For GAN, there are different flavors: Wasserstein-GAN (Facebook), Cramer-GAN (DeepMind), Optimal Transport-GAN (OpenAI), Coulomb-GAN (Linz University), although at the end, maybe they are all equal (Google).
You can also find more in the Natural Language Processing literature (and apply them to SMILES):
Finally, it will be interesting to design a systematic procedure for testing hyperparameter values. These methods are often very sensitive to hyperparameter choice (another suggestion from an anonymous referee of my previous paper).