课后练习之第二课——Week7

笨笨sg

已于 2024-11-14 20:22:11 修改

阅读量734

点赞数 23

分类专栏：吴恩达——机器学习+深度学习文章标签：课后练习第二课——Week6

于 2024-11-11 11:32:53 首次发布

本文链接：https://blog.csdn.net/a131529/article/details/143676982

版权

吴恩达——机器学习+深度学习专栏收录该内容

24 篇文章

订阅专栏

2.8 在TensorFlow中构建你的第一个神经网络

实现正向传播

计算代价函数

训练模型

1 前言

本文为2021吴恩达学习笔记deeplearning.ai《深度学习专项课程》篇——“第二课——Week7”章节的课后练习，完整内容参见：

深度学习入门指南——2021吴恩达学习笔记deeplearning.ai《深度学习专项课程》篇-CSDN博客

2 Tensorflow 简单使用

欢迎来到本周的编程作业！到目前为止，你一直使用Numpy来构建神经网络，但本周你将探索一个深度学习框架，它可以让你更轻松地构建神经网络。机器学习框架，如TensorFlow、PaddlePaddle、Torch、Caffe、Keras等，可以显著加快机器学习的发展.

2.1 导包

import h5py
import numpy as np
import tensorflow as tf
import matplotlib.pyplot as plt
from tensorflow.python.framework.ops import EagerTensor
from tensorflow.python.ops.resource_variable_ops import ResourceVariable
import time

improv_utils.py：

import h5py
import numpy as np
import tensorflow as tf
import math

def load_dataset():
    train_dataset = h5py.File('datasets/train_signs.h5', "r")
    train_set_x_orig = np.array(train_dataset["train_set_x"][:]) # your train set features
    train_set_y_orig = np.array(train_dataset["train_set_y"][:]) # your train set labels

    test_dataset = h5py.File('datasets/test_signs.h5', "r")
    test_set_x_orig = np.array(test_dataset["test_set_x"][:]) # your test set features
    test_set_y_orig = np.array(test_dataset["test_set_y"][:]) # your test set labels

    classes = np.array(test_dataset["list_classes"][:]) # the list of classes
    
    train_set_y_orig = train_set_y_orig.reshape((1, train_set_y_orig.shape[0]))
    test_set_y_orig = test_set_y_orig.reshape((1, test_set_y_orig.shape[0]))
    
    return train_set_x_orig, train_set_y_orig, test_set_x_orig, test_set_y_orig, classes


def random_mini_batches(X, Y, mini_batch_size = 64, seed = 0):
    """
    Creates a list of random minibatches from (X, Y)
    
    Arguments:
    X -- input data, of shape (input size, number of examples)
    Y -- true "label" vector (containing 0 if cat, 1 if non-cat), of shape (1, number of examples)
    mini_batch_size - size of the mini-batches, integer
    seed -- this is only for the purpose of grading, so that you're "random minibatches are the same as ours.
    
    Returns:
    mini_batches -- list of synchronous (mini_batch_X, mini_batch_Y)
    """
    
    m = X.shape[1]                  # number of training examples
    mini_batches = []
    np.random.seed(seed)
    
    # Step 1: Shuffle (X, Y)
    permutation = list(np.random.permutation(m))
    shuffled_X = X[:, permutation]
    shuffled_Y = Y[:, permutation].reshape((Y.shape[0],m))

    # Step 2: Partition (shuffled_X, shuffled_Y). Minus the end case.
    num_complete_minibatches = math.floor(m/mini_batch_size) # number of mini batches of size mini_batch_size in your partitionning
    for k in range(0, num_complete_minibatches):
        mini_batch_X = shuffled_X[:, k * mini_batch_size : k * mini_batch_size + mini_batch_size]
        mini_batch_Y = shuffled_Y[:, k * mini_batch_size : k * mini_batch_size + mini_batch_size]
        mini_batch = (mini_batch_X, mini_batch_Y)
        mini_batches.append(mini_batch)
    
    # Handling the end case (last mini-batch < mini_batch_size)
    if m % mini_batch_size != 0:
        mini_batch_X = shuffled_X[:, num_complete_minibatches * mini_batch_size : m]
        mini_batch_Y = shuffled_Y[:, num_complete_minibatches * mini_batch_size : m]
        mini_batch = (mini_batch_X, mini_batch_Y)
        mini_batches.append(mini_batch)
    
    return mini_batches

def convert_to_one_hot(Y, C):
    Y = np.eye(C)[Y.reshape(-1)].T
    return Y

def predict(X, parameters):
    
    W1 = tf.convert_to_tensor(parameters["W1"])
    b1 = tf.convert_to_tensor(parameters["b1"])
    W2 = tf.convert_to_tensor(parameters["W2"])
    b2 = tf.convert_to_tensor(parameters["b2"])
    W3 = tf.convert_to_tensor(parameters["W3"])
    b3 = tf.convert_to_tensor(parameters["b3"])
    
    params = {"W1": W1,
              "b1": b1,
              "W2": W2,
              "b2": b2,
              "W3": W3,
              "b3": b3}
    
    x = tf.placeholder("float", [12288, 1])
    
    z3 = forward_propagation(x, params)
    p = tf.argmax(z3)
    
    with tf.Session() as sess:
        prediction = sess.run(p, feed_dict = {x: X})
        
    return prediction
    

def create_placeholders(n_x, n_y):
    """
    Creates the placeholders for the tensorflow session.
    
