[None][chore] Remove executor_config in create_py_executor_instance #7463
Conversation
Signed-off-by: leslie-fang25 <[email protected]>
leslie-fang25
force-pushed
the
leslie/remove_executor_config_in_create_py_executor_instance
branch
from
eb64e57 to
e159b4e
Compare
|
/bot run |
|
PR_Github #17309 [ run ] triggered by Bot |
|
PR_Github #17309 [ run ] completed with state |
📝 WalkthroughWalkthroughRefactors PyExecutor creation to remove the top-level executor_config parameter. Adds granular parameters (batch/beam/token limits and specific config objects) to creation utilities and PyExecutor. Updates call sites to pass these parameters and captures potential peft_cache_config mutations from the created PyExecutor back into the original config holder. Changes
Sequence Diagram(s)sequenceDiagram
autonumber
participant Creator as py_executor_creator.drafting_loop_wrapper
participant Util as _util.create_py_executor_instance
participant Exec as PyExecutor
participant Cfg as executor_config (holder)
Note over Creator: First executor creation (granular params)
Creator->>Util: create_py_executor_instance(max_batch_size, max_beam_width,<br/>max_num_tokens, peft_cache_config, scheduler_config,<br/>cache_transceiver_config, ...)
rect rgba(220,240,255,0.4)
Util->>Exec: new PyExecutor(..., peft_cache_config)
Exec-->>Util: instance (may have mutated peft_cache_config)
end
Util-->>Creator: PyExecutor
Note over Creator: Surface potential PEFT cache mutation
Creator->>Cfg: Cfg.peft_cache_config = Exec.peft_cache_config
Note over Creator: Second executor creation with same granular params
Creator->>Util: create_py_executor_instance(..., peft_cache_config, scheduler_config,<br/>cache_transceiver_config, ...)
Util->>Exec: new PyExecutor(..., peft_cache_config)
Exec-->>Util: instance
Util-->>Creator: PyExecutor
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~25 minutes Possibly related PRs
Suggested reviewers
✨ 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. CodeRabbit Commands (Invoked using PR/Issue comments)Type Other keywords and placeholders
Status, Documentation and Community
|
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 2
🧹 Nitpick comments (5)
tensorrt_llm/_torch/pyexecutor/_util.py (4)
10-16: Deduplicate ModelConfig import to avoid ambiguityBoth absolute and relative imports bring in ModelConfig; keep one.
Apply:
-from ..model_config import ModelConfigAlso applies to: 24-24
533-533: Fix confusing log message (duplicate/max_num_requests label)Currently logs max_batch_size twice and uses “max_num_requests” ambiguously.
Apply:
- f"max_seq_len={max_seq_len}, max_num_requests={max_batch_size}, max_num_tokens={max_num_tokens}, max_batch_size={max_batch_size}" + f"max_seq_len={max_seq_len}, max_batch_size={max_batch_size}, max_beam_width={max_beam_width}, max_num_tokens={max_num_tokens}"
585-599: PEFT cache conversion path looks correct; minor naming nitFlow from PeftCacheConfig (pydantic) → _to_pybind() is good. Consider renaming the reassigned peft_cache_config variable to avoid shadowing the input param for readability.
Apply:
- peft_cache_config = peft_cache_config_model._to_pybind() + peft_cache_config_pybind = peft_cache_config_model._to_pybind()And update uses below in this function:
- peft_cache_manager = PeftCacheManager( - peft_cache_config=peft_cache_config, + peft_cache_manager = PeftCacheManager( + peft_cache_config=peft_cache_config_pybind, ... - return PyExecutor( + return PyExecutor( ... - peft_cache_config=peft_cache_config) + peft_cache_config=peft_cache_config_pybind)
652-652: Remove no-op self-assignmentRedundant statement.
Apply:
- cache_transceiver_config = cache_transceiver_configtensorrt_llm/_torch/pyexecutor/py_executor.py (1)
161-168: Document new parameters/attributes in constructor docstringAdd brief notes for max_seq_len and peft_cache_config to keep public API discoverable.
📜 Review details
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
💡 Knowledge Base configuration:
- MCP integration is disabled by default for public repositories
- Jira integration is disabled by default for public repositories
- Linear integration is disabled by default for public repositories
You can enable these sources in your CodeRabbit configuration.
