Deep learning in the browser

Official repository of the book Deep learning in the browser released August 2018 and published by Bleeding Edge Press. Here you will find all of the source code of the demos in the book.

Content of this repository:

• chapter 1: Introduction to deep learning - activation functions
• chapter 3: Deep learning frameworks for JavaScript - Tensorflow.js
• chapter 5: GPU acceleration with WebGL
• chapter 6: Extracting data from the browser
• chapter 7: Recipes for advanced data manipulation
• chapter 8: Building applications with TensorFlow.js

Getting Started

Clone the repo and all submodules.

$ git clone git@github.com:backstopmedia/deep-learning-browser.git
$ cd deep-learning-browser
$ git submodule update --init --recursive

You can serve the code of the chapters using a simple static webserver.

$ npm install http-server -g
$ http-server chapter6 --cors -p 8081

Navigate to localhost:8081 to run the code from the selected chapter.

License

The whole content of this repository is released under MIT License (see /LICENSE.txt), except:

• Third party libraries (e.g. in /lib subdirectory)
• Third party data (e.g. in /data subdirectory)