摘要：
记录MindSpore AI框架使用ResNet50神经网络模型，选择Bottleneck残差网络结构对CIFAR-10数据集进行分类的过程、步骤和方法。包括环境准备、下载数据集、数据集加载和预处理、构建模型、模型训练、模型测试等。
一、 概 念
1. 图像分类
最基础的计算机视觉应用
有监督学习类别
给定一张图像(猫、狗、飞机、汽车等等)
判断图像所属的类别
使用ResNet50网络
对CIFAR-10数据集进行分类
2.ResNet网络
ResNet50网络
2015年微软实验室提出
ILSVRC2015图像分类竞赛第一名
传统卷积神经网络
一系列卷积层和池化层堆叠
堆叠到一定深度时会出现退化问题
56层网络与20层网络训练误差和测试误差图
CIFAR-10数据集
56层网络比20层网络训练误差和测试误差更大
随着网络加深，误差并没有减小
3. 残差网络结构
Residual Network
减轻退化问题
实现搭建较深的网络结构（突破1000层）
ResNet网络在CIFAR-10数据集上的训练误差与测试误差图
虚线 训练误差
实线 测试误差
网络层数越深，训练误差和测试误差越小
二、环境准备
%%capture captured_output
# 实验环境已经预装了mindspore==2.2.14，如需更换mindspore版本，可更改下面mindspore的版本号
!pip uninstall mindspore -y
!pip install -i https://pypi.mirrors.ustc.edu.cn/simple mindspore==2.2.14
# 查看当前 mindspore 版本
!pip show mindspore输出：
Name: mindspore
Version: 2.2.14
Summary: MindSpore is a new open source deep learning training/inference framework that could be used for mobile, edge and cloud scenarios.
Home-page: https://www.mindspore.cn
Author: The MindSpore Authors
Author-email: [email protected]
License: Apache 2.0
Location: /home/nginx/miniconda/envs/jupyter/lib/python3.9/site-packages
Requires: asttokens, astunparse, numpy, packaging, pillow, protobuf, psutil, scipy
Required-by: 三、 数据集准备与加载
1.数据集
60000张32*32的彩色图像
50000张训练图片
10000张评估图片
10个类别
每类有6000张图
2.下载数据集
download接口
下载
解压
仅支持解析二进制版本的CIFAR-10文件（CIFAR-10 binary version）
from download import download
url = "https://mindspore-website.obs.cn-north-4.myhuaweicloud.com/notebook/datasets/cifar-10-binary.tar.gz"
download(url, "./datasets-cifar10-bin", kind="tar.gz", replace=True)输出：
Downloading data from https://mindspore-website.obs.cn-north-4.myhuaweicloud.com/notebook/datasets/cifar-10-binary.tar.gz (162.2 MB)
file_sizes: 100%|█████████████████████████████| 170M/170M [00:01<00:00, 113MB/s]
Extracting tar.gz file...
Successfully downloaded / unzipped to ./datasets-cifar10-bin
'./datasets-cifar10-bin'数据集目录结构：
datasets-cifar10-bin/cifar-10-batches-bin
├── batches.meta.text
├── data_batch_1.bin
├── data_batch_2.bin
├── data_batch_3.bin
├── data_batch_4.bin
├── data_batch_5.bin
├── readme.html
└── test_batch.bin3.加载 数据 集
mindspore.dataset.Cifar10Dataset接口
加载数据集
图像增强操作
import mindspore as ms
import mindspore.dataset as ds
import mindspore.dataset.vision as vision
import mindspore.dataset.transforms as transforms
from mindspore import dtype as mstype
data_dir = "./datasets-cifar10-bin/cifar-10-batches-bin" # 数据集根目录
batch_size = 256 # 批量大小
image_size = 32 # 训练图像空间大小
workers = 4 # 并行线程个数
num_classes = 10 # 分类数量
def create_dataset_cifar10(dataset_dir, usage, resize, batch_size, workers):
data_set = ds.Cifar10Dataset(dataset_dir=dataset_dir,
usage=usage,
num_parallel_workers=workers,
shuffle=True)
trans = []
if usage == "train":
trans += [
vision.RandomCrop((32, 32), (4, 4, 4, 4)),
vision.RandomHorizontalFlip(prob=0.5)
]
trans += [
vision.Resize(resize),
vision.Rescale(1.0 / 255.0, 0.0),
vision.Normalize([0.4914, 0.4822, 0.4465], [0.2023, 0.1994, 0.2010]),
vision.HWC2CHW()
]
target_trans = transforms.TypeCast(mstype.int32)
# 数据映射操作
data_set = data_set.map(operations=trans,
input_columns='image',
num_parallel_workers=workers)
