📌  相关文章
📜  Tensorflow.js tf.initializers.leCunNormal()函数

📅  最后修改于: 2022-05-13 01:56:28.130000             🧑  作者: Mango

Tensorflow.js tf.initializers.leCunNormal()函数

Tensorflow.js是谷歌开发的一个开源库,用于在浏览器或节点环境中运行机器学习模型和深度学习神经网络。它还可以帮助开发人员用 JavaScript 语言开发 ML 模型,并且可以直接在浏览器或 Node.js 中使用 ML。

tf.initializers.leCunNormal()函数从截断正态分布中提取样本,该正态分布以 stddev = sqrt(1 / fanIn) 为中心为零。请注意, fanIn是张量权重中的输入数量。

句法:

tf.initializers.leCunNormal(arguments).

参数:

  • arguments:它是一个包含种子(一个数字)的对象,它是随机数生成器的种子/数字。

返回值:它返回tf.initializers.Initializer。

示例 1:

Javascript
// Importing the tensorflow.Js library
import * as tf from "@tensorflow/tfjs"
 
// Initializing the .initializers.leCunNormal() function
const geek = tf.initializers.leCunNormal(3)
 
// Printing gain
 
console.log(geek);
console.log('\nIndividual values:\n');
console.log(geek.scale);
console.log(geek.mode);
console.log(geek.distribution);


Javascript
// Importing the tensorflow.Js library
import * as tf from "@tensorflow/tfjs"
 
// Defining the input value
let inputValue = tf.input({ shape: [4] });
 
// Initializing tf.initializers.leCunNormal()
// function
let funcValue = tf.initializers.leCunNormal(7)
 
// Creating dense layer 1
let dense_layer_1 = tf.layers.dense({
    units: 5,
    activation: 'relu',
    kernelInitialize: funcValue
});
 
// Creating dense layer 2
let dense_layer_2 = tf.layers.dense({
    units: 7,
    activation: 'softmax'
});
 
// Output
let outputValue = dense_layer_2.apply(
    dense_layer_1.apply(inputValue)
);
 
// Creation the model.
let model = tf.model({
    inputs: inputValue,
    outputs: outputValue
});
 
// Predicting the output
let finalOutput = model.predict(tf.ones([2, 4]));
finalOutput.print();



输出:

{
    "scale": 1,
    "mode": "fanIn",
    "distribution": "normal"
}

Individual values:

1
fanIn
normal

示例 2:

Javascript

// Importing the tensorflow.Js library
import * as tf from "@tensorflow/tfjs"
 
// Defining the input value
let inputValue = tf.input({ shape: [4] });
 
// Initializing tf.initializers.leCunNormal()
// function
let funcValue = tf.initializers.leCunNormal(7)
 
// Creating dense layer 1
let dense_layer_1 = tf.layers.dense({
    units: 5,
    activation: 'relu',
    kernelInitialize: funcValue
});
 
// Creating dense layer 2
let dense_layer_2 = tf.layers.dense({
    units: 7,
    activation: 'softmax'
});
 
// Output
let outputValue = dense_layer_2.apply(
    dense_layer_1.apply(inputValue)
);
 
// Creation the model.
let model = tf.model({
    inputs: inputValue,
    outputs: outputValue
});
 
// Predicting the output
let finalOutput = model.predict(tf.ones([2, 4]));
finalOutput.print();

输出:

Tensor
    [[0.0666204, 0.1171203, 0.2322821, 0.1056982, 
            0.2149536, 0.1846998, 0.0786256],
     [0.0666204, 0.1171203, 0.2322821, 0.1056982, 
            0.2149536, 0.1846998, 0.0786256]]

参考: https://js.tensorflow.org/api/latest/#initializers.leCunNormal