[TRTLLM-6342][feat] Factory TP sharding of quantized models #8123
Conversation
Signed-off-by: greg-kwasniewski1 <[email protected]>
Signed-off-by: greg-kwasniewski1 <[email protected]>
📝 WalkthroughWalkthroughIntroduces a Starcoder2Config patch to set base_model_tp_plan for MLP projection. Extends node filtering to accept an iterable of predicates. Updates sharding detection to include fake-quantized linear ops and adds a fallback simple shard when encountering unsupported sharding actions, with adjusted log messages and early loop break preserved. Changes
Sequence Diagram(s)sequenceDiagram
autonumber
actor User
participant FactoryConfig
participant ShardingDetector as detect_sharding_from_factory_config
participant NodeFilter as filtered_nodes
participant LinNode as Linear/FQ Linear Node
participant Plan as TP Plan Builder
User->>FactoryConfig: Provide sharding factory config
FactoryConfig->>ShardingDetector: Start detection
ShardingDetector->>NodeFilter: Filter nodes by [is_linear_op, is_fake_quantized_linear_op]
loop For each filtered node
NodeFilter-->>ShardingDetector: LinNode
ShardingDetector->>ShardingDetector: Match config action for LinNode
alt Supported action
ShardingDetector->>Plan: Append specified shard transform
ShardingDetector-->>ShardingDetector: break
else Unsupported action
ShardingDetector->>ShardingDetector: Log "Unsupported sharding action {config}"
ShardingDetector->>Plan: Append simple shard (column split, all_gather, min_local_shape=1)
ShardingDetector-->>ShardingDetector: break
end
end
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~25 minutes Pre-merge checks and finishing touches❌ Failed checks (1 warning)
✅ Passed checks (2 passed)
✨ Finishing touches
🧪 Generate unit tests
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: 4
🧹 Nitpick comments (2)
tensorrt_llm/_torch/auto_deploy/models/patches/starcoder.py (1)
4-4: Consider the implications of mutating a class-level attribute.Direct mutation of
Starcoder2Config.base_model_tp_planmodifies a global class attribute that may affect all instances. Ensure this side effect is intentional and that the patch is applied early enough in the initialization flow to avoid race conditions or unexpected behavior in concurrent usage.tensorrt_llm/_torch/auto_deploy/transform/library/sharding.py (1)
390-402: Fallback to simple shard improves resilience.Adding a default simple shard for unrecognized config values prevents silent failures and provides predictable behavior. However, ensure that a simple shard (column split + all_gather) is semantically appropriate for all potential future config values, or document the expected behavior.
Consider adding a validation step earlier in the pipeline or documenting the fallback behavior in the function docstring to clarify the assumptions.
📜 Review details
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (3)
tensorrt_llm/_torch/auto_deploy/models/patches/starcoder.py(1 hunks)tensorrt_llm/_torch/auto_deploy/transform/library/sharding.py(3 hunks)tensorrt_llm/_torch/auto_deploy/utils/node_utils.py(1 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/auto_deploy/utils/node_utils.pytensorrt_llm/_torch/auto_deploy/models/patches/starcoder.pytensorrt_llm/_torch/auto_deploy/transform/library/sharding.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/auto_deploy/utils/node_utils.pytensorrt_llm/_torch/auto_deploy/models/patches/starcoder.pytensorrt_llm/_torch/auto_deploy/transform/library/sharding.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/auto_deploy/utils/node_utils.pytensorrt_llm/_torch/auto_deploy/models/patches/starcoder.pytensorrt_llm/_torch/auto_deploy/transform/library/sharding.py
🧬 Code graph analysis (2)
tensorrt_llm/_torch/auto_deploy/utils/node_utils.py (1)
tensorrt_llm/_torch/auto_deploy/transform/library/fused_moe.py (1)
target(359-360)
tensorrt_llm/_torch/auto_deploy/transform/library/sharding.py (2)
tensorrt_llm/_torch/auto_deploy/utils/node_utils.py (3)
filtered_nodes(205-252)is_linear_op(255-265)is_fake_quantized_linear_op(268-274)tensorrt_llm/_torch/auto_deploy/utils/sharding_utils.py (4)
TPShardingInfo(245-288)from_node(253-258)from_node(753-758)SplitDimension(198-206)
⏰ 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/auto_deploy/transform/library/sharding.py (1)
371-371: LGTM: Warning message for unsupported local sharding action.The warning message clearly indicates that the unsupported sharding action is being skipped, which is appropriate for the local-shard branch where no fallback is applied.
Signed-off-by: greg-kwasniewski1 <[email protected]>
|
/bot run |
|
PR_Github #21103 [ run ] triggered by Bot |
|
PR_Github #21103 [ run ] completed with state |
|
/bot run |
|
PR_Github #21105 [ run ] triggered by Bot |
|
PR_Github #21105 [ run ] completed with state |
|
/bot run --reuse-pipeline |
|
PR_Github #21237 Bot args parsing error: usage: /bot [-h] |
|
/bot reuse-pipeline |
|
PR_Github #21249 [ reuse-pipeline ] triggered by Bot |
|
PR_Github #21249 [ reuse-pipeline ] completed with state |
) Signed-off-by: greg-kwasniewski1 <[email protected]> Co-authored-by: Suyog Gupta <[email protected]>
) Signed-off-by: greg-kwasniewski1 <[email protected]> Co-authored-by: Suyog Gupta <[email protected]> Signed-off-by: yufeiwu-nv <[email protected]>
) Signed-off-by: greg-kwasniewski1 <[email protected]> Co-authored-by: Suyog Gupta <[email protected]>
) Signed-off-by: greg-kwasniewski1 <[email protected]> Co-authored-by: Suyog Gupta <[email protected]>
) Signed-off-by: greg-kwasniewski1 <[email protected]> Co-authored-by: Suyog Gupta <[email protected]>
) Signed-off-by: greg-kwasniewski1 <[email protected]> Co-authored-by: Suyog Gupta <[email protected]>
3 participants
Description
Previously,
detect_sharding_from_factory_configapplied sharding strategy to all linear nodes filtered byis_linear_op:is_linear_opdoes not capture quantized linear nodes such astorch.ops.auto_deploy.torch_fake_quant_fp8_linear.This PR extents
detect_sharding_from_factory_configto:An update to
filtered_nodeslogic was needed to achieve it.Summary by CodeRabbit
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.
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.