[None][fix] fix scaffolding dynasor test #7070
Conversation
Signed-off-by: Zhenhuan Chen <chenzhh3671@gmail.com>
📝 WalkthroughWalkthroughReplaces the legacy test harness with a single main() entry point. Switches from proposer_worker to generation_worker (TRTLLMWorker). Integrates DynasorGenerationController with optional MajorityVoteController. Adds both streaming (async) and non-streaming output handling and updates shutdown to use generation_worker. Changes
Sequence Diagram(s)sequenceDiagram
autonumber
participant U as User (CLI)
participant M as main()
participant GW as generation_worker (TRTLLMWorker)
participant L as ScaffoldingLlm
participant C as Controller<br/>(DynasorGenerationController / MajorityVoteController)
Note over U,M: Start application
U->>M: Run script
M->>GW: Initialize
M->>C: Instantiate controller (streaming or non-streaming)
M->>L: Create LLM with generation_worker
alt Streaming
M->>L: generate_async(prompts)
L->>C: Control flow (streaming)
C->>GW: Generate tokens (stream)
loop token stream
GW-->>C: Partial output
C-->>L: Current output
L-->>M: cur_output updates
M->>U: Print progress
end
L-->>M: Final result
M->>U: Print final output
else Non-streaming
M->>L: generate(prompts)
L->>C: Control flow (batch)
C->>GW: Generate output
GW-->>C: Result
C-->>L: Result
L-->>M: results.outputs[0].text
M->>U: Print output
end
M->>GW: shutdown()
M-->>U: Exit
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~25 minutes Tip 🔌 Remote MCP (Model Context Protocol) integration is now available!Pro plan users can now connect to remote MCP servers from the Integrations page. Connect with popular remote MCPs such as Notion and Linear to add more context to your reviews and chats. ✨ 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
|
|
/bot skip --comment "only modify scaffolding example file" |
dc3671
changed the title
![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
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (2)
examples/scaffolding/contrib/Dynasor/scaffolding_dynasor_run.py (2)
65-90: Make shutdown reliable and fix potential NameError in streaming path.Two issues:
- result is referenced after the async-for loop; if the generator yields nothing, this raises NameError.
- Wrap generation in try/finally so we always shut down LLM/worker on error, avoiding GPU/resource leaks.
- if args.streaming: - - async def task(prompt: str): - i = 0 - async for result in llm.generate_async(prompt): - i += 1 - print(">>>", i, result) - async for output in result.cur_output: - print(">>>", i, len(output.outputs[0].token_ids), "\n", - output.outputs[0].text) - print(f">>> final output {len(result.outputs[0].token_ids)}\n", - result.outputs[0].text) - - # Need to provide LLM's event loop to get results in the middle of the whole process. - asyncio.run_coroutine_threadsafe(task(prompts[0]), llm.loop).result() - else: - results = llm.generate(prompts) - for result in results: - print(result.outputs[0].text) - - print(f"main shutting down...") - llm.shutdown() - print(f"worker shutting down...") - generation_worker.shutdown() - print(f"main shut down done") + try: + if args.streaming: + async def task(prompt: str): + i = 0 + last_result = None + async for result in llm.generate_async(prompt): + i += 1 + print(">>>", i, result) + async for output in result.cur_output: + print( + ">>>", + i, + len(output.outputs[0].token_ids), + "\n", + output.outputs[0].text, + ) + last_result = result + if last_result is not None: + print( + f">>> final output {len(last_result.outputs[0].token_ids)}\n", + last_result.outputs[0].text, + ) + else: + print(">>> final output: <no output produced>") + + # Need to provide LLM's event loop to get results in the middle of the whole process. + asyncio.run_coroutine_threadsafe(task(prompts[0]), llm.loop).result() + else: + results = llm.generate(prompts) + for result in results: + print(result.outputs[0].text) + finally: + print("main shutting down...") + llm.shutdown() + print("worker shutting down...") + generation_worker.shutdown() + print("main shut down done")
1-1: Add NVIDIA SPDX copyright header to Python sourcePer repository convention, prepend the following two-line SPDX banner to the top of examples/scaffolding/contrib/Dynasor/scaffolding_dynasor_run.py (before any imports):
+ # SPDX-FileCopyrightText: Copyright (c) 2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved. + # SPDX-License-Identifier: Apache-2.0 import argparse import asyncio
- File: examples/scaffolding/contrib/Dynasor/scaffolding_dynasor_run.py
- Location: Insert immediately above the first
importstatement.
