search

TangXu-Group / Remote-Sensing-Images-Classification

the official TangXu's group released codes about the Remote Sensing images classificaiton
24 stars 5 forks source link

您好，我设置的40轮为什么到第36就结束啦？ #1

Open 201811510210hjx opened 1 year ago

201811510210hjx commented 1 year ago

您好，我设置的40轮为什么到第36就结束啦？

显示这个：

epoch 36, train 100/100, lr=0.0000100, loss_total=1.6982,loss_proto=0.8535,loss_refined_proto=0.8523,R_fc1=0.1117,cf_acc=1.0000,proto_acc=1.0000 test_accuracy: 0.6672 best_accuracy:0.7042 ETA:38m/42m EarlyStopping counter: 15 out of 15 Early stopping!

lixingyi19 commented 11 months ago

您好，请问您能提供一下您的源码吗？我在训练的时候遇到了Function 'XlogyBackward0' returned nan values in its 1th output.的错误

liuyan-design commented 4 months ago

您好，我设置的40轮为什么到第36就结束啦？

显示这个：

epoch 36, train 100/100, lr=0.0000100, loss_total=1.6982,loss_proto=0.8535,loss_refined_proto=0.8523,R_fc1=0.1117,cf_acc=1.0000,proto_acc=1.0000 test_accuracy: 0.6672 best_accuracy:0.7042 ETA:38m/42m EarlyStopping counter: 15 out of 15 Early stopping!

你好，我还没开始训练，数据读取就有问题(https://github.com/TangXu-Group/Remote-Sensing-Images-Classification/issues/5#issue-2407441229). 可以分享一下数据目录结构么？

liuyan-design commented 4 months ago

您好，请问您能提供一下您的源码吗？我在训练的时候遇到了Function 'XlogyBackward0' returned nan values in its 1th output.的错误

你好，我还没开始训练，数据读取就有问题(https://github.com/TangXu-Group/Remote-Sensing-Images-Classification/issues/5#issue-2407441229). 可以分享一下数据目录结构么？