mnist.py

from __future__ import print_function
import argparse
import os
import subprocess
import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
from torchvision import datasets, transforms
from torch.utils.data import Dataset, DataLoader
from torch.utils.data.distributed import DistributedSampler
from torch.nn.parallel import DistributedDataParallel


def dist_init(host_addr, rank, local_rank, world_size, port=23456):
    host_addr_full = 'tcp://' + host_addr + ':' + str(port)
    torch.distributed.init_process_group("gloo", init_method=host_addr_full,
                                         rank=rank, world_size=world_size)
    assert torch.distributed.is_initialized()

class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.conv1 = nn.Conv2d(1, 20, 5, 1)
        self.conv2 = nn.Conv2d(20, 50, 5, 1)
        self.fc1 = nn.Linear(4*4*50, 500)
        self.fc2 = nn.Linear(500, 10)

    def forward(self, x):
        x = F.relu(self.conv1(x))
        x = F.max_pool2d(x, 2, 2)
        x = F.relu(self.conv2(x))
        x = F.max_pool2d(x, 2, 2)
        x = x.view(-1, 4*4*50)
        x = F.relu(self.fc1(x))
        x = self.fc2(x)
        return F.log_softmax(x, dim=1)

def train(args, model, local_rank, train_loader, optimizer, epoch):
    model.train()
    for batch_idx, (data, target) in enumerate(train_loader):
        data, target = data.to(local_rank), target.to(local_rank)
        optimizer.zero_grad()
        output = model(data)
        loss = F.nll_loss(output, target)
        loss.backward()
        optimizer.step()
        if batch_idx % args.log_interval == 0:
            print('Train Epoch: {} [{}/{} ({:.0f}%)]\tLoss: {:.6f}'.format(
                epoch, batch_idx * len(data), len(train_loader.dataset),
                100. * batch_idx / len(train_loader), loss.item()))

def test(args, model, local_rank, test_loader, world_size):
    model.eval()
    test_loss = 0
    correct = 0
    length = len(test_loader.dataset)/world_size
    with torch.no_grad():
        for data, target in test_loader:
            data, target = data.to(local_rank), target.to(local_rank)
            output = model(data)
            test_loss += F.nll_loss(output, target, reduction='sum').item() # sum up batch loss
            pred = output.argmax(dim=1, keepdim=True) # get the index of the max log-probability
            correct += pred.eq(target.view_as(pred)).sum().item()

    test_loss /= length

    print('\nTest set: Average loss: {:.4f}, Accuracy: {}/{} ({:.0f}%)\n'.format(
        test_loss, correct, length,
        100. * correct / length))

def main():

    # Training settings
    parser = argparse.ArgumentParser(description='PyTorch MNIST Example')
    parser.add_argument('--batch-size', type=int, default=64, metavar='N',
                        help='input batch size for training (default: 64)')
    parser.add_argument('--test-batch-size', type=int, default=1000, metavar='N',
                        help='input batch size for testing (default: 1000)')
    parser.add_argument('--epochs', type=int, default=10, metavar='N',
                        help='number of epochs to train (default: 10)')
    parser.add_argument('--lr', type=float, default=0.01, metavar='LR',
                        help='learning rate (default: 0.01)')
    parser.add_argument('--momentum', type=float, default=0.5, metavar='M',
                        help='SGD momentum (default: 0.5)')
    parser.add_argument('--no-cuda', action='store_true', default=False,
                        help='disables CUDA training')
    parser.add_argument('--seed', type=int, default=1, metavar='S',
                        help='random seed (default: 1)')
    parser.add_argument('--log-interval', type=int, default=10, metavar='N',
                        help='how many batches to wait before logging training status')
    parser.add_argument('--save-model', action='store_true', default=False,
                        help='For Saving the current Model')
    parser.add_argument('--datasetDir',
                        help='Please add your dataset directory')

    args = parser.parse_args()

    # use_cuda = not args.no_cuda and torch.cuda.is_available()
    # torch.manual_seed(args.seed)
    # device = torch.device("cuda" if use_cuda else "cpu")
    # kwargs = {'num_workers': 1, 'pin_memory': True} if use_cuda else {}

    rank = int(os.environ['SLURM_PROCID'])
    local_rank = int(os.environ['SLURM_LOCALID'])
    world_size = int(os.environ['SLURM_NTASKS'])
    iplist = os.environ['SLURM_JOB_NODELIST']
    ip = subprocess.getoutput(f"scontrol show hostname {iplist} | head -n1")

    dist_init(ip, rank, local_rank, world_size)
    train_dataset = datasets.MNIST(args.datasetDir, train=True, download=False,
                       transform=transforms.Compose([
                           transforms.ToTensor(),
                           transforms.Normalize((0.1307,), (0.3081,))
                       ]))
    train_sampler = DistributedSampler(train_dataset, num_replicas=world_size, rank=rank)
    train_loader = torch.utils.data.DataLoader(
        train_dataset,
        batch_size=args.batch_size,
        sampler=train_sampler
        )
    test_dataset = datasets.MNIST(args.datasetDir, train=False, transform=transforms.Compose([
                           transforms.ToTensor(),
                           transforms.Normalize((0.1307,), (0.3081,))
                       ]))
    test_sampler = DistributedSampler(test_dataset, num_replicas=world_size, rank=rank)

    test_loader = torch.utils.data.DataLoader(
        test_dataset,
        batch_size=args.test_batch_size,
        sampler=test_sampler
        )

    model = Net().to(local_rank)
    model = DistributedDataParallel(model)
    optimizer = optim.SGD(model.parameters(), lr=args.lr, momentum=args.momentum)

    for epoch in range(1, args.epochs + 1):
        train(args, model, local_rank, train_loader, optimizer, epoch)
        test(args, model, local_rank, test_loader, world_size)

    if (args.save_model):
        torch.save(model.state_dict(),"mnist_cnn.pt")
if __name__ == "__main__":
    main()