简要概览

  pytorch官方提供的数据并行类为：

torch.nn.DataParallel(module, device_ids=None, output_device=None, dim=0)

  当给定model时，主要实现功能是将input数据依据batch的这个维度，将数据划分到指定的设备上。其他的对象(objects)复制到每个设备上。在前向传播的过程中，module被复制到每个设备上，每个复制的副本处理一部分输入数据。在反向传播过程中，每个副本module的梯度被汇聚到原始的module上计算(一般为第0块GPU)。

并且这里要注意的一点是，这里官方推荐是用DistributedDataParallel，因为DistributedDataParallel使用的是多进程方式，而DataParallel使用的是多线程的方式。如果使用的是DistributedDataParallel，你需要使用torch.distributed.launch去launch程序，参考Distributed Communication Package - Torch.Distributed。

  batch size的大小一定要大于GPU的数量，我在实践过程中batch size的大小一般设置为GPU块数的倍数。在数据分配到不同的机器上的时候，传入module的数据同样都可以传入DataParallel(并行之后的module类型)中，但是tensor默认按照dim=0分配到不同的机器上，tuple, list，dict类型的数据被浅拷贝到不同的GPU上，其它类型的数据将会被分配到不同的进程中。

  在调用DataParallel之前，module必须要具有他自己的参数(能获取到模型的参数)，还需要在指定的GPU上具有buffer(不然会报内存出错)。

在前向传播的过程中，module被复制到每个设备上，因此在前线传播过程中的任何更新都会丢失。举例来说，如果module有一个counter属性，在每次前线传播过程中都会加1，它将会保留在初始值状态，因为更新在副本上，但是副本前线传播完就被销毁了。然而在DataParallel中，device[0]上的副本将其参数和内存数据与并行的module共享，因此在device[0]上更新数据将会被记录。

返回的结果是来自各个device上的数据的汇总。默认是dim 0维度上的汇总。因此在处理RNN时序数据时就需要注意这一点。My recurrent network doesn’t work with data parallelism

torch.nn.DataParallel(module, device_ids=None, output_device=None, dim=0)

  torch.nn.DataParallel()函数的参数主要有module，device_ids，output_device这三个。

1. module为需要并行的module。
2. device_ids为一个list，默认为所有可操作的devices。
3. output_device为需要输出汇总的指定GPU，默认为device_ids[0]号。

  简单的举例为：

>>> net = torch.nn.DataParallel(model, device_ids=[0, 1, 2])
>>> output = net(input_var)  # input_var can be on any device, including CPU

源码解析

  data_parallel.py的源码地址为：https://github.com/pytorch/pytorch/blob/master/torch/nn/parallel/data_parallel.py

  源码注释

import operator
import torch
import warnings
from itertools import chain
from ..modules import Module
from .scatter_gather import scatter_kwargs, gather
from .replicate import replicate
from .parallel_apply import parallel_apply
from torch._utils import (
    _get_all_device_indices,
    _get_available_device_type,
    _get_device_index,
    _get_devices_properties
)

def _check_balance(device_ids):
    imbalance_warn = """
    There is an imbalance between your GPUs. You may want to exclude GPU {} which
    has less than 75% of the memory or cores of GPU {}. You can do so by setting
    the device_ids argument to DataParallel, or by setting the CUDA_VISIBLE_DEVICES
    environment variable."""
    device_ids = [_get_device_index(x, True) for x in device_ids]
    dev_props = _get_devices_properties(device_ids)

    def warn_imbalance(get_prop):
        values = [get_prop(props) for props in dev_props]
        min_pos, min_val = min(enumerate(values), key=operator.itemgetter(1))
        max_pos, max_val = max(enumerate(values), key=operator.itemgetter(1))
        if min_val / max_val < 0.75:
            warnings.warn(imbalance_warn.format(device_ids[min_pos], device_ids[max_pos]))
            return True
        return False

    if warn_imbalance(lambda props: props.total_memory):
        return
    if warn_imbalance(lambda props: props.multi_processor_count):
        return

DataParallel类初始化：

class DataParallel(Module):
    # TODO: update notes/cuda.rst when this class handles 8+ GPUs well

    def __init__(self, module, device_ids=None, output_device=None, dim=0):
        super(DataParallel, self).__init__()
		
		# 通过调用torch.cuda.is_available()判断是返回“cuda”还是None。
        device_type = _get_available_device_type() 
        if device_type is None: # 检查是否有GPU
        	# 如果没有GPU的话，module就不能够并行，直接赋值，设备id置空
            self.module = module
            self.device_ids = []
            return

        if device_ids is None: # 如果没有指定GPU，则默认使用所有可用的GPU
        	# 获取所有可用的设备ID，为一个list。
            device_ids = _get_all_device_indices()

        if output_device is None: # 判断输出设备是否指定
            output_device = device_ids[0] # 默认为指定设备的第一个

        self.dim = dim
        self.module = module # self.module就是传入的module。
        self.device_ids = [_get_device_index(x, True) for x in device_ids]
        self.output_device = _get_device_index(output_device, True)
        self.src_device_obj = torch.device(device_type, self.device_ids[0])

