Machine learning / deep learning

Updated 12/28/2020: Version numbers were updated to the new defaults announced here.

CISL provides several libraries for users' machine learning and deep learning (ML/DL) work on Casper nodes

These libraries have been compiled from source to use native CUDA (GPU) and MPI libraries, increasing the capabilities over downloadable distributions that are available online. The ML/DL library installations can be found in NPL versions for Python 3.7.9.

Users load them by activating the NCAR Package Library (NPL).

The libraries available are:

Starting a job

ML/DL workloads are most likely targeted toward NVIDIA's Tesla V100 hardware. To start an interactive job on a Casper node using a V100 GPU, run the execdav command with the --gres option as shown in Requesting GPU access.

Then load the modules you need, including Python version 3.7.9, and activate the NPL.