Face Generation

In this project, you'll use generative adversarial networks to generate new images of faces.

Get the Data

You'll be using two datasets in this project: - MNIST - CelebA

Since the celebA dataset is complex and you're doing GANs in a project for the first time, we want you to test your neural network on MNIST before CelebA. Running the GANs on MNIST will allow you to see how well your model trains sooner.

If you're using FloydHub, set data_dir to "/input" and use the FloydHub data ID "R5KrjnANiKVhLWAkpXhNBe".

data_dir = './data'

# FloydHub - Use with data ID "R5KrjnANiKVhLWAkpXhNBe"
#data_dir = '/input'


"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import helper

helper.download_extract('mnist', data_dir)
helper.download_extract('celeba', data_dir)

Found mnist Data
Found celeba Data

Explore the Data

MNIST

As you're aware, the MNIST dataset contains images of handwritten digits. You can view the first number of examples by changing show_n_images.

show_n_images = 25

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
%matplotlib inline
import os
from glob import glob
from matplotlib import pyplot

mnist_images = helper.get_batch(glob(os.path.join(data_dir, 'mnist/*.jpg'))[:show_n_images], 28, 28, 'L')
pyplot.imshow(helper.images_square_grid(mnist_images, 'L'), cmap='gray')

<matplotlib.image.AxesImage at 0x7f5ca6429390>

png

CelebA

The CelebFaces Attributes Dataset (CelebA) dataset contains over 200,000 celebrity images with annotations. Since you're going to be generating faces, you won't need the annotations. You can view the first number of examples by changing show_n_images.

show_n_images = 25

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
mnist_images = helper.get_batch(glob(os.path.join(data_dir, 'img_align_celeba/*.jpg'))[:show_n_images], 28, 28, 'RGB')
pyplot.imshow(helper.images_square_grid(mnist_images, 'RGB'))

<matplotlib.image.AxesImage at 0x7f5ca634ff60>

png

Preprocess the Data

Since the project's main focus is on building the GANs, we'll preprocess the data for you. The values of the MNIST and CelebA dataset will be in the range of -0.5 to 0.5 of 28x28 dimensional images. The CelebA images will be cropped to remove parts of the image that don't include a face, then resized down to 28x28.

The MNIST images are black and white images with a single [color channel](https://en.wikipedia.org/wiki/Channel_(digital_image%29) while the CelebA images have [3 color channels (RGB color channel)](https://en.wikipedia.org/wiki/Channel_(digital_image%29#RGB_Images).

Build the Neural Network

You'll build the components necessary to build a GANs by implementing the following functions below: - model_inputs - discriminator - generator - model_loss - model_opt - train

Check the Version of TensorFlow and Access to GPU

This will check to make sure you have the correct version of TensorFlow and access to a GPU

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
from distutils.version import LooseVersion
import warnings
import tensorflow as tf

# Check TensorFlow Version
assert LooseVersion(tf.__version__) >= LooseVersion('1.0'), 'Please use TensorFlow version 1.0 or newer.  You are using {}'.format(tf.__version__)
print('TensorFlow Version: {}'.format(tf.__version__))

# Check for a GPU
if not tf.test.gpu_device_name():
    warnings.warn('No GPU found. Please use a GPU to train your neural network.')
else:
    print('Default GPU Device: {}'.format(tf.test.gpu_device_name()))

TensorFlow Version: 1.0.0
Default GPU Device: /gpu:0

Input

Implement the model_inputs function to create TF Placeholders for the Neural Network. It should create the following placeholders: - Real input images placeholder with rank 4 using image_width, image_height, and image_channels. - Z input placeholder with rank 2 using z_dim. - Learning rate placeholder with rank 0.

Return the placeholders in the following the tuple (tensor of real input images, tensor of z data)

import problem_unittests as tests

def model_inputs(image_width, image_height, image_channels, z_dim):
    """
    Create the model inputs
    :param image_width: The input image width
    :param image_height: The input image height
    :param image_channels: The number of image channels
    :param z_dim: The dimension of Z
    :return: Tuple of (tensor of real input images, tensor of z data, learning rate)
    """
    # TODO: Implement Function


    #1. Real input images placeholder with rank 4 using image_width, image_height, and image_channels.
    input_real = tf.placeholder(tf.float32, [None, image_width,\
                                             image_height,\
                                             image_channels],\
                                             name = "input_real")



    #2. Z input placeholder with rank 2 using z_dim.
    input_z = tf.placeholder(tf.float32, [None, z_dim],\
                                       name = "input_z")


