原官方网页：https://pytorch.org/tutorials/beginner/transfer_learning_tutorial.html
通过本教程，你将学到如何使用迁移学习训练你的网络。你可以在cs231n notes了解更多关于迁移学习
引用一些笔记：
- 实际中，基本没有人会从零开始（随机初始化）训练一个完整的卷积网络，因为相对于网络，很难得到一个足够大的数据集[网络很深, 需要足够大数据集]。通常的做法是在一个很大的数据集上进行预训练得到卷积网络
ConvNet, 然后将这个ConvNet的参数作为目标任务的初始化参数或者固定这些参数
以下是应用迁移学习的两种场景：
- 微调
Convnet：使用预训练的网络(如在imagenet 1000上训练而来的网络)来初始化自己的网络，而不是随机初始化。其他的训练步骤不变。 - 将
Convnet看成固定的特征提取器。首先固定ConvNet除了最后的全连接层外的其他所有层。最后的全连接层被替换成一个新的随机初始化的层，只有这个新的层会被训练[只有这层参数会在反向传播时更新]
# License: BSD
# Author: Sasank Chilamkurthy
from __future__ import print_function, division
import torch
import torch.nn as nn
import torch.optim as optim
from torch.optim import lr_scheduler
import numpy as np
import torchvision
from torchvision import datasets, models, transforms
import matplotlib.pyplot as plt
import time
import os
import copy
plt.ion() # interactive mode
1. 数据加载
我们通常会使用torchvision和torch .utils.data包来加载数据
今天要解决的问题是训练一个模型来分类蚂蚁ants和蜜蜂bees。ants和bees各有约120张训练图片。每个类有75张验证图片。从零开始在如此小的数据集上进行训练通常是很难泛化的。由于我们使用迁移学习，模型的泛化能力会相当好
这个数据集是imagenet的子集，可以在这里下载
# 训练集数据增广和归一化
# 在验证集上仅仅归一化
data_transforms = {
'train': transforms.Compose([
transforms.RandomResizedCrop(224), # 随机裁剪一个area之后再resize
transforms.RandomHorizontalFlip(), # 随机水平翻转
transforms.ToTensor(),
transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
]),
'val': transforms.Compose([
transforms.Resize(256),
transforms.CenterCrop(224),
transforms.ToTensor(),
transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
]),
}
data_dir = 'hymenoptera_data'
image_datasets = {x: datasets.ImageFolder(os.path.join(data_dir, x),
data_transforms[x])
for x in ['train', 'val']}
dataloaders = {x: torch.utils.data.DataLoader(image_datasets[x], batch_size=4,
shuffle=True, num_workers=4)
for x in ['train', 'val']}
dataset_sizes = {x: len(image_datasets[x]) for x in ['train', 'val']}
class_names = image_datasets['train'].classes
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
1.1 可视化一些数据
我们可视化了一些训练图片来明白数据增广操作
def imshow(inp, title=None):
"""Imshow for Tensor."""
inp = inp.numpy().transpose((1, 2, 0))
mean = np.array([0.485, 0.456, 0.406])
std = np.array([0.229, 0.224, 0.225])
inp = std * inp + mean
inp = np.clip(inp, 0, 1)
plt.imshow(inp)
if title is not None:
plt.title(title)
plt.pause(0.001) # pause a bit so that plots are updated
# Get a batch of training data
inputs, classes = next(iter(dataloaders['train']))
# Make a grid from batch
out = torchvision.utils.make_grid(inputs)
imshow(out, title=[class_names[x] for x in classes])
可视化结果
2. 训练网络
现在我们写一个通用的函数来训练网络。我们将展示:
- 调整学习速率
- 保存最好的模型
如下，参数scheduler是一个来自torch.optim.lr_scheduler的学习速率调整类的对象(LR scheduler object)
def train_model(model, criterion, optimizer, scheduler, num_epochs=25):
since = time.time()
best_model_wts = copy.deepcopy(model.state_dict())
best_acc = 0.0
for epoch in range(num_epochs):
print('Epoch {}/{}'.format(epoch, num_epochs - 1))
print('-' * 10)
# Each epoch has a training and validation phase
for phase in ['train', 'val']:
if phase == 'train':
scheduler.step()
model.train() # Set model to training mode
else:
