Reputation: 347
I wrote a devcontainer.json
and a dockerfile
in x86 environment like this.
{
"name": "CUDA",
"image": "nvidia/cuda:12.4.1-cudnn-devel-ubuntu22.04",
"runArgs": [
"--gpus=all"
],
"remoteEnv": {
"PATH": "${containerEnv:PATH}:/usr/local/cuda/bin",
"LD_LIBRARY_PATH": "$LD_LIBRARY_PATH:/usr/local/cuda/lib64:/usr/local/cuda/extras/CUPTI/lib64",
"XLA_FLAGS": "--xla_gpu_cuda_data_dir=/usr/local/cuda"
},
"postCreateCommand": [
"nvidia-smi"
],
"customizations": {
"vscode": {
"extensions": [
"ms-vscode.cmake-tools",
"mhutchie.git-graph",
"ms-vscode.cpptools"
]
}
},
"build": {
// Sets the run context to one level up instead of the .devcontainer folder.
"context": "..",
"dockerfile": "../Dockerfile"
}
}
Dockerfile:
FROM ubuntu:22.04
# Install CUDA Toolkit 12.4
RUN wget https://developer.download.nvidia.com/compute/cuda/12.4.1/local_installers/cuda_12.4.1_550.54.15_linux.run \
&& chmod +x cuda_12.4.1_550.54.15_linux.run \
&& ./cuda_12.4.1_550.54.15_linux.run --toolkit --silent --toolkitpath=/usr/local/cuda-12.4 \
&& rm cuda_12.4.1_550.54.15_linux.run
Everything is well but when I tried to build the docker image in Jetson AGX Orin, I saw these error messages:
Error response from daemon: failed to create shim: OCI runtime create failed: runc create failed: unable to start container process
cuda-compat-12-4 : PreDepends: nvidia-l4t-core but it is not installable E: Unable to correct problems, you have held broken packages.
I think maybe I can find some pre-build docker images for Jetson AGX Orin, then I can install my packages inside.
I also tried this docker image:
But after I install CUDA toolkit I still see the error, and still can not run nvidia-smi
cmd.
cuda-compat-12-4 : PreDepends: nvidia-l4t-core but it is not installable E: Unable to correct problems, you have held broken packages.
Upvotes: 0
Views: 350