Predicting Multiple Time Steps in a Time Series Using LSTM in Keras

I want to predict, for example, the k

following points in a time series using the LSTM in Keras. I am creating a dataset starting at the beginning of a list containing all points, choosing 0:p-1

points as input functions and next k

ie points p:p+k-1

as output functions. I continue this procedure by taking 1:p

both input functions and ... Finally, I get two dataframes X

, input txp

and y

output txk

. So my problem has a many-to-many structure based on here .

X = X.values.reshape(X.shape[0], 1, X.shape[1])
y = y.values.reshape(y.shape[0], 1, y.shape[1]) 

      

        
        
        
      

    

and then the first layer of my network:

model.add(LSTM(neurons, input_shape=(X.shape[1], X.shape[2]), return_sequences=True))

      

        
        
        
      

    

But here the time step is 1. My question is how can I increase the time intervals. Do I have to replicate some lines in X

and y

? Am I doing it right?