经典CNN网络之Resnet 图像分类网络实战项目：25种宝石图像分类数据集（迁移学习）

[train hyper-parameters: Namespace(batch_size=16, lr=0.001, lrf=0.01)] [epoch: 1] train loss:0.1832 test accuracy:0.5490 [epoch: 2] train loss:0.0935 test accuracy:0.8039 [epoch: 3] train loss:0.0453 test accuracy:0.8235 [epoch: 4] train loss:0.0316 test accuracy:0.8954 [epoch: 5] train loss:0.0213 test accuracy:0.8824 [epoch: 6] train loss:0.0169 test accuracy:0.8889 [epoch: 7] train loss:0.0138 test accuracy:0.8954 [epoch: 8] train loss:0.0131 test accuracy:0.8889 [epoch: 9] train loss:0.0117 test accuracy:0.9085 [epoch: 10] train loss:0.0111 test accuracy:0.8824