Pytorch官方教程学习笔记6：完整训练一个模型
发布时间
阅读量：
阅读量
数据处理
PyTorch 有两个用于处理数据的基本工具：torch.utils.data.DataLoader 和 torch.utils.data.Dataset。Dataset 用于存储样本及其对应的标签，而 DataLoader 则为 Dataset 提供一个可迭代的包装器
import torch
from torch import nn
from torch.utils.data import DataLoader
from torchvision import datasets
from torchvision.transforms import ToTensor
PyTorch 提供了特定领域的库，例如 TorchText、TorchVision 和 TorchAudio，所有这些库都包含数据集。本教程中，我们将使用一个 TorchVision 数据集。
torchvision.datasets 模块包含许多真实视觉数据的 Dataset 对象，例如 CIFAR 和 COCO（完整列表请查看此处）。在本教程中，我们使用 FashionMNIST 数据集。每个 TorchVision 数据集都包含两个参数：transform 和 target_transform，分别用于修改样本和标签。
# 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(),
)
我们已经下载了数据集FashionMNIST训练集和测试集并分别下载入training_data和test_data。接下来我们将数据集 (Dataset) 作为参数传递给 DataLoader。DataLoader 为数据集提供一个可迭代的包装，并支持自动批处理、采样、打乱顺序以及多进程数据加载。在这里，我们定义了一个批大小为64，即 dataloader 中的每个元素将返回包含64个特征和标签的批次。
batch_size = 64
# Create data loaders.
train_dataloader = DataLoader(training_data, batch_size=batch_size)
test_dataloader = DataLoader(test_data, batch_size=batch_size)
for X, y in test_dataloader:
print(f"Shape of X [N, C, H, W]: {X.shape}")
print(f"Shape of y: {y.shape} {y.dtype}")
break
输出结果：
Shape of X [N, C, H, W]: torch.Size([64, 1, 28, 28])
Shape of y: torch.Size([64]) torch.int64
创建模型
在 PyTorch 中定义神经网络时，我们创建一个继承自 nn.Module 的类。在 __init__ 函数中定义网络的各层结构，并在 forward 函数中指定数据如何通过网络。为加速神经网络中的运算，如果有可用的 GPU 或 MPS（苹果芯片的加速器），我们会将网络移动到该设备上运行。
# Get cpu, gpu or mps device for training.
device = (
"cuda"
if torch.cuda.is_available()
else "mps"
if torch.backends.mps.is_available()
else "cpu"
)
print(f"Using {device} device")
# Define model
class NeuralNetwork(nn.Module):
def __init__(self):
super().__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)
运行结果：
Using cuda device
NeuralNetwork(
(flatten): Flatten(start_dim=1, end_dim=-1)
(linear_relu_stack): Sequential(
(0): Linear(in_features=784, out_features=512, bias=True)
(1): ReLU()
(2): Linear(in_features=512, out_features=512, bias=True)
(3): ReLU()
(4): Linear(in_features=512, out_features=10, bias=True)
)
)
优化模型参数
要训练模型，我们需要一个损失函数和一个优化器。
loss_fn = nn.CrossEntropyLoss()
optimizer = torch.optim.SGD(model.parameters(), lr=1e-3)
在一次训练循环中，模型对训练数据集（以批次输入）进行预测，并将预测误差反向传播，以调整模型的参数。
def train(dataloader, model, loss_fn, optimizer):
size = len(dataloader.dataset)
model.train()
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
loss.backward()
optimizer.step()
optimizer.zero_grad()
if batch % 100 == 0:
loss, current = loss.item(), (batch + 1) * len(X)
print(f"loss: {loss:>7f} [{current:>5d}/{size:>5d}]")
训练过程分为多个迭代（epoch）进行。在每个 epoch 中，模型通过学习参数来提高预测效果。我们会在每个 epoch 输出模型的准确率和损失，希望看到准确率逐渐提高，损失逐渐降低。
epochs = 5
for t in range(epochs):
print(f"Epoch {t+1}\n-------------------------------")
train(train_dataloader, model, loss_fn, optimizer)
test(test_dataloader, model, loss_fn)
print("Done!")
