Python 如何将通过model（）建立的神经网络转换为sequential（）？

我使用TensorFlow 2.0构建了一个完全连接的神经网络，一个去噪自动编码器。网络结构为128-64-32-64-128（神经元数），经过五层网络处理，输入1000个数据，输出1000个数据

我的代码如下（Model（））：

以及相应的model.summary（）如下：

Model: "model_1"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
input_1 (InputLayer)         (None, 1000)              0         
_________________________________________________________________
dense_1 (Dense)              (None, 128)               128128    
_________________________________________________________________
dense_2 (Dense)              (None, 64)                8256      
_________________________________________________________________
dense_3 (Dense)              (None, 32)                2080      
_________________________________________________________________
dense_4 (Dense)              (None, 64)                2112      
_________________________________________________________________
dense_5 (Dense)              (None, 128)               8320      
_________________________________________________________________
dense_6 (Dense)              (None, 1000)              129000    
=================================================================
Total params: 277,896
Trainable params: 277,896
Non-trainable params: 0

input_size_1000 = 1000
hidden_size_128 = 128
hidden_size_64 = 64
code_size_32 = 32
output_size_1000 = 1000

# using sequential()
model = tf.keras.models.Sequential() 
# model.add(tf.keras.layers.Input(input_size_1000,))
model.add(tf.keras.layers.Dense(hidden_size_128,input_dim=1000,activation=tf.nn.relu))
model.add(tf.keras.layers.Dense(hidden_size_64,activation=tf.nn.relu))
model.add(tf.keras.layers.Dense(code_size_32,activation=tf.nn.relu))
model.add(tf.keras.layers.Dense(hidden_size_64,activation=tf.nn.relu))
model.add(tf.keras.layers.Dense(hidden_size_128,activation=tf.nn.relu))
model.add(tf.keras.layers.Dense(output_size_1000,activation=tf.nn.sigmoid))

model.compile(optimizer='adam', loss='mse', metrics=["mse"])

Model: "sequential"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
dense (Dense)                (None, 128)               128128    
_________________________________________________________________
dense_1 (Dense)              (None, 64)                8256      
_________________________________________________________________
dense_2 (Dense)              (None, 32)                2080      
_________________________________________________________________
dense_3 (Dense)              (None, 64)                2112      
_________________________________________________________________
dense_4 (Dense)              (None, 128)               8320      
_________________________________________________________________
dense_5 (Dense)              (None, 1000)              129000    
=================================================================
Total params: 277,896
Trainable params: 277,896
Non-trainable params: 0
_________________________________________________________________
None

我想用sequential（）模式重写这个神经网络（因为我需要的Python库必须使用sequential（）。我的代码如下：

Model: "model_1"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
input_1 (InputLayer)         (None, 1000)              0         
_________________________________________________________________
dense_1 (Dense)              (None, 128)               128128    
_________________________________________________________________
dense_2 (Dense)              (None, 64)                8256      
_________________________________________________________________
dense_3 (Dense)              (None, 32)                2080      
_________________________________________________________________
dense_4 (Dense)              (None, 64)                2112      
_________________________________________________________________
dense_5 (Dense)              (None, 128)               8320      
_________________________________________________________________
dense_6 (Dense)              (None, 1000)              129000    
=================================================================
Total params: 277,896
Trainable params: 277,896
Non-trainable params: 0

input_size_1000 = 1000
hidden_size_128 = 128
hidden_size_64 = 64
code_size_32 = 32
output_size_1000 = 1000

# using sequential()
model = tf.keras.models.Sequential() 
# model.add(tf.keras.layers.Input(input_size_1000,))
model.add(tf.keras.layers.Dense(hidden_size_128,input_dim=1000,activation=tf.nn.relu))
model.add(tf.keras.layers.Dense(hidden_size_64,activation=tf.nn.relu))
model.add(tf.keras.layers.Dense(code_size_32,activation=tf.nn.relu))
model.add(tf.keras.layers.Dense(hidden_size_64,activation=tf.nn.relu))
model.add(tf.keras.layers.Dense(hidden_size_128,activation=tf.nn.relu))
model.add(tf.keras.layers.Dense(output_size_1000,activation=tf.nn.sigmoid))

model.compile(optimizer='adam', loss='mse', metrics=["mse"])

