[https://nvbugs/5556020][fix] test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_eagle3 dimension mismatch #8517

sunnyqgg · 2025-10-21T03:54:43Z

… py_seq_slot

Summary by CodeRabbit

Refactor
- Optimized token indexing logic for improved alignment during speculative decoding.
Tests
- Enabled previously skipped test to expand test coverage for disaggregated serving scenarios.

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)
Any new dependencies have been scanned for license and vulnerabilities
CODEOWNERS updated if ownership changes
Documentation updated as needed
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

/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)]

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. Examples: "A10-PyTorch-1, xxx". 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. Examples: --extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx".

--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.

For guidance on mapping tests to stage names, see docs/source/reference/ci-overview.md
and the scripts/test_to_stage_mapping.py helper.

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.

… py_seq_slot Signed-off-by: qgai <qgai@nvidia.com>

sunnyqgg · 2025-10-21T03:56:16Z

/bot run --disable-fail-fast

coderabbitai · 2025-10-21T03:58:50Z

📝 Walkthrough

Walkthrough

Two changes: (1) Modified draft/target token indexing in model_drafter.py to use py_seq_slot instead of py_batch_idx, altering token alignment from batch-dimension to sequence-slot-dimension mapping; (2) Removed skip decorator from test_ngram in the test suite, enabling it to run.

Changes

Cohort / File(s)	Change Summary
Speculative model token indexing `tensorrt_llm/_torch/speculative/model_drafter.py`	Replaced draft_idx and target_idx indexing logic to use `request.py_seq_slot` instead of `request.py_batch_idx` when not in static draft loop, changing token placement from batch-based to sequence-slot-based alignment during speculative drafting.
Test suite enabling `tests/integration/defs/accuracy/test_disaggregated_serving.py`	Removed skip decorator from test_ngram method, allowing test to execute as part of the test suite.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~12 minutes

Pre-merge checks and finishing touches

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Description Check	⚠️ Warning	The PR description is substantially incomplete against the required template. The author has provided only a brief commit message ("… py_seq_slot") and left the critical sections empty: the Description section (which should explain the issue and solution) and the Test Coverage section (which should list relevant tests) are both blank. While the PR checklist is marked as complete with an [x], this appears premature given that the description lacks any actual explanation of what was changed and why, or documentation of test coverage. The author referenced "@coderabbitai summary" but did not include the generated summary or provide a human-written alternative explanation.	The author should provide a substantive Description section explaining the dimension mismatch issue, the root cause, and how the py_seq_slot indexing change fixes it. Additionally, the Test Coverage section must clearly list which tests validate these changes (the test_ngram test that was unskipped and any tests that verify the model_drafter.py indexing fix). The PR checklist should only be marked complete after these required sections are properly filled out.

✅ Passed checks (2 passed)

Check name	Status	Explanation
Title Check	✅ Passed	The PR title follows the required template format with a valid NVBugs ID, type indicator [fix], and a clear summary of the primary issue being addressed: a dimension mismatch in test_eagle3. While the title describes the bug being fixed rather than the specific implementation details (using py_seq_slot for token mapping), it is sufficiently clear and directly related to the main purpose of the changeset. A developer reviewing the history would understand that this PR addresses a dimension mismatch issue. The title is not misleading, though it could be more technically specific.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.

✨ Finishing touches

📝 Generate docstrings

🧪 Generate unit tests (beta)

Create PR with unit tests
Post copyable unit tests in a comment

📜 Recent review details

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between ab4b996 and 218b8af.

📒 Files selected for processing (2)

tensorrt_llm/_torch/speculative/model_drafter.py (1 hunks)
tests/integration/defs/accuracy/test_disaggregated_serving.py (0 hunks)

💤 Files with no reviewable changes (1)

tests/integration/defs/accuracy/test_disaggregated_serving.py

🧰 Additional context used

📓 Path-based instructions (3)

**/*.{h,hpp,hh,hxx,cpp,cxx,cc,cu,cuh,py}

📄 CodeRabbit inference engine (CODING_GUIDELINES.md)

Use only spaces, no tabs; indent with 4 spaces.

Files:

tensorrt_llm/_torch/speculative/model_drafter.py

**/*.py

📄 CodeRabbit inference engine (CODING_GUIDELINES.md)

**/*.py: Python code must target Python 3.8+.
Indent Python code with 4 spaces; do not use tabs.
Maintain module namespace when importing; prefer 'from package.subpackage import foo' then 'foo.SomeClass()' instead of importing the class directly.
Python filenames should be snake_case (e.g., some_file.py).
Python classes use PascalCase names.
Functions and methods use snake_case names.
Local variables use snake_case; prefix 'k' for variables that start with a number (e.g., k_99th_percentile).
Global variables use upper SNAKE_CASE prefixed with 'G' (e.g., G_MY_GLOBAL).
Constants use upper SNAKE_CASE (e.g., MY_CONSTANT).
Avoid shadowing variables from an outer scope.
Initialize all externally visible members of a class in the constructor.
Prefer docstrings for interfaces that may be used outside a file; comments for in-function or file-local interfaces.
Use Google-style docstrings for classes and functions (Sphinx-parsable).
Document attributes and variables inline so they render under the class/function docstring.
Avoid reflection when a simpler, explicit approach suffices (e.g., avoid dict(**locals()) patterns).
In try/except, catch the most specific exceptions possible.
For duck-typing try/except, keep the try body minimal and use else for the main logic.

Files:

tensorrt_llm/_torch/speculative/model_drafter.py

**/*.{cpp,cxx,cc,h,hpp,hh,hxx,cu,cuh,py}

📄 CodeRabbit inference engine (CODING_GUIDELINES.md)

Prepend the NVIDIA Apache-2.0 copyright header with current year to the top of all source files (e.g., .cpp, .h, .cu, .py).

Files:

tensorrt_llm/_torch/speculative/model_drafter.py

⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (1)

GitHub Check: Pre-commit Check

🔇 Additional comments (1)

tensorrt_llm/_torch/speculative/model_drafter.py (1)

502-504: Fix is correct and complete - dimension mismatch properly addressed.

Verification confirms the changes at lines 502-504 correctly resolve the dimension mismatch. The fix properly distinguishes between static and dynamic draft loops:

Dynamic mode (line 502): Uses request.py_seq_slot for tensor indexing, matching sequence slot assignment

Static mode (line 502): Uses req_idx from batch enumeration

Static-only method process_static_draft_outputs (lines 568-569): Correctly uses req_idx throughout since it only executes in static mode

The tensor indexing is now consistent across all code paths. No py_batch_idx remains in critical indexing operations, and the codebase properly separates static vs. dynamic indexing strategies.

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

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

tensorrt-cicd · 2025-10-21T04:01:26Z

PR_Github #21994 [ run ] triggered by Bot. Commit: 218b8af

tensorrt-cicd · 2025-10-21T08:01:28Z

PR_Github #21994 [ run ] completed with state SUCCESS. Commit: 218b8af
/LLM/main/L0_MergeRequest_PR pipeline #16584 completed with status: 'FAILURE'

sunnyqgg · 2025-10-21T08:09:45Z

/bot run --disable-fail-fast

tensorrt-cicd · 2025-10-21T08:17:43Z

PR_Github #22030 [ run ] triggered by Bot. Commit: 218b8af

tensorrt-cicd · 2025-10-21T09:46:54Z

PR_Github #22030 [ run ] completed with state SUCCESS. Commit: 218b8af
/LLM/main/L0_MergeRequest_PR pipeline #16610 completed with status: 'SUCCESS'
Pipeline passed with automatic retried tests. Check the rerun report for details.

