迁移学习(Transfer Learning)
在完成60分钟入门之后,接下来有六节tutorials和五节关于文本处理的tutorials。争取一天一节。不过重点是关注神经网络构建和数据处理部分。
这个教程中,要学习的是如何使用迁移学习,可以在cs231n课程查看。
什么是迁移学习?
在实践中,很少有人从头开始训练整个卷积网络(随机初始化),因为拥有足够大小的数据集是相对罕见的。相反,通常在非常大的数据集(例如 ImageNet,其包含具有1000个类别的120万个图像)上预先训练 ConvNet,然后使用 ConvNet 作为感兴趣任务的初始化或固定特征提取器。
以下时两个主要的迁移学习使用场景:
- Finetuning the convnet:我们使用预训练网络初始化网络,而不是随机初始化,就像在imagenet 1000数据集上训练的网络一样。其余训练看起来像往常一样。
- ConvNet as fixed feature extractor: 在这里,我们将冻结除最终完全连接层之外的所有网络的权重。最后一个全连接层被替换为具有随机权重的新层,并且仅训练该层。
教程目录:
- 库准备
- 加载数据
- 部分图片的可视化
- 训练网络
- 模型预测的可视化
- Finetuning the convnet
- 训练和评估
- ConvNet as fixed feature extractor
- 训练和评估
库准备
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
加载数据
使用torchvision和torch.utils.data包加载数据。
这次解决的问题是训练一个能够分类蚂蚁和蜜蜂的模型,蚂蚁和蜜蜂的训练集各120张图片和75张验证集图片。如果从头开始训练的话,这是一个很小的数据集,但是使用迁移学习的话,可以用小的数据集推广。(这是imagenet的小的子集)
数据集下载地址
data_transforms = {
'train': transforms.Compose([
transforms.RandomResizedCrop(224),
transforms.RandomHorizontalFlip(),
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 = 'data/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")
部分图片的可视化
现在就可以看看数据集
def imshow(inp, title=None):
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(5) #显示的时长
if __name__ == '__main__':
inputs, classes = next(iter(dataloaders['train']))
out = torchvision.utils.make_grid(inputs)
imshow(out, title=[class_names[x] for x in classes])
训练网络
现在,做一个通用函数来训练模型。
- 调度学习率
- 保存最佳的学习模型
下面函数中, scheduler 参数是 torch.optim.lr_scheduler 中的 LR scheduler 对象。
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)
for phase in ['train', 'val']:
if phase == 'train':
scheduler.step()
model.train()
else:
model.eval()
running_loss = 0.0
running_correct = 0
for inputs, labels in dataloaders[phase]:
inputs = inputs.to(device)
labels = labels.to(device)
optimizer.zero_grad()
with torch.set_grad_enabled(phase == 'train'):
outputs = model(inputs)
_, preds = torch.max(outputs, 1)
loss = criterion(outputs, labels)
if phase == 'train':
loss.backward()
optimizer.step()
running_loss += loss.item() * inputs.size(0)
running_correct += torch.sum(preds == labels.data)
epoch_loss = running_loss / dataset_sizes[phase]
epoch_acc = running_correct.double() / dataset_sizes[phase]
print('{} Loss: {:.4f} Acc: {:.4f}'.format(phase, epoch_loss, epoch_acc))
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))
model.load_state_dict(best_model_wts)
return model
模型预测的可视化
用于显示少量图像预测的通用功能
def visualize_model(model, num_images=6):
was_training = model.training
model.val()
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)
Finetuning the convnet
加载一个预训练模型并重置最终的全连接层
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()
optimizer_ft=optim.SGD(model_ft.parameters(),lr=0.001,momentum=0.9)
exp_lr_scheduler=lr_scheduler.StepLR(optimizer_ft,step_size=7,gamma=0.1)
训练和评估
CPU上需要大约15-25分钟。但是在GPU上,它只需不到一分钟。
可惜我的GPU没法用,丢……
model_ft = train_model(model_ft, criterion, optimizer_ft, exp_lr_scheduler, num_epochs=25)
ConvNet as fixed feature extractor
在这里,我们需要冻结除最后一层之外的所有网络。我们需要设置 requires_grad == False 冻结参数,以便在 backward() 中不计算梯度。
model_conv = torchvision.models.resnet18(pretrained=True)
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
# opposed 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)
训练和评估
在CPU上,与前一个场景相比,这将花费大约一半的时间。这是预期的,因为不需要为大多数网络计算梯度。但是,前向传递需要计算梯度。
model_conv = train_model(model_conv, criterion, optimizer_conv, exp_lr_scheduler, num_epochs=25)