I\'m researching this topic as furiously as possible, but can\'t quite solve the issue of nan loss values when training.
(I\'m rather new to ML, so you\'ll have to fo
