1) Importing Python Packages for GAN

In [6]:

from keras.datasets import mnist

from keras.models import Sequential

from keras.layers import BatchNormalization
from keras.layers import Dense, Reshape, Flatten
from keras.layers import LeakyReLU
from tensorflow.keras.optimizers import Adam

import numpy as np
!mkdir generated_images

2) Variables for Neural Networks & Data

In [7]:

img_width = 28
img_height = 28
channels = 1
img_shape = (img_width, img_height, channels)
latent_dim = 100
adam = Adam(lr=0.0001)

/usr/local/lib/python3.7/dist-packages/keras/optimizers/optimizer_v2/adam.py:110: UserWarn
ing: The `lr` argument is deprecated, use `learning_rate` instead.
super(Adam, self).__init__(name, **kwargs)

3) Building Generator
In [8]:

def build_generator():
model = Sequential()

model.add(Dense(256, input_dim=latent_dim))
model.add(LeakyReLU(alpha=0.2))
model.add(BatchNormalization(momentum=0.8))

model.add(Dense(256))
model.add(LeakyReLU(alpha=0.2))
model.add(BatchNormalization(momentum=0.8))

model.add(Dense(256))
model.add(LeakyReLU(alpha=0.2))
model.add(BatchNormalization(momentum=0.8))

model.add(Dense(np.prod(img_shape), activation='tanh'))
model.add(Reshape(img_shape))

model.summary()
return model

generator = build_generator()

Model: "sequential"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
dense (Dense) (None, 256) 25856

leaky_re_lu (LeakyReLU) (None, 256) 0

 batch_normalization (BatchN (None, 256) 1024
ormalization)

dense_1 (Dense) (None, 256) 65792

leaky_re_lu_1 (LeakyReLU) (None, 256) 0

batch_normalization_1 (Batc (None, 256) 1024

hNormalization)

dense_2 (Dense) (None, 256) 65792

leaky_re_lu_2 (LeakyReLU) (None, 256) 0

batch_normalization_2 (Batc (None, 256) 1024

hNormalization)

dense_3 (Dense) (None, 784) 201488

reshape (Reshape) (None, 28, 28, 1) 0

=================================================================
Total params: 362,000
Trainable params: 360,464
Non-trainable params: 1,536
_________________________________________________________________

4) Building Discriminator
In [9]:
def build_discriminator():
model = Sequential()

model.add(Flatten(input_shape=img_shape))
model.add(Dense(512))
model.add(LeakyReLU(alpha=0.2))
model.add(Dense(256))
model.add(Dense(1, activation='sigmoid'))

model.summary()
return model

discriminator = build_discriminator()
discriminator.compile(loss='binary_crossentropy', optimizer=adam, metrics=['accuracy'])

Model: "sequential_1"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
flatten (Flatten) (None, 784) 0

dense_4 (Dense) (None, 512) 401920

leaky_re_lu_3 (LeakyReLU) (None, 512) 0

dense_5 (Dense) (None, 256) 131328

dense_6 (Dense) (None, 1) 257

=================================================================
Total params: 533,505
Trainable params: 533,505
Non-trainable params: 0
_________________________________________________________________

5) Connecting Neural Networks to build GAN

In [10]:
GAN = Sequential()
discriminator.trainable = False
GAN.add(generator)
GAN.add(discriminator)

GAN.compile(loss='binary_crossentropy', optimizer=adam)
GAN.summary()

Model: "sequential_2"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
sequential (Sequential) (None, 28, 28, 1) 362000

sequential_1 (Sequential) (None, 1) 533505

=================================================================
Total params: 895,505
Trainable params: 360,464
Non-trainable params: 535,041
_________________________________________________________________

6) Outputting Images
In [11]:

#@title
## **7) Outputting Images**
import matplotlib.pyplot as plt
import glob
import imageio
import PIL

save_name = 0.00000000

def save_imgs(epoch):
r, c = 5, 5
noise = np.random.normal(0, 1, (r * c, latent_dim))
gen_imgs = generator.predict(noise)
global save_name
save_name += 0.00000001
print("%.8f" % save_name)

