[docs]defis_available()->bool:""" Returns ``True`` if the distributed package is available. Otherwise, ``torch.distributed`` does not expose any other APIs. Currently, ``torch.distributed`` is available on Linux, MacOS and Windows. Set ``USE_DISTRIBUTED=1`` to enable it when building PyTorch from source. Currently, the default value is ``USE_DISTRIBUTED=1`` for Linux and Windows, ``USE_DISTRIBUTED=0`` for MacOS. """returnhasattr(torch._C,"_c10d_init")
ifis_available()andnottorch._C._c10d_init():raiseRuntimeError("Failed to initialize torch.distributed")ifis_available():fromtorch._C._distributed_c10dimport(Store,FileStore,TCPStore,ProcessGroup,PrefixStore,Reducer,Logger,BuiltinCommHookType,GradBucket,Workas_Work,_DEFAULT_FIRST_BUCKET_BYTES,_register_comm_hook,_register_builtin_comm_hook,_broadcast_coalesced,_compute_bucket_assignment_by_size,_verify_params_across_processes,_test_python_store,DebugLevel,get_debug_level,set_debug_level,set_debug_level_from_env,)ifsys.platform!="win32":fromtorch._C._distributed_c10dimport(HashStore,_round_robin_process_groups,)from.distributed_c10dimport*# noqa: F403# Variables prefixed with underscore are not auto imported# See the comment in `distributed_c10d.py` above `_backend` on why we expose# this.from.distributed_c10dimport(_backend,_all_gather_base,_reduce_scatter_base,_create_process_group_wrapper,_rank_not_in_group,)from.remote_deviceimport_remote_deviceset_debug_level_from_env()
Docs
Access comprehensive developer documentation for PyTorch
To analyze traffic and optimize your experience, we serve cookies on this site. By clicking or navigating, you agree to allow our usage of cookies. As the current maintainers of this site, Facebook’s Cookies Policy applies. Learn more, including about available controls: Cookies Policy.