dgl.multiprocessing.pytorch.call_once_and_shareο
- dgl.multiprocessing.pytorch.call_once_and_share(func, shape, dtype, rank=0)[source]ο
Invoke the function in a single process of the PyTorch distributed process group, and share the result with other processes.
- Parameters:
func (callable) β Any callable that accepts no arguments and returns an arbitrary object.
shape (tuple[int]) β The shape of the shared tensor. Must match the output of
func
.dtype (torch.dtype) β The data type of the shared tensor. Must match the output of
func
.rank (int, optional) β The process ID to actually execute the function.