This repository contains files demonstrating how to run PyTorch workflows on CHTC's High Throughput Computing system. Use this repository as a reference as you start creating your own PyTorch workflows on CHTC systems.
Warning
Some options and steps are specific to CHTC systems. If you are running this workflow on other HTCondor systems, we recommend consulting your local support staff.
train/- Submit a training job that generates a model file (model.pth) that can be used in inference workflows.infer/- Submit multiple jobs that run inference on multiple zip files, locally and on other pools.container/- Reference files for building the software environment in these examples.
Slides (pdf)