[TRTLLM-7279][test] add accuracy test for deepseek-r1 with chunked_prefill #7365
Conversation
📝 WalkthroughWalkthroughAdds new and updated PyTorch accuracy tests for NVFP4 and FP8 block-scale with chunked prefill across multi-GPU setups, parameterized over MOE backends and parallelism configs. Expands public test signatures to include moe_backend, updates skip gating, asserts selected MOE backend and quantization, and registers new tests in QA test lists. Changes
Sequence Diagram(s)sequenceDiagram
autonumber
participant T as PyTest
participant C as Test Case (DeepSeek R1/V3 Lite)
participant L as LLM Init
participant Q as Quant Selector
participant M as MOE Backend
participant E as GSM8K Eval
Note over C: Parameters: {tp, pp, ep, ... , enable_chunked_prefill, moe_backend}
T->>C: Run parameterized test
C->>L: Instantiate LLM (enable_chunked_prefill, moe_backend, configs)
L->>Q: Determine quantization algorithm
Q-->>L: NVFP4 or FP8_BLOCK_SCALES
L->>M: Select/Configure MOE backend
C->>C: Assert MOE backend and QuantAlgo
C->>E: Execute GSM8K task
E-->>C: Results
C-->>T: Test pass/fail
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~25 minutes Possibly related PRs
Suggested labels
Suggested reviewers
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
|
![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
🧹 Nitpick comments (2)
tests/integration/defs/accuracy/test_llm_api_pytorch.py (2)
1864-1908: Bound FP8 chunked prefill window for stability.Set max_num_tokens to constrain prefill chunking and reduce variance/risks in CI.
- with LLM(f"{llm_models_root()}/DeepSeek-R1/DeepSeek-R1", - max_batch_size=max_batch_size, + with LLM(f"{llm_models_root()}/DeepSeek-R1/DeepSeek-R1", + max_batch_size=max_batch_size, + max_num_tokens=256, tensor_parallel_size=tp_size, pipeline_parallel_size=pp_size, moe_expert_parallel_size=ep_size, kv_cache_config=kv_cache_config, **pytorch_config, enable_attention_dp=attention_dp, speculative_config=mtp_config, enable_chunked_prefill=True) as llm:
1757-1818: Add amax_num_tokensargument for deterministic, bounded chunked prefill. No need to dropmtp_nextn>0—NVFP4+MTP chunked prefill is supported on Blackwell (guarded by@skip_pre_blackwell), so keep the existingmtp_nextn=3.
📜 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)
tests/integration/defs/accuracy/test_llm_api_pytorch.py(3 hunks)tests/integration/test_lists/qa/llm_function_full.txt(2 hunks)tests/integration/test_lists/qa/llm_function_sanity.txt(1 hunks)
🧰 Additional context used
📓 Path-based instructions (3)
**/*.{cpp,cc,cxx,cu,py,h,hpp,hh,hxx,cuh}
📄 CodeRabbit inference engine (CODING_GUIDELINES.md)
Use spaces only; indent 4 spaces
Files:
tests/integration/defs/accuracy/test_llm_api_pytorch.py
**/*.py
📄 CodeRabbit inference engine (CODING_GUIDELINES.md)
**/*.py: Python code must target Python 3.8+
Indent with 4 spaces; do not use tabs
Preserve module namespaces in imports; prefer from package.subpackage import module then module.Symbol
Python file names use snake_case (e.g., some_file.py)
Class names use PascalCase
Function and method names use snake_case
Local variable names use snake_case; if starting with a number, prefix k (e.g., k_99th_percentile)
Global variables use G_ prefix and UPPER_SNAKE_CASE (e.g., G_MY_GLOBAL)
Constants use UPPER_SNAKE_CASE
Avoid shadowing variables from an outer scope
Initialize all externally visible members of a class in init
For interfaces used outside a file, prefer docstrings; reserve comments for internal code or local interfaces
Use Google-style docstrings for classes and functions; document attributes/variables inline as shown
Avoid reflection when simple, explicit code suffices (e.g., prefer def make_complex(x,y) over locals()/dict tricks)
Catch the narrowest exceptions possible in try/except
For duck-typing try/except, keep try body minimal and use else for main logic
Files:
tests/integration/defs/accuracy/test_llm_api_pytorch.py
**/*.{cpp,cc,cxx,cu,h,hpp,hh,hxx,cuh,py}
📄 CodeRabbit inference engine (CODING_GUIDELINES.md)
Prepend NVIDIA copyright header (current year) to all source files (.cpp, .h, .cu, .py, etc.)
Files:
tests/integration/defs/accuracy/test_llm_api_pytorch.py
🧠 Learnings (1)
📚 Learning: 2025-07-28T17:06:08.621Z
Learnt from: moraxu
PR: NVIDIA/TensorRT-LLM#6303
File: tests/integration/test_lists/qa/examples_test_list.txt:494-494
Timestamp: 2025-07-28T17:06:08.621Z
Learning: In TensorRT-LLM testing, it's common to have both CLI flow tests (test_cli_flow.py) and PyTorch API tests (test_llm_api_pytorch.py) for the same model. These serve different purposes: CLI flow tests validate the traditional command-line workflow, while PyTorch API tests validate the newer LLM API backend. Both are legitimate and should coexist.
Applied to files:
tests/integration/test_lists/qa/llm_function_sanity.txttests/integration/test_lists/qa/llm_function_full.txt
🧬 Code graph analysis (1)
tests/integration/defs/accuracy/test_llm_api_pytorch.py (6)
tensorrt_llm/llmapi/llm_args.py (4)
KvCacheConfig(946-1077)CudaGraphConfig(108-165)MoeConfig(168-196)MTPDecodingConfig(521-556)tensorrt_llm/llmapi/llm.py (1)
LLM(1011-1027)tests/integration/defs/conftest.py (2)
llm_models_root(77-83)get_sm_version(1857-1860)tensorrt_llm/quantization/mode.py (1)
QuantAlgo(23-46)tensorrt_llm/evaluate/lm_eval.py (2)
GSM8K(440-488)evaluate(385-417)tests/integration/defs/accuracy/accuracy_core.py (3)
GSM8K(293-308)evaluate(147-206)evaluate(707-717)
⏰ 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 (2)
tests/integration/test_lists/qa/llm_function_sanity.txt (1)
38-42: Approve code changes: Verified thattest_nvfp4_multi_gpus_chunked_prefill(line 1786) andtest_fp8_blockscale_chunked_prefill(line 1871) exist in tests/integration/defs/accuracy/test_llm_api_pytorch.py; entries are consistent.tests/integration/test_lists/qa/llm_function_full.txt (1)
501-501: Verified: all new parameterized tests appear in both full and sanity QA lists
All four test IDs are present inllm_function_full.txt(lines 518–522) andllm_function_sanity.txt(lines 38–42).
|
/bot run |
|
PR_Github #17141 [ run ] triggered by Bot |
|
PR_Github #17141 [ run ] completed with state |
|
/bot run |
|
PR_Github #17347 [ run ] triggered by Bot |
|
PR_Github #17347 [ run ] completed with state |
Signed-off-by: Ivy Zhang <[email protected]>
Signed-off-by: Ivy Zhang <[email protected]>
|
/bot run |
|
PR_Github #17625 [ run ] triggered by Bot |
|
PR_Github #17625 [ run ] completed with state |
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.