Tensorflow 理解Keras模型体系结构（张量指数）

此脚本使用

产生以下输出

____________________________________________________________________________________________________
Layer (type)                     Output Shape          Param #     Connected to                     
====================================================================================================
A_input (InputLayer)             (None, 100)           0                                            
____________________________________________________________________________________________________
B_dense (Dense)                  (None, 20)            2020        A_input[0][0]                    
____________________________________________________________________________________________________
C_dense_shared (Dense)           (None, 20)            420         B_dense[0][0]                    
                                                                   B_dense[0][0]                    
____________________________________________________________________________________________________
D_concat (Concatenate)           (None, 40)            0           C_dense_shared[0][0]             
                                                                   C_dense_shared[1][0]             
====================================================================================================

我的问题涉及连接到列的

的内容。
我明白

在这种情况下，
C\u densite\u shared
有两个节点，
D\u concat
连接到这两个节点（
C\u densite\u shared[0][0]
和
C\u densite\u shared[1][0]
）。所以第一个索引（node_index

）对我来说很清楚。但第二个指数意味着什么？从中我了解到这是张量索引：

layer_name[node_index][tensor_index]

但是，张量指数是什么意思？在什么情况下，它可以有一个不同于

0

的值？

我认为

节点

类的属性非常清楚：

    tensor_indices: a list of integers,
        the same length as `inbound_layers`.
        `tensor_indices[i]` is the index of `input_tensors[i]` within the
        output of the inbound layer
        (necessary since each inbound layer might
        have multiple tensor outputs, with each one being
        independently manipulable).

如果一个层有多个输出张量，

tensor\u index

将非零。这不同于多个“数据流”（例如层共享）的情况，其中层具有多个出站节点。例如，如果给定

return\u state=True

，则

LSTM

层将返回3个张量：

上一个时间步的隐藏状态，或所有隐藏状态（如果

return\u sequences=True

上一时间步的隐藏状态

最后一个时间步的存储单元

作为另一个示例，特征转换可以实现为

Lambda

层：

def generate_powers(x):
    return [x, K.sqrt(x), K.square(x)]

model_input = Input(shape=(10,))
powers = Lambda(generate_powers)(model_input)
x = Concatenate()(powers)
x = Dense(10, activation='relu')(x)
x = Dense(1, activation='sigmoid')(x)
model = Model(model_input, x)

从

model.summary（）

，您可以看到

concatenate_5

连接到

lambda_7[0][0]

、

lambda_7[0][1]

和

lambda_7[0][2]

：

____________________________________________________________________________________________________
Layer (type)                     Output Shape          Param #     Connected to                     
====================================================================================================
input_7 (InputLayer)             (None, 10)            0                                            
____________________________________________________________________________________________________
lambda_7 (Lambda)                [(None, 10), (None, 1 0           input_7[0][0]                    
____________________________________________________________________________________________________
concatenate_5 (Concatenate)      (None, 30)            0           lambda_7[0][0]                   
                                                                   lambda_7[0][1]                   
                                                                   lambda_7[0][2]                   
____________________________________________________________________________________________________
dense_8 (Dense)                  (None, 10)            310         concatenate_5[0][0]              
____________________________________________________________________________________________________
dense_9 (Dense)                  (None, 1)             11          dense_8[0][0]                    
====================================================================================================
Total params: 321 
Trainable params: 321 
Non-trainable params: 0
____________________________________________________________________________________________________

---

谢谢。这是有道理的。我还没有使用Lambda或LSTM。我会在14小时内把赏金判给你，因为这是不可能提前完成的。

____________________________________________________________________________________________________
Layer (type)                     Output Shape          Param #     Connected to                     
====================================================================================================
input_7 (InputLayer)             (None, 10)            0                                            
____________________________________________________________________________________________________
lambda_7 (Lambda)                [(None, 10), (None, 1 0           input_7[0][0]                    
____________________________________________________________________________________________________
concatenate_5 (Concatenate)      (None, 30)            0           lambda_7[0][0]                   
                                                                   lambda_7[0][1]                   
                                                                   lambda_7[0][2]                   
____________________________________________________________________________________________________
dense_8 (Dense)                  (None, 10)            310         concatenate_5[0][0]              
____________________________________________________________________________________________________
dense_9 (Dense)                  (None, 1)             11          dense_8[0][0]                    
====================================================================================================
Total params: 321 
Trainable params: 321 
Non-trainable params: 0
____________________________________________________________________________________________________