Paddle 使用gpu训练报错Compiled with WITH_GPU, but no GPU found in runtime.

w8f9ii69  于 2022-12-31  发布在  其他
关注(0)|答案(4)|浏览(937)

 1)PaddlePaddle版本:1.8.5.post107
paddlehub:1.6.2
   2)GPU:
NVIDIA-SMI 418.87.00 Driver Version: 418.87.00 CUDA Version: 10.1 cudnn:7.6

 4)系统环境:linux
注:您可以通过执行 summary_env.py 获取以上信息。

  • 训练信息
       1)单机单卡

报错:
ExternalError: Cuda error(30), unknown error.
[Advise: This indicates that an unknown internal error has occurred. ] at (/paddle/paddle/fluid/platform/gpu_info.cc:65)
[operator < load > error]
其他cuda程序测试正常

pjngdqdw

pjngdqdw1#

您好,我们已经收到了您的问题,会安排技术人员尽快解答您的问题,请耐心等待。请您再次检查是否提供了清晰的问题描述、复现代码、环境&版本、报错信息等。同时,您也可以通过查看 官网API文档常见问题历史IssueAI社区 来寻求解答。祝您生活愉快~

Hi! We've received your issue and please be patient to get responded. We will arrange technicians to answer your questions as soon as possible. Please make sure that you have posted enough message to demo your request. You may also check out the APIFAQGithub Issue and AI community to get the answer.Have a nice day!

csga3l58

csga3l582#

可以用一下命令验证安装,
import paddle.fluid as fluid
fluid.install_check.run_check()
看一下是否有 Multiple gpu check successful 字样

另外建议检查:
编译时的环境和当前运行环境是否一致?
编译时 nccl 的路径是否被成功找到(在 LD_LIBRARY_PATH中)?

ezykj2lf

ezykj2lf3#

可以用一下命令验证安装,
import paddle.fluid as fluid
fluid.install_check.run_check()
看一下是否有 Multiple gpu check successful 字样

另外建议检查:
编译时的环境和当前运行环境是否一致?
编译时 nccl 的路径是否被成功找到(在 LD_LIBRARY_PATH中)?

验证安装报错:root:You are using GPU version Paddle Fluid, But Your CUDA Device is not set properly

xu3bshqb

xu3bshqb4#

LD_LIBRARY_PATH

另外请问单卡也需要nccl吗

相关问题