我们在博客《分布式机器学习:同步并行SGD算法的实现与复杂度分析(PySpark)》和博客《分布式机器学习:模型平均MA与弹性平均EASGD(PySpark) 》中介绍的都是同步算法。同步算法的共性是所有的节点会以一定的频率进行全局同步。然而,当工作节点的计算性能存在差异,或者某些工作节点无法正常工作(比如死机)的时候,分布式系统的整体运行效率不好,甚至无法完成训练任务。为了解决此问题,人们提出了异步的并行算法。在异步的通信模式下,各个工作节点不需要互相等待,而是以一个或多个全局服务器做为中介,实现对全局模型的更新和读取。这样可以显著减少通信时间,从而获得更好的多机扩展性。
2 异步SGD
2.1 算法描述与实现
异步SGD[9]是最基础的异步算法,其流畅如下图所示。粗略地讲,ASGD的参数更新发生在工作节点,而模型的更新发生在服务器端。当参数服务器接收到来自某个工作节点的参数梯度时,就直接将其加到全局模型上,而无需等待其它工作节点的梯度信息。
下面我们用Pytorch实现的训练代码(采用RPC进行进程间通信)。首先,我们设置初始化多个进程,其中0号进程做为参数服务器,其余进程做为worker来对模型进行训练,则总的通信域(world_size)大小为workers的数量+1。这里我们设置参数服务器IP地址为localhost
,端口号29500
。
def run(rank, world_size):
os.environ['MASTER_ADDR'] = 'localhost'
os.environ['MASTER_PORT'] = '29500'
options=rpc.TensorPipeRpcBackendOptions(
num_worker_threads=16,
rpc_timeout=0 # infinite timeout
)
if rank == 0:
rpc.init_rpc(
"ps",
rank=rank,
world_size=world_size,
rpc_backend_options=options
)
run_ps([f"trainer{r}" for r in range(1, world_size)])
else:
rpc.init_rpc(
f"trainer{rank}",
rank=rank,
world_size=world_size,
rpc_backend_options=options
)
# trainer passively waiting for ps to kick off training iterations
# block until all rpcs finish
rpc.shutdown()
if __name__=="__main__":
world_size = n_workers + 1
mp.spawn(run, args=(world_size, ), nprocs=world_size, join=True)
下面我们定义参数服务器的所要完成工作流程,包括将训练数据划分到各个worker,异步调用所有worker的训练流程,最后训练完毕后在参数服务器完成对模型的评估。
def run_trainer(ps_rref, train_dataset):
trainer = Trainer(ps_rref)
trainer.train(train_dataset)
def run_ps(trainers):
transform=transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.1307,), (0.3081,))
])
train_dataset = datasets.MNIST('./data', train=True, download=True,
transform=transform)
local_train_datasets = dataset_split(train_dataset, n_workers)
print(f"{datetime.now().strftime('%H:%M:%S')} Start training")
ps = ParameterServer()
ps_rref = rpc.RRef(ps)
futs = []
for idx, trainer in enumerate(trainers):
futs.append(
rpc.rpc_async(trainer, run_trainer, args=(ps_rref, local_train_datasets[idx]))
)
torch.futures.wait_all(futs)
print(f"{datetime.now().strftime('%H:%M:%S')} Finish training")
ps.evaluation()
这里数据集的划分代码采用我们在《Pytorch:单卡多进程并行训练》中所述的数据划分方式:
class CustomSubset(Subset):
'''A custom subset class with customizable data transformation'''
def __init__(self, dataset, indices, subset_transform=None):
super().__init__(dataset, indices)
self.subset_transform = subset_transform
def __getitem__(self, idx):
x, y = self.dataset[self.indices[idx]]
if self.subset_transform:
x = self.subset_transform(x)
return x, y
def __len__(self):
return len(self.indices)
def dataset_split(dataset, n_workers):
n_samples = len(dataset)
n_sample_per_workers = n_samples // n_workers
local_datasets = []
for w_id in range(n_workers):
if w_id < n_workers - 1:
local_datasets.append(CustomSubset(dataset, range(w_id * n_sample_per_workers, (w_id + 1) * n_sample_per_workers)))
else:
local_datasets.append(CustomSubset(dataset, range(w_id * n_sample_per_workers, n_samples)))
return local_datasets
以下是参数服务器类ParameterServer
的定义:
class ParameterServer(object):
def __init__(self, n_workers=n_workers):
self.model = Net().to(device)
