utils.py

import torch
import torchvision
import torchvision.transforms as transforms

def get_dataloader(train):
    transform_train = transforms.Compose([
        transforms.RandomCrop(32, padding=4),
        transforms.RandomHorizontalFlip(),
        transforms.RandAugment(num_ops=2, magnitude=9),
        transforms.ToTensor(),
        transforms.Normalize((0.4914, 0.4822, 0.4465), (0.2023, 0.1994, 0.2010)),
        transforms.RandomErasing(p=0.25, scale=(0.02, 0.2), ratio=(0.3, 3.3)),
    ])

    transform_test = transforms.Compose([
        transforms.ToTensor(),
        transforms.Normalize((0.4914, 0.4822, 0.4465), (0.2023, 0.1994, 0.2010)),
    ])
    
    transform = transform_train if train else transform_test
    
    dataset = torchvision.datasets.CIFAR100(
        root='./data', train=train, download=True, transform=transform)
    dataloader = torch.utils.data.DataLoader(
        dataset, batch_size=128, shuffle=train, num_workers=2)
    
    return dataloader