# Rescale images 0 - 1
gen_imgs = 0.5 * gen_imgs + 0.5

fig, axs = plt.subplots(r, c)

cnt = 0
for i in range(r):
for j in range(c):
axs[i,j].imshow(gen_imgs[cnt, :,:,0], cmap='gray')
# axs[i,j].imshow(gen_imgs[cnt])
axs[i,j].axis('off')
cnt += 1
fig.savefig("generated_images/%.8f.png" % save_name)
print('saved')
plt.close()

7) Training GAN
In [12]:
def train(epochs, batch_size=64, save_interval=200):
(X_train, _), (_, _) = mnist.load_data()

# print(X_train.shape)
#Rescale data between -1 and 1
 X_train = X_train / 127.5 -1.
# X_train = np.expand_dims(X_train, axis=3)
# print(X_train.shape)

#Create our Y for our Neural Networks

valid = np.ones((batch_size, 1))
fakes = np.zeros((batch_size, 1))

for epoch in range(epochs):

#Get Random Batch
idx = np.random.randint(0, X_train.shape[0], batch_size)
imgs = X_train[idx]

#Generate Fake Images

noise = np.random.normal(0, 1, (batch_size, latent_dim))
gen_imgs = generator.predict(noise)

#Train discriminator
d_loss_real = discriminator.train_on_batch(imgs, valid)
d_loss_fake = discriminator.train_on_batch(gen_imgs, fakes)
d_loss = 0.5 * np.add(d_loss_real, d_loss_fake)

noise = np.random.normal(0, 1, (batch_size, latent_dim))

#inverse y label
g_loss = GAN.train_on_batch(noise, valid)

print("******* %d [D loss: %f , acc: %.2f%%] [G loss: %f ]" % (epoch, d_loss[0], 100* d

_loss[1], g_loss))

if (epoch % save_interval) == 0:
save_imgs(epoch)

# print(valid)

train(30000, batch_size=64, save_interval=200)

Streaming output truncated to the last 5000 lines.