self.lock = threading.Lock()
self.future_model = torch.futures.Future()
self.n_workers = n_workers
self.curr_update_size = 0
self.optimizer = optim.SGD(self.model.parameters(), lr=0.001, momentum=0.9)
for p in self.model.parameters():
p.grad = torch.zeros_like(p)
self.test_loader = torch.utils.data.DataLoader(
datasets.MNIST('../data', train=False,
transform=transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.1307,), (0.3081,))
])),
batch_size=32, shuffle=True)
def get_model(self):
# TensorPipe RPC backend only supports CPU tensors,
# so we move your tensors to CPU before sending them over RPC
return self.model.to("cpu")
@staticmethod
@rpc.functions.async_execution
def update_and_fetch_model(ps_rref, grads):
self = ps_rref.local_value()
for p, g in zip(self.model.parameters(), grads):
p.grad += g
with self.lock:
self.curr_update_size += 1
fut = self.future_model
if self.curr_update_size >= self.n_workers:
for p in self.model.parameters():
p.grad /= self.n_workers
self.curr_update_size = 0
self.optimizer.step()
self.optimizer.zero_grad()
fut.set_result(self.model)
self.future_model = torch.futures.Future()
return fut
def evaluation(self):
self.model.eval()
self.model = self.model.to(device)
test_loss = 0
correct = 0
with torch.no_grad():
for data, target in self.test_loader:
output = self.model(data.to(device))
test_loss += F.nll_loss(output, target.to(device), reduction='sum').item() # sum up batch loss
pred = output.max(1)[1] # get the index of the max log-probability
correct += pred.eq(target.to(device)).sum().item()
test_loss /= len(self.test_loader.dataset)
print('\nTest result - Accuracy: {}/{} ({:.0f}%)\n'.format(
correct, len(self.test_loader.dataset), 100. * correct / len(self.test_loader.dataset)))
以下是Trainer
类的定义:
class Trainer(object):
def __init__(self, ps_rref):
self.ps_rref = ps_rref
self.model = Net().to(device)
def train(self, train_dataset):
train_loader = torch.utils.data.DataLoader(train_dataset, batch_size=batch_size, shuffle=True)
model = self.ps_rref.rpc_sync().get_model().cuda()
pid = os.getpid()
for epoch in range(epochs):
for batch_idx, (data, target) in enumerate(train_loader):
output = model(data.to(device))
loss = F.nll_loss(output, target.to(device))
loss.backward()
model = rpc.rpc_sync(
self.ps_rref.owner(),
ParameterServer.update_and_fetch_model,
args=(self.ps_rref, [p.grad for p in model.cpu().parameters()]),
).cuda()
if batch_idx % log_interval == 0:
print('{}\tTrain Epoch: {} [{}/{} ({:.0f}%)]\tLoss: {:.6f}'.format(
pid, epoch + 1, batch_idx * len(data), len(train_loader.dataset),
100. * batch_idx / len(train_loader), loss.item()))
完整代码我已经上传到了GitHub仓库 [Distributed-Algorithm-PySpark],感兴趣的童鞋可以前往查看。
运行该代码得到的评估结果为:
Test result - Accuracy: 9696/10000 (97%)
可见该训练算法是收敛的,但在10个epoch下在测试集上只能达到97%的精度,不如我们下面提到的在10个epoch就能在测试集上达到99%精度的Hogwild!算法。注意,ASGD和Hogwild都是异步算法,但ASGD是分布式算法(当然我们这里是单机多进程模拟),进程间采用RPC通信,不会出现同步错误的问题,根本不需要考虑加不加锁。而Hogwild!算法是单机算法,进程/线程间采用共享内存通信,需要考虑加不加锁的问题,不过Hogwild!算法为了提高训练过程中的数据吞吐量,直接采用了无锁的全局内存访问。
2.2 收敛性分析
