【PyTorch】PyTorch模型定义
本节内容学习了：
- 模型自定义的三种方式：Sequential, Modile List,Module dict；
- 当模型有重复出现的层结构，我们可以构建模型块实现复用来构建复杂模型；
- 如果要对模型进行修改：通过实例化来修改特定的层、通过在forward中增加参数来增加输入（要注意修改函数体内增加的输入是如何起作用的、对应的模型的定义也要进行修改）、通过return特定的值来实现额外输出；
- 模型的读取和保存要考虑2种训练模式2种保存类型。
1 模型自定义的三种方式：
- Sequential：层的排列，适用于层数较少的情况，但不需要定义forward。
Direct List：直接列出
## Sequential: Direct list
import torch.nn as nn
net1 = nn.Sequential(
nn.Linear(784, 256),
nn.ReLU(),
nn.Linear(256, 10),
)
print(net1)
Sequential(
(0): Linear(in_features=784, out_features=256, bias=True)
(1): ReLU()
(2): Linear(in_features=256, out_features=10, bias=True)
)
Ordered Dict：有序字典（Python中的字典无序，但网络是有序的，因此用Ordered Dict）
## Sequential: Ordered Dict
import collections
import torch.nn as nn
net2 = nn.Sequential(collections.OrderedDict([
('fc1', nn.Linear(784, 256)), # fully connect
('relu1', nn.ReLU()),
('fc2', nn.Linear(256, 10))
]))
print(net2)
Sequential(
(fc1): Linear(in_features=784, out_features=256, bias=True)
(relu1): ReLU()
(fc2): Linear(in_features=256, out_features=10, bias=True)
)
用OrderedDict可以实现层的命名。
- Module List ：对于模块比较重复的网络比较实用，但是需要定义forward并实例化。
一个错误展示：
# ModuleList
net3 = nn.ModuleList([nn.Linear(784, 256), nn.ReLU()]) #[nn.Linear(x,x) for i in range(5)]
net3.append(nn.Linear(256, 10)) # # 类似List的append操作
print(net3[-1]) # 类似List的索引访问，输出最后一层
print(net3)
注意：ModuleList定义后并没有形成网络，只是将不同的模块存储到一起，因此不能直接输入数据，因此还需要定义forward()。正确方式如下：
class Net3(nn.Module):
def __init__(self):
super().__init__()
self.modulelist = nn.ModuleList([nn.Linear(784, 256), nn.ReLU()])
self.modulelist.append(nn.Linear(256, 10))
def forward(self, x):
for layer in self.modulelist:
x = layer(x)
return x
net3_ = Net3()
out3_ = net3_(a)
print(out3_.shape)
- Module Dict：对于模块比较重复的网络比较实用，但是需要定义forward并实例化。
## ModuleDict
net4 = nn.ModuleDict({
'linear': nn.Linear(784, 256),
'act': nn.ReLU(),
})
net4['output'] = nn.Linear(256, 10) # 添加
print(net4['linear']) # 访问
print(net4.output)
同样地，ModuleDict并没有定义一个网络，它只是将不同的模块储存在一起。此处应报错。正确使用方式同上。
2 利用模型块快速搭建复杂网络
重复出现的层成为模块。以U-Net为例，介绍如何构建模型块，以及如何利用模型块快速搭建复杂模型。
组成U-Net的模型块主要有如下几个部分：
1）每个子块内部的两次卷积（Double Convolution）
2）左侧模型块之间的下采样连接，即最大池化（Max pooling）
3）右侧模型块之间的上采样连接（Up sampling）
4）输出层的处理
除模型块外，还有模型块之间的横向连接，输入和U-Net底部的连接等计算，这些单独的操作可以通过forward函数来实现。 （参考：https://github.com/milesial/Pytorch-UNet ）
import os
import numpy as np
import collections
import torch
import torch.nn as nn
import torch.nn.functional as F
import torchvision
class DoubleConv(nn.Module):
"""(convolution => [BN] => ReLU) * 2"""
def __init__(self, in_channels, out_channels, mid_channels=None):
super().__init__()
if not mid_channels:
mid_channels = out_channels
self.double_conv = nn.Sequential(
nn.Conv2d(in_channels, mid_channels, kernel_size=3, padding=1, bias=False),
nn.BatchNorm2d(mid_channels),
nn.ReLU(inplace=True),
nn.Conv2d(mid_channels, out_channels, kernel_size=3, padding=1, bias=False),
nn.BatchNorm2d(out_channels),
nn.ReLU(inplace=True)
)
def forward(self, x):
return self.double_conv(x)
class Down(nn.Module):
