Tensorflow.js tf.registerBackend() Function

Tensorflow.js is an open-source library developed by Google for running machine learning models and deep learning neural networks in the browser or node environment.

Tensorflow.js tf.registerBackend() function is used to register a global backend at the time of importing module files. The registered backend is used for modular builds.

Syntax:

tf.registerBackend (name, factory, priority?)

Parameters:

• name (string): It specifies name of the backend to be registered.
• factory: It specifies the backend factory function. This function returns a backend instance or a promise of an instance.
• priority (number): It specifies the priority of this backend. More importance is given to a backend with a higher priority. Default value of this parameter is 1.

Return value: It returns a boolean value.

Example 1:

Javascript

// Importing the tensorflow.js library
const tf = require("@tensorflow/tfjs");

// Try to register a 'webgl' backend
const status = tf.registerBackend("webgl")

// Print status
console.log(status)

Output:

true

Example 2:

Javascript

// Importing the tensorflow.js library
const tf = require("@tensorflow/tfjs");

// Try to register a 'cpu' backend
const status = tf.registerBackend("cpu")

// Print status
console.log(status)

Output:

false

Reference: https://js.tensorflow.org/api/latest/#registerBackend