2/2 [==============================] - 0s 5ms/step
******* 27518 [D loss: 0.484928, acc: 77.34%] [G loss: 1.363376]
2/2 [==============================] - 0s 5ms/step
******* 27519 [D loss: 0.600889, acc: 71.09%] [G loss: 1.262974]
2/2 [==============================] - 0s 6ms/step
******* 27520 [D loss: 0.628647, acc: 65.62%] [G loss: 1.172042]
2/2 [==============================] - 0s 6ms/step
******* 27521 [D loss: 0.578565, acc: 68.75%] [G loss: 1.006741]
2/2 [==============================] - 0s 4ms/step
******* 27522 [D loss: 0.675294, acc: 62.50%] [G loss: 1.118249]
2/2 [==============================] - 0s 7ms/step
******* 27523 [D loss: 0.606178, acc: 71.09%] [G loss: 1.202019]
2/2 [==============================] - 0s 5ms/step
******* 27524 [D loss: 0.577705, acc: 72.66%] [G loss: 1.358428]
2/2 [==============================] - 0s 8ms/step
******* 27525 [D loss: 0.614934, acc: 67.19%] [G loss: 1.408203]
2/2 [==============================] - 0s 4ms/step
******* 27526 [D loss: 0.624310, acc: 64.84%] [G loss: 1.455104]
2/2 [==============================] - 0s 5ms/step
******* 27527 [D loss: 0.637720, acc: 65.62%] [G loss: 1.206390]
2/2 [==============================] - 0s 7ms/step
******* 27528 [D loss: 0.580422, acc: 74.22%] [G loss: 1.186193]
2/2 [==============================] - 0s 5ms/step
******* 27529 [D loss: 0.616732, acc: 64.84%] [G loss: 1.137124]
2/2 [==============================] - 0s 5ms/step
******* 27530 [D loss: 0.537695, acc: 71.88%] [G loss: 1.224113]
2/2 [==============================] - 0s 7ms/step
******* 27531 [D loss: 0.503431, acc: 72.66%] [G loss: 1.385631]
2/2 [==============================] - 0s 6ms/step
******* 27532 [D loss: 0.543676, acc: 71.09%] [G loss: 1.528680]
2/2 [==============================] - 0s 6ms/step
******* 27533 [D loss: 0.601877, acc: 68.75%] [G loss: 1.466871]
2/2 [==============================] - 0s 6ms/step
******* 27534 [D loss: 0.654370, acc: 63.28%] [G loss: 1.062838]
2/2 [==============================] - 0s 9ms/step
******* 29863 [D loss: 0.623452, acc: 67.19%] [G loss: 1.409948]
2/2 [==============================] - 0s 4ms/step
******* 29864 [D loss: 0.542609, acc: 74.22%] [G loss: 1.172149]
2/2 [==============================] - 0s 7ms/step
******* 29865 [D loss: 0.625219, acc: 65.62%] [G loss: 1.020908]
2/2 [==============================] - 0s 5ms/step
******* 29866 [D loss: 0.583065, acc: 73.44%] [G loss: 1.155057]
2/2 [==============================] - 0s 4ms/step
******* 29867 [D loss: 0.576574, acc: 70.31%] [G loss: 1.256038]
2/2 [==============================] - 0s 6ms/step
******* 29868 [D loss: 0.568287, acc: 69.53%] [G loss: 1.301167]
2/2 [==============================] - 0s 8ms/step
******* 29869 [D loss: 0.655900, acc: 66.41%] [G loss: 1.149893]
2/2 [==============================] - 0s 5ms/step
******* 29870 [D loss: 0.608220, acc: 72.66%] [G loss: 1.271738]
2/2 [==============================] - 0s 5ms/step
******* 29871 [D loss: 0.594480, acc: 67.97%] [G loss: 1.228074]
2/2 [==============================] - 0s 6ms/step
******* 29872 [D loss: 0.540134, acc: 71.88%] [G loss: 1.311851]
2/2 [==============================] - 0s 8ms/step
******* 29873 [D loss: 0.610670, acc: 68.75%] [G loss: 1.280084]
2/2 [==============================] - 0s 6ms/step
******* 29874 [D loss: 0.496260, acc: 77.34%] [G loss: 1.255110]
2/2 [==============================] - 0s 6ms/step
******* 29875 [D loss: 0.701481, acc: 56.25%] [G loss: 1.188344]
2/2 [==============================] - 0s 6ms/step
******* 29876 [D loss: 0.661403, acc: 62.50%] [G loss: 1.132108]
2/2 [==============================] - 0s 9ms/step