data_set = data_set.map(operations=target_trans,
input_columns='label',
num_parallel_workers=workers)
# 批量操作
data_set = data_set.batch(batch_size)
return data_set
# 获取处理后的训练与测试数据集
dataset_train = create_dataset_cifar10(dataset_dir=data_dir,
usage="train",
resize=image_size,
batch_size=batch_size,
workers=workers)
step_size_train = dataset_train.get_dataset_size()
dataset_val = create_dataset_cifar10(dataset_dir=data_dir,
usage="test",
resize=image_size,
batch_size=batch_size,
workers=workers)
step_size_val = dataset_val.get_dataset_size()4.显示 CIFAR-10训练数据集
import matplotlib.pyplot as plt
import numpy as np
data_iter = next(dataset_train.create_dict_iterator())
images = data_iter["image"].asnumpy()
labels = data_iter["label"].asnumpy()
print(f"Image shape: {images.shape}, Label shape: {labels.shape}")
# 训练数据集中，前六张图片所对应的标签
print(f"Labels: {labels[:6]}")
classes = []
with open(data_dir + "/batches.meta.txt", "r") as f:
for line in f:
line = line.rstrip()
if line:
classes.append(line)
# 训练数据集的前六张图片
plt.figure()
for i in range(6):
plt.subplot(2, 3, i + 1)
image_trans = np.transpose(images[i], (1, 2, 0))
mean = np.array([0.4914, 0.4822, 0.4465])
std = np.array([0.2023, 0.1994, 0.2010])
image_trans = std * image_trans + mean
image_trans = np.clip(image_trans, 0, 1)
plt.title(f"{classes[labels[i]]}")
plt.imshow(image_trans)
plt.axis("off")
plt.show()输出：
Image shape: (256, 3, 32, 32), Label shape: (256,)
Labels: [3 3 6 4 7 4]
四、 构建网络
残差网络结构(Residual Network)
有效减轻ResNet退化问题
实现更深的网络结构设计
提高网络的训练精度
堆叠残差网络构建ResNet50网络
1. 构建残差网络结构
残差网络结构图
残差网络由两个分支构成
主分支
堆叠系列卷积操作得到
输出的特征矩阵()
shortcuts（图中弧线表示）
从输入直接到输出
主分支F(x)加上shortcuts输出的特征矩阵x得到F(x)+x
Relu激活函数
输出
残差网络结构主要由两种
Building Block
用于较浅的ResNet网络，如ResNet18和ResNet34
Bottleneck
用于层数较深的ResNet网络，如ResNet50、ResNet101和ResNet152
Building Block
Building Block结构图
主分支有两层卷积网络结构：
第一层网络
输入channel为64
3×33×3卷积层
Batch Normalization层
Relu激活函数层
输出channel为64
第二层网络
输入channel为64
3×33×3的卷积层
Batch Normalization层
输出channel为64
融合
主分支输出的特征矩阵
shortcuts输出的特征矩阵
保证shape相同
输出Relu激活函数
主分支与shortcuts输出的特征矩阵相加
如果shape不相同
如输出channel是输入channel的一倍时，
shortcuts需要使用数量与输出channel相等
大小为1×11×1的卷积核进行卷积操作
若输出的图像较输入图像缩小一倍
设置shortcuts中卷积操作中的stride为2
主分支第一层卷积操作的stride也需设置为2
定义ResidualBlockBase类实现Building Block结构代码:
from typing import Type, Union, List, Optional
import mindspore.nn as nn
from mindspore.common.initializer import Normal
# 初始化卷积层与BatchNorm的参数
weight_init = Normal(mean=0, sigma=0.02)
gamma_init = Normal(mean=1, sigma=0.02)
class ResidualBlockBase(nn.Cell):
expansion: int = 1 # 最后一个卷积核数量与第一个卷积核数量相等
def __init__(self, in_channel: int, out_channel: int,
stride: int = 1, norm: Optional[nn.Cell] = None,
down_sample: Optional[nn.Cell] = None) -> None:
super(ResidualBlockBase, self).__init__()
if not norm:
self.norm = nn.BatchNorm2d(out_channel)
else:
self.norm = norm
self.conv1 = nn.Conv2d(in_channel, out_channel,
kernel_size=3, stride=stride,
weight_init=weight_init)
self.conv2 = nn.Conv2d(in_channel, out_channel,
kernel_size=3, weight_init=weight_init)
self.relu = nn.ReLU()
self.down_sample = down_sample
def construct(self, x):
"""ResidualBlockBase construct."""
