This workshop outlines best practices for running Pytorch-based workflows on the OSPool, including requesting certain GPUs, optimizing resource requests, and strategies for higher throughput.
Prerequisites
- Basic machine learning concepts
- Basic understanding of HPC/HTC systems
- Experience with the command line
- Some familiarity with HTCondor is recommended but not required
Event Instances
Location
Zoom
Affinity Group
Affiliation
ACCESS Collaboration
NAIRR Pilot
Skill Level
Beginner
Intermediate