Bugfix/fix nemotron nas lora support #6380
Conversation
📝 WalkthroughWalkthroughThis change simplifies the logic for configuring the number of key-value attention heads in model configuration and LoRA handling. It streamlines per-layer and uniform KV head assignment, updates LoRA compatibility checks, and re-enables a previously skipped LoRA test for the Nemotron model. Changes
Sequence Diagram(s)sequenceDiagram
participant User
participant ModelConfig
participant ModelConfigCpp
User->>ModelConfig: get_bindings_model_config(...)
ModelConfig->>ModelConfigCpp: set num_key_value_heads (per-layer or uniform)
Note right of ModelConfigCpp: Applies TP/CP scaling as needed
sequenceDiagram
participant PyExecutorUtil
participant ModelConfigCpp
PyExecutorUtil->>ModelConfigCpp: get num_kv_heads_per_layer
alt per-layer KV heads (non-uniform)
PyExecutorUtil->>PyExecutorUtil: log warning, use max KV heads
else uniform KV heads
PyExecutorUtil->>PyExecutorUtil: use uniform KV heads
end
Estimated code review effort🎯 2 (Simple) | ⏱️ ~8 minutes Possibly related PRs
Suggested reviewers
Note ⚡️ Unit Test Generation is now available in beta!Learn more here, or try it out under "Finishing Touches" below. ✨ 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. 🪧 TipsChatThere are 3 ways to chat with CodeRabbit:
SupportNeed 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)
Other keywords and placeholders
Documentation and Community
|
|
@coderabbitai help |
CodeRabbit Commands (Invoked using PR comments)
|
|
@coderabbitai review |
✅ Actions performedReview triggered.
|
|
Caution An unexpected error occurred while opening a pull request: Reference update failed - https://docs.github.com/rest/git/refs#create-a-reference |
|
/bot run |
|
PR_Github #13101 [ run ] triggered by Bot |
|
/bot run --disable-fail-fast |
|
PR_Github #13108 [ run ] triggered by Bot |
|
PR_Github #13101 [ run ] completed with state |
|
PR_Github #13108 [ run ] completed with state |
|
/bot run |
|
/bot run |
|
PR_Github #13194 [ run ] triggered by Bot |
|
PR_Github #13194 [ run ] completed with state |
|
/bot run |
|
PR_Github #13291 [ run ] triggered by Bot |
|
/bot run |
|
PR_Github #13497 [ run ] triggered by Bot |
Signed-off-by: Shahar Mor <[email protected]>
|
PR_Github #13497 [ run ] completed with state |
Signed-off-by: Shahar Mor <[email protected]>
shaharmor98
force-pushed
the
bugfix/fix-nemotron-nas-lora-support
branch
from
f31f069 to
c3931f6
Compare
|
/bot run --disable-fail-fast |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 0
🧹 Nitpick comments (1)
tensorrt_llm/_torch/pyexecutor/_util.py (1)
453-462: Fix line length and clarify the uniformity check.The simplified approach to handle KV heads per layer is good, but there are minor improvements needed:
- Line 457 exceeds the 120-character limit
- The uniformity check could be more explicit
Apply this diff to address both issues:
- num_kv_attention_heads_per_layer = model_binding_config.num_kv_heads_per_layer - if max(num_kv_attention_heads_per_layer) != min( - num_kv_attention_heads_per_layer): - logger.warning( - "Defining LORA with per-layer KV heads is not supported for LORA, using the max number of KV heads per layer" - ) - num_kv_attention_heads = max(num_kv_attention_heads_per_layer) - else: - # all layers have the same number of KV heads - num_kv_attention_heads = num_kv_attention_heads_per_layer[0] + num_kv_attention_heads_per_layer = model_binding_config.num_kv_heads_per_layer + # Check if all layers have uniform KV heads + has_uniform_kv_heads = (max(num_kv_attention_heads_per_layer) == + min(num_kv_attention_heads_per_layer)) + + if not has_uniform_kv_heads: + logger.warning( + "Non-uniform KV heads per layer detected. LoRA will use the maximum " + "number of KV heads across all layers for compatibility." + ) + num_kv_attention_heads = max(num_kv_attention_heads_per_layer) + else: + # All layers have the same number of KV heads + num_kv_attention_heads = num_kv_attention_heads_per_layer[0]
📜 Review details
Configuration used: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (3)
tensorrt_llm/_torch/model_config.py(1 hunks)tensorrt_llm/_torch/pyexecutor/_util.py(1 hunks)tests/unittest/llmapi/test_llm_pytorch.py(0 hunks)
💤 Files with no reviewable changes (1)
- tests/unittest/llmapi/test_llm_pytorch.py
🚧 Files skipped from review as they are similar to previous changes (1)
- tensorrt_llm/_torch/model_config.py
🧰 Additional context used
📓 Path-based instructions (2)
**/*.py
📄 CodeRabbit Inference Engine (CODING_GUIDELINES.md)
**/*.py: The code developed for TensorRT-LLM should conform to Python 3.8+.
Indent Python code with 4 spaces. Do not use tabs.
