[https://nvbugs/5569713][fix] Disable fp8 deep gemm for EXAONE-4.0-32B-FP8 #8429
Conversation
📝 WalkthroughWalkthroughThis PR adds a Changes
Sequence DiagramsequenceDiagram
actor Init as Initialization
participant DL as Exaone4DecoderLayer
participant QC as quant_config
participant Attn as Exaone4Attention
participant MLP as GatedMLP
Init->>DL: __init__(quant_config, ...)
DL->>QC: check quant_algo
alt quant_algo == FP8_BLOCK_SCALES
QC-->>DL: True
DL->>DL: disable_deep_gemm = True
else Other algo
QC-->>DL: False
DL->>DL: disable_deep_gemm = False
end
DL->>Attn: __init__(..., disable_deep_gemm)
DL->>MLP: __init__(..., disable_deep_gemm)
Attn-->>DL: initialized
MLP-->>DL: initialized
Estimated code review effort🎯 2 (Simple) | ⏱️ ~12 minutes The changes involve straightforward parameter additions and propagation within a single file. The logic is cohesive and localized to configuration-driven initialization flows. Moderate complexity arises from understanding the quantization algorithm condition and how the flag affects attention behavior, but the edits are consistent and follow a clear pattern. Pre-merge checks and finishing touches❌ Failed checks (2 warnings)
✅ Passed checks (1 passed)
✨ Finishing touches
🧪 Generate unit tests (beta)
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. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
📜 Review details
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (1)
tensorrt_llm/_torch/models/modeling_exaone4.py(5 hunks)
🧰 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/models/modeling_exaone4.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/models/modeling_exaone4.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/models/modeling_exaone4.py
🧬 Code graph analysis (1)
tensorrt_llm/_torch/models/modeling_exaone4.py (1)
tensorrt_llm/quantization/mode.py (1)
QuantAlgo(23-47)
⏰ 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 (4)
tensorrt_llm/_torch/models/modeling_exaone4.py (4)
58-59: LGTM! Parameter addition is correct.The
disable_deep_gemmparameter is properly added with a sensible default ofFalsefor backward compatibility and correctly forwarded to the parent class. This enables selective disabling of deep_gemm for specific quantization configurations.Also applies to: 93-93
134-139: LGTM! Logic correctly identifies when to disable deep_gemm.The implementation correctly:
- Uses defensive programming with
getattrto safely access potentially missing attributes- Specifically targets
FP8_BLOCK_SCALESquantization algorithm as mentioned in the PR objective- Includes a clear comment explaining the workaround for the CUDA illegal memory access issue
144-144: LGTM! Parameter correctly propagated to Exaone4Attention.The
disable_deep_gemmflag is correctly passed toExaone4Attention, which we confirmed accepts this parameter.
154-154: No action needed: GatedMLP init already includesdisable_deep_gemm.
|
/bot run |
|
PR_Github #21580 [ run ] triggered by Bot |
|
PR_Github #21580 [ run ] completed with state |
yechank-nvidia
left a comment
yechank-nvidia
left a comment
There was a problem hiding this comment.
LGTM. Left one question below.
