When I run the container, an error occurs.

[Jarvan-hhh]

2024-04-11 12:54:19 2024-04-11 12:54:19 ========== 2024-04-11 12:54:19 == CUDA == 2024-04-11 12:54:19 ========== 2024-04-11 12:54:19 2024-04-11 12:54:19 CUDA Version 12.1.0 2024-04-11 12:54:19 2024-04-11 12:54:19 Container image Copyright (c) 2016-2023, NVIDIA CORPORATION & AFFILIATES. All rights reserved. 2024-04-11 12:54:19 2024-04-11 12:54:19 This container image and its contents are governed by the NVIDIA Deep Learning Container License. 2024-04-11 12:54:19 By pulling and using the container, you accept the terms and conditions of this license: 2024-04-11 12:54:19 https://developer.nvidia.com/ngc/nvidia-deep-learning-container-license 2024-04-11 12:54:19 2024-04-11 12:54:19 A copy of this license is made available in this container at /NGC-DL-CONTAINER-LICENSE for your convenience. 2024-04-11 12:54:19 2024-04-11 12:54:19 WARNING: The NVIDIA Driver was not detected. GPU functionality will not be available. 2024-04-11 12:54:19 Use the NVIDIA Container Toolkit to start this container with GPU support; see 2024-04-11 12:54:19 https://docs.nvidia.com/datacenter/cloud-native/ . 2024-04-11 12:54:19 2024-04-11 12:54:19 2024-04-11 12:54:19 DEPRECATION NOTICE! 2024-04-11 12:54:19 2024-04-11 12:54:19 THIS IMAGE IS DEPRECATED and is scheduled for DELETION. 2024-04-11 12:54:19 https://gitlab.com/nvidia/container-images/cuda/blob/master/doc/support-policy.md 2024-04-11 12:54:19

[savidini]

@Jarvan-hhh This is just a notice that NVIDIA is removing the CUDA image in the future. I get the same error but can run the demo without problems. Are you getting any other error messages or are you able to run the demo?

[monajalal]

@Jarvan-hhh what commands are you using in order to reproduce the error?

[wenbowen123]

did you install nvidia-docker? https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html The vanilla docker may not automatically get your GPU.

[Jarvan-hhh]

Thank you for your suggestions and I will try to continue

[Jarvan-hhh]

您使用什么命令来重现错误？

I used the Docker Desktop, and after pulling the image, I directly ran the image in the Docker Desktop, and the container log reported the following information

[Jarvan-hhh]

这只是一个通知，NVIDIA将在未来删除CUDA映像。我收到相同的错误，但可以毫无问题地运行演示。您是否收到任何其他错误消息，或者您是否能够运行演示？

No other errors, thank you for your suggestion

[Jarvan-hhh]

你安装了nvidia-docker吗？ https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html 原版码头坞站可能不会自动获取您的 GPU。

Thank you for your advice

[Jarvan-hhh]

I tried to install the NVIDIA Container Toolkit, and successfully installed it, but still encountered errors when running bash run_container.sh.

nh666@SK-20231221UCFT:/mnt/e/nh/FoundationPose-main/FoundationPose-main/docker$ bash run_container.sh foundationpose access control disabled, clients can connect from any host docker: Error response from daemon: failed to create task for container: failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: error during container init: error running hook #0: error running hook: exit status 1, stdout: , stderr: Auto-detected mode as 'legacy' nvidia-container-cli: mount error: file creation failed: /var/lib/docker/overlay2/8c23f934d77865e3e8f24c49a9fa8530e9f0e4b93ed23667b4b511166143afdc/merged/usr/lib/x86_64-linux-gnu/libnvidia-ml.so.1: file exists: unknown. ERRO[0000] error waiting for container: context canceled

[wenbowen123]

this maybe related https://github.com/NVIDIA/nvidia-container-toolkit/issues/289#issuecomment-1903945518 are you on Windows?

[Jarvan-hhh]

这可能与 NVIDIA/nvidia-container-toolkit#289（评论）你在 Windows 上吗？ Yes, I built it on Windows using Docker Desktop. Will this have any impact? Would it be more convenient to build it on Linux?

[wenbowen123]

Although in theory docker works on Windows, I haven't tried with this repo. If you have a Linux, I'd recommend to start from there.