输出如下：
Epoch 1
-------------------------------
loss: 2.303494 [ 64/60000]
loss: 2.294637 [ 6464/60000]
loss: 2.277102 [12864/60000]
loss: 2.269977 [19264/60000]
loss: 2.254235 [25664/60000]
loss: 2.237146 [32064/60000]
loss: 2.231055 [38464/60000]
loss: 2.205037 [44864/60000]
loss: 2.203240 [51264/60000]
loss: 2.170889 [57664/60000]
Test Error:
Accuracy: 53.9%, Avg loss: 2.168588
Epoch 2
-------------------------------
loss: 2.177787 [ 64/60000]
loss: 2.168083 [ 6464/60000]
loss: 2.114910 [12864/60000]
loss: 2.130412 [19264/60000]
loss: 2.087473 [25664/60000]
loss: 2.039670 [32064/60000]
loss: 2.054274 [38464/60000]
loss: 1.985457 [44864/60000]
loss: 1.996023 [51264/60000]
loss: 1.917241 [57664/60000]
Test Error:
Accuracy: 60.2%, Avg loss: 1.920374
Epoch 3
-------------------------------
loss: 1.951705 [ 64/60000]
loss: 1.919516 [ 6464/60000]
loss: 1.808730 [12864/60000]
loss: 1.846550 [19264/60000]
loss: 1.740618 [25664/60000]
loss: 1.698733 [32064/60000]
loss: 1.708889 [38464/60000]
loss: 1.614436 [44864/60000]
loss: 1.646475 [51264/60000]
loss: 1.524308 [57664/60000]
Test Error:
Accuracy: 61.4%, Avg loss: 1.547092
Epoch 4
-------------------------------
loss: 1.612695 [ 64/60000]
loss: 1.570870 [ 6464/60000]
loss: 1.424730 [12864/60000]
loss: 1.489542 [19264/60000]
loss: 1.367256 [25664/60000]
loss: 1.373464 [32064/60000]
loss: 1.376744 [38464/60000]
loss: 1.304962 [44864/60000]
loss: 1.347154 [51264/60000]
loss: 1.230661 [57664/60000]
Test Error:
Accuracy: 62.7%, Avg loss: 1.260891
Epoch 5
-------------------------------
loss: 1.337803 [ 64/60000]
loss: 1.313278 [ 6464/60000]
loss: 1.151837 [12864/60000]
loss: 1.252142 [19264/60000]
loss: 1.123048 [25664/60000]
loss: 1.159531 [32064/60000]
loss: 1.175011 [38464/60000]
loss: 1.115554 [44864/60000]
loss: 1.160974 [51264/60000]
loss: 1.062730 [57664/60000]
Test Error:
Accuracy: 64.6%, Avg loss: 1.087374
Done!
保存模型
保存模型的一种常见方法是序列化内部的状态字典（包含模型参数）。
torch.save(model.state_dict(), "model.pth")
print("Saved PyTorch Model State to model.pth")
输出：
Saved PyTorch Model State to model.pth
加载模型
加载模型的过程包括重新创建模型结构，并将状态字典加载到模型中。
model = NeuralNetwork().to(device)
model.load_state_dict(torch.load("model.pth", weights_only=True))
输出：
<All keys matched successfully>
现在这个模型可以用来做预测了。
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():
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"
总结
您今天需要运行的代码：
分两个文件，一个文件用于训练模型，另一个用于加载模型和使用模型，注意两个模型需要在同一目录下。
import torch
from torch import nn
from torch.utils.data import DataLoader
from torchvision import datasets
from torchvision.transforms import ToTensor
# 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(),
)
batch_size = 64
# Create data loaders.
train_dataloader = DataLoader(training_data, batch_size=batch_size)
test_dataloader = DataLoader(test_data, batch_size=batch_size)
for X, y in test_dataloader:
print(f"Shape of X [N, C, H, W]: {X.shape}")
print(f"Shape of y: {y.shape} {y.dtype}")
break
# Get cpu, gpu or mps device for training.
device = (
"cuda"
if torch.cuda.is_available()
else "mps"
if torch.backends.mps.is_available()
else "cpu"
)
print(f"Using {device} device")
# Define model
class NeuralNetwork(nn.Module):
def __init__(self):
super().__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)
loss_fn = nn.CrossEntropyLoss()
optimizer = torch.optim.SGD(model.parameters(), lr=1e-3)
def train(dataloader, model, loss_fn, optimizer):
size = len(dataloader.dataset)
model.train()
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
loss.backward()
optimizer.step()
optimizer.zero_grad()
if batch % 100 == 0:
loss, current = loss.item(), (batch + 1) * len(X)
print(f"loss: {loss:>7f} [{current:>5d}/{size:>5d}]")
def test(dataloader, model, loss_fn):
size = len(dataloader.dataset)
num_batches = len(dataloader)
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 /= num_batches
correct /= size
print(f"Test Error: \n Accuracy: {(100*correct):>0.1f}%, Avg loss: {test_loss:>8f} \n")
epochs = 5
for t in range(epochs):
print(f"Epoch {t+1}\n-------------------------------")
train(train_dataloader, model, loss_fn, optimizer)
test(test_dataloader, model, loss_fn)
print("Done!")
