Sample code for "Tensorflow and deep learning, without a PhD" presentation and code lab.
Python
Latest commit 1669551 Jan 12, 2017 Martin Gorner better image URL
Permalink
Failed to load latest commit information.
cloudml cosmetic fixes after initial code review Jan 5, 2017
.gitignore small bug fix in batch norm Oct 27, 2016
CONTRIBUTING.md initial commit 2 Jan 5, 2017
INSTALL.txt Python 2 allowance in INSTALL.txt Jan 5, 2017
LICENSE Initial commit May 9, 2016
README.md better image URL Jan 12, 2017
mnist_1.0_softmax.py compatibility with Tensorflow 0.11 Oct 13, 2016
mnist_2.0_five_layers_sigmoid.py changed kwarg targets into labels in softmax_cross_entropy_with_logit… Dec 28, 2016
mnist_2.1_five_layers_relu_lrdecay.py changed kwarg targets into labels in softmax_cross_entropy_with_logit… Dec 28, 2016
mnist_2.2_five_layers_relu_lrdecay_dropout.py changed kwarg targets into labels in softmax_cross_entropy_with_logit… Dec 28, 2016
mnist_3.0_convolutional.py changed kwarg targets into labels in softmax_cross_entropy_with_logit… Dec 28, 2016
mnist_3.1_convolutional_bigger_dropout.py changed kwarg targets into labels in softmax_cross_entropy_with_logit… Dec 28, 2016
mnist_4.0_batchnorm_five_layers_sigmoid.py changed kwarg targets into labels in softmax_cross_entropy_with_logit… Dec 28, 2016
mnist_4.1_batchnorm_five_layers_relu.py changed kwarg targets into labels in softmax_cross_entropy_with_logit… Dec 28, 2016
mnist_4.2_batchnorm_convolutional.py changed kwarg targets into labels in softmax_cross_entropy_with_logit… Dec 28, 2016
tensorflowvisu.mplstyle added dpi setting to Tensorflow visu initialisation. Removed alternat… May 10, 2016
tensorflowvisu.py added dpi setting to Tensorflow visu initialisation. Removed alternat… May 10, 2016
tensorflowvisu_digits.py cosmetic fixes after initial code review Jan 5, 2017

README.md

Image

This is support code for the codelab "Tensorflow and deep learning - without a PhD"

The presentation explaining the underlying concepts is here and you will find codelab instructions to follow on its last slide. Do not forget to open the speaker notes in the presentation, a lot of the explanations are there.

The lab takes 2.5 hours and takes you through the design and optimisation of a neural network for recognising handwritten digits, from the simplest possible solution all the way to a recognition accuracy above 99%. It covers dense and convolutional networks, as well as techniques such as learning rate decay and dropout.

Installation instructions here. The short version is: install Python3, then pip3 install tensorflow and matplotlib.


Disclaimer: This is not an official Google product but sample code provided for an educational purpose