torch语法学习1

import torch

# 设置数据类型和设备
dtype = torch.float  # 张量数据类型为浮点型
device = torch.device("cpu")  # 本次计算在 CPU 上进行

# 创建并打印两个随机张量 a 和 b
a = torch.randn(2, 3, device=device, dtype=dtype)  # 创建一个 2x3 的随机张量
b = torch.randn(2, 3, device=device, dtype=dtype)  # 创建另一个 2x3 的随机张量

print("张量 a:")
print(a)

print("张量 b:")
print(b)

# 逐元素相乘并输出结果
print("a 和 b 的逐元素乘积:")
print(a * b)

# 输出张量 a 所有元素的总和
print("张量 a 所有元素的总和:")
print(a.sum())

# 输出张量 a 中第 2 行第 3 列的元素（注意索引从 0 开始）
print("张量 a 第 2 行第 3 列的元素:")
print(a[1, 2])

# 输出张量 a 中的最大值
print("张量 a 中的最大值:")
print(a.max())

import numpy as np
numpy_array = np.array([[1, 2], [3, 4]])
tensor_from_numpy = torch.from_numpy(numpy_array)
print(tensor_from_numpy)

import torch
tensor = torch.tensor([1, 2, 3])
print(tensor)

x.numpy()

# 创建一个需要梯度的张量
tensor_requires_grad = torch.tensor([1.0], requires_grad=True)
# 进行一些操作
tensor_result = tensor_requires_grad * 2
# 计算梯度
tensor_result.backward()
print(tensor_requires_grad.grad)  # 输出梯度

with torch.no_grad():
    y = x * 2

import torch.nn as nn
import torch.optim as optim
# 定义一个简单的全连接神经网络
class SimpleNN(nn.Module):
    def __init__(self):
        super(SimpleNN, self).__init__()
        self.fc1 = nn.Linear(2, 2)  # 输入层到隐藏层
        self.fc2 = nn.Linear(2, 1)  # 隐藏层到输出层
    
    def forward(self, x):
        x = torch.relu(self.fc1(x))  # ReLU 激活函数
        x = self.fc2(x)
        return x
# 创建网络实例
model = SimpleNN()
# 打印模型结构
print(model)

# 随机输入
x = torch.randn(1, 2)
# 前向传播
output = model(x)
print(output)
# 定义损失函数（例如均方误差 MSE）
criterion = nn.MSELoss()
# 假设目标值为 1
target = torch.randn(1, 1)
# 计算损失
loss = criterion(output, target)
# 定义优化器（使用 Adam 优化器）
optimizer = optim.Adam(model.parameters(), lr=0.001)
# 训练步骤
optimizer.zero_grad()  # 清空梯度
loss.backward()  # 反向传播
optimizer.step()  # 更新参数

import torch
import torch.nn as nn
import torch.optim as optim

# 1. 定义一个简单的神经网络模型
class SimpleNN(nn.Module):
    def __init__(self):
        super(SimpleNN, self).__init__()
        self.fc1 = nn.Linear(2, 2)  # 输入层到隐藏层
        self.fc2 = nn.Linear(2, 1)  # 隐藏层到输出层
    
    def forward(self, x):
        x = torch.relu(self.fc1(x))  # ReLU 激活函数
        x = self.fc2(x)
        return x

# 2. 创建模型实例
model = SimpleNN()

# 3. 定义损失函数和优化器
criterion = nn.MSELoss()  # 均方误差损失函数
optimizer = optim.Adam(model.parameters(), lr=0.001)  # Adam 优化器

# 4. 假设我们有训练数据 X 和 Y
X = torch.randn(10, 2)  # 10 个样本，2 个特征
Y = torch.randn(10, 1)  # 10 个目标值

# 5. 训练循环
for epoch in range(100):  # 训练 100 轮
    optimizer.zero_grad()  # 清空之前的梯度
    output = model(X)  # 前向传播
    loss = criterion(output, Y)  # 计算损失
    loss.backward()  # 反向传播
    optimizer.step()  # 更新参数
    
    # 每 10 轮输出一次损失
    if (epoch+1) % 10 == 0:
        print(f'Epoch [{epoch+1}/100], Loss: {loss.item():.4f}')

# 导入PyTorch库
import torch
import torch.nn as nn

# 定义输入层大小、隐藏层大小、输出层大小和批量大小
n_in, n_h, n_out, batch_size = 10, 5, 1, 10

# 创建虚拟输入数据和目标数据
x = torch.randn(batch_size, n_in)  # 随机生成输入数据
y = torch.tensor([[1.0], [0.0], [0.0], 
                 [1.0], [1.0], [1.0], [0.0], [0.0], [1.0], [1.0]])  # 目标输出数据

