Swin-Transformer 图像分类网络实战项目：6种生活垃圾图像分类数据集（迁移学习）

[train hyper-parameters: Namespace(batch_size=16, num_classes=6, lr=0.001, lrf=0.01)] [epoch: 1] train loss:0.0451 test accuracy:0.8966 [epoch: 2] train loss:0.0149 test accuracy:0.8807 [epoch: 3] train loss:0.0075 test accuracy:0.9225 [epoch: 4] train loss:0.0051 test accuracy:0.9304 [epoch: 5] train loss:0.0026 test accuracy:0.9344 [epoch: 6] train loss:0.0017 test accuracy:0.9423 [epoch: 7] train loss:0.0011 test accuracy:0.9423 [epoch: 8] train loss:0.0009 test accuracy:0.9463 [epoch: 9] train loss:0.0007 test accuracy:0.9463 [epoch: 10] train loss:0.0006 test accuracy:0.9443