    Arguments:
    n_x -- scalar, size of an image vector (num_px * num_px = 64 * 64 * 3 = 12288)
    n_y -- scalar, number of classes (from 0 to 5, so -> 6)
    
    Returns:
    X -- placeholder for the data input, of shape [n_x, None] and dtype "float"
    Y -- placeholder for the input labels, of shape [n_y, None] and dtype "float"
    
    Tips:
    - You will use None because it let's us be flexible on the number of examples you will for the placeholders.
      In fact, the number of examples during test/train is different.
    """

    ### START CODE HERE ### (approx. 2 lines)
    X = tf.placeholder("float", [n_x, None])
    Y = tf.placeholder("float", [n_y, None])
    ### END CODE HERE ###
    
    return X, Y


def initialize_parameters():
    """
    Initializes parameters to build a neural network with tensorflow. The shapes are:
                        W1 : [25, 12288]
                        b1 : [25, 1]
                        W2 : [12, 25]
                        b2 : [12, 1]
                        W3 : [6, 12]
                        b3 : [6, 1]
    
    Returns:
    parameters -- a dictionary of tensors containing W1, b1, W2, b2, W3, b3
    """
    
    tf.set_random_seed(1)                              # so that your "random" numbers match ours
        
    ### START CODE HERE ### (approx. 6 lines of code)
    W1 = tf.get_variable("W1", [25,12288], initializer = tf.contrib.layers.xavier_initializer(seed = 1))
    b1 = tf.get_variable("b1", [25,1], initializer = tf.zeros_initializer())
    W2 = tf.get_variable("W2", [12,25], initializer = tf.contrib.layers.xavier_initializer(seed = 1))
    b2 = tf.get_variable("b2", [12,1], initializer = tf.zeros_initializer())
    W3 = tf.get_variable("W3", [6,12], initializer = tf.contrib.layers.xavier_initializer(seed = 1))
    b3 = tf.get_variable("b3", [6,1], initializer = tf.zeros_initializer())
    ### END CODE HERE ###

    parameters = {"W1": W1,
                  "b1": b1,
                  "W2": W2,
                  "b2": b2,
                  "W3": W3,
                  "b3": b3}
    
    return parameters


def compute_cost(z3, Y):
    """
    Computes the cost
    
    Arguments:
    z3 -- output of forward propagation (output of the last LINEAR unit), of shape (10, number of examples)
    Y -- "true" labels vector placeholder, same shape as z3
    
    Returns:
    cost - Tensor of the cost function
    """
    
    # to fit the tensorflow requirement for tf.nn.softmax_cross_entropy_with_logits()
    logits = tf.transpose(z3)
    labels = tf.transpose(Y)
    
    ### START CODE HERE ### (1 line of code)
    cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits = logits, labels = labels))
    ### END CODE HERE ###
    
    return cost







def model(X_train, Y_train, X_test, Y_test, learning_rate = 0.0001,
          num_epochs = 1500, minibatch_size = 32, print_cost = True):
    """
    Implements a three-layer tensorflow neural network: LINEAR->RELU->LINEAR->RELU->LINEAR->SOFTMAX.
    
    Arguments:
    X_train -- training set, of shape (input size = 12288, number of training examples = 1080)
    Y_train -- test set, of shape (output size = 6, number of training examples = 1080)
    X_test -- training set, of shape (input size = 12288, number of training examples = 120)
    Y_test -- test set, of shape (output size = 6, number of test examples = 120)
    learning_rate -- learning rate of the optimization
    num_epochs -- number of epochs of the optimization loop
    minibatch_size -- size of a minibatch
    print_cost -- True to print the cost every 100 epochs
    
    Returns:
    parameters -- parameters learnt by the model. They can then be used to predict.
    """
    
    ops.reset_default_graph()                         # to be able to rerun the model without overwriting tf variables
    tf.set_random_seed(1)                             # to keep consistent results
    seed = 3                                          # to keep consistent results
    (n_x, m) = X_train.shape                          # (n_x: input size, m : number of examples in the train set)
    n_y = Y_train.shape[0]                                  # n_y : output size
    costs = []                                        # To keep track of the cost
    
