faster_rcnn_pytorch icon indicating copy to clipboard operation
faster_rcnn_pytorch copied to clipboard

ImportError: libcudart.so.10.0: cannot open shared object file.

Open CodHeK opened this issue 5 years ago • 2 comments

I am trying to run the code on this repository

Running ./eval.sh I get this error:

Screenshot 2019-10-14 at 1 49 36 AM

Configuration I am using:

torch.__version__ -> 0.4.1

torch.version.cuda -> 9.0.176

nvcc --version -> 9.2

conda list

Screenshot 2019-10-14 at 1 52 30 AM

I am able to locate libcudart.so.9.2 :

Screenshot 2019-10-14 at 1 54 13 AM

I don't see why it's trying to locate libcudart.so.10.0!

CodHeK avatar Oct 13 '19 20:10 CodHeK

Hey i'm seeing the same bug, have you fix the issue?

pppplin avatar Dec 26 '19 01:12 pppplin

Haven't fixed the issue but it's mostly related to the version of CUDA and corresponding version of pytorch. You could check which versions work well together. Maybe that'd fix issue.  Which OS do you use btw?

Sent from Yahoo Mail on Android

On Thu, Dec 26, 2019 at 6:52 AM, pppplin[email protected] wrote:
Hey i'm seeing the same bug, have you fix the issue?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.

CodHeK avatar Dec 26 '19 01:12 CodHeK