|
/bot run |
|
PR_Github #21651 [ run ] triggered by Bot |
|
PR_Github #21651 [ run ] completed with state |
|
/bot run |
|
PR_Github #21979 [ run ] triggered by Bot. Commit: |
|
PR_Github #21979 [ run ] completed with state |
…B-FP8 Signed-off-by: Junyi Xu <219237550+JunyiXu-nv@users.noreply.github.com>
Signed-off-by: Junyi Xu <219237550+JunyiXu-nv@users.noreply.github.com>
Signed-off-by: Junyi Xu <219237550+JunyiXu-nv@users.noreply.github.com>
Signed-off-by: Junyi Xu <219237550+JunyiXu-nv@users.noreply.github.com>
Signed-off-by: Junyi Xu <219237550+JunyiXu-nv@users.noreply.github.com>
JunyiXu-nv
force-pushed
the
dev-junyi-bug-5569713
branch
from
3cc3a3d to
6173df7
Compare
|
/bot run |
|
PR_Github #22061 [ run ] triggered by Bot. Commit: |
|
PR_Github #22061 [ run ] completed with state |
…B-FP8 (NVIDIA#8429) Signed-off-by: Junyi Xu <219237550+JunyiXu-nv@users.noreply.github.com>
…B-FP8 (NVIDIA#8429) Signed-off-by: Junyi Xu <219237550+JunyiXu-nv@users.noreply.github.com> Signed-off-by: Mike Iovine <6158008+mikeiovine@users.noreply.github.com>
…B-FP8 (NVIDIA#8429) Signed-off-by: Junyi Xu <219237550+JunyiXu-nv@users.noreply.github.com> Signed-off-by: Mike Iovine <6158008+mikeiovine@users.noreply.github.com>
…B-FP8 (NVIDIA#8429) Signed-off-by: Junyi Xu <219237550+JunyiXu-nv@users.noreply.github.com> Signed-off-by: Mike Iovine <6158008+mikeiovine@users.noreply.github.com>
…B-FP8 (NVIDIA#8429) Signed-off-by: Junyi Xu <219237550+JunyiXu-nv@users.noreply.github.com> Signed-off-by: Mike Iovine <6158008+mikeiovine@users.noreply.github.com>
…B-FP8 (NVIDIA#8429) Signed-off-by: Junyi Xu <219237550+JunyiXu-nv@users.noreply.github.com> Signed-off-by: Mike Iovine <6158008+mikeiovine@users.noreply.github.com>
…B-FP8 (NVIDIA#8429) Signed-off-by: Junyi Xu <219237550+JunyiXu-nv@users.noreply.github.com> Signed-off-by: Mike Iovine <6158008+mikeiovine@users.noreply.github.com>
…B-FP8 (NVIDIA#8429) Signed-off-by: Junyi Xu <219237550+JunyiXu-nv@users.noreply.github.com> Signed-off-by: Mike Iovine <6158008+mikeiovine@users.noreply.github.com>
…B-FP8 (NVIDIA#8429) Signed-off-by: Junyi Xu <219237550+JunyiXu-nv@users.noreply.github.com> Signed-off-by: Mike Iovine <6158008+mikeiovine@users.noreply.github.com>
…B-FP8 (NVIDIA#8429) Signed-off-by: Junyi Xu <219237550+JunyiXu-nv@users.noreply.github.com> Signed-off-by: Mike Iovine <6158008+mikeiovine@users.noreply.github.com>
…B-FP8 (NVIDIA#8429) Signed-off-by: Junyi Xu <219237550+JunyiXu-nv@users.noreply.github.com> Signed-off-by: Mike Iovine <6158008+mikeiovine@users.noreply.github.com> Signed-off-by: Mike Iovine <miovine@nvidia.com>
…B-FP8 (#8429) Signed-off-by: Junyi Xu <219237550+JunyiXu-nv@users.noreply.github.com> Signed-off-by: Mike Iovine <6158008+mikeiovine@users.noreply.github.com> Signed-off-by: Mike Iovine <miovine@nvidia.com>
4 participants
Summary by CodeRabbit
disable_deep_gemmconfiguration option in the Exaone4 model, which automatically activates when using FP8 block scales quantization. This option controls computation behavior in the attention and feedforward components, enabling optimized performance with specific quantization settings.Description
CUDA illegal memory access issue appears when using deep_gemm fp8 gemm in EXAONE-4.0-32B-FP8.
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 thestage-listparameter 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.mdand the
scripts/test_to_stage_mapping.pyhelper.kill
killKill all running builds associated with pull request.
skip
skip --comment COMMENTSkip 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-pipelineReuse 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.