# vm.torch.distributed¶

## Functions¶

all_reduce(…) : Reduce the tensor across all nodes in a group.

broadcast(…) : Broadcast the tensor from source node in a group.