Tuning Large Neural Networks via Zero-Shot Hyperparameter Transfergithub.com3 points·wizardcat··0 commentsOpen articleSaveView on HN