[#9717][chore] Standardize MoE weights interface #10295
Conversation
Signed-off-by: Tal Cherckez <127761168+tcherckez-nvidia@users.noreply.github.com>
|
/bot run |
📝 WalkthroughWalkthroughThis pull request refactors MoE (Mixture of Experts) weight handling from dual-format support (stacked/Llama4 weights) to a unified per-expert-list approach. The changes remove stacked weight format code paths, introduce state-dict split hooks for decomposing stacked weights on load, and update MLP construction, fusion logic, and sharding operations accordingly. Changes
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~25 minutes Pre-merge checks and finishing touches❌ Failed checks (2 warnings)
✅ 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 |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 3
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (1)
tensorrt_llm/_torch/auto_deploy/custom_ops/fused_moe/torch_moe.py (1)
127-149: Inconsistent dtype casting between gated and non-gated MLP paths.The gated MLP path (lines 134-136) explicitly casts input to weight dtype with
inp.to(W1.dtype)andinp.to(W3.dtype), but the non-gated MLP path (line 145) does not perform any dtype casting. This inconsistency could cause dtype mismatch errors when using the non-gated path with mixed precision weights.🔎 Proposed fix - add dtype casting to non-gated path
else: # Standard per-expert list format with simple MLP def make_mlp(i: int): W_up = w1_weight[i] # (I, H) W_down = w2_weight[i] # (H, I) - return lambda inp: F.linear(torch_act_fn(F.linear(inp, W_up)), W_down) + return lambda inp: F.linear(torch_act_fn(F.linear(inp.to(W_up.dtype), W_up)), W_down) mlps = [make_mlp(i) for i in range(len(w1_weight))]
🧹 Nitpick comments (4)
tensorrt_llm/_torch/auto_deploy/transform/library/fused_moe.py (1)
51-51: Consider addingstrict=Truetozip()for safety.Since
w1_keysandw3_keysshould always have the same length as the number of experts, addingstrict=Truewould catch any mismatch early.🔎 Proposed fix
- for w1_key, w3_key, w1, w3 in zip(w1_keys, w3_keys, w1_experts, w3_experts): + for w1_key, w3_key, w1, w3 in zip(w1_keys, w3_keys, w1_experts, w3_experts, strict=True):tests/unittest/_torch/auto_deploy/unit/singlegpu/transformations/library/test_bmm_moe_hooks.py (1)
12-27: Consider parameterizing fixtures for broader coverage.The fixtures use fixed dimensions. While the parameterized tests in the class cover different configurations, the standalone fixtures (
gate_up_stacked_weight,down_stacked_weight) could benefit from parameterization if they're meant to be used more broadly.tensorrt_llm/_torch/auto_deploy/transform/library/sharding.py (1)
1448-1448: Cosmetic: Ambiguous multiplication sign in comment.The comment uses
×(Unicode multiplication sign) instead ofx. While this doesn't affect functionality, it could cause issues with some editors or search tools.🔎 Proposed fix
- for i in range(len(scale_names) * 3): # 3 layers (w1, w2, w3) × #scale_names per layer + for i in range(len(scale_names) * 3): # 3 layers (w1, w2, w3) x #scale_names per layertests/unittest/_torch/auto_deploy/unit/singlegpu/custom_ops/test_ad_moe_op.py (1)
65-104: Removesetup_bmm_moe_testfixture as it is unused.The
setup_bmm_moe_testfunction (lines 65–104) is no longer called anywhere in the codebase after the removal oftest_bmm_based_moe_op_run. Delete this function to eliminate dead code.
📜 Review details
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (5)
tensorrt_llm/_torch/auto_deploy/custom_ops/fused_moe/torch_moe.pytensorrt_llm/_torch/auto_deploy/transform/library/fused_moe.pytensorrt_llm/_torch/auto_deploy/transform/library/sharding.pytests/unittest/_torch/auto_deploy/unit/singlegpu/custom_ops/test_ad_moe_op.pytests/unittest/_torch/auto_deploy/unit/singlegpu/transformations/library/test_bmm_moe_hooks.py
🧰 Additional context used
📓 Path-based instructions (2)
**/*.py
📄 CodeRabbit inference engine (CODING_GUIDELINES.md)
**/*.py: Code developed for TensorRT-LLM should conform to Python 3.8+
Indent Python code with 4 spaces. Do not use tabs
Always maintain the namespace when importing in Python, even if only one class or function from a module is used
Python files should use snake_case naming:some_file.py
Python classes should use PascalCase naming:class SomeClass
Python functions and methods should use snake_case naming:def my_awesome_function():
Python local variables should use snake_case naming:my_variable = ...
