Nrthugu

i am new in deep learning and i want a help for my thesis!
I want todo supervised deep learning for text. it is a classification issue. I have a trainset of author for gender and the same trainset for sexual preference. I want to build a rnn model to predict if an author is female or male and gay or lesbian.
I have to do this with multi tasking techique. This is my code:

   left_branch = Sequential()

   left_branch.add(LSTM(32, input_dim=100))







   right_branch = Sequential()

   right_branch.add(LSTM(32, input_dim=100))



   f=add([left_branch.output, right_branch.output])



   final_model = Sequential()



   final_model.add(Dense(2,activation='sigmoid'))

   model_output = final_model(f)

   model = Model([left_branch.input, right_branch.input], model_output)

   final_model.compile(loss='categorical_crossentropy',

          optimizer='adam',

          metrics=['accuracy'])



   history = final_model.fit([x_train,x_train2],[y_train,y_train2],

                batch_size=batch_size,

                epochs=30,

                verbose=2

               )





  results=final_model.evaluate([x_test,x_test2],[y_test,y_test2],verbose=2)

  print(results)

but with this code i have the following error:

str(data)[:200] + '...')
ValueError: Error when checking model input: the list of Numpy arrays that you are passing to your model is not the size the model expected. Expected to see 1 array(s), but instead got the following list of 2 arrays: [array([[0. , 0. , 0. , ..., 0.00045344, 0.00211607,
0.00634821],
[0. , 0. , 0. , ..., 0.00049548, 0.0094142 ,