Smiles LSTM With fine-tuning Train on ChEMBL 28 Fine-tune on user-provided set of molecules Draws from this GitHub repo and the associated Medium post WORK IN PROGRESS