Model: "sequential"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
dense (Dense)                (None, 128)               128128    
_________________________________________________________________
dense_1 (Dense)              (None, 64)                8256      
_________________________________________________________________
dense_2 (Dense)              (None, 32)                2080      
_________________________________________________________________
dense_3 (Dense)              (None, 64)                2112      
_________________________________________________________________
dense_4 (Dense)              (None, 128)               8320      
_________________________________________________________________
dense_5 (Dense)              (None, 1000)              129000    
=================================================================
Total params: 277,896
Trainable params: 277,896
Non-trainable params: 0
_________________________________________________________________
None

以及相应的model.summary（）如下：

Model: "model_1"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
input_1 (InputLayer)         (None, 1000)              0         
_________________________________________________________________
dense_1 (Dense)              (None, 128)               128128    
_________________________________________________________________
dense_2 (Dense)              (None, 64)                8256      
_________________________________________________________________
dense_3 (Dense)              (None, 32)                2080      
_________________________________________________________________
dense_4 (Dense)              (None, 64)                2112      
_________________________________________________________________
dense_5 (Dense)              (None, 128)               8320      
_________________________________________________________________
dense_6 (Dense)              (None, 1000)              129000    
=================================================================
Total params: 277,896
Trainable params: 277,896
Non-trainable params: 0

input_size_1000 = 1000
hidden_size_128 = 128
hidden_size_64 = 64
code_size_32 = 32
output_size_1000 = 1000

# using sequential()
model = tf.keras.models.Sequential() 
# model.add(tf.keras.layers.Input(input_size_1000,))
model.add(tf.keras.layers.Dense(hidden_size_128,input_dim=1000,activation=tf.nn.relu))
model.add(tf.keras.layers.Dense(hidden_size_64,activation=tf.nn.relu))
model.add(tf.keras.layers.Dense(code_size_32,activation=tf.nn.relu))
model.add(tf.keras.layers.Dense(hidden_size_64,activation=tf.nn.relu))
model.add(tf.keras.layers.Dense(hidden_size_128,activation=tf.nn.relu))
model.add(tf.keras.layers.Dense(output_size_1000,activation=tf.nn.sigmoid))

model.compile(optimizer='adam', loss='mse', metrics=["mse"])

Model: "sequential"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
dense (Dense)                (None, 128)               128128    
_________________________________________________________________
dense_1 (Dense)              (None, 64)                8256      
_________________________________________________________________
dense_2 (Dense)              (None, 32)                2080      
_________________________________________________________________
dense_3 (Dense)              (None, 64)                2112      
_________________________________________________________________
dense_4 (Dense)              (None, 128)               8320      
_________________________________________________________________
dense_5 (Dense)              (None, 1000)              129000    
=================================================================
Total params: 277,896
Trainable params: 277,896
Non-trainable params: 0
_________________________________________________________________
None

我想知道为什么在重写的代码中没有输入层（我添加了input_dim=1000）。首先我添加了

model.add（tf.keras.layers.Input（Input\u size\u 1000，）

---

，但它不起作用。然后我命令这行，加上inpu\u dim=1000，仍然没有输入层。顺便问一下，我的重写有什么问题吗？因为我将要使用的库没有输入层，但它提供了一个展平层，如果删除输入层会有任何影响吗？如果我使用展平层而不是输入层（我的数据是一维列表）（因为MNIST图片展平后库中的样本很密集），会怎么样？谢谢

必须指定密集层的输入尺寸。在您的顺序模型中，它是

input\u dim

。使用

print（model.inputs）

可以看到输入张量

---

如果使用展平层，则必须指定展平层的

input_shape

，以便下一层（密集）的输入维度清晰，例如：

tf.keras.layers.flatte（input_shape=[24,24]）

必须指定密集层的输入维度。在您的顺序模型中，它是

input\u dim

。使用

print（model.inputs）

可以看到输入张量

---

如果使用展平层，则必须指定展平层的

input\u shape

，以便下一层（密集）的输入维度是清晰的，例如：

tf.keras.layers.flatte（input\u shape=[24,24]）

看起来输入层被注释掉了，这是您代码中的情况吗？对不起，我没有解释清楚。我注释的行对代码的model.summary（）没有任何影响，因此我尝试再次添加输入，但仍然无效。我还在问题中添加了此脚注。看起来输入层被注释掉了，您的代码中是这样吗？很抱歉，我没有解释清楚。我的注释行对代码的model.summary（）没有任何影响，因此我再次尝试添加输入，但仍然无效。我还在问题中添加了此脚注。