[https://nvbugs/5467531][fix] Fix moe test and wide ep fake impl #8883
Conversation
1. Adjust HIDDEN_SIZE to 4096. The DeepEP version we use does not support 2560 hidden size. 2. Adjust test impl. The per rank function now accepts CPU inputs and tensor to avoid a weird integer overflow of MPI execution(Maybe caused by too many GPU memory are captured). 3. Fix Wide EP fake impl 4. Apply a reduce in the test if topk dimension is not reduced for MNNVL Signed-off-by: Jin Li <[email protected]>
liji-nv
force-pushed
the
dev-liji-fused-moe-test
branch
from
4d0ea39 to
403b035
Compare
📝 WalkthroughWalkthroughAdds Changes
Sequence Diagram(s)sequenceDiagram
autonumber
participant Caller
participant WideEPMoE
participant Parent (MoE Base)
Caller->>WideEPMoE: forward_fake(x, router_logits, ...)
WideEPMoE->>Parent: forward_fake(..., output_dtype=bfloat16, ...)
Parent-->>WideEPMoE: moe_output
alt alltoall_method_type == MNNVL
Note right of WideEPMoE #f0f8ff: Placeholder branch for MNNVL
WideEPMoE-->>Caller: empty_tensor(shape=[batch, experts_per_token, features])
else other alltoall methods
Note right of WideEPMoE #f9fbe7: Preserve parent output
WideEPMoE-->>Caller: moe_output
end
Estimated code review effort🎯 4 (Complex) | ⏱️ ~45 minutes
Pre-merge checks and finishing touches❌ Failed checks (1 warning)
✅ Passed checks (1 passed)
✨ Finishing touches
🧪 Generate unit tests (beta)
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 |
|
PR_Github #23388 [ run ] triggered by Bot. Commit: |
![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)
tests/unittest/_torch/modules/test_fused_moe.py (1)
288-289: Drop stray debug print.Please remove the leftover
print("XXX"); it’s noisy in regular test runs and doesn’t contribute to the assertions.
📜 Review details
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (2)
tensorrt_llm/_torch/modules/fused_moe/fused_moe_wide_ep.py(1 hunks)tests/unittest/_torch/modules/test_fused_moe.py(7 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/modules/fused_moe/fused_moe_wide_ep.pytests/unittest/_torch/modules/test_fused_moe.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/modules/fused_moe/fused_moe_wide_ep.pytests/unittest/_torch/modules/test_fused_moe.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/modules/fused_moe/fused_moe_wide_ep.pytests/unittest/_torch/modules/test_fused_moe.py
🧠 Learnings (4)
📓 Common learnings
Learnt from: djns99
Repo: NVIDIA/TensorRT-LLM PR: 6915
File: cpp/tensorrt_llm/kernels/cutlass_kernels/moe_gemm/moe_kernels.cu:4010-4012
Timestamp: 2025-08-14T23:23:27.449Z
Learning: For MOE (Mixture of Experts) code reviews in TensorRT-LLM, avoid repeatedly suggesting finalize fusion validation checks and safety assertions. The user djns99 has indicated these suggestions are repetitive and unwanted across multiple MOE-related changes.
Learnt from: sklevtsov-nvidia
Repo: NVIDIA/TensorRT-LLM PR: 3294
File: cpp/tensorrt_llm/kernels/cutlass_kernels/moe_gemm/moe_gemm_tma_warp_specialized_input.cu:118-127
Timestamp: 2025-08-09T20:57:04.084Z
Learning: In the CUTLASS MoE finalize fusion implementation (cpp/tensorrt_llm/kernels/cutlass_kernels/moe_gemm/moe_gemm_tma_warp_specialized_input.cu), when setting `fused_finalize_epilogue.stride_final_output` with shape `(hidden_size, num_output_tokens, 1)`, the `num_rows_in_final_output` should be set to `num_output_tokens` (not `hidden_size`) because of a swap+transpose operation that maps rows of the output tensor to `hidden_size` and columns to `num_output_tokens`.
Learnt from: jhaotingc
Repo: NVIDIA/TensorRT-LLM PR: 7856
File: cpp/tensorrt_llm/thop/fp8BlockScaleMoe.cpp:159-166
Timestamp: 2025-09-19T21:28:13.751Z
Learning: In TensorRT-LLM blockScaleMoe routing (cpp/tensorrt_llm/kernels/trtllmGenKernels/blockScaleMoe/runner.cu), the DeepSeek routing method performs reinterpret_cast<float*>(routingLogits) at line 89, which could cause issues if routing_logits are BF16. However, Qwen3-FP8 models use RenormalizeNaive routing method and are not affected by this dtype casting issue.
📚 Learning: 2025-08-14T23:23:27.449Z
Learnt from: djns99
Repo: NVIDIA/TensorRT-LLM PR: 6915
File: cpp/tensorrt_llm/kernels/cutlass_kernels/moe_gemm/moe_kernels.cu:4010-4012
Timestamp: 2025-08-14T23:23:27.449Z
Learning: For MOE (Mixture of Experts) code reviews in TensorRT-LLM, avoid repeatedly suggesting finalize fusion validation checks and safety assertions. The user djns99 has indicated these suggestions are repetitive and unwanted across multiple MOE-related changes.