    # Create Placeholders of shape (n_x, n_y)
    ### START CODE HERE ### (1 line)
    X, Y = create_placeholders(n_x, n_y)
    ### END CODE HERE ###

    # Initialize parameters
    ### START CODE HERE ### (1 line)
    parameters = initialize_parameters()
    ### END CODE HERE ###
    
    # Forward propagation: Build the forward propagation in the tensorflow graph
    ### START CODE HERE ### (1 line)
    z3 = forward_propagation(X, parameters)
    ### END CODE HERE ###
    
    # Cost function: Add cost function to tensorflow graph
    ### START CODE HERE ### (1 line)
    cost = compute_cost(z3, Y)
    ### END CODE HERE ###
    
    # Backpropagation: Define the tensorflow optimizer. Use an AdamOptimizer.
    ### START CODE HERE ### (1 line)
    optimizer = tf.train.AdamOptimizer(learning_rate = learning_rate).minimize(cost)
    ### END CODE HERE ###
    
    # Initialize all the variables
    init = tf.global_variables_initializer()

    # Start the session to compute the tensorflow graph
    with tf.Session() as sess:
        
        # Run the initialization
        sess.run(init)
        
        # Do the training loop
        for epoch in range(num_epochs):

            minibatch_cost = 0.
            num_minibatches = int(m / minibatch_size) # number of minibatches of size minibatch_size in the train set
            seed = seed + 1
            minibatches = random_mini_batches(X_train, Y_train, minibatch_size, seed)

            for minibatch in minibatches:

                # Select a minibatch
                (minibatch_X, minibatch_Y) = minibatch
                
                # IMPORTANT: The line that runs the graph on a minibatch.
                # Run the session to execute the optimizer and the cost, the feedict should contain a minibatch for (X,Y).
                ### START CODE HERE ### (1 line)
                _ , temp_cost = sess.run([optimizer, cost], feed_dict={X: minibatch_X, Y: minibatch_Y})
                ### END CODE HERE ###
                
                minibatch_cost += temp_cost / num_minibatches

            # Print the cost every epoch
            if print_cost == True and epoch % 100 == 0:
                print ("Cost after epoch %i: %f" % (epoch, minibatch_cost))
            if print_cost == True and epoch % 5 == 0:
                costs.append(minibatch_cost)
                
        # plot the cost
        plt.plot(np.squeeze(costs))
        plt.ylabel('cost')
        plt.xlabel('iterations (per tens)')
        plt.title("Learning rate =" + str(learning_rate))
        plt.show()

        # lets save the parameters in a variable
        parameters = sess.run(parameters)
        print ("Parameters have been trained!")

        # Calculate the correct predictions
        correct_prediction = tf.equal(tf.argmax(z3), tf.argmax(Y))

        # Calculate accuracy on the test set
        accuracy = tf.reduce_mean(tf.cast(correct_prediction, "float"))

        print ("Train Accuracy:", accuracy.eval({X: X_train, Y: Y_train}))
        print ("Test Accuracy:", accuracy.eval({X: X_test, Y: Y_test}))
        
        return parameters

test_utils.py：

import numpy as np


def datatype_check(expected_output, target_output, error):
    success = 0
    if isinstance(target_output, dict):
        for key in target_output.keys():
            try:
                success += datatype_check(expected_output[key],
                                          target_output[key], error)
            except:
                print("Error: {} in variable {}. Got {} but expected type {}".format(error,
                                                                                     key, type(target_output[key]), type(expected_output[key])))
        if success == len(target_output.keys()):
            return 1
        else:
            return 0
    elif isinstance(target_output, tuple) or isinstance(target_output, list):
        for i in range(len(target_output)):
            try:
                success += datatype_check(expected_output[i],
                                          target_output[i], error)
            except:
                print("Error: {} in variable {}, expected type: {}  but expected type {}".format(error,
                                                                                                 i, type(target_output[i]), type(expected_output[i])))
        if success == len(target_output):
            return 1
        else:
            return 0

    else:
        assert isinstance(target_output, type(expected_output))
        return 1


def equation_output_check(expected_output, target_output, error):
    success = 0
    if isinstance(target_output, dict):
        for key in target_output.keys():
            try:
                success += equation_output_check(expected_output[key],
                                                 target_output[key], error)
            except:
                print("Error: {} for variable {}.".format(error,
                                                          key))
        if success == len(target_output.keys()):
            return 1
        else:
            return 0
    elif isinstance(target_output, tuple) or isinstance(target_output, list):
        for i in range(len(target_output)):
            try:
                success += equation_output_check(expected_output[i],
                                                 target_output[i], error)
            except:
                print("Error: {} for variable in position {}.".format(error, i))
        if success == len(target_output):
            return 1
        else:
            return 0

    else:
        if hasattr(target_output, 'shape'):
            np.testing.assert_array_almost_equal(
                target_output, expected_output)
        else:
            assert target_output == expected_output
        return 1


