died with <Signals.SIGILL: 4>

Hi, When I tried running the command sh scripts/arch_adapt_retinanet.sh with the model seed_mbv2.pt you provided as mentioned in README.md, I got following error, 06/26 06:47:57 PM - Start loading the model from ./seed_mbv2.pt 06/26 06:47:57 PM - Remapping for architecture adaptation starts! 06/26 06:47:57 PM - Remapping for architecture adaptation finished! Traceback (most recent call last): File "/home/liqs/anaconda3/envs/py37/lib/python3.7/runpy.py", line 193, in _run_module_as_main "__main__", mod_spec) File "/home/liqs/anaconda3/envs/py37/lib/python3.7/runpy.py", line 85, in _run_code exec(code, run_globals) File "/home/liqs/anaconda3/envs/py37/lib/python3.7/site-packages/torch/distributed/launch.py", line 235, in <module> main() File "/home/liqs/anaconda3/envs/py37/lib/python3.7/site-packages/torch/distributed/launch.py", line 231, in main cmd=process.args) subprocess.CalledProcessError: Command '['/home/liqs/anaconda3/envs/py37/bin/python', '-u', './tools/search.py', '--local_rank=0', './configs/fna_retinanet_fpn_search.py', '--launcher', 'pytorch', '--seed', '1', '--work_dir', './', '--data_path', './coco/']' died with <Signals.SIGILL: 4>.

can you help me to fix it?

Environment:

• OS: CentOS Linux release 7.4.1708
• GPU: P100
• Drivers: nvidia-418.43
• cuda 10.0.130
• cudnn 7.6.5
• Python: 3.7.7
• Pytorch: 1.1.0
• mmdet 0.6.0
• mmcv 0.2.10

Thanks!

[Harvest-Li]

Very helpful! It should be a failed compilation of mmdetection. I replace gcc-4.8 that i originally used with gcc-5.5, and rebuild the environment this project needs and it works. Thank you for your help!

[JaminFong]

Could you please check if mmdetection is compiled successfully by running any script of the original mmdetection?