TensorFlow.js

TensorFlow.js is an open-source hardware-accelerated JavaScript library for training and deploying machine learning models.

Develop ML in the Browser
Use flexible and intuitive APIs to build models from scratch using the low-level JavaScript linear algebra library or the high-level layers API.

Run Existing models
Use TensorFlow.js model converters to run pre-existing TensorFlow models right in the browser.

Retrain Existing models
Retrain pre-existing ML models using sensor data connected to the browser, or other client-side data.

Importing

You can import TensorFlow.js directly via yarn or npm: yarn add @tensorflow/tfjs or npm install @tensorflow/tfjs.

Alternatively you can use a script tag. The library will be available as a global variable named tf:

<script src="https://cdn.jsdelivr.net/npm/@tensorflow/tfjs@latest"></script>
<!-- or -->
<script src="https://unpkg.com/@tensorflow/tfjs@latest"></script>

You can also specify which version to load replacing @latest with a specific version string (e.g. 0.6.0).

About this repo

This repository contains the logic and scripts that combine two packages:

TensorFlow.js Core, a flexible low-level API, formerly known as deeplearn.js.
TensorFlow.js Layers, a high-level API which implements functionality similar to Keras.

If you care about bundle size, you can import those packages individually.

Examples

Check out our examples repository and our tutorials.

Migrating from deeplearn.js

See these release notes for how to migrate from deeplearn.js to TensorFlow.js.

Getting started

Let's add a scalar value to a vector. TensorFlow.js supports broadcasting the value of scalar over all the elements in the tensor.

import * as tf from '@tensorflow/tfjs'; // If not loading the script as a global

const a = tf.tensor1d([1, 2, 3]);
const b = tf.scalar(2);

const result = a.add(b); // a is not modified, result is a new tensor
result.data().then(data => console.log(data)); // Float32Array([3, 4, 5]

// Alternatively you can use a blocking call to get the data.
// However this might slow your program down if called repeatedly.
console.log(result.dataSync()); // Float32Array([3, 4, 5]

See the core-concepts tutorial for more.

Now, let's build a toy model to perform linear regression.

import * as tf from '@tensorflow/tfjs';

// A sequential model is a container which you can add layers to.
const model = tf.sequential();

// Add a dense layer with 1 output unit.
model.add(tf.layers.dense({units: 1, inputShape: [1]}));

// Specify the loss type and optimizer for training.
model.compile({loss: 'meanSquaredError', optimizer: 'sgd'});

// Generate some synthetic data for training.
const xs = tf.tensor2d([[1], [2], [3], [4]], [4, 1]);
const ys = tf.tensor2d([[1], [3], [5], [7]], [4, 1]);

// Train the model.
await model.fit(xs, ys, {epochs: 500});

// After the training, perform inference.
const output = model.predict(tf.tensor2d([[5]], [1, 1]));
output.print();

For a deeper dive into building models, see the MNIST tutorial

Importing pre-trained models

We support porting pre-trained models from:

Find out more

TensorFlow.js is a part of the TensorFlow ecosystem. For more info:

mei3/tfjs