Distributed-TensorFlow-Guide/Synchronous-SGD at master · tmulc18/Distributed-TensorFlow-Guide · GitHub
Skip to content

Latest commit

 

History

History

Folders and files

SSGD (Synchronous SGD)

Have workers send their updates to a ps, but only update the variables on the ps after N updates have been accumulated. If the number of workers is M and M>N, then this is known as dropping the last M-N stale gradients.