        _check_balance(self.device_ids)

        if len(self.device_ids) == 1:
            self.module.to(self.src_device_obj)

前向传播

    def forward(self, *inputs, **kwargs):
    	# 如果没有可用的GPU则使用原来的module来计算
        if not self.device_ids:
            return self.module(*inputs, **kwargs)
		
		# 这里应该是判断模型的参数和buffer都要有。
        for t in chain(self.module.parameters(), self.module.buffers()):
            if t.device != self.src_device_obj:
                raise RuntimeError("module must have its parameters and buffers "
                                   "on device {} (device_ids[0]) but found one of "
                                   "them on device: {}".format(self.src_device_obj, t.device))

        # 用scatter函数将input平均分配到每个GPU上
        inputs, kwargs = self.scatter(inputs, kwargs, self.device_ids) 
        # for forward function without any inputs, empty list and dict will be created
        # so the module can be executed on one device which is the first one in device_ids
        if not inputs and not kwargs:
            inputs = ((),)
            kwargs = ({
    },)

        if len(self.device_ids) == 1: # 只有一个给定的GPU的话，就直接调用未并行的module，否者进入下一步
            return self.module(*inputs[0], **kwargs[0])
        replicas = self.replicate(self.module, self.device_ids[:len(inputs)]) # replicate函数主要讲模型复制到多个GPU上
        outputs = self.parallel_apply(replicas, inputs, kwargs) # 并行地在多个GPU上计算模型。
        return self.gather(outputs, self.output_device) # 将数据聚合到一起，传送到output_device上，默认也是dim 0维度聚合。

    def replicate(self, module, device_ids):
        return replicate(module, device_ids, not torch.is_grad_enabled())

    def scatter(self, inputs, kwargs, device_ids):
        return scatter_kwargs(inputs, kwargs, device_ids, dim=self.dim)

    def parallel_apply(self, replicas, inputs, kwargs):
        return parallel_apply(replicas, inputs, kwargs, self.device_ids[:len(replicas)])

    def gather(self, outputs, output_device):
        return gather(outputs, output_device, dim=self.dim)

• scatter函数：

def scatter(inputs, target_gpus, dim=0):
    r"""
    Slices tensors into approximately equal chunks and
    distributes them across given GPUs. Duplicates
    references to objects that are not tensors.
    """
    def scatter_map(obj):
        if isinstance(obj, torch.Tensor):
            return Scatter.apply(target_gpus, None, dim, obj)
        if isinstance(obj, tuple) and len(obj) > 0:
            return list(zip(*map(scatter_map, obj)))
        if isinstance(obj, list) and len(obj) > 0:
            return list(map(list, zip(*map(scatter_map, obj))))
        if isinstance(obj, dict) and len(obj) > 0:
            return list(map(type(obj), zip(*map(scatter_map, obj.items()))))
        return [obj for targets in target_gpus]

    # After scatter_map is called, a scatter_map cell will exist. This cell
    # has a reference to the actual function scatter_map, which has references
    # to a closure that has a reference to the scatter_map cell (because the
    # fn is recursive). To avoid this reference cycle, we set the function to
    # None, clearing the cell
    try:
        res = scatter_map(inputs)
    finally:
        scatter_map = None
    return res

  在前向传播中，数据需要通过scatter函数分配到每个GPU上，代码在scatter_gather.py文件下，如果输入的类型不是tensor的话，会依据数据类型处理一下变成tensor，再递归调用scatter_map，最后调用Scatter.apply方法将数据依据给定的GPU给划分好返回。

• replicate函数：

  replicate函数需要将模型给复制到每个GPU上。如果你定义的模型是ScriptModule的话，也就是在编写自己model的时候不是继承的nn.Module，而是继承的nn.ScriptModule，就不能复制，会报错。

  这个函数主要就是将模型参数、buffer等需要共享的信息，复制到每个GPU上，感兴趣的自己看吧。

data_parallel

def data_parallel(module, inputs, device_ids=None, output_device=None, dim=0, module_kwargs=None):
    r"""Evaluates module(input) in parallel across the GPUs given in device_ids.

    This is the functional version of the DataParallel module.

