Keras Sequential 类

2024 年 8 月 28 日 | 阅读 2 分钟

Sequential 类

Keras sequential 类有助于形成一个层集群，该集群线性堆叠到 tf.keras.Model 中。顺序类向此模型提供训练和推理的功能。

示例

>>> # The first layer can optionally receive an `input_shape` argument:  
>>> model = tf.keras.Sequential()
>>> model.add(tf.keras.layers.Dense(8, input_shape=(16,)))
>>> # Next we can perform automatic shape inference:  
>>> model.add(tf.keras.layers.Dense(4))

>>> # This is identical to the following:  
>>> model = tf.keras.Sequential()
>>> model.add(tf.keras.Input(shape=(16,)))
>>> model.add(tf.keras.layers.Dense(8))

>>> # It can be noted that you can also omit the `input_shape` argument.  
>>> # In that case the model doesn't have any weights until the first call  
>>> # to a training/evaluation method (since it isn't yet built):  
>>> model = tf.keras.Sequential()
>>> model.add(tf.keras.layers.Dense(8))
>>> model.add(tf.keras.layers.Dense(4))
>>> # model.weights not created yet  

>>> # Whereas if the input shape is specified, then in that case the model will built up  
>>> # constantly as long as you add layers:  
>>> model = tf.keras.Sequential()
>>> model.add(tf.keras.layers.Dense(8, input_shape=(16,)))
>>> model.add(tf.keras.layers.Dense(4))
>>> len(model.weights)
4

>>> # When you make use of a delayed-build pattern (no input shape specified), you can  
>>> # manually choose to  build your model by calling  
>>> # `build(batch_input_shape)`:  
>>> model = tf.keras.Sequential()
>>> model.add(tf.keras.layers.Dense(8))
>>> model.add(tf.keras.layers.Dense(4))
>>> model.build((None, 16))
>>> len(model.weights)
4

# Note that when you utilize the delayed-build pattern (i.e., no input shape specified),
# then the model will be built in the first time when you call `fit,` `eval,` or `predict,`
# or if you call the model on some input data.
model = tf.keras.Sequential()
model.add(tf.keras.layers.Dense(8))
model.add(tf.keras.layers.Dense(1))
model.compile(optimizer='sgd', loss='mse')
# It helps to build the model for the very first time:
model.fit(x, y, batch_size=32, epochs=10)