ResNet18加SENet实现cifar10分类training准确率95.66test准确率90.77 基于TensorFlow2利用ResNet18+SENet 实现cifar10分类 training准确率95.66%,test准确率90.77%