迁移学习教程
译者：片刻
校对者：cluster
在本教程中，您将学习如何使用迁移学习来训练您的网络。您可以在 cs231n 笔记 上阅读更多关于迁移学习的信息
引用这些笔记：
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 对感兴趣的任务进行初始化或用作固定特征提取器。
如下是两个主要的迁移学习场景：
- 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。(此微调过程对应引用中所说的初始化)
- 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。(此步对应引用中的固定特征提取器)
# 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
加载数据
我们将使用 torchvision 和 torch.utils.data 包来加载数据。
我们今天要解决的问题是训练一个模型来对 蚂蚁 和 蜜蜂 进行分类。我们有大约120个训练图像，每个图像用于 蚂蚁 和 蜜蜂。每个类有75个验证图像。通常，如果从头开始训练，这是一个非常小的数据集。由于我们正在使用迁移学习，我们应该能够合理地泛化。
该数据集是 imagenet 的一个非常小的子集。
注意
从 此处 下载数据并将其解压缩到当前目录。
# Data augmentation and normalization for training
# Just normalization for validation
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")
可视化一些图像
让我们可视化一些训练图像，以便了解数据增强。
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])
训练模型
现在, 让我们编写一个通用函数来训练模型. 这里, 我们将会举例说明:
- 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)
# 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
可视化模型预测
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.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)
微调卷积网络
加载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()
# 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)
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。
model_ft = train_model(model_ft, criterion, optimizer_ft, exp_lr_scheduler,
num_epochs=25)
Out:
Epoch 0/24
----------
train Loss: 0.6022 Acc: 0.6844
val Loss: 0.1765 Acc: 0.9412
Epoch 1/24
----------
train Loss: 0.4156 Acc: 0.8238
val Loss: 0.2380 Acc: 0.9216
Epoch 2/24
----------
train Loss: 0.5010 Acc: 0.7951
val Loss: 0.2571 Acc: 0.8954
Epoch 3/24
----------
train Loss: 0.7152 Acc: 0.7705
val Loss: 0.2060 Acc: 0.9346
Epoch 4/24
----------
train Loss: 0.5779 Acc: 0.8033
val Loss: 0.4542 Acc: 0.8889
Epoch 5/24
----------
train Loss: 0.5653 Acc: 0.7951
val Loss: 0.3167 Acc: 0.8824
Epoch 6/24
----------
train Loss: 0.4948 Acc: 0.8074
val Loss: 0.3238 Acc: 0.8758
Epoch 7/24
----------
train Loss: 0.3712 Acc: 0.8361
val Loss: 0.2284 Acc: 0.9020
Epoch 8/24
----------
train Loss: 0.2982 Acc: 0.8730
val Loss: 0.3488 Acc: 0.8497
Epoch 9/24
----------
train Loss: 0.2491 Acc: 0.8934
val Loss: 0.2405 Acc: 0.8889
Epoch 10/24
----------
train Loss: 0.3498 Acc: 0.8238
val Loss: 0.2435 Acc: 0.8889
Epoch 11/24
----------
train Loss: 0.3042 Acc: 0.8648
val Loss: 0.3021 Acc: 0.8627
Epoch 12/24
----------
train Loss: 0.2500 Acc: 0.8852
val Loss: 0.2340 Acc: 0.8954
Epoch 13/24
----------
train Loss: 0.3246 Acc: 0.8730
val Loss: 0.2236 Acc: 0.9020
Epoch 14/24
----------
train Loss: 0.2976 Acc: 0.8566
val Loss: 0.2928 Acc: 0.8562
Epoch 15/24
----------
train Loss: 0.2733 Acc: 0.8934
val Loss: 0.2370 Acc: 0.8954
Epoch 16/24
----------
train Loss: 0.3502 Acc: 0.8361
val Loss: 0.2792 Acc: 0.8824
Epoch 17/24
----------
train Loss: 0.2215 Acc: 0.8975
val Loss: 0.2790 Acc: 0.8497
Epoch 18/24
----------
train Loss: 0.3929 Acc: 0.8484
val Loss: 0.2648 Acc: 0.8824
Epoch 19/24
----------
train Loss: 0.3227 Acc: 0.8607