📒 Files selected for processing (3)
tensorrt_llm/_torch/pyexecutor/_util.py(7 hunks)tensorrt_llm/_torch/pyexecutor/py_executor.py(2 hunks)tensorrt_llm/_torch/pyexecutor/py_executor_creator.py(2 hunks)
🧰 Additional context used
📓 Path-based instructions (4)
**/*
📄 CodeRabbit inference engine (CODING_GUIDELINES.md)
Filenames compiled into a target must be case-insensitively unique
Files:
tensorrt_llm/_torch/pyexecutor/py_executor.pytensorrt_llm/_torch/pyexecutor/_util.pytensorrt_llm/_torch/pyexecutor/py_executor_creator.py
**/*.{h,hpp,hh,hxx,cc,cpp,cxx,cu,cuh,py}
📄 CodeRabbit inference engine (CODING_GUIDELINES.md)
Use spaces, not tabs; indent 4 spaces
Files:
tensorrt_llm/_torch/pyexecutor/py_executor.pytensorrt_llm/_torch/pyexecutor/_util.pytensorrt_llm/_torch/pyexecutor/py_executor_creator.py
**/*.py
📄 CodeRabbit inference engine (CODING_GUIDELINES.md)
**/*.py: Code must target Python 3.8+
Indent with 4 spaces; do not use tabs (Python)
Maintain module namespace on import: prefer from package.subpackage import foo; use foo.Symbol()
Python filenames use snake_case
Python class names use PascalCase
Python functions and methods use snake_case
Python local variables use snake_case; if starting with a number concept, prefix with k (e.g., k_99th_percentile)
Python global variables use G_ prefix with UPPER_SNAKE_CASE
Python constants use UPPER_SNAKE_CASE
Avoid shadowing variables from outer scopes
Initialize all externally visible class members in init
For public interfaces, prefer docstrings over comments; comments should be for in-function or file-local interfaces
Use Google-style docstrings for classes and functions (Sphinx-parsable)
Document attributes and variables inline with docstrings immediately after assignment
Avoid reflection when a non-reflective approach suffices
Limit except clauses to specific exceptions where possible
When using try/except for duck-typing, keep try body minimal and move logic to else
Files:
tensorrt_llm/_torch/pyexecutor/py_executor.pytensorrt_llm/_torch/pyexecutor/_util.pytensorrt_llm/_torch/pyexecutor/py_executor_creator.py
**/*.{cpp,cc,cxx,h,hpp,hh,hxx,cu,cuh,py}
📄 CodeRabbit inference engine (CODING_GUIDELINES.md)
Prepend NVIDIA copyright header (current year) to all source files
Files:
tensorrt_llm/_torch/pyexecutor/py_executor.pytensorrt_llm/_torch/pyexecutor/_util.pytensorrt_llm/_torch/pyexecutor/py_executor_creator.py
🧠 Learnings (1)
📚 Learning: 2025-08-26T06:07:02.166Z
Learnt from: shaharmor98
PR: NVIDIA/TensorRT-LLM#7231
File: tensorrt_llm/_torch/pyexecutor/_util.py:504-509
Timestamp: 2025-08-26T06:07:02.166Z
Learning: In tensorrt_llm/_torch/pyexecutor/_util.py, when calling model_engine.set_lora_model_config(), pass model_binding_config.mlp_hidden_size directly without multiplying by mapping.tp_size, as the mlp_hidden_size from get_bindings_model_config() is already the per-TP rank value needed for LoRA weight packaging.
Applied to files:
tensorrt_llm/_torch/pyexecutor/_util.pytensorrt_llm/_torch/pyexecutor/py_executor_creator.py
🧬 Code graph analysis (1)
tensorrt_llm/_torch/pyexecutor/py_executor.py (2)
cpp/include/tensorrt_llm/executor/types.h (3)
tensorrt_llm(36-39)tensorrt_llm(41-663)RequestStage(368-603)tensorrt_llm/llmapi/llm_args.py (1)
PeftCacheConfig(829-895)
🔇 Additional comments (7)
tensorrt_llm/_torch/pyexecutor/_util.py (3)
620-624: max_num_sequences computation LGTMUsing pp_size multiplier is consistent with existing semantics.
665-676: Plumbing of sizing and PEFT args into PyExecutor looks correctmax_batch_size/max_beam_width/max_seq_len and peft_cache_config are correctly forwarded.
639-646: Remove unnecessary None check forscheduler_config
scheduler_configis always instantiated (with a defaultcapacity_scheduler_policy) and neverNone, so guarding before.capacity_scheduler_policyis not needed.Likely an incorrect or invalid review comment.
tensorrt_llm/_torch/pyexecutor/py_executor.py (2)
27-31: Importing _PeftCacheConfig alias here is appropriateKeeps binding types explicit and avoids confusion with llm_args.PeftCacheConfig.
161-168: Expose peft_cache_config on PyExecutorAddition is straightforward and low-risk.
tensorrt_llm/_torch/pyexecutor/py_executor_creator.py (2)
516-526: Creation call updated to granular params; PEFT config propagation back to executor_config is sensibleThis preserves the previous implicit mutation behavior while aligning with the new API shape.
498-522: Confirm no lingering executor_config usageI did not find any calls still passing an executor_config param to create_py_executor_instance, but please manually verify across all call sites that no caller still uses the old executor_config argument.
Signed-off-by: leslie-fang25 <[email protected]>
|
/bot run |
|
PR_Github #17596 [ run ] triggered by Bot |
|
PR_Github #17596 [ run ] completed with state |
|
/bot run |
|
PR_Github #17769 [ run ] triggered by Bot |
|
PR_Github #17769 [ run ] completed with state |
…VIDIA#7463) Signed-off-by: leslie-fang25 <[email protected]>
3 participants
Summary by CodeRabbit
New Features
Refactor
Bug Fixes
Notes
Description
This PR removes the use of
executor_configincreate_py_executor_instance, replacing it with individual field inputs.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.