    Args:
        module (Module): the module to evaluate in parallel
        inputs (Tensor): inputs to the module
        device_ids (list of int or torch.device): GPU ids on which to replicate module
        output_device (list of int or torch.device): GPU location of the output  Use -1 to indicate the CPU.
            (default: device_ids[0])
    Returns:
        a Tensor containing the result of module(input) located on
        output_device
    """
    if not isinstance(inputs, tuple):
        inputs = (inputs,) if inputs is not None else ()

    device_type = _get_available_device_type()

    if device_ids is None:
        device_ids = _get_all_device_indices()

    if output_device is None:
        output_device = device_ids[0]

    device_ids = [_get_device_index(x, True) for x in device_ids]
    output_device = _get_device_index(output_device, True)
    src_device_obj = torch.device(device_type, device_ids[0])

    for t in chain(module.parameters(), module.buffers()):
        if t.device != src_device_obj:
            raise RuntimeError("module must have its parameters and buffers "
                               "on device {} (device_ids[0]) but found one of "
                               "them on device: {}".format(src_device_obj, t.device))

    inputs, module_kwargs = scatter_kwargs(inputs, module_kwargs, device_ids, dim)
    # for module without any inputs, empty list and dict will be created
    # so the module can be executed on one device which is the first one in device_ids
    if not inputs and not module_kwargs:
        inputs = ((),)
        module_kwargs = ({
    },)

    if len(device_ids) == 1:
        return module(*inputs[0], **module_kwargs[0])
    used_device_ids = device_ids[:len(inputs)]
    replicas = replicate(module, used_device_ids)
    outputs = parallel_apply(replicas, inputs, module_kwargs, used_device_ids)
    return gather(outputs, output_device, dim)

  并行的模型也有了，数据也有了，之后就是利用并行的模型和并行的数据来做计算了。

• parallel_apply函数：

def parallel_apply(modules, inputs, kwargs_tup=None, devices=None):
	# 判断模型数和输入数据数是否相等
    assert len(modules) == len(inputs)
    if kwargs_tup is not None:
        assert len(modules) == len(kwargs_tup)
    else:
        kwargs_tup = ({
    },) * len(modules)
    if devices is not None:
        assert len(modules) == len(devices)
    else:
        devices = [None] * len(modules)
    devices = list(map(lambda x: _get_device_index(x, True), devices))
    lock = threading.Lock()
    results = {
    }
    grad_enabled, autocast_enabled = torch.is_grad_enabled(), torch.is_autocast_enabled()

    def _worker(i, module, input, kwargs, device=None):
        torch.set_grad_enabled(grad_enabled)
        if device is None:
            device = get_a_var(input).get_device()
        try:
            with torch.cuda.device(device), autocast(enabled=autocast_enabled):
                # this also avoids accidental slicing of `input` if it is a Tensor
                if not isinstance(input, (list, tuple)):
                    input = (input,)
                output = module(*input, **kwargs)
            with lock:
                results[i] = output
        except Exception:
            with lock:
                results[i] = ExceptionWrapper(
                    where="in replica {} on device {}".format(i, device))

    if len(modules) > 1:
        threads = [threading.Thread(target=_worker,
                                    args=(i, module, input, kwargs, device))
                   for i, (module, input, kwargs, device) in
                   enumerate(zip(modules, inputs, kwargs_tup, devices))]

        for thread in threads:
            thread.start()
        for thread in threads:
            thread.join()
    else:
        _worker(0, modules[0], inputs[0], kwargs_tup[0], devices[0])

    outputs = []
    for i in range(len(inputs)):
        output = results[i]
        if isinstance(output, ExceptionWrapper):
            output.reraise()
        outputs.append(output)
    return outputs

  先判断一下数据的长度是否符合要求。之后利用多线程来处理数据。最后将所有的数据gather在一起，默认是从第0个维度gather在一起。

实例

import torch
import torch.nn as nn
import torch.optim as optim
from torch.autograd import Variable
from torch.utils.data import Dataset, DataLoader


class RandomDataset(Dataset):
    def __init__(self, size, length):
        self.len = length
        self.data = torch.randn(length, size)

    def __getitem__(self, index):
        return self.data[index]

    def __len__(self):
        return self.len


class Model(nn.Module):
    def __init__(self, input_size, output_size):
        super(Model, self).__init__()
        self.fc = nn.Linear(input_size, output_size)
        self.sigmoid = nn.Sigmoid()
        # self.modules = [self.fc, self.sigmoid]

    def forward(self, input):
        return self.sigmoid(self.fc(input))


if __name__ == '__main__':
    # Parameters and DataLoaders
    input_size = 5
    output_size = 1
    batch_size = 30
    data_size = 100

    rand_loader = DataLoader(dataset=RandomDataset(input_size, data_size),
                             batch_size=batch_size, shuffle=True)

    model = Model(input_size, output_size)
    if torch.cuda.device_count() > 1:
        print("Let's use", torch.cuda.device_count(), "GPUs!")
        model = nn.DataParallel(model).cuda()

    optimizer = optim.SGD(params=model.parameters(), lr=1e-3)
    cls_criterion = nn.BCELoss()

    for data in rand_loader:
        targets = torch.empty(data.size(0)).random_(2).view(-1, 1)

        if torch.cuda.is_available():
            input = Variable(data.cuda())
            with torch.no_grad():
                targets = Variable(targets.cuda())
        else:
            input = Variable(data)
            with torch.no_grad():
                targets = Variable(targets)

        output = model(input)

        optimizer.zero_grad()
        loss = cls_criterion(output, targets)
        loss.backward()
        optimizer.step()