Accelerating Distributed ML Training via Selective Synchronization (Poster Abstract)

share
Conference Paper