torch.distributed.elastic

若API未标明支持情况,则代表该API的支持情况待验证。

API名称

是否支持

限制与说明

torch.distributed.elastic.agent.server.ElasticAgent

是

  

torch.distributed.elastic.agent.server.ElasticAgent.get_worker_group

是

  

torch.distributed.elastic.agent.server.ElasticAgent.run

是

  

torch.distributed.elastic.agent.server.WorkerSpec

是

  

torch.distributed.elastic.agent.server.WorkerSpec.get_entrypoint_name

是

  

torch.distributed.elastic.agent.server.WorkerState

是

  

torch.distributed.elastic.agent.server.WorkerState.is_running

是

  

torch.distributed.elastic.agent.server.Worker

是

  

torch.distributed.elastic.agent.server.WorkerGroup

是

  

torch.distributed.elastic.agent.server.local_elastic_agent.LocalElasticAgent

是

  

torch.distributed.elastic.agent.server.SimpleElasticAgent

是

  

torch.distributed.elastic.agent.server.SimpleElasticAgent._assign_worker_ranks

是

  

torch.distributed.elastic.agent.server.SimpleElasticAgent._exit_barrier

是

  

torch.distributed.elastic.agent.server.SimpleElasticAgent._initialize_workers

是

  

torch.distributed.elastic.agent.server.SimpleElasticAgent._monitor_workers

是

  

torch.distributed.elastic.agent.server.SimpleElasticAgent._rendezvous

是

  

torch.distributed.elastic.agent.server.SimpleElasticAgent._restart_workers

是

  

torch.distributed.elastic.agent.server.SimpleElasticAgent._shutdown

是

  

torch.distributed.elastic.agent.server.SimpleElasticAgent._start_workers

是

  

torch.distributed.elastic.agent.server.SimpleElasticAgent._stop_workers

是

  

torch.distributed.elastic.agent.server.api.RunResult

是

  

torch.distributed.elastic.multiprocessing.start_processes

是

  

torch.distributed.elastic.multiprocessing.api.PContext

是

  

torch.distributed.elastic.multiprocessing.api.MultiprocessContext

是

  

torch.distributed.elastic.multiprocessing.api.SubprocessContext

是

  

torch.distributed.elastic.multiprocessing.api.RunProcsResult

是

  

torch.distributed.elastic.multiprocessing.errors.record

是

  

torch.distributed.elastic.multiprocessing.errors.ChildFailedError

是

  

torch.distributed.elastic.multiprocessing.errors.ErrorHandler

是

  

torch.distributed.elastic.multiprocessing.errors.ProcessFailure

是

  

torch.distributed.elastic.rendezvous.RendezvousParameters

是

  

torch.distributed.elastic.rendezvous.RendezvousParameters.get

是

  

torch.distributed.elastic.rendezvous.RendezvousParameters.get_as_bool

是

  

torch.distributed.elastic.rendezvous.RendezvousParameters.get_as_int

是

  

torch.distributed.elastic.rendezvous.RendezvousHandlerRegistry

是

  

torch.distributed.elastic.rendezvous.RendezvousHandlerRegistry.create_handler

是

  

torch.distributed.elastic.rendezvous.RendezvousHandlerRegistry.register

是

  

torch.distributed.elastic.rendezvous.RendezvousHandler

是

  

torch.distributed.elastic.rendezvous.RendezvousHandler.get_backend

是

  

torch.distributed.elastic.rendezvous.RendezvousHandler.get_run_id

是

  

torch.distributed.elastic.rendezvous.RendezvousHandler.is_closed

是

  

torch.distributed.elastic.rendezvous.RendezvousHandler.next_rendezvous

是

  

torch.distributed.elastic.rendezvous.RendezvousHandler.num_nodes_waiting

是

  

torch.distributed.elastic.rendezvous.RendezvousHandler.set_closed

是

  

torch.distributed.elastic.rendezvous.RendezvousHandler.shutdown

是

  

torch.distributed.elastic.rendezvous.RendezvousError

是

  

torch.distributed.elastic.rendezvous.RendezvousClosedError

是

  

torch.distributed.elastic.rendezvous.RendezvousTimeoutError

是

  

torch.distributed.elastic.rendezvous.RendezvousConnectionError

是

  

torch.distributed.elastic.rendezvous.RendezvousStateError

是

  

torch.distributed.elastic.rendezvous.dynamic_rendezvous.create_handler

是

  

torch.distributed.elastic.rendezvous.dynamic_rendezvous.DynamicRendezvousHandler

是

  

torch.distributed.elastic.rendezvous.dynamic_rendezvous.DynamicRendezvousHandler.from_backend