def shape_check(expected_output, target_output, error):
    success = 0
    if isinstance(target_output, dict):
        for key in target_output.keys():
            try:
                success += shape_check(expected_output[key],
                                       target_output[key], error)
            except:
                print("Error: {} for variable {}.".format(error, key))
        if success == len(target_output.keys()):
            return 1
        else:
            return 0
    elif isinstance(target_output, tuple) or isinstance(target_output, list):
        for i in range(len(target_output)):
            try:
                success += shape_check(expected_output[i],
                                       target_output[i], error)
            except:
                print("Error: {} for variable {}.".format(error, i))
        if success == len(target_output):
            return 1
        else:
            return 0

    else:
        if hasattr(target_output, 'shape'):
            assert target_output.shape == expected_output.shape
        return 1


def single_test(test_cases, target):
    success = 0
    for test_case in test_cases:
        try:
            if test_case['name'] == "datatype_check":
                assert isinstance(target(*test_case['input']),
                                  type(test_case["expected"]))
                success += 1
            if test_case['name'] == "equation_output_check":
                assert np.allclose(test_case["expected"],
                                   target(*test_case['input']))
                success += 1
            if test_case['name'] == "shape_check":
                assert test_case['expected'].shape == target(
                    *test_case['input']).shape
                success += 1
        except:
            print("Error: " + test_case['error'])

    if success == len(test_cases):
        print("\033[92m All tests passed.")
    else:
        print('\033[92m', success, " Tests passed")
        print('\033[91m', len(test_cases) - success, " Tests failed")
        raise AssertionError(
            "Not all tests were passed for {}. Check your equations and avoid using global variables inside the function.".format(target.__name__))


def multiple_test(test_cases, target):
    success = 0
    for test_case in test_cases:
        try:
            target_answer = target(*test_case['input'])
            if test_case['name'] == "datatype_check":
                success += datatype_check(test_case['expected'],
                                          target_answer, test_case['error'])
            if test_case['name'] == "equation_output_check":
                success += equation_output_check(
                    test_case['expected'], target_answer, test_case['error'])
            if test_case['name'] == "shape_check":
                success += shape_check(test_case['expected'],
                                       target_answer, test_case['error'])
        except:
            print("Error: " + test_case['error'])

    if success == len(test_cases):
        print("\033[92m All tests passed.")
    else:
        print('\033[92m', success, " Tests passed")
        print('\033[91m', len(test_cases) - success, " Tests failed")
        raise AssertionError(
            "Not all tests were passed for {}. Check your equations and avoid using global variables inside the function.".format(target.__name__))

tf_utils.py：

import h5py
import numpy as np
import tensorflow as tf
import math

def load_dataset():
    train_dataset = h5py.File('datasets/train_signs.h5', "r")
    train_set_x_orig = np.array(train_dataset["train_set_x"][:]) # your train set features
    train_set_y_orig = np.array(train_dataset["train_set_y"][:]) # your train set labels

    test_dataset = h5py.File('datasets/test_signs.h5', "r")
    test_set_x_orig = np.array(test_dataset["test_set_x"][:]) # your test set features
    test_set_y_orig = np.array(test_dataset["test_set_y"][:]) # your test set labels

    classes = np.array(test_dataset["list_classes"][:]) # the list of classes
    
    train_set_y_orig = train_set_y_orig.reshape((1, train_set_y_orig.shape[0]))
    test_set_y_orig = test_set_y_orig.reshape((1, test_set_y_orig.shape[0]))
    
    return train_set_x_orig, train_set_y_orig, test_set_x_orig, test_set_y_orig, classes


def random_mini_batches(X, Y, mini_batch_size = 64, seed = 0):
    """
    Creates a list of random minibatches from (X, Y)
    
    Arguments:
    X -- input data, of shape (input size, number of examples)
    Y -- true "label" vector (containing 0 if cat, 1 if non-cat), of shape (1, number of examples)
    mini_batch_size - size of the mini-batches, integer
    seed -- this is only for the purpose of grading, so that you're "random minibatches are the same as ours.
    
    Returns:
    mini_batches -- list of synchronous (mini_batch_X, mini_batch_Y)
    """
    
    m = X.shape[1]                  # number of training examples
    mini_batches = []
    np.random.seed(seed)
    
    # Step 1: Shuffle (X, Y)
    permutation = list(np.random.permutation(m))
    shuffled_X = X[:, permutation]
    shuffled_Y = Y[:, permutation].reshape((Y.shape[0],m))

    # Step 2: Partition (shuffled_X, shuffled_Y). Minus the end case.
    num_complete_minibatches = math.floor(m/mini_batch_size) # number of mini batches of size mini_batch_size in your partitionning
    for k in range(0, num_complete_minibatches):
        mini_batch_X = shuffled_X[:, k * mini_batch_size : k * mini_batch_size + mini_batch_size]
        mini_batch_Y = shuffled_Y[:, k * mini_batch_size : k * mini_batch_size + mini_batch_size]
        mini_batch = (mini_batch_X, mini_batch_Y)
        mini_batches.append(mini_batch)
    
