Making outputs of sub-models the inputs of other models LSTM

I’m trying to create a neural network that is composed of different sub-models. Each sub-model have their own inputs, which some of them can be shared between the sub-models.

For example, sub-model 1 could be something like this:

S1 = Input(shape=(1,50),name="Stats")
S2 = LSTM(300,return_sequences = True)(S1)
S3 = Dropout(0.5)(S2)
S4 = LSTM(200,return_sequences = False)(S3)
S5 = Dropout(0.5)(S4)
S6 = Dense(100,kernel_initializer='uniform',activation='relu')(S5)
S7 = Dense(1,kernel_initializer='uniform',activation='relu')(S6)

And then, the output of this sub-model I want it to go as an input of a hidden layer. The thing is that the hidden layer I’m trying to do is also with LSTM and I read in some forums that using keras concatenate can do this.

So, it would go like this:

merge = concatenate([S7,A7,B7,...],name="conca")

At last, I tried to gave this merge object to a new model that would be treated as a “hidden layer”.

It would go like this:

hidden_1_1 = Input(shape=(1,6),name="Hidden_Layer_1")(merge)
hidden_1_2 = LSTM(300,return_sequences = True)(hidden_1_1)
hidden_1_3 = Dropout(0.5)(hidden_1_2)
hidden_1_4 = LSTM(200,return_sequences = False)(hidden_1_3)
hidden_1_5 = Dropout(0.5)(hidden_1_4)
hidden_1_6 = Dense(100,kernel_initializer='uniform',activation='relu')(hidden_1_5)
hidden_1_7 = Dense(1,kernel_initializer='uniform',activation='relu')(hidden_1_6)

And I’m getting this error: TypeError: 'Tensor' object is not callable.

So…how can I make that the outputs of the sub-models to be the inputs of the other models?

In an image, what I’m trying to do is something like this, but the thing is that instead of a neuron, each neuron would be a model.

Thank you in advance!

hidden_1_2 = LSTM(300,return_sequences = True)(merge)
hidden_1_3 = Dropout(0.5)(hidden_1_2)
hidden_1_4 = LSTM(200,return_sequences = False)(hidden_1_3)
hidden_1_5 = Dropout(0.5)(hidden_1_4)
hidden_1_6 = Dense(100,kernel_initializer='uniform',activation='relu')(hidden_1_5)
hidden_1_7 = Dense(1,kernel_initializer='uniform',activation='relu')(hidden_1_6)