tensorflow-build-archived
tensorflow-build-archived copied to clipboard
TF 2.0 for macOS Catalina?
Is there any way I could request you to upload the file for TF 2.0 (macOS Catalina) ? It'd be greatly appreciated! Thanks!
Hi there, I have tried building TF2.0 on Catalina but I keep receiving build failures from protobuf
ERROR: /private/var/tmp/_bazel_lakshayg/3fd65ba6e1ece8419032ab8d359c9789/external/com_google_protobuf/BUILD:294:1: C++ compilation of rule '@com_google_protobuf//:protoc_lib' failed (Exit 1)
external/com_google_protobuf/src/google/protobuf/compiler/command_line_interface.cc:450:3: error: use of undeclared identifier 'STLDeleteValues'
STLDeleteValues(&files_);
^
external/com_google_protobuf/src/google/protobuf/compiler/command_line_interface.cc:899:30: error: cannot initialize a variable of type 'google::protobuf::compiler::CommandLineInterface::GeneratorContextImpl **' with an rvalue of type 'std::__1::unordered_map<std::__1::basic_string<char>, std::__1::unique_ptr<google::protobuf::compiler::CommandLineInterface::GeneratorContextImpl, std::__1::default_delete<google::protobuf::compiler::CommandLineInterface::GeneratorContextImpl> >, std::__1::hash<std::__1::basic_string<char> >, std::__1::equal_to<std::__1::basic_string<char> >, std::__1::allocator<std::__1::pair<const std::__1::basic_string<char>, std::__1::unique_ptr<google::protobuf::compiler::CommandLineInterface::GeneratorContextImpl, std::__1::default_delete<google::protobuf::compiler::CommandLineInterface::GeneratorContextImpl> > > > >::mapped_type *' (aka 'std::__1::unique_ptr<google::protobuf::compiler::CommandLineInterface::GeneratorContextImpl, std::__1::default_delete<google::protobuf::compiler::CommandLineInterface::GeneratorContextImpl> > *')
GeneratorContextImpl** map_slot = &output_directories[output_location];
^ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
external/com_google_protobuf/src/google/protobuf/compiler/command_line_interface.cc:907:9: error: use of undeclared identifier 'STLDeleteValues'
STLDeleteValues(&output_directories);
^
external/com_google_protobuf/src/google/protobuf/compiler/command_line_interface.cc:917:27: error: no viable conversion from 'std::__1::unique_ptr<google::protobuf::compiler::CommandLineInterface::GeneratorContextImpl, std::__1::default_delete<google::protobuf::compiler::CommandLineInterface::GeneratorContextImpl> >' to 'google::protobuf::compiler::CommandLineInterface::GeneratorContextImpl *'
GeneratorContextImpl* directory = iter->second;
^ ~~~~~~~~~~~~
external/com_google_protobuf/src/google/protobuf/compiler/command_line_interface.cc:920:9: error: use of undeclared identifier 'STLDeleteValues'
STLDeleteValues(&output_directories);
^
external/com_google_protobuf/src/google/protobuf/compiler/command_line_interface.cc:929:9: error: use of undeclared identifier 'STLDeleteValues'
STLDeleteValues(&output_directories);
^
external/com_google_protobuf/src/google/protobuf/compiler/command_line_interface.cc:943:3: error: use of undeclared identifier 'STLDeleteValues'
STLDeleteValues(&output_directories);
^
external/com_google_protobuf/src/google/protobuf/compiler/command_line_interface.cc:1961:27: error: no viable conversion from 'const std::__1::unique_ptr<google::protobuf::compiler::CommandLineInterface::GeneratorContextImpl, std::__1::default_delete<google::protobuf::compiler::CommandLineInterface::GeneratorContextImpl> >' to 'google::protobuf::compiler::CommandLineInterface::GeneratorContextImpl *'
GeneratorContextImpl* directory = iter->second;
^ ~~~~~~~~~~~~
8 errors generated.
I haven't been able to find time to dig deeper into the issue so I'm just waiting for some bugfixes from the TF team
Hi there!
I built Tensorflow 2.1 for macOS catalina with support for AVX, AVX2, FMA, SSE4.1 and SSE4.2. You can find the wheel file here: https://github.com/reuank/tensorflow-wheels-macOS/releases/tag/tensorflow-2.1-catalina
Same error while running
bazel build tensorflow/tools/graph_transforms:transform_graph
https://github.com/tensorflow/tensorflow/tree/master/tensorflow/tools/graph_transforms#inspecting-graphs
Is there anyway to use the transform_graph
command line tool from the tensorflow/tensorflow
Docker image ?