TensorFlow

open-source repositories online

Google Brain Team open-sourced TensorFlow to accelerate machine learning research

For full details and full GPU install goto:

NOTE: Currently there is not full support for Python 3.6 or higher, it is being worked on and was claimed to work in TensorFlow 1.2 but even in TensorFlow 1.3 still there is not a working official release (:

With TensorFlow 1.0 and Python 3.5 in windows Installing the CPU version of TensorFlow is very straight forward in Windows, make sure you have installed Python version 3.5 and are connected to the internet.

open a command box and type in

pip3 install --upgrade tensorflow

To test tensorflow is installed correctly open python and type in the following:

>>> import tensorflow as tf

>>> hello = tf.constant('Hello, TensorFlow!')

>>> sess = tf.Session()

>>> print(sess.run(hello))

This should display

Hello, TensorFlow!

My experience of installing and using TensorFlow has not been good as it is still very much in a state of development, changing the way it works, there are often many hoops you will have to work through to even getting it working with existing examples and tutorials also issues with running with the latest versions of Python, .

Any problems please refer to the TensorFlow website

Using GPU with TensorFlow

Installing CPU and GPU TensorFlow on Windows

Some changes in TesorFlow since 2016

classifier

TensorFlow has updated the Clissifiers, example to work remove "TensorFlow" example skflow.TensorFlowLinearClassifier now skflow.LinearClassifier

Example Python code:

from tensorflow.contrib.learn.python import SKCompat

classifier = SKCompat(skflow.LinearClassifier([10, 20, 10],n_classes=3))