[https://nvbugs/5474409][fix] Disable concurrent loading by default #7663
Conversation
📝 WalkthroughWalkthroughChanged the default value passed to os.environ.get for TRT_LLM_DISABLE_LOAD_WEIGHTS_IN_PARALLEL in tensorrt_llm/_torch/models/modeling_utils.py from False to True; the membership check against ["True","true","1","yes","y"] remains unchanged, so behavior is unchanged when the env var is unset. Changes
Estimated code review effort🎯 2 (Simple) | ⏱️ ~10 minutes Pre-merge checks (1 passed, 2 warnings)❌ Failed checks (2 warnings)
✅ Passed checks (1 passed)
📜 Recent review detailsConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Pro 📒 Files selected for processing (1)
✅ Files skipped from review due to trivial changes (1)
⏰ 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)
✨ 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 |
|
/bot run |
![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
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (3)
tensorrt_llm/_torch/models/modeling_utils.py (3)
833-839: Inverted env flag + wrong default: this enables concurrency by default (contrary to PR title).With the env var unset,
os.environ.get(..., False)yieldsFalse(bool), which is not in the string list["False","false","0","no","n"], so the code takes the concurrent branch. Also, comparing a possibly-bool to strings is brittle. Make serial (non-concurrent) the default and normalize parsing.Apply this diff here to use a shared helper:
- if os.environ.get("TRT_LLM_DISABLE_LOAD_WEIGHTS_IN_PARALLEL", - False) in ["False", "false", "0", "no", "n"]: + if _is_parallel_loading_disabled_by_default(): for name, module in tqdm(list(model.named_modules()), desc="Loading weights"): load_single_module(name, module) else:Add this helper (place near other module-level helpers, e.g., above
_load_weights_impl):def _is_parallel_loading_disabled_by_default() -> bool: """ Return True to load weights serially. Default is True (concurrency disabled) unless the env explicitly opts in. Controlled by TRT_LLM_DISABLE_LOAD_WEIGHTS_IN_PARALLEL. """ val = os.getenv("TRT_LLM_DISABLE_LOAD_WEIGHTS_IN_PARALLEL") if val is None: return True return val.strip().lower() in {"1", "true", "yes", "y", "on"}
901-907: Inconsistent gating vs. above block.This branch still uses the old “truthy → serial” check, while the earlier block (Lines 833–839) was changed. Unify both on the same helper so behavior is consistent across v1/v2 loaders.
- if os.environ.get("TRT_LLM_DISABLE_LOAD_WEIGHTS_IN_PARALLEL", - False) in ["True", "true", "1", "yes", "y"]: + if _is_parallel_loading_disabled_by_default(): for name, module in tqdm(list(model.named_modules()), desc="Loading weights"): load_single_module(name, module) else:
833-839: Use the_is_parallel_loading_disabled_by_defaulthelper for flag checks
- In
tensorrt_llm/_torch/models/modeling_utils.py(lines 833–839 and 899–903), replace the directchecks with calls toos.environ.get("TRT_LLM_DISABLE_LOAD_WEIGHTS_IN_PARALLEL", False) in […]_is_parallel_loading_disabled_by_default()to ensure a single, consistent implementation of the flag logic.
🧹 Nitpick comments (3)
tensorrt_llm/_torch/models/modeling_utils.py (3)
749-754: Avoid mutable default args (skip_modules=[]).Using a shared list across calls risks cross-call contamination (e.g., v2 calls
weight_mapper.add_skip_modules). PreferNonedefault and initialize inside.-def _load_weights_impl(model: Union[nn.Module, DecoderModelForCausalLM], - weights: Dict, - skip_modules: List[str] = [], +def _load_weights_impl(model: Union[nn.Module, DecoderModelForCausalLM], + weights: Dict, + skip_modules: Optional[List[str]] = None, params_map: Optional[Dict[str, str]] = None, preload_weight_modules: Optional[List[str]] = None):if params_map is not None: weights = rename_weights_with_regex(params_map, weights) logger.info(f"Renamed weights with params_map: {params_map}") + if skip_modules is None: + skip_modules = []-def _load_weights_impl_v2(model: Union[nn.Module, DecoderModelForCausalLM], - weights: Dict, - weight_mapper: "BaseWeightMapper", - skip_modules: List[str] = [], +def _load_weights_impl_v2(model: Union[nn.Module, DecoderModelForCausalLM], + weights: Dict, + weight_mapper: "BaseWeightMapper", + skip_modules: Optional[List[str]] = None, params_map: Optional[Dict[str, str]] = None, preload_weight_modules: Optional[List[str]] = None):weight_mapper.add_skip_modules(skip_modules) + if skip_modules is None: + skip_modules = []Also applies to: 765-769, 862-868, 871-877
1-1: Missing NVIDIA Apache-2.0 header (per repo guidelines).Add the 2025 copyright header at the top.