    # Handling the end case (last mini-batch < mini_batch_size)
    if m % mini_batch_size != 0:
        mini_batch_X = shuffled_X[:, num_complete_minibatches * mini_batch_size : m]
        mini_batch_Y = shuffled_Y[:, num_complete_minibatches * mini_batch_size : m]
        mini_batch = (mini_batch_X, mini_batch_Y)
        mini_batches.append(mini_batch)
    
    return mini_batches

def convert_to_one_hot(Y, C):
    Y = np.eye(C)[Y.reshape(-1)].T
    return Y


def predict(X, parameters):
    
    W1 = tf.convert_to_tensor(parameters["W1"])
    b1 = tf.convert_to_tensor(parameters["b1"])
    W2 = tf.convert_to_tensor(parameters["W2"])
    b2 = tf.convert_to_tensor(parameters["b2"])
    W3 = tf.convert_to_tensor(parameters["W3"])
    b3 = tf.convert_to_tensor(parameters["b3"])
    
    params = {"W1": W1,
              "b1": b1,
              "W2": W2,
              "b2": b2,
              "W3": W3,
              "b3": b3}
    
    x = tf.placeholder("float", [12288, 1])
    
    z3 = forward_propagation_for_predict(x, params)
    p = tf.argmax(z3)
    
    sess = tf.Session()
    prediction = sess.run(p, feed_dict = {x: X})
        
    return prediction

def forward_propagation_for_predict(X, parameters):
    """
    Implements the forward propagation for the model: LINEAR -> RELU -> LINEAR -> RELU -> LINEAR -> SOFTMAX
    
    Arguments:
    X -- input dataset placeholder, of shape (input size, number of examples)
    parameters -- python dictionary containing your parameters "W1", "b1", "W2", "b2", "W3", "b3"
                  the shapes are given in initialize_parameters

    Returns:
    Z3 -- the output of the last LINEAR unit
    """
    
    # Retrieve the parameters from the dictionary "parameters" 
    W1 = parameters['W1']
    b1 = parameters['b1']
    W2 = parameters['W2']
    b2 = parameters['b2']
    W3 = parameters['W3']
    b3 = parameters['b3'] 
                                                           # Numpy Equivalents:
    Z1 = tf.add(tf.matmul(W1, X), b1)                      # Z1 = np.dot(W1, X) + b1
    A1 = tf.nn.relu(Z1)                                    # A1 = relu(Z1)
    Z2 = tf.add(tf.matmul(W2, A1), b2)                     # Z2 = np.dot(W2, a1) + b2
    A2 = tf.nn.relu(Z2)                                    # A2 = relu(Z2)
    Z3 = tf.add(tf.matmul(W3, A2), b3)                     # Z3 = np.dot(W3,Z2) + b3
    
    return Z3

2.2 检查 TensorFlow 版本

您将在本次作业中使用v2.3+，以获得最大的速度和效率。

tf.__version__

'2.4.0'

2.3 基本优化与梯度计算图

TensorFlow 2的美妙之处在于它的简单。基本上，您所需要做的就是通过计算图实现前向传播。TensorFlow将通过“GradientTape”记录的图形向后移动，为您计算导数。剩下你要做的就是指定你想要使用的代价函数和优化器！

在编写TensorFlow程序时，要使用和转换的主要对象是“tf.Tensor”。这些张量相当于Numpy数组，即给定数据类型的多维数组，也包含有关计算图的信息。

下面，你将使用“tf”。变量’来存储变量的状态。变量只能创建一次，因为它的初始值定义了变量的形状和类型。此外，`tf `中的`dtype`参数。变量'可以设置为允许将数据转换为该类型。但是如果没有指定，如果初始值是张量，则数据类型将被保留，或者‘ convert_to_tensor ’将决定。通常最好是直接指定，这样就不会出错！

在这里，您将调用在HDF5文件上创建的TensorFlow数据集，您可以使用它来代替Numpy数组来存储数据集。你可以把它想象成一个TensorFlow数据生成器！

您将使用手势数据集，该数据集由形状为64x64x3的图像组成。

train_dataset = h5py.File('datasets/train_signs.h5', "r")
test_dataset = h5py.File('datasets/test_signs.h5', "r")
train_dataset["train_set_x"]

