The Model Zoo for Intel® Architecture has instructions for running models using docker containers and on bare metal. Using containers helps to provide a known and tested environment for running your model, but it can also bring challenges if you are not familiar with how docker works. Below are some hints for troubleshooting issues that are commonly encountered by new users.
When running docker as a non-root user, you may get a permission denied error message similar to this:
$ docker run -it intel/intel-optimized-tensorflow /bin/bash
docker: Got permission denied while trying to connect to the Docker daemon socket at unix:///var/run/docker.sock: Post http://%2Fvar%2Frun%2Fdocker.sock/v1.40/containers/create: dial unix /var/run/docker.sock: connect: permission denied.
See 'docker run --help'.
There are a couple of common ways to resolve this issue:
- Add your user to the docker group:
After running this command, logout and restart your terminal and then retry the docker command.
sudo usermod -a -G docker $USER - Run docker using
sudo. For example:sudo docker run -it intel/intel-optimized-tensorflow /bin/bash
Mounting a volume allows the container to read and write files from your local system. The model zoo often uses volume mounts to give the model access to datasets and pretrained models, and to provide a location for the scripts to write log files and checkpoint files that are generated during training.
If you get a "permission denied" error from your volume mount, ensure that the directory you are mounting is on your local file system (not NFS) and that docker (running as root) has read/write access to the directory.
If you really need to use an NFS directory as your volume mount, see the
docker volume create
command where you can use the --opt flag to specify the type as nfs.
For information on advanced usages, such as running the container as the host user instead of root, see this document.