pytorch.反向传播算法和优化器

在训练神经网络时，最常见的算法就是反向传播

为了支持反向传播，pytorch有一个内置的分类引擎，叫做TORCH.AUTOGRAD

import torch

x = torch.ones(5)  # input tensor
y = torch.zeros(3)  # expected output
w = torch.randn(5, 3, requires_grad=True)  # 如果需要反向传播就打开这个参数
b = torch.randn(3, requires_grad=True)  # 如果需要反向传播就打开这个参数
z = torch.matmul(x, w) + b
# print(f"x:{x}\ny:{y}\nw:{w}\nb:{b}\nz:{z}\n")

loss = torch.nn.functional.binary_cross_entropy_with_logits(z, y)
# print(loss)

print(f"Gradient function for z = {z.grad_fn}")
print(f"Gradient function for loss = {loss.grad_fn}")

loss.backward()
print(w.grad)
print(b.grad)

# 怎么样把反向传播的选项关掉呢，为什么要关掉，当我们训练好了网络，不在需要训练网络时，可以关掉
z = torch.matmul(x, w) + b
print(z.requires_grad)

with torch.no_grad():
    z = torch.matmul(x, w) + b
print(z.requires_grad)

# 也可以对tensor下手
z = torch.matmul(x, w) + b
z_det = z.detach()
print(z_det.requires_grad)

inp = torch.eye(5, requires_grad=True)
out = (inp+1).pow(2)
out.backward(torch.ones_like(inp), retain_graph=True)
print(f"First call\n{inp.grad}")
out.backward(torch.ones_like(inp), retain_graph=True)
print(f"\nSecond call\n{inp.grad}")
inp.grad.zero_()
out.backward(torch.ones_like(inp), retain_graph=True)
print(f"\nCall after zeroing gradients\n{inp.grad}")

---

我们首先需要创建一个神经网络，并导入一些训练数据

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(f"Using {device} device")

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)


class NeuralNetwork(nn.Module):
    def __init__(self):
        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),
        )

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


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

# 学习率
learning_rate = 1e-3
# 每次导入的数据量
batch_size = 64
# 训练轮数
epochs = 5

常见的损失函数包括nn.MSELoss(均方误差)

nn.NLLLoss(负对数，用于分类问题)

nn.CrossEntropyLoss(结合了softmax和NULLLoss)

优化器Optimizer

优化是每次训练过程中调整参数减小模型误差的过程，优化算法定义了这个过程是怎么进行的，我们使用梯度下降法

optimizer = torch.optim.SGD(model.parameters(), lr=learning_rate)

def train_loop(dataloader, model, loss_fn, optimizer):
    size = len(dataloader.dataset)
    for batch, (X, y) in enumerate(dataloader):
        X = X.to(device)
        y = y.to(device)
        # Compute prediction and loss
        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}]")


def test_loop(dataloader, model, loss_fn):
    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")

# 损失函数
loss_fn = nn.CrossEntropyLoss()
# 优化器
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!")