pytorch訓練模型大體流程

學習記錄：

常用api:

#卷積層
conv2d = nn.Conv2d(in_channels,out_channels,kernel_size=3,stride=1,padding=1)
#輸入通道數，輸出通道數，卷積核大小，移動步長，補0大小

#批量歸一化
bn = nn.BatchNorm2d(out_channels) #緊跟在卷積層後，參數爲卷積層輸出通道數

#激活層
relu = nn.ReLU(inplace=True)

#池化層
pool = nn.MaxPool2d(kernel_size=2,stride=2)

#反捲積層
deconv = nn.ConvTranspose2d(in_channels,out_channels,kernel_size=3,stride=2,padding=1)

訓練大體流程

# -*- coding:utf-8 -*-
import torch
import torchvision
from torch import nn,optim
from torchvision import datasets
from torchvision import transforms
from torch.utils.data import DataLoader
from torch.nn import functional as F

#外部定義
criterion = nn.NLLLoss().cuda()
optimizer = optim.Adam(net.parameters(),lr=LR)
#訓練過程
optimizer.zero_grad() # 梯度清0
out = net(imgData)
out = F.log_softmax(out,dim=1)
loss = criterion(out,imgLabel)
loss.backward()  #反向傳播
optimizer.step() #更新梯度信息