ASGD避开了同步开销,但会给模型更新增加一些延迟。我们下面将ASGD的工作流程用下图加以剖析来解释这一点。用\(\text{worker}(k)\)来代表第\(k\)个工作节点,用\(w^t\)来代表第\(t\)轮迭代时服务端的全局模型。按照时间顺序,首先\(\text{worker}(k)\)先从参数服务器获取全局模型\(w^t\),再根据本地数据计算模型梯度\(g(w_t)\)并将其发往参数服务器。一段时间后,\(\text{worker}(k')\)也从参数服务器取回当时的全局模型\(w^{t+1}\),并同样依据它的本地数据计算模型的梯度\(f(w^{t+1})\)。注意,在\(\text{worker}(k')\)取回参数并进行计算的过程中,其它工作节点(比如\(\text{worker}(k)\))可能已经将它的梯度提交给服务器并进行更新了。所以当\(\text{worker}(k')\)将其梯度\(g(w^{t+1})\)发给服务器时,全局模型已经不再是\(w^{t+1}\),而已经是被更新过的版本。
我们将上面这种现象称为梯度和模型的失配,也即我们用一个比较旧的参数计算了梯度,而将这个“延迟”的梯度更新到了模型参数上。这种延迟使得ASGD和SGD之间在参数更新规则上存在偏差,可能导致模型在某些特定的更新点上出现严重抖动,设置优化过程出错,无法收敛。后面我们会介绍克服延迟问题的手段。
3 Hogwild!算法
3.1 算法描述与实现
异步并行算法既可以在多机集群上开展,也可以在多核系统下通过多线程开展。当我们把ASGD算法应用在多线程环境中时,因为不再有参数服务器这一角色,算法的细节会发生一些变化。特别地,因为全局模型存储在共享内存中,所以当异步的模型更新发生时,我们需要讨论是否将内存加锁,以保证模型写入的一致性。
Hogwild!算法[2]为了提高训练过程中的数据吞吐量,选择了无锁的全局模型访问,其工作逻辑如下所示:
这里使用我们在《Pytorch:单卡多进程并行训练》所提到的torch.multiprocessing
来进行多进程并行训练。多进程原本内存空间是独立的,这里我们显式调用model.share_memory()
来讲模型设置在共享内存中以进行进程间通信。不过注意,如果我们采用GPU训练,则GPU直接就做为了多进程的共享内存,此时model.share_memory()
实际上为空操作(no-op)。
我们用Pytorch实现的训练代码如下:
from __future__ import print_function
import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.multiprocessing as mp
from torchvision import datasets, transforms
import os
import torch
import torch.optim as optim
import torch.nn.functional as F
batch_size = 64 # input batch size for training
test_batch_size = 1000 # input batch size for testing
epochs = 10 # number of global epochs to train
lr = 0.01 # learning rate
momentum = 0.5 # SGD momentum
seed = 1 # random seed
log_interval = 10 # how many batches to wait before logging training status
n_workers = 4 # how many training processes to use
cuda = True # enables CUDA training
mps = False # enables macOS GPU training
dry_run = False # quickly check a single pass
def train(rank, model, device, dataset, dataloader_kwargs):
torch.manual_seed(seed + rank)
train_loader = torch.utils.data.DataLoader(dataset, **dataloader_kwargs)
optimizer = optim.SGD(model.parameters(), lr=lr, momentum=momentum)
for epoch in range(1, epochs + 1):
model.train()
pid = os.getpid()
for batch_idx, (data, target) in enumerate(train_loader):
optimizer.zero_grad()
output = model(data.to(device))
loss = F.nll_loss(output, target.to(device))
loss.backward()
optimizer.step()
if batch_idx % log_interval == 0:
print('{}\tTrain Epoch: {} [{}/{} ({:.0f}%)]\tLoss: {:.6f}'.format(
pid, epoch, batch_idx * len(data), len(train_loader.dataset),
100. * batch_idx / len(train_loader), loss.item()))
if dry_run:
break
def test(model, device, dataset, dataloader_kwargs):
torch.manual_seed(seed)
test_loader = torch.utils.data.DataLoader(dataset, **dataloader_kwargs)
model.eval()
test_loss = 0
correct = 0
with torch.no_grad():
for data, target in test_loader:
output = model(data.to(device))
test_loss += F.nll_loss(output, target.to(device), reduction='sum').item() # sum up batch loss
pred = output.max(1)[1] # get the index of the max log-probability
correct += pred.eq(target.to(device)).sum().item()
test_loss /= len(test_loader.dataset)
print('\nTest set: Global loss: {:.4f}, Accuracy: {}/{} ({:.0f}%)\n'.format(
test_loss, correct, len(test_loader.dataset),
100. * correct / len(test_loader.dataset)))
