经典CNN网络之AlexNet 对9种纸币面额识别【包括数据、代码、训练好了结果】

[train hyper-parameters: Namespace(batch_size=32, lr=0.001, lrf=0.01)] [epoch: 1] train loss:0.0483 test accuracy:0.1900 [epoch: 2] train loss:0.0245 test accuracy:0.6300 [epoch: 3] train loss:0.0088 test accuracy:0.4900 [epoch: 4] train loss:0.0079 test accuracy:0.3000 [epoch: 5] train loss:0.0058 test accuracy:0.2800 [epoch: 6] train loss:0.0022 test accuracy:0.1800 [epoch: 7] train loss:0.0016 test accuracy:0.3400 [epoch: 8] train loss:0.0014 test accuracy:0.3400 [epoch: 9] train loss:0.0014 test accuracy:0.2500 [epoch: 10] train loss:0.0009 test accuracy:0.3900 [epoch: 11] train loss:0.0007 test accuracy:0.2900 [epoch: 12] train loss:0.0004 test accuracy:0.3000 [epoch: 13] train loss:0.0004 test accuracy:0.3500 [epoch: 14] train loss:0.0004 test accuracy:0.3400 [epoch: 15] train loss:0.0003 test accuracy:0.3200 [epoch: 16] train loss:0.0004 test accuracy:0.3100 [epoch: 17] train loss:0.0002 test accuracy:0.3200 [epoch: 18] train loss:0.0003 test accuracy:0.3200 [epoch: 19] train loss:0.0004 test accuracy:0.3100 [epoch: 20] train loss:0.0003 test accuracy:0.3100