The basic goal of my project is to use an RNN to parse through hundreds of thousands of valid drug compounds (as SMILES sequences - drugs represented as strings), and use th