docker_pi.sh

git clone https://github.com/tensorflow/tensorflow tensorflow
cd tensorflow
docker build -t "tf-pi" -f tensorflow/tools/ci_build/Dockerfile.cpu tensorflow/tools/ci_build/
docker run -it tf-pi
# Commands after this are run on the Docker instance.
dpkg --add-architecture armhf
echo 'deb [arch=armhf] http://ports.ubuntu.com/ trusty main restricted universe multiverse' >> /etc/apt/sources.list.d/armhf.list
echo 'deb [arch=armhf] http://ports.ubuntu.com/ trusty-updates main restricted universe multiverse' >> /etc/apt/sources.list.d/armhf.list
echo 'deb [arch=armhf] http://ports.ubuntu.com/ trusty-security main restricted universe multiverse' >> /etc/apt/sources.list.d/armhf.list
echo 'deb [arch=armhf] http://ports.ubuntu.com/ trusty-backports main restricted universe multiverse' >> /etc/apt/sources.list.d/armhf.list
sed -i 's#deb http://archive.ubuntu.com/ubuntu/#deb [arch=amd64] http://archive.ubuntu.com/ubuntu/#g' /etc/apt/sources.list
apt-get update
apt-get install -y libpython-all-dev:armhf
echo "deb [arch=amd64] http://storage.googleapis.com/bazel-apt stable jdk1.8" | sudo tee /etc/apt/sources.list.d/bazel.list
curl https://bazel.build/bazel-release.pub.gpg | sudo apt-key add -
apt-get update
rm -rf /usr/local/bin/bazel
apt-get install -y bazel python python-numpy python-dev python-pip
git clone https://github.com/tensorflow/tensorflow tensorflow
cd tensorflow
third_party/toolchains/cpus/arm/build_raspberry_pi.sh
# On the host OS (change the path to the actual temp directory shown by the build output).
docker cp `docker ps -alq`:/tmp/tmp.smXLh20Qhj/tensorflow-1.3.0rc0-cp27-none-any.whl ~/Downloads/

I think the second to last command is not necessary, but the last command needs to know the true tmpdir

You're right, I just had it because I was having to restart the bazel build manually after hitting RAM limits and the OOM killer, so the rest of the script was being skipped. I've updated it.

You'll also need to apply this patch to the source to get NEON Pi 2/3 builds working, until the Eigen version is updated:
lissyx/tensorflow@f7def94

tried this guy to build a PI_ONE wheel. successfully generated and successfully installed, but on python import i'm seeing

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/__init__.py", line 24, in <module>
    from tensorflow.python import *
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/__init__.py", line 63, in <module>
    from tensorflow.python.framework.framework_lib import *
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/framework_lib.py", line 102, in <module>
    from tensorflow.python.framework.importer import import_graph_def
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/importer.py", line 30, in <module>
    from tensorflow.python.framework import function
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/function.py", line 34, in <module>
    from tensorflow.python.ops import variable_scope as vs
  File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/variable_scope.py", line 191, in <module>
    """
AttributeError: 'int' object attribute '__doc__' is read-only

i previously saw an import error ImportError: No module named enum but was able to move past it with a pip install

was able to comment that __doc__ line out and everything is running fine now. so slowww tho, but that's another matter entirely