In [1]: import numpy as np

import matplotlib.pyplot as plt

import tensorflow as tf
from tensorflow.keras.layers import Input, Dense
from tensorflow.keras.models import Model
from tensorflow.keras.datasets import mnist

# Load MNIST dataset

(x_train, _), (x_test, _) = mnist.load_data()

# Normalize the data to values between 0 and 1

x_train = x_train.astype('float32') / 255.
x_test = x_test.astype('float32') / 255.

# Flatten the 28x28 images into vectors of size 784

x_train = x_train.reshape((len(x_train), np.prod(x_train.shape[1:])))
x_test = x_test.reshape((len(x_test), np.prod(x_test.shape[1:])))

# Define the size of the encoded representation

encoding_dim = 32

# Input placeholder
input_img = Input(shape=(784,))

# Encoded representation of the input

encoded = Dense(encoding_dim, activation='relu')(input_img)

# Reconstruction of the input

decoded = Dense(784, activation='sigmoid')(encoded)

# Model that maps an input to its reconstruction

autoencoder = Model(input_img, decoded)

# Encoder model
encoder = Model(input_img, encoded)

# Decoder model
encoded_input = Input(shape=(encoding_dim,))
decoder_layer = autoencoder.layers[-1]
decoder = Model(encoded_input, decoder_layer(encoded_input))

# Compile the autoencoder

autoencoder.compile(optimizer='adam', loss='binary_crossentropy')

# Train the autoencoder

autoencoder.fit(x_train, x_train,
epochs=50,
batch_size=256,
shuffle=True,
validation_data=(x_test, x_test))

# Encode and decode some digits

encoded_imgs = encoder.predict(x_test)
decoded_imgs = decoder.predict(encoded_imgs)

# Visualize the original and reconstructed images

n = 10
plt.figure(figsize=(20, 4))
for i in range(n):
# Display original
ax = plt.subplot(2, n, i + 1)
plt.imshow(x_test[i].reshape(28, 28))
plt.gray()
ax.get_xaxis().set_visible(False)
ax.get_yaxis().set_visible(False)

# Display reconstruction
ax = plt.subplot(2, n, i + 1 + n)
plt.imshow(decoded_imgs[i].reshape(28, 28))
plt.gray()
ax.get_xaxis().set_visible(False)
ax.get_yaxis().set_visible(False)
plt.show()

Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/mnist.npz

11490434/11490434 ━━━━━━━━━━━━━━━━━━━━ 8s 1us/step
Epoch 1/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 5s 11ms/step - loss: 0.3844 - val_loss: 0.1883
Epoch 2/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 9ms/step - loss: 0.1795 - val_loss: 0.1534
Epoch 3/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.1490 - val_loss: 0.1325
Epoch 4/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.1304 - val_loss: 0.1203
Epoch 5/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.1196 - val_loss: 0.1115
Epoch 6/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 3s 10ms/step - loss: 0.1116 - val_loss: 0.1060
Epoch 7/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.1060 - val_loss: 0.1017
Epoch 8/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 3s 10ms/step - loss: 0.1023 - val_loss: 0.0986
Epoch 9/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0993 - val_loss: 0.0966
Epoch 10/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0973 - val_loss: 0.0951
Epoch 11/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 3s 11ms/step - loss: 0.0960 - val_loss: 0.0941
Epoch 12/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 3s 10ms/step - loss: 0.0951 - val_loss: 0.0934
Epoch 13/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 3s 10ms/step - loss: 0.0946 - val_loss: 0.0931
Epoch 14/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0942 - val_loss: 0.0927
Epoch 15/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0938 - val_loss: 0.0925
Epoch 16/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 3s 10ms/step - loss: 0.0937 - val_loss: 0.0924
Epoch 17/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 3s 10ms/step - loss: 0.0935 - val_loss: 0.0922
Epoch 18/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0935 - val_loss: 0.0922
Epoch 19/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0934 - val_loss: 0.0920
Epoch 20/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0933 - val_loss: 0.0920
Epoch 21/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0931 - val_loss: 0.0919
Epoch 22/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0930 - val_loss: 0.0919
Epoch 23/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 3s 10ms/step - loss: 0.0930 - val_loss: 0.0918
Epoch 24/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0931 - val_loss: 0.0918
Epoch 25/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0929 - val_loss: 0.0919
Epoch 26/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0929 - val_loss: 0.0917
Epoch 27/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0929 - val_loss: 0.0917
Epoch 28/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0930 - val_loss: 0.0917
Epoch 29/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0927 - val_loss: 0.0917
Epoch 30/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0929 - val_loss: 0.0917
Epoch 31/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0926 - val_loss: 0.0916
Epoch 32/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0927 - val_loss: 0.0916
Epoch 33/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 3s 10ms/step - loss: 0.0928 - val_loss: 0.0916
Epoch 34/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0929 - val_loss: 0.0916
Epoch 35/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0928 - val_loss: 0.0916
Epoch 36/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0927 - val_loss: 0.0916
Epoch 37/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0925 - val_loss: 0.0916
Epoch 38/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0928 - val_loss: 0.0916
Epoch 39/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0927 - val_loss: 0.0915
Epoch 40/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0927 - val_loss: 0.0914
Epoch 41/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0926 - val_loss: 0.0915
Epoch 42/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0927 - val_loss: 0.0915
Epoch 43/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0926 - val_loss: 0.0915
Epoch 44/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0927 - val_loss: 0.0915
Epoch 45/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 3s 10ms/step - loss: 0.0925 - val_loss: 0.0914
Epoch 46/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0925 - val_loss: 0.0916
Epoch 47/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 3s 11ms/step - loss: 0.0927 - val_loss: 0.0915
Epoch 48/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0927 - val_loss: 0.0915
Epoch 49/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0925 - val_loss: 0.0914
Epoch 50/50
235/235 ━━━━━━━━━━━━━━━━━━━━ 2s 10ms/step - loss: 0.0927 - val_loss: 0.0915
313/313 ━━━━━━━━━━━━━━━━━━━━ 1s 2ms/step
313/313 ━━━━━━━━━━━━━━━━━━━━ 1s 2ms/step
In [ ]: