CIFAR10 image classification using CNN with pytorch gpu. This is a simple network and accuracy reaches to 77% on 10 epochs.

PyTorchImageClassificaitonGPU.py

import torch
torch.manual_seed(0)
import numpy as np
np.random.seed(0)
import random
random.seed(0)
torch.backends.cudnn.deterministic = True
torch.backends.cudnn.benchmark = False

#torch.cudnn.benchmark = True
#torch.cudnn.enabled = True


import torchvision
import torchvision.transforms as transforms
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim



if __name__ == '__main__':

    device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
    print(device)

    transform = transforms.Compose(
        [transforms.ToTensor(),
         transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))])

    trainset = torchvision.datasets.CIFAR10(root='./data', train=True,
                                            download=True, transform=transform)
    trainloader = torch.utils.data.DataLoader(trainset, batch_size=4,
                                              shuffle=True, num_workers=4)

    testset = torchvision.datasets.CIFAR10(root='./data', train=False,
                                           download=True, transform=transform)
    testloader = torch.utils.data.DataLoader(testset, batch_size=4,
                                             shuffle=False, num_workers=4)

    classes = ('plane', 'car', 'bird', 'cat',
               'deer', 'dog', 'frog', 'horse', 'ship', 'truck')




    l1 = 64
    l2 = 128
    l3 = 256



    class Net(nn.Module):
        def __init__(self):
            super(Net, self).__init__()


            # input channel, output filter, kernel size
            self.conv1 = nn.Conv2d(3, l1, 5, padding=2)
            self.conv2 = nn.Conv2d(l1, l2, 5, padding=2)
            self.conv3 = nn.Conv2d(l2, l3, 3)

            self.BatchNorm2d1 = nn.BatchNorm2d(l1)
            self.BatchNorm2d2 = nn.BatchNorm2d(l2)
            self.BatchNorm2d3 = nn.BatchNorm2d(l3)

            self.BatchNorm2d4 = nn.BatchNorm2d(120)



            self.pool = nn.MaxPool2d(2, 2)
            self.dropout = nn.Dropout(p=0)

            self.fc1 = nn.Linear(l3 * 3 * 3, 120)
            self.fc2 = nn.Linear(120, 84)
            self.fc3 = nn.Linear(84, 10)

        def forward(self, x):
            x = self.BatchNorm2d1(self.pool(F.leaky_relu(self.conv1(x))))
            x = self.BatchNorm2d2(self.pool(F.leaky_relu(self.conv2(x))))
            x = self.BatchNorm2d3(self.pool(F.leaky_relu(self.conv3(x))))
            x = x.view(-1, l3 * 3 * 3)
            x = self.dropout(F.leaky_relu(self.fc1(x)))
            x = self.dropout(F.leaky_relu(self.fc2(x)))
            #x = self.dropout(F.leaky_relu(self.fc4(x)))
            x = self.fc3(x)
            return x



    net = Net()
    net = net.to(device)
    print(net)

    criterion = nn.CrossEntropyLoss()
    #optimizer = optim.SGD(net.parameters(), lr=0.001, momentum=0.9)
    #optimizer = optim.Adam(net.parameters(), lr=0.001)
    optimizer = optim.Adam(net.parameters(), lr=0.001)


    def main():

        for epoch in range(5):

            running_loss = 0.0

            for i, data in enumerate(trainloader, 0):
                inputs, labels = data
                inputs, labels = inputs.to(device), labels.to(device)
                optimizer.zero_grad()

                outputs = net(inputs)
                loss = criterion(outputs, labels)

                loss.backward()
                optimizer.step()

                running_loss += loss.item()
                if i % 2000 == 1999:
                    print('[%d, %5d] loss: %.3f' %
                          (epoch + 1, i + 1, running_loss / 2000))
                    running_loss = 0.0
        print('Finished Training')

        dataiter = iter(testloader)
        images, labels = dataiter.next()
        images, labels = images.to(device), labels.to(device)

        print('Ground truth: ', ' '.join('%5s' % classes[labels[j]] for j in
                                         range(4)))

        outputs = net(images)
        _, predicted = torch.max(outputs, 1)
        print('Predicted: ', ' '.join('%5s' % classes[predicted[j]]
                                      for j in range(4)))

        correct = 0
        total = 0
        with torch.no_grad():
            for data in testloader:
                images, labels = data
                images, labels = images.to(device), labels.to(device)
                outputs = net(images)
                _, predicted = torch.max(outputs.data, 1)
                total += labels.size(0)
                correct += (predicted == labels).sum().item()
        print('Accuracy of the network on the 10000 test images: %d %%' % (
                100 * correct / total))

        class_correct = list(0. for i in range(10))
        class_total = list(0. for i in range(10))
        with torch.no_grad():
            for data in testloader:
                images, labels = data
                images, labels = images.to(device), labels.to(device)
                outputs = net(images)
                _, predicted = torch.max(outputs, 1)
                c = (predicted == labels).squeeze()
                for i in range(4):
                    label = labels[i]
                    class_correct[label] += c[i].item()
                    class_total[label] += 1

        for i in range(10):
            print('Accuracy of %5s : %2d %%' % (
                classes[i], 100 * class_correct[i] / class_total[i]))


    main()