140 lines
5.4 KiB
Python
140 lines
5.4 KiB
Python
# Copyright The PyTorch Lightning team.
|
|
#
|
|
# 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.
|
|
import os
|
|
from abc import ABC, abstractmethod
|
|
from contextlib import contextmanager
|
|
from typing import Any, List, Optional
|
|
|
|
import torch
|
|
from torch.nn.parallel import DistributedDataParallel
|
|
|
|
import pytorch_lightning as pl
|
|
from pytorch_lightning.overrides.base import unwrap_lightning_module
|
|
from pytorch_lightning.plugins.environments.cluster_environment import ClusterEnvironment
|
|
from pytorch_lightning.plugins.io.checkpoint_plugin import CheckpointIO
|
|
from pytorch_lightning.plugins.precision import PrecisionPlugin
|
|
from pytorch_lightning.strategies.training_type_plugin import Strategy
|
|
from pytorch_lightning.utilities import _XLA_AVAILABLE
|
|
from pytorch_lightning.utilities.distributed import all_gather_ddp_if_available, ReduceOp
|
|
|
|
|
|
class ParallelStrategy(Strategy, ABC):
|
|
"""Plugin for training with multiple processes in parallel."""
|
|
|
|
def __init__(
|
|
self,
|
|
accelerator: Optional["pl.accelerators.accelerator.Accelerator"] = None,
|
|
parallel_devices: Optional[List[torch.device]] = None,
|
|
cluster_environment: Optional[ClusterEnvironment] = None,
|
|
checkpoint_io: Optional[CheckpointIO] = None,
|
|
precision_plugin: Optional[PrecisionPlugin] = None,
|
|
):
|
|
super().__init__(accelerator=accelerator, checkpoint_io=checkpoint_io, precision_plugin=precision_plugin)
|
|
self.parallel_devices = parallel_devices
|
|
self.cluster_environment = cluster_environment
|
|
|
|
@property
|
|
@abstractmethod
|
|
def root_device(self) -> torch.device:
|
|
"""Return the root device."""
|
|
|
|
@property
|
|
def on_gpu(self) -> bool:
|
|
return self.root_device.type == "cuda" and torch.cuda.is_available()
|
|
|
|
@property
|
|
def on_tpu(self) -> bool:
|
|
return self.root_device.type == "xla" and _XLA_AVAILABLE
|
|
|
|
@property
|
|
def lightning_module(self) -> Optional["pl.LightningModule"]:
|
|
return unwrap_lightning_module(self.model) if self.model is not None else None
|
|
|
|
@property
|
|
def global_rank(self) -> int:
|
|
return self.cluster_environment.global_rank() if self.cluster_environment is not None else 0
|
|
|
|
@property
|
|
def local_rank(self) -> int:
|
|
return self.cluster_environment.local_rank() if self.cluster_environment is not None else 0
|
|
|
|
@property
|
|
def node_rank(self) -> int:
|
|
return self.cluster_environment.node_rank() if self.cluster_environment is not None else 0
|
|
|
|
@property
|
|
def world_size(self) -> int:
|
|
return self.cluster_environment.world_size() if self.cluster_environment is not None else 1
|
|
|
|
@property
|
|
def is_global_zero(self) -> bool:
|
|
return self.global_rank == 0
|
|
|
|
@property
|
|
def distributed_sampler_kwargs(self):
|
|
distributed_sampler_kwargs = dict(num_replicas=len(self.parallel_devices), rank=self.global_rank)
|
|
return distributed_sampler_kwargs
|
|
|
|
def reconciliate_processes(self, trace: str):
|
|
"""Function to re-conciliate processes on failure."""
|
|
|
|
def all_gather(self, tensor: torch.Tensor, group: Optional[Any] = None, sync_grads: bool = False) -> torch.Tensor:
|
|
"""Perform a all_gather on all processes."""
|
|
return all_gather_ddp_if_available(tensor, group=group, sync_grads=sync_grads)
|
|
|
|
def reduce_boolean_decision(self, decision: bool) -> bool:
|
|
decision = torch.tensor(int(decision), device=self.lightning_module.device)
|
|
decision = self.reduce(decision, reduce_op=ReduceOp.SUM)
|
|
decision = bool(decision == self.world_size)
|
|
return decision
|
|
|
|
@property
|
|
def torch_distributed_backend(self):
|
|
torch_backend = os.getenv("PL_TORCH_DISTRIBUTED_BACKEND")
|
|
if torch_backend is None:
|
|
torch_backend = "nccl" if self.on_gpu else "gloo"
|
|
return torch_backend
|
|
|
|
@staticmethod
|
|
def configure_sync_batchnorm(model: "pl.LightningModule") -> "pl.LightningModule":
|
|
"""Add global batchnorm for a model spread across multiple GPUs and nodes.
|
|
|
|
Override to synchronize batchnorm between specific process groups instead
|
|
of the whole world or use a different sync_bn like `apex`'s version.
|
|
|
|
Args:
|
|
model: pointer to current :class:`LightningModule`.
|
|
|
|
Return:
|
|
LightningModule with batchnorm layers synchronized between process groups
|
|
"""
|
|
return torch.nn.SyncBatchNorm.convert_sync_batchnorm(model)
|
|
|
|
@contextmanager
|
|
def block_backward_sync(self):
|
|
"""Blocks ddp sync gradients behaviour on backwards pass.
|
|
|
|
This is useful for skipping sync when accumulating gradients, reducing communication overhead
|
|
Returns: context manager with sync behaviour off
|
|
"""
|
|
if isinstance(self.model, DistributedDataParallel):
|
|
with self.model.no_sync():
|
|
yield None
|
|
else:
|
|
yield None
|
|
|
|
def teardown(self) -> None:
|
|
self.cluster_environment.teardown()
|
|
super().teardown()
|