identity = x # shortcuts分支
out = self.conv1(x) # 主分支第一层：3*3卷积层
out = self.norm(out)
out = self.relu(out)
out = self.conv2(out) # 主分支第二层：3*3卷积层
out = self.norm(out)
if self.down_sample is not None:
identity = self.down_sample(x)
out += identity # 输出为主分支与shortcuts之和
out = self.relu(out)
return outBottleneck
Bottleneck结构图
Bottleneck结构的参数数量更少
更适合层数较深的网络
主分支有三层卷积结构
1×11×1的卷积层
输入channel为256
通过数量为64
降维
Batch Normalization层
Relu激活函数层
输出channel为64
3×33×3卷积层
通过数量为64
Batch Normalization层
Relu激活函数层
输出channel为64
1×11×1的卷积层
升维
通过数量为256
Batch Normalization层
输出channel为256
融合
主分支输出的特征矩阵
shortcuts输出的特征矩阵
保证特征矩阵shape相同
输出Relu激活函数
主分支与shortcuts输出的特征矩阵相加
如果shape不相同
如输出channel是输入channel的一倍时，
shortcuts上需要使用数量与输出channel相等
大小为1×11×1的卷积核进行卷积操作
如输出的图像较输入图像缩小一倍
设置shortcuts中卷积操作中的stride为2
主分支第二层卷积操作的stride也需设置为2
定义ResidualBlock类实现Bottleneck结构代码：
class ResidualBlock(nn.Cell):
expansion = 4 # 最后一个卷积核的数量是第一个卷积核数量的4倍
def __init__(self, in_channel: int, out_channel: int,
stride: int = 1, down_sample: Optional[nn.Cell] = None) -> None:
super(ResidualBlock, self).__init__()
self.conv1 = nn.Conv2d(in_channel, out_channel,
kernel_size=1, weight_init=weight_init)
self.norm1 = nn.BatchNorm2d(out_channel)
self.conv2 = nn.Conv2d(out_channel, out_channel,
kernel_size=3, stride=stride,
weight_init=weight_init)
self.norm2 = nn.BatchNorm2d(out_channel)
self.conv3 = nn.Conv2d(out_channel, out_channel * self.expansion,
kernel_size=1, weight_init=weight_init)
self.norm3 = nn.BatchNorm2d(out_channel * self.expansion)
self.relu = nn.ReLU()
self.down_sample = down_sample
def construct(self, x):
identity = x # shortscuts分支
out = self.conv1(x) # 主分支第一层：1*1卷积层
out = self.norm1(out)
out = self.relu(out)
out = self.conv2(out) # 主分支第二层：3*3卷积层
out = self.norm2(out)
out = self.relu(out)
out = self.conv3(out) # 主分支第三层：1*1卷积层
out = self.norm3(out)
if self.down_sample is not None:
identity = self.down_sample(x)
out += identity # 输出为主分支与shortcuts之和
out = self.relu(out)
return out2. 构建ResNet50网络
ResNet网络层结构
以输入彩色图像224×224为例
通过数量64
卷积核大小为7×7
stride为2的卷积层conv1
该层输出图片大小为112×112
输出channel为64；
通过3×3最大下采样池化层
输出图片大小为56×56
输出channel为64；
再堆叠4个残差网络块
conv2_x
conv3_x
conv4_x
conv5_x
输出图片大小为7×7
输出channel为2048
最后通过
平均池化层
全连接层
Softmax
得到分类概率。
以ResNet50网络中的conv2_x为例
每个残差网络块
由3个Bottleneck结构堆叠而成
每个Bottleneck
输入channel为64
输出channel为256
示例
定义make_layer实现残差块的构建，
参数如下:
last_out_channel：上一个残差网络输出的通道数
block： 残差网络的类别
ResidualBlockBase
ResidualBlock
channel： 残差网络输入的通道数
block_nums： 残差网络块堆叠的个数
stride： 卷积移动的步幅
def make_layer(last_out_channel, block: Type[Union[ResidualBlockBase, ResidualBlock]],
channel: int, block_nums: int, stride: int = 1):
down_sample = None # shortcuts分支
if stride != 1 or last_out_channel != channel * block.expansion:
down_sample = nn.SequentialCell([
nn.Conv2d(last_out_channel, channel * block.expansion,
kernel_size=1, stride=stride, weight_init=weight_init),
nn.BatchNorm2d(channel * block.expansion, gamma_init=gamma_init)
])
layers = []
layers.append(block(last_out_channel, channel, stride=stride, down_sample=down_sample))
in_channel = channel * block.expansion
