[TRTLLM-12173][tests] Add E2E accuracy test for nano v3 omni by 2ez4bz · Pull Request #13616 · NVIDIA/TensorRT-LLM

2ez4bz · 2026-04-29T17:13:19Z

Description

As title states.

Test Coverage

Added new E2E MMMU accuracy test for BF16 / FP8 / NVFP4.

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)
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.

Summary by CodeRabbit

New Features
- Extended multimodal model testing support for Nemotron-3-Nano-Omni models across BF16, FP8, and NVFP4 quantization configurations.
Tests
- Added comprehensive accuracy test coverage for multimodal models on H100 and B200 GPUs with multiple precision configurations.

2ez4bz · 2026-04-29T17:13:31Z

/bot help

github-actions · 2026-04-29T17:13:40Z

GitHub Bot Help

/bot [-h] ['run', 'kill', 'skip', 'reuse-pipeline'] ...

Provide a user friendly way for developers to interact with a Jenkins server.

Run /bot [-h|--help] to print this help message.

See details below for each supported subcommand.

Details

run [--reuse-test (optional)pipeline-id --disable-fail-fast --skip-test --stage-list "A10-PyTorch-1, xxx" --gpu-type "A30, H100_PCIe" --test-backend "pytorch, cpp" --add-multi-gpu-test --only-multi-gpu-test --disable-multi-gpu-test --post-merge --extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx" --detailed-log --debug(experimental) --high-priority]

Launch build/test pipelines. All previously running jobs will be killed.

--reuse-test (optional)pipeline-id (OPTIONAL) : Allow the new pipeline to reuse build artifacts and skip successful test stages from a specified pipeline or the last pipeline if no pipeline-id is indicated. If the Git commit ID has changed, this option will be always ignored. The DEFAULT behavior of the bot is to reuse build artifacts and successful test results from the last pipeline.

--disable-reuse-test (OPTIONAL) : Explicitly prevent the pipeline from reusing build artifacts and skipping successful test stages from a previous pipeline. Ensure that all builds and tests are run regardless of previous successes.

--disable-fail-fast (OPTIONAL) : Disable fail fast on build/tests/infra failures.

--skip-test (OPTIONAL) : Skip all test stages, but still run build stages, package stages and sanity check stages. Note: Does NOT update GitHub check status.

--stage-list "A10-PyTorch-1, xxx" (OPTIONAL) : Only run the specified test stages. Supports wildcard * for pattern matching (e.g., "*PerfSanity*" matches all stages containing PerfSanity). Examples: "A10-PyTorch-1, xxx", "PerfSanity". Note: Does NOT update GitHub check status.

--gpu-type "A30, H100_PCIe" (OPTIONAL) : Only run the test stages on the specified GPU types. Examples: "A30, H100_PCIe". Note: Does NOT update GitHub check status.

--test-backend "pytorch, cpp" (OPTIONAL) : Skip test stages which don't match the specified backends. Only support [pytorch, cpp, tensorrt, triton]. Examples: "pytorch, cpp" (does not run test stages with tensorrt or triton backend). Note: Does NOT update GitHub pipeline status.

--only-multi-gpu-test (OPTIONAL) : Only run the multi-GPU tests. Note: Does NOT update GitHub check status.

--disable-multi-gpu-test (OPTIONAL) : Disable the multi-GPU tests. Note: Does NOT update GitHub check status.

--add-multi-gpu-test (OPTIONAL) : Force run the multi-GPU tests in addition to running L0 pre-merge pipeline.

--post-merge (OPTIONAL) : Run the L0 post-merge pipeline instead of the ordinary L0 pre-merge pipeline.

--extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx" (OPTIONAL) : Run the ordinary L0 pre-merge pipeline and specified test stages. Supports wildcard * for pattern matching. Examples: --extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx", --extra-stage "Post-Merge".

--detailed-log (OPTIONAL) : Enable flushing out all logs to the Jenkins console. This will significantly increase the log volume and may slow down the job.

