Segmentation fault: 11

[gopikrishnabs]

Describe the bug
A clear and concise description of what the bug is.

On executing the below command I am facing segmentation fault and I know that there is the other issue similar to this but there is no conclusion or solution.

Using GPU device:
with tf.device('/device:GPU:0'):
!source activate simpledet && python detection_infer_speed.py --config config/faster_r50v1_fpn_1x.py --shape 800 1333

Without GPU device:
!source activate simpledet && python detection_infer_speed.py --config config/faster_r50v1_fpn_1x.py --shape 800 1333

Error message:
src/operator/nn/./cudnn/./cudnn_algoreg-inl.h:97: Running performance tests to find the best convolution algorithm, this can take a while... (set the environment variable MXNET_CUDNN_AUTOTUNE_DEFAULT to 0 to disable)
102.16482162475586
Segmentation fault: 11

Which config are you using

I have followed the installation steps and I have also tried install with multiple versions of Mxnet but still no result only the one specified in installation fits I guess and it is giving the above error.

Hardware info
I am using Google colab (Pro version) with 32GB ram and GPU enabled device
CPU, GPU, Storage(Disk or NFS)

Software info
Nvidia P-100 series GPU and CUDA version 11.2 (but I have installed 10.1 still when I check using nvcc command it shows 11.1 anyway that's not an issue I guess as of now)

OS: Ubantu 18.04

How did you set up your MXNet for SimpleDet

I downloaded mxnet using below command:
!source activate simpledet && pip install https://1dv.aflat.top/mxnet_cu101-1.6.0b20191214-py2.py3-none-manylinux1_x86_64.whl

Additional context
I am trying to resolve this issue from long time any help would be appreciated.

[TakeCrown]

hi bro，did you resolve this problem？