Apologies in advance for the text dump
I am attempting to compile Tensorflow in OSX 10.13.4 using Bazel based on these instructions: https://medium.com/xplore-ai/nvidia-egpu-macos-tensorflow-gpu-the-definitive-setup-guide-to-avoid-headaches-f40e831f26ea
I attempted the build using the following command:
bazel build --config=cuda --config=opt --cxxopt="-D_GLIBCXX_USE_CXX11_ABI=0" --action_env PATH --action_env LD_LIBRARY_PATH --action_env DYLD_LIBRARY_PATH //tensorflow/tools/pip_package:build_pip_package
Here are the versions of all relevant programs: Python: 3.6 CUDA: 10.0 cuDNN: 7.4 Tensorflow: https://github.com/zylo117/tensorflow-gpu-macosx GPU: NVIDIA GTX 980 Bazel Version: 0.16.1
Here is the verbose error information along with some run info:
5 warnings generated.
ERROR: /Users/brianmoser/tensorflow-gpu-macosx/tensorflow/core/kernels/BUILD:3423:1: error while parsing .d file: /private/var/tmp/_bazel_brianmoser/9608e82147ff56f68b42fd19bee93cb0/execroot/org_tensorflow/bazel-out/darwin-opt/bin/tensorflow/core/kernels/_objs/bincount_op_gpu/bincount_op_gpu.cu.d (No such file or directory) nvcc fatal : The version ('10.0') of the host compiler ('Apple clang') is not supported Target //tensorflow/tools/pip_package:build_pip_package failed to build Use --verbose_failures to see the command lines of failed build steps. INFO: Elapsed time: 1407.017s, Critical Path: 63.10s INFO: 2389 processes: 2389 local. FAILED: Build did NOT complete successfully
Additionally, here is the final step the build conducted:
tensorflow/stream_executor/cuda/cuda_dnn.cc:1506:19: warning: private field 'data_type_' is not used [-Wunused-private-field] cudnnDataType_t data_type_;
Finally, here is the final relevant information:
git rev-parse HEAD 795d7c36152a1f96adca0c48f4537e500f5ad36b
bazel version Build label: 0.16.1
Build target:
bazel-out/darwin-opt/bin/src/main/java/com/google/devtools/build/lib/bazel/BazelServer_deploy.jar
Does anybody have any idea what could be the reason for this error?