0. PyTorch Quickstart

Learn the Basics

Machine Learning Workflow: data → modeling → optimizing parameters → save model

Example data: FashionMNIST (10 classes: T-shirt/top, Trouser, Pullover, Dress, Coat, Sandal, Shirt, Sneaker, Bag, or Ankle boot)

---

Quickstart

Working with Data

(1) `torch.utils.data.Dataset` : stores the samples and their corresponding labels
(2) `torch.utils.data.DataLoader` : wraps an iterable around the `Dataset`

import torch
from torch import nn
from torch.utils.data import DataLoader
from torchvision import datasets
from torchvision.transforms import ToTensor, Lambda, Compose
import matplotlib.pyplot as plt

Some domain-specific libraries: `TorchText`, `TorchVision`, `TorchAudio`

`torchvision.datasets` includes `CIFAR`, `COCO` etc. 

Every TorchVision `Dataset` includes two arguments:
(1) `transform` : modify samples
(2) `target_transform` : modify labels

# Download training data from open datasets.
training_data = datasets.FashionMNIST(
    root="data",
    train=True,
    download=True,
    transform=ToTensor(),
)

# Download test data from open datasets.
test_data = datasets.FashionMNIST(
    root="data",
    train=False,
    download=True,
    transform=ToTensor(),
)

Pass `Dataset` as an argument to `DataLoader` → wraps an iterable over the dataset
(1) automatic batching (2) sampling (3) shuffling (4) multiprocess data loading

Below example, batch_size = 64.

batch_size = 64

# Create data loaders.
train_dataloader = DataLoader(training_data, batch_size=batch_size) # 화려한 로더가 데이터를 감싸네
test_dataloader = DataLoader(test_data, batch_size=batch_size)

for X, y in test_dataloader:
    print("Shape of X [N, C, H, W]: ", X.shape)
    print("Shape of y: ", y.shape, y.dtype)
    break

---

Creating Models

Defining a model → a class that inherits from `nn.Module`.

Layers → `__init__` function
Data flow → `forward` function

※ Use GPU if available.

# Get cpu or gpu device for training.
device = "cuda" if torch.cuda.is_available() else "cpu"
print("Using {} device".format(device))

# Define model
class NeuralNetwork(nn.Module):
    def __init__(self): # 모델 레이어는 여기에 정의
        super(NeuralNetwork, self).__init__()
        self.flatten = nn.Flatten()
        self.linear_relu_stack = nn.Sequential(
            nn.Linear(28*28, 512),
            nn.ReLU(),
            nn.Linear(512, 512),
            nn.ReLU(),
            nn.Linear(512, 10),
            nn.ReLU()
        )

    def forward(self, x): # 정의된 레이어에 데이터 들어감
        x = self.flatten(x)
        logits = self.linear_relu_stack(x)
        return logits

model = NeuralNetwork().to(device)
print(model)

---

Optimizing the Model Parameters

Want to train model? You need:

(1) loss function
(2) optimizer

loss_fn = nn.CrossEntropyLoss()
optimizer = torch.optim.SGD(model.parameters(), lr=1e-3)

1 training loop → (1) batch_size training dataset make prediction (2) backpropagation adjusts parameters

def train(dataloader, model, loss_fn, optimizer):
    size = len(dataloader.dataset)
    for batch, (X, y) in enumerate(dataloader):
        X, y = X.to(device), y.to(device)

        # Compute prediction error
        pred = model(X)
        loss = loss_fn(pred, y)

        # Backpropagation
        optimizer.zero_grad()
        loss.backward()
        optimizer.step()

        if batch % 100 == 0:
            loss, current = loss.item(), batch * len(X)
            print(f"loss: {loss:>7f}  [{current:>5d}/{size:>5d}]")

test function, refer to below.

def test(dataloader, model):
    size = len(dataloader.dataset)
    model.eval()
    test_loss, correct = 0, 0
    with torch.no_grad():
        for X, y in dataloader:
            X, y = X.to(device), y.to(device)
            pred = model(X)
            test_loss += loss_fn(pred, y).item()
            correct += (pred.argmax(1) == y).type(torch.float).sum().item()
    test_loss /= size
    correct /= size
    print(f"Test Error: \n Accuracy: {(100*correct):>0.1f}%, Avg loss: {test_loss:>8f} \n")

Define epoch and print loss.

epochs = 5
for t in range(epochs):
    print(f"Epoch {t+1}\n-------------------------------")
    train(train_dataloader, model, loss_fn, optimizer)
    test(test_dataloader, model)
print("Done!")

So, this is how you can train your model.

---

Saving Models

Done training? Save the model by serializing the internal state dictionary (which contains the model parameters).

torch.save(model.state_dict(), "model.pth")
print("Saved PyTorch Model State to model.pth")

Loading Models

Saved model and want to use it again? Load it.

model = NeuralNetwork()
model.load_state_dict(torch.load("model.pth"))

Use the model for prediction.

classes = [
    "T-shirt/top",
    "Trouser",
    "Pullover",
    "Dress",
    "Coat",
    "Sandal",
    "Shirt",
    "Sneaker",
    "Bag",
    "Ankle boot",
]

model.eval()
x, y = test_data[0][0], test_data[0][1]
with torch.no_grad():
    pred = model(x)
    predicted, actual = classes[pred[0].argmax(0)], classes[y]
    print(f'Predicted: "{predicted}", Actual: "{actual}"')

---

Reference: Quickstart — PyTorch Tutorials 1.8.1+cu102 documentation