# 堆叠残差网络
for _ in range(1, block_nums):
layers.append(block(in_channel, channel))
return nn.SequentialCell(layers)ResNet50网络
5个卷积结构
1个平均池化层
1个全连接层
以CIFAR-10数据集为例：
conv1
输入图片大小为32×32
输入channel为3
卷积层
卷积核数量为64
卷积核大小为7×7
stride为2
Batch Normalization层
Reul激活函数
输出feature map大小为16×16
输出channel为64
conv2_x
输入feature map大小为16×16
输入channel为64
最大下采样池化操作
卷积核大小为3×3
stride为2
堆叠3个Bottleneck
1×1，64； 3×3，64； 1×1，256\]结构 输出feature map大小为8×8 输出channel为256 conv3_x 输入feature map大小为8×8 输入channel为256 堆叠4个Bottleneck \[1×1，128； 3×3，128； 1×1，512\]结构 输出feature map大小为4×4 输出channel为512 conv4_x 输入feature map大小为4×4 输入channel为512 堆叠6个Bottleneck \[1×1，256； 3×3，256； 1×1，1024\]结构 输出feature map大小为2×2 输出channel为1024 conv5_x 输入feature map大小为2×2 输入channel为1024 堆叠3个Bottleneck \[1×1，512； 3×3，512； 1×1，2048\]结构。 输出feature map大小为1×1 输出channel为2048 average pool \& fc 输入channel为2048 输出channel为分类的类别数
ResNet50模型构建代码
函数resnet50参数：
num_classes： 分类的类别数
默认类别数为1000。
Pretrained ： 下载对应的训练模型
加载预训练模型中的参数
from mindspore import load_checkpoint, load_param_into_net
class ResNet(nn.Cell):
def __init__(self, block: Type[Union[ResidualBlockBase, ResidualBlock]],
layer_nums: List[int], num_classes: int, input_channel: int) -> None:
super(ResNet, self).__init__()
self.relu = nn.ReLU()
# 第一个卷积层，输入channel为3（彩色图像），输出channel为64
self.conv1 = nn.Conv2d(3, 64, kernel_size=7, stride=2, weight_init=weight_init)
self.norm = nn.BatchNorm2d(64)
# 最大池化层，缩小图片的尺寸
self.max_pool = nn.MaxPool2d(kernel_size=3, stride=2, pad_mode='same')
# 各个残差网络结构块定义
self.layer1 = make_layer(64, block, 64, layer_nums[0])
self.layer2 = make_layer(64 * block.expansion, block, 128, layer_nums[1], stride=2)
self.layer3 = make_layer(128 * block.expansion, block, 256, layer_nums[2], stride=2)
self.layer4 = make_layer(256 * block.expansion, block, 512, layer_nums[3], stride=2)
# 平均池化层
self.avg_pool = nn.AvgPool2d()
# flattern层
self.flatten = nn.Flatten()
# 全连接层
self.fc = nn.Dense(in_channels=input_channel, out_channels=num_classes)
def construct(self, x):
x = self.conv1(x)
x = self.norm(x)
x = self.relu(x)
x = self.max_pool(x)
x = self.layer1(x)
x = self.layer2(x)
x = self.layer3(x)
x = self.layer4(x)
x = self.avg_pool(x)
x = self.flatten(x)
x = self.fc(x)
return x
def _resnet(model_url: str, block: Type[Union[ResidualBlockBase, ResidualBlock]],
layers: List[int], num_classes: int, pretrained: bool, pretrained_ckpt: str,
input_channel: int):
model = ResNet(block, layers, num_classes, input_channel)
if pretrained:
# 加载预训练模型
download(url=model_url, path=pretrained_ckpt, replace=True)
param_dict = load_checkpoint(pretrained_ckpt)
load_param_into_net(model, param_dict)
return model
def resnet50(num_classes: int = 1000, pretrained: bool = False):
"""ResNet50模型"""
resnet50_url = "https://mindspore-website.obs.cn-north-4.myhuaweicloud.com/notebook/models/application/resnet50_224_new.ckpt"
resnet50_ckpt = "./LoadPretrainedModel/resnet50_224_new.ckpt"
return _resnet(resnet50_url, ResidualBlock, [3, 4, 6, 3], num_classes,
pretrained, resnet50_ckpt, 2048)五、 模型训练与评估
调用resnet50构造ResNet50模型
