Demo: Decoupled Momentum Optimization – training neural networks in parallelarxiv.org3 points·sergiotapia··0 commentsOpen articleSaveView on HN