Using Archai to search for best Face Segmentation model on Azure ML

  • Chris Lovett

This video shows how to use the Archai search framework to find the best possible face segmentation model on Azure ML including a check that the models work well on Qualcomm Snapdragon hardware.
– 00:00 – intro
– 02:08 – aml.py pipeline entry point
– 03:12 – demo
– 05_20 – parallel training on a gpu cluster
– 06:15 – the status table
– 08:34 – snapdragon quantization pipeline

See https://github.com/microsoft/archai

Speaker Details

Chris is a Principal Software Developer who joined Microsoft in ’97 and has worked on everything from XML, IE, .NET, SQL, Visual Studio, and Windows and joined Microsoft Research in 2015.

Watch Next