About Deep Learning General (DLG)

DLG is for discussion about training and designing models. This can range from model architecture, to hyperparameter search, to .

Example DLG topics could include:

  • Why isn’t my network converging?
  • What batch size should I use?
  • What layer type is the best for X?

Topics such as discussion about new research papers belongs in MLR. Discussion about system administration or hardware belong in SNE.