Always maintain the namespace when importing in Python, even if only one class or function from a module is used.
Python filenames should use snake_case (e.g., some_file.py).
Python classes should use PascalCase (e.g., class SomeClass).
Python functions and methods should use snake_case (e.g., def my_awesome_function():).
Python local variables should use snake_case. Prefix k for variable names that start with a number (e.g., k_99th_percentile).
Python global variables should use upper snake_case and prefix G (e.g., G_MY_GLOBAL).
Python constants should use upper snake_case (e.g., MY_CONSTANT).
Avoid shadowing variables declared in an outer scope in Python.
Initialize all externally visible members of a Python class in the constructor.
For interfaces that may be used outside a file, prefer docstrings over comments in Python.
Comments in Python should be reserved for code within a function, or interfaces that are local to a file.
Use Google style docstrings for Python classes and functions, which can be parsed by Sphinx.
Attributes and variables in Python can be documented inline; attribute docstrings will be rendered under the docstring for the class.
Avoid using reflection in Python when functionality can be easily achieved without reflection.
When using try-except blocks in Python, limit the except to the smallest set of errors possible.
When using try-except blocks to handle multiple possible variable types in Python, keep the body of the try as small as possible, using the else block to implement the logic.
Files:
tensorrt_llm/_torch/pyexecutor/_util.py
**/*.{cpp,h,cu,py}
📄 CodeRabbit Inference Engine (CODING_GUIDELINES.md)
All TensorRT-LLM Open Source Software code should contain an NVIDIA copyright header that includes the current year. This includes .cpp, .h, .cu, .py, and any other source files which are compiled or interpreted.
Files:
tensorrt_llm/_torch/pyexecutor/_util.py
🧠 Learnings (2)
📓 Common learnings
Learnt from: amitz-nv
PR: NVIDIA/TensorRT-LLM#5616
File: tensorrt_llm/executor/worker.py:375-384
Timestamp: 2025-07-17T09:01:27.402Z
Learning: In tensorrt_llm/executor/worker.py, the LoRA adapter cache optimization logic that checks `is_adapter_in_cpu_cache()` and conditionally passes None for weights/config has a known race condition issue that cannot be solved with simple error handling or verification checks. This is a known limitation that requires a more comprehensive solution.
tensorrt_llm/_torch/pyexecutor/_util.py (1)
Learnt from: amitz-nv
PR: #5616
File: tensorrt_llm/executor/worker.py:375-384
Timestamp: 2025-07-17T09:01:27.402Z
Learning: In tensorrt_llm/executor/worker.py, the LoRA adapter cache optimization logic that checks is_adapter_in_cpu_cache() and conditionally passes None for weights/config has a known race condition issue that cannot be solved with simple error handling or verification checks. This is a known limitation that requires a more comprehensive solution.
🪛 Ruff (0.12.2)
tensorrt_llm/_torch/pyexecutor/_util.py
457-457: Line too long (125 > 120)
(E501)
⏰ 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/pyexecutor/_util.py (1)
453-462: LGTM! Solid simplification of KV heads determination.This change effectively simplifies the logic by directly using
model_binding_config.num_kv_heads_per_layerinstead of the previous approach. The handling of both uniform and non-uniform KV heads scenarios is appropriate, and using the maximum value as a fallback for LoRA compatibility is a reasonable approach.
|
PR_Github #13537 [ run ] triggered by Bot |
|
PR_Github #13537 [ run ] completed with state |
|
/bot run |
|
PR_Github #13614 [ run ] triggered by Bot |
|
PR_Github #13614 [ run ] completed with state |
|
/bot run --disable-fail-fast |
|
PR_Github #13644 [ run ] triggered by Bot |
|
/bot run --disable-fail-fast |
There was a problem hiding this comment.
Actionable comments posted: 0
🧹 Nitpick comments (1)
tensorrt_llm/_torch/pyexecutor/_util.py (1)
458-458: Fix line length violation.Line 458 exceeds the 120-character limit. Consider splitting the warning message for better readability.
