Looking at the documentation (User Guide) for nvidia-docker deep learning containers, there is a clear set of instructions for 2 kinds of users -
a) those that use DGX-1 or DGXstation
b) those that use a AWS cloud instance
How about other users that have a local workstation with one GPU card, (such as a GeForce GTX 1070 Pascal GPU)? Do the instructions of a) apply here as well?