feat: binding type build argument (pybind, nanobind)

[Linda-Stadter]

binding type build argument (pybind, nanobind)

• Adds a build argument to specify which library to use for the Python/C++ bindings (pybind or nanobind, default: pybind)
• Adds nanobind as a 3rdparty submodule
• Uses nanobind's stubgen tool when building with nanobind. This creates only one bindings.pyi file in contrast to pybind which creates a folder structure containing multiple *.pyi files (see also changed sanity check)
• build_wheel.py reads the previous binding_type from cache and cleans up build artifacts related to bindings and triggers a cmake_configure_command when the binding_type changed
• The nanobind bindings are currently under development and will be added in an upcoming PR

Testing

Building with nanobind:
python scripts/build_wheel.py --binding_type nanobind --trt_root /usr/local/tensorrt --cuda_architectures "native" -D FAST_BUILD=ON --use_ccache

Verifying binding type:
cd tensorrt_llm && python -c "import bindings; print("bindings.binding_type")"
should return
nanobind

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 [--disable-fail-fast --skip-test --stage-list "A10-1, xxx" --gpu-type "A30, H100_PCIe" --add-multi-gpu-test --only-multi-gpu-test --disable-multi-gpu-test --post-merge --extra-stage "H100_PCIe-[Post-Merge]-1, xxx"]

Launch build/test pipelines. All previously running jobs will be killed.

--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 "A10-1, xxx" (OPTIONAL) : Only run the specified test stages. Examples: "A10-1, 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.

--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. Will also run 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-[Post-Merge]-1, xxx" (OPTIONAL) : Run the ordinary L0 pre-merge pipeline and specified test stages. Examples: --extra-stage "H100_PCIe-[Post-Merge]-1, xxx".

For guidance on mapping tests to stage names, see docs/source/reference/ci-overview.md.

kill

kill

Kill all running builds associated with pull request.

skip

skip --comment COMMENT

Skip 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-pipeline

Reuse 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.

[Linda-Stadter]

/bot run

[tensorrt-cicd]

PR_Github #11175 [ run ] triggered by Bot

[tensorrt-cicd]

PR_Github #11175 [ run ] completed with state SUCCESS
/LLM/main/L0_MergeRequest_PR pipeline #8266 completed with status: 'FAILURE'

[Linda-Stadter]

/bot run

[tensorrt-cicd]

PR_Github #11254 [ run ] triggered by Bot

[tensorrt-cicd]

PR_Github #11254 [ run ] completed with state SUCCESS
/LLM/main/L0_MergeRequest_PR pipeline #8323 completed with status: 'SUCCESS'
Pipeline passed with automatic retried tests. Check the rerun report for details.

[Linda-Stadter]

/bot run

[tensorrt-cicd]

PR_Github #11314 [ run ] triggered by Bot

[tensorrt-cicd]

PR_Github #11314 [ run ] completed with state SUCCESS
/LLM/main/L0_MergeRequest_PR pipeline #8367 completed with status: 'SUCCESS'
Pipeline passed with automatic retried tests. Check the rerun report for details.

[Linda-Stadter]

/bot run

[tensorrt-cicd]

PR_Github #11417 [ run ] triggered by Bot

[tensorrt-cicd]

PR_Github #11416 [ run ] triggered by Bot

[tensorrt-cicd]

PR_Github #11417 [ run ] completed with state ABORTED

[tensorrt-cicd]

PR_Github #11416 [ run ] completed with state SUCCESS
/LLM/main/L0_MergeRequest_PR pipeline #8443 completed with status: 'FAILURE'

[Linda-Stadter]

/bot run

[tensorrt-cicd]

PR_Github #11439 [ run ] triggered by Bot

[tensorrt-cicd]

PR_Github #11439 [ run ] completed with state FAILURE
/LLM/main/L0_MergeRequest_PR pipeline #8461 completed with status: 'FAILURE'

[Linda-Stadter]

/bot run

[tensorrt-cicd]

PR_Github #11462 [ run ] triggered by Bot

[tensorrt-cicd]

PR_Github #11462 [ run ] completed with state SUCCESS
/LLM/main/L0_MergeRequest_PR pipeline #8482 completed with status: 'FAILURE'

[Linda-Stadter]

/bot run

[tensorrt-cicd]

PR_Github #11534 [ run ] triggered by Bot

[tensorrt-cicd]

PR_Github #11534 [ run ] completed with state SUCCESS
/LLM/main/L0_MergeRequest_PR pipeline #8538 completed with status: 'SUCCESS'

[ziyixiong-nv]

@Linda-Stadter Could you also update the section "Compile and Install TensorRT-LLM" in https://github.com/NVIDIA/TensorRT-LLM/blob/main/docs/source/blogs/Best_perf_practice_on_DeepSeek-R1_in_TensorRT-LLM.md .

[Linda-Stadter]

@Linda-Stadter Could you also update the section "Compile and Install TensorRT-LLM" in https://github.com/NVIDIA/TensorRT-LLM/blob/main/docs/source/blogs/Best_perf_practice_on_DeepSeek-R1_in_TensorRT-LLM.md .

Thanks for the comment! I haven't done any doc updates because this build flag is only temporary. I want to completely remove our pybind bindings and have nanobind as the only option. And currently, nanobind is not fully functional yet and can cause problems