Chapter 56
Conclusions
NotebookPython 351 cells
This is a companion notebook for the book Deep Learning with Python, Second Edition. For readability, it only contains runnable code blocks and section titles, and omits everything else in the book: text paragraphs, figures, and pseudocode.
If you want to be able to follow what's going on, I recommend reading the notebook side by side with your copy of the book.
This notebook was generated for TensorFlow 2.6.
Conclusions
Key concepts in review
Various approaches to AI
What makes deep learning special within the field of machine learning
How to think about deep learning
Key enabling technologies
The universal machine-learning workflow
Key network architectures
Densely connected networks
In [0]python · cell 11
python
from tensorflow import keras
from tensorflow.keras import layers
inputs = keras.Input(shape=(num_input_features,))
x = layers.Dense(32, activation="relu")(inputs)
x = layers.Dense(32, activation="relu")(x)
outputs = layers.Dense(1, activation="sigmoid")(x)
model = keras.Model(inputs, outputs)
model.compile(optimizer="rmsprop", loss="binary_crossentropy")In [0]python · cell 12
python
inputs = keras.Input(shape=(num_input_features,))
x = layers.Dense(32, activation="relu")(inputs)
x = layers.Dense(32, activation="relu")(x)
outputs = layers.Dense(num_classes, activation="softmax")(x)
model = keras.Model(inputs, outputs)
model.compile(optimizer="rmsprop", loss="categorical_crossentropy")In [0]python · cell 13
python
inputs = keras.Input(shape=(num_input_features,))
x = layers.Dense(32, activation="relu")(inputs)
x = layers.Dense(32, activation="relu")(x)
outputs = layers.Dense(num_classes, activation="sigmoid")(x)
model = keras.Model(inputs, outputs)
model.compile(optimizer="rmsprop", loss="binary_crossentropy")In [0]python · cell 14
python
inputs = keras.Input(shape=(num_input_features,))
x = layers.Dense(32, activation="relu")(inputs)
x = layers.Dense(32, activation="relu")(x)
outputs layers.Dense(num_values)(x)
model = keras.Model(inputs, outputs)
model.compile(optimizer="rmsprop", loss="mse")Convnets
In [0]python · cell 16
python
inputs = keras.Input(shape=(height, width, channels))
x = layers.SeparableConv2D(32, 3, activation="relu")(inputs)
x = layers.SeparableConv2D(64, 3, activation="relu")(x)
x = layers.MaxPooling2D(2)(x)
x = layers.SeparableConv2D(64, 3, activation="relu")(x)
x = layers.SeparableConv2D(128, 3, activation="relu")(x)
x = layers.MaxPooling2D(2)(x)
x = layers.SeparableConv2D(64, 3, activation="relu")(x)
x = layers.SeparableConv2D(128, 3, activation="relu")(x)
x = layers.GlobalAveragePooling2D()(x)
x = layers.Dense(32, activation="relu")(x)
outputs = layers.Dense(num_classes, activation="softmax")(x)
model = keras.Model(inputs, outputs)
model.compile(optimizer="rmsprop", loss="categorical_crossentropy")RNNs
In [0]python · cell 18
python
inputs = keras.Input(shape=(num_timesteps, num_features))
x = layers.LSTM(32)(inputs)
outputs = layers.Dense(num_classes, activation="sigmoid")(x)
model = keras.Model(inputs, outputs)
model.compile(optimizer="rmsprop", loss="binary_crossentropy")In [0]python · cell 19
python
inputs = keras.Input(shape=(num_timesteps, num_features))
x = layers.LSTM(32, return_sequences=True)(inputs)
x = layers.LSTM(32, return_sequences=True)(x)
x = layers.LSTM(32)(x)
outputs = layers.Dense(num_classes, activation="sigmoid")(x)
model = keras.Model(inputs, outputs)
model.compile(optimizer="rmsprop", loss="binary_crossentropy")In [0]python · cell 21
python
encoder_inputs = keras.Input(shape=(sequence_length,), dtype="int64")
x = PositionalEmbedding(sequence_length, vocab_size, embed_dim)(encoder_inputs)
encoder_outputs = TransformerEncoder(embed_dim, dense_dim, num_heads)(x)
decoder_inputs = keras.Input(shape=(None,), dtype="int64")
x = PositionalEmbedding(sequence_length, vocab_size, embed_dim)(decoder_inputs)
x = TransformerDecoder(embed_dim, dense_dim, num_heads)(x, encoder_outputs)
decoder_outputs = layers.Dense(vocab_size, activation="softmax")(x)
transformer = keras.Model([encoder_inputs, decoder_inputs], decoder_outputs)
transformer.compile(optimizer="rmsprop", loss="categorical_crossentropy")In [0]python · cell 22
python
inputs = keras.Input(shape=(sequence_length,), dtype="int64")
x = PositionalEmbedding(sequence_length, vocab_size, embed_dim)(inputs)
x = TransformerEncoder(embed_dim, dense_dim, num_heads)(x)
x = layers.GlobalMaxPooling1D()(x)
outputs = layers.Dense(1, activation="sigmoid")(x)
model = keras.Model(inputs, outputs)
model.compile(optimizer="rmsprop", loss="binary_crossentropy")