是

  

torch.distributed.elastic.rendezvous.dynamic_rendezvous.RendezvousBackend

是

  

torch.distributed.elastic.rendezvous.dynamic_rendezvous.RendezvousBackend.get_state

是

  

torch.distributed.elastic.rendezvous.dynamic_rendezvous.RendezvousBackend.name

     

torch.distributed.elastic.rendezvous.dynamic_rendezvous.RendezvousBackend.set_state

是

  

torch.distributed.elastic.rendezvous.dynamic_rendezvous.RendezvousTimeout

是

  

torch.distributed.elastic.rendezvous.dynamic_rendezvous.RendezvousTimeout.close

     

torch.distributed.elastic.rendezvous.dynamic_rendezvous.RendezvousTimeout.heartbeat

     

torch.distributed.elastic.rendezvous.dynamic_rendezvous.RendezvousTimeout.join

     

torch.distributed.elastic.rendezvous.dynamic_rendezvous.RendezvousTimeout.last_call

     

torch.distributed.elastic.rendezvous.c10d_rendezvous_backend.create_backend

是

  

torch.distributed.elastic.rendezvous.c10d_rendezvous_backend.C10dRendezvousBackend

是

  

torch.distributed.elastic.rendezvous.c10d_rendezvous_backend.C10dRendezvousBackend.get_state

是

  

torch.distributed.elastic.rendezvous.c10d_rendezvous_backend.C10dRendezvousBackend.name

     

torch.distributed.elastic.rendezvous.c10d_rendezvous_backend.C10dRendezvousBackend.set_state

是

  

torch.distributed.elastic.rendezvous.etcd_rendezvous_backend.create_backend

     

torch.distributed.elastic.rendezvous.etcd_rendezvous_backend.EtcdRendezvousBackend

     

torch.distributed.elastic.rendezvous.etcd_rendezvous_backend.EtcdRendezvousBackend.get_state

     

torch.distributed.elastic.rendezvous.etcd_rendezvous_backend.EtcdRendezvousBackend.name

     

torch.distributed.elastic.rendezvous.etcd_rendezvous_backend.EtcdRendezvousBackend.set_state

     

torch.distributed.elastic.rendezvous.etcd_rendezvous.EtcdRendezvousHandler

     

torch.distributed.elastic.rendezvous.etcd_store.EtcdStore

     

torch.distributed.elastic.rendezvous.etcd_store.EtcdStore.add

     

torch.distributed.elastic.rendezvous.etcd_store.EtcdStore.check

     

torch.distributed.elastic.rendezvous.etcd_store.EtcdStore.get

     

torch.distributed.elastic.rendezvous.etcd_store.EtcdStore.set

     

torch.distributed.elastic.rendezvous.etcd_store.EtcdStore.wait

     

torch.distributed.elastic.rendezvous.etcd_server.EtcdServer

     

torch.distributed.elastic.timer.configure

是

  

torch.distributed.elastic.timer.expires

是

  

torch.distributed.elastic.timer.LocalTimerServer

是

  

torch.distributed.elastic.timer.LocalTimerClient

是

  

torch.distributed.elastic.timer.FileTimerServer

     

torch.distributed.elastic.timer.FileTimerClient

     

torch.distributed.elastic.timer.TimerRequest

是

  

torch.distributed.elastic.timer.TimerServer

是

  

torch.distributed.elastic.timer.TimerServer.clear_timers

是

  

torch.distributed.elastic.timer.TimerServer.get_expired_timers

是

  

torch.distributed.elastic.timer.TimerServer.register_timers

是

  

torch.distributed.elastic.timer.TimerClient

是

  

torch.distributed.elastic.timer.TimerClient.acquire

是

  

torch.distributed.elastic.timer.TimerClient.release

是

  

torch.distributed.elastic.metrics.api.MetricHandler

是

  

torch.distributed.elastic.metrics.api.ConsoleMetricHandler

是

  

torch.distributed.elastic.metrics.api.NullMetricHandler

是

  

torch.distributed.elastic.metrics.configure

是

  

torch.distributed.elastic.metrics.prof

是

  

torch.distributed.elastic.metrics.put_metric

是

  

torch.distributed.elastic.events.record

是

  

torch.distributed.elastic.events.get_logging_handler

是

  

torch.distributed.elastic.events.api.Event

是

  

torch.distributed.elastic.events.api.EventSource

是

  

torch.distributed.elastic.events.api.EventMetadataValue

是