设置pretrained参数为True
自动下载ResNet50预训练模型
加载预训练模型中的参数到网络中。
定义优化器和损失函数
逐epoch打印训练的损失值和评估精度
保存评估精度最高的ckpt文件（resnet50-best.ckpt）到当前路径的./BestCheckPoint
预训练模型
全连接层（fc）输出大小（对应参数num_classes）默认为1000
CIFAR10数据集共有10个分类
重置全连接层输出大小为10
展示5个epochs的训练过程
# 定义ResNet50网络
network = resnet50(pretrained=True)
# 全连接层输入层的大小
in_channel = network.fc.in_channels
fc = nn.Dense(in_channels=in_channel, out_channels=10)
# 重置全连接层
network.fc = fc输出：
Downloading data from https://mindspore-website.obs.cn-north-4.myhuaweicloud.com/notebook/models/application/resnet50_224_new.ckpt (97.7 MB)
file_sizes: 100%|█████████████████████████████| 102M/102M [00:00<00:00, 109MB/s]
Successfully downloaded file to ./LoadPretrainedModel/resnet50_224_new.ckpt
# 设置学习率
num_epochs = 5
lr = nn.cosine_decay_lr(min_lr=0.00001, max_lr=0.001, total_step=step_size_train * num_epochs,
step_per_epoch=step_size_train, decay_epoch=num_epochs)
# 定义优化器和损失函数
opt = nn.Momentum(params=network.trainable_params(), learning_rate=lr, momentum=0.9)
loss_fn = nn.SoftmaxCrossEntropyWithLogits(sparse=True, reduction='mean')
def forward_fn(inputs, targets):
logits = network(inputs)
loss = loss_fn(logits, targets)
return loss
grad_fn = ms.value_and_grad(forward_fn, None, opt.parameters)
def train_step(inputs, targets):
loss, grads = grad_fn(inputs, targets)
opt(grads)
return loss
import os
# 创建迭代器
data_loader_train = dataset_train.create_tuple_iterator(num_epochs=num_epochs)
data_loader_val = dataset_val.create_tuple_iterator(num_epochs=num_epochs)
# 最佳模型存储路径
best_acc = 0
best_ckpt_dir = "./BestCheckpoint"
best_ckpt_path = "./BestCheckpoint/resnet50-best.ckpt"
if not os.path.exists(best_ckpt_dir):
os.mkdir(best_ckpt_dir)
import mindspore.ops as ops
def train(data_loader, epoch):
"""模型训练"""
losses = []
network.set_train(True)
for i, (images, labels) in enumerate(data_loader):
loss = train_step(images, labels)
if i % 100 == 0 or i == step_size_train - 1:
print('Epoch: [%3d/%3d], Steps: [%3d/%3d], Train Loss: [%5.3f]' %
(epoch + 1, num_epochs, i + 1, step_size_train, loss))
losses.append(loss)
return sum(losses) / len(losses)
def evaluate(data_loader):
"""模型验证"""
network.set_train(False)
correct_num = 0.0 # 预测正确个数
total_num = 0.0 # 预测总数
for images, labels in data_loader:
logits = network(images)
pred = logits.argmax(axis=1) # 预测结果
correct = ops.equal(pred, labels).reshape((-1, ))
correct_num += correct.sum().asnumpy()
total_num += correct.shape[0]
acc = correct_num / total_num # 准确率
return acc
# 开始循环训练
print("Start Training Loop ...")
for epoch in range(num_epochs):
curr_loss = train(data_loader_train, epoch)
curr_acc = evaluate(data_loader_val)
print("-" * 50)
print("Epoch: [%3d/%3d], Average Train Loss: [%5.3f], Accuracy: [%5.3f]" % (
epoch+1, num_epochs, curr_loss, curr_acc
))
print("-" * 50)
# 保存当前预测准确率最高的模型
if curr_acc > best_acc:
best_acc = curr_acc
ms.save_checkpoint(network, best_ckpt_path)
print("=" * 80)
print(f"End of validation the best Accuracy is: {best_acc: 5.3f}, "
f"save the best ckpt file in {best_ckpt_path}", flush=True)输出：
Start Training Loop ...
