INNER CODE UNIT · Python

test_images

mrdbourke/tensorflow-deep-learning · extras/image_classification_test.py:25

train_images, test_images = train_images / 255.0, test_images / 255.0 

# Check shape of input data
# print(train_images.shape)
# print(train_labels.shape)

# Build model
model = tf.keras.Sequential([
    # Reshape inputs to be compatible with Conv2D layer
    layers.Lambda(lambda x: tf.expand_dims(x, axis=-1)),
    layers.Conv2D(32, 3, activation="relu"),
    layers.MaxPool2D(),
    layers.Conv2D(32, 3, activation="relu"),
    layers.MaxPool2D(),
    layers.Conv2D(32, 3, activation="relu"),
    layers.Flatten(), # flatten outputs of final Conv layer to be suited for final Dense layer
    layers.Dense(10, activation="softmax")
])

View source record →