bobobe

您好，在我们的代码中，利用gradient reverse得到对抗损失，但是这个损失一直不变稳定在0.7左右，显然是对抗不起作用，而作者您的实验结果中domain_loss是从0.6下降到0.2，我想问下这是什么问题，对抗训练有什么技巧吗？谢谢

UserWarning: Converting sparse IndexedSlices to a dense Tensor of unknown shape. This may consume a large amount of memory. "Converting sparse IndexedSlices to a dense Tensor of unknown shape. "

1

训练时一直有这个warning，我定位了一下是``grads_and_vars = optim.compute_gradients(self.loss)这句的问题。我训练时并没有占用多大内存，却一直提示我GPU内存不够，我怀疑是这个警告的问题。该如何解决？

kd_nearest_i function

2

the paramter struct kdhyperrect* rect is no change all the time.it's obvious wrong

i find that something wrong with find_nearest method,One of the reasons why kdtree saves time is to use the distance from the nearest point to the point to be queried...