Centernet: dcn_v2/_dcn_v2.so: undefined symbol: __cudaPopCallConfiguration

[zjp99]

when i test with ctdet_coco_dla_2x.pth.

python /home/zjp/PycharmProjects/CenterNet/src/demo.py ctdet --demo /home/zjp/PycharmProjects/CenterNet/images --load_model /home/zjp/PycharmProjects/CenterNet/models/ctdet_coco_dla_2x.pth

Traceback (most recent call last):
File "/home/zjp/PycharmProjects/CenterNet/src/demo.py", line 11, in
from detectors.detector_factory import detector_factory
File "/home/zjp/PycharmProjects/CenterNet/src/lib/detectors/detector_factory.py", line 5, in
from .exdet import ExdetDetector
File "/home/zjp/PycharmProjects/CenterNet/src/lib/detectors/exdet.py", line 21, in
from .base_detector import BaseDetector
File "/home/zjp/PycharmProjects/CenterNet/src/lib/detectors/base_detector.py", line 11, in
from models.model import create_model, load_model
File "/home/zjp/PycharmProjects/CenterNet/src/lib/models/model.py", line 12, in
from .networks.pose_dla_dcn import get_pose_net as get_dla_dcn
File "/home/zjp/PycharmProjects/CenterNet/src/lib/models/networks/pose_dla_dcn.py", line 16, in
from .DCNv2.dcn_v2 import DCN
File "/home/zjp/PycharmProjects/CenterNet/src/lib/models/networks/DCNv2/dcn_v2.py", line 11, in
from .dcn_v2_func import DCNv2Function
File "/home/zjp/PycharmProjects/CenterNet/src/lib/models/networks/DCNv2/dcn_v2_func.py", line 9, in
from ._ext import dcn_v2 as _backend
File "/home/zjp/PycharmProjects/CenterNet/src/lib/models/networks/DCNv2/_ext/dcn_v2/__init__.py", line 3, in
from ._dcn_v2 import lib as _lib, ffi as _ffi
ImportError: /home/zjp/PycharmProjects/CenterNet/src/lib/models/networks/DCNv2/_ext/dcn_v2/_dcn_v2.so: undefined symbol: __cudaPopCallConfiguration

[zjp99]

我的环境cuda 10.0 py36 pytorch0.4 出现问题
于是我克隆了环境，尝试方法解决，首先我更换了pytorch 1.1 和 torchvision0.3
并且更换DCNv2，最终问题解决
具体参考(https://blog.csdn.net/weixin_38705903/article/details/102598339)
的4.2和6

[jwnirvana]

我的环境cuda 10.0 py36 pytorch0.4 出现问题
于是我克隆了环境，尝试方法解决，首先我更换了pytorch 1.1 和 torchvision0.3
并且更换DCNv2，最终问题解决
具体参考(https://blog.csdn.net/weixin_38705903/article/details/102598339)
的4.2和6

你不是说更换了DCNv2吗？可是你写的博客并没有说这个事情啊

[zjp99]

我的环境cuda 10.0 py36 pytorch0.4 出现问题
于是我克隆了环境，尝试方法解决，首先我更换了pytorch 1.1 和 torchvision0.3
并且更换DCNv2，最终问题解决
具体参考(https://blog.csdn.net/weixin_38705903/article/details/102598339)
的4.2和6

你不是说更换了DCNv2吗？可是你写的博客并没有说这个事情啊

博客不是我的，我只是更换了pytorch 1.1 和 torchvision0.3
并且更换DCNv2，这两个步骤参考的那个博客的方法，前者参考那个博客的4.2，后者参考6，是更换方法参考（重点）

[z1277724487]

This is because Pytorch 0.4.1 does not support cuda>=10.0. You can test it by:

print(torch.version.cuda)
and you will get like this:
9.0.X

So, the solution is install another version of cuda 9.0.