******* 29877 [D loss: 0.654212, acc: 65.62%] [G loss: 1.086925]
2/2 [==============================] - 0s 13ms/step
******* 29878 [D loss: 0.618295, acc: 67.19%] [G loss: 1.164721]
2/2 [==============================] - 0s 5ms/step
******* 29879 [D loss: 0.520297, acc: 75.78%] [G loss: 1.123332]
2/2 [==============================] - 0s 9ms/step
******* 29880 [D loss: 0.553012, acc: 71.88%] [G loss: 1.283132]
2/2 [==============================] - 0s 8ms/step
******* 29881 [D loss: 0.547159, acc: 70.31%] [G loss: 1.374631]
2/2 [==============================] - 0s 4ms/step
******* 29882 [D loss: 0.536171, acc: 74.22%] [G loss: 1.275265]
2/2 [==============================] - 0s 6ms/step
******* 29883 [D loss: 0.629597, acc: 59.38%] [G loss: 1.039468]
2/2 [==============================] - 0s 6ms/step
******* 29884 [D loss: 0.593872, acc: 74.22%] [G loss: 1.073526]
2/2 [==============================] - 0s 11ms/step
******* 29885 [D loss: 0.525715, acc: 71.88%] [G loss: 1.254506]
2/2 [==============================] - 0s 4ms/step
******* 29886 [D loss: 0.536416, acc: 74.22%] [G loss: 1.422001]
2/2 [==============================] - 0s 6ms/step
******* 29887 [D loss: 0.630727, acc: 67.19%] [G loss: 1.484703]
2/2 [==============================] - 0s 6ms/step
******* 29888 [D loss: 0.639953, acc: 63.28%] [G loss: 1.229375]
2/2 [==============================] - 0s 4ms/step
******* 29889 [D loss: 0.613540, acc: 66.41%] [G loss: 0.990796]
2/2 [==============================] - 0s 5ms/step
******* 29890 [D loss: 0.555331, acc: 72.66%] [G loss: 1.112627]
2/2 [==============================] - 0s 12ms/step
******* 29891 [D loss: 0.542436, acc: 76.56%] [G loss: 1.129661]
2/2 [==============================] - 0s 6ms/step
******* 29892 [D loss: 0.580063, acc: 67.19%] [G loss: 1.215682]
2/2 [==============================] - 0s 4ms/step
******* 29893 [D loss: 0.515409, acc: 78.12%] [G loss: 1.328687]
2/2 [==============================] - 0s 7ms/step
******* 29894 [D loss: 0.631865, acc: 67.97%] [G loss: 1.267977]
2/2 [==============================] - 0s 14ms/step
******* 29895 [D loss: 0.613308, acc: 64.06%] [G loss: 1.121264]
2/2 [==============================] - 0s 6ms/step
******* 29896 [D loss: 0.480486, acc: 77.34%] [G loss: 1.309521]
2/2 [==============================] - 0s 9ms/step
******* 29897 [D loss: 0.554778, acc: 74.22%] [G loss: 1.268197]
2/2 [==============================] - 0s 6ms/step
******* 29898 [D loss: 0.664903, acc: 64.06%] [G loss: 1.332328]
2/2 [==============================] - 0s 7ms/step
2/2 [==============================] - 0s 7ms/step
******* 29899 [D loss: 0.605389, acc: 70.31%] [G loss: 1.193784]
2/2 [==============================] - 0s 5ms/step
******* 29900 [D loss: 0.640663, acc: 64.84%] [G loss: 1.235355]
2/2 [==============================] - 0s 4ms/step
******* 29901 [D loss: 0.575814, acc: 71.09%] [G loss: 1.221311]
2/2 [==============================] - 0s 10ms/step
******* 29902 [D loss: 0.626534, acc: 71.88%] [G loss: 1.253850]
2/2 [==============================] - 0s 5ms/step
******* 29903 [D loss: 0.603055, acc: 67.19%] [G loss: 1.261721]
2/2 [==============================] - 0s 7ms/step
******* 29904 [D loss: 0.735876, acc: 62.50%] [G loss: 1.220604]
2/2 [==============================] - 0s 6ms/step
******* 29905 [D loss: 0.605115, acc: 65.62%] [G loss: 1.112600]
2/2 [==============================] - 0s 5ms/step
******* 29906 [D loss: 0.625115, acc: 66.41%] [G loss: 1.220238]
2/2 [==============================] - 0s 8ms/step
******* 29907 [D loss: 0.577993, acc: 67.97%] [G loss: 1.159944]
2/2 [==============================] - 0s 8ms/step
******* 29908 [D loss: 0.550957, acc: 74.22%] [G loss: 1.290124]
2/2 [==============================] - 0s 10ms/step