Python variable names that start with a number should be prefixed with 'k':k_99th_percentile = ...
Python global variables should use upper snake_case with prefix 'G':G_MY_GLOBAL = ...
Python constants should use upper snake_case naming:MY_CONSTANT = ...
Avoid shadowing variables declared in an outer scope in Python
Initialize all externally visible members of a Python class in the constructor
For Python interfaces that may be used outside a file, prefer docstrings over comments
Python comments should be reserved for code within a function, or interfaces that are local to a file
Use Google style docstrings in Python for classes and functions, which can be parsed by Sphinx
Python attributes and variables can be documented inline with type and description
Avoid using reflection in Python when functionality can be easily achieved without reflection
When using try-except blocks in Python, limit the except to the smallest set of errors possible
When using try-except blocks in Python to handle multiple possible variable types (duck-typing), keep the body of the try as small as possible, using the else block for logic
Files:
tests/unittest/_torch/auto_deploy/unit/singlegpu/transformations/library/test_bmm_moe_hooks.pytests/unittest/_torch/auto_deploy/unit/singlegpu/custom_ops/test_ad_moe_op.pytensorrt_llm/_torch/auto_deploy/custom_ops/fused_moe/torch_moe.pytensorrt_llm/_torch/auto_deploy/transform/library/sharding.pytensorrt_llm/_torch/auto_deploy/transform/library/fused_moe.py
**/*.{cpp,h,cu,cuh,py}
📄 CodeRabbit inference engine (CODING_GUIDELINES.md)
All TensorRT-LLM Open Source Software code should contain an NVIDIA copyright header that includes the year of its latest meaningful modification
Files:
tests/unittest/_torch/auto_deploy/unit/singlegpu/transformations/library/test_bmm_moe_hooks.pytests/unittest/_torch/auto_deploy/unit/singlegpu/custom_ops/test_ad_moe_op.pytensorrt_llm/_torch/auto_deploy/custom_ops/fused_moe/torch_moe.pytensorrt_llm/_torch/auto_deploy/transform/library/sharding.pytensorrt_llm/_torch/auto_deploy/transform/library/fused_moe.py
🧠 Learnings (10)
📓 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.
📚 Learning: 2025-10-20T17:09:21.560Z
Learnt from: nvchenghaoz
Repo: NVIDIA/TensorRT-LLM PR: 8469
File: tensorrt_llm/_torch/auto_deploy/transform/library/rms_norm.py:180-182
Timestamp: 2025-10-20T17:09:21.560Z
Learning: In tensorrt_llm/_torch/auto_deploy/transform/library/rms_norm.py, the _gated_rmsnorm_replacement function does not need to cast the output of torch.ops.auto_deploy.torch_rmsnorm_gated back to the input dtype, even though the custom op returns fp32. The dtype handling is managed elsewhere or the fp32 output is acceptable for downstream consumers.
Applied to files:
tests/unittest/_torch/auto_deploy/unit/singlegpu/custom_ops/test_ad_moe_op.py
📚 Learning: 2025-08-27T16:22:10.695Z
Learnt from: Fridah-nv
Repo: NVIDIA/TensorRT-LLM PR: 7227
File: tensorrt_llm/_torch/auto_deploy/utils/quantization_utils.py:94-100
Timestamp: 2025-08-27T16:22:10.695Z
Learning: When there are inconsistent operator detection methods (like custom_op() vs target_op()), removing one method and standardizing on the other is often cleaner than supporting both methods simultaneously.
Applied to files:
tests/unittest/_torch/auto_deploy/unit/singlegpu/custom_ops/test_ad_moe_op.py
📚 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:
tests/unittest/_torch/auto_deploy/unit/singlegpu/custom_ops/test_ad_moe_op.pytensorrt_llm/_torch/auto_deploy/custom_ops/fused_moe/torch_moe.pytensorrt_llm/_torch/auto_deploy/transform/library/fused_moe.py
📚 Learning: 2025-08-20T07:43:36.447Z
Learnt from: ChristinaZ
Repo: NVIDIA/TensorRT-LLM PR: 7068
File: cpp/tensorrt_llm/kernels/moeTopKFuncs.cuh:169-172
Timestamp: 2025-08-20T07:43:36.447Z
Learning: In TensorRT-LLM MOE kernels, when processing up to 128 experts across 32 threads, each thread handles at most 4 experts (N < 5 constraint), where N represents candidates per thread rather than total system capacity.
