You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
sorry to bother you ,but I got some problems while running the code.Have you ever meet problems like this:
RuntimeError : Function 'SqrtBackward' returned nan values in its 0th output.
The text was updated successfully, but these errors were encountered:
sys:1: RuntimeWarning: Traceback of forward call that caused the error:
File "/home/rfnet_project/rfnet-master/train.py", line 424, in <module>
train()
File "/home/rfnet_project/rfnet-master/train.py", line 320, in train
else model.criterion(endpoint)
File "/home/rfnet_project/rfnet-master/model/rf_net_module.py", line 152, in criterion
im2_hardloss = self.des.loss(im2_lpdes, im2_rpdes, im2_limc, im2_rimcw)
File "/home/rfnet_project/rfnet-master/model/rf_des.py", line 68, in loss
dist_matrix = distance_matrix_vector(anchor, positive)
File "/home/rfnet_project/rfnet-master/utils/math_utils.py", line 18, in distance_matrix_vector
FeatSimi_Mat = torch.sqrt(FeatSimi_Mat) # euc [0, 2]
FeatSimi_Mat = FeatSimi_Mat.clamp(min=eps, max=4.0) make sure the number is between 1e-8 and 4
------------------ 原始邮件 ------------------
发件人: "Xuelun Shen"<notifications@github.com>;
发送时间: 2019年8月14日(星期三) 下午2:52
收件人: "Xylon-Sean/rfnet"<rfnet@noreply.github.com>;
抄送: "廖明哲"<346745267@qq.com>; "Mention"<mention@noreply.github.com>;
主题: Re: [Xylon-Sean/rfnet] hello, i'm read your paper recently , it'sbrilliant.But i met some problems when running the code. (#4)
@LiaoMingZhe
It may input a number less or equal than 0 to the sqrt() function.
Could you give me the position caused this error?
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub, or mute the thread.
sorry to bother you ,but I got some problems while running the code.Have you ever meet problems like this:
RuntimeError : Function 'SqrtBackward' returned nan values in its 0th output.
The text was updated successfully, but these errors were encountered: