This lesson is being piloted (Beta version)

SMU SuperPOD 101

Instructor

  • Instructor: Tue Vu, PhD
  • Office: 119 Ford Hall, SMU
  • Email: tuev@smu.edu

SMU SuperPOD 101

Course Outline

Topic

Description

Setup Preparing for the course
1. Introduction to SMU SuperPOD
2. Working with Conda Environment How to create personal conda environment in SuperPOD
3. Using NGC Container in SuperPOD How to use NGC Container in SuperPOD?
4. Using Jupyter Lab in SuperPOD How to use Jupter Lab in SuperPOD?
5. Using Batch script in SuperPOD How to run Batch script in SuperPOD
6. Job queueing and control in SuperPOD How to run control Job in SuperPOD
7. Data Science workflow with GPUs using RAPIDS How to install and use RAPIDS
8. Sample Application of NEMO for Sentiment Analysis How to use NEMO in container
9. Sample Applications of MultiGPUs for Computer Vision using Horovod How to utilize MultiGPUs in SuperPOD
10. Using YOLOv5 for object detection How to use train YOLOv5 to detect objects
11. Using Transfer Learning with ResNet50 How to use apply transfer learning to detect object
12. Using Stable Diffusion with SuperPOD How to use Stable Diffusion model
13. Using Pre-trained model from HuggingFace How to use pre-trained model already available from Hugging Face hub
Finish

The actual schedule may vary slightly depending on the topics and exercises chosen by the instructor.