<HDF5 dataset "train_set_x": shape (1080, 64, 64, 3), type "|u1">

# tf.data.Dataset.from_tensor_slices( list_or_numpy_array ) creates TensorFlow Datasets
x_train = tf.data.Dataset.from_tensor_slices(train_dataset['train_set_x'])
y_train = tf.data.Dataset.from_tensor_slices(train_dataset['train_set_y'])

x_test = tf.data.Dataset.from_tensor_slices(test_dataset['test_set_x'])
y_test = tf.data.Dataset.from_tensor_slices(test_dataset['test_set_y'])
type(x_train)

tensorflow.python.data.ops.dataset_ops.TensorSliceDataset

由于TensorFlow数据集是生成器，你不能直接访问内容，除非你在for循环中迭代它们，或者通过使用‘ iter ’显式创建Python迭代器并使用‘ next ’。此外，您可以使用“element_spec”属性检查每个元素的“形状”和“dtype”。

print(x_train.element_spec)
print(next(iter(x_train)))

TensorSpec(shape=(64, 64, 3), dtype=tf.uint8, name=None)
tf.Tensor(
[[[227 220 214]
[227 221 215]
[227 222 215]
...
[232 230 224]
[231 229 222]
[230 229 221]]

[[227 221 214]
[227 221 215]
[228 221 215]
...
[232 230 224]
[231 229 222]
[231 229 221]]

[[227 221 214]
[227 221 214]
[227 221 215]
...
[232 230 224]
[231 229 223]
[230 229 221]]

...

[[119 81 51]
[124 85 55]
[127 87 58]
...
[210 211 211]
[211 212 210]
[210 211 210]]

[[119 79 51]
[124 84 55]
[126 85 56]
...
[210 211 210]
[210 211 210]
[209 210 209]]

[[119 81 51]
[123 83 55]
[122 82 54]
...
[209 210 210]
[209 210 209]
[208 209 209]]], shape=(64, 64, 3), dtype=uint8)

for element in x_train:
    print(element)
    break

tf.Tensor(
[[[227 220 214]
[227 221 215]
[227 222 215]
...
[232 230 224]
[231 229 222]
[230 229 221]]

[[227 221 214]
[227 221 215]
[228 221 215]
...
[232 230 224]
[231 229 222]
[231 229 221]]

[[227 221 214]
[227 221 214]
[227 221 215]
...
[232 230 224]
[231 229 223]
[230 229 221]]

...

[[119 81 51]
[124 85 55]
[127 87 58]
...
[210 211 211]
[211 212 210]
[210 211 210]]

[[119 79 51]
[124 84 55]
[126 85 56]
...
[210 211 210]
[210 211 210]
[209 210 209]]

[[119 81 51]
[123 83 55]
[122 82 54]
...
[209 210 210]
[209 210 209]
[208 209 209]]], shape=(64, 64, 3), dtype=uint8)

TensorFlow数据集和Numpy数组之间还有一个额外的区别：如果您需要转换一个数据集，您将调用‘ map ’方法来将作为参数传递的函数应用于每个元素。

def normalize(image):
    image = tf.cast(image, tf.float32) / 256.0
    image = tf.reshape(image, [-1,1])
    return image
new_train = x_train.map(normalize)
new_test = x_test.map(normalize)
new_train.element_spec

TensorSpec(shape=(12288, 1), dtype=tf.float32, name=None)

print(next(iter(new_train)))

tf.Tensor(
[[0.88671875]
[0.859375 ]
[0.8359375 ]
...
[0.8125 ]
[0.81640625]
[0.81640625]], shape=(12288, 1), dtype=float32)

2.4 线性函数

# GRADED FUNCTION: linear_function

def linear_function():

    np.random.seed(1)
    
    X = tf.constant(np.random.randn(3,1))
    W = tf.constant(np.random.randn(4,3))
    b = tf.constant(np.random.randn(4,1))
    Y = tf.add(tf.matmul(W,X),b)
    
    return Y

result = linear_function()
print(result)

assert type(result) == EagerTensor, "Use the TensorFlow API"
assert np.allclose(result, [[-2.15657382], [ 2.95891446], [-1.08926781], [-0.84538042]]), "Error"
print("\033[92mAll test passed")

tf.Tensor(
[[-2.15657382]
[ 2.95891446]
[-1.08926781]
[-0.84538042]], shape=(4, 1), dtype=float64)
All test passed

2.5 计算Sigmoid

# GRADED FUNCTION: sigmoid

def sigmoid(z):
    
    z = tf.cast(z,tf.float32)
    a = tf.keras.activations.sigmoid(z)
    return a

result = sigmoid(-1)
print ("type: " + str(type(result)))
print ("dtype: " + str(result.dtype))
print ("sigmoid(-1) = " + str(result))
print ("sigmoid(0) = " + str(sigmoid(0.0)))
print ("sigmoid(12) = " + str(sigmoid(12)))

def sigmoid_test(target):
    result = target(0)
    assert(type(result) == EagerTensor)
    assert (result.dtype == tf.float32)
    assert sigmoid(0) == 0.5, "Error"
    assert sigmoid(-1) == 0.26894143, "Error"
    assert sigmoid(12) == 0.9999938, "Error"

    print("\033[92mAll test passed")

sigmoid_test(sigmoid)

type: <class 'tensorflow.python.framework.ops.EagerTensor'>
dtype: <dtype: 'float32'>
sigmoid(-1) = tf.Tensor(0.26894143, shape=(), dtype=float32)
sigmoid(0) = tf.Tensor(0.5, shape=(), dtype=float32)
sigmoid(12) = tf.Tensor(0.9999938, shape=(), dtype=float32)
All test passed

2.6 使用独热编码

# GRADED FUNCTION: one_hot_matrix
def one_hot_matrix(label, depth=6):

    one_hot = tf.one_hot(label,depth,axis=0)
    one_hot = tf.reshape(one_hot,[depth,1])

    return one_hot

def one_hot_matrix_test(target):
    label = tf.constant(1)
    depth = 4
    result = target(label, depth)
    print(result)
    assert result.shape[0] == depth, "Use the parameter depth"
    assert result.shape[1] == 1, f"Reshape to have only 1 column"
    assert np.allclose(result,  [[0.], [1.], [0.], [0.]] ), "Wrong output. Use tf.one_hot"
    result = target(3, depth)
    assert np.allclose(result, [[0.], [0.], [0.], [1.]] ), "Wrong output. Use tf.one_hot"
    
    print("\033[92mAll test passed")

one_hot_matrix_test(one_hot_matrix)

tf.Tensor(
[[0.]
[1.]
[0.]
[0.]], shape=(4, 1), dtype=float32)
All test passed

new_y_test = y_test.map(one_hot_matrix)
new_y_train = y_train.map(one_hot_matrix)
print(next(iter(new_y_test)))

tf.Tensor(
[[1.]
[0.]
[0.]
[0.]
[0.]
[0.]], shape=(6, 1), dtype=float32)

2.7 初始化参数

# GRADED FUNCTION: initialize_parameters

def initialize_parameters():
                                
    initializer = tf.keras.initializers.GlorotNormal(seed=1)   

    W1 = tf.Variable(initializer(shape=(25, 12288)))
    b1 = tf.Variable(initializer(shape=(25, 1)))
    W2 = tf.Variable(initializer(shape=(12, 25)))
    b2 = tf.Variable(initializer(shape=(12, 1)))
    W3 = tf.Variable(initializer(shape=(6,12)))
    b3 = tf.Variable(initializer(shape=(6, 1)))

    parameters = {"W1": W1,
                  "b1": b1,
                  "W2": W2,
                  "b2": b2,
                  "W3": W3,
                  "b3": b3}
    
    return parameters

def initialize_parameters_test(target):
    parameters = target()

    values = {"W1": (25, 12288),
              "b1": (25, 1),
              "W2": (12, 25),
              "b2": (12, 1),
              "W3": (6, 12),
              "b3": (6, 1)}

    for key in parameters:
        print(f"{key} shape: {tuple(parameters[key].shape)}")
        assert type(parameters[key]) == ResourceVariable, "All parameter must be created using tf.Variable"
        assert tuple(parameters[key].shape) == values[key], f"{key}: wrong shape"
        assert np.abs(np.mean(parameters[key].numpy())) < 0.5,  f"{key}: Use the GlorotNormal initializer"
        assert np.std(parameters[key].numpy()) > 0 and np.std(parameters[key].numpy()) < 1, f"{key}: Use the GlorotNormal initializer"

    print("\033[92mAll test passed")
    
initialize_parameters_test(initialize_parameters)

W1 shape: (25, 12288)
b1 shape: (25, 1)
W2 shape: (12, 25)
b2 shape: (12, 1)
W3 shape: (6, 12)
b3 shape: (6, 1)
All test passed

parameters = initialize_parameters()

2.8 在TensorFlow中构建你的第一个神经网络

实现正向传播

# GRADED FUNCTION: forward_propagation

@tf.function
def forward_propagation(X, parameters):
    
    # Retrieve the parameters from the dictionary "parameters" 
    W1 = parameters['W1']
    b1 = parameters['b1']
    W2 = parameters['W2']
    b2 = parameters['b2']
    W3 = parameters['W3']
    b3 = parameters['b3']
    