val Loss: 0.2643 Acc: 0.8693
Epoch 20/24
----------
train Loss: 0.3816 Acc: 0.8484
val Loss: 0.2395 Acc: 0.9085
Epoch 21/24
----------
train Loss: 0.2904 Acc: 0.8975
val Loss: 0.2399 Acc: 0.8889
Epoch 22/24
----------
train Loss: 0.3375 Acc: 0.8648
val Loss: 0.2380 Acc: 0.9020
Epoch 23/24
----------
train Loss: 0.2107 Acc: 0.9139
val Loss: 0.2251 Acc: 0.9085
Epoch 24/24
----------
train Loss: 0.3243 Acc: 0.8525
val Loss: 0.2545 Acc: 0.8824
Training complete in 1m 7s
Best val Acc: 0.941176
visualize_model(model_ft)
ConvNet 作为固定特征提取器
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。
您可以在 此处 的文档中阅读更多相关信息。
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)
Out:
Epoch 0/24
----------
train Loss: 0.5666 Acc: 0.6967
val Loss: 0.2794 Acc: 0.8824
Epoch 1/24
----------
train Loss: 0.5590 Acc: 0.7582
val Loss: 0.1473 Acc: 0.9477
Epoch 2/24
----------
train Loss: 0.4187 Acc: 0.8156
val Loss: 0.3534 Acc: 0.8693
Epoch 3/24
----------
train Loss: 0.5248 Acc: 0.7459
val Loss: 0.1848 Acc: 0.9477
Epoch 4/24
----------
train Loss: 0.4315 Acc: 0.8115
val Loss: 0.1640 Acc: 0.9477
Epoch 5/24
----------
train Loss: 0.3948 Acc: 0.8238
val Loss: 0.1609 Acc: 0.9542
Epoch 6/24
----------
train Loss: 0.3359 Acc: 0.8648
val Loss: 0.1734 Acc: 0.9608
Epoch 7/24
----------
train Loss: 0.3681 Acc: 0.8443
val Loss: 0.1715 Acc: 0.9477
Epoch 8/24
----------
train Loss: 0.4034 Acc: 0.8361
val Loss: 0.1602 Acc: 0.9477
Epoch 9/24
----------
train Loss: 0.2983 Acc: 0.8811
val Loss: 0.1561 Acc: 0.9542
Epoch 10/24
----------
train Loss: 0.4516 Acc: 0.7992
val Loss: 0.1660 Acc: 0.9477
Epoch 11/24
----------
train Loss: 0.3516 Acc: 0.8484
val Loss: 0.1551 Acc: 0.9542
Epoch 12/24
----------
train Loss: 0.3592 Acc: 0.8238
val Loss: 0.1525 Acc: 0.9477
Epoch 13/24
----------
train Loss: 0.2982 Acc: 0.8648
val Loss: 0.1772 Acc: 0.9542
Epoch 14/24
----------
train Loss: 0.3352 Acc: 0.8484
val Loss: 0.1583 Acc: 0.9542
Epoch 15/24
----------
train Loss: 0.2981 Acc: 0.8770
val Loss: 0.2133 Acc: 0.9412
Epoch 16/24
----------
train Loss: 0.2778 Acc: 0.8811
val Loss: 0.1934 Acc: 0.9542
Epoch 17/24
----------
train Loss: 0.3678 Acc: 0.8156
val Loss: 0.1846 Acc: 0.9477
Epoch 18/24
----------
train Loss: 0.3520 Acc: 0.8197
val Loss: 0.1577 Acc: 0.9542
Epoch 19/24
----------
train Loss: 0.3342 Acc: 0.8402
val Loss: 0.1734 Acc: 0.9542
Epoch 20/24
----------
train Loss: 0.3649 Acc: 0.8361
val Loss: 0.1554 Acc: 0.9412
Epoch 21/24
----------
train Loss: 0.2948 Acc: 0.8566
val Loss: 0.1878 Acc: 0.9542
Epoch 22/24
----------
train Loss: 0.3047 Acc: 0.8811
val Loss: 0.1760 Acc: 0.9477
Epoch 23/24
----------
train Loss: 0.3363 Acc: 0.8648
val Loss: 0.1660 Acc: 0.9542
Epoch 24/24
----------
train Loss: 0.2745 Acc: 0.8770
val Loss: 0.1853 Acc: 0.9542
Training complete in 0m 34s
Best val Acc: 0.960784
visualize_model(model_conv)
plt.ioff()
plt.show()
脚本总运行时间: (1分54.087秒)
Download Python source code: transfer_learning_tutorial.pyDownload Jupyter notebook: transfer_learning_tutorial.ipynb