******* 29909 [D loss: 0.637799, acc: 64.06%] [G loss: 1.455835]
2/2 [==============================] - 0s 5ms/step
******* 29910 [D loss: 0.635876, acc: 65.62%] [G loss: 1.216055]
2/2 [==============================] - 0s 7ms/step
******* 29911 [D loss: 0.633757, acc: 66.41%] [G loss: 0.844575]
2/2 [==============================] - 0s 7ms/step
******* 29912 [D loss: 0.544526, acc: 71.09%] [G loss: 1.100386]
2/2 [==============================] - 0s 8ms/step
******* 29913 [D loss: 0.604366, acc: 67.19%] [G loss: 1.279352]
2/2 [==============================] - 0s 5ms/step
******* 29914 [D loss: 0.623141, acc: 61.72%] [G loss: 1.315183]
2/2 [==============================] - 0s 4ms/step
******* 29915 [D loss: 0.593884, acc: 68.75%] [G loss: 1.161430]
2/2 [==============================] - 0s 4ms/step
******* 29916 [D loss: 0.562622, acc: 67.97%] [G loss: 1.314931]
2/2 [==============================] - 0s 8ms/step
******* 29917 [D loss: 0.551400, acc: 76.56%] [G loss: 1.188707]
2/2 [==============================] - 0s 6ms/step
******* 29918 [D loss: 0.542793, acc: 75.00%] [G loss: 1.134578]
2/2 [==============================] - 0s 6ms/step
******* 29919 [D loss: 0.542166, acc: 73.44%] [G loss: 1.142911]
2/2 [==============================] - 0s 4ms/step
******* 29920 [D loss: 0.610307, acc: 64.84%] [G loss: 1.203183]
2/2 [==============================] - 0s 4ms/step
******* 29921 [D loss: 0.540264, acc: 73.44%] [G loss: 1.316001]
2/2 [==============================] - 0s 4ms/step
******* 29922 [D loss: 0.603281, acc: 68.75%] [G loss: 1.390278]
2/2 [==============================] - 0s 4ms/step
******* 29923 [D loss: 0.695459, acc: 64.06%] [G loss: 1.254979]
2/2 [==============================] - 0s 4ms/step
******* 29924 [D loss: 0.595394, acc: 71.88%] [G loss: 1.107866]
2/2 [==============================] - 0s 5ms/step
******* 29925 [D loss: 0.661634, acc: 63.28%] [G loss: 0.985243]
2/2 [==============================] - 0s 3ms/step
******* 29926 [D loss: 0.596458, acc: 64.84%] [G loss: 1.160980]
2/2 [==============================] - 0s 5ms/step
******* 29927 [D loss: 0.610984, acc: 67.19%] [G loss: 1.224946]
2/2 [==============================] - 0s 8ms/step
******* 29928 [D loss: 0.577828, acc: 71.09%] [G loss: 1.238572]
2/2 [==============================] - 0s 5ms/step
******* 29929 [D loss: 0.589287, acc: 69.53%] [G loss: 1.334740]
2/2 [==============================] - 0s 7ms/step
******* 29930 [D loss: 0.607593, acc: 68.75%] [G loss: 1.156895]
2/2 [==============================] - 0s 8ms/step
******* 29931 [D loss: 0.589881, acc: 66.41%] [G loss: 1.029689]
2/2 [==============================] - 0s 7ms/step
******* 29932 [D loss: 0.569632, acc: 67.19%] [G loss: 1.152985]
2/2 [==============================] - 0s 7ms/step
******* 29933 [D loss: 0.615932, acc: 66.41%] [G loss: 1.269793]
2/2 [==============================] - 0s 6ms/step
******* 29934 [D loss: 0.556390, acc: 72.66%] [G loss: 1.141622]
2/2 [==============================] - 0s 12ms/step
2/2 [==============================] - 0s 12ms/step
******* 29935 [D loss: 0.608344, acc: 67.19%] [G loss: 1.203616]
2/2 [==============================] - 0s 8ms/step
******* 29936 [D loss: 0.656705, acc: 63.28%] [G loss: 1.087211]
2/2 [==============================] - 0s 4ms/step
******* 29937 [D loss: 0.589256, acc: 69.53%] [G loss: 1.240844]
2/2 [==============================] - 0s 6ms/step
******* 29938 [D loss: 0.669647, acc: 57.81%] [G loss: 1.171465]
2/2 [==============================] - 0s 6ms/step
******* 29939 [D loss: 0.617651, acc: 69.53%] [G loss: 1.204986]
2/2 [==============================] - 0s 8ms/step
