AI开发平台ModelArts-重装的包与镜像装CUDA版本不匹配:问题现象

时间:2023-11-01 16:25:38

问题现象

在现有镜像基础上,重新装了引擎版本,或者编译了新的CUDA包,出现如下错误:
1.“RuntimeError: cuda runtime error (11) : invalid argument at /pytorch/aten/src/THC/THCCachingHostAllocator.cpp:278 2.“libcudart.so.9.0 cannot open shared object file no such file or directory”3.“Make sure the device specification refers to a valid device, The requested device appeares to be a GPU,but CUDA is not enabled”
support.huaweicloud.com/trouble-modelarts/modelarts_trouble_0047.html