single-machine-and-multi-GPU-DataParallel.py

import torch
from torch import nn
from torch.utils.data import DataLoader
from torchvision import datasets
from torchvision.transforms import ToTensor
from model import NeuralNetwork


def train(dataloader, model, loss_fn, optimizer, device):
    size = len(dataloader.dataset)
    model.train()
    for batch, (X, y) in enumerate(dataloader):
        X, y = X.to(device), y.to(device)  # copy data from cpu to gpu

        # Compute prediction error
        pred = model(X)
        loss = loss_fn(pred, y)

        # Backpropagation
        optimizer.zero_grad()
        loss.backward()
        optimizer.step()

        if batch % 100 == 0:
            loss, current = loss.item(), batch * len(X)
            print(f"loss: {loss:>7f}  [{current:>5d}/{size:>5d}]")


def test(dataloader, model, loss_fn, device):
    size = len(dataloader.dataset)
    num_batches = len(dataloader)
    model.eval()
    test_loss, correct = 0, 0
    with torch.no_grad():
        for X, y in dataloader:
            X, y = X.to(device), y.to(device)  # copy data from cpu to gpu
            pred = model(X)
            test_loss += loss_fn(pred, y).item()
            correct += (pred.argmax(1) == y).type(torch.float).sum().item()
    test_loss /= num_batches
    correct /= size
    print(f"Test Error: \n Accuracy: {(100*correct):>0.1f}%, Avg loss: {test_loss:>8f} \n")


if __name__ == '__main__':
    # initialize dataset
    training_data = datasets.FashionMNIST(root="data", train=True, download=True, transform=ToTensor())
    test_data = datasets.FashionMNIST(root="data", train=False, download=True, transform=ToTensor())

    # initialize data loader
    batch_size = 64
    train_dataloader = DataLoader(training_data, batch_size=batch_size)
    test_dataloader = DataLoader(test_data, batch_size=batch_size)

    # [*] Get multiple GPU device for training.
    n_gpu = torch.cuda.device_count()
    device = torch.device('cuda:0' if n_gpu > 0 else 'cpu')
    device_ids = list(range(n_gpu))
    print("n_gpu: {}".format(n_gpu))

    # initialize model
    model = NeuralNetwork().to(device)  # copy model from cpu to gpu
    # [*] copy model to multi-GPU
    if len(device_ids) > 1:
        model = torch.nn.DataParallel(model, device_ids=device_ids)
    print(model)

    # initialize optimizer
    loss_fn = nn.CrossEntropyLoss()
    optimizer = torch.optim.SGD(model.parameters(), lr=1e-3)

    # train on multiple-GPU
    epochs = 5
    for t in range(epochs):
        print(f"Epoch {t + 1}\n-------------------------------")
        train(train_dataloader, model, loss_fn, optimizer, device)
        test(test_dataloader, model, loss_fn, device)
    print("Done!")

    # save model
    # [*] save model with multi-GPU
    if isinstance(model, torch.nn.DataParallel):
        model_state_dict = model.module.state_dict()
    else:
        model_state_dict = model.state_dict()
    torch.save(model_state_dict, "model.pth")
    print("Saved PyTorch Model State to model.pth")