Applied to files:
tensorrt_llm/_torch/modules/fused_moe/fused_moe_wide_ep.pytests/unittest/_torch/modules/test_fused_moe.py
📚 Learning: 2025-08-09T20:57:04.084Z
Learnt from: sklevtsov-nvidia
Repo: NVIDIA/TensorRT-LLM PR: 3294
File: cpp/tensorrt_llm/kernels/cutlass_kernels/moe_gemm/moe_gemm_tma_warp_specialized_input.cu:118-127
Timestamp: 2025-08-09T20:57:04.084Z
Learning: In the CUTLASS MoE finalize fusion implementation (cpp/tensorrt_llm/kernels/cutlass_kernels/moe_gemm/moe_gemm_tma_warp_specialized_input.cu), when setting `fused_finalize_epilogue.stride_final_output` with shape `(hidden_size, num_output_tokens, 1)`, the `num_rows_in_final_output` should be set to `num_output_tokens` (not `hidden_size`) because of a swap+transpose operation that maps rows of the output tensor to `hidden_size` and columns to `num_output_tokens`.
Applied to files:
tensorrt_llm/_torch/modules/fused_moe/fused_moe_wide_ep.py
📚 Learning: 2025-10-13T19:45:03.518Z
Learnt from: nv-lschneider
Repo: NVIDIA/TensorRT-LLM PR: 7910
File: tests/unittest/_torch/multi_gpu/test_nccl_device.py:138-149
Timestamp: 2025-10-13T19:45:03.518Z
Learning: In test_nccl_device.py, the NCCL device AllReduce implementation compares the entire residual tensor on each rank, unlike the UB implementation which compares per-rank chunks. The residual chunking calculations in the test are intentionally overridden to reflect this design difference.
Applied to files:
tests/unittest/_torch/modules/test_fused_moe.py
🧬 Code graph analysis (2)
tensorrt_llm/_torch/modules/fused_moe/fused_moe_wide_ep.py (5)
tensorrt_llm/_torch/modules/fused_moe/fused_moe_cutlass.py (1)
forward_fake(564-593)tensorrt_llm/_torch/modules/fused_moe/fused_moe_trtllm_gen.py (1)
forward_fake(475-505)tensorrt_llm/_torch/modules/fused_moe/interface.py (1)
forward_fake(215-232)tensorrt_llm/_torch/utils.py (2)
Fp4QuantizedTensor(100-107)shape(106-107)tensorrt_llm/_torch/models/modeling_qwen3_moe.py (1)
routing_method(67-80)
tests/unittest/_torch/modules/test_fused_moe.py (5)
tensorrt_llm/mapping.py (1)
Mapping(32-519)tensorrt_llm/_torch/modules/fused_moe/fused_moe_wide_ep.py (1)
forward_fake(1045-1070)tensorrt_llm/_torch/modules/fused_moe/fused_moe_cutlass.py (1)
forward_fake(564-593)tensorrt_llm/_torch/modules/fused_moe/fused_moe_trtllm_gen.py (1)
forward_fake(475-505)tensorrt_llm/_torch/modules/fused_moe/interface.py (1)
forward_fake(215-232)
🪛 Ruff (0.14.2)
tensorrt_llm/_torch/modules/fused_moe/fused_moe_wide_ep.py
1051-1051: Unused method argument: output_dtype
(ARG002)
tests/unittest/_torch/modules/test_fused_moe.py
474-475: zip() without an explicit strict= parameter
Add explicit value for parameter strict=
(B905)
⏰ 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 #23388 [ run ] completed with state |
|
/bot run |
|
PR_Github #23448 [ run ] triggered by Bot. Commit: |
|
PR_Github #23448 [ run ] completed with state |
|
|
||
| w1_weight_nvfp4, w1_sf_block = torch.ops.trtllm.fp4_quantize( | ||
| w1_weight, w3_w1_global, SCALING_VECTOR_SIZE, False) | ||
| w1_weight.cuda(), w3_w1_global.cuda(), SCALING_VECTOR_SIZE, |
There was a problem hiding this comment.
What's the meaning of removing all the device="cuda" above and add them back here?
There was a problem hiding this comment.
Is it sth relevant to Adjust test impl. The per rank function now accepts CPU inputs and tensor to avoid a weird integer overflow of MPI execution(Maybe caused by too many GPU memory are captured).?
|
|
||
| x_list = x_list_world[mapping.rank] | ||
| weights = weights_world[mapping.rank] | ||
| weights = {k: v.cuda() for k, v in weights.items()} |
There was a problem hiding this comment.
Same here. Why you replace all the .cuda(i) with .cpu() and add them back here? It may consume more time for weight preprocessing.
4 participants
Summary by CodeRabbit
Refactor
Tests
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.