可视化数据、模型和训练过程
【目标】
- 学习读入数据并转换
- 建立 TensorBoard
- 写入 TensorBoard
- 使用 TensorBoard 检查模型架构
- 使用 TensorBoard 创建可视化交互版本
- 检查训练数据的几种方法
- 跟踪训练中的模型性能
- 训练后模型性能的评估
开始
import matplotlib.pyplot as plt
import numpy as np
import torch
import torchvision
import torchvision.transforms as transforms
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
## 转换
transform = transforms.Compose(
[transforms.ToTensor(),
transforms.Normalize((0.5), (0.5))]
)
## 数据集
trainset = torchvision.datasets.FashionMNIST(
root = "../../datasets",
download = True,
train = True,
transform = transform
)
testset = torchvision.datasets.FashionMNIST(
root = "../../datasets",
download = True,
train = False,
transform = transform
)
## dataloader
trainloader = torch.utils.data.DataLoader(trainset, batch_size = 4, shuffle = True)
testloader = torch.utils.data.DataLoader(testset, batch_size = 4, shuffle = True)
## 类别
classes = ('T-shirt/top', 'Trouser', 'Pullover', 'Dress', 'Coat',
'Sandal', 'Shirt', 'Sneaker', 'Bag', 'Ankle Boot')
## 显示数据
def matplotlib_show(img, one_channel=False):
if one_channel:
img = img.mean(dim=0)
img = img / 2 + 0.5
npimg = img.numpy()
if one_channel:
plt.imshow(npimg, cmap='Greys')
else:
plt.imshow(np.transpose(npimg, (1, 2, 0)))
## 模型
class Net(nn.Module):
def __init__(self):
super(Net, self).__init__()
self.conv1 = nn.Conv2d(1, 6, 5)
self.pool1 = nn.MaxPool2d(2, 2)
self.conv2 = nn.Conv2d(6, 16, 5)
self.pool2 = nn.MaxPool2d(2, 2)
self.fc1 = nn.Linear(16 * 4 * 4, 120)
self.fc2 = nn.Linear(120, 84)
self.fc3 = nn.Linear(84, 10)
def forward(self, x):
x = self.pool1(F.relu(self.conv1(x)))
x = self.pool2(F.relu(self.conv2(x)))
x = x.view(-1, 16 * 4 * 4)
x = F.relu(self.fc1(x))
x = F.relu(self.fc2(x))
x = self.fc3(x)
return x
## 实例化
net = Net()
criterion = nn.CrossEntropyLoss()
optimizer = optim.SGD(net.parameters(), lr = 0.001, momentum = 0.9)
建立 TensorBoard
from torch.utils.tensorboard import SummaryWriter
writer = SummaryWriter("../runs/fashion_mnist_exp")
写入 TensorBoard
让我们将图像写入到 TensorBoard, 用 make_grid
dataiter = iter(trainloader)
images, labels = next(dataiter)
image_grid = torchvision.utils.make_grid(images)
matplotlib_show(image_grid, one_channel=True)
writer.add_image('images', image_grid)
jupyter notebook
可以支持 TensorBoard
.
- 运行完后,点击 启动TensorBoard 会话(也可以在命令行执行)
- 如果是脚本文件的话,通过命令行执行
tensorboard --logdir=runs
,runs
为路径名
使用 TensorBoard 检查模型架构
TensorBoard
一个非常强的功能是可视化复杂的模型结构。
writer.add_graph(net, images)
writer.close()
执行以上语句,并刷新 TensorBoard
, 就可以看到多了一个 GRAPH
。
双击 Net
就可以展开网络结构;TensorBoard 有一个非常方便的功能就是在低维空间可视化高维数据。
添加一个 Projector 到 TensorBoard 中
可视化高维数据在低维空间的表达 add_embedding 方法。
跟踪模型训练
现在,我们将把运行 loss 记录到 TensorBoard ,并查看模型通过 plot_classes_preds 函数所做的预测。
def images_to_probs(net, images):
'''
Generates predictions and corresponding probabilities from a trained
network and a list of images
'''
outputs = net(images)
_, preds_tensor = torch.max(outputs, 1)
preds = np.squeeze(preds_tensor.numpy())
return preds, [F.softmax(el, dim=0)[i].item() for i, el in zip(preds, outputs)]
def plot_classes_preds(net, images, labels):
'''
Generates matplotlib Figure using a trained network, along with images
and labels from a batch, that shows the network's top prediction along
with its probability, alongside the actual label, coloring this
information based on whether the prediction was correct or not.
Uses the "images_to_probs" function.
'''
preds, probs = images_to_probs(net, images)
# plot the images in the batch, along with predicted and true labels
fig = plt.figure(figsize=(12, 48))
for idx in np.arange(4):
ax = fig.add_subplot(1, 4, idx+1, xticks=[], yticks=[])
matplotlib_show(images[idx], one_channel=True)
ax.set_title("{0}, {1:.1f}%\n(label: {2})".format(
classes[preds[idx]],
probs[idx] * 100.0,
classes[labels[idx]]),
color=("green" if preds[idx] == labels[idx].item() else "red"))
return fig
running_loss = 0.0
for epoch in range(10): # loop over the dataset multiple times
for i, data in enumerate(trainloader, 0):
# get the inputs; data is a list of [inputs, labels]
inputs, labels = data
# zero the parameter gradients
optimizer.zero_grad()
# forward + backward + optimize
outputs = net(inputs)
loss = criterion(outputs, labels)
loss.backward()
optimizer.step()
running_loss += loss.item()
if i % 1000 == 999: # every 1000 mini-batches...
# ...log the running loss
writer.add_scalar('training loss',
running_loss / 1000,
epoch * len(trainloader) + i)
# ...log a Matplotlib Figure showing the model's predictions on a
# random mini-batch
writer.add_figure('predictions vs. actuals',
plot_classes_preds(net, inputs, labels),
global_step=epoch * len(trainloader) + i)
running_loss = 0.0
print('Finished Training')
模型评估
class_probs = []
class_label = []
with torch.no_grad():
for data in testloader:
images, labels = data
output = net(images)
class_probs_batch = [F.softmax(el, dim=0) for el in output]
class_probs.append(class_probs_batch)
class_label.append(labels)
test_probs = torch.cat([torch.stack(batch) for batch in class_probs])
test_label = torch.cat(class_label)
def add_pr_curve_tensorboard(class_index, test_probs, test_label, global_step = 0):
tensorboard_truth = test_label == class_index
tensorboard_probs = test_probs[:, class_index]
writer.add_pr_curve(
classes[class_index],
tensorboard_truth,
tensorboard_probs,
global_step=global_step)
writer.close()
# plot all pr curves
for i in range(len(classes)):
add_pr_curve_tensorboard(i, test_probs, test_label)
可以在上面的菜单栏中看到多了一个菜单 PR CURVES
, 每个类别的 PR 曲线都不一样。