pytorch快速入门

一、Tensors

Tensors贯穿pytorch始终

和多维数组很相似,一个特点是可以加速硬件加速

tensors的初始化(四种):

  1. 直接给值:
data = [[1,2],[3,4]]
x_data = torch.tensor(data)
  1. 从numpy数组转化
np_arr = np.array(data)
x_np = torch.from_numpy(np_array)
  1. 从另一个tensor
x_np = torch.from_numpy(np_array)
  1. 赋01或随机值
shape = (2,3,)
rand_tensor = torch.rand(shape)
ones_tensor = torch.ones(shape)
zeros_tensor = torch.zeros(shape)

tensors的属性

使用GPU方式

if torch.cuda_is_available():
    tensor = tensor.to("cuda")
  1. 索引和切片
tensor = torch.ones(4,4)
print(tensor[0]) #第一行(0开始)
print(tensor[;,0]) #第一列(0开始)
print(tensor[...,-1]) #最后一列
  1. 连接
t1 = troch.cat([tensor, tensor],dim = 1) # 沿着第一维的方向拼接
  1. 矩阵乘法(三种办法,类似于运算符重载、成员函数和非成员函数)
y1 = tensor @ tensor
y2 = tensor.matmul(tensor.T)
y3 = torch.rand_like(tensor)
torch.matmul(tensor, tensor.T, out=y3)

  1. 点乘
z1 = tensor * tensor
z2 = tensor.mul(tensor)
z3 = tensor.rand_like(tensor)
torch.mul(tensor, tensor, out=z3)
  1. 单元素tensor求值
agg = tensor.sum()
agg_item = agg.item()
print(agg_item, type(agg_item))
  1. In-place操作(就是会改变成员内容的成员函数,以下划线结尾)
tensor.add_(5) #每个元素都+5

节约内存,但是会丢失计算前的值,不推荐使用

和numpy的联系

  1. tensor转numpy数组
t = torch.ones(5)
n = t.numpy()

类似于引用,并没有新建内存,二者同时修改同步

  1. numpy数组转tensor
n = np.ones(5)
t = torch.from_numpy(n)

同样为引用,一个的修改会对另一个有影响

二、数据集的数据加载器

加载数据集

以FasnionMNIST为例,需要四个参数

  • root是路径
  • train区分训练集还是测试集
  • download表示如果root找不到,就得从网上下载
  • transform表明数据的下载格式
import torch
from torch.utils.data import Dataset
from torchvision import datasets
from torchvision.transforms import ToTensor
import matplotlib.pyplot as plt

training_data = datasets.FansionMNIST(
	root = "data",
    train = True,
    download = True,
    transform = ToTensor()
)

test_data = datasets.FansionMNIST(
	root = "data",
    train = Flase,
    download = True,
    transform = ToTensor()
)

标号和可视化

labels_map = {
    0: "T-Shirt",
    1: "Trouser",
    2: "Pullover",
    3: "Dress",
    4: "Coat",
    5: "Sandal",
    6: "Shirt",
    7: "Sneaker",
    8: "Bag",
    9: "Ankle Boot",
}

figure = plt.figure(figsize=(8,8))

cols, rows = 3, 3
for i in range(1, cols * row + 1):
    sample_idx = torch.randint(len(training_data), size=(1,)).item()
    img, label = training_data[sample_idx]
    figure.add_subplot(rows, cols, i)
    plt.title(labels_map[label])
    plt.axis("off")
    plt.imshow(img.squeeze(), cmap="gray")
    
plt.show()

自己创建数据集

必须实现三个函数__init__,__len__,__getitem__

import os
import pandas as pd
from torchvision.io import read_image

class CustomImageDataset(Dataset):
    def __init__(self, annotations_file, img_dir, transform=None, target_transform=None):
        self.img_labels = pd.read_csv(annotations_file)
        self.img_dir = img_dir
        self.transform = transform
        self.target_transform = target_transform
    def __len__(self):
        return len(self.img_labels)
    
    def __getitem__(self, idx):
        img_path = os.path.join(self.img_dir, self.img_labels.iloc[idx, 0])
        image = read_image(img_path)
        label = self.img_labels.iloc[idx, 1]
        if self.transform:
            image = self.transform(image)
        if self.target_transform:
            label = self.target_transform(label)
        
        return image, label
          
  • __init__类似于构造函数
  • __len__求数据个数
  • __getitem__按下标找数据和标签,类似重载[]

