加载cifar10和cifar100数据集

import numpy as np
import tensorflow as tf
from tensorflow.keras import layers, models
# Load and normalize CIFAR-10 and CIFAR-100 datasets
(train_images_10, train_labels_10), (test_images_10, test_labels_10) = tf.keras.datasets.cifar10.load_data()
(train_images_100, train_labels_100), (test_images_100, test_labels_100) = tf.keras.datasets.cifar100.load_data()

train_images = np.concatenate((train_images_10, train_images_100), axis=0) / 255.0
test_images = np.concatenate((test_images_10, test_images_100), axis=0) / 255.0
train_labels = np.concatenate((train_labels_10, train_labels_100 + 10), axis=0)
test_labels = np.concatenate((test_labels_10, test_labels_100 + 10), axis=0)

使用函数 ‘tf.keras.datasets.cifar10.load_data()’ 函数加载cifar10的训练（样本、标签） 测试（样本、标签），同理 cifar100

使用np.concatenate 拼接两个训练数据集， train_10 和train_100 , axis=0就是垂直方向拼接 ，

对每个像素执行归一化， 将每个像素/255 ， 缩放到【0，1】

同理， 对标签进行合并。