model.eval() # Set model to evaluate mode
running_loss = 0.0
running_corrects = 0
# Iterate over data.
for inputs, labels in dataloaders[phase]:
inputs = inputs.to(device)
labels = labels.to(device)
# zero the parameter gradients
optimizer.zero_grad()
# forward
# track history if only in train
with torch.set_grad_enabled(phase == 'train'):
outputs = model(inputs)
_, preds = torch.max(outputs, 1)
loss = criterion(outputs, labels)
# backward + optimize only if in training phase
if phase == 'train':
loss.backward()
optimizer.step()
# statistics
running_loss += loss.item() * inputs.size(0)
running_corrects += torch.sum(preds == labels.data)
epoch_loss = running_loss / dataset_sizes[phase]
epoch_acc = running_corrects.double() / dataset_sizes[phase]
print('{} Loss: {:.4f} Acc: {:.4f}'.format(
phase, epoch_loss, epoch_acc))
# deep copy the model
if phase == 'val' and epoch_acc > best_acc:
best_acc = epoch_acc
best_model_wts = copy.deepcopy(model.state_dict())
print()
time_elapsed = time.time() - since
print('Training complete in {:.0f}m {:.0f}s'.format(
time_elapsed // 60, time_elapsed % 60))
print('Best val Acc: {:4f}'.format(best_acc))
# load best model weights
model.load_state_dict(best_model_wts)
return model
2.1 可视化模型的预测结果
一个通用的展示少量预测图片的函数
def visualize_model(model, num_images=6):
was_training = model.training
model.eval()
images_so_far = 0
fig = plt.figure()
with torch.no_grad():
for i, (inputs, labels) in enumerate(dataloaders['val']):
inputs = inputs.to(device)
labels = labels.to(device)
outputs = model(inputs)
_, preds = torch.max(outputs, 1)
for j in range(inputs.size()[0]):
images_so_far += 1
ax = plt.subplot(num_images//2, 2, images_so_far)
ax.axis('off')
ax.set_title('predicted: {}'.format(class_names[preds[j]]))
imshow(inputs.cpu().data[j])
if images_so_far == num_images:
model.train(mode=was_training)
return
model.train(mode=was_training)
3. 微调convnent
加载预训练模型并且重置最后一个全连接层
model_ft = models.resnet18(pretrained=True)
num_ftrs = model_ft.fc.in_features
model_ft.fc = nn.Linear(num_ftrs, 2)
model_ft = model_ft.to(device)
criterion = nn.CrossEntropyLoss()
# Observe that all parameters are being optimized
optimizer_ft = optim.SGD(model_ft.parameters(), lr=0.001, momentum=0.9)
# Decay LR by a factor of 0.1 every 7 epochs
exp_lr_scheduler = lr_scheduler.StepLR(optimizer_ft, step_size=7, gamma=0.1)
3.1 训练并评估
在CPU上将耗时大约15-25分钟，在GPU上将花少于1分钟的时间
- 训练
model_ft = train_model(model_ft, criterion, optimizer_ft, exp_lr_scheduler,
num_epochs=25)
- output
Epoch 0/24
----------
train Loss: 0.6849 Acc: 0.6762
val Loss: 0.2146 Acc: 0.9281
.
.
.
Epoch 23/24
----------
train Loss: 0.2282 Acc: 0.9139
val Loss: 0.2709 Acc: 0.8954
Epoch 24/24
----------
train Loss: 0.3081 Acc: 0.8566
val Loss: 0.3045 Acc: 0.9020
Training complete in 0m 58s
Best val Acc: 0.928105
- 可视化
visualize_model(model_ft)
训练结果
4. 将convnent看成特征提取器
这里，我们将冻结全部网络，除了最后一层。我们应该将需要设置欲冻结的参数的requires_grad == False，这样在反向传播backward()的时候他们的梯度就不会被计算
更多关于grad的文档在这里
model_conv = torchvision.models.resnet18(pretrained=True)
# 最重要的一步
for param in model_conv.parameters():
param.requires_grad = False
# Parameters of newly constructed modules have requires_grad=True by default
num_ftrs = model_conv.fc.in_features
model_conv.fc = nn.Linear(num_ftrs, 2)
model_conv = model_conv.to(device)
criterion = nn.CrossEntropyLoss()
# Observe that only parameters of final layer are being optimized as
# opoosed to before.
optimizer_conv = optim.SGD(model_conv.fc.parameters(), lr=0.001, momentum=0.9)
# Decay LR by a factor of 0.1 every 7 epochs
exp_lr_scheduler = lr_scheduler.StepLR(optimizer_conv, step_size=7, gamma=0.1)
4.1 训练和评估
在CPU上，固定参数相比于之前的作为初始化参数的做法，会节约大约一半的时间。这是可以预期的，因为网络的绝大部分参数的梯度不会在反向传播中计算。(但是这些参数是参与前向传播的)
- 训练
model_conv = train_model(model_conv, criterion, optimizer_conv,
exp_lr_scheduler, num_epochs=25)
- output
Epoch 0/24
----------
train Loss: 0.6421 Acc: 0.6557
val Loss: 0.4560 Acc: 0.7451
Epoch 1/24
----------
train Loss: 0.4694 Acc: 0.7746
val Loss: 0.1616 Acc: 0.9608
Epoch 2/24
----------
train Loss: 0.4500 Acc: 0.7746
val Loss: 0.3041 Acc: 0.8627
.
.
.
Epoch 24/24
----------
train Loss: 0.3382 Acc: 0.8566
val Loss: 0.1605 Acc: 0.9542
Training complete in 0m 46s
Best val Acc: 0.967320
- 可视化
visualize_model(model_conv)
plt.ioff()
plt.show()
训练结果
