It depends on how the parallelism is implemented, e.g. distributed data parallel (DDP) to synchronize gradients: https://pytorch.org/tutorials/intermediate/ddp_

It depends on how the parallelism is implemented, e.g. distributed data parallel (DDP) to synchronize gradients: https://pytorch.org/tutorials/intermediate/ddp_tutorial.html https://pytorch.org/tutorials/intermediate/ddp_tutorial.html It's a rabbit hole I stay away from for pragmatic reasons.