torch.save(model.state_dict(), "model.pth")
print("Saved PyTorch Model State to model.pth")
输出如下：
Shape of X [N, C, H, W]: torch.Size([64, 1, 28, 28])
Shape of y: torch.Size([64]) torch.int64
Using cuda device
NeuralNetwork(
(flatten): Flatten(start_dim=1, end_dim=-1)
(linear_relu_stack): Sequential(
(0): Linear(in_features=784, out_features=512, bias=True)
(1): ReLU()
(2): Linear(in_features=512, out_features=512, bias=True)
(3): ReLU()
(4): Linear(in_features=512, out_features=10, bias=True)
)
)
Epoch 1
-------------------------------
loss: 2.321039 [ 64/60000]
loss: 2.300082 [ 6464/60000]
loss: 2.276550 [12864/60000]
loss: 2.261506 [19264/60000]
loss: 2.253393 [25664/60000]
loss: 2.225407 [32064/60000]
loss: 2.232775 [38464/60000]
loss: 2.197544 [44864/60000]
loss: 2.204715 [51264/60000]
loss: 2.160194 [57664/60000]
Test Error:
Accuracy: 47.8%, Avg loss: 2.158097
Epoch 2
-------------------------------
loss: 2.177010 [ 64/60000]
loss: 2.158526 [ 6464/60000]
loss: 2.095876 [12864/60000]
loss: 2.105602 [19264/60000]
loss: 2.056697 [25664/60000]
loss: 2.007255 [32064/60000]
loss: 2.030588 [38464/60000]
loss: 1.952178 [44864/60000]
loss: 1.966830 [51264/60000]
loss: 1.876572 [57664/60000]
Test Error:
Accuracy: 56.8%, Avg loss: 1.879299
Epoch 3
-------------------------------
loss: 1.920324 [ 64/60000]
loss: 1.881016 [ 6464/60000]
loss: 1.757274 [12864/60000]
loss: 1.795695 [19264/60000]
loss: 1.687385 [25664/60000]
loss: 1.645998 [32064/60000]
loss: 1.667249 [38464/60000]
loss: 1.569485 [44864/60000]
loss: 1.599562 [51264/60000]
loss: 1.485055 [57664/60000]
Test Error:
Accuracy: 61.0%, Avg loss: 1.506500
Epoch 4
-------------------------------
loss: 1.573652 [ 64/60000]
loss: 1.539013 [ 6464/60000]
loss: 1.383209 [12864/60000]
loss: 1.455913 [19264/60000]
loss: 1.338859 [25664/60000]
loss: 1.340163 [32064/60000]
loss: 1.354653 [38464/60000]
loss: 1.285331 [44864/60000]
loss: 1.317009 [51264/60000]
loss: 1.218405 [57664/60000]
Test Error:
Accuracy: 63.4%, Avg loss: 1.243824
Epoch 5
-------------------------------
loss: 1.316245 [ 64/60000]
loss: 1.301429 [ 6464/60000]
loss: 1.130637 [12864/60000]
loss: 1.240115 [19264/60000]
loss: 1.112968 [25664/60000]
loss: 1.142092 [32064/60000]
loss: 1.165823 [38464/60000]
loss: 1.113276 [44864/60000]
loss: 1.144331 [51264/60000]
loss: 1.062266 [57664/60000]
Test Error:
Accuracy: 64.9%, Avg loss: 1.082635
Done!
Saved PyTorch Model State to model.pth
Process finished with exit code 0
等待train.py训练好模型后，运行apply.py:
import torch
from torch import nn
from torch.utils.data import DataLoader
from torchvision import datasets
from torchvision.transforms import ToTensor
# Get cpu, gpu or mps device for training.
# 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(),
)
device = (
"cuda"
if torch.cuda.is_available()
else "mps"
if torch.backends.mps.is_available()
else "cpu"
)
print(f"Using {device} device")
# Define model
class NeuralNetwork(nn.Module):
def __init__(self):
super().__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)
model = NeuralNetwork().to(device)
model.load_state_dict(torch.load("model.pth", weights_only=True))
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():
x = x.to(device)
pred = model(x)
predicted, actual = classes[pred[0].argmax(0)], classes[y]
print(f'Predicted: "{predicted}", Actual: "{actual}"')
输出如下：
Using cuda device
Predicted: "Ankle boot", Actual: "Ankle boot"
Process finished with exit code 0
您当然可以省略存储模型，加载模型这一步，直接在一个文件中训练完后直接使用，但当您需要反复使用这个模型时，上述方法可以替您省去反复训练模型的时间。
全部评论 (0)
还没有任何评论哟~