--debug (OPTIONAL) : Experimental feature. Enable access to the CI container for debugging purpose. Note: Specify exactly one stage in the stage-list parameter to access the appropriate container environment. Note: Does NOT update GitHub check status.

--high-priority (OPTIONAL) : Run the pipeline with high priority. This option is restricted to authorized users only and will route the job to a high-priority queue.

kill

kill

Kill all running builds associated with pull request.

skip

skip --comment COMMENT

Skip testing for latest commit on pull request. --comment "Reason for skipping build/test" is required. IMPORTANT NOTE: This is dangerous since lack of user care and validation can cause top of tree to break.

reuse-pipeline

reuse-pipeline

Reuse a previous pipeline to validate current commit. This action will also kill all currently running builds associated with the pull request. IMPORTANT NOTE: This is dangerous since lack of user care and validation can cause top of tree to break.

2ez4bz · 2026-04-29T17:18:55Z

/bot run

coderabbitai · 2026-04-29T17:19:59Z

📝 Walkthrough

Walkthrough

Extends multimodal LLM evaluation by modifying MultimodalLmEvalWrapper to forward additional parameters (chat_template_kwargs, model_type, is_force_single_image) to the base class constructor. Introduces a new test suite for Nemotron-3-Nano-Omni models across BF16, FP8, and NVFP4 quantization configurations with platform-specific GPU constraints and deterministic sampling settings.

Changes

Cohort / File(s)	Summary
Wrapper Parameter Forwarding `tensorrt_llm/evaluate/lm_eval.py`	Modified `MultimodalLmEvalWrapper.__init__` to explicitly pass `chat_template_kwargs`, `model_type`, and `is_force_single_image` to base class constructor alongside existing parameters.
Model Reference Documentation `tests/integration/defs/accuracy/references/mmmu.yaml`	Updated MMMU test comments, broadened Nemotron model specification, and added accuracy reference entries for Nemotron-3-Nano-Omni-30B-A3B-Reasoning with FP8 and NVFP4 quantization variants.
Nemotron Model Test Suite `tests/integration/defs/accuracy/test_llm_api_pytorch_multimodal.py`	Added `TestNanoV3Omni` class with parametrized test covering BF16, FP8, and NVFP4 configurations, deterministic sampling, chunked prefill optimization, and GPU-generation-specific skip markers.
Test Selection Lists `tests/integration/test_lists/qa/llm_function_core.txt`, `tests/integration/test_lists/test-db/l0_b200.yml`, `tests/integration/test_lists/test-db/l0_h100.yml`	Added `TestNanoV3Omni::test_auto_dtype` test entries for BF16, FP8, and NVFP4 variants across multiple GPU-specific test selection profiles (QA core, B200, H100).

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~20 minutes

🚥 Pre-merge checks | ✅ 3 | ❌ 2

❌ Failed checks (1 warning, 1 inconclusive)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 33.33% 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 description lacks substantial detail; it states 'As title states' instead of explaining the purpose, implementation, and test coverage comprehensively.	Expand the description section to explain what the test does, why it's needed, the test configurations (BF16/FP8/NVFP4), and implementation details beyond just repeating the title.

✅ Passed checks (3 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly and specifically summarizes the main change: adding an end-to-end accuracy test for the nano v3 omni model.
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

_{Review rate limit: 9/10 reviews remaining, refill in 6 minutes.}

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

coderabbitai

Actionable comments posted: 3

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Inline comments:
In `@tests/integration/defs/accuracy/references/mmmu.yaml`:
- Line 25: Remove the unresolved merge-time question comment "# Can this be
removed?" from the canonical reference data and either delete the comment
entirely or replace it with a clear, actionable placeholder linking to a tracked
issue (e.g., "TODO: remove after ISSUE-1234 - <issue URL>") so the file contains
no ambiguity; update the corresponding reference entry so it remains valid and
commit the change.

