speech-command Unable to load the model using ensureModelLoaded

System information

• Have I written custom code (as opposed to using a stock example script provided in TensorFlow.js): No
• OS Platform and Distribution (e.g., Linux Ubuntu 16.04): Windows10
• TensorFlow.js installed from (npm or script link): npm
• TensorFlow.js version (use command below): 3.0.0
• Browser version: Version 88.0.4324.150 (Official Build) (x86_64)

Describe the current behavior Error when trying to load the model using ensureModelLoaded method: Uncaught (in promise) Error: No backend found in registry.

Describe the expected behavior Expected to load the model.

Standalone code to reproduce the issue Example in Codesandbox

Other info / logs It was working fine in speech-command 0.4.2, but faced this issue after using the latest version 0.5.1

Full error:

Uncaught (in promise) Error: No backend found in registry.
    at ut.getSortedBackends (speech-commands.esm.js:17)
    at ut.initializeBackendsAndReturnBest (speech-commands.esm.js:17)
    at ut.get backend [as backend] (speech-commands.esm.js:17)
    at ut.makeTensor (speech-commands.esm.js:17)
    at Ss (speech-commands.esm.js:17)
    at Po.apply (speech-commands.esm.js:17)
    at Sh.addWeight (speech-commands.esm.js:17)
    at Sh.build (speech-commands.esm.js:17)
    at speech-commands.esm.js:17
    at ho (speech-commands.esm.js:17)
    at Sh.apply (speech-commands.esm.js:17)
    at $u.add (speech-commands.esm.js:17)
    at fromConfig (speech-commands.esm.js:17)
    at _a (speech-commands.esm.js:17)
    at Ml (speech-commands.esm.js:17)
    at speech-commands.esm.js:17

[rthadur]

Here is the link which explains why we need add https://github.com/tensorflow/tfjs/releases/tag/tfjs-v2.0.0

thank you

[pyu10055]

Look like it might have been treeshaked, here what I did:

import React from "react";
import * as tf from "@tensorflow/tfjs";
import * as speechCommands from "@tensorflow-models/speech-commands";
import { useState } from "react";

console.log(tf.engine().backend);