./build.sh lattice_img fails

[capaulson]

I'm having some difficultly building the docker image. The final few lines I see are:

 C10_DEPRECATED inline MemoryFormat get_contiguous_memory_format() {
                                               ^~~~~~~~~~~~~~~~~~~~~       
[1585/2460] Building NVCC (Device) object caffe2/CMakeFiles/torch.dir/__/aten/src/ATen/native/cuda/torch_generated_BinaryCompareKernel.cu.o
[1586/2460] Building NVCC (Device) object caffe2/CMakeFiles/torch.dir/__/aten/src/ATen/native/cuda/torch_generated_SpectralOps.cu.o
[1587/2460] Building NVCC (Device) object caffe2/CMakeFiles/torch.dir/__/aten/src/ATen/native/cuda/torch_generated_PointwiseOpsKernel.cu.o
The command '/bin/sh -c MAX_JOBS=8 USE_NINJA=1 BUILD_TEST=0 BUILD_JNI=0  python3 setup.py install' returned a non-zero code: 137

There are many warnings, but I didn't see any errors.

[capaulson]

There's a bit of error text far above the end in some of the python install stuff:

Building wheels for collected packages: pyyaml, typing
  Running setup.py bdist_wheel for pyyaml: started
  Running setup.py bdist_wheel for pyyaml: finished with status 'error'
  Complete output from command /usr/bin/python3 -u -c "import setuptools, tokenize;__file__='/tmp/pip-build-j2x8nd7t/pyyaml/setup.py';f=getattr(tokenize, 'open', open)(__file__);code=f.read().replace('\r\n', '\n');f.close();exec(compile(code, __file__, 'exec'))" bdist_wheel -d /tmp/tmpuov9lf7_pip-wheel- --python-tag cp36:
  usage: -c [global_opts] cmd1 [cmd1_opts] [cmd2 [cmd2_opts] ...]
     or: -c --help [cmd1 cmd2 ...]
     or: -c --help-commands
     or: -c cmd --help
  
  error: invalid command 'bdist_wheel'
  
  ----------------------------------------
  Failed building wheel for pyyaml
  Running setup.py clean for pyyaml
  Running setup.py bdist_wheel for typing: started
  Running setup.py bdist_wheel for typing: finished with status 'error'
  Complete output from command /usr/bin/python3 -u -c "import setuptools, tokenize;__file__='/tmp/pip-build-j2x8nd7t/typing/setup.py';f=getattr(tokenize, 'open', open)(__file__);code=f.read().replace('\r\n', '\n');f.close();exec(compile(code, __file__, 'exec'))" bdist_wheel -d /tmp/tmpgpxvhh2rpip-wheel- --python-tag cp36:
  usage: -c [global_opts] cmd1 [cmd1_opts] [cmd2 [cmd2_opts] ...]
     or: -c --help [cmd1 cmd2 ...]
     or: -c --help-commands
     or: -c cmd --help
  
  error: invalid command 'bdist_wheel'
  
  ----------------------------------------
  Failed building wheel for typing
  Running setup.py clean for typing
Failed to build pyyaml typing
Installing collected packages: numpy, pyyaml, intel-openmp, mkl, mkl-include, typing
  Running setup.py install for pyyaml: started
    Running setup.py install for pyyaml: finished with status 'done'
  Running setup.py install for typing: started
    Running setup.py install for typing: finished with status 'done'
Successfully installed intel-openmp-2020.0.133 mkl-2019.0 mkl-include-2019.0 numpy-1.19.1 pyyaml-5.3.1 typing-3.7.4.3

[RaduAlexandru]

This may be indicative of an out-of-memory error.
What system are you running on?
If it happens to be MacOS or Windows you can try increasing docker's memory resources in the docker tab->preferences->advanced.
Also in the Dockerfile you may try decreasing the number of threads used to build pytorch and torchvision by changing MAX_JOBS=8 to something like 4 or 2.

[scho419]

Is it possible to run this on macOS? would be great to know how the code was tested and the OS it was on.