# 创建顺序模型，包含线性层、ReLU激活函数和Sigmoid激活函数
model = nn.Sequential(
   nn.Linear(n_in, n_h),  # 输入层到隐藏层的线性变换
   nn.ReLU(),            # 隐藏层的ReLU激活函数
   nn.Linear(n_h, n_out),  # 隐藏层到输出层的线性变换
   nn.Sigmoid()           # 输出层的Sigmoid激活函数
)

# 定义均方误差损失函数和随机梯度下降优化器
criterion = torch.nn.MSELoss()
optimizer = torch.optim.SGD(model.parameters(), lr=0.01)  # 学习率为0.01

# 执行梯度下降算法进行模型训练
for epoch in range(50):  # 迭代50次
   y_pred = model(x)  # 前向传播，计算预测值
   loss = criterion(y_pred, y)  # 计算损失
   print('epoch: ', epoch, 'loss: ', loss.item())  # 打印损失值

   optimizer.zero_grad()  # 清零梯度
   loss.backward()  # 反向传播，计算梯度
   optimizer.step()  # 更新模型参数

import torch
from torch.utils.data import Dataset

# 自定义数据集类
class MyDataset(Dataset):
    def __init__(self, X_data, Y_data):
        """
        初始化数据集，X_data 和 Y_data 是两个列表或数组
        X_data: 输入特征
        Y_data: 目标标签
        """
        self.X_data = X_data
        self.Y_data = Y_data

    def __len__(self):
        """返回数据集的大小"""
        return len(self.X_data)

    def __getitem__(self, idx):
        """返回指定索引的数据"""
        x = torch.tensor(self.X_data[idx], dtype=torch.float32)  # 转换为 Tensor
        y = torch.tensor(self.Y_data[idx], dtype=torch.float32)
        return x, y

# 示例数据
X_data = [[1, 2], [3, 4], [5, 6], [7, 8]]  # 输入特征
Y_data = [1, 0, 1, 0]  # 目标标签

# 创建数据集实例
dataset = MyDataset(X_data, Y_data)

from torch.utils.data import DataLoader

# 创建 DataLoader 实例，batch_size 设置每次加载的样本数量
dataloader = DataLoader(dataset, batch_size=2, shuffle=True)

# 打印加载的数据
for epoch in range(1):
    for batch_idx, (inputs, labels) in enumerate(dataloader):
        print(f'Batch {batch_idx + 1}:')
        print(f'Inputs: {inputs}')
        print(f'Labels: {labels}')

import torchvision.datasets as datasets
import torchvision.transforms as transforms

# 定义预处理操作
transform = transforms.Compose([
    transforms.ToTensor(),
    transforms.Normalize((0.5,), (0.5,))  # 对灰度图像进行标准化
])

# 下载并加载 MNIST 数据集
train_dataset = datasets.MNIST(root='./data', train=True, download=True, transform=transform)
test_dataset = datasets.MNIST(root='./data', train=False, download=True, transform=transform)

# 创建 DataLoader
train_loader = DataLoader(train_dataset, batch_size=64, shuffle=True)
test_loader = DataLoader(test_dataset, batch_size=64, shuffle=False)

# 迭代训练数据
for inputs, labels in train_loader:
    print(inputs.shape)  # 每个批次的输入数据形状
    print(labels.shape)  # 每个批次的标签形状

import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
from torchvision import datasets, transforms


# 1. 数据加载与预处理
transform = transforms.Compose([
    transforms.ToTensor(),  # 转为张量
    transforms.Normalize((0.5,), (0.5,))  # 归一化到 [-1, 1]
])

# 加载 MNIST 数据集
train_dataset = datasets.MNIST(root='./data', train=True, transform=transform, download=True)
test_dataset = datasets.MNIST(root='./data', train=False, transform=transform, download=True)

train_loader = torch.utils.data.DataLoader(dataset=train_dataset, batch_size=64, shuffle=True)
test_loader = torch.utils.data.DataLoader(dataset=test_dataset, batch_size=64, shuffle=False)

# 2. 定义 CNN 模型
class SimpleCNN(nn.Module):
    def __init__(self):
        super(SimpleCNN, self).__init__()
        # 定义卷积层
        self.conv1 = nn.Conv2d(1, 32, kernel_size=3, stride=1, padding=1)  # 输入1通道，输出32通道
        self.conv2 = nn.Conv2d(32, 64, kernel_size=3, stride=1, padding=1)  # 输入32通道，输出64通道
        # 定义全连接层
        self.fc1 = nn.Linear(64 * 7 * 7, 128)  # 展平后输入到全连接层
        self.fc2 = nn.Linear(128, 10)  # 10 个类别

    def forward(self, x):
        x = F.relu(self.conv1(x))  # 第一层卷积 + ReLU
        x = F.max_pool2d(x, 2)     # 最大池化
        x = F.relu(self.conv2(x))  # 第二层卷积 + ReLU
        x = F.max_pool2d(x, 2)     # 最大池化
        x = x.view(-1, 64 * 7 * 7) # 展平
        x = F.relu(self.fc1(x))    # 全连接层 + ReLU
        x = self.fc2(x)            # 最后一层输出
        return x

