Skip to content

build.sh: don't require user to set --use_full_protobuf with --use_tensorrt option. we can set it implicitly. #780

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
snnn merged 2 commits into from
Apr 6, 2019
Merged

build.sh: don't require user to set --use_full_protobuf with --use_tensorrt option. we can set it implicitly. #780

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
d28af4c
use_full_protobuf if tensorrt build option is enabled.
jywu-msft Apr 5, 2019
2c84bc2
update BUILD.md sections on MKLDNN and TensorRT/full_protobuf option
jywu-msft Apr 5, 2019
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 4 additions & 4 deletions BUILD.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -113,8 +113,9 @@ _Side note: If you have multiple versions of CUDA installed on a Windows machine
e.g. C:\Program Files (x86)\Microsoft Visual Studio\2017\Enterprise\Common7\IDE\VC\VCTargets\BuildCustomizations\.
If you want to build with an earlier version, you must temporarily remove the 'CUDA x.y.*' files for later versions from this directory._

### MKL-DNN
To build ONNX Runtime with MKL-DNN support, build it with `./build.sh --use_mkldnn --use_mklml`
### MKL-DNN/MKLML
To build ONNX Runtime with MKL-DNN support, build it with `./build.sh --use_mkldnn`
To build ONNX Runtime using MKL-DNN built with dependency on MKL small libraries, build it with `./build.sh --use_mkldnn --use_mklml`

### TensorRT
ONNX Runtime supports the TensorRT execution provider (released as preview). You will need to download and install [CUDA](https://developer.nvidia.com/cuda-toolkit), [CUDNN](https://developer.nvidia.com/cudnn) and [TensorRT](https://developer.nvidia.com/nvidia-tensorrt-download).
Expand All @@ -125,12 +126,11 @@ The TensorRT execution provider for ONNX Runtime is built and tested with CUDA 9
- The path to the CUDA `bin` directory must be added to the PATH environment variable so that `nvcc` is found.
- The path to the CUDNN installation (path to folder that contains libcudnn.so) must be provided via the CUDNN_PATH environment variable, or `--cudnn_home parameter`.
- The path to TensorRT installation must be provided via the `--tensorrt_home parameter`.
- There are two kind of protobufs: protobuf and protobuf-lite. Currently TensorRT execution provider only supports protobuf by using `--use_full_protobuf` option.

You can build from source on Linux by using the following `cmd` from the onnxruntime directory:

```
./build.sh --use_full_protobuf --cudnn_home <path to CUDNN e.g. /usr/lib/x86_64-linux-gnu/> --cuda_home <path to folder for CUDA e.g. /usr/local/cuda> --use_tensorrt --tensorrt_home <path to TensorRT home> (Linux)
./build.sh --cudnn_home <path to CUDNN e.g. /usr/lib/x86_64-linux-gnu/> --cuda_home <path to folder for CUDA e.g. /usr/local/cuda> --use_tensorrt --tensorrt_home <path to TensorRT home> (Linux)

```

Expand Down
3 changes: 2 additions & 1 deletion tools/ci_build/build.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -320,7 +320,8 @@ def generate_build_tree(cmake_path, source_dir, build_dir, cuda_home, cudnn_home
# By default - we currently support only cross compiling for ARM/ARM64 (no native compilation supported through this script)
"-Donnxruntime_CROSS_COMPILING=" + ("ON" if args.arm64 or args.arm else "OFF"),
"-Donnxruntime_BUILD_x86=" + ("ON" if args.x86 else "OFF"),
"-Donnxruntime_USE_FULL_PROTOBUF=" + ("ON" if args.use_full_protobuf else "OFF"),
# TensorRT provider currently only supports full_protobuf option.
"-Donnxruntime_USE_FULL_PROTOBUF=" + ("ON" if args.use_full_protobuf or args.use_tensorrt else "OFF"),
"-Donnxruntime_DISABLE_CONTRIB_OPS=" + ("ON" if args.disable_contrib_ops else "OFF"),
]
if args.use_brainslice:
Expand Down
2 changes: 1 addition & 1 deletion tools/ci_build/github/azure-pipelines/linux-gpu-tensorrt-ci-pipeline.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,7 @@ jobs:
- template: templates/set-test-data-variables-step.yml

# There are some tests in 20190130.zip that TensorRT can't run. Instead here use 20181210 opset8 for TensorRT test.
- script: 'tools/ci_build/github/linux/run_dockerbuild.sh -o ubuntu16.04 -d tensorrt -r $(Build.BinariesDirectory) -x "--use_full_protobuf --test_data_url https://onnxruntimetestdata.blob.core.windows.net/models/20181210.zip --test_data_checksum a966def7447f4ff04f5665bca235b3f3"'
- script: 'tools/ci_build/github/linux/run_dockerbuild.sh -o ubuntu16.04 -d tensorrt -r $(Build.BinariesDirectory) -x "--test_data_url https://onnxruntimetestdata.blob.core.windows.net/models/20181210.zip --test_data_checksum a966def7447f4ff04f5665bca235b3f3"'

displayName: 'Command Line Script'

Expand Down
6 changes: 3 additions & 3 deletions tools/ci_build/github/azure-pipelines/win-gpu-tensorrt-ci-pipeline.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -66,7 +66,7 @@ jobs:
displayName: 'Download test data and generate cmake config'
inputs:
filename: '$(Build.BinariesDirectory)\packages\python\python.exe'
arguments: '$(Build.SourcesDirectory)\tools\ci_build\build.py --config Debug Release --build_dir $(Build.BinariesDirectory) --cmake_path $(Build.BinariesDirectory)\cmake\bin\cmake.exe --ctest_path $(Build.BinariesDirectory)\cmake\bin\ctest.exe --enable_pybind --use_openmp --use_mkldnn --use_mkldnn --build_shared_lib --enable_onnx_tests --cuda_home="C:\local\cuda_10.0.130_win10" --cudnn_home="C:\local\cudnn-10.0-windows10-x64-v7.3.1.20\cuda" --use_full_protobuf --use_tensorrt --tensorrt_home="C:\local\TensorRT-5.0.4.3" --test_data_url $(TestDataUrl) --test_data_checksum $(TestDataChecksum) --update --msvc_toolset=14.11'
arguments: '$(Build.SourcesDirectory)\tools\ci_build\build.py --config Debug Release --build_dir $(Build.BinariesDirectory) --cmake_path $(Build.BinariesDirectory)\cmake\bin\cmake.exe --ctest_path $(Build.BinariesDirectory)\cmake\bin\ctest.exe --enable_pybind --use_openmp --use_mkldnn --build_shared_lib --enable_onnx_tests --cuda_home="C:\local\cuda_10.0.130_win10" --cudnn_home="C:\local\cudnn-10.0-windows10-x64-v7.3.1.20\cuda" --use_tensorrt --tensorrt_home="C:\local\TensorRT-5.0.4.3" --test_data_url $(TestDataUrl) --test_data_checksum $(TestDataChecksum) --update --msvc_toolset=14.11'
workingDirectory: "$(Build.BinariesDirectory)"

- task: VSBuild@1
Expand All @@ -83,7 +83,7 @@ jobs:
displayName: 'Test Debug'
inputs:
filename: '$(Build.BinariesDirectory)\packages\python\python.exe'
arguments: '$(Build.SourcesDirectory)\tools\ci_build\build.py --config Debug --build_dir $(Build.BinariesDirectory) --cmake_path $(Build.BinariesDirectory)\cmake\bin\cmake.exe --ctest_path $(Build.BinariesDirectory)\cmake\bin\ctest.exe --enable_pybind --use_openmp --use_mkldnn --use_mkldnn --build_shared_lib --enable_onnx_tests --cuda_home="C:\local\cuda_10.0.130_win10" --cudnn_home="C:\local\cudnn-10.0-windows10-x64-v7.3.1.20\cuda" --use_full_protobuf --use_tensorrt --tensorrt_home="C:\local\TensorRT-5.0.4.3" --test_data_url $(TestDataUrl) --test_data_checksum $(TestDataChecksum) --test'
arguments: '$(Build.SourcesDirectory)\tools\ci_build\build.py --config Debug --build_dir $(Build.BinariesDirectory) --cmake_path $(Build.BinariesDirectory)\cmake\bin\cmake.exe --ctest_path $(Build.BinariesDirectory)\cmake\bin\ctest.exe --enable_pybind --use_openmp --use_mkldnn --build_shared_lib --enable_onnx_tests --cuda_home="C:\local\cuda_10.0.130_win10" --cudnn_home="C:\local\cudnn-10.0-windows10-x64-v7.3.1.20\cuda" --use_tensorrt --tensorrt_home="C:\local\TensorRT-5.0.4.3" --test_data_url $(TestDataUrl) --test_data_checksum $(TestDataChecksum) --test'
workingFolder: '$(Build.BinariesDirectory)'
- task: VSBuild@1
displayName: 'Build C# Debug'
Expand Down Expand Up @@ -119,7 +119,7 @@ jobs:
displayName: 'Test Release'
inputs:
filename: '$(Build.BinariesDirectory)\packages\python\python.exe'
arguments: '$(Build.SourcesDirectory)\tools\ci_build\build.py --config Release --build_dir $(Build.BinariesDirectory) --cmake_path $(Build.BinariesDirectory)\cmake\bin\cmake.exe --ctest_path $(Build.BinariesDirectory)\cmake\bin\ctest.exe --enable_pybind --use_openmp --use_mkldnn --build_shared_lib --enable_onnx_tests --cuda_home="C:\local\cuda_10.0.130_win10" --cudnn_home="C:\local\cudnn-10.0-windows10-x64-v7.3.1.20\cuda" --use_full_protobuf --use_tensorrt --tensorrt_home="C:\local\TensorRT-5.0.4.3" --test_data_url $(TestDataUrl) --test_data_checksum $(TestDataChecksum) --test'
arguments: '$(Build.SourcesDirectory)\tools\ci_build\build.py --config Release --build_dir $(Build.BinariesDirectory) --cmake_path $(Build.BinariesDirectory)\cmake\bin\cmake.exe --ctest_path $(Build.BinariesDirectory)\cmake\bin\ctest.exe --enable_pybind --use_openmp --use_mkldnn --build_shared_lib --enable_onnx_tests --cuda_home="C:\local\cuda_10.0.130_win10" --cudnn_home="C:\local\cudnn-10.0-windows10-x64-v7.3.1.20\cuda" --use_tensorrt --tensorrt_home="C:\local\TensorRT-5.0.4.3" --test_data_url $(TestDataUrl) --test_data_checksum $(TestDataChecksum) --test'
workingFolder: "$(Build.BinariesDirectory)"

- task: VSBuild@1
Expand Down