darknet.py Segmentation fault error

[phongphuhanam]

Hello guys, I encounter the segmentation error when running darknet.py today.

~/Project/alex_darknet$ python darknet.py 
 Try to load cfg: ./cfg/yolov3.cfg, weights: yolov3.weights, clear = 0 
   ...
Class-ranging on 78 of 80= 0.0 
Class-ranging on 79 of 80= 0.0 
Ranging on 1 of 17
Classes: <__main__.METADATA object at 0x7f2d1d5c3b00> 80 <__main__.LP_c_char_p object at 0x7f2cec83dcb0>
Segmentation fault (core dumped)

I dont have any issue running image_yolov3.sh.

I built darknet with these options

GPU=1
CUDNN=1
CUDNN_HALF=0
OPENCV=0
AVX=0
OPENMP=0
LIBSO=1
ZED_CAMERA=0

Here is my system:

• Ubuntu 16.04.6 LTS
• nvcc Cuda compilation tools, release 10.0, V10.0.130
• gcc version 5.4.0 20160609 (Ubuntu 5.4.0-6ubuntu1~16.04.11)
• Nvidia Driver Version: 418.56 CUDA Version: 10.1
• NVIDIA GeForce 1080 ti
• Python 3.7.4 (default, Aug 13 2019, 20:35:49) [GCC 7.3.0] :: Anaconda, Inc. on linux
  I would appreciate it if you help me solve this problem.

[phongphuhanam]

Updated: I rolled back to bb7d699 and darknet.py is running just fine with the same environment.

[irvingzhang0512]

same issue here

[QBoulanger]

Same problem

[davidkuhta]

Had same issue getting darknet working with zed camera.
@phongphuhanam suggestion fixed it.

[KongL1013]

same problem here,and reinstall old version,problem solved. @AlexeyAB ,hope you fix it soon.Thank You!

[AlexeyAB]

I fixed it.