2018-08-09tensorlow installation

other toolkits
caffe, DeepLearning4j, Microsoft Cognitve Toolkit, MXNet, PaddlePaddle(baidu), etc.

dependency

• Protocal Buffer: processing structured data, XML and json are tools for that. sequenced data is bytes but not string,schema is required to be defined before
• Bazel: automatically building tool (compared to Makefile、ant、Maven), WORKSPACE file is needed in the tensorflow project's root dir.,which define the dependency to the external resources.

installation

docker(virtual technique)+tensorflow

pip:

1.pip installation

#linux
sudo apt-get install python-pip python-dev
# Mac OS
sudo easy_install pip
sudo easy_install --upgrade six

2. pip install
   cuda+tensorflow-gpu
   tensorflow

install via source code

1.bazel
jdk8

sudo apt-get install software-properities-common
sudo add-apt-repository ppa:webupd8team/java
sudo apt-get update
sudo apt-get install oracle-java8-installer

other dependency for bazel

sudo apt-get install pkg-config zip g++ zliblg-dev unzip

code installation

chmod +x <your installer.sh>
<your installer.sh> -user
export PATH="$PATH:$HOME/bin"

2.other dependency

sudo apt-get install python3-numpy swig python3-dev python3-wheel
#for GPU additional cuda is needed
tar -xvzf  <cuda_name.tgz>
sudo cp <cuda_name/cudnn.h> /usr/local/cuda/include
sudo cp <cuda_name/libcudnn*> /usr/local/cuda/lib64
sudo chmod a+r /usr/local/cuda/include/cudnn.h \
/usr/local/cuda/lib64/libcudnn*

3.tensorflow

git clone https://github.om/tensorflow/tensorflow
cd tensorflow
./configure
#waiting it to be completed
bazel build -c opt --config=cuda /tensorflow/tools/pip-package:build_ pip-package
bazel-bin/tensorflow/tools/pip_package:build_pip_package /tmp/tensorflow_ pkg
pip install /tmp/tensorflow_pkg/<tensorflow_name>.whl