import tensorflow.keras as tfk # Define architecture model = tfk.Sequential() model.add(tfk.layers.Dense(units=16, activation='relu', input_shape=(100,))) model.add(tfk.layers.Dense(units=8, activation='relu')) model.add(tfk.layers.Dense(units=1, activation='sigmoid')) # Compile the model model.compile(optimizer='rmsprop', loss='binary_crossentropy')In this example, we have added 3 layers to our `Sequential` model. The first layer is a `Dense` layer with 16 hidden units and ReLU activation function. The second layer is another `Dense` layer with 8 hidden units and ReLU activation function. The last layer is a `Dense` layer with 1 unit and sigmoid activation function. We have also specified the input shape of our data for the first layer using the `input_shape` argument. Finally, we compile the model using RSMprop optimizer and binary cross-entropy loss function. In summary, `tensorflow.keras` is a package library for building and training deep learning models. `Sequential` model is a linear stack of layers where we can add multiple layers to it using `.add()` method. We can compile our model specifying optimizer and loss function using `compile()` method.