+# Copyright (c) 2025, NVIDIA CORPORATION. All rights reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License.
833-860: (Optional) Log chosen mode for easier debugging.Emitting a single INFO line (“serial” vs “concurrent”) helps triage load-time issues.
- if _is_parallel_loading_disabled_by_default(): + if _is_parallel_loading_disabled_by_default(): + logger.info("TRT-LLM: loading weights serially (concurrency disabled).") for name, module in tqdm(list(model.named_modules()), desc="Loading weights"): load_single_module(name, module) else: + logger.info("TRT-LLM: loading weights concurrently (concurrency enabled).")- if _is_parallel_loading_disabled_by_default(): + if _is_parallel_loading_disabled_by_default(): + logger.info("TRT-LLM: loading weights serially (concurrency disabled).") for name, module in tqdm(list(model.named_modules()), desc="Loading weights"): load_single_module(name, module) else: + logger.info("TRT-LLM: loading weights concurrently (concurrency enabled).")Also applies to: 901-927
📜 Review details
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (1)
tensorrt_llm/_torch/models/modeling_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/models/modeling_utils.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_utils.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_utils.py
🧠 Learnings (1)
📚 Learning: 2025-08-19T12:45:11.997Z
Learnt from: amitz-nv
PR: NVIDIA/TensorRT-LLM#7033
File: tensorrt_llm/_torch/pyexecutor/model_engine.py:0-0
Timestamp: 2025-08-19T12:45:11.997Z
Learning: In tensorrt_llm/_torch/pyexecutor/model_engine.py, DoRA (Delta Orthogonal Rank Adaptation) functionality was removed from the PyTorch flow to eliminate issues with inverted DoRA detection logic. The original is_dora condition was checking if scaling_vec_pointer == 0, which was potentially incorrect.
Applied to files:
tensorrt_llm/_torch/models/modeling_utils.py
⏰ 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
|
PR_Github #18287 [ run ] triggered by Bot |
Signed-off-by: nv-guomingz <[email protected]>
nv-guomingz
force-pushed
the
user/guomingz/revert_concurrent_load
branch
from
476fa4f to
ea01cb9
Compare
|
/bot run |
|
PR_Github #18354 [ run ] triggered by Bot |
|
PR_Github #18354 [ run ] completed with state |
|
/bot run |
|
PR_Github #18363 [ run ] triggered by Bot |
|
PR_Github #18363 [ run ] completed with state |
…VIDIA#7663) Signed-off-by: nv-guomingz <[email protected]>
…VIDIA#7663) Signed-off-by: nv-guomingz <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
…VIDIA#7663) Signed-off-by: nv-guomingz <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
…VIDIA#7663) Signed-off-by: nv-guomingz <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
…VIDIA#7663) Signed-off-by: nv-guomingz <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
…VIDIA#7663) Signed-off-by: nv-guomingz <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
…VIDIA#7663) Signed-off-by: nv-guomingz <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
…VIDIA#7663) Signed-off-by: nv-guomingz <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
…VIDIA#7663) Signed-off-by: nv-guomingz <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
…VIDIA#7663) Signed-off-by: nv-guomingz <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
…VIDIA#7663) Signed-off-by: nv-guomingz <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
…VIDIA#7663) Signed-off-by: nv-guomingz <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
…VIDIA#7663) Signed-off-by: nv-guomingz <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
…VIDIA#7663) Signed-off-by: nv-guomingz <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
…VIDIA#7663) Signed-off-by: nv-guomingz <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
…VIDIA#7663) Signed-off-by: nv-guomingz <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
…VIDIA#7663) Signed-off-by: nv-guomingz <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
…VIDIA#7663) Signed-off-by: nv-guomingz <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
…VIDIA#7663) Signed-off-by: nv-guomingz <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
…VIDIA#7663) Signed-off-by: nv-guomingz <[email protected]>
…VIDIA#7663) Signed-off-by: nv-guomingz <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
…VIDIA#7663) Signed-off-by: nv-guomingz <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
…VIDIA#7663) Signed-off-by: nv-guomingz <[email protected]> Signed-off-by: Wangshanshan <[email protected]>
4 participants
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.
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.