Build TF from source on Windows

Install a Python 3.5.x or Python 3.6.x 64-bit release for Windows at here. Select pip as an optional feature and add it to your %PATH% environmental variable.

1.3. Install python dependencies via pip

pip3 install six numpy wheel
pip3 install keras_applications==1.0.6 --no-deps
pip3 install keras_preprocessing==1.0.5 --no-deps

Or even install Visual Studio (optional) VS2015 Update 3 or newer is required. For students, VS Community 2015 is appropriate version, which can be downloaded at here. You should restart your computer to complete this step.

Note that, verify the existence of folder C:\Program Files (x86)\Microsoft Visual Studio 14.0\VC

1.4. Install MSYS2

Install MSYS2 for the bin tools needed to build TensorFlow. If MSYS2 is installed to C:\tools\msys64, add C:\tools\msys64\usr\bin to your %PATH% environment variable.

pacman -S git patch unzip

1.5. Download and install Bazel

Copy the file bazel-0.20.0-windows-x86_64.exe to folder C:\tools\bazel Rename the file from bazel-0.20.0-windows-x86_64.exe to bazel.exe

1.6. Install GPU support (Optional)

See the Windows GPU support guide to install the drivers and additional software required to run TensorFlow on a GPU.

2. Build pip-package of TensorFlow from source

2.1. Download source code of TensorFlow

git clone https://github.com/tensorflow/tensorflow.git
cd tensorflow
git checkout branch_name  # r1.9, r1.10, r1.11, r1.12, etc.

2.2. Configure building process

python ./configure.py

python ./configure.py
Starting local Bazel server and connecting to it...
................
You have bazel 0.15.0 installed.
Please specify the location of python. [Default is C:\python36\python.exe]: 

Found possible Python library paths:
  C:\python36\lib\site-packages
Please input the desired Python library path to use.  Default is [C:\python36\lib\site-packages]

Do you wish to build TensorFlow with CUDA support? [y/N]: Y
CUDA support will be enabled for TensorFlow.

Please specify the CUDA SDK version you want to use. [Leave empty to default to CUDA 9.0]:

Please specify the location where CUDA 9.0 toolkit is installed. Refer to README.md for more details. [Default is C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v9.0]:

Please specify the cuDNN version you want to use. [Leave empty to default to cuDNN 7.0]: 7.0

Please specify the location where cuDNN 7 library is installed. Refer to README.md for more details. [Default is C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v9.0]: C:\tools\cuda

Please specify a list of comma-separated Cuda compute capabilities you want to build with.
You can find the compute capability of your device at: https://developer.nvidia.com/cuda-gpus.
Please note that each additional compute capability significantly increases your build time and binary size. [Default is: 3.5,7.0]: 3.7

Please specify optimization flags to use during compilation when bazel option "--config=opt" is specified [Default is /arch:AVX]: 

Would you like to override eigen strong inline for some C++ compilation to reduce the compilation time? [Y/n]:
Eigen strong inline overridden.

Configuration finished

2.3. Build pip-package

If your computer has a limit memory, please insert the following argument to the compile command

--local_resources 2048,.5,1.0

bazel build --define=no_tensorflow_py_deps=true --incompatible_remove_native_http_archive=false --cpu=x64_windows --compiler=msvc-cl --copt=-nvcc_options=disable-warnings --config=opt --define=no_tensorflow_py_deps=true //tensorflow/tools/pip_package:build_pip_package

bazel build --define=no_tensorflow_py_deps=true --incompatible_remove_native_http_archive=false --cpu=x64_windows --compiler=msvc-cl --copt=-nvcc_options=disable-warnings --config=opt --config=cuda --define=no_tensorflow_py_deps=true //tensorflow/tools/pip_package:build_pip_package

bazel-bin\tensorflow\tools\pip_package\build_pip_package C:/tmp/tensorflow_pkg

3. Install from pip-package

3.1. Directly install to python on system

pip install C:/tmp/tensorflow_pkg/tensorflow-version-cp36-cp36m-win_amd64.whl

3.2. Install to a virtual environment of python

In case you do NOT want to install and overwrite a new compiled TF to your python on system. Or you want to verify that new compiled TF.

mkdir my_project_folder
virtualenv --system-site-packages <full path to my_project_folder>

<full path to my_project_folder>\Scripts\activate.bat

pip install six numpy wheel
pip install keras_applications==1.0.5 --no-deps
pip install keras_preprocessing==1.0.3 --no-deps

4. Verify installed TF

import tensorflow as tf
hello = tf.constant('Hello, TensorFlow!')
session = tf.Session()
print(session.run(hello))

Then, run the script file by python. If you installed TF in virtual environment, please active virtual environment and run script file from that.

5. Errors during build from source:

**System information**
    OS Platform and Distribution (e.g., Linux Ubuntu 16.04): Arch Linux 4.15 x86_64
    Mobile device (e.g. iPhone 8, Pixel 2, Samsung Galaxy) if the issue happens on mobile device:
    TensorFlow installed from (source or binary): source, building ci_build.sh from TF distro
    TensorFlow version: latest from master
    Python version: tried both 2.7 and 3
    Installed using virtualenv? pip? conda?:
    Bazel version (if compiling from source): tried both 0.20.0 and 0.19.2
    GCC/Compiler version (if compiling from source):
    CUDA/cuDNN version:
    GPU model and memory:

**Describe the problem**

**Provide the exact sequence of commands / steps that you executed before running into the problem**

**Any other info / logs**

0. Environment

1. Prerequisite

1.1. OS environment

1.2. Install Python 3