"""Downscaling with maxpool then double conv"""
def __init__(self, in_channels, out_channels):
super().__init__()
self.maxpool_conv = nn.Sequential(
nn.MaxPool2d(2),
DoubleConv(in_channels, out_channels)
)
def forward(self, x):
return self.maxpool_conv(x)
class Up(nn.Module):
"""Upscaling then double conv"""
def __init__(self, in_channels, out_channels, bilinear=True):
super().__init__()
# if bilinear, use the normal convolutions to reduce the number of channels
if bilinear:
self.up = nn.Upsample(scale_factor=2, mode='bilinear', align_corners=True)
self.conv = DoubleConv(in_channels, out_channels, in_channels // 2)
else:
self.up = nn.ConvTranspose2d(in_channels, in_channels // 2, kernel_size=2, stride=2)
self.conv = DoubleConv(in_channels, out_channels)
def forward(self, x1, x2):
x1 = self.up(x1)
# input is CHW
diffY = x2.size()[2] - x1.size()[2]
diffX = x2.size()[3] - x1.size()[3]
x1 = F.pad(x1, [diffX // 2, diffX - diffX // 2,
diffY // 2, diffY - diffY // 2])
# if you have padding issues, see
# https://github.com/HaiyongJiang/U-Net-Pytorch-Unstructured-Buggy/commit/0e854509c2cea854e247a9c615f175f76fbb2e3a
# https://github.com/xiaopeng-liao/Pytorch-UNet/commit/8ebac70e633bac59fc22bb5195e513d5832fb3bd
x = torch.cat([x2, x1], dim=1)
return self.conv(x)
class OutConv(nn.Module):
def __init__(self, in_channels, out_channels):
super(OutConv, self).__init__()
self.conv = nn.Conv2d(in_channels, out_channels, kernel_size=1)
def forward(self, x):
return self.conv(x)
## 组装
class UNet(nn.Module):
def __init__(self, n_channels, n_classes, bilinear=True):
super(UNet, self).__init__()
self.n_channels = n_channels
self.n_classes = n_classes
self.bilinear = bilinear
self.inc = DoubleConv(n_channels, 64)
self.down1 = Down(64, 128)
self.down2 = Down(128, 256)
self.down3 = Down(256, 512)
factor = 2 if bilinear else 1
self.down4 = Down(512, 1024 // factor)
self.up1 = Up(1024, 512 // factor, bilinear)
self.up2 = Up(512, 256 // factor, bilinear)
self.up3 = Up(256, 128 // factor, bilinear)
self.up4 = Up(128, 64, bilinear)
self.outc = OutConv(64, n_classes)
def forward(self, x):
x1 = self.inc(x)
x2 = self.down1(x1)
x3 = self.down2(x2)
x4 = self.down3(x3)
x5 = self.down4(x4)
x = self.up1(x5, x4)
x = self.up2(x, x3)
x = self.up3(x, x2)
x = self.up4(x, x1)
logits = self.outc(x)
return logits
unet = UNet(3,1) #实例化
unet
3 模型修改
- 修改特定的层
## 修改特定层
import copy
unet1 = copy.deepcopy(unet)#复制
unet1.outc
尝试进行修改：
b = torch.rand(1,3,224,224)#1个batch,3个通道，大小224*224
out_unet1 = unet1(b)
print(out_unet1.shape)
torch.Size([1, 1, 224, 224])
这里输出了1个通道，如果我想让它输出5个通道该如何做呢？
unet1.outc = OutConv(64, 5) #重新实例化
unet1.outc
OutConv(
(conv): Conv2d(64, 5, kernel_size=(1, 1), stride=(1, 1))
)
这里，输出就变成了5个通道。
- 添加额外输入
只需要修改两个地方：forward中增加一个输入，函数体内根据需要修改add_variable如何起作用的；修改对应的模型的定义。
## 添加额外输入
class UNet2(nn.Module):
def __init__(self, n_channels, n_classes, bilinear=True):
super(UNet2, self).__init__()
self.n_channels = n_channels
self.n_classes = n_classes
self.bilinear = bilinear