# 创建模型实例
model = SimpleCNN()

# 3. 定义损失函数与优化器
criterion = nn.CrossEntropyLoss()  # 多分类交叉熵损失
optimizer = optim.SGD(model.parameters(), lr=0.01, momentum=0.9)

# 4. 模型训练
num_epochs = 5
model.train()  # 设置模型为训练模式

for epoch in range(num_epochs):
    total_loss = 0
    for images, labels in train_loader:
        outputs = model(images)  # 前向传播
        loss = criterion(outputs, labels)  # 计算损失

        optimizer.zero_grad()  # 清空梯度
        loss.backward()  # 反向传播
        optimizer.step()  # 更新参数

        total_loss += loss.item()

    print(f"Epoch [{epoch+1}/{num_epochs}], Loss: {total_loss / len(train_loader):.4f}")

# 5. 模型测试
model.eval()  # 设置模型为评估模式
correct = 0
total = 0

with torch.no_grad():  # 关闭梯度计算
    for images, labels in test_loader:
        outputs = model(images)
        _, predicted = torch.max(outputs, 1)
        total += labels.size(0)
        correct += (predicted == labels).sum().item()

accuracy = 100 * correct / total
print(f"Test Accuracy: {accuracy:.2f}%")

import torch
import torch.nn as nn
import torch.optim as optim
import matplotlib.pyplot as plt
import numpy as np

# 数据集：字符序列预测（Hello -> Elloh）
char_set = list("hello")
char_to_idx = {c: i for i, c in enumerate(char_set)}
idx_to_char = {i: c for i, c in enumerate(char_set)}

# 数据准备
input_str = "hello"
target_str = "elloh"
input_data = [char_to_idx[c] for c in input_str]
target_data = [char_to_idx[c] for c in target_str]

# 转换为独热编码
input_one_hot = np.eye(len(char_set))[input_data]

# 转换为 PyTorch Tensor
inputs = torch.tensor(input_one_hot, dtype=torch.float32)
targets = torch.tensor(target_data, dtype=torch.long)

# 模型超参数
input_size = len(char_set)
hidden_size = 8
output_size = len(char_set)
num_epochs = 200
learning_rate = 0.1

# 定义 RNN 模型
class RNNModel(nn.Module):
    def __init__(self, input_size, hidden_size, output_size):
        super(RNNModel, self).__init__()
        self.rnn = nn.RNN(input_size, hidden_size, batch_first=True)
        self.fc = nn.Linear(hidden_size, output_size)

    def forward(self, x, hidden):
        out, hidden = self.rnn(x, hidden)
        out = self.fc(out)  # 应用全连接层
        return out, hidden

model = RNNModel(input_size, hidden_size, output_size)

# 定义损失函数和优化器
criterion = nn.CrossEntropyLoss()
optimizer = optim.Adam(model.parameters(), lr=learning_rate)

# 训练 RNN
losses = []
hidden = None  # 初始隐藏状态为 None
for epoch in range(num_epochs):
    optimizer.zero_grad()

    # 前向传播
    outputs, hidden = model(inputs.unsqueeze(0), hidden)
    hidden = hidden.detach()  # 防止梯度爆炸

    # 计算损失
    loss = criterion(outputs.view(-1, output_size), targets)
    loss.backward()
    optimizer.step()
    losses.append(loss.item())

    if (epoch + 1) % 20 == 0:
        print(f"Epoch [{epoch + 1}/{num_epochs}], Loss: {loss.item():.4f}")

# 测试 RNN
with torch.no_grad():
    test_hidden = None
    test_output, _ = model(inputs.unsqueeze(0), test_hidden)
    predicted = torch.argmax(test_output, dim=2).squeeze().numpy()

    print("Input sequence: ", ''.join([idx_to_char[i] for i in input_data]))
    print("Predicted sequence: ", ''.join([idx_to_char[i] for i in predicted]))

# 可视化损失
plt.plot(losses, label="Training Loss")
plt.xlabel("Epoch")
plt.ylabel("Loss")
plt.title("RNN Training Loss Over Epochs")
plt.legend()
plt.show()