【偷偷卷死小伙伴Pytorch20天】-【day8】-【低阶API示范】

系统教程20天拿下Pytorch​ 最近和中哥、会哥进行一个小打卡活动，20天pytorch，这是第8天。欢迎一键三连。

本章我们介绍Pytorch中5个不同的层次结构：即硬件层，内核层，低阶API，中阶API，高阶API【torchkeras】。并以线性回归和DNN二分类模型为例，直观对比展示在不同层级实现模型的特点。

Pytorch的层次结构从低到高可以分成如下五层。

最底层为硬件层，Pytorch支持CPU、GPU加入计算资源池。

第二层为C++实现的内核。

第三层为Python实现的操作符，提供了封装C++内核的低级API指令，主要包括各种张量操作算子、自动微分、变量管理.
如torch.tensor,torch.cat,torch.autograd.grad,nn.Module.
如果把模型比作一个房子，那么第三层API就是【模型之砖】。

第四层为Python实现的模型组件，对低级API进行了函数封装，主要包括各种模型层，损失函数，优化器，数据管道等等。
如torch.nn.Linear,torch.nn.BCE,torch.optim.Adam,torch.utils.data.DataLoader.
如果把模型比作一个房子，那么第四层API就是【模型之墙】。

第五层为Python实现的模型接口。Pytorch没有官方的高阶API。为了便于训练模型，作者仿照keras中的模型接口，使用了不到300行代码，封装了Pytorch的高阶模型接口torchkeras.Model。如果把模型比作一个房子，那么第五层API就是模型本身，即【模型之屋】。

文章目录

下面的范例使用Pytorch的低阶API实现

线性回归模型和

DNN二分类模型。

低阶API主要包括张量操作，计算图和自动微分。

一、线性回归

1，准备数据

import numpy as np 
import pandas as pd
from matplotlib import pyplot as plt 
import torch
from torch import nn


#样本数量
n = 400

# 生成测试用数据集
X = 10*torch.rand([n,2])-5.0  #torch.rand是均匀分布 
w0 = torch.tensor([[2.0],[-3.0]])
b0 = torch.tensor([[10.0]])
Y = X@w0 + b0 + torch.normal( 0.0,2.0,size = [n,1])  # @表示矩阵乘法,增加正态扰动

torch.rand Returns a tensor filled with random numbers from a uniform distribution on the interval [0, 1)[0,1)

# 数据可视化

%matplotlib inline
%config InlineBackend.figure_format = 'svg'

plt.figure(figsize = (12,5))
ax1 = plt.subplot(121)
ax1.scatter(X[:,0].numpy(),Y[:,0].numpy(), c = "b",label = "samples")
ax1.legend()
plt.xlabel("x1")
plt.ylabel("y",rotation = 0)

ax2 = plt.subplot(122)
ax2.scatter(X[:,1].numpy(),Y[:,0].numpy(), c = "g",label = "samples")
ax2.legend()
plt.xlabel("x2")
plt.ylabel("y",rotation = 0)
plt.show()

ax.legend()作用：在图上标明一个图例，用于说明每条曲线的文字显示

# 构建数据管道迭代器
def data_iter(features, labels, batch_size=8):
    num_examples = len(features)
    indices = list(range(num_examples))
    np.random.shuffle(indices)  #样本的读取顺序是随机的
    for i in range(0, num_examples, batch_size):
        indexs = torch.LongTensor(indices[i: min(i + batch_size, num_examples)])
        yield  features.index_select(0, indexs), labels.index_select(0, indexs)
        
# 测试数据管道效果   
batch_size = 8
(features,labels) = next(data_iter(X,Y,batch_size))
print(features)
print(labels)

2，定义模型

# 定义模型
class LinearRegression: 
    
    def __init__(self):
        self.w = torch.randn_like(w0,requires_grad=True)
        self.b = torch.zeros_like(b0,requires_grad=True)
        
    #正向传播
    def forward(self,x): 
        return x@self.w + self.b

    # 损失函数 MSE
    def loss_func(self,y_pred,y_true):  
        return torch.mean((y_pred - y_true)**2/2)

model = LinearRegression()

3，训练模型

def train_step(model, features, labels):
    
    predictions = model.forward(features)
    loss = model.loss_func(predictions,labels)
    
    print(model.w.grad) #一开始是梯度为None
    print(model.b.grad)

        
    # 反向传播求梯度
    loss.backward()
    
    # 使用torch.no_grad()避免梯度记录，也可以通过操作 model.w.data 实现避免梯度记录 
    with torch.no_grad():
        # 梯度下降法更新参数
        model.w -= 0.001*model.w.grad
        model.b -= 0.001*model.b.grad