用DataLoaders准备数据用于训练

# 展示图像和标签
train_features, train_labels = next(iter(train_dataloader))
print(f"Feature batch shape: {train_features.size()}")
print(f"Labels batch shape: {train_labels.size()}")

img = train_features[0].squeeze()
label = train_labels[0]
plt.imshow(img, cmap="gray")
plt.show()
print(f"Label: {label}")

Transforms

让数据转换成需要的形式

transform指定feature的转换

target_transform指定标签的转换

比如,需要数据从PTL Image变成Tensors,标签从整数变成one-hot encoded tensors

上述用到了两个技术,ToTensor()Lambda表达式

ToTensor()将PIL images或Numpy数组转化成FloatTensor,每个像素的灰度转化到[0,1]范围内

Lambda类似C++里的Lambda表达式,我们需要将整数转化为one-hot encoded tensor, 就先创建一个长度为数据标签类型的全0的Tensor,然后用scatter_()把第y个值改为1。此时,scatter的index接受的参数也是Tensor。

三、神经网络

神经网络是一些层或者模块,对数据进行处理。

torch.nn提供了诸多构造神经网络的模块,模块化的结构方便了管理复杂结构。

接下来以在FashionMNIST上构造一个图像分类器为例。

import os
import torch
from torch.utils.data import DataLoader
from torchvision import datasets, tranforms

准备训练设备

  • GPU
  • CPU
device = "cuda" if torch.cuda.is_available() else "cpu"
print(f"Using {device} device")

定义网络的类

nn.Module继承

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

创建一个实例(对象),放到device上

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

跑一下结果

Using cpu 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)
    )
)

结果是返回值的softmax,是个10维的概率,找到最大的就是预测结果

X = torch.rand(1, 28, 28, device=device)
logits = model(X)
pred_probab = nn.Softmax(dim=1)(logits)
y_pred = pred_probab.argmax(1)
print(f"Predicted class: {y_pred}")

模型的layers

以3张28*28的图像为例,分析它在network里的状态

input_image = torch.rand(3, 28, 28)
print(input_image.size())

'''
torch.Size([3,28,28])
'''

nn.Flatten

Flatten顾名思义,扁平化,用于将2维tensor转为1维

flatten = nn.flatten()
flat_image = flatten(input_image)
print(flag_image.size())

'''
torch.Size([3,784])
'''

nn.Linear

Linear,做线性变换

layer1 = nn.Linear(in_features=28*28, out_features=20)
hidden1 = layer1(flag_image)
print(hidden1.size())

'''
torch.Size([3,20])
'''

nn.ReLU

非线性激活函数,在Linear层后,增加非线性,让神经网络学到更多的信息

hidden1 = nn.ReLU()(hidden1)

nn.Sequential

Sequential,序列的,类似于把layers一层一层的摆着

seq_modules = nn.Sequential(
	flatten,
    layer1,
    nn.ReLU(),
    nn.Linear(20, 10)
)
input_image = torch.rand(3, 28, 28)
logits = seq_modules(imput_image)

nn.Softmax

最后一层的结果返回一个在[-inf,inf]的值logits,通过softmax层后,映射到[0,1]

这样[0,1]的值可以作为概率输出,dim指定和为1的维度

softmax = nn.Softmax(dim=1)
pred_probab = softmax(logits)

模型的参数

这些layers是参数化的,就是说在训练中weights和biases不断被优化

下述代码输出这个模型的所有参数值

for name, param in model.named_parameters():
    print(name, param.size(), param[:2])

四、自动求导

训练神经网络的时候,最常用的是反向传播,模型参数根据loss function的梯度进行调整。

为了求梯度,也就是求导,使用torch.autograd

考虑到就一个layer的网络,输入x,参数w和b,以及一个loss function,也就是

import torch

x = torch.ones(5) # input tensor
y = torch.zeros(3) # expected output
w = torch.randn(5, 3, requires_grad=True)
b = torch.randn(3, requires_grad=True)
z = torch.matmul(x, w) + b
loss = torch.nn.functional.binary_cross_entropy_with_logits(z, y)

