Les traductions sont fournies par des outils de traduction automatique. En cas de conflit entre le contenu d'une traduction et celui de la version originale en anglais, la version anglaise prévaudra.

Exécution d'une tâche d'entraînement distribué avec parallélisme de modèles SageMaker

Dans cette section, vous allez apprendre :

Pour en savoir plus sur les modules smdistributed.modelparallel, consultez SageMaker model parallel APIs (API de parallélisme de modèles SageMaker) dans la documentation du kit SDK Python SageMaker.

Parallélisme de tenseur seul

Voici un exemple d'option d'entraînement distribué permettant d'activer uniquement le parallélisme de tenseur, sans parallélisme de pipeline. Configurez les dictionnaires mpi_options et smp_options pour spécifier des options d'entraînement distribué pour l'estimateur PyTorch SageMaker.

Configuration d'un estimateur PyTorch SageMaker

mpi_options = {
    "enabled" : True,
    "processes_per_host" : 8,               # 8 processes
    "custom_mpi_options" : "--mca btl_vader_single_copy_mechanism none "
}
               
smp_options = {
    "enabled":True,
    "parameters": {
        "pipeline_parallel_degree": 1,    # alias for "partitions"
        "placement_strategy": "cluster",
        "tensor_parallel_degree": 4,      # tp over 4 devices
        "ddp": True
    }
}
              
smp_estimator = PyTorch(
    entry_point='your_training_script.py', # Specify
    role=role,
    instance_type='ml.p3.16xlarge',
    sagemaker_session=sagemaker_session,
    framework_version='1.13.1',
    py_version='py36',
    instance_count=1,
    distribution={
        "smdistributed": {"modelparallel": smp_options},
        "mpi": mpi_options
    },
    base_job_name="SMD-MP-demo",
)

smp_estimator.fit('s3://my_bucket/my_training_data/')

Adaptation du script d'entraînement PyTorch

L'exemple de script d'entraînement suivant montre comment adapter la bibliothèque de parallélisme de modèles SageMaker à un script d'entraînement. Dans cet exemple, on suppose que le script est nommé your_training_script.py.

import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
from torchnet.dataset import SplitDataset
from torchvision import datasets

import smdistributed.modelparallel.torch as smp

class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.conv1 = nn.Conv2d(1, 32, 3, 1)
        self.conv2 = nn.Conv2d(32, 64, 3, 1)
        self.fc1 = nn.Linear(9216, 128)
        self.fc2 = nn.Linear(128, 10)

    def forward(self, x):
        x = self.conv1(x)
        x = F.relu(x)
        x = self.conv2(x)
        x = F.relu(x)
        x = F.max_pool2d(x, 2)
        x = torch.flatten(x, 1)
        x = self.fc1(x)
        x = F.relu(x)
        x = self.fc2(x)
        return F.log_softmax(x, 1)

def train(model, device, train_loader, optimizer):
    model.train()
    for batch_idx, (data, target) in enumerate(train_loader):
        # smdistributed: Move input tensors to the GPU ID used by
        # the current process, based on the set_device call.
        data, target = data.to(device), target.to(device)
        optimizer.zero_grad()
        output = model(data)
        loss = F.nll_loss(output, target, reduction="mean")
        loss.backward()
        optimizer.step()

# smdistributed: Initialize the backend
smp.init()

# smdistributed: Set the device to the GPU ID used by the current process.
# Input tensors should be transferred to this device.
torch.cuda.set_device(smp.local_rank())
device = torch.device("cuda")

# smdistributed: Download only on a single process per instance.
# When this is not present, the file is corrupted by multiple processes trying
# to download and extract at the same time
if smp.local_rank() == 0:
    dataset = datasets.MNIST("../data", train=True, download=False)
smp.barrier()

# smdistributed: Shard the dataset based on data parallel ranks
if smp.dp_size() > 1:
    partitions_dict = {f"{i}": 1 / smp.dp_size() for i in range(smp.dp_size())}
    dataset = SplitDataset(dataset, partitions=partitions_dict)
    dataset.select(f"{smp.dp_rank()}")

train_loader = torch.utils.data.DataLoader(dataset, batch_size=64)

# smdistributed: Enable tensor parallelism for all supported modules in the model
# i.e., nn.Linear in this case. Alternatively, we can use
# smp.set_tensor_parallelism(model.fc1, True)
# to enable it only for model.fc1
with smp.tensor_parallelism():
    model = Net()

# smdistributed: Use the DistributedModel wrapper to distribute the
# modules for which tensor parallelism is enabled
model = smp.DistributedModel(model)

optimizer = optim.AdaDelta(model.parameters(), lr=4.0)
optimizer = smp.DistributedOptimizer(optimizer)

train(model, device, train_loader, optimizer)

Parallélisme de tenseur associé au parallélisme de pipeline

Voici un exemple d'option d'entraînement distribué qui active le parallélisme de tenseur combiné au parallélisme de pipeline. Configurez les paramètres mpi_options et smp_options pour spécifier les options de parallélisme de modèles avec un parallélisme de tenseur lorsque vous configurez un estimateur PyTorch SageMaker.