        # # 梯度清零
        model.w.grad.zero_()
        model.b.grad.zero_()
    return loss

# 测试train_step效果
batch_size = 10
(features,labels) = next(data_iter(X,Y,batch_size))
train_step(model,features,labels)

grad_fn用来记录变量是怎么来的，方便计算梯度

def train_model(model,epochs):
    for epoch in range(1,epochs+1):
        for features, labels in data_iter(X,Y,10):
            loss = train_step(model,features,labels)

        if epoch%200==0:
            # printbar()
            print("epoch =",epoch,"loss = ",loss.item())
            print("model.w =",model.w.data)
            print("model.b =",model.b.data)

train_model(model,epochs = 1000)

4，结果可视化

# 结果可视化

%matplotlib inline
%config InlineBackend.figure_format = 'svg'

plt.figure(figsize = (12,5))
ax1 = plt.subplot(121)
ax1.scatter(X[:,0].numpy(),Y[:,0].numpy(), c = "b",label = "samples")
ax1.plot(X[:,0].numpy(),(model.w[0].data*X[:,0]+model.b[0].data).numpy(),"-r",linewidth = 5.0,label = "model")
ax1.legend()
plt.xlabel("x1")
plt.ylabel("y",rotation = 0)


ax2 = plt.subplot(122)
ax2.scatter(X[:,1].numpy(),Y[:,0].numpy(), c = "g",label = "samples")
ax2.plot(X[:,1].numpy(),(model.w[1].data*X[:,1]+model.b[0].data).numpy(),"-r",linewidth = 5.0,label = "model")
ax2.legend()
plt.xlabel("x2")
plt.ylabel("y",rotation = 0)

plt.show()

二、DNN二分类模型

1，准备数据

%matplotlib inline
%config InlineBackend.figure_format = 'svg'

import numpy as np 
import pandas as pd 
from matplotlib import pyplot as plt
import torch
from torch import nn



#正负样本数量
n_positive,n_negative = 2000,2000

#生成正样本, 小圆环分布
r_p = 5.0 + torch.normal(0.0,1.0,size = [n_positive,1]) 
theta_p = 2*np.pi*torch.rand([n_positive,1])
Xp = torch.cat([rp*torch.cos(theta_p),r_p*torch.sin(theta_p)],axis = 1)
Yp = torch.ones_like(r_p)

#生成负样本, 大圆环分布
r_n = 8.0 + torch.normal(0.0,1.0,size = [n_negative,1]) 
theta_n = 2*np.pi*torch.rand([n_negative,1])
Xn = torch.cat([r_n*torch.cos(theta_n),r_n*torch.sin(theta_n)],axis = 1)
Yn = torch.zeros_like(r_n)

#汇总样本
X = torch.cat([Xp,Xn],axis = 0)
Y = torch.cat([Yp,Yn],axis = 0)


#可视化
plt.figure(figsize = (6,6))
plt.scatter(Xp[:,0].numpy(),Xp[:,1].numpy(),c = "r")
plt.scatter(Xn[:,0].numpy(),Xn[:,1].numpy(),c = "g")
plt.legend(["positive","negative"]);

C=torch.cat((A,B),0) #按维数0（行）拼接

# 构建数据管道迭代器
def data_iter(features, labels, batch_size=8):
    num_examples = len(features)
    indices = list(range(num_examples))
    np.random.shuffle(indices)  #样本的读取顺序是随机的
    for i in range(0, num_examples, batch_size):
        indexs = torch.LongTensor(indices[i: min(i + batch_size, num_examples)])
        yield  features.index_select(0, indexs), labels.index_select(0, indexs)
        
# 测试数据管道效果   
batch_size = 8
(features,labels) = next(data_iter(X,Y,batch_size))
print(features)
print(labels)

2，定义模型

此处范例我们利用nn.Module来组织模型变量。

class DNNModel(nn.Module):
    def __init__(self):
        super(DNNModel, self).__init__()
        self.w1 = nn.Parameter(torch.randn(2,4))
        self.b1 = nn.Parameter(torch.zeros(1,4))
        self.w2 = nn.Parameter(torch.randn(4,8))
        self.b2 = nn.Parameter(torch.zeros(1,8))
        self.w3 = nn.Parameter(torch.randn(8,1))
        self.b3 = nn.Parameter(torch.zeros(1,1))

    # 正向传播
    def forward(self,x):
        x = torch.relu(x@self.w1 + self.b1)
        x = torch.relu(x@self.w2 + self.b2)
        y = torch.sigmoid(x@self.w3 + self.b3)
        return y
    
    # 损失函数(二元交叉熵)
    def loss_func(self,y_pred,y_true):  
        #将预测值限制在1e-7以上, 1- (1e-7)以下，避免log(0)错误
        eps = 1e-7
        y_pred = torch.clamp(y_pred,eps,1.0-eps)
        bce = - y_true*torch.log(y_pred) - (1-y_true)*torch.log(1-y_pred)
        return torch.mean(bce)
    
    # 评估指标(准确率)
    def metric_func(self,y_pred,y_true):
        y_pred = torch.where(y_pred>0.5,torch.ones_like(y_pred,dtype = torch.float32),
                          torch.zeros_like(y_pred,dtype = torch.float32))
        acc = torch.mean(1-torch.abs(y_true-y_pred))
        return acc
    
model = DNNModel()

# 测试模型结构
batch_size = 10
(features,labels) = next(data_iter(X,Y,batch_size))

predictions = model(features)

loss = model.loss_func(labels,predictions)
metric = model.metric_func(labels,predictions)

print("init loss:", loss.item())
print("init metric:", metric.item())

3，训练模型

def train_step(model, features, labels):   
    
    # 正向传播求损失
    predictions = model.forward(features)
    loss = model.loss_func(predictions,labels)
    metric = model.metric_func(predictions,labels)
        
    # 反向传播求梯度
    loss.backward()
    
    # 梯度下降法更新参数
    for param in model.parameters():
        #注意是对param.data进行重新赋值,避免此处操作引起梯度记录
        param.data = (param.data - 0.01*param.grad.data) 
        
    # 梯度清零
    model.zero_grad()
        
    return loss.item(),metric.item()
 

def train_model(model,epochs):
    for epoch in range(1,epochs+1):
        loss_list,metric_list = [],[]
        for features, labels in data_iter(X,Y,20):
            lossi,metrici = train_step(model,features,labels)
            loss_list.append(lossi)
            metric_list.append(metrici)
        loss = np.mean(loss_list)
        metric = np.mean(metric_list)

        if epoch%100==0:
            print("epoch =",epoch,"loss = ",loss,"metric = ",metric)
        
train_model(model,epochs = 1000)

4，结果可视化

# 结果可视化
fig, (ax1,ax2) = plt.subplots(nrows=1,ncols=2,figsize = (12,5))
ax1.scatter(Xp[:,0],Xp[:,1], c="r")
ax1.scatter(Xn[:,0],Xn[:,1],c = "g")
ax1.legend(["positive","negative"]);
ax1.set_title("y_true");

Xp_pred = X[torch.squeeze(model.forward(X)>=0.5)]
Xn_pred = X[torch.squeeze(model.forward(X)<0.5)]

ax2.scatter(Xp_pred[:,0],Xp_pred[:,1],c = "r")
ax2.scatter(Xn_pred[:,0],Xn_pred[:,1],c = "g")
ax2.legend(["positive","negative"]);
ax2.set_title("y_pred");

总结

线性回归总结

1.torch.rand Returns a tensor filled with random numbers from a uniform distribution on the interval [0, 1)[0,1)
2.ax1.legend() 显示图例
3.构建数据管道迭代器示范代码

# 构建数据管道迭代器
def data_iter(features, labels, batch_size=8):
    num_examples = len(features)
    indices = list(range(num_examples))
    np.random.shuffle(indices)  #样本的读取顺序是随机的
    for i in range(0, num_examples, batch_size):
        indexs = torch.LongTensor(indices[i: min(i + batch_size, num_examples)])
        yield  features.index_select(0, indexs), labels.index_select(0, indexs)
        
# 测试数据管道效果   
batch_size = 8
(features,labels) = next(data_iter(X,Y,batch_size))
print(features)
print(labels)

4.训练模型切记梯度清0 with torch.no_grad():

```python
def train_step(model, features, labels):
    
    predictions = model.forward(features)
    loss = model.loss_func(predictions,labels)
    
    print(model.w.grad) #一开始梯度为None
    print(model.b.grad)

        
    # 反向传播求梯度
    loss.backward()
    
    # 使用torch.no_grad()避免梯度记录，也可以通过操作 model.w.data 实现避免梯度记录 
    with torch.no_grad():
        # 梯度下降法更新参数
        model.w -= 0.001*model.w.grad
        model.b -= 0.001*model.b.grad

        # # 梯度清零
        model.w.grad.zero_()
        model.b.grad.zero_()
    return loss

DNN总结

1.torch.cat((A,B),0) #按维数0（行）拼接
2.构建数据管道示范代码

# 构建数据管道迭代器
def data_iter(features, labels, batch_size=8):
    num_examples = len(features)
    indices = list(range(num_examples))
    np.random.shuffle(indices)  #样本的读取顺序是随机的
    for i in range(0, num_examples, batch_size):
        indexs = torch.LongTensor(indices[i: min(i + batch_size, num_examples)])
        yield  features.index_select(0, indexs), labels.index_select(0, indexs)
        
# 测试数据管道效果   
batch_size = 8
(features,labels) = next(data_iter(X,Y,batch_size))
print(features)
print(labels)

3.梯度清0 param.data = (param.data - 0.01*param.grad.data)

def train_step(model, features, labels):   
    
    # 正向传播求损失
    predictions = model.forward(features)
    loss = model.loss_func(predictions,labels)
    metric = model.metric_func(predictions,labels)
        
    # 反向传播求梯度
    loss.backward()
    
    # 梯度下降法更新参数
    for param in model.parameters():
        #注意是对param.data进行重新赋值,避免此处操作引起梯度记录
        param.data = (param.data - 0.01*param.grad.data) 
        
    # 梯度清零
    model.zero_grad()
        
    return loss.item(),metric.item()

4.torch.squeeze（）降维