menoh

menoh copied to clipboard

Failure of MSVC build on Appveyor

2

MSVC build on Appveyor fails with the following error: https://ci.appveyor.com/project/pfnet-research/menoh/builds/34608060/job/awhcqncaxowfypvm ``` onnx_menoh_onnx.pb.cc onnx-operators_menoh_onnx.pb.cc C:\protobuf-3.6.0-msvc\include\google/protobuf/stubs/logging.h(100): error C2220: warning treated as error - no 'object' file generated (compiling source file C:\projects\menoh\build\external\onnx\onnx\onnx-operators_menoh_onnx.pb.cc) [C:\projects\menoh\build\external\onnx\onnx_proto.vcxproj]...

msakai

Menoh error loading model from onnx file

5

Hi I'm trying to run Menoh with the ONNX file converted from the Keras/Tensorflow model. The following error occurs when calling the menoh function "make_model_data_from_onnx". menoh invalid attribute type error:...

kmorito

Please allow to use the external onnx

1

I have onnx as a package on FreeBSD, and would like to just use it instead of bundling it. Could you please create the cmake option that would allow to...

yurivict

error: no type named 'format' in 'mkldnn::memory'

3

I have mkl-dnn-1.0.1 installed, and clang8 complains: ``` In file included from /usr/ports/math/menoh/work/menoh-1.1.1-169-gc00e8d5/menoh/composite_backend/backend/mkldnn/memory_cache.cpp:1: In file included from /usr/ports/math/menoh/work/menoh-1.1.1-169-gc00e8d5/menoh/composite_backend/backend/mkldnn/memory_cache.hpp:10: /usr/ports/math/menoh/work/menoh-1.1.1-169-gc00e8d5/menoh/composite_backend/backend/mkldnn/memory_conversion.hpp:12:49: error: no type named 'format' in 'mkldnn::memory' bool is_data_format(mkldnn::memory::format format); ~~~~~~~~~~~~~~~~^...

yurivict

support ARM NN backend (https://developer.arm.com/products/processors/machine-learning/arm-nn) 1. MKLDNN is not required 2. [ARM NN v18.05.02](https://github.com/ARM-software/armnn/tree/v18.05.02) and [ARM Compute Library : v18.05](https://github.com/ARM-software/ComputeLibrary/tree/v18.05) I tested it in the following environment. (example_arm/vgg16_example_in_cpp_arm.cpp) - x86-64、not...

vengineer

This PR adds TensorRT backend. It enables fast inference by using NVIDIA's TensorRT. - [x] Modify CMakeLists.txt for merging - [x] Add a license - [ ] Testing

okdshin

add more libraries to Libs.private for static linking

2

We need to list all privately used libraries in `Libs.private` for static linking to work. Note that `-lstdc++` and `-lm` seem to work also on macOS.

msakai

fix install command for menoh_onnx and menoh_onnx_proto

1

Modification for #215

okdshin

make install fails for static library build

1

If `-DBUILD_SHARED_LIBS=OFF` is specified, `make install` fails with following error: ``` $ cmake .. -DBUILD_SHARED_LIBS=OFF -DENABLE_TEST=OFF -DENABLE_EXAMPLE=ON && make -- The C compiler identification is AppleClang 10.0.0.10001044 -- The CXX...

msakai

CMake enforce Protobuf 2.6.1

3

WilliamTambellini