Epoch: [ 1/ 5], Steps: [ 1/196], Train Loss: [2.412]
Epoch: [ 1/ 5], Steps: [101/196], Train Loss: [1.384]
Epoch: [ 1/ 5], Steps: [196/196], Train Loss: [0.991]
--------------------------------------------------
Epoch: [ 1/ 5], Average Train Loss: [1.590], Accuracy: [0.606]
--------------------------------------------------
Epoch: [ 2/ 5], Steps: [ 1/196], Train Loss: [1.030]
Epoch: [ 2/ 5], Steps: [101/196], Train Loss: [0.892]
Epoch: [ 2/ 5], Steps: [196/196], Train Loss: [0.968]
--------------------------------------------------
Epoch: [ 2/ 5], Average Train Loss: [0.994], Accuracy: [0.692]
--------------------------------------------------
Epoch: [ 3/ 5], Steps: [ 1/196], Train Loss: [0.774]
Epoch: [ 3/ 5], Steps: [101/196], Train Loss: [0.950]
Epoch: [ 3/ 5], Steps: [196/196], Train Loss: [0.642]
--------------------------------------------------
Epoch: [ 3/ 5], Average Train Loss: [0.836], Accuracy: [0.721]
--------------------------------------------------
Epoch: [ 4/ 5], Steps: [ 1/196], Train Loss: [0.804]
Epoch: [ 4/ 5], Steps: [101/196], Train Loss: [0.824]
Epoch: [ 4/ 5], Steps: [196/196], Train Loss: [0.924]
--------------------------------------------------
Epoch: [ 4/ 5], Average Train Loss: [0.766], Accuracy: [0.737]
--------------------------------------------------
Epoch: [ 5/ 5], Steps: [ 1/196], Train Loss: [0.843]
Epoch: [ 5/ 5], Steps: [101/196], Train Loss: [0.756]
Epoch: [ 5/ 5], Steps: [196/196], Train Loss: [0.965]
--------------------------------------------------
Epoch: [ 5/ 5], Average Train Loss: [0.737], Accuracy: [0.738]
--------------------------------------------------
================================================================================
End of validation the best Accuracy is: 0.738, save the best ckpt file in ./BestCheckpoint/resnet50-best.ckpt六、可视化模型预测
定义visualize_model函数
使用验证精度最高的模型
预测CIFAR-10测试数据集
预测结果可视化
预测字体颜色为蓝色 表示预测正确
预测字体颜色为红色 表示预测错误
5 epochs模型在验证数据集的预测准确率在70%左右
6张图片中会有2张预测失败
要达到理想的训练效果，训练80个epochs
import matplotlib.pyplot as plt
def visualize_model(best_ckpt_path, dataset_val):
num_class = 10 # 对狼和狗图像进行二分类
net = resnet50(num_class)
# 加载模型参数
param_dict = ms.load_checkpoint(best_ckpt_path)
ms.load_param_into_net(net, param_dict)
# 加载验证集的数据进行验证
data = next(dataset_val.create_dict_iterator())
images = data["image"]
labels = data["label"]
# 预测图像类别
output = net(data['image'])
pred = np.argmax(output.asnumpy(), axis=1)
# 图像分类
classes = []
with open(data_dir + "/batches.meta.txt", "r") as f:
for line in f:
line = line.rstrip()
if line:
classes.append(line)
# 显示图像及图像的预测值
plt.figure()
for i in range(6):
plt.subplot(2, 3, i + 1)
# 若预测正确，显示为蓝色；若预测错误，显示为红色
color = 'blue' if pred[i] == labels.asnumpy()[i] else 'red'
plt.title('predict:{}'.format(classes[pred[i]]), color=color)
picture_show = np.transpose(images.asnumpy()[i], (1, 2, 0))
mean = np.array([0.4914, 0.4822, 0.4465])
std = np.array([0.2023, 0.1994, 0.2010])
picture_show = std * picture_show + mean
picture_show = np.clip(picture_show, 0, 1)
plt.imshow(picture_show)
plt.axis('off')
plt.show()
# 使用测试数据集进行验证
visualize_model(best_ckpt_path=best_ckpt_path, dataset_val=dataset_val)输出：
