5. 用PyTorch实现线性回归

B站 刘二大人老师的课程代码
PyTorch 深度学习实践

1.回顾

import torch
# 样本数据
x_data = [1.0, 2.0, 3.0]  # 输入样本
y_data = [2.0, 4.0, 6.0]  # 输出样本

w = torch.tensor([1.0]) #权重初始值
w.requires_grad = True #计算梯度，默认不计算

def forward(x):
    return x * w  


def loss(x, y):  #构建计算图
    y_pred = forward(x)
    return (y_pred - y) ** 2

print("predict (before training)", 4, forward(4).item())


for epoch in range(100):
   
    for x, y in zip(x_data, y_data):
        l = loss(x, y) # l是一个张量，tensor主要是在建立计算图 forward, compute the loss
        l.backward()
        print('\tgrad:', x, y, w.grad.item())
        w.data = w.data - 0.01 * w.grad.data # 权重更新时，注意grad也是一个tensor
        w.grad.data.zero_() # after update, remember set the grad to zero
    print('progress:', epoch, l.item())  # 取出loss使用l.item，不要直接使用l（l是tensor会构建计算图）
    
    
print("predict (after training)", 4, forward(4).item())

2. 构建线性模型 SGD

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

x_data=torch.tensor([[1.0],[2.0],[3.0]])
y_data=torch.tensor([[4.0],[5.0],[6.0]])

class LinearModel(torch.nn.Module):
    def __init__(self):
        super(LinearModel,self).__init__()
        self.linear=torch.nn.Linear(1,1)
        
    def forward(self,x):
        y_pred=self.linear(x)
        return y_pred
    
model=LinearModel() #可调用

#构造损失函数
criterion=torch.nn.MSELoss(size_average=False)
optimizer=torch.optim.SGD(model.parameters(),lr=0.01)

#训练过程
epoch_list=[]
loss_list=[]

for epoch in range(100):
    y_pred=model(x_data)
    loss=criterion(y_pred,y_data)
    print(epoch,loss)
    epoch_list.append([epoch])
    loss_list.append([loss.detach().numpy()])
    
    optimizer.zero_grad()
    loss.backward()
    optimizer.step()
    
#打印权重
print('w=',model.linear.weight.item())
print('b=',model.linear.bias.item())

#测试模型
x_test=torch.tensor([[4.0]])
y_test=model(x_test)
print("y_pred=",y_test.data)

plt.plot(epoch_list,loss_list)
plt.xlabel('epoch')
plt.ylabel('loss')
plt.show()

可变参数

class Foobar:
    def __init__(self):
        pass
    def __call__(self,*args,**kwargs):
        print("Hello"+str(args[0]))
        
foobar=Foobar()
foobar(1,2,3)

def func(*args,x,y):
    print(args)
    
func(1,2,4,3,x=3,y=5)

def func(*args,**kwargs):
    print(args)
    print(kwargs)
    
func(1,2,4,3,x=5,y=6)

3. 其他优化函数

在构造损失函数部分直接将优化函数改为需要的优化函数。
如果想用Adam，直接把SGD改为Adam

4. 练习

用三阶多项式拟合sin(x)

4.1 numpy实现

import numpy as np
import math 

# 创建输入和输出数据
x=np.linspace(-math.pi,math.pi,2000)
y=np.sin(x)

#参数随机初始化
a=np.random.randn()
b=np.random.randn()
c=np.random.randn()
d=np.random.randn()

#学习率
learining_rate=1e-6

#正向迭代2000次
for t in range(2000):
    #计算预测
    y_pred=a+b*x+c*x**2+d*x**3
    

    #计算损失
    loss=((y_pred-y)**2).sum()
    if t%100==99:
        print(t,loss)
        
    #反向传播
    #计算参数的梯度
    grad_y_pred=2*(y_pred-y)
    grad_a=grad_y_pred.sum()
    """grad_b=(grad_y_pred*x).sum()
    grad_c=(grad_y_pred*x**2).sum()
    grad_d=(grad_y_pred*x**3).sum()"""

    grad_b=grad_y_pred@x.T
    grad_c=grad_y_pred@(x**2).T
    grad_d=grad_y_pred@(x**3).T

    #参数更新
    a=a-learining_rate*grad_a
    b=b-learining_rate*grad_b
    c=c-learining_rate*grad_c
    d=d-learining_rate*grad_d

#输出
print(f'result: y={a}+{b}x+{c}x^2+{d}x^3')

4.2 张量tensor实现

# -- coding: utf-8 --
import math
import numpy as np

dtype = torch.float
device = torch.device("cpu")
# device = torch.device("cuda:0") # Uncomment this to run on GPU

# 创建输入和输出数据
x = torch.linspace(-math.pi, math.pi, 2000, device=device, dtype=dtype)
y = torch.sin(x)

# 参数随机初始化
a = torch.randn((), device=device, dtype=dtype)
b = torch.randn((), device=device, dtype=dtype)
c = torch.randn((), device=device, dtype=dtype)
d = torch.randn((), device=device, dtype=dtype)

# 学习率
learning_rate = 1e-6

# 正向迭代2000次
for t in range(2000):
    # 计算预测
    y_pred = a + b * x + c * x ** 2 + d * x ** 3

    # 计算损失
    loss = ((y_pred - y) ** 2).pow(2).sum().item()
    if t % 100 == 99:
        print(t, loss)

    # 反向传播
    # 计算参数的梯度
    grad_y_pred = 2 * (y_pred - y)
    grad_a = grad_y_pred.sum()
    grad_b = (grad_y_pred * x).sum()
    grad_c = (grad_y_pred * x ** 2).sum()
    grad_d = (grad_y_pred * x ** 3).sum()

"""    grad_b=grad_y_pred@x.T
    grad_c=grad_y_pred@(x**2).T
    grad_d=grad_y_pred@(x**3).T"""

# 参数更新
a -= learning_rate * grad_a
b -= learning_rate * grad_b
c -= learning_rate * grad_c
d -= learning_rate * grad_d

# 输出
print(f'result: y={a.item()}+{b.item()}x+{c.item()}x^2+{d.item()}x^3')

4.3 自动更新梯度Autograd

# -- coding: utf-8 --
import math
import numpy as np


# 创建输入和输出数据
x = torch.linspace(-math.pi, math.pi, 2000, device=device, dtype=dtype)
y = torch.sin(x)

# 输出y是(x, x^2, x^3)的线性函数
p=torch.tensor([1,2,3])
xx=x.unsqueeze(-1).pow(p)


model=torch.nn.Sequential(torch.nn.Linear(3,1),torch.nn.Flatten(0,1))
loss_fn=torch.nn.MSELoss(reduction='sum')

# 学习率
learning_rate = 1e-6

# 正向迭代2000次
for t in range(2000):
    # 计算预测
    y_pred = model(xx)

    # 计算损失
    loss = loss_fn(y_pred,y)
    if t % 100 == 99:
        print(t, loss)

    # 反向传播
    # 计算参数的梯度
    model.zero_grad()
    loss.backward()
    
    #参数更新
    with torch.no_grad():
        for param in model.parameters():
            param-=learning_rate * param.grad

linear_layer=model[0]

# 输出
print(f'Result: y = {linear_layer.bias.item()} + {linear_layer.weight[:, 0].item()} x + {linear_layer.weight[:, 1].item()} x^2 + {linear_layer.weight[:, 2].item()} x^3')