Applied to files:
tests/unittest/_torch/auto_deploy/unit/singlegpu/custom_ops/test_ad_moe_op.py
📚 Learning: 2025-08-21T02:39:12.009Z
Learnt from: djns99
Repo: NVIDIA/TensorRT-LLM PR: 7104
File: cpp/tensorrt_llm/kernels/cutlass_kernels/moe_gemm/moe_kernels.cu:1475-1480
Timestamp: 2025-08-21T02:39:12.009Z
Learning: The min latency mode functionality in TensorRT-LLM MOE kernels (cpp/tensorrt_llm/kernels/cutlass_kernels/moe_gemm/moe_kernels.cu) is deprecated and no longer being maintained/updated, as confirmed by djns99. Bug reports and optimization suggestions for the computeStridesTmaWarpSpecializedLowLatencyKernel and related min latency code paths should be deprioritized.
Applied to files:
tensorrt_llm/_torch/auto_deploy/custom_ops/fused_moe/torch_moe.py
📚 Learning: 2025-10-20T16:54:09.824Z
Learnt from: nvchenghaoz
Repo: NVIDIA/TensorRT-LLM PR: 8469
File: tensorrt_llm/_torch/auto_deploy/custom_ops/rms_norm.py:6-6
Timestamp: 2025-10-20T16:54:09.824Z
Learning: In tensorrt_llm/_torch/auto_deploy/custom_ops/rms_norm.py, the import `from ...modules.mamba.layernorm_gated import _layer_norm_fwd` is correct and should not be changed to modules.fla.layernorm_gated. The _layer_norm_fwd function exists in both modules/mamba/layernorm_gated.py and modules/fla/layernorm_gated.py, but the mamba version is the intended implementation for this use case.
Applied to files:
tensorrt_llm/_torch/auto_deploy/custom_ops/fused_moe/torch_moe.py
📚 Learning: 2025-10-20T17:07:18.745Z
Learnt from: nvchenghaoz
Repo: NVIDIA/TensorRT-LLM PR: 8469
File: tensorrt_llm/_torch/auto_deploy/models/patches/nemotron_h.py:98-116
Timestamp: 2025-10-20T17:07:18.745Z
Learning: In NemotronH models (tensorrt_llm/_torch/auto_deploy/models/patches/nemotron_h.py), the gate (self.gate) returns topk_indices and topk_weights that are already in the correct shape to be passed directly to torch_ops.auto_deploy.torch_moe without needing to reshape them when hidden_states is flattened.
Applied to files:
tensorrt_llm/_torch/auto_deploy/custom_ops/fused_moe/torch_moe.pytensorrt_llm/_torch/auto_deploy/transform/library/fused_moe.py
📚 Learning: 2025-08-19T12:45:11.997Z
Learnt from: amitz-nv
Repo: NVIDIA/TensorRT-LLM PR: 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/auto_deploy/custom_ops/fused_moe/torch_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/auto_deploy/custom_ops/fused_moe/torch_moe.py
🧬 Code graph analysis (3)
tests/unittest/_torch/auto_deploy/unit/singlegpu/transformations/library/test_bmm_moe_hooks.py (1)
tensorrt_llm/_torch/auto_deploy/transform/library/fused_moe.py (2)
_bmm_moe_down_split_hook(56-77)_bmm_moe_gate_up_split_hook(24-53)
tests/unittest/_torch/auto_deploy/unit/singlegpu/custom_ops/test_ad_moe_op.py (1)
tests/unittest/_torch/helpers.py (1)
reference_moe_torch(96-121)
tensorrt_llm/_torch/auto_deploy/transform/library/fused_moe.py (2)
tensorrt_llm/_torch/auto_deploy/utils/node_utils.py (3)
is_op(219-242)extract_op_args(557-594)shape(920-923)tensorrt_llm/_torch/auto_deploy/custom_ops/fused_moe/torch_moe.py (1)
torch_moe(90-149)
🪛 Ruff (0.14.10)
tensorrt_llm/_torch/auto_deploy/transform/library/sharding.py
1448-1448: Comment contains ambiguous × (MULTIPLICATION SIGN). Did you mean x (LATIN SMALL LETTER X)?
(RUF003)
tensorrt_llm/_torch/auto_deploy/transform/library/fused_moe.py
27-27: Unused function argument: args
(ARG001)
51-51: zip() without an explicit strict= parameter
Add explicit value for parameter strict=
(B905)
59-59: Unused function argument: args
(ARG001)
76-76: zip() without an explicit strict= parameter
Add explicit value for parameter strict=
(B905)
1169-1171: Avoid specifying long messages outside the exception class
(TRY003)
1179-1182: Avoid specifying long messages outside the exception class
(TRY003)
⏰ 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 (10)
tensorrt_llm/_torch/auto_deploy/transform/library/fused_moe.py (3)
100-119: LGTM - Clean refactoring of MoE node processing.The extraction of the processing loop with immediate dead code elimination is a good pattern for managing GPU memory during large model transformations.
