神经网络模型---ResNet

一、ResNet

1.导入包

import tensorflow as tf
from tensorflow.keras import layers, models, datasets, optimizers

optimizers是用于更新模型参数以最小化损失函数的算法

2.加载数据集、归一化、转为独热编码的内容一致

3.增加颜色通道

train_images = train_images[..., tf.newaxis].astype("float32")
test_images = test_images[..., tf.newaxis].astype("float32")

在train_images和test_imag