TensorFlow与Pytorch的转换——2手写数字识别

数据处理

import tensorflow as tf

mnist = tf.keras.datasets.mnist
(x_train, y_train), (x_test, y_test) = mnist.load_data()

# 将像素值缩放到0到1之间
x_train, x_test = x_train / 255.0, x_test / 255.0

# 将标签转换为one-hot编码
y_train = tf.keras.utils.to_categorical(y_train, num_classes=10)
y_test = tf.keras.utils.to_categorical(y_test, num_classes=10)

构建模型

model = tf.keras.models.Sequential([
    tf.keras.layers.Flatten(input_shape=(28, 28)),
    tf.keras.layers.Dense(128, activation='relu'),
    tf.keras.layers.Dense(64, activation='relu'),
    tf.keras.layers.Dense(10, activation='softmax')
])

训练模型

model.compile(optimizer='adam',
              loss='categorical_crossentropy',
              metrics=['accuracy'])
model.fit(x_train, y_train, epochs=10)

测试模型

test_loss, test_acc = model.evaluate(x_test, y_test)
print('Test accuracy:', test_acc)

import matplotlib.pyplot as plt
import numpy as np

predictions = model.predict(x_test)


# 随机选择一些测试图像
indices = np.random.choice(range(len(x_test)), 10)

predictions = model.predict(x_test)
fig, axs = plt.subplots(2,5, figsize=(20,8))

# 可视化测试图像及其预测标签
for i, ax in zip(indices, axs.flatten()):
    ax.imshow(x_test[i], cmap='gray')
    ax.set_title(f"Predicted label: {np.argmax(predictions[i])}")
plt.show()

Pytorch版本

import torch  
import torch.nn as nn  
import torch.optim as optim  
import torchvision  
import torchvision.transforms as transforms  
import matplotlib.pyplot as plt  
import numpy as np  
  
# 加载MNIST数据集  
transform = transforms.Compose([transforms.ToTensor(), transforms.Normalize((0.5,), (0.5,))])  
trainset = torchvision.datasets.MNIST(root='./data', train=True, download=True, transform=transform)  
trainloader = torch.utils.data.DataLoader(trainset, batch_size=64, shuffle=True)  
  
testset = torchvision.datasets.MNIST(root='./data', train=False, download=True, transform=transform)  
testloader = torch.utils.data.DataLoader(testset, batch_size=64, shuffle=False)  
  
# 定义神经网络模型  
class Net(nn.Module):  
    def __init__(self):  
        super(Net, self).__init__()  
        self.fc1 = nn.Linear(28*28, 128)  
        self.fc2 = nn.Linear(128, 64)  
        self.fc3 = nn.Linear(64, 10)  
  
    def forward(self, x):  
        x = x.view(-1, 28*28)  # 将图像展平为向量  
        x = torch.relu(self.fc1(x))  
        x = torch.relu(self.fc2(x))  
        x = torch.softmax(self.fc3(x), dim=1)  # 使用softmax输出概率分布  
        return x  
  
net = Net()  
  
# 定义损失函数和优化器  
criterion = nn.CrossEntropyLoss()  # 注意：CrossEntropyLoss内部进行了log_softmax操作，因此输出层不需要再softmax  
optimizer = optim.Adam(net.parameters(), lr=0.001)  
  
# 训练模型  
for epoch in range(10):  # 迭代10个epoch  
    running_loss = 0.0  
    for i, data in enumerate(trainloader, 0):  
        inputs, labels = data  
        optimizer.zero_grad()  # 清空梯度  
        outputs = net(inputs)  # 前向传播  
        loss = criterion(outputs, labels)  # 计算损失  
        loss.backward()  # 反向传播  
        optimizer.step()  # 更新参数  
        running_loss += loss.item()  
    print(f'Epoch [{epoch+1}/10], Loss: {running_loss/len(trainloader):.4f}')  
  
# 在测试集上评估模型  
correct = 0  
total = 0  
with torch.no_grad():  # 评估模式，不需要计算梯度  
    for data in testloader:  
        images, labels = data  
        outputs = net(images)  
        _, predicted = torch.max(outputs.data, 1)  
        total += labels.size(0)  
        correct += (predicted == labels).sum().item()  
  
print(f'Accuracy of the network on the 10000 test images: {100 * correct / total:.2f}%')  
  
# 可视化测试图像及其预测标签  
predictions = []  
test_images, test_labels = next(iter(testloader))  # 一次性加载整个测试集可能会占用大量内存，这里只取一个batch  
with torch.no_grad():  
    test_outputs = net(test_images)  
    _, predicted_labels = torch.max(test_outputs, 1)  
    predictions.append(predicted_labels.numpy())  
  
predictions = np.concatenate(predictions)  # 虽然这里只有一个batch，但为了与TensorFlow代码风格一致，仍然使用concatenate  
indices = np.random.choice(range(len(test_images)), 10)  
  
fig, axs = plt.subplots(2, 5, figsize=(20, 8))  
for i, ax in zip(indices, axs.flatten()):  
    ax.imshow(test_images[i].squeeze().numpy(), cmap='gray')  # 转换回numpy数组并去除多余的维度  
    ax.set_title(f"Predicted label: {predictions[i]}")  
plt.show()