******* 29940 [D loss: 0.589308, acc: 67.19%] [G loss: 1.119389]
2/2 [==============================] - 0s 7ms/step
******* 29941 [D loss: 0.611637, acc: 66.41%] [G loss: 1.123769]
2/2 [==============================] - 0s 8ms/step
******* 29942 [D loss: 0.608144, acc: 67.19%] [G loss: 1.237388]
2/2 [==============================] - 0s 5ms/step
******* 29943 [D loss: 0.663953, acc: 60.94%] [G loss: 1.246949]
2/2 [==============================] - 0s 5ms/step
******* 29944 [D loss: 0.651479, acc: 64.06%] [G loss: 1.238617]
2/2 [==============================] - 0s 4ms/step
******* 29945 [D loss: 0.654096, acc: 60.94%] [G loss: 1.230141]
2/2 [==============================] - 0s 6ms/step
******* 29946 [D loss: 0.540784, acc: 72.66%] [G loss: 1.229845]
2/2 [==============================] - 0s 4ms/step
******* 29947 [D loss: 0.570492, acc: 74.22%] [G loss: 1.312420]
2/2 [==============================] - 0s 9ms/step
******* 29948 [D loss: 0.595098, acc: 69.53%] [G loss: 1.156778]
2/2 [==============================] - 0s 6ms/step
******* 29949 [D loss: 0.672523, acc: 59.38%] [G loss: 1.016071]
2/2 [==============================] - 0s 6ms/step
******* 29950 [D loss: 0.638359, acc: 67.19%] [G loss: 1.065424]
2/2 [==============================] - 0s 6ms/step
******* 29951 [D loss: 0.551705, acc: 76.56%] [G loss: 1.198018]
2/2 [==============================] - 0s 7ms/step
******* 29952 [D loss: 0.545436, acc: 71.09%] [G loss: 1.248866]
2/2 [==============================] - 0s 9ms/step
******* 29953 [D loss: 0.568265, acc: 71.09%] [G loss: 1.240091]
2/2 [==============================] - 0s 6ms/step
******* 29954 [D loss: 0.642307, acc: 64.06%] [G loss: 1.193786]
2/2 [==============================] - 0s 7ms/step
******* 29955 [D loss: 0.688531, acc: 60.16%] [G loss: 1.334040]
2/2 [==============================] - 0s 4ms/step
******* 29956 [D loss: 0.654302, acc: 63.28%] [G loss: 1.203178]
2/2 [==============================] - 0s 7ms/step
******* 29957 [D loss: 0.718338, acc: 63.28%] [G loss: 1.135725]
2/2 [==============================] - 0s 5ms/step
******* 29958 [D loss: 0.664015, acc: 62.50%] [G loss: 1.160858]
2/2 [==============================] - 0s 6ms/step
******* 29959 [D loss: 0.562412, acc: 67.19%] [G loss: 1.203674]
2/2 [==============================] - 0s 7ms/step
******* 29960 [D loss: 0.572852, acc: 70.31%] [G loss: 1.049615]
2/2 [==============================] - 0s 4ms/step
******* 29961 [D loss: 0.656629, acc: 64.84%] [G loss: 1.124449]
2/2 [==============================] - 0s 6ms/step
******* 29962 [D loss: 0.640985, acc: 68.75%] [G loss: 1.235999]
2/2 [==============================] - 0s 5ms/step
******* 29963 [D loss: 0.593317, acc: 66.41%] [G loss: 1.239542]
2/2 [==============================] - 0s 6ms/step
******* 29964 [D loss: 0.606864, acc: 67.19%] [G loss: 1.275741]
2/2 [==============================] - 0s 8ms/step
******* 29965 [D loss: 0.646582, acc: 62.50%] [G loss: 1.194611]
2/2 [==============================] - 0s 5ms/step
******* 29966 [D loss: 0.587659, acc: 64.06%] [G loss: 1.146992]
2/2 [==============================] - 0s 6ms/step
******* 29967 [D loss: 0.547437, acc: 72.66%] [G loss: 1.137243]
2/2 [==============================] - 0s 5ms/step
******* 29968 [D loss: 0.621019, acc: 64.84%] [G loss: 1.065349]
2/2 [==============================] - 0s 4ms/step
******* 29969 [D loss: 0.629466, acc: 67.19%] [G loss: 1.321414]
2/2 [==============================] - 0s 4ms/step
******* 29970 [D loss: 0.597810, acc: 65.62%] [G loss: 1.307625]
2/2 [==============================] - 0s 4ms/step
******* 29971 [D loss: 0.607508, acc: 72.66%] [G loss: 1.364951]
2/2 [==============================] - 0s 4ms/step
