Hello. I have a dataset of 60 000 timesteps, which is divided into 2 columns (so 2 features, with a length of 30 000 timesteps for each feature). Question, how do I divide it into smaller chunks of 500 time steps each (each sequences have a distance of 2 timesteps) to feed into my LSTM neural network, as I’ve heard that LSTM networks can have problems with data of very high lengths?
Here’s an illustration of what I’m trying to do: