This guide provides step-by-step instructions for installing TensorFlow 1.4 on 64-bit Windows.
For more information check out our video:
This guide explains how to install TensorFlow 1.4 for CPU-only and GPU support on Debian 9. Stretch has only partial support for the new ABI. Therefore, we need the following libraries from testing:
libcudnn7=7.0.3-1+cuda9.0 libnvinfer4=4.1.2-1+cuda9.0 libnvinfer-dev=4.1.2-1+cuda9.0
To get these libraries, edit your /etc/apt/sources.list file and add the following lines:
deb http://ftp.debian.org/debian stretch main contrib non-free
deb http://security.debian.org/debian-security stretch/updates main contrib non-free
deb http://ftp.debian.org/debian testing main contrib non-free
TensorFlow 1.4 requires Python 3.3–3.6. We recommend that you use pip to install TensorFlow 1.4.
Before you install TensorFlow 1.4, we recommend that you create a new virtual environment. This will make it easy to keep your development environment separate from your other projects.
To create a new virtual environment, open a terminal and enter the following commands:
$ virtualenv – system-site-packages ~/tensorflow1.4
$ source ~/tensorflow1.4/bin/activate
(tensorflow1.4)$ # Your prompt should change
With your virtual environment active, enter the following command to install TensorFlow 1.4:
(tensorflow1.4)$ pip install – upgrade https://storage.googleapis.com/tensorflow/mac/cpu/tensorflow-1.4.0-py3-none-any.whl
There are a few ways to configure your TensorFlow 1.4 environment. If you’re just getting started, we recommend the CPU-only version of TensorFlow 1.4. For more information on different editions of TensorFlow 1.4, check out our release notes.
To get started:
1. Download and install Anaconda 3.6 for your platform (Linux, MacOS, or Windows). Note that TensorFlow 1.4 requires Python 3.5–3.6.
2a. Create a new conda environment named tensorflow by running `conda create -n tensorflow python=3.5` (MacOS/Linux) or `activate tensorflow` (Windows). This command will install Python in your new environment.
2b. Alternatively, you can use pip to install TensorFlow within an existing virtualenv environment: `pip install – upgrade https://storage.googleapis.com/tensorflow/mac/cpu/tensorflow-1.–cp35-cp35m-macosx_10_11_x86_64.whl` (MacOS) or `pip install – upgrade https://storage.googleapisC:\Users\nguyen\AppData\Local\Programs\Python\Python35\Scripts\.virtualenvs\tensorflow2b-ExKItR7s>tensorboard – logdir=”./graphs”unzip DLLs_and_libcupti_for_Anaconda3″win32″ for Windows>. Alternatively, use the fully containerized version of TensorFlow provided by Docker by following the instructions here . PyCharm remote interpreter with WSL . Install Intel MKL . nohup
CPU -only: For detailed instructions, please refer to: Ubuntu installation , macOS installation , or Windows installation . GPU : See the GPU guide for CUDA®-enabled cards
TensorFlow is an open source software library for numerical computation using data flow graphs. Nodes in the graph represent mathematical operations, while the graph edges represent the multidimensional data arrays (tensors) that flow between them. The flexible architecture allows you to deploy computation to one or more CPUs or GPUs in a desktop, server, or mobile device with a single API.
TensorFlow 1.4 has aRequirement class that allows for specifying conditional dependencies. If you have anaconda installed, you can specify a dependency on it with:
pip install ‘tensorflow==1.4.0’
If you don’t have anaconda installed, you can still use TensorFlow 1.4 by specifying a different version of the package:
pip install ‘tensorflow==1.4.0’
Tips and Tricks
Installing TensorFlow can be a challenge, especially if you’re new to Python and don’t have a lot of experience with programming. In this guide, we’ll show you some tips and tricks to help you get TensorFlow 1.4 up and running on your computer.
First, make sure you have Python 3.5 or 3.6 installed. TensorFlow 1.4 is not compatible with Python 2.7.
Next, you’ll need to install the following dependencies:
If you’re using Windows, we recommend installing the Anaconda distribution of Python, which comes with all of these dependencies pre-installed.
Once you have all the dependencies installed, you can install TensorFlow 1.4 by running the following command:
pip install tensorflow==1.4
Q: Which versions of TensorFlow support Python 3?
A: TensorFlow 1.4 or later requires Python 3.4 or later.
If you’re having trouble installing TensorFlow 1.4, here are some troubleshooting tips that might help.
-Make sure you have Python 3.5 or 3.6 installed, as TensorFlow 1.4 is not compatible with earlier versions of Python.
-If you’re using Anaconda, update to the latest version of conda before proceeding.
-If you’re using pip, make sure you have the latest version installed. You can update pip by running “pip install -U pip”.
-If you’re still having trouble, try installing TensorFlow in a virtual environment. This can help isolate any issues with your installation and allow you to use different versions of TensorFlow on the same system.
If you have feedback, please let us know either through the [TensorFlow discussion forum](https://www.tensorflow.org/community), or on GitHub by [filing a new issue](https://github.com/tensorflow/tensorflow/issues/new).
Keyword: How to Install TensorFlow 1.4