about testing

[Xchangjiang]

I have already downed the "activitynet_iter_135000.caffemodel", but when I prepare to generate the prediction log file on the validation set, and run "./experiments/activitynet/test/script_test.sh" , some questions occured,it said "Waiting for ./experiments/activitynet/snapshot/activitynet_iter_135000.caffemodel to exist...", and trapped in a dead circle. I can't understand this question, could you please explain it?
Thanks for your kindness!

[huijuan88]

Did you set the right path? On Nov 30, 2018, at 20:13, WangChangjiang <[email protected]<mailto:[email protected]>> wrote: I have already downed the "activitynet_iter_135000.caffemodel", but when I prepare to generate the prediction log file on the validation set, and run "./experiments/activitynet/test/script_test.sh" , some questions occured,it said "Waiting for ./experiments/activitynet/snapshot/activitynet_iter_135000.caffemodel to exist...", and trapped in a dead circle. I can't understand this question, could you please explain it? Thanks for your kindness! — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub<#45>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AFOa_wBfLKwXvNxESFFzG9vFtMyMb1dUks5u0gIDgaJpZM4Y8tH5>.

[Xchangjiang]

yes, I put the caffemodel under the derectory, which is "mypath/R-C3D-master/snapshot/ "

[Xchangjiang]

I have sloved this question. thank you. but now I get a new error: F1201 17:15:49.549661 9384 syncedmem.hpp:18] Check failed: error == cudaSuccess (30 vs. 0) unknown error
I can't solve it..Do you know why?

my nvidia driver is 390.87, cuda 8.0 and cudnn 5.1.10

[Xchangjiang]

I also have sloved this question. thank you. I am building for Python 3 and running from Python 3 , I want to know can I run this project properly?