[fix] Cherry pick "[TRTLLM-6262] Fix Llama4 Scout FP4 crash issue" #6267

mikeiovine · 2025-07-22T20:56:40Z

Description

Cherry pick this fix from main. Original PR: #5834

The scale is unset, which will cause a bad_optional_access exception inside the custom op.

Test Coverage

Existing tests

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.

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.

Summary by CodeRabbit

New Features
- Improved attention mechanism with enhanced output scaling support in the Llama4 model.

Signed-off-by: Mike Iovine <[email protected]>

coderabbitai · 2025-07-22T20:56:46Z

Walkthrough

A new local variable, out_scale_sf, is introduced in the _forward_nope method of the Llama4Attention class. It is conditionally set based on two attributes and is passed as an extra keyword argument to a method call. No other logic or control flow is changed.

Changes

File	Change Summary
tensorrt_llm/_torch/models/modeling_llama.py	Added `out_scale_sf` variable in `_forward_nope` and passed it as a keyword argument to a method.

Estimated code review effort

1 (~2 minutes)

Poem

A bunny hops through lines anew,
With out_scale_sf in gentle view.
Condition checked, the call enhanced,
In llama code, the bits now danced.
A tiny tweak, a careful touch—
For forward flows that mean so much!

📜 Recent review details

Configuration used: .coderabbit.yaml
Review profile: CHILL
Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 1209001 and efae635.

📒 Files selected for processing (1)

tensorrt_llm/_torch/models/modeling_llama.py (1 hunks)

⏰ 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 (3)

tensorrt_llm/_torch/models/modeling_llama.py (3)

162-162: LGTM! Proper variable initialization.

The initialization of out_scale_sf to None follows the same pattern as the existing out_scale variable and ensures a safe default value.

166-167: Excellent fix for the FP4 crash issue.

This conditional assignment directly addresses the bad_optional_access exception mentioned in the PR objectives. When both NVFP4 quantization is enabled and NVFP4 output is supported, the scale factor is properly set from self.o_proj.input_scale, preventing the crash from an unset scale value.

175-175: LGTM! Completes the FP4 fix by passing the scale factor.

The addition of out_scale_sf=out_scale_sf as a keyword argument ensures the computed scale factor is properly passed to the attention forward method, completing the fix for the crash issue.

✨ Finishing Touches

📝 Generate Docstrings

🪧 Tips

Chat

There are 3 ways to chat with CodeRabbit:

Review comments: Directly reply to a review comment made by CodeRabbit. Example:
- I pushed a fix in commit <commit_id>, please review it.
- Explain this complex logic.
- Open a follow-up GitHub issue for this discussion.
Files and specific lines of code (under the "Files changed" tab): Tag @coderabbitai in a new review comment at the desired location with your query. Examples:
- @coderabbitai explain this code block.
- @coderabbitai modularize this function.
PR comments: Tag @coderabbitai in a new PR comment to ask questions about the PR branch. For the best results, please provide a very specific query, as very limited context is provided in this mode. Examples:
- @coderabbitai gather interesting stats about this repository and render them as a table. Additionally, render a pie chart showing the language distribution in the codebase.
- @coderabbitai read src/utils.ts and explain its main purpose.
- @coderabbitai read the files in the src/scheduler package and generate a class diagram using mermaid and a README in the markdown format.
- @coderabbitai help me debug CodeRabbit configuration file.

Support

Need help? Create a ticket on our support page for assistance with any issues or questions.

Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments.

CodeRabbit Commands (Invoked using PR comments)

@coderabbitai pause to pause the reviews on a PR.
@coderabbitai resume to resume the paused reviews.
@coderabbitai review to trigger an incremental review. This is useful when automatic reviews are disabled for the repository.
@coderabbitai full review to do a full review from scratch and review all the files again.
@coderabbitai summary to regenerate the summary of the PR.
@coderabbitai generate docstrings to generate docstrings for this PR.
@coderabbitai generate sequence diagram to generate a sequence diagram of the changes in this PR.
@coderabbitai resolve resolve all the CodeRabbit review comments.
@coderabbitai configuration to show the current CodeRabbit configuration for the repository.
@coderabbitai help to get help.

Other keywords and placeholders

Add @coderabbitai ignore anywhere in the PR description to prevent this PR from being reviewed.
Add @coderabbitai summary to generate the high-level summary at a specific location in the PR description.
Add @coderabbitai anywhere in the PR title to generate the title automatically.

Documentation and Community

Visit our Documentation for detailed information on how to use CodeRabbit.
Join our Discord Community to get help, request features, and share feedback.
Follow us on X/Twitter for updates and announcements.

mikeiovine · 2025-07-22T20:57:29Z

/bot run

tensorrt-cicd · 2025-07-22T21:02:56Z

PR_Github #12608 [ run ] triggered by Bot

tensorrt-cicd · 2025-07-23T01:58:15Z

PR_Github #12608 [ run ] completed with state SUCCESS
/LLM/release-0.21/L0_MergeRequest_PR pipeline #259 completed with status: 'FAILURE'

[fix] Cherry pick "[TRTLLM-6262] Fix Llama4 Scout FP4 crash issue"

efae635

Signed-off-by: Mike Iovine <[email protected]>

mikeiovine requested a review from yilin-void July 22, 2025 20:56

mikeiovine requested a review from a team as a code owner July 22, 2025 20:56

mikeiovine requested a review from symphonylyh July 22, 2025 20:56

litaotju approved these changes Jul 22, 2025

View reviewed changes

schetlur-nv approved these changes Jul 23, 2025

View reviewed changes

schetlur-nv merged commit a55c631 into NVIDIA:release/0.21 Jul 23, 2025
3 of 4 checks passed

mikeiovine deleted the fix-llama4-crash branch July 23, 2025 18:03

This was referenced Aug 14, 2025

[https://nvbugs/5455836][fix] Fix llama 4 FP4 #6911

Merged

[https://nvbugs/5427801][fix] Torch compile support for Llama4 and Ea… #6978

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

[fix] Cherry pick "[TRTLLM-6262] Fix Llama4 Scout FP4 crash issue" #6267

[fix] Cherry pick "[TRTLLM-6262] Fix Llama4 Scout FP4 crash issue" #6267

mikeiovine commented Jul 22, 2025 •

edited by coderabbitai bot

Loading

Uh oh!

coderabbitai bot commented Jul 22, 2025 •

edited

Loading

Chat

Support

CodeRabbit Commands (Invoked using PR comments)

Other keywords and placeholders

Documentation and Community

Uh oh!

mikeiovine commented Jul 22, 2025

Uh oh!

tensorrt-cicd commented Jul 22, 2025

Uh oh!

tensorrt-cicd commented Jul 23, 2025

Uh oh!

Uh oh!

Uh oh!

[fix] Cherry pick "[TRTLLM-6262] Fix Llama4 Scout FP4 crash issue" #6267

[fix] Cherry pick "[TRTLLM-6262] Fix Llama4 Scout FP4 crash issue" #6267

Conversation

mikeiovine commented Jul 22, 2025 • edited by coderabbitai bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Description

Test Coverage

GitHub Bot Help

kill

skip

reuse-pipeline

Summary by CodeRabbit

Uh oh!

coderabbitai bot commented Jul 22, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Walkthrough

Changes

Estimated code review effort

Poem

Chat

Support

CodeRabbit Commands (Invoked using PR comments)

Other keywords and placeholders

Documentation and Community

Uh oh!

mikeiovine commented Jul 22, 2025

Uh oh!

tensorrt-cicd commented Jul 22, 2025

Uh oh!

tensorrt-cicd commented Jul 23, 2025

Uh oh!

Uh oh!

Uh oh!

mikeiovine commented Jul 22, 2025 •

edited by coderabbitai bot

Loading

coderabbitai bot commented Jul 22, 2025 •

edited

Loading