dragen1860 / deep-learning-with-pytorch-tutorials Goto Github PK

你好，龙老师，关于 lesson66 - 图卷积网络GCN，什么时候会在网易云课堂上更新呢

Early-stopping, Dropout这个章节的代码没有使用dropout，和visdom可视化那个章节的代码相同，请老师修正

您好,当我在学习lesson63迁移学习-自定义数据集实战时,运行train_transfer.py,遇到了如下的错误:
Process finished with exit code 139 (interrupted by signal 11: SIGSEGV).
我尝试去解决去它,但是都没能成功!
你能帮助我解决它!
感谢!

这一节，你举例lenet，卷积核是6x5x5，所以可以把1x32x32的图片-》6x28x28，你视频里讲成并写成3x3

When I run your handwritten numbers , I don't know what should I do after that and how can I try my own images? Please reply me . Thank you very much . ^^

老师您好，Lesson30 Visdom 可视化程序内没有visdom相应内容，只是MLP对MNIST的分类程序

I met this problem(One of the differentiated Tensors appears to not have been used in the graph.),how can i solve it ?

So how to use data in the latest version of torchtext?

In version 0.9, Torchtext.data.Field was moved to torchtext.legacy.data.Field.here

In version 0.12, they had removed the legacy folder in this release.here

龙龙老师，numpy实战bp的源码貌似没有，我想参考一下如何加载mnist数据集。

课程讲得很好很容易理解，但发现了个问题，第8节，分类问题的引入-2中的one hot编码应该是 [0 0 0 1 ... 0] 吧，视频中写成了[0 0 0 3 ... 0]