I have created a data generator for testing purposes that will create data whose category can be easily recognized by a quick glance.
And then created three simple scripts for a feedorward, bi-lstm, rnn, and pre-processing, by following the tutorials, hoping to better understand the them.
I must have made some glaring mistake given that none of my models seem to do better than a coin toss, on a data-set that should be easily predictable. Could someone have a look in case I am missing something when it comes to how PyTorch works.