hope you\'re doing good. I\'m working on an autoencoder architecture where I have 7 variables. I decided to use a sequence of 5 , so my dataset became -> (n_lines, 5 , 7)