In `@tests/integration/defs/accuracy/test_llm_api_pytorch_multimodal.py`:
- Line 444: The placeholder comment "# Note: these need to be changed to refer
to the models on scratch before merging." is an outstanding pre-merge TODO and
must be resolved: either replace the comment with the actual scratch model
identifiers used in this test (update the model name strings referenced in the
test setup/fixtures so tests point to the correct models) or remove the note and
add a concise tracked-issue reference (e.g., "TODO: track in ISSUE-xxx") if the
change will be done later; locate the placeholder comment in
tests/integration/defs/accuracy/test_llm_api_pytorch_multimodal.py and update
the surrounding test fixture/model variables (the test's model name constants)
accordingly so no pre-merge action item remains.
- Around line 445-480: MODEL_PARAMS is defined as a mutable list (and other
nearby mutable containers) which triggers RUF012; change those to immutable
containers by replacing the list with a tuple of pytest.param entries and
convert any mutable dict-like configs to immutable mappings (e.g.,
types.MappingProxyType or frozen dataclass/tuple) so tests cannot mutate shared
state; update the definition of MODEL_PARAMS and any related config holders
(referencing MODEL_PARAMS and KvCacheConfig usages) to use tuple/immutable
mapping equivalents and ensure all callers expecting sequence still work with a
tuple.

🪄 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: 8fe1b4d6-64f6-4142-8fb1-070f603fe3d9

📥 Commits

Reviewing files that changed from the base of the PR and between 3b7af1c and 5b4a104.

📒 Files selected for processing (6)

tensorrt_llm/evaluate/lm_eval.py
tests/integration/defs/accuracy/references/mmmu.yaml
tests/integration/defs/accuracy/test_llm_api_pytorch_multimodal.py
tests/integration/test_lists/qa/llm_function_core.txt
tests/integration/test_lists/test-db/l0_b200.yml
tests/integration/test_lists/test-db/l0_h100.yml

2ez4bz · 2026-04-29T17:22:00Z

/bot kill

tensorrt-cicd · 2026-04-29T17:25:27Z

PR_Github #46194 [ run ] triggered by Bot. Commit: 5b4a104 Link to invocation

tensorrt-cicd · 2026-04-29T17:28:51Z

PR_Github #46197 [ kill ] triggered by Bot. Commit: 5b4a104 Link to invocation

tensorrt-cicd · 2026-04-29T17:28:53Z

PR_Github #46194 [ run ] completed with state ABORTED. Commit: 5b4a104

Link to invocation

tensorrt-cicd · 2026-04-29T17:29:24Z

PR_Github #46197 [ kill ] completed with state SUCCESS. Commit: 5b4a104
Successfully killed previous jobs for commit 5b4a104

Link to invocation

2ez4bz · 2026-04-29T18:59:18Z

/bot run

tensorrt-cicd · 2026-04-29T19:05:43Z

PR_Github #46204 [ run ] triggered by Bot. Commit: 76688c5 Link to invocation

tensorrt-cicd · 2026-04-29T23:34:50Z

PR_Github #46204 [ run ] completed with state SUCCESS. Commit: 76688c5
/LLM/main/L0_MergeRequest_PR pipeline #36317 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

2ez4bz · 2026-04-29T23:42:19Z

/bot run --disable-fail-fast

tensorrt-cicd · 2026-04-29T23:53:33Z

PR_Github #46226 [ run ] triggered by Bot. Commit: 76688c5 Link to invocation

xinhe-nv · 2026-04-30T01:18:47Z

@coderabbitai review the code as a QA engineer.

tensorrt-cicd · 2026-05-05T09:17:25Z

PR_Github #46744 [ run ] completed with state SUCCESS. Commit: bee433d
/LLM/main/L0_MergeRequest_PR pipeline #36774 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

Link to invocation

2ez4bz · 2026-05-05T16:26:23Z

/bot run

tensorrt-cicd · 2026-05-05T16:32:27Z

PR_Github #46833 [ run ] triggered by Bot. Commit: bee433d Link to invocation

2ez4bz · 2026-05-05T21:08:44Z

/bot help

github-actions · 2026-05-05T21:09:05Z