    #3. Learning rate placeholder with rank 0.
    learning_rate = tf.placeholder(tf.float32, None,\
                                   name = "learning_rate")


    return input_real, input_z, learning_rate



"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_inputs(model_inputs)

Tests Passed

Discriminator

Implement discriminator to create a discriminator neural network that discriminates on images. This function should be able to reuse the variabes in the neural network. Use tf.variable_scope with a scope name of "discriminator" to allow the variables to be reused. The function should return a tuple of (tensor output of the generator, tensor logits of the generator).

def discriminator(images, reuse=False):
    """
    Create the discriminator network
    :param image: Tensor of input image(s)
    :param reuse: Boolean if the weights should be reused
    :return: Tuple of (tensor output of the discriminator, tensor logits of the discriminator)
    """
    # TODO: Implement Function

    alpha= 0.1
    sdev= 0.02
    with tf.variable_scope('discriminator', reuse=reuse):

        # input layer with image size(28*28*3)
        x1 = tf.layers.conv2d(images, 32, 5, 2,\
                        padding='same',\
                        kernel_initializer=tf.random_normal_initializer(stddev=sdev))
        lrelu1 = tf.maximum(alpha * x1, x1)
        # Layer 1 out: 14x14x32



        # Layer 2: 14x14x32
        x2 = tf.layers.conv2d(lrelu1, 64, 5, 2,\
                        padding='same',\
                        kernel_initializer=tf.random_normal_initializer(stddev=sdev))
        bn2 = tf.layers.batch_normalization(x2, training=True)
        lrelu2 = tf.maximum(alpha * bn2, bn2)
        # Layer 2 out: 7x7x64



        x3 = tf.layers.conv2d(lrelu2, 128, 5, 2,\
                        padding='same',\
                        kernel_initializer=tf.random_normal_initializer(stddev=sdev))
        bn3 = tf.layers.batch_normalization(x3, training=True)
        lrelu3 = tf.maximum(alpha * bn3, bn3)
        # Layer 3 out: 4x4x128



        x4 = tf.layers.conv2d(lrelu3, 256, 5, 2,\
                        padding='same',\
                        kernel_initializer=tf.random_normal_initializer(stddev=sdev))
        bn4 = tf.layers.batch_normalization(x4, training=True)
        lrelu4 = tf.maximum(alpha * bn4, bn4)
        # Layer 4 out: 2x2x256



        flattened = tf.reshape(lrelu4, (-1, 2*2*256))
        logits = tf.layers.dense(flattened, 1,\
                            kernel_initializer=tf.random_normal_initializer(stddev=sdev))
        output = tf.sigmoid(logits)

        return output, logits


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_discriminator(discriminator, tf)

Tests Passed

Generator

Implement generator to generate an image using z. This function should be able to reuse the variabes in the neural network. Use tf.variable_scope with a scope name of "generator" to allow the variables to be reused. The function should return the generated 28 x 28 x out_channel_dim images.

def generator(z, out_channel_dim, is_train=True):
    """
    Create the generator network
    :param z: Input z
    :param out_channel_dim: The number of channels in the output image
    :param is_train: Boolean if generator is being used for training
    :return: The tensor output of the generator
    """
    # TODO: Implement Function

    alpha = 0.2
    sdev = 0.02
    # variable scope for generator
    with tf.variable_scope('generator', reuse=not is_train):
        #fake with fully connected

        # Layer 1 in: 7x7x256
        x1 = tf.layers.dense(z, 4 * 4 * 512)
        x1 = tf.reshape(x1, (-1, 4, 4, 512))
        lrelu1 = tf.maximum(alpha * x1, x1)
        # Layer 1 out: 4x4x512


        x2 = tf.layers.conv2d_transpose(lrelu1, 128, 4, 1,\
                                padding='valid',\
                                kernel_initializer=tf.random_normal_initializer(stddev=sdev))
        bn2 = tf.layers.batch_normalization(x2, training=is_train)
        lrelu2 = tf.maximum(alpha * bn2, bn2)
        # Layer 2 out: 8x8x128


        x3 = tf.layers.conv2d_transpose(lrelu2, 64, 5, 2,\
                                padding='same', \
                                kernel_initializer=tf.random_normal_initializer(stddev=sdev))
        bn3 = tf.layers.batch_normalization(x3, training=is_train)
        lrelu3 = tf.maximum(alpha * bn3, bn3)
        # Layer 3 out: 16x16x64



        x4 = tf.layers.conv2d_transpose(lrelu3, 32, 5, 2,\
                                padding='same',\
                                kernel_initializer=tf.random_normal_initializer(stddev=sdev))
        bn4 = tf.layers.batch_normalization(x4, training=is_train)
        lrelu4 = tf.maximum(alpha * bn4, bn4)
        # Layer 4 out: 32x32x32



        logits = tf.layers.conv2d_transpose(lrelu4, out_channel_dim, 3, 1,\
                                    padding='same',\
                                    kernel_initializer=tf.random_normal_initializer(stddev=sdev))
        output = tf.tanh(logits)


        return output




"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_generator(generator, tf)