Configuration d'un estimateur PyTorch SageMaker

mpi_options = {
    "enabled" : True,
    "processes_per_host" : 8,               # 8 processes
    "custom_mpi_options" : "--mca btl_vader_single_copy_mechanism none "
}
               
smp_options = {
    "enabled":True,
    "parameters": {
    "microbatches": 4,
        "pipeline_parallel_degree": 2,    # alias for "partitions"
        "placement_strategy": "cluster",
        "tensor_parallel_degree": 2,      # tp over 2 devices
        "ddp": True
    }
}
              
smp_estimator = PyTorch(
    entry_point='your_training_script.py', # Specify
    role=role,
    instance_type='ml.p3.16xlarge',
    sagemaker_session=sagemaker_session,
    framework_version='1.13.1',
    py_version='py36',
    instance_count=1,
    distribution={
        "smdistributed": {"modelparallel": smp_options},
        "mpi": mpi_options
    },
    base_job_name="SMD-MP-demo",
)

smp_estimator.fit('s3://my_bucket/my_training_data/')  

Adaptation du script d'entraînement PyTorch

L'exemple de script d'entraînement suivant montre comment adapter la bibliothèque de parallélisme de modèles SageMaker à un script d'entraînement. Notez que le script d'entraînement inclut désormais le décorateur smp.step :

import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
from torchnet.dataset import SplitDataset
from torchvision import datasets

import smdistributed.modelparallel.torch as smp

class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.conv1 = nn.Conv2d(1, 32, 3, 1)
        self.conv2 = nn.Conv2d(32, 64, 3, 1)
        self.fc1 = nn.Linear(9216, 128)
        self.fc2 = nn.Linear(128, 10)

    def forward(self, x):
        x = self.conv1(x)
        x = F.relu(x)
        x = self.conv2(x)
        x = F.relu(x)
        x = F.max_pool2d(x, 2)
        x = torch.flatten(x, 1)
        x = self.fc1(x)
        x = F.relu(x)
        x = self.fc2(x)
        return F.log_softmax(x, 1)


# smdistributed: Define smp.step. Return any tensors needed outside.
@smp.step
def train_step(model, data, target):
    output = model(data)
    loss = F.nll_loss(output, target, reduction="mean")
    model.backward(loss)
    return output, loss

def train(model, device, train_loader, optimizer):
    model.train()
    for batch_idx, (data, target) in enumerate(train_loader):
        # smdistributed: Move input tensors to the GPU ID used by
        # the current process, based on the set_device call.
        data, target = data.to(device), target.to(device)
        optimizer.zero_grad()
        # Return value, loss_mb is a StepOutput object
        _, loss_mb = train_step(model, data, target)

        # smdistributed: Average the loss across microbatches.
        loss = loss_mb.reduce_mean()

        optimizer.step()

# smdistributed: Initialize the backend
smp.init()

# smdistributed: Set the device to the GPU ID used by the current process.
# Input tensors should be transferred to this device.
torch.cuda.set_device(smp.local_rank())
device = torch.device("cuda")

# smdistributed: Download only on a single process per instance.
# When this is not present, the file is corrupted by multiple processes trying
# to download and extract at the same time
if smp.local_rank() == 0:
    dataset = datasets.MNIST("../data", train=True, download=False)
smp.barrier()

# smdistributed: Shard the dataset based on data parallel ranks
if smp.dp_size() > 1:
    partitions_dict = {f"{i}": 1 / smp.dp_size() for i in range(smp.dp_size())}
    dataset = SplitDataset(dataset, partitions=partitions_dict)
    dataset.select(f"{smp.dp_rank()}")

# smdistributed: Set drop_last=True to ensure that batch size is always divisible
# by the number of microbatches
train_loader = torch.utils.data.DataLoader(dataset, batch_size=64, drop_last=True)

model = Net()

# smdistributed: enable tensor parallelism only for model.fc1
smp.set_tensor_parallelism(model.fc1, True)

# smdistributed: Use the DistributedModel container to provide the model
# to be partitioned across different ranks. For the rest of the script,
# the returned DistributedModel object should be used in place of
# the model provided for DistributedModel class instantiation.
model = smp.DistributedModel(model)

optimizer = optim.AdaDelta(model.parameters(), lr=4.0)
optimizer = smp.DistributedOptimizer(optimizer)

train(model, device, train_loader, optimizer)