Ecosyste.ms: Issues
An open API service for providing issue and pull request metadata for open source projects.
GitHub / ldoublev/semi-supervised-gan issues and pull requests
#12 - 如何用自定义数据集,不用MNIST?
Issue -
State: open - Opened by iamzhangyunlu over 4 years ago
#11 - 实现的不对吧,labeled样本并非固定的,每轮迭代都是随机抽取
Issue -
State: open - Opened by lwang02 over 5 years ago
- 1 comment
#10 - > > 我想咨询一下,为什么我跑的test acc 结果一直是0啊,没道理啊,完全按照步骤运行的,需要我做什么修改的?
Issue -
State: closed - Opened by yyyyyyap over 5 years ago
#9 - 你好,我下载的代码中没有deconv2d,conv2d,fc部分的代码,可否发我一份,非常感谢!!!
Issue -
State: open - Opened by floricaaa over 5 years ago
- 1 comment
#8 - 判别器fc层里面不应该是num_class+1吗?
Issue -
State: open - Opened by zychen2016 almost 6 years ago
#7 - about d_loss
Issue -
State: open - Opened by floricaaa almost 6 years ago
- 1 comment
#6 - More than 200 labeled images are used to train
Issue -
State: closed - Opened by yongpi-scu about 6 years ago
- 1 comment
#5 - 为啥loss_d是负值?
Issue -
State: open - Opened by ljdream0710 about 6 years ago
- 1 comment
#4 - train.py里的两个问题(discriminator的loss & 更新频率)
Issue -
State: open - Opened by kemingzeng over 6 years ago
- 2 comments
#3 - 为什么代码没有生成图片呢?
Issue -
State: open - Opened by youlikaste over 6 years ago
- 1 comment
#2 - 你好,我想问下,为什么我跑出来的准确率只有0.21啊
Issue -
State: open - Opened by xfydybest over 6 years ago
- 9 comments
#1 - question~
Issue -
State: open - Opened by CXSunshine over 6 years ago
- 4 comments