# Copyright 2020 Petuum, Inc. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""
This module contains simple collective communications primitives which operate
on arbitrary python objects. It is meant to be general but *non-performant*.
Only use these primitives if you are synchronizing small objects which can be
efficiently pickled and operated on. For larger objects, use framework-specific
functions, such as those provided by `torch.distributed`.
The functions in this module should be invoked *in the same order* across all
replicas in the current job. Otherwise, their behavior is undefined and you may
encounter unexpected bugs and errors.
"""
# TODO: Merge the reducer into this module once the previous trainer APIs
# are removed.
import adaptdl.env
from .reducer import Reducer, default_reduce_fn
_REDUCER = None
[docs]def initialize(master_addr=None,
master_port=None,
replica_rank=None,
num_replicas=None):
"""
Initialize this module, must be invoked before calling any other functions.
This function will block until it has been invoked from all replicas.
Arguments:
master_addr: address of the replica with rank 0.
master_port: free port of the replica with rank 0.
replica_rank: rank of the current replica.
num_replicas: total number of replicas.
Raises:
RuntimeError: If this module had already been initialized.
"""
global _REDUCER
if replica_rank is None:
replica_rank = adaptdl.env.replica_rank()
if num_replicas is None:
num_replicas = adaptdl.env.num_replicas()
if _REDUCER is not None:
raise RuntimeError("{} is already initialized".format(__name__))
if master_addr is None:
master_addr = adaptdl.env.master_addr()
if master_port is None:
master_port = adaptdl.env.master_port()
_REDUCER = Reducer(replica_rank,
num_replicas,
master_addr,
master_port)
[docs]def teardown():
"""
Teardown this module, will block until this function has been invoked from
all replicas.
Raises:
RuntimeError: If this module has not been initialized.
"""
raise NotImplementedError # TODO
[docs]def allreduce(value, reduce_fn=default_reduce_fn):
"""
Reduces a value across all replicas in such a way that they all get the
final result. Blocks until this function is invoked by all replicas.
Arguments:
value (object): The object which will be reduced together with all
other replicas.
reduce_fn (Function): A reduction function which two objects as
arguments, and returns the resulting reduced object.
Returns:
object: Resulting value after being reduced across all replicas.
Raises:
RuntimeError: If this module has not been initialized.
"""
if _REDUCER is None:
raise RuntimeError("{} has not been initialized".format(__name__))
return _REDUCER.allreduce(value, reduce_fn)
[docs]def allreduce_async(value, reduce_fn=default_reduce_fn):
"""
Asynchronous version of the `allreduce` function. Does not block, instead
returns a future which can be used to obtain the result later.
Arguments:
value (object): The object which will be reduced together with all
other replicas.
reduce_fn (Function): A reduction function which two objects as
arguments, and returns the resulting reduced object.
Returns:
Future: Object from which the result can be obtained later.
Raises:
RuntimeError: If this module has not been initialized.
"""
if _REDUCER is None:
raise RuntimeError("{} has not been initialized".format(__name__))
return _REDUCER.allreduce_async(value, reduce_fn)
[docs]def broadcast(value):
"""
Broadcasts a value from the replica of rank 0 to all replicas. Blocks until
this function is invoked by all replicas.
Arguments:
value (object): The object which will be broadcasted from replica 0.
Ignored on all other replicas.
Returns:
object: The value broadcasted from replica 0.
Raises:
RuntimeError: If this module has not been initialized.
"""
if _REDUCER is None:
raise RuntimeError("{} has not been initialized".format(__name__))
return _REDUCER.broadcast(value)