密集层没有给出预期的输出形状

Posted 2023-02-16

【中文标题】密集层没有给出预期的输出形状

【英文标题】：Dense layer does not give expected Output shape

【发布时间】：2022-01-12 20:35:33

【问题描述】：

我正在尝试复制模型架构。在原始模型架构中，应用最后一个 Dense 层后，输出形状为 (None, 3)，参数为 300。如图

 _________________________________________________________________
 Layer (type)                Output Shape              Param #   
 =================================================================

 dense_Dense1 (Dense)        (None, 100)               128100    
                                                             
 dense_Dense2 (Dense)        (None, 3)                 300       

但是当我应用密集输出形状时，我得到的是 (None, 3) 和 303 个参数。如下图

 _________________________________________________________________
 Layer (type)                Output Shape              Param #   
 =================================================================

 dense_35 (Dense)                (None, 100)          128100  

 dense_36 (Dense)                (None, 3)            303  

这是我为此部分编写的代码：

x = GlobalAveragePooling2D()(x)
x = Dense(100, activation="relu")(x)
prediction = Dense(3, activation='softmax')(x)

【参考方案1】：

您尝试复制的架构是否可能不使用偏见？试试not using bias：

Dense(3, activation='softmax', use_bias=False)

Model: "sequential_5"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
global_average_pooling2d_3 ( (None, 8)                 0         
_________________________________________________________________
dense_9 (Dense)              (None, 100)               900       
_________________________________________________________________
dense_10 (Dense)             (None, 3)                 300       
=================================================================
Total params: 1,200
Trainable params: 1,200
Non-trainable params: 0
_________________________________________________________________