class Net(nn.Module):
def __init__(self):
super(Net, self).__init__()
self.conv1 = nn.Conv2d(1, 10, kernel_size=5)
self.conv2 = nn.Conv2d(10, 20, kernel_size=5)
self.conv2_drop = nn.Dropout2d()
self.fc1 = nn.Linear(320, 50)
self.fc2 = nn.Linear(50, 10)
def forward(self, x):
x = F.relu(F.max_pool2d(self.conv1(x), 2))
x = F.relu(F.max_pool2d(self.conv2_drop(self.conv2(x)), 2))
x = x.view(-1, 320)
x = F.relu(self.fc1(x))
x = F.dropout(x, training=self.training)
x = self.fc2(x)
return F.log_softmax(x, dim=1)
if __name__ == '__main__':
use_cuda = cuda and torch.cuda.is_available()
use_mps = mps and torch.backends.mps.is_available()
if use_cuda:
device = torch.device("cuda")
elif use_mps:
device = torch.device("mps")
else:
device = torch.device("cpu")
print(device)
transform=transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.1307,), (0.3081,))
])
train_dataset = datasets.MNIST('../data', train=True, download=True,
transform=transform)
test_dataset = datasets.MNIST('../data', train=False,
transform=transform)
kwargs = {'batch_size': batch_size,
'shuffle': True}
if use_cuda:
kwargs.update({'num_workers': 1,
'pin_memory': True,
})
torch.manual_seed(seed)
mp.set_start_method('spawn', force=True)
model = Net().to(device)
model.share_memory() # gradients are allocated lazily, so they are not shared here
processes = []
for rank in range(n_workers):
p = mp.Process(target=train, args=(rank, model, device,
train_dataset, kwargs))
# We first train the model across `n_workers` processes
p.start()
processes.append(p)
for p in processes:
p.join()
# Once training is complete, we can test the model
test(model, device, test_dataset, kwargs)
运行得到的评估结果为:
Test set: Global loss: 0.0325, Accuracy: 9898/10000 (99%)
可见该训练算法是收敛的。
3.2 收敛性分析
当采用不带锁的多线程的写入(即在更新\(w_j\)的时候不用先获取对\(w_j\)的访问权限),而这可能会导致导致同步错误[10]的问题。比如在线程\(1\)加载全局参数\(w^t_j\)之后,线程\(2\)还没等线程\(1\)存储全局参数更新后的值,就也对全局参数\(w^t_j\)进行加载,这样导致每个线程都会存储值为\(w^t_j - \eta^t g(w^t_j)\)的更新后的全局参数值,这样就导致其中一个线程的更新实际上在做“无用功”。直观的感觉是这应该会对学习的过程产生负面影响。不过,当我们对模型访问的稀疏性(sparity)做一定的限定后,这种访问冲突实际上是非常有限的。这正是Hogwild!算法收敛性存在的理论依据。
假设我们要最小化的损失函数为\(l: \mathcal{W}\rightarrow \mathbb{R}\),对于特定的训练样本集合,损失函数\(l\)是由一系列稀疏子函数组合而来的:
也就是说,实际的学习过程中,每个训练样本涉及的参数组合\(e\)只是全体参数集合中的一个很小的子集。我们可以用一个超图\(G=(V, E)\)来表述这个学习过程中参数和参数之间的关系,其中节点\(v\)表示参数,而超边\(e\)表示训练样本涉及的参数组合。那么,稀疏性可以用下面几个统计量加以表示:
其中,\(\Omega\)表达了最大超边的大小,也就是单个样本最多涉及的参数个数;\(\Delta\)反映的是一个参数最多可以涉及多少个不同的超边;而\(\rho\)则反映了给定任意一个超边,与其共享参数的超边个数。这三个值的取值越小,则优化问题越稀疏。在\(\Omega\)、\(\Delta\)、\(\rho\)都比较小的条件下,Hogwild!算法的收敛性保证还需要假设损失函数是凸函数,并且是Lipschitz连续的,详细的理论证明和定量关系请参考文献[2]。
参考
-
[1] Agarwal A, Duchi J C. Distributed delayed stochastic optimization[J]. Advances in neural information processing systems, 2011, 24.
-
[2] Recht B, Re C, Wright S, et al. Hogwild!: A lock-free approach to parallelizing stochastic gradient descent[J]. Advances in neural information processing systems, 2011, 24.
-
[3] 刘浩洋,户将等. 最优化:建模、算法与理论[M]. 高教出版社, 2020.
-
[4] 刘铁岩,陈薇等. 分布式机器学习:算法、理论与实践[M]. 机械工业出版社, 2018.
-
[5] Stanford CME 323: Distributed Algorithms and Optimization (Lecture 7)
-
[6] Bryant R E等.《深入理解计算机系统》[M]. 机械工业出版社, 2015.