最近开始学习pytorch，从mnist入手
一、先记录下几个知识点：
1、torch中的tensor与numpy中的array，tensor与array是可以相互转换的， Variable中包含两部分信息，分别是data and grad:
tensor to array : a = torch.FloatTensor(3,3) b = a.numpy()
numpy to tensor : a = np.ones(5) b = torch.from_numpy(a)
Variable to array : a = Variable(torch.FloatTensor(3,3)) b = a.data.numpy()
array to Variable : a = np.ones(5) b = Variable(torch.from_numpy(a))
2、tensor.size() 返回一个张量的shape,可以看成是一个tuple:
x = torch.Tensor(5, 3) # construct a 5x3 matrix, uninitialized
x.size() # out: torch.Size([5, 3])
x.size(0) # out: 5
二、MNIST训练：
# -*- coding: utf-8 -*-
"""
Created on Wed Dec 5 10:14:25 2018
@author: Administrator
"""
import torchvision
from torchvision import transforms,datasets
import torch
from torch.utils.data import DataLoader
from torch.autograd import Variable
# 定义超参数
batch_size = 32
learning_rate = 1e-3
num_epoches = 100
# 下载训练集 MNIST 手写数字训练集
train_dataset = datasets.MNIST(root='./data', train=True,
transform=transforms.ToTensor(),
download=True)
test_dataset = datasets.MNIST(root='./data', train=False,
transform=transforms.ToTensor())
#导入图片
train_loader = DataLoader(train_dataset, batch_size=batch_size, shuffle=True)
test_loader = DataLoader(test_dataset, batch_size=batch_size, shuffle=False)
class Logstic_Regression(torch.nn.Module):
def __init__(self, in_dim, n_class):
super(Logstic_Regression, self).__init__()
self.logstic = torch.nn.Linear(in_dim, n_class)
def forward(self, x):
out = self.logstic(x)
return out
model = Logstic_Regression(28*28, 10) # 图片大小是28x28
model = model.cuda()
criterion = torch.nn.CrossEntropyLoss()
optimizer = torch.optim.SGD(model.parameters(), lr=learning_rate)
for epoch in range(num_epoches):
running_loss = 0.0
running_acc = 0.0
total = 0
time = 0
for i, data in enumerate(train_loader, 1): #可以定制从第几个开始枚举
img, label = data
img = img.view(img.size(0), -1) # 将图片展开成 28x28
if torch.cuda.is_available():
img = Variable(img).cuda()
label = Variable(label).cuda()
else:
img = Variable(img)
label = Variable(label)
# 向前传播
out = model(img)
loss = criterion(out, label)
running_loss += loss.data * label.size(0)
_, pred = torch.max(out, 1) #按行获取最大值，并返回所在列的索引值
num_correct = (pred == label).sum().item() # item()将一个值的张量变为标量
total += label.size(0)
running_acc += num_correct
# 向后传播
optimizer.zero_grad()
loss.backward()
optimizer.step()
if (epoch + 1) % 20 == 0:
time += 1
print('Epoch[{}/{}], loss: {:.6f}, Acc: {:.6f}'.format(epoch + 1,num_epoches,loss.data, running_acc/total))
#if time:
#print("time:",time)
# 测试
total = 0
running_acc = 0.0
for i,data in enumerate(test_loader, 1):
img , label = data
img = img.view(img.size(0), -1)
img = Variable(img).cuda()
label = Variable(label).cuda()
model.eval()
out = model(img)
_, pred = torch.max(out, 1)
num_correct = (pred == label).sum().item()
running_acc += num_correct
total += label.size(0)
if total:
print(' Acc: {:.6f}'.format( running_acc/total))
结果：
Epoch[100/100], loss: 0.198154, Acc: 0.911670
Epoch[100/100], loss: 0.165920, Acc: 0.911684
Epoch[100/100], loss: 0.276448, Acc: 0.911698
Epoch[100/100], loss: 0.155880, Acc: 0.911728
Epoch[100/100], loss: 0.185178, Acc: 0.911759
Epoch[100/100], loss: 0.256010, Acc: 0.911756
Epoch[100/100], loss: 0.309416, Acc: 0.911753
Epoch[100/100], loss: 0.244036, Acc: 0.911767
Acc: 0.916600