******* 29972 [D loss: 0.622211, acc: 68.75%] [G loss: 1.222455]
2/2 [==============================] - 0s 4ms/step
******* 29973 [D loss: 0.679487, acc: 59.38%] [G loss: 1.047787]
2/2 [==============================] - 0s 5ms/step
******* 29974 [D loss: 0.548078, acc: 71.88%] [G loss: 1.229700]
2/2 [==============================] - 0s 5ms/step
******* 29975 [D loss: 0.679821, acc: 62.50%] [G loss: 1.126660]
2/2 [==============================] - 0s 4ms/step
******* 29976 [D loss: 0.605525, acc: 69.53%] [G loss: 1.222195]
2/2 [==============================] - 0s 4ms/step
******* 29977 [D loss: 0.541331, acc: 72.66%] [G loss: 1.410168]
2/2 [==============================] - 0s 3ms/step
******* 29978 [D loss: 0.647117, acc: 62.50%] [G loss: 1.244983]
2/2 [==============================] - 0s 6ms/step
******* 29979 [D loss: 0.675301, acc: 60.16%] [G loss: 1.098966]
2/2 [==============================] - 0s 5ms/step
******* 29980 [D loss: 0.680653, acc: 61.72%] [G loss: 1.047067]
2/2 [==============================] - 0s 6ms/step
******* 29981 [D loss: 0.610733, acc: 66.41%] [G loss: 1.356553]
2/2 [==============================] - 0s 6ms/step
******* 29982 [D loss: 0.656336, acc: 63.28%] [G loss: 1.248860]
2/2 [==============================] - 0s 4ms/step
******* 29983 [D loss: 0.546962, acc: 76.56%] [G loss: 1.215411]
2/2 [==============================] - 0s 4ms/step
******* 29984 [D loss: 0.709845, acc: 53.91%] [G loss: 1.122472]
2/2 [==============================] - 0s 4ms/step
******* 29985 [D loss: 0.613870, acc: 65.62%] [G loss: 1.108248]
2/2 [==============================] - 0s 5ms/step
******* 29986 [D loss: 0.604478, acc: 67.19%] [G loss: 1.210661]
2/2 [==============================] - 0s 4ms/step
******* 29987 [D loss: 0.568745, acc: 69.53%] [G loss: 1.281371]
2/2 [==============================] - 0s 6ms/step
******* 29988 [D loss: 0.648623, acc: 69.53%] [G loss: 1.187950]
2/2 [==============================] - 0s 6ms/step
******* 29989 [D loss: 0.681824, acc: 64.06%] [G loss: 1.060595]
2/2 [==============================] - 0s 4ms/step
******* 29990 [D loss: 0.568510, acc: 70.31%] [G loss: 1.049071]
2/2 [==============================] - 0s 4ms/step
******* 29991 [D loss: 0.637266, acc: 63.28%] [G loss: 1.062266]
2/2 [==============================] - 0s 4ms/step
******* 29992 [D loss: 0.706634, acc: 57.81%] [G loss: 1.090986]
2/2 [==============================] - 0s 7ms/step
******* 29993 [D loss: 0.611674, acc: 65.62%] [G loss: 1.212140]
2/2 [==============================] - 0s 4ms/step
******* 29994 [D loss: 0.603303, acc: 65.62%] [G loss: 1.340815]
2/2 [==============================] - 0s 5ms/step
******* 29995 [D loss: 0.658634, acc: 64.84%] [G loss: 1.209580]
2/2 [==============================] - 0s 4ms/step
******* 29996 [D loss: 0.635326, acc: 64.06%] [G loss: 1.228579]
2/2 [==============================] - 0s 5ms/step
******* 29997 [D loss: 0.610135, acc: 65.62%] [G loss: 1.113297]
2/2 [==============================] - 0s 5ms/step
******* 29998 [D loss: 0.562490, acc: 76.56%] [G loss: 1.191264]
2/2 [==============================] - 0s 5ms/step
******* 29999 [D loss: 0.607785, acc: 65.62%] [G loss: 1.229325]

8) Making GIF

In [13]:
# Display a single image using the epoch number
# def display_image(epoch_no):
# return PIL.Image.open('generated_images/%.8f.png'.format(epoch_no))

anim_file = 'dcgan.gif'

with imageio.get_writer(anim_file, mode='I') as writer:

filenames = glob.glob('generated_images/*.png')
 filenames = sorted(filenames)
for filename in filenames:
image = imageio.imread(filename)
writer.append_data(image)
image = imageio.imread(filename)
writer.append_data(image)

In [13]: