1+ import tensorflow as tf # deep learning library. Tensors are just multi-dimensional arrays
2+ 3+ mnist = tf .keras .datasets .mnist # mnist is a dataset of 28x28 images of handwritten digits and their labels
4+ (x_train , y_train ),(x_test , y_test ) = mnist .load_data () # unpacks images to x_train/x_test and labels to y_train/y_test
5+ 6+ x_train = tf .keras .utils .normalize (x_train , axis = 1 ) # scales data between 0 and 1
7+ x_test = tf .keras .utils .normalize (x_test , axis = 1 ) # scales data between 0 and 1
8+ 9+ model = tf .keras .models .Sequential () # a basic feed-forward model
10+ model .add (tf .keras .layers .Flatten ()) # takes our 28x28 and makes it 1x784
11+ model .add (tf .keras .layers .Dense (128 , activation = tf .nn .relu )) # a simple fully-connected layer, 128 units, relu activation
12+ model .add (tf .keras .layers .Dense (128 , activation = tf .nn .relu )) # a simple fully-connected layer, 128 units, relu activation
13+ model .add (tf .keras .layers .Dense (10 , activation = tf .nn .softmax )) # our output layer. 10 units for 10 classes. Softmax for probability distribution
14+ 15+ model .compile (optimizer = 'adam' , # Good default optimizer to start with
16+ loss = 'sparse_categorical_crossentropy' , # how will we calculate our "error." Neural network aims to minimize loss.
17+ metrics = ['accuracy' ]) # what to track
18+ 19+ model .fit (x_train , y_train , epochs = 3 ) # train the model
20+ 21+ val_loss , val_acc = model .evaluate (x_test , y_test ) # evaluate the out of sample data with model
22+ print (val_loss ) # model's loss (error)
23+ print (val_acc ) # model's accuracy
0 commit comments