Tensorflow.js tf.data.Dataset class .prefetch() Method
Last Updated :
22 Apr, 2022
Improve
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.
The tf.data.Dataset class .prefetch() function is used to produce a dataset that prefetches the specified elements from this given dataset.
Syntax:
prefetch (bufferSize)
Parameters: This function accepts a parameter which is illustrated below:
- bufferSize: It is an integer value that specifies the number of elements to be prefetched.
Return Value: It returns a dataset of elements.
Example 1:
// Importing the tensorflow.js library
import * as tf from "@tensorflow/tfjs"
// Calling the .prefetch() function over
// the specified dataset of some elements
const a = tf.data.array([5, 10, 15, 20]).prefetch(4);
// Getting the dataset of prefetched elements
await a.forEachAsync(a => console.log(a));
Output:
5 10 15 20
Example 2:
// Importing the tensorflow.js library
import * as tf from "@tensorflow/tfjs"
// Specifying a dataset of some elements
const a = tf.data.array(["a", "b", "c", "d", "e"]);
// Calling the .prefetch() function over
// the above dataset along with the
// batch of size 2
const b = a.batch(2)
const c = b.prefetch(2)
// Getting the dataset of prefetched elements
await c.forEachAsync(c => console.log(c));
Output:
Tensor
['a', 'b']
Tensor
['c', 'd']
Tensor
['e']
Reference: https://js.tensorflow.org/api/latest/#tf.data.Dataset.prefetch