186-224: Dtype casting and routing logic look correct.The dtype handling ensures activation dtype matches weight dtype for the fused kernel, and the
apply_routing_on_inputlogic correctly substitutesones_liketo prevent double-application of routing weights.
996-1020: Good defensive validation for Llama4-shaped weights.The shape validation ensures the BMM MoE pattern only matches tensors with the expected dimensional relationships before proceeding with the transformation.
tests/unittest/_torch/auto_deploy/unit/singlegpu/custom_ops/test_ad_moe_op.py (1)
4-4: Import change looks correct.The import now correctly references only
reference_moe_torchwhich is used by the remaining tests.tensorrt_llm/_torch/auto_deploy/custom_ops/fused_moe/torch_moe.py (1)
112-124: Docstring updates are clear and accurate.The updated parameter documentation correctly describes the per-expert weight list format and the routing behavior options.
tests/unittest/_torch/auto_deploy/unit/singlegpu/transformations/library/test_bmm_moe_hooks.py (1)
1-249: Comprehensive test coverage for new BMM MoE hooks.The test suite provides good coverage including:
- Parameterized tests for various expert/dimension configurations
- Content validation ensuring correct tensor splitting and transposition
- Edge case handling for missing source keys
- Module prefix handling
- Integration test for the complete flow
Note: These tests confirm the hooks work without
num_expertsparameter, which indicates thepartial()calls infused_moe.pythat passnum_expertsare incorrect.tensorrt_llm/_torch/auto_deploy/transform/library/sharding.py (4)
1401-1412: Clean partition helper with correct remainder handling.The
get_partitionfunction correctly handles the case wherenum_experts % world_size != 0by assigning remainder experts to the last rank.
1466-1470: Good cleanup of unused expert weights.The code properly eliminates dead code and removes the unused expert weight attributes, which is important for memory management with large MoE models.
2448-2451: MLPType inference logic is straightforward.The heuristic checking if
args[5](w3_weight/gate list) is non-empty to determine GATED_MLP vs MLP is clear and aligns with the per-expert weight list convention.
1414-1416: Code correctly handles uneven expert distribution; no action needed.The code already explicitly accounts for cases where
ep_sizedoesn't evenly dividenum_experts. Theget_partitionfunction (lines 1406–1410) assigns remaining experts to the last rank via conditional logic. Similarly, the selected_experts sharding (lines 1383–1384) usestorch.gefor the last rank instead oftorch.eq, correctly masking experts for the asymmetric case. The final_scales follow the same logic. Downstream all_reduce (line 1458) sums outputs across ranks, which is mathematically correct regardless of asymmetry.
|
PR_Github #29928 [ run ] triggered by Bot. Commit: |
Signed-off-by: Tal Cherckez <127761168+tcherckez-nvidia@users.noreply.github.com>
|
/bot run |
|
PR_Github #29935 [ run ] triggered by Bot. Commit: |
|
PR_Github #29935 [ run ] completed with state
|
lucaslie
left a comment
lucaslie
left a comment
There was a problem hiding this comment.
Overall interface looks good!
|
|
||
| # Now create get_attr nodes for each expert weight | ||
| # These must be created within the insertion context for proper graph ordering | ||
| with graph.inserting_before(output_node): |
There was a problem hiding this comment.
Let's insert them with the other get_attr nodes. The graph is usually sorted such that all input nodes appear first, then all the get_attr nodes, then all the compute nodes.
You can get the last get_attr node like this:
last_get_attr_node = graph.find_nodes(target="get_attr")[-1]
| gm.graph.eliminate_dead_code() | ||
| for expert in ( | ||
| w_up_list_to_remove + w_down_list_to_remove + w_gate_list_to_remove + scales_to_remove | ||
| ): | ||
| delattr(gm, expert.target) | ||
|
|
There was a problem hiding this comment.
why is this explicit clean up needed rather than using the generic utility in the outer call function?
| if args[5] and len(args[5]) > 0: | ||
| mlp_type = MLPType.GATED_MLP |
There was a problem hiding this comment.
why not use the argument that you now have for this?
There was a problem hiding this comment.
can we now also remove torch_moe_dense_mlp ? I see it's still being used in mxfp4_moe.py
4 participants
Summary by CodeRabbit
Refactor
Tests
✏️ Tip: You can customize this high-level summary in your review settings.
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
Update tava architecture diagram if there is a significant design change in PR.
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.