Product Manager at Microsoft
Sabrina is a Product Manager at Microsoft working with PyTorch and ONNX Runtime.
Title: Five ways to increase your model performance with PyTorch Profiler
We all like speed and want our models to run FASTER! This is where PIP INSTALL TORCH-TB-PROFILER comes in.
PyTorch Profiler v1.9 has been released! The goal of this new release is to provide you with new state-of-the-art tools to help diagnose and fix machine learning performance issues to help you run your models faster regardless of whether you are working on one or numerous machines. The objective is to target the execution steps that are the most costly in time and/or memory, and visualize the workload distribution between GPUs and CPUs. We all want to run our models faster and cheaper!
In this session, we will go over the new PyTorch Profiler release and how you can start leveraging this performance tool. We’ll be covering how to use Distributed Training view, Memory view, GPU Utilization Visualization, and Cloud Storage Support.