Tensors,Functions and Computational Graph

考虑这个过程的Computational Graph,如下

img

这个一定是DAG (有向无环图)

为了计算loss在w和b方向上的梯度,我们给他们设置requires_grad

w.requires_grad(True)
b.requires_grad(True)

Functions实际上是对象,有计算正向值和反向导数的成员。

print(z.grad_fn)
print(loss.grad_fn)

计算梯度

要计算losswb的偏导,只需使用

loss.backward()

于是就得到了

print(w.grad)
print(b.grad)

注意:

  • 只能计算图里叶子的梯度,内部的点不能算
  • 一张图只能计算一次梯度,要保留节点的话,backward要传retain_graph=True
import torch

x = torch.randn((1, 4), dtype=torch.float32, requires_grad=True)
y = x ** 2
z = y * 4

print(x)
print(y)
print(z)
loss1 = z.mean()
loss2 = z.sum()
print(loss1, loss2)
loss1.backward() # 这个代码执行正常,但是执行完中间变量都free了,所以下一个会出现问题
print(loss1, loss2)
loss2.backward() # 这时会引发错误

要保留节点的话,可把loss1的那行改成

loss1.backward(retain_graph=True)

不计算梯度

有时我们不需要计算梯度,比如模型已经训练好了,只需要正向使用

这时算梯度就很拖累时间,所以要禁用梯度

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

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

'''
True
False
'''

另一个办法就是用.detach()

z = torch.matmul(x, w) + b
z_det = z.detach()
print(z_det.requires_grad)

'''
False
'''

tensor输出和雅可比积

如果函数的输出是tensor,就不能简单算梯度了

结果是一个矩阵(其实就是依次遍历x和y的分量,求偏导)

img

具体来说,把v当参数传进去

inp = torch.eye(5, requires_grad=True)
out = (inp + 1).pow(2)
out.backward(torch.ones_like(inp), retain_graph=True)

五、优化模型参数

有了模型,接下来要进行训练、验证和测试

前置代码

首先要加载数据,建立模型

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

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

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

超参数

定义三个超参数

  • Epochs数:数据集迭代次数
  • Batch size:单次训练样本数
  • Learning Rate:学习速度

优化训练

接下来,进行多轮的优化,每一轮称为一个epoch

每个epoch包含两部分,训练loop和验证/测试loop

Loss Function

PyTorch提供常见的Loss Functions

  • nn.MSLoss(Mean Square Error)
  • nn.NLLLoss(Negative Log Likelihood)
  • nn.CrossEntropyLoss(交叉熵)

使用交叉熵,把原始结果logits放进去

loss_fn = nn.CrossEntropyLoss()

Optimizer

初始化优化器,给它需要优化的参数,和超参数Learning Rate

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

优化器在每个epoch里做三件事

  • optimizer.zero_grad()将梯度清零
  • loss.backward()进行反向传播
  • optimizer.step()根据梯度调整参数

完整实现

train_loop里训练,test_loop里测试

import torch
from torch import nn
from torch.utils.data import DataLoader
from torchvision import datasets
from torchvision.transforms import ToTensor, Lambda
 
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()
 
learning_rate = 1e-3
batch_size = 64
epochs = 5
 
# Initialize the loss function
loss_fn = nn.CrossEntropyLoss()
 
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):
        # 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:
            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")
 
loss_fn = nn.CrossEntropyLoss()
optimizer = torch.optim.SGD(model.parameters(), lr=learning_rate)
 
epochs = 10
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!")

六、保存和加载模型

import torch
import torchvison.models as models

保存和加载模型权重1

通过torch.save方法,可以将模型保存到state_dict类型的字典里。

model = models.vgg16(pretrained=True)
torch.save(model.state_dict(), 'model_weights.pth')

而要加载的话,需要先构造相同类型的模型,然后把参数加载进去

model = models.vgg16()
model.load_state_dict(torch.load('model_weights.pth'))
model.eval()

一定要调一下model.eval(),防止后续出错

保存和加载模型权重2

上一种方法里,需要先实例化模型,再导入权值

有没有办法直接保存和加载整个模型呢?

使用不传model.state_dict()参数,改成model

保存方式:

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

加载方式:

model = torch.load('model.pth')

参考链接:

  1. https://zhuanlan.zhihu.com/p/515043918
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值