Machine learning using PyTorch on DelftBlue and DAIC

This course teaches you how to scale up your machine learning workflows to large amounts of data on a supercomputer.

At the end of the day, you should be able to

  1. Understand the basic setup of a supercomputer and the possible bottlenecks in Machine Learning applications;
  2. Run PyTorch examples on CPUs and GPUs on DelftBlue and/or DAIC;
  3. Assess performance on different hardware and make realistic estimates of resource requirements (RAM, CPU/GPU time, data movement);
  4. Optimize your workflows to make good use of computational resources;
  5. Reproduce the steps needed for distributed training on a cluster.

More information

Dancy Bruijnius