist #2845 [OMNIML-6178] [1/11] Run the unit tests in parallel with pytest-xdist (merged)
2. [OMNIML-6178] [2/11] Use the sysmon coverage core in CI #2846 [OMNIML-6178] [2/11] Use the sysmon coverage core in CI (merged)
3. [OMNIML-6179] [3/11] Run manual-marked tests on nightly and on-demand GPU runs #2847 [OMNIML-6179] [3/11] Run manual-marked tests on nightly and on-demand GPU runs
4. [OMNIML-6181] [4/11] Give gpu_megatron's slow cold starts more time #2848 [OMNIML-6181] [4/11] Give gpu_megatron's slow cold starts more time
5. [OMNIML-6179] [5/11] Make the ONNX tests cheaper: CPU-only ones to the unit lane, small models #2849 [OMNIML-6179] [5/11] Make the ONNX tests cheaper: CPU-only ones to the unit lane, small models (merged)
6. [OMNIML-6179] [6/11] Use local tokenizers and calibration text in GPU tests, shrink puzzletron data #2850 [OMNIML-6179] [6/11] Use local tokenizers and calibration text in GPU tests, shrink puzzletron data
7. [OMNIML-6180] [7/11] Trim the llm_eval, llm_qat, EAGLE and megatron_bridge example tests #2851 [OMNIML-6180] [7/11] Trim the llm_eval, llm_qat, EAGLE and megatron_bridge example tests
8. [OMNIML-6180] [8/11] Trim the DFlash regression test #2852 [OMNIML-6180] [8/11] Trim the DFlash regression test (merged)
9. [OMNIML-6180] [9/11] Trim the torch_onnx example lane #2853 [OMNIML-6180] [9/11] Trim the torch_onnx example lane (this PR)
10. [OMNIML-6181] [10/11] Make the distributed worker pool fail fast when a worker dies #2854 [OMNIML-6181] [10/11] Make the distributed worker pool fail fast when a worker dies (merged)
11. [OMNIML-6181] [11/11] Share one Megatron worker pool across the gpu_megatron modules #2855 [OMNIML-6181] [11/11] Share one Megatron worker pool across the gpu_megatron modules

Summary by CodeRabbit

• Updates
  • Quantization examples now use zero-worker calibration loading with device-resident samples.
  • TensorRT example builds perform a single inference iteration without warmup or timing.
• Tests
  • Pull-request validation skips selected SwinV2 Tiny formats, while scheduled validation continues to cover them.
  • Pull-request detection uses the repository’s CI reference environment.
• Impact
  • These changes affect example workflows and validation coverage only; no application user-facing functionality is changed.

[copy-pr-bot]

Auto-sync is disabled for draft pull requests in this repository. Workflows must be run manually.

Contributors can view more details about this message here.

[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.

• Last reviewed commit: 7c8634b
• Latest commit, not reviewed: ced00bc

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: Repository: NVIDIA/Model-Optimizer/.coderabbit.yaml
• Review profile: CHILL
• Plan: Enterprise
• Run ID: 3515e3eb-a21f-458a-8dd0-6000b13fe607

📥 Commits

Reviewing files that changed from the base of the PR and between 9aff571 and 7c8634b.

📒 Files selected for processing: 2

• examples/torch_onnx/torch_quant_to_onnx.py
• tests/examples/torch_onnx/test_torch_quant_to_onnx.py

💤 Files with no reviewable changes: 1

• examples/torch_onnx/torch_quant_to_onnx.py

Included review availability: This review used your included allowance. Your plan provides up to 12 included reviews per hour; 0 remain after this review.

📝 Walkthrough

The ONNX example now uses single-process calibration and one TensorRT inference iteration. PR CI skips selected swinv2_tiny formats, while scheduled CI continues to run them.

Changes: ONNX execution and CI coverage

Layer / File(s)	Summary
Calibration and TensorRT execution settings examples/torch_onnx/torch_quant_to_onnx.py	The example removes forced multiprocessing setup, sets calibration loaders to num_workers=0, and configures TensorRT without warm-up or timed duration and with one inference iteration.
PR-specific model test selection tests/examples/torch_onnx/test_torch_quant_to_onnx.py	PR CI skips selected swinv2_tiny formats. Scheduled CI does not apply these skips.

Priority: ⬇️ Low

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

Change: Other

Suggested reviewers: ajrasane

Merge Risk: ⚪ Minimal · up to 7c863

This change shortens the torch_onnx example CI lane. PR runs skip three SwinV2 cases that scheduled runs still cover, and calibration and TensorRT checks run with less overhead. The PR detection matches how this repository triggers PR builds. No merge-blocking issue remains.

Pre-merge checks: 5 · 1

Failed checks: 1 warning

Check name	Status	Explanation	Resolution
Docstring Coverage		Docstring coverage is 50.00% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 4 functions across 2 files.	Write docstrings for the functions missing them to satisfy the coverage threshold.

Passed checks: 5 passed

Check name	Status	Explanation
Title check		The title clearly summarizes the main change: reducing the torch_onnx example CI lane runtime. It is concise and specific.
Linked Issues check		Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check		Check skipped because no linked issues were found for this pull request.
Security Anti-Patterns		PASS. The PR changes only two Python files and adds no model-loading, deserialization, dynamic eval/exec, # nosec, or dependency patterns covered by the check. The only model.eval() occurrence…
Description Check		Check skipped - CodeRabbit’s high-level summary is enabled.

Full details: Security Anti-Patterns

Explanation

PASS. The PR changes only two Python files and adds no model-loading, deserialization, dynamic eval/exec, # nosec, or dependency patterns covered by the check. The only model.eval() occurrence is an existing method call, not dynamic code execution. No pyproject.toml or requirements.txt file changed.

✨ Finishing Touches: 💡 2

📝 Generate docstrings: 💡

• Commit to this branch
• Create a new PR

🛠️ Fix failing CI checks: 💡

• Commit to this branch
• Create a new PR

🧪 Generate unit tests (beta)

• Commit to this branch
• Create a new PR

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

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

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 70.16%. Comparing base (6d76b59) to head (ced00bc).

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #2853   +/-   ##
=======================================
  Coverage   70.16%   70.16%           
=======================================
  Files         659      659           
  Lines       74081    74081           
=======================================
  Hits        51978    51978           
  Misses      22103    22103           

Flag	Coverage Δ
examples-torch_onnx	21.38% <ø> (-0.03%)	⬇️
examples-torch_trt	15.24% <ø> (-0.01%)	⬇️
unit	60.59% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[shengliangxu]

/ok to test 595189c

[shengliangxu]

/ok to test cae95d3

[cjluo-nv]

Bot review (bedrock-claude-opus-5-5) — DM the bot to share feedback.

Approving: the three cuts are correct, small (+14 lines), and match the repo's existing CI patterns.

No action needed.

No action needed:

• Dropping mp.set_start_method("spawn") is safe. The calibration loaders now use num_workers=0, and the --evaluate workers in examples/onnx_ptq/evaluation.py only run CPU-side PIL/torchvision transforms with no CUDA, so the default fork start method is fine.
• The PR-only skip uses the same refs/heads/pull-request/ check that _example_tests_runner.yml already uses for hf_ptq. Test ids are unchanged, scheduled runs keep every combination, and swin_tiny still covers every format on PRs.
• The trtexec flags --warmUp=0 --duration=0 --iterations=1 still run the engine once, and the script ignored the benchmark output anyway.
• No existing assertions were changed and there are no licensing changes.

[github-actions]

PR Preview Action v1.8.1
Preview removed because the pull request was closed.
2026-10-21 17:49 UTC