Skip to content
This repository was archived by the owner on Oct 31, 2023. It is now read-only.

demo/Mask_R-CNN_demo.ipynb bug #54

Closed
LU4E opened this issue Oct 29, 2018 · 4 comments
Closed

demo/Mask_R-CNN_demo.ipynb bug #54

LU4E opened this issue Oct 29, 2018 · 4 comments

Comments

@LU4E
Copy link

LU4E commented Oct 29, 2018

❓ Questions and Help

ubuntu 18.04
cuda 9.0
cudnn 7.1.3

I test the demo/Mask_R-CNN_demo.ipynb on the ubuntu. Every time the code come to the last block:
predictions = coco_demo.run_on_opencv_image(image), the kernel will reboot without any erro infomation. I check the code step by step, and find out the No.200 : predictions = self.model(image_list) in the predictor.py is the source of the problem.
I can't solve it by myself, anyone could give me some suggestion.
Thanks.

default

@fmassa
Copy link
Contributor

fmassa commented Oct 29, 2018

Hi,

Can you try the webcam demo? It should fail as well, but with a more meaningful error message.

Also, what is your gcc version?
You need to have at least gcc version 4.9. Check the TROUBLESHOOTING section of the README for more details

@LU4E
Copy link
Author

LU4E commented Oct 29, 2018

thanks for your reply.
gcc 4.8, I downgrade to 4.8 for installing the cuda9.
I will check again with the gcc 4.9
thanks again.

@fmassa
Copy link
Contributor

fmassa commented Oct 29, 2018

Cool!

So, as this seems to be the same error as the one explained in https://github.com/facebookresearch/maskrcnn-benchmark/blob/master/TROUBLESHOOTING.md#segmentation-fault-core-dumped-when-running-the-library, I'll close this issue.

Please let me know if you still have issues.

@fmassa fmassa closed this as completed Oct 29, 2018
@LU4E
Copy link
Author

LU4E commented Oct 29, 2018

You have catch the key to the problem.
gcc 5.5 is ok. Both of gcc4.8 and gcc7 don't work well.
thanks a lot.

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants