ValueError: 层 model_2 需要 2 个输入，但它接收到 1 个输入张量

Question

I have this model I built and it is throwing the error in the title at model3.add(graph) . From what I have read and understood, the second model which is model3 here expects two inputs at model3.add(graph) but it is only received one. Why does it need 2 inputs? Am I overlooking something? Please help?

inputs3 = model.inputs[:2]  # We are getting all layers EXCEPT last 2 layers
layer_output3 = model.get_layer('Encoder-12-FeedForward-Norm')).output  #this is a layer from a pretrained BERT model
removed_layer = RemoveMask()(layer_output3)    #the previous layer contains masks which are not compatible with a CNN layer in Keras
conv_blocks = [] 
filter_sizes = (2,3,4)
for fx in filter_sizes:
    conv_layer = Conv1D(100, kernel_size=fx,
                                    activation= 'softsign'), data_format='channels_first')(removed_layer)  
    maxpool_layer = MaxPooling1D(pool_size=2)(conv_layer)
    flat_layer = Flatten()(maxpool_layer)
    conv_blocks.append(flat_layer)
conc_layer = concatenate(conv_blocks, axis=1)
restored_layer = RestoreMask()([conc_layer, layer_output3])
graph = Model(input=inputs3, outputs=restored_layer)

model3 = Sequential()
model3.add(graph)
model3.add(Dropout(0.1))
model3.add(Dense(3, activation='softmax'))
model3.compile(loss='sparse_categorical_crossentropy', optimizer='adam', metrics=['accuracy'])
model3.summary()

Answer 1

You are combining a Functional Model(graph) with Sequential Model(model3). Either convert both the model to be Functional Model(like graph) OR convert both the model to be Sequential Model(like model3).

You can find solution for converting Functional model to Sequential Model and vice versa here .