Parallel Training of Deep Networks with Local Updatesarxiv.org1 point·adamnemecek··0 commentsOpen articleSaveView on HN