🧹 Nitpick comments (3)
examples/scaffolding/contrib/Dynasor/scaffolding_dynasor_run.py (3)
9-22: Add type hints and short docstrings to public entry points.Improves readability and aligns with the docstring/type-hint guidelines.
-def parse_arguments(): +def parse_arguments() -> argparse.Namespace: + """Parse CLI arguments for the Dynasor scaffolding example. + + Returns: + argparse.Namespace: Parsed command-line arguments. + """ parser = argparse.ArgumentParser() @@ - return args + return args @@ -def main(): - args = parse_arguments() +def main() -> None: + """Entry point for the Dynasor scaffolding example.""" + args = parse_arguments()Also applies to: 25-26
42-63: De-duplicate ScaffoldingLlm construction.Reduce duplication by selecting the prototype_controller once, then constructing ScaffoldingLlm in a single place.
# If majority voting is requested, wrap the controller in MajorityVoteController if args.majority_vote: majority_vote_controller = MajorityVoteController( generation_controller=dynasor_generation_controller, default_sample_num=args.sample_num, ) - llm = ScaffoldingLlm( - prototype_controller=majority_vote_controller, - workers={ - DynasorGenerationController.WorkerTag.GENERATION: - generation_worker - }, - ) - else: - # Otherwise Use Dynasor controller directly - llm = ScaffoldingLlm( - prototype_controller=dynasor_generation_controller, - workers={ - DynasorGenerationController.WorkerTag.GENERATION: - generation_worker - }, - ) + # Choose controller and initialize the LLM once + prototype_controller = ( + majority_vote_controller + if args.majority_vote + else dynasor_generation_controller + ) + llm = ScaffoldingLlm( + prototype_controller=prototype_controller, + workers={ + DynasorGenerationController.WorkerTag.GENERATION: generation_worker + }, + )
65-79: Optional: stream all prompts (not just the first) or make it explicit.Currently, the streaming path runs only prompts[0], while the non-streaming path handles all prompts. Either iterate through all prompts in streaming mode or add a comment flag making the intentional single-prompt behavior explicit.
Would you like a small helper to stream prompts sequentially (or concurrently) with asyncio.gather?
📜 Review details
Configuration used: .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 (1)
examples/scaffolding/contrib/Dynasor/scaffolding_dynasor_run.py(4 hunks)
🧰 Additional context used
📓 Path-based instructions (2)
**/*.py
📄 CodeRabbit Inference Engine (CODING_GUIDELINES.md)
**/*.py: Python code must target Python 3.8+
Python indentation: 4 spaces, no tabs
Maintain module namespace in imports (from package.subpackage import foo; then use foo.SomeClass())
Python file names use snake_case
Python class names use PascalCase
Python functions/methods and local variables use snake_case; variables starting with a number get k_ prefix (e.g., k_99th_percentile)
Global variables use G_ prefixed UPPER_SNAKE_CASE (e.g., G_MY_GLOBAL)
Constants use UPPER_SNAKE_CASE in Python
Avoid shadowing variables from outer scopes in Python
Initialize all externally visible members of a Python class in init
Prefer docstrings for interfaces used outside a file; comments for local code
Use Google-style docstrings for classes and functions (Sphinx-parsable)
Document attributes/variables inline with short docstrings
Avoid reflection when simple alternatives exist (e.g., prefer explicit parameters over dict(**locals()))
In try/except, catch the narrowest exceptions possible
For duck-typing with try/except, keep try body minimal and put logic in else
Files:
examples/scaffolding/contrib/Dynasor/scaffolding_dynasor_run.py
**/*.{cpp,cxx,cc,cu,h,hpp,hxx,hh,cuh,py}
📄 CodeRabbit Inference Engine (CODING_GUIDELINES.md)
Prepend NVIDIA copyright header (current year) to all source files
Files:
examples/scaffolding/contrib/Dynasor/scaffolding_dynasor_run.py
🪛 Ruff (0.12.2)
examples/scaffolding/contrib/Dynasor/scaffolding_dynasor_run.py
29-29: Line too long (174 > 120)
(E501)
30-30: Line too long (151 > 120)
(E501)
31-31: Line too long (128 > 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
|
PR_Github #15865 [ skip ] triggered by Bot |
|
PR_Github #15865 [ skip ] completed with state |
3 participants
Summary by CodeRabbit
New Features
Improvements
Refactor
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.