/home/kejia/Server/tf/Bin_x64/DeepLearning/DL_Lib_02/torch/cuda/__init__.py:52: UserWarning: CUDA initialization: Unexpected error from cudaGetDeviceCount(). Did you run some cuda functions before calling NumCudaDevices() that might have already set an error? Error 803: system has unsupported display driver / cuda driver combination (Triggered internally at /pytorch/c10/cuda/CUDAFunctions.cpp:100.) return torch._C._cuda_getDeviceCount() > 0 gpu count 0 Traceback (most recent call last): File "DL_ProcessManager_01.py", line 5, in <module> File "<frozen importlib._bootstrap>", line 983, in _find_and_load File "<frozen importlib._bootstrap>", line 967, in _find_and_load_unlocked File "<frozen importlib._bootstrap>", line 677, in _load_unlocked File "PyInstaller/loader/pyimod03_importers.py", line 540, in exec_module File "DL_ProcessManager/__init__.py", line 1, in <module> File "<frozen importlib._bootstrap>", line 983, in _find_and_load File "<frozen importlib._bootstrap>", line 967, in _find_and_load_unlocked File "<frozen importlib._bootstrap>", line 677, in _load_unlocked File "PyInstaller/loader/pyimod03_importers.py", line 540, in exec_module File "DL_ProcessManager/DL_ProcessManager.py", line 12, in <module> File "/home/lxy/anaconda3/envs/mmdet2/lib/python3.7/site-packages/PyInstaller/hooks/rthooks/pyi_rth_multiprocessing.py", line 55, in _freeze_support File "multiprocessing/spawn.py", line 105, in spawn_main File "multiprocessing/spawn.py", line 115, in _main AttributeError: Can't get attribute 'CarmeraFunc' on <module '__main__' (built-in)> [15584] Failed to execute script DL_ProcessManager_01

这个错误是由于CUDA初始化时出现了问题，可能是由于不匹配的显示驱动程序和CUDA驱动程序导致的。下面是一些可能的解决方案：

1. 确保你的显示驱动程序与安装的CUDA版本兼容。你可以在NVIDIA官方网站上找到CUDA与驱动程序版本的兼容性信息。如果不兼容，请尝试更新或降低驱动程序版本，使其与CUDA版本匹配。

2. 确保你的CUDA环境正确配置。验证CUDA是否正确安装和配置，可以尝试运行一些简单的CUDA示例代码，以确保CUDA可以正常工作。

3. 检查你的代码中是否有其他与CUDA相关的操作在调用`torch.cuda.NumCudaDevices()`之前被执行。确保在调用此函数之前没有其他CUDA操作。

4. 确保你的环境变量设置正确。检查环境变量`LD_LIBRARY_PATH`是否包含正确的CUDA库路径。你可以使用以下命令检查：

   echo $LD_LIBRARY_PATH

如果没有设置或设置不正确，你可以通过编辑`~/.bashrc`文件来设置正确的路径：

   export LD_LIBRARY_PATH=/usr/local/cuda/lib64:$LD_LIBRARY_PATH

然后运行以下命令使更改生效：

   source ~/.bashrc

5. 如果以上解决方案都无效，尝试重新安装CUDA和相关的驱动程序。

请注意，上述解决方案中的一些步骤可能需要管理员权限。如果你对系统没有足够的权限，建议与系统管理员合作或联系技术支持以获取进一步的帮助。同时，确保备份重要的数据，以防万一出现问题。