- logger.warning( - "Defining LORA with per-layer KV heads is not supported for LORA, using the max number of KV heads per layer" - ) + logger.warning( + "Defining LORA with per-layer KV heads is not supported for LORA, " + "using the max number of KV heads per layer" + )
📜 Review details
Configuration used: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (3)
tensorrt_llm/_torch/model_config.py(1 hunks)tensorrt_llm/_torch/pyexecutor/_util.py(1 hunks)tests/unittest/llmapi/test_llm_pytorch.py(0 hunks)
💤 Files with no reviewable changes (1)
- tests/unittest/llmapi/test_llm_pytorch.py
🚧 Files skipped from review as they are similar to previous changes (1)
- tensorrt_llm/_torch/model_config.py
🧰 Additional context used
📓 Path-based instructions (2)
**/*.py
📄 CodeRabbit Inference Engine (CODING_GUIDELINES.md)
**/*.py: Python code should conform to Python 3.8+.
Indent Python code with 4 spaces. Do not use tabs.
Always maintain the namespace when importing in Python, even if only one class or function from a module is used.
Python filenames should use snake_case (e.g., some_file.py).
Python classes should use PascalCase (e.g., class SomeClass).
Python functions and methods should use snake_case (e.g., def my_awesome_function():).
Python local variables should use snake_case. Prefix k for variable names that start with a number (e.g., k_99th_percentile).
Python global variables should use upper snake_case and prefix G (e.g., G_MY_GLOBAL).
Python constants should use upper snake_case (e.g., MY_CONSTANT).
Avoid shadowing variables declared in an outer scope in Python.
Initialize all externally visible members of a Python class in the constructor.
For interfaces that may be used outside a file, prefer docstrings over comments in Python.
Comments in Python should be reserved for code within a function, or interfaces that are local to a file.
Use Google style docstrings for Python classes and functions, which can be parsed by Sphinx.
Attributes and variables in Python can be documented inline, and attribute docstrings will be rendered under the docstring for the class.
Avoid using reflection in Python when functionality can be easily achieved without it.
When using try-except blocks in Python, limit the except to the smallest set of errors possible.
When using try-except blocks to handle multiple possible variable types in Python, keep the body of the try as small as possible, using the else block to implement the logic.
Files:
tensorrt_llm/_torch/pyexecutor/_util.py
**/*.{cpp,h,hpp,cc,cxx,cu,py}
📄 CodeRabbit Inference Engine (CODING_GUIDELINES.md)
All TensorRT-LLM Open Source Software code should contain an NVIDIA copyright header that includes the current year. The block of text should be prepended to the top of all files, including .cpp, .h, .cu, .py, and any other source files which are compiled or interpreted.
Files:
tensorrt_llm/_torch/pyexecutor/_util.py
🧠 Learnings (2)
📓 Common learnings
Learnt from: amitz-nv
PR: NVIDIA/TensorRT-LLM#5616
File: tensorrt_llm/executor/worker.py:375-384
Timestamp: 2025-07-17T09:01:27.402Z
Learning: In tensorrt_llm/executor/worker.py, the LoRA adapter cache optimization logic that checks `is_adapter_in_cpu_cache()` and conditionally passes None for weights/config has a known race condition issue that cannot be solved with simple error handling or verification checks. This is a known limitation that requires a more comprehensive solution.
tensorrt_llm/_torch/pyexecutor/_util.py (1)
Learnt from: amitz-nv
PR: #5616
File: tensorrt_llm/executor/worker.py:375-384
Timestamp: 2025-07-17T09:01:27.402Z
Learning: In tensorrt_llm/executor/worker.py, the LoRA adapter cache optimization logic that checks is_adapter_in_cpu_cache() and conditionally passes None for weights/config has a known race condition issue that cannot be solved with simple error handling or verification checks. This is a known limitation that requires a more comprehensive solution.
🪛 Ruff (0.12.2)
tensorrt_llm/_torch/pyexecutor/_util.py
458-458: Line too long (125 > 120)
(E501)
⏰ 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/pyexecutor/_util.py (1)
454-463: LGTM! Improved LoRA KV heads determination logic.This change simplifies the approach by directly using the per-layer KV heads from the model binding configuration, which is more efficient than the previous method of querying individual attention layers. The uniformity check using max/min comparison is correct, and the fallback logic appropriately handles non-uniform configurations.
|
PR_Github #13644 [ run ] completed with state |
|
PR_Github #13667 [ run ] triggered by Bot |
|
PR_Github #13667 [ run ] completed with state |
Signed-off-by: Shahar Mor <[email protected]> Signed-off-by: Lanyu Liao <[email protected]>
Signed-off-by: Shahar Mor <[email protected]>
Summary by CodeRabbit
Bug Fixes
Tests
test_nemotron_nas_loratest, allowing it to run on systems with sufficient GPU memory.Description
Test Coverage
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.