…a3_1_8BInstruct::test_eagle3 dimension mismatch (NVIDIA#8517) Signed-off-by: qgai <qgai@nvidia.com> Signed-off-by: yufeiwu-nv <230315618+yufeiwu-nv@users.noreply.github.com>

…a3_1_8BInstruct::test_eagle3 dimension mismatch (NVIDIA#8517) Signed-off-by: qgai <qgai@nvidia.com>

request.py_batch_idx is None under disaggreated model so change it to…

218b8af

… py_seq_slot Signed-off-by: qgai <qgai@nvidia.com>

sunnyqgg requested a review from a team as a code owner October 21, 2025 03:54

sunnyqgg requested a review from zheyuf October 21, 2025 03:54

sunnyqgg requested review from mikeiovine and ziyixiong-nv October 21, 2025 03:57

ziyixiong-nv approved these changes Oct 21, 2025

View reviewed changes

mikeiovine approved these changes Oct 21, 2025

View reviewed changes

zheyuf approved these changes Oct 22, 2025

View reviewed changes

longlee0622 merged commit 90080e0 into NVIDIA:main Oct 22, 2025
8 of 9 checks passed

dominicshanshan pushed a commit to dominicshanshan/TensorRT-LLM that referenced this pull request Nov 1, 2025

[https://nvbugs/5556020][fix] test_disaggregated_serving.py::TestLlam…

0156f69

…a3_1_8BInstruct::test_eagle3 dimension mismatch (NVIDIA#8517) Signed-off-by: qgai <qgai@nvidia.com>

dominicshanshan pushed a commit to dominicshanshan/TensorRT-LLM that referenced this pull request Nov 3, 2025

[https://nvbugs/5556020][fix] test_disaggregated_serving.py::TestLlam…

81da222

…a3_1_8BInstruct::test_eagle3 dimension mismatch (NVIDIA#8517) Signed-off-by: qgai <qgai@nvidia.com>

dominicshanshan pushed a commit to dominicshanshan/TensorRT-LLM that referenced this pull request Nov 3, 2025

[https://nvbugs/5556020][fix] test_disaggregated_serving.py::TestLlam…

9ed8113

…a3_1_8BInstruct::test_eagle3 dimension mismatch (NVIDIA#8517) Signed-off-by: qgai <qgai@nvidia.com>

dominicshanshan pushed a commit to dominicshanshan/TensorRT-LLM that referenced this pull request Nov 3, 2025

[https://nvbugs/5556020][fix] test_disaggregated_serving.py::TestLlam…

2f810c8

…a3_1_8BInstruct::test_eagle3 dimension mismatch (NVIDIA#8517) Signed-off-by: qgai <qgai@nvidia.com>

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

[https://nvbugs/5556020][fix] test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_eagle3 dimension mismatch #8517

[https://nvbugs/5556020][fix] test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_eagle3 dimension mismatch #8517

Uh oh!

sunnyqgg commented Oct 21, 2025 •

edited by coderabbitai bot

Loading

Uh oh!

sunnyqgg commented Oct 21, 2025

Uh oh!

coderabbitai bot commented Oct 21, 2025

Walkthrough

Changes

Estimated code review effort

Uh oh!

tensorrt-cicd commented Oct 21, 2025

Uh oh!

tensorrt-cicd commented Oct 21, 2025

Uh oh!

sunnyqgg commented Oct 21, 2025

Uh oh!

tensorrt-cicd commented Oct 21, 2025

Uh oh!

tensorrt-cicd commented Oct 21, 2025

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

[https://nvbugs/5556020][fix] test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_eagle3 dimension mismatch #8517

[https://nvbugs/5556020][fix] test_disaggregated_serving.py::TestLlama3_1_8BInstruct::test_eagle3 dimension mismatch #8517

Uh oh!

Conversation

sunnyqgg commented Oct 21, 2025 • 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

kill

skip

reuse-pipeline

Uh oh!

sunnyqgg commented Oct 21, 2025

Uh oh!

coderabbitai bot commented Oct 21, 2025

Walkthrough

Changes

Estimated code review effort

Pre-merge checks and finishing touches

Uh oh!

tensorrt-cicd commented Oct 21, 2025

Uh oh!

tensorrt-cicd commented Oct 21, 2025

Uh oh!

sunnyqgg commented Oct 21, 2025

Uh oh!

tensorrt-cicd commented Oct 21, 2025

Uh oh!

tensorrt-cicd commented Oct 21, 2025

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

sunnyqgg commented Oct 21, 2025 •

edited by coderabbitai bot

Loading