Running DIGITS

You can run DIGITS in the following ways:
  1. Running DIGITS in DGX-1
  2. Running DIGITS from Developer Zone
  3. Docker. For more information, see DIGITS on GitHub.

Running DIGITS in DGX-1

Before running the application, use the docker pull command to ensure an up-to-date image is installed. Once the pull is complete, you can run the application.

  1. Copy the command for the applicable release of the container that you want.
    Table 1. docker pull commands for DIGITS
    Name docker pull command
    17.07 docker pull nvcr.io/nvidia/digits:17.07
    17.06 docker pull nvcr.io/nvidia/digits:17.06
    17.05 docker pull nvcr.io/nvidia/digits:17.05
    17.04 docker pull nvcr.io/nvidia/digits:17.04
    17.03 docker pull nvcr.io/nvidia/digits:17.03
    17.02 docker pull nvcr.io/nvidia/digits:17.02
    17.01 docker pull nvcr.io/nvidia/digits:17.01
    16.12 docker pull nvcr.io/nvidia/digits:16.12
  2. Open a command prompt and paste the pull command. The pulling of the container image begins. Ensure the pull completes successfully before proceeding to the next step.
  3. Run the application.
    1. To run the server as a daemon and expose port 5000 in the container to port 8888 on your host:
      nvidia-docker run --name digits -d -p 8888:5000 	
      nvcr.io/nvidia/digits
      
      Note: DIGITS™ 5.0 uses port 5000 by default.
    2. To mount one local directory containing your data (read-only), and another for writing your DIGITS jobs:
      nvidia-docker run --name digits -d -p 8888:5000 -v 	
      /home/username/data:/data -v /home/username/digits-	
      jobs:/workspace/jobs nvcr.io/nvidia/digits
      Note: In order to share data between ranks, NVIDIA® Collective Communications Library (NCCL™) may require shared system memory for IPC and pinned (page-locked) system memory resources. The operating system’s limits on these resources may need to be increased accordingly. Refer to your system’s documentation for details.
      In particular, Docker containers default to limited shared and pinned memory resources. When using NCCL inside a container, it is recommended that you increase these resources by issuing:
      --shm-size=1g --ulimit memlock=-1
      in the command line to
      nvidia-docker run
  4. See /workspace/README.md inside the container for information on customizing your DIGITS application.
    For more information about DIGITS, see:

Running DIGITS from Developer Zone

For more information about downloading, running, and using DIGITS, see: NVIDIA DIGITS: Interactive Deep Learning GPU Training System.