I have a dataset with undefined dimensionality (None, 2). And I need to feed this dataset into a recurrent neural network.
Briefly about the dataset: it is data from several measurements from one instrument, which have a total influence, but each time this number of measurements is different.
My idea is to zero-fill the data so that the dimension of the maximal vector is instead of None (i.e. use padding with a mask). However, I am having trouble using the mask in my task. And without the mask, the model does not detect a pattern
I would be glad if you could give me a concrete example of how to solve this problem in the form of code