保存和加载模型

import torch
from torch import nn
from torch.utils.data import DataLoader
from torchvision import datasets
from torchvision.transforms import ToTensor, Lambda

device = 'cuda' if torch.cuda.is_available() else 'cpu'
print(device)

# ======================= 数据 =======================
training_data = datasets.FashionMNIST(
    root="data",
    train=True,
    download=True,
    transform=ToTensor()
)

test_data = datasets.FashionMNIST(
    root="data",
    train=False,
    download=True,
    transform=ToTensor()
)

train_dataloader = DataLoader(training_data, batch_size=64)
test_dataloader = DataLoader(test_data, batch_size=64)


# ======================= 模型 =======================
class NeuralNetwork(nn.Module):
    def __init__(self):
        super(NeuralNetwork, self).__init__()

        self.linear_relu_stack = nn.Sequential(
            nn.Linear(28 * 28, 512),
            nn.ReLU(),
            nn.Linear(512, 512),
            nn.ReLU(),
            nn.Linear(512, 10),
        )

    def forward(self, x):
        x = x.flatten(1)
        logits = self.linear_relu_stack(x)
        return logits


def train_loop(dataloader, model, loss_fn, optimizer):
    '''训练循环'''
    size = len(dataloader.dataset)
    for batch, (X, y) in enumerate(dataloader):
        X = X.to(device)
        y = y.to(device)
        # 计算估计值与损失
        pred = model(X)
        loss = loss_fn(pred, y)

        # 反向传播
        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}]")

def test_loop(dataloader, model, loss_fn):
    '''测试循环'''
    size = len(dataloader.dataset)
    num_batches = len(dataloader)
    test_loss, correct = 0, 0

    with torch.no_grad():
        for X, y in dataloader:
            X = X.to(device)
            y = 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 /= num_batches
    correct /= size
    print(f"Test Error: \n Accuracy: {(100 * correct):>0.1f}%, Avg loss: {test_loss:>8f} \n")

# 超参数 
learning_rate = 1e-3
batch_size = 64
epochs = 10

# 模型实例
model = NeuralNetwork().to(device)
# 损失函数实例
loss_fn = nn.CrossEntropyLoss()
# 优化器实例
optimizer = torch.optim.SGD(model.parameters(), lr=learning_rate)

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

在之前的章节中，我们介绍了如何用Pytorch搭建模型以及训练。在本节中，我们将了解到如何保存、加载模型。

保存加载模型参数

PyTorch模型可以将学习到的参数存储在内部状态字典（称为state_dict）中。我们可以通过torch.save保存这些参数：

torch.save(model.state_dict(), "model_state_dict.pth")

要加载模型参数，需要先创建同一模型的实例，然后使用load_state_dict()方法加载。

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

使用加载的模型进行预测：

# model.eval()
classes = [
    "T-shirt/top",
    "Trouser",
    "Pullover",
    "Dress",
    "Coat",
    "Sandal",
    "Shirt",
    "Sneaker",
    "Bag",
    "Ankle boot",
]
x, y = test_data[0][0], test_data[0][1]
with torch.no_grad():
    x = x.to(device)
    pred = model(x)
    predicted, actual = classes[pred[0].argmax(0)], classes[y]
    print(f'Predicted: "{predicted}", Actual: "{actual}"')

输出：

Predicted: "Ankle boot", Actual: "Ankle boot"

注意：
如果在训练模型时使用到dropout或batch normalization，则需要使用model.eval()方法将网络设置为测试模式，否则将产生不一致的推断结果。

保存加载模型和参数

加载模型参数时，我们需要首先实例化模型类，因为该类定义了网络的结构。我们还可以将此类的结构与模型一起保存：将model（而不是model.state_dict()）传递给torch.save：

torch.save(model, 'model.pth')

然后我们可以像这样加载模型：

model = torch.load("model.pth")
# model.eval()
classes = [
    "T-shirt/top",
    "Trouser",
    "Pullover",
    "Dress",
    "Coat",
    "Sandal",
    "Shirt",
    "Sneaker",
    "Bag",
    "Ankle boot",
]
x, y = test_data[0][0], test_data[0][1]
with torch.no_grad():
    x = x.to(device)
    pred = model(x)
    predicted, actual = classes[pred[0].argmax(0)], classes[y]
    print(f'Predicted: "{predicted}", Actual: "{actual}"')

Predicted: "Ankle boot", Actual: "Ankle boot"

参考：
[1] https://pytorch.org/tutorials/beginner/basics/saveloadrun_tutorial.html