- Brief overview of Data, Model and Pipeline Parallelism
- Systems Optimization for Neural Network Training and Inference using Pytorch Lightning, Ray Tune, and the Pytorch Intel Extension for CPU
For Evening of Python Coding Meetups that were held in Austin TX alongside the Austin python meetup and DC Python meetup. In this gathering, python programmers gathered and presented demos of various python tools and libraries. This page collects links to those demos and meetup pages in hope those demos will be useful.
- Gathering - where folk gathered and setup their computers
- Announcements - brief statements by attendees about events, and other python related issues
- Demo - short presentation with many questions
- Working groups - break out into multiple smaller working group such as python learners, web, data science, automation, and whatever attendees suggested
- Explain the paradigm of rising demand of GPU vs lowering cost of compute
- Parallelism Techniques
- Brief Kubernetes Demo of 2 Python workers and 1 Redis store on the control plane
- Python Class vs Ray Actor/Ray Task
- Pytorch
- Pytorch Lighting
- Pytorch Lighting + Ray Tune
- Pytorch + Ray Train
- Other optimization Opportunities Ahead