practical-02
October 3, 2024
[6]: import matplotlib.pyplot as plt
import numpy as np
import torch
import torchvision
import torchvision.transforms as transforms
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
0.1 Check if GPU is available and use it
[7]: device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
0.2 Transformations
[8]: # Transformations
transform = transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))
])
1 Set parameters
[9]: batch_size = 4
num_workers = 2
2 Load training data and Load test data
[10]: trainset = torchvision.datasets.CIFAR10(root='./data', train=True,␣
↪download=True, transform=transform)
trainloader = torch.utils.data.DataLoader(trainset, batch_size=batch_size,␣
↪shuffle=True, num_workers=num_workers)
1
testset = torchvision.datasets.CIFAR10(root='./data', train=False,␣
↪download=True, transform=transform)
testloader = torch.utils.data.DataLoader(testset, batch_size=batch_size,␣
↪shuffle=False, num_workers=num_workers)
Files already downloaded and verified
Files already downloaded and verified
3 Class names for CIFAR-10
[11]: # Class names for CIFAR-10
classes = ('plane', 'car', 'bird', 'cat', 'deer', 'dog', 'frog', 'horse',␣
↪'ship', 'truck')
4 Function to show images
[12]: # Function to show images
def imshow(img):
img = img / 2 + 0.5 # unnormalize
npimg = img.numpy()
plt.imshow(np.transpose(npimg, (1, 2, 0)))
plt.show()
# Get random training images
dataiter = iter(trainloader)
images, labels = next(dataiter) # Use next() instead of .next()
5 Call function to show images and Define the CNN model
[13]: imshow(torchvision.utils.make_grid(images))
print(' '.join('%s' % classes[labels[j]] for j in range(batch_size)))
class Net(nn.Module):
def __init__(self):
super(Net, self).__init__()
self.conv1 = nn.Conv2d(3, 6, 5)
self.pool = nn.MaxPool2d(2, 2)
self.conv2 = nn.Conv2d(6, 16, 5)
self.fc1 = nn.Linear(16 * 5 * 5, 120)
self.fc2 = nn.Linear(120, 84)
self.fc3 = nn.Linear(84, 10)
def forward(self, x):
2
x = self.pool(F.relu(self.conv1(x)))
x = self.pool(F.relu(self.conv2(x)))
x = x.view(-1, 16 * 5 * 5)
x = F.relu(self.fc1(x))
x = F.relu(self.fc2(x))
x = self.fc3(x)
return x
horse deer truck ship
6 Instantiate the model and move it to the device
[15]: net = Net().to(device)
print(net)
# Define loss function and optimizer
criterion = nn.CrossEntropyLoss()
optimizer = optim.SGD(net.parameters(), lr=0.001, momentum=0.9)
Net(
(conv1): Conv2d(3, 6, kernel_size=(5, 5), stride=(1, 1))
(pool): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1,
ceil_mode=False)
(conv2): Conv2d(6, 16, kernel_size=(5, 5), stride=(1, 1))
(fc1): Linear(in_features=400, out_features=120, bias=True)
(fc2): Linear(in_features=120, out_features=84, bias=True)
(fc3): Linear(in_features=84, out_features=10, bias=True)
)
3
7 Start training
[17]: for epoch in range(2):
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')
[1, 2000] loss: 2.230
[1, 4000] loss: 1.874
[1, 6000] loss: 1.667
[1, 8000] loss: 1.575
[1, 10000] loss: 1.529
[1, 12000] loss: 1.455
[2, 2000] loss: 1.388
[2, 4000] loss: 1.388
[2, 6000] loss: 1.375
[2, 8000] loss: 1.314
[2, 10000] loss: 1.323
[2, 12000] loss: 1.299
Finished Training
8 Test the model
[18]: # Test the model
dataiter = iter(testloader)
images, labels = next(dataiter)
# Move images to CPU for plotting
imshow(torchvision.utils.make_grid(images.cpu()))
print('GroundTruth: ', ' '.join('%s' % classes[labels[j]] for j in range(4)))
4
GroundTruth: cat ship ship plane