self.inc = DoubleConv(n_channels, 64)
self.down1 = Down(64, 128)
self.down2 = Down(128, 256)
self.down3 = Down(256, 512)
factor = 2 if bilinear else 1
self.down4 = Down(512, 1024 // factor)
self.up1 = Up(1024, 512 // factor, bilinear)
self.up2 = Up(512, 256 // factor, bilinear)
self.up3 = Up(256, 128 // factor, bilinear)
self.up4 = Up(128, 64, bilinear)
self.outc = OutConv(64, n_classes)
def forward(self, x, add_variable):
x1 = self.inc(x)
x2 = self.down1(x1)
x3 = self.down2(x2)
x4 = self.down3(x3)
x5 = self.down4(x4)
x = self.up1(x5, x4)
x = self.up2(x, x3)
x = self.up3(x, x2)
x = self.up4(x, x1)
x = x + add_variable #修改点
logits = self.outc(x)
return logits
unet2 = UNet2(3,1)
c = torch.rand(1,1,224,224)
out_unet2 = unet2(b, c)
print(out_unet2.shape)
- 添加额外的输出
只需要修改return。假设想利用Unet的bottom layer做分类，只需要return中增加x5。
## 添加额外输出
class UNet3(nn.Module):
def __init__(self, n_channels, n_classes, bilinear=True):
super(UNet3, self).__init__()
self.n_channels = n_channels
self.n_classes = n_classes
self.bilinear = bilinear
self.inc = DoubleConv(n_channels, 64)
self.down1 = Down(64, 128)
self.down2 = Down(128, 256)
self.down3 = Down(256, 512)
factor = 2 if bilinear else 1
self.down4 = Down(512, 1024 // factor)
self.up1 = Up(1024, 512 // factor, bilinear)
self.up2 = Up(512, 256 // factor, bilinear)
self.up3 = Up(256, 128 // factor, bilinear)
self.up4 = Up(128, 64, bilinear)
self.outc = OutConv(64, n_classes)
def forward(self, x):
x1 = self.inc(x)
x2 = self.down1(x1)
x3 = self.down2(x2)
x4 = self.down3(x3)
x5 = self.down4(x4)
x = self.up1(x5, x4)
x = self.up2(x, x3)
x = self.up3(x, x2)
x = self.up4(x, x1)
logits = self.outc(x)
return logits, x5 # 修改点
unet3 = UNet3(3,1)
c = torch.rand(1,1,224,224)
out_unet3, mid_out = unet3(b)
print(out_unet3.shape, mid_out.shape)
模型保存和读取
- 模型保存
需要考虑两种两种训练模式下的两种保存类型：- 两种训练模式：单卡、多卡
- 两种保存类型：保存整个模型、保存模型权重
unet.state_dict()用字典的方式将每个层的权重保存下来，
## CPU或单卡：保存&读取整个模型
torch.save(unet, "./unet_example.pth")
loaded_unet = torch.load("./unet_example.pth")
loaded_unet.state_dict()
## CPU或单卡：保存&读取模型权重
torch.save(unet.state_dict(), "./unet_weight_example.pth")
loaded_unet_weights = torch.load("./unet_weight_example.pth")
unet.load_state_dict(loaded_unet_weights)
unet.state_dict()
## 多卡：保存&读取整个模型。注意模型层名称前多了module
## 不建议，因为保存模型的GPU_id等信息和读取后训练环境可能不同，尤其是要把保存的模型交给另一用户使用的情况
os.environ['CUDA_VISIBLE_DEVICES'] = '2,3'
unet_mul = copy.deepcopy(unet)
unet_mul = nn.DataParallel(unet_mul).cuda()
unet_mul
torch.save(unet_mul, "./unet_mul_example.pth")
loaded_unet_mul = torch.load("./unet_mul_example.pth")
loaded_unet_mul
## 多卡：保存&读取模型权重。
torch.save(unet_mul.state_dict(), "./unet_weight_mul_example.pth")
loaded_unet_weights_mul = torch.load("./unet_weight_mul_example.pth")
unet_mul.load_state_dict(loaded_unet_weights_mul)
unet_mul = nn.DataParallel(unet_mul).cuda()
unet_mul.state_dict()
# 另外，如果保存的是整个模型，也建议采用提取权重的方式构建新的模型：
unet_mul.state_dict = loaded_unet_mul.state_dict
unet_mul = nn.DataParallel(unet_mul).cuda()
unet_mul.state_dict()