    Z1 = tf.add(tf.matmul(W1,X),b1)    # Z1 = np.dot(W1, X) + b1
    A1 = tf.keras.activations.relu(Z1) # A1 = relu(Z1)
    Z2 = tf.add(tf.matmul(W2,A1),b2)   # Z2 = np.dot(W2, A1) + b2
    A2 = tf.keras.activations.relu(Z2) # A2 = relu(Z2)
    Z3 = tf.add(tf.matmul(W3,A2),b3)   # Z3 = np.dot(W3, A2) + b3
    
    return Z3

def forward_propagation_test(target, examples):
    for batch in examples:
        forward_pass = target(batch, parameters)
        assert type(forward_pass) == EagerTensor, "Your output is not a tensor"
        assert forward_pass.shape == (6, 1), "Last layer must use W3 and b3"
        assert np.any(forward_pass < 0), "Don't use a ReLu layer at end of your network"
        assert np.allclose(forward_pass, 
                           [[-0.13082162],
                           [ 0.21228778],
                           [ 0.7050022 ],
                           [-1.1224034 ],
                           [-0.20386729],
                           [ 0.9526217 ]]), "Output does not match"
        print(forward_pass)
        break
    

    print("\033[92mAll test passed")

forward_propagation_test(forward_propagation, new_train)

tf.Tensor(
[[-0.1308215 ]
[ 0.21228784]
[ 0.7050022 ]
[-1.1224034 ]
[-0.20386738]
[ 0.9526218 ]], shape=(6, 1), dtype=float32)
All test passed

计算代价函数

# GRADED FUNCTION: compute_cost 

@tf.function
def compute_cost(logits, labels):
    
    cost = tf.reduce_mean(tf.keras.losses.binary_crossentropy(labels,logits,from_logits=True))
    
    return cost

def compute_cost_test(target):
    labels = np.array([[0., 1.], [0., 0.], [1., 0.]])
    logits = np.array([[0.6, 0.4], [0.4, 0.6], [0.4, 0.6]])
    result = compute_cost(logits, labels)
    print(result)
    assert(type(result) == EagerTensor), "Use the TensorFlow API"
    assert (np.abs(result - (0.7752516 +  0.9752516 + 0.7752516) / 3.0) < 1e-7), "Test does not match. Did you get the mean of your cost functions?"

    print("\033[92mAll test passed")

compute_cost_test(compute_cost)

tf.Tensor(0.8419182681095857, shape=(), dtype=float64)
All test passed

训练模型

def model(X_train, Y_train, X_test, Y_test, learning_rate = 0.0001,
          num_epochs = 1500, minibatch_size = 32, print_cost = True):
    
    costs = []                                        # To keep track of the cost
    
    # Initialize your parameters
    parameters = initialize_parameters()

    W1 = parameters['W1']
    b1 = parameters['b1']
    W2 = parameters['W2']
    b2 = parameters['b2']
    W3 = parameters['W3']
    b3 = parameters['b3']

    optimizer = tf.keras.optimizers.SGD(learning_rate)

    X_train = X_train.batch(minibatch_size, drop_remainder=True).prefetch(8)# <<< extra step    
    Y_train = Y_train.batch(minibatch_size, drop_remainder=True).prefetch(8) # loads memory faster 

    # Do the training loop
    for epoch in range(num_epochs):

        epoch_cost = 0.
        
        for (minibatch_X, minibatch_Y) in zip(X_train, Y_train):
            # Select a minibatch
            with tf.GradientTape() as tape:
                # 1. predict
                Z3 = forward_propagation(minibatch_X, parameters)
                # 2. loss
                minibatch_cost = compute_cost(Z3, minibatch_Y)
                
            trainable_variables = [W1, b1, W2, b2, W3, b3]
            grads = tape.gradient(minibatch_cost, trainable_variables)
            optimizer.apply_gradients(zip(grads, trainable_variables))
            epoch_cost += minibatch_cost / minibatch_size

        # Print the cost every epoch
        if print_cost == True and epoch % 10 == 0:
            print ("Cost after epoch %i: %f" % (epoch, epoch_cost))
        if print_cost == True and epoch % 5 == 0:
            costs.append(epoch_cost)

    # Plot the cost
    plt.plot(np.squeeze(costs))
    plt.ylabel('cost')
    plt.xlabel('iterations (per fives)')
    plt.title("Learning rate =" + str(learning_rate))
    plt.show()

    # Save the parameters in a variable
    print ("Parameters have been trained!")

    return parameters

model(new_train, new_y_train, new_test, new_y_test, num_epochs=200)

Cost after epoch 0: 0.742591
Cost after epoch 10: 0.614557
Cost after epoch 20: 0.598900
Cost after epoch 30: 0.588907
Cost after epoch 40: 0.579898
Cost after epoch 50: 0.570628
Cost after epoch 60: 0.560898
Cost after epoch 70: 0.550808
Cost after epoch 80: 0.540497
Cost after epoch 90: 0.488141
Cost after epoch 100: 0.478272
Cost after epoch 110: 0.472865
Cost after epoch 120: 0.468991
Cost after epoch 130: 0.466015
Cost after epoch 140: 0.463661
Cost after epoch 150: 0.461677
Cost after epoch 160: 0.459951
Cost after epoch 170: 0.458392
Cost after epoch 180: 0.456970
Cost after epoch 190: 0.455647