TensorFlow.js Examples

This repository contains a set of examples implemented in TensorFlow.js.

Each example directory is standalone so the directory can be copied to another project.

Overview of Examples

Example name	Demo link	Input data type	Task type	Model type	Training	Inference	API type	Save-load operations
addition-rnn	🔗	Text	Sequence-to-sequence	RNN: SimpleRNN, GRU and LSTM	Browser	Browser	Layers
baseball-node		Numeric	Multiclass classification	Multilayer perceptron	Node.js	Node.js	Layers
boston-housing	🔗	Numeric	Regression	Multilayer perceptron	Browser	Browser	Layers
cart-pole	🔗		Reinforcement learning	Policy gradient	Browser	Browser	Layers	IndexedDB
custom-layer	🔗		(Illustrates how to define and use a custom Layer subtype)			Browser	Layers
iris	🔗	Numeric	Multiclass classification	Multilayer perceptron	Browser	Browser	Layers
lstm-text-generation	🔗	Text	Sequence-to-prediction	RNN: LSTM	Browser	Browser	Layers	IndexedDB
mnist	🔗	Image	Multiclass classification	Convolutional neural network	Browser	Browser	Layers
mnist-acgan	🔗	Image	Generative Adversarial Network (GAN)	Convolutional neural network; GAN	Node.js	Browser	Layers	Saving to filesystem from Node.js and loading it in the browser
mnist-core	🔗	Image	Multiclass classification	Convolutional neural network	Browser	Browser	Core (Ops)
mnist-node		Image	Multiclass classification	Convolutional neural network	Node.js	Node.js	Layers	Saving to filesystem
mnist-transfer-cnn	🔗	Image	Multiclass classification (transfer learning)	Convolutional neural network	Browser	Browser	Layers	Loading pretrained model
mobilenet	🔗	Image	Multiclass classification	Convolutional neural network		Browser	Layers	Loading pretrained model
polynomial-regression	🔗	Numeric	Regression	Shallow neural network	Browser	Browser	Layers
polynomial-regression-core	🔗	Numeric	Regression	Shallow neural network	Browser	Browser	Core (Ops)
sentiment	🔗	Text	Sequence-to-regression	LSTM, 1D convnet		Browser	Layers	Loading model converted from Keras
simple-object-detection	🔗	Image	Object detection	Convolutional neural network (transfer learning)	Node.js	Browser	Layers	Save a trained model from tfjs-node and load it in the browser
translation	🔗	Text	Sequence-to-sequence	LSTM encoder and decoder		Browser	Layers	Loading model converted from Keras
tsne-mnist-canvas			Dimension reduction and data visualization	tSNE	Browser	Browser	Core (Ops)
webcam-transfer-learning	🔗	Image	Multiclass classification (transfer learning)	Convolutional neural network	Browser	Browser	Layers	Loading pretrained model
website-phishing	🔗	Numeric	Binary classification	Multilayer perceptron	Browser	Browser	Layers

Dependencies

Except for getting_started, all the examples require the following dependencies to be installed.

Node.js version 8.9 or higher
NPM cli OR Yarn

How to build an example

cd into the directory

If you are using yarn:

cd mnist-core
yarn
yarn watch

If you are using npm:

cd mnist-core
npm install
npm run watch

Details

The convention is that each example contains two scripts:

yarn watch or npm run watch: starts a local development HTTP server which watches the filesystem for changes so you can edit the code (JS or HTML) and see changes when you refresh the page immediately.
yarn build or npm run build: generates a dist/ folder which contains the build artifacts and can be used for deployment.

Contributing