1.Mnist分类任务
-
网络基本构建与训练方法,常用函数解析
-
torch.nn.functional模块
-
nn.Module模块
学习方法:边用边查,多打印,duo'gua
使用jupyter的优点,可以打印出每一个步骤。
2.读取数据集
自动下载
%matplotlib inline
#查看本机torch的版本
import torch
print(torch.__version__)#打印torch的版本
加载并读取数据集
from pathlib import Path
import requests
DATA_PATH = Path("data")
PATH = DATA_PATH / "mnist"
PATH.mkdir(parents=True, exist_ok=True)
URL = "http://deeplearning.net/data/mnist/"
FILENAME = "mnist.pkl.gz"
if not (PATH / FILENAME).exists():
content = requests.get(URL + FILENAME).content
(PATH / FILENAME).open("wb").write(content)
import pickle
import gzip
with gzip.open((PATH / FILENAME).as_posix(), "rb") as f:
((x_train, y_train), (x_valid, y_valid), _) = pickle.load(f, encoding="latin-1")
观察数据的结构
784是mnist数据集每个样本的像素点个数
print(x_train[0].shape)
print(x_train.shape)
y_train
显示一个记录的灰度图
from matplotlib import pyplot
import numpy as np
pyplot.imshow(x_train[0].reshape((28, 28)), cmap="gray")
print(x_train.shape)
numpy和torch的区别:
torch->gpu->tensor
numpy->cpu->ndarray
数据转化:
import torch
x_train, y_train, x_valid, y_valid = map(
torch.tensor, (x_train, y_train, x_valid, y_valid)
)
n, c = x_train.shape
x_train, x_train.shape, y_train.min(), y_train.max()
print(x_train, y_train)
print(x_train.shape)
print(y_train.min(), y_train.max())
3 torch.nn.functional 很多层和函数在这里都会见到
torch.nn.functional中有很多功能,后续会常用的。那什么时候使用nn.Module,什么时候使用nn.functional呢?一般情况下,如果模型有可学习的参数,最好用nn.Module,其他情nn.functional相对更简单一些
import torch.nn.functional as F
loss_func = F.cross_entropy
def model(xb):
return xb.mm(weights) + bias
bs = 64
xb = x_train[0:bs] # a mini-batch from x
yb = y_train[0:bs]
weights = torch.randn([784, 10], dtype = torch.float, requires_grad = True)
#线性代数的相关知识,weights与输入想乘之后,需要输出的格式为10分类,所以的weights的矩阵为(784,10)
bs = 64
bias = torch.zeros(10, requires_grad=True)#偏执的设置,常数值作为初始化,因为这个东西对模型的影###响不是很大。
print(loss_func(model(xb), yb)):#计算真实值和预测值之间的误差
4 创建一个model来更简化代码
- 必须继承nn.Module且在其构造函数中需调用nn.Module的构造函数
- 无需写反向传播函数,nn.Module能够利用autograd自动实现反向传播
- Module中的可学习参数可以通过named_parameters()或者parameters()返回迭代器
from torch import nn
class Mnist_NN(nn.Module):
def __init__(self):
super().__init__()
self.hidden1 = nn.Linear(784, 128)
self.hidden2 = nn.Linear(128, 256)
self.out = nn.Linear(256, 10)
#定义前向传播,torch有一个优点:前向传播自己定义,反向传播自动实现。
def forward(self, x):
x = F.relu(self.hidden1(x))
x = F.relu(self.hidden2(x))
x = self.out(x)
return x
net = Mnist_NN()
print(net)
可以打印我们定义好名字里的权重和偏置项
for name, parameter in net.named_parameters():
print(name, parameter,parameter.size())
5 使用TensorDataset和DataLoader来简化
from torch.utils.data import TensorDataset
from torch.utils.data import DataLoader
train_ds = TensorDataset(x_train, y_train)
train_dl = DataLoader(train_ds, batch_size=bs, shuffle=True)#shuffle=True:洗牌的操作。
valid_ds = TensorDataset(x_valid, y_valid)
valid_dl = DataLoader(valid_ds, batch_size=bs * 2)
def get_data(train_ds, valid_ds, bs):
return (
DataLoader(train_ds, batch_size=bs, shuffle=True),
DataLoader(valid_ds, batch_size=bs * 2),
)
- 一般在训练模型时加上model.train(),这样会正常使用Batch Normalization和 Dropout
- 测试的时候一般选择model.eval(),这样就不会使用Batch Normalization和 Dropout