更改Keras层输出形状

Question

我正在开发一个程序，该程序用一个无方向的Conv2D层（跨距=1）替换跨距的Conv2D层（跨距=2），然后在激活层之后添加一个AveragePooling2D层（跨距=2）。换句话说，AveragePooling2D层将减少输出维度，而不是让Conv2D层减少输出维度。

我使用本文中描述的方法将Conv2D层替换为非结构化版本，并在激活后插入averagepoolig2d层。替换Conv2D层效果很好。问题是，激活层的新输出将具有与以前不同的形状。这篇文章解释说，“多重”输出形状是因为现在有两个输出节点连接到激活层，一个是跨步卷积中的旧形状，另一个是无边卷积中的新形状。

下面是说明问题的代码：

# Origninal model with strided convolution
inp = tf.keras.layers.Input((32,32,3))
x = tf.keras.layers.Conv2D(filters=1,kernel_size=3,strides=2, padding='same')(inp)
x = tf.nn.relu(x)
x = tf.keras.layers.GlobalAveragePooling2D()(x)
x = tf.keras.layers.Dense(units=10, activation='linear')(x)
outp = x
model = tf.keras.models.Model(inp,outp)

# Clone original model in order to modify the clone
model_clone = tf.keras.models.clone_model(model)

# Customize the cloned model - this function is where the convolution
# is replaced and the average pooling layer is added
model_custom = MyCustomFunction(model_clone)

原始模型概述：

Model: "model_2"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
input_2 (InputLayer)         [(None, 32, 32, 3)]       0         
_________________________________________________________________
conv2d_2 (Conv2D)            (None, 16, 16, 1)         28        
_________________________________________________________________
tf.nn.relu_1 (TFOpLambda)    (None, 16, 16, 1)         0         
_________________________________________________________________
global_average_pooling2d_1 ( (None, 1)                 0         
_________________________________________________________________
dense_1 (Dense)              (None, 10)                20        
=================================================================
Total params: 48
Trainable params: 48
Non-trainable params: 0
_________________________________________________________________

model\u自定义摘要：

Model: "model_6"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
input_3 (InputLayer)         [(None, 32, 32, 3)]       0         
_________________________________________________________________
conv2d_6 (Conv2D)            (None, 32, 32, 1)         28        
_________________________________________________________________
tf.nn.relu_2 (TFOpLambda)    multiple                  0         
_________________________________________________________________
average_pooling2d (AveragePo (None, 16, 16, 1)         0         
_________________________________________________________________
global_average_pooling2d_2 ( (None, 1)                 0         
_________________________________________________________________
dense_2 (Dense)              (None, 10)                20        
=================================================================
Total params: 48
Trainable params: 48
Non-trainable params: 0
_________________________________________________________________

当我尝试训练新model_custom时，我收到这个错误：

InvalidArgumentError:  logits and labels must have the same first dimension, got logits shape [32768,1] and labels shape [32]

我认为这是因为它仍在查看激活层的旧输出形状（None，16，16，1），而不是新的输出形状（None，32，32，1）。

以下是我用于培训的代码，仅供参考：

model_custom.compile(optimizer='adam',
              loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),
              metrics=['accuracy'])

history = model_custom.fit(train_images, train_labels, epochs=2,
                    validation_data=(test_images, test_labels))

当我将相同的代码应用于原始模型时，它可以正常工作。

任何帮助将不胜感激！

Answer 1

否，错误InvalidArgumentError:logits和labels必须具有相同的第一个维度，got logits shape[32768,1]和labels shape[32]与模型架构无关，而是因为批次大小维度。请参阅此。

您输入了形状[32768,1]的模型数据，而您应用的损失需要[32]的形状，以单独进行热编码。