Skip to main content
 首页 » 编程设计

python之Keras:训练时出现值错误

2024年12月31日9mfrbuaa

我有这样的深度学习架构:

main_input_1 = Input(shape=(50,1), dtype='float32', name='main_input_1') 
main_input_2 = Input(shape=(50,1), dtype='float32', name='main_input_2') 
lstm_out=LSTM(32,activation='tanh',recurrent_activation='sigmoid',return_sequences=True) 
mean_pooling=AveragePooling1D(pool_size=2,strides=2,padding='valid') 
 
lstm_out_1=lstm_out(main_input_1) 
lstm_out_2=lstm_out(main_input_2) 
mean_pooling_1=mean_pooling(lstm_out_1) 
mean_pooling_2=mean_pooling(lstm_out_2) 
 
concatenate_layer=Concatenate()([mean_pooling_1,mean_pooling_2]) 
 
logistic_regression_output=Dense(1,activation='softmax',name='main_output')(concatenate_layer) 
 
 
model = Model(inputs=[main_input_1, main_input_2], outputs=[main_output]) 

我有平行运行的层(两侧具有相同的结构)。我正在使用 Keras 的功能性 api 来做同样的事情。但是在运行它时出现以下错误:

Traceback (most recent call last): 
  File "Main_Architecture.py", line 38, in <module> 
    model = Model(inputs=[main_input_1, main_input_2], outputs=[main_output]) 
  File "/home/tpradhan/anaconda3/lib/python3.6/site-packages/keras/legacy/interfaces.py", line 91, in wrapper 
    return func(*args, **kwargs) 
  File "/home/tpradhan/anaconda3/lib/python3.6/site-packages/keras/engine/network.py", line 91, in __init__ 
    self._init_graph_network(*args, **kwargs) 
  File "/home/tpradhan/anaconda3/lib/python3.6/site-packages/keras/engine/network.py", line 192, in _init_graph_network 
    'Found: ' + str(x)) 
ValueError: Output tensors to a Model must be the output of a TensorFlow `Layer` (thus holding past layer metadata). Found: [0.00000000e+00 5.09370000e-06 8.19930500e-04 ... 9.61476653e-02 
 3.62692160e-03 3.62692160e-03] 

我已经阅读了类似错误的问题,但没有一个对我有用。请帮助我解决这个问题。

请您参考如下方法:

您正在为输出参数传递图层名称。您应该传递层(换句话说,参数值应该是引用输出层的变量)。

model = Model(inputs=[main_input_1, main_input_2], outputs=[logistic_regression_output])