python-tensorflowHow can I use Python and TensorFlow to create a MNIST example?

To use Python and TensorFlow to create a MNIST example, first install the necessary libraries and packages:

pip install tensorflow
pip install numpy
pip install matplotlib

Then, import the necessary libraries and packages into your Python script:

import tensorflow as tf
import numpy as np
import matplotlib.pyplot as plt

Next, load the MNIST data set and normalize it:

mnist = tf.keras.datasets.mnist
(x_train, y_train), (x_test, y_test) = mnist.load_data()
x_train = tf.keras.utils.normalize(x_train, axis=1)
x_test = tf.keras.utils.normalize(x_test, axis=1)

After that, create the model architecture, compile it, and fit it to the training data:

model = tf.keras.models.Sequential()
model.add(tf.keras.layers.Flatten())
model.add(tf.keras.layers.Dense(128, activation=tf.nn.relu))
model.add(tf.keras.layers.Dense(128, activation=tf.nn.relu))
model.add(tf.keras.layers.Dense(10, activation=tf.nn.softmax))

model.compile(optimizer='adam',
              loss='sparse_categorical_crossentropy',
              metrics=['accuracy'])

model.fit(x_train, y_train, epochs=3)

Finally, evaluate the model with the test data and print the results:

val_loss, val_acc = model.evaluate(x_test, y_test)
print(val_loss, val_acc)

0.09817703712594568 0.9714

Code explanation

• pip install tensorflow: Installs the TensorFlow library.
• pip install numpy: Installs the NumPy library.
• pip install matplotlib: Installs the Matplotlib library.
• import tensorflow as tf: Imports the TensorFlow library into the script.
• import numpy as np: Imports the NumPy library into the script.
• import matplotlib.pyplot as plt: Imports the Matplotlib library into the script.
• mnist = tf.keras.datasets.mnist: Loads the MNIST data set.
• (x_train, y_train), (x_test, y_test) = mnist.load_data(): Splits the MNIST data set into training and test sets.
• x_train = tf.keras.utils.normalize(x_train, axis=1): Normalizes the training data.
• x_test = tf.keras.utils.normalize(x_test, axis=1): Normalizes the test data.
• model = tf.keras.models.Sequential(): Creates a sequential model.
• model.add(tf.keras.layers.Flatten()): Adds a flatten layer to the model.
• model.add(tf.keras.layers.Dense(128, activation=tf.nn.relu)): Adds a dense layer with 128 neurons and ReLU activation to the model.
• model.add(tf.keras.layers.Dense(128, activation=tf.nn.relu)): Adds a dense layer with 128 neurons and ReLU activation to the model.
• model.add(tf.keras.layers.Dense(10, activation=tf.nn.softmax)): Adds a dense output layer with 10 neurons and softmax activation to the model.
• model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['accuracy']): Compiles the model with the Adam optimizer, sparse categorical cross-entropy loss, and accuracy metric.
• model.fit(x_train, y_train, epochs=3): Fits the model to the training data for 3 epochs.
• val_loss, val_acc = model.evaluate(x_test, y_test): Evaluates the model with the test data.
• print(val_loss, val_acc): Prints the loss and accuracy of the model.

Helpful links

• TensorFlow Documentation
• NumPy Documentation
• Matplotlib Documentation
• MNIST Tutorial

Edit this code on GitHub