Tests Passed

Loss

Implement model_loss to build the GANs for training and calculate the loss. The function should return a tuple of (discriminator loss, generator loss). Use the following functions you implemented: - discriminator(images, reuse=False) - generator(z, out_channel_dim, is_train=True)

def model_loss(input_real, input_z, out_channel_dim):
    """
    Get the loss for the discriminator and generator
    :param input_real: Images from the real dataset
    :param input_z: Z input
    :param out_channel_dim: The number of channels in the output image
    :return: A tuple of (discriminator loss, generator loss)
    """
    # TODO: Implement Function

    g_model = generator(input_z, out_channel_dim)

    d_model_real, d_logits_real = discriminator(input_real)
    d_model_fake, d_logits_fake = discriminator(g_model, reuse=True)



    d_loss_real = tf.reduce_mean(
        tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_real,\
                                                labels=tf.ones_like(d_model_real)))
    d_loss_fake = tf.reduce_mean(
        tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_fake,\
                                                labels=tf.zeros_like(d_model_fake)))
    g_loss = tf.reduce_mean(
        tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_fake,\
                                                labels=tf.ones_like(d_model_fake)))



    d_loss = d_loss_real + d_loss_fake

    return d_loss, g_loss


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_loss(model_loss)

Tests Passed

Optimization

Implement model_opt to create the optimization operations for the GANs. Use tf.trainable_variables to get all the trainable variables. Filter the variables with names that are in the discriminator and generator scope names. The function should return a tuple of (discriminator training operation, generator training operation).

def model_opt(d_loss, g_loss, learning_rate, beta1):
    """
    Get optimization operations
    :param d_loss: Discriminator loss Tensor
    :param g_loss: Generator loss Tensor
    :param learning_rate: Learning Rate Placeholder
    :param beta1: The exponential decay rate for the 1st moment in the optimizer
    :return: A tuple of (discriminator training operation, generator training operation)
    """
    # TODO: Implement Function


    t_vars = tf.trainable_variables()
    d_vars = [var for var in t_vars if var.name.startswith('discriminator')]
    g_vars = [var for var in t_vars if var.name.startswith('generator')]


    # Optimize
    d_train_opt = tf.train.AdamOptimizer(learning_rate,\
                                    beta1=beta1).minimize(d_loss, var_list=d_vars)


    with tf.control_dependencies(tf.get_collection(tf.GraphKeys.UPDATE_OPS)): 
        g_train_opt = tf.train.AdamOptimizer(learning_rate,\
                                    beta1=beta1).minimize(g_loss, var_list=g_vars)

    return d_train_opt, g_train_opt




"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_opt(model_opt, tf)

Tests Passed

Neural Network Training

Show Output

Use this function to show the current output of the generator during training. It will help you determine how well the GANs is training.

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import numpy as np

def show_generator_output(sess, n_images, input_z, out_channel_dim, image_mode):
    """
    Show example output for the generator
    :param sess: TensorFlow session
    :param n_images: Number of Images to display
    :param input_z: Input Z Tensor
    :param out_channel_dim: The number of channels in the output image
    :param image_mode: The mode to use for images ("RGB" or "L")
    """
    cmap = None if image_mode == 'RGB' else 'gray'
    z_dim = input_z.get_shape().as_list()[-1]
    example_z = np.random.uniform(-1, 1, size=[n_images, z_dim])

    samples = sess.run(
        generator(input_z, out_channel_dim, False),
        feed_dict={input_z: example_z})

    images_grid = helper.images_square_grid(samples, image_mode)
    pyplot.imshow(images_grid, cmap=cmap)
    pyplot.show()

Train

Implement train to build and train the GANs. Use the following functions you implemented: - model_inputs(image_width, image_height, image_channels, z_dim) - model_loss(input_real, input_z, out_channel_dim) - model_opt(d_loss, g_loss, learning_rate, beta1)

Use the show_generator_output to show generator output while you train. Running show_generator_output for every batch will drastically increase training time and increase the size of the notebook. It's recommended to print the generator output every 100 batches.

def train(epoch_count, batch_size, z_dim, learning_rate, beta1, get_batches, data_shape, data_image_mode):
    """
    Train the GAN
    :param epoch_count: Number of epochs
    :param batch_size: Batch Size
    :param z_dim: Z dimension
    :param learning_rate: Learning Rate
    :param beta1: The exponential decay rate for the 1st moment in the optimizer
    :param get_batches: Function to get batches
    :param data_shape: Shape of the data
    :param data_image_mode: The image mode to use for images ("RGB" or "L")
    """
    # TODO: Build Model

    n_samples, width, height, channels = data_shape
    input_real, input_z, learn_rate = model_inputs(width, height, channels, z_dim)
    d_loss, g_loss = model_loss(input_real, input_z, channels)
    d_opt, g_opt = model_opt(d_loss, g_loss, learn_rate, beta1)

    steps = 0
    show_every = 50
    print_every = 25

    with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())
        for epoch_i in range(epoch_count):
            for batch_images in get_batches(batch_size):
                batch_images *= 2

                # TODO: Train Model
                steps += 1

                # Sample random noise for G
                batch_z = np.random.uniform(-1, 1, size=(batch_size, z_dim))


                # Run optimizers
                _ = sess.run(d_opt, feed_dict={input_real: batch_images,\
                                               input_z: batch_z,\
                                               learn_rate: learning_rate})

                _ = sess.run(g_opt, feed_dict={input_real: batch_images,\
                                               input_z: batch_z,\
                                               learn_rate: learning_rate})




                # show_generator_output to show generator output
                if steps % show_every == 0:
                    n_images = 16
                    show_generator_output(sess, n_images, input_z, channels, data_image_mode)




                if steps % print_every == 0:
                    train_loss_d = d_loss.eval({input_z: batch_z, input_real: batch_images})
                    train_loss_g = g_loss.eval({input_z: batch_z})

                    print("Epoch {}/{}...".format(epoch_i, epoch_count),
                          "Discriminator Loss: {:.4f}...".format(train_loss_d),
                          "Generator Loss: {:.4f}".format(train_loss_g))

MNIST

Test your GANs architecture on MNIST. After 2 epochs, the GANs should be able to generate images that look like handwritten digits. Make sure the loss of the generator is lower than the loss of the discriminator or close to 0.

batch_size = 64
z_dim = 100
learning_rate = 0.0005
beta1 = 0.5


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
epochs = 2

mnist_dataset = helper.Dataset('mnist', glob(os.path.join(data_dir, 'mnist/*.jpg')))
with tf.Graph().as_default():
    train(epochs, batch_size, z_dim, learning_rate, beta1, mnist_dataset.get_batches,
          mnist_dataset.shape, mnist_dataset.image_mode)

Epoch 0/2... Discriminator Loss: 0.1132... Generator Loss: 3.8164

png

Epoch 0/2... Discriminator Loss: 0.4083... Generator Loss: 1.6668
Epoch 0/2... Discriminator Loss: 0.2134... Generator Loss: 2.3080

png

Epoch 0/2... Discriminator Loss: 0.8935... Generator Loss: 0.7820
Epoch 0/2... Discriminator Loss: 0.9974... Generator Loss: 0.9753

png

Epoch 0/2... Discriminator Loss: 1.4598... Generator Loss: 0.4637
Epoch 0/2... Discriminator Loss: 1.0837... Generator Loss: 0.9862

png

Epoch 0/2... Discriminator Loss: 1.0771... Generator Loss: 1.2634
Epoch 0/2... Discriminator Loss: 1.5585... Generator Loss: 0.3205

png

Epoch 0/2... Discriminator Loss: 1.0825... Generator Loss: 0.6745
Epoch 0/2... Discriminator Loss: 0.9129... Generator Loss: 1.0191

png

Epoch 0/2... Discriminator Loss: 0.9434... Generator Loss: 0.8767
Epoch 0/2... Discriminator Loss: 0.9463... Generator Loss: 0.9866

png

Epoch 0/2... Discriminator Loss: 1.1447... Generator Loss: 0.6476
Epoch 0/2... Discriminator Loss: 1.3758... Generator Loss: 0.4319

png

Epoch 0/2... Discriminator Loss: 0.9714... Generator Loss: 0.6623
Epoch 0/2... Discriminator Loss: 1.5291... Generator Loss: 0.3816

png

Epoch 0/2... Discriminator Loss: 1.3693... Generator Loss: 2.0586
Epoch 0/2... Discriminator Loss: 1.7561... Generator Loss: 0.2967

png

Epoch 0/2... Discriminator Loss: 1.5004... Generator Loss: 0.3576
Epoch 0/2... Discriminator Loss: 0.7120... Generator Loss: 1.2979

png

Epoch 0/2... Discriminator Loss: 0.8413... Generator Loss: 1.0782
Epoch 0/2... Discriminator Loss: 1.1858... Generator Loss: 0.5920

png

Epoch 0/2... Discriminator Loss: 1.0060... Generator Loss: 0.6120
Epoch 0/2... Discriminator Loss: 1.2757... Generator Loss: 0.4474

png

Epoch 0/2... Discriminator Loss: 1.5774... Generator Loss: 0.3005
Epoch 0/2... Discriminator Loss: 1.2721... Generator Loss: 0.4242

png

Epoch 0/2... Discriminator Loss: 1.0972... Generator Loss: 0.5191
Epoch 0/2... Discriminator Loss: 0.8444... Generator Loss: 1.2999

png

Epoch 0/2... Discriminator Loss: 0.8549... Generator Loss: 0.9284
Epoch 0/2... Discriminator Loss: 0.8498... Generator Loss: 0.9417

png

Epoch 0/2... Discriminator Loss: 0.6521... Generator Loss: 1.2973
Epoch 0/2... Discriminator Loss: 0.9847... Generator Loss: 0.6543

png

Epoch 0/2... Discriminator Loss: 1.2542... Generator Loss: 0.4832
Epoch 0/2... Discriminator Loss: 0.9244... Generator Loss: 1.4905

png

Epoch 0/2... Discriminator Loss: 1.0157... Generator Loss: 0.5582
Epoch 0/2... Discriminator Loss: 0.5705... Generator Loss: 1.7848

png

Epoch 1/2... Discriminator Loss: 1.0804... Generator Loss: 0.7368
Epoch 1/2... Discriminator Loss: 1.3714... Generator Loss: 2.7104

png

Epoch 1/2... Discriminator Loss: 0.7393... Generator Loss: 1.0792
Epoch 1/2... Discriminator Loss: 0.8901... Generator Loss: 0.9193

png

Epoch 1/2... Discriminator Loss: 1.3041... Generator Loss: 0.4482
Epoch 1/2... Discriminator Loss: 1.0246... Generator Loss: 1.6754

png

Epoch 1/2... Discriminator Loss: 1.4366... Generator Loss: 0.4005
Epoch 1/2... Discriminator Loss: 0.7496... Generator Loss: 0.9064

png

Epoch 1/2... Discriminator Loss: 1.0066... Generator Loss: 0.6571
Epoch 1/2... Discriminator Loss: 0.6344... Generator Loss: 1.0048

png

Epoch 1/2... Discriminator Loss: 0.9111... Generator Loss: 1.4843
Epoch 1/2... Discriminator Loss: 0.8954... Generator Loss: 1.0413

png

Epoch 1/2... Discriminator Loss: 2.1049... Generator Loss: 0.1755
Epoch 1/2... Discriminator Loss: 1.0708... Generator Loss: 0.5250

png

Epoch 1/2... Discriminator Loss: 1.4498... Generator Loss: 0.3438
Epoch 1/2... Discriminator Loss: 0.8951... Generator Loss: 0.7148

png

Epoch 1/2... Discriminator Loss: 0.9677... Generator Loss: 0.6380
Epoch 1/2... Discriminator Loss: 2.0528... Generator Loss: 0.1594

png

Epoch 1/2... Discriminator Loss: 0.5354... Generator Loss: 1.1890
Epoch 1/2... Discriminator Loss: 1.5721... Generator Loss: 0.5104

png

Epoch 1/2... Discriminator Loss: 0.7788... Generator Loss: 0.9999
Epoch 1/2... Discriminator Loss: 0.8880... Generator Loss: 0.6436

png

Epoch 1/2... Discriminator Loss: 0.6986... Generator Loss: 1.4282
Epoch 1/2... Discriminator Loss: 1.2309... Generator Loss: 0.4484

png

Epoch 1/2... Discriminator Loss: 1.0756... Generator Loss: 0.6155
Epoch 1/2... Discriminator Loss: 3.1573... Generator Loss: 3.8849

png

Epoch 1/2... Discriminator Loss: 0.8739... Generator Loss: 0.6847
Epoch 1/2... Discriminator Loss: 1.7689... Generator Loss: 0.2244

png

Epoch 1/2... Discriminator Loss: 0.8876... Generator Loss: 0.7815
Epoch 1/2... Discriminator Loss: 1.1661... Generator Loss: 0.4628

png

Epoch 1/2... Discriminator Loss: 1.5133... Generator Loss: 0.3179
Epoch 1/2... Discriminator Loss: 0.6079... Generator Loss: 1.3311

png

Epoch 1/2... Discriminator Loss: 0.7159... Generator Loss: 1.5858
Epoch 1/2... Discriminator Loss: 0.7321... Generator Loss: 0.8564

png

Epoch 1/2... Discriminator Loss: 1.2747... Generator Loss: 0.4073
Epoch 1/2... Discriminator Loss: 1.3412... Generator Loss: 0.3678

png

Epoch 1/2... Discriminator Loss: 0.8914... Generator Loss: 0.7424

CelebA

Run your GANs on CelebA. It will take around 20 minutes on the average GPU to run one epoch. You can run the whole epoch or stop when it starts to generate realistic faces.

batch_size = 64
z_dim = 100
learning_rate = 0.0005
beta1 = 0.5


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
epochs = 1

celeba_dataset = helper.Dataset('celeba', glob(os.path.join(data_dir, 'img_align_celeba/*.jpg')))
with tf.Graph().as_default():
    train(epochs, batch_size, z_dim, learning_rate, beta1, celeba_dataset.get_batches,
          celeba_dataset.shape, celeba_dataset.image_mode)

Epoch 0/1... Discriminator Loss: 0.6341... Generator Loss: 1.5425

png

Epoch 0/1... Discriminator Loss: 1.0661... Generator Loss: 1.4051
Epoch 0/1... Discriminator Loss: 1.4774... Generator Loss: 0.4814

png

Epoch 0/1... Discriminator Loss: 0.5296... Generator Loss: 2.0486
Epoch 0/1... Discriminator Loss: 0.4144... Generator Loss: 1.7049

png

Epoch 0/1... Discriminator Loss: 0.8953... Generator Loss: 1.3714
Epoch 0/1... Discriminator Loss: 1.2676... Generator Loss: 2.1938

png

Epoch 0/1... Discriminator Loss: 0.6703... Generator Loss: 1.9268
Epoch 0/1... Discriminator Loss: 1.3622... Generator Loss: 0.9100

png

Epoch 0/1... Discriminator Loss: 0.8473... Generator Loss: 1.2377
Epoch 0/1... Discriminator Loss: 0.7923... Generator Loss: 1.3153

png

Epoch 0/1... Discriminator Loss: 2.3745... Generator Loss: 3.9981
Epoch 0/1... Discriminator Loss: 1.3498... Generator Loss: 0.9091

png

Epoch 0/1... Discriminator Loss: 0.9757... Generator Loss: 1.3260
Epoch 0/1... Discriminator Loss: 0.9970... Generator Loss: 0.8662

png

Epoch 0/1... Discriminator Loss: 1.2872... Generator Loss: 0.7643
Epoch 0/1... Discriminator Loss: 0.9423... Generator Loss: 1.0006

png

Epoch 0/1... Discriminator Loss: 0.6227... Generator Loss: 1.2088
Epoch 0/1... Discriminator Loss: 0.5900... Generator Loss: 1.5671

png

Epoch 0/1... Discriminator Loss: 1.0243... Generator Loss: 0.9068
Epoch 0/1... Discriminator Loss: 1.5452... Generator Loss: 0.3467

png

Epoch 0/1... Discriminator Loss: 0.9187... Generator Loss: 1.3025
Epoch 0/1... Discriminator Loss: 0.9896... Generator Loss: 0.9055

png

Epoch 0/1... Discriminator Loss: 0.9112... Generator Loss: 1.2679
Epoch 0/1... Discriminator Loss: 1.0865... Generator Loss: 1.3766

png

Epoch 0/1... Discriminator Loss: 1.0086... Generator Loss: 0.7582
Epoch 0/1... Discriminator Loss: 0.8523... Generator Loss: 1.3466

png

Epoch 0/1... Discriminator Loss: 0.7004... Generator Loss: 1.6204
Epoch 0/1... Discriminator Loss: 1.1147... Generator Loss: 1.0215

png

Epoch 0/1... Discriminator Loss: 0.9817... Generator Loss: 0.9707
Epoch 0/1... Discriminator Loss: 0.8524... Generator Loss: 0.7910

png

Epoch 0/1... Discriminator Loss: 1.3500... Generator Loss: 0.6876
Epoch 0/1... Discriminator Loss: 1.1861... Generator Loss: 0.7478

png

Epoch 0/1... Discriminator Loss: 1.4163... Generator Loss: 0.4300
Epoch 0/1... Discriminator Loss: 1.3214... Generator Loss: 0.6985

png

Epoch 0/1... Discriminator Loss: 1.6581... Generator Loss: 0.3179
Epoch 0/1... Discriminator Loss: 0.8624... Generator Loss: 2.0154

png

Epoch 0/1... Discriminator Loss: 0.7325... Generator Loss: 1.5246
Epoch 0/1... Discriminator Loss: 1.3200... Generator Loss: 0.4783

png

Epoch 0/1... Discriminator Loss: 0.9317... Generator Loss: 0.8402
Epoch 0/1... Discriminator Loss: 1.2700... Generator Loss: 1.6259

png

Epoch 0/1... Discriminator Loss: 1.6959... Generator Loss: 1.6403
Epoch 0/1... Discriminator Loss: 0.9565... Generator Loss: 1.8874

png

Epoch 0/1... Discriminator Loss: 0.7762... Generator Loss: 1.6332
Epoch 0/1... Discriminator Loss: 0.9902... Generator Loss: 0.6471

png

Epoch 0/1... Discriminator Loss: 1.0775... Generator Loss: 0.6818
Epoch 0/1... Discriminator Loss: 1.0928... Generator Loss: 0.9076

png

Epoch 0/1... Discriminator Loss: 1.2636... Generator Loss: 0.8888
Epoch 0/1... Discriminator Loss: 1.9158... Generator Loss: 0.1960

png

Epoch 0/1... Discriminator Loss: 1.0883... Generator Loss: 0.6902
Epoch 0/1... Discriminator Loss: 1.0772... Generator Loss: 0.6896

png

Epoch 0/1... Discriminator Loss: 1.1797... Generator Loss: 0.5554
Epoch 0/1... Discriminator Loss: 1.1049... Generator Loss: 0.9699

png

Epoch 0/1... Discriminator Loss: 1.2788... Generator Loss: 2.0408
Epoch 0/1... Discriminator Loss: 0.7399... Generator Loss: 1.1408

png

Epoch 0/1... Discriminator Loss: 1.3329... Generator Loss: 0.4466
Epoch 0/1... Discriminator Loss: 0.9657... Generator Loss: 0.9572

png

Epoch 0/1... Discriminator Loss: 1.0361... Generator Loss: 0.7271
Epoch 0/1... Discriminator Loss: 1.0897... Generator Loss: 0.8504

png

Epoch 0/1... Discriminator Loss: 1.2010... Generator Loss: 1.1790
Epoch 0/1... Discriminator Loss: 1.6883... Generator Loss: 1.1168

png

Epoch 0/1... Discriminator Loss: 1.0092... Generator Loss: 0.6636
Epoch 0/1... Discriminator Loss: 1.3417... Generator Loss: 0.5133

png

Epoch 0/1... Discriminator Loss: 0.9791... Generator Loss: 0.9111
Epoch 0/1... Discriminator Loss: 0.8803... Generator Loss: 1.1090

png

Epoch 0/1... Discriminator Loss: 0.8513... Generator Loss: 1.2366
Epoch 0/1... Discriminator Loss: 0.7197... Generator Loss: 1.5789

png

Epoch 0/1... Discriminator Loss: 0.9334... Generator Loss: 0.7907
Epoch 0/1... Discriminator Loss: 1.1299... Generator Loss: 0.5871

png

Epoch 0/1... Discriminator Loss: 0.9805... Generator Loss: 1.1100
Epoch 0/1... Discriminator Loss: 1.1180... Generator Loss: 2.0649

png

Epoch 0/1... Discriminator Loss: 0.9413... Generator Loss: 0.6843
Epoch 0/1... Discriminator Loss: 1.1723... Generator Loss: 1.0662

png

Epoch 0/1... Discriminator Loss: 1.3740... Generator Loss: 0.4719
Epoch 0/1... Discriminator Loss: 1.2226... Generator Loss: 0.7428

png

Epoch 0/1... Discriminator Loss: 1.0916... Generator Loss: 1.5043
Epoch 0/1... Discriminator Loss: 1.2517... Generator Loss: 2.0049

png

Epoch 0/1... Discriminator Loss: 1.1599... Generator Loss: 0.7082
Epoch 0/1... Discriminator Loss: 1.0402... Generator Loss: 0.8053

png

Epoch 0/1... Discriminator Loss: 1.0036... Generator Loss: 0.9750
Epoch 0/1... Discriminator Loss: 0.9274... Generator Loss: 0.9230

