mnist_hf.py

import argparse
import os
import time

import torch
import torch.nn.functional as F
import torch.optim as optim
from accelerate import Accelerator
from accelerate.state import AcceleratorState
from torch.optim.lr_scheduler import StepLR
from torchvision import datasets, transforms

from net import Net

accelerator = Accelerator()
local_rank = int(os.environ.get("LOCAL_RANK", -1))


def train(args, model, device, train_loader, optimizer, epoch):
    model.train()
    for batch_idx, (data, target) in enumerate(train_loader):
        optimizer.zero_grad()
        output = model(data)
        loss = F.nll_loss(output, target)
        accelerator.backward(loss)
        optimizer.step()
        if batch_idx % args.log_interval == 0:
            if accelerator.is_main_process:
                print(
                    "Train Epoch: {} [{}/{} ({:.0f}%)]\tLoss: {:.6f}".format(
                        epoch,
                        AcceleratorState().num_processes * batch_idx * len(data),
                        len(train_loader.dataset),
                        100.0 * batch_idx / len(train_loader),
                        loss.item(),
                    )
                )
            if args.dry_run:
                break


def test(model, device, test_loader):
    model.eval()
    test_loss = 0
    correct = 0
    with torch.no_grad():
        for data, target in test_loader:
            data, target = data.to(device), target.to(device)
            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 /= len(test_loader.dataset)

    if accelerator.is_main_process:
        print('\nTest set: Average loss: {:.4f}, Accuracy: {}/{} ({:.0f}%)\n'.format(
            test_loss, correct, len(test_loader.dataset),
            100. * correct / len(test_loader.dataset)))


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=14, metavar='N',
                        help='number of epochs to train (default: 14)')
    parser.add_argument('--lr', type=float, default=1.0, metavar='LR',
                        help='learning rate (default: 1.0)')
    parser.add_argument('--gamma', type=float, default=0.7, metavar='M',
                        help='Learning rate step gamma (default: 0.7)')
    parser.add_argument('--no-cuda', action='store_true', default=False,
                        help='disables CUDA training')
    parser.add_argument('--no-mps', action='store_true', default=False,
                        help='disables macOS GPU training')
    parser.add_argument('--dry-run', action='store_true', default=False,
                        help='quickly check a single pass')
    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')
    args = parser.parse_args()
    use_cuda = not args.no_cuda and torch.cuda.is_available()

    torch.manual_seed(args.seed)

    if use_cuda:
        device = accelerator.device
    else:
        device = torch.device("cpu")

    train_kwargs = {'batch_size': args.batch_size}
    test_kwargs = {'batch_size': args.test_batch_size}
    if use_cuda:
        cuda_kwargs = {'num_workers': 1,
                       'pin_memory': True,
                       'shuffle': True}
        train_kwargs.update(cuda_kwargs)
        test_kwargs.update(cuda_kwargs)

    transform = transforms.Compose([
        transforms.ToTensor(),
        transforms.Normalize((0.1307,), (0.3081,))
    ])

    if not accelerator.is_main_process:
        # might be downloading mnist data, let rank 0 download first
        accelerator.wait_for_everyone()

    dataset1 = datasets.MNIST('./data', train=True, download=True, transform=transform)

    if accelerator.is_main_process:
        # mnist data is downloaded, indicate other ranks can proceed
        accelerator.wait_for_everyone()

    dataset2 = datasets.MNIST('./data', train=False, transform=transform)
    train_loader = torch.utils.data.DataLoader(dataset1, **train_kwargs)
    test_loader = torch.utils.data.DataLoader(dataset2, **test_kwargs)

    model = Net().to(device)
    optimizer = optim.Adadelta(model.parameters(), lr=args.lr)
    scheduler = StepLR(optimizer, step_size=1, gamma=args.gamma)

    model, optimizer, train_loader, scheduler = accelerator.prepare(
        model, optimizer, train_loader, scheduler
    )

    total_time = 0.

    for epoch in range(1, args.epochs + 1):
        start = time.time()
        train(args, model, device, train_loader, optimizer, epoch)
        total_time += time.time() - start
        test(model, device, test_loader)
        scheduler.step()

    if args.save_model:
        torch.save(model.state_dict(), "mnist_cnn.pt")

    return total_time


if __name__ == '__main__':
    print(f'[{local_rank}] Total time elapsed: {main()} seconds')