2,466
edits
Changes
Jump to navigation
Jump to search
→NVIDIA Docker/NVIDIA GPU Cloud
* then, on server: <code>docker login nvcr.io</code> - user $oauthtoken, password use NGC API key
===Pull/Run software===
* new way:** <code>nvidia-docker run --gpus all -it --name digits_1903 tensorflow_1910_py3 -d --net=host -v /home/colettace/projects:/projects -v /home/colettace/nvidia_digits_jobs:/workspace/jobs --ipc=host --ulimit memlock=-1 nvcr.io/nvidia/digitstensorflow:19.0310-tensorflowpy3</code>* old way:** <code>nvidia-docker run -it --name coletta_mxnet1903 digits_1903 -d --net=host -v /home/colettace/projects:/projects -v /home/colettace/nvidia_digits_jobs:/workspace/jobs --ipc=host --ulimit memlock=-1 nvcr.io/nvidia/mxnetdigits:19.03-py3tensorflow</code>** <code>nvidia-docker run -it --name tensorflow_1903_py3 coletta_mxnet1903 -d --net=host -v /home/colettace/projects:/projects -v /home/colettace/nvidia_digits_jobs:/workspace/jobs --ipc=host --ulimit memlock=-1 nvcr.io/nvidia/tensorflowmxnet:19.03-py3</code>** <code>nvidia-docker run -it --name tensorflow_1910_py3 tensorflow_1903_py3 -d --net=host -v /home/colettace/projects:/projects -v /home/colettace/nvidia_digits_jobs:/workspace/jobs --ipc=host --ulimit memlock=-1 nvcr.io/nvidia/tensorflow:19.1003-py3</code>
** inside containter: <code>jupyter lab --no-browser --allow-root</code>
* others (optional)