png

Epoch 0/1... Discriminator Loss: 1.1304... Generator Loss: 1.1508
Epoch 0/1... Discriminator Loss: 0.7811... Generator Loss: 1.4115

png

Epoch 0/1... Discriminator Loss: 1.0645... Generator Loss: 0.8931
Epoch 0/1... Discriminator Loss: 1.4299... Generator Loss: 2.5213

png

Epoch 0/1... Discriminator Loss: 1.0405... Generator Loss: 0.6912
Epoch 0/1... Discriminator Loss: 1.1746... Generator Loss: 0.6077

png

Epoch 0/1... Discriminator Loss: 0.9569... Generator Loss: 2.3023
Epoch 0/1... Discriminator Loss: 0.9257... Generator Loss: 0.9512

png

Epoch 0/1... Discriminator Loss: 0.9541... Generator Loss: 0.7513
Epoch 0/1... Discriminator Loss: 0.9034... Generator Loss: 0.9518

png

Epoch 0/1... Discriminator Loss: 1.1156... Generator Loss: 0.8511
Epoch 0/1... Discriminator Loss: 0.9349... Generator Loss: 1.1430

png

Epoch 0/1... Discriminator Loss: 1.0691... Generator Loss: 0.6279
Epoch 0/1... Discriminator Loss: 1.1409... Generator Loss: 0.9605

png

Epoch 0/1... Discriminator Loss: 1.4045... Generator Loss: 0.3858
Epoch 0/1... Discriminator Loss: 1.0732... Generator Loss: 0.7521

png

Epoch 0/1... Discriminator Loss: 1.0664... Generator Loss: 0.9785
Epoch 0/1... Discriminator Loss: 1.1484... Generator Loss: 1.4160

png

Epoch 0/1... Discriminator Loss: 1.2619... Generator Loss: 0.5863
Epoch 0/1... Discriminator Loss: 0.9650... Generator Loss: 0.9150

png

Epoch 0/1... Discriminator Loss: 0.9211... Generator Loss: 1.4288
Epoch 0/1... Discriminator Loss: 0.9854... Generator Loss: 0.7512

png

Epoch 0/1... Discriminator Loss: 0.8539... Generator Loss: 0.9985
Epoch 0/1... Discriminator Loss: 0.9036... Generator Loss: 1.0608

png

Epoch 0/1... Discriminator Loss: 1.0430... Generator Loss: 0.7103
Epoch 0/1... Discriminator Loss: 0.8404... Generator Loss: 0.9762

png

Epoch 0/1... Discriminator Loss: 1.2528... Generator Loss: 0.5786
Epoch 0/1... Discriminator Loss: 1.2818... Generator Loss: 0.4568

png

Epoch 0/1... Discriminator Loss: 1.0006... Generator Loss: 0.8900
Epoch 0/1... Discriminator Loss: 1.3815... Generator Loss: 0.3886

png

Epoch 0/1... Discriminator Loss: 1.2679... Generator Loss: 0.5873
Epoch 0/1... Discriminator Loss: 0.9585... Generator Loss: 0.9575

png

Epoch 0/1... Discriminator Loss: 1.2240... Generator Loss: 0.7966
Epoch 0/1... Discriminator Loss: 0.8579... Generator Loss: 1.4551

png

Epoch 0/1... Discriminator Loss: 1.3028... Generator Loss: 0.5394
Epoch 0/1... Discriminator Loss: 1.1112... Generator Loss: 0.7762

png

Epoch 0/1... Discriminator Loss: 1.1230... Generator Loss: 0.9227
Epoch 0/1... Discriminator Loss: 1.2362... Generator Loss: 0.5178

png

Epoch 0/1... Discriminator Loss: 1.0913... Generator Loss: 0.7929
Epoch 0/1... Discriminator Loss: 0.8596... Generator Loss: 0.9705

png

Epoch 0/1... Discriminator Loss: 1.1329... Generator Loss: 0.6999
Epoch 0/1... Discriminator Loss: 1.0311... Generator Loss: 0.8163

png

Epoch 0/1... Discriminator Loss: 0.8289... Generator Loss: 1.1358
Epoch 0/1... Discriminator Loss: 1.2079... Generator Loss: 0.5730

png

Epoch 0/1... Discriminator Loss: 0.9879... Generator Loss: 0.7466

Submitting This Project

When submitting this project, make sure to run all the cells before saving the notebook. Save the notebook file as "dlnd_face_generation.ipynb" and save it as a HTML file under "File" -> "Download as". Include the "helper.py" and "problem_unittests.py" files in your submission.