Getting started with Azure Machine Learning services
summary
Live demo trains a custom vision model in about one minute from ~50 images, yields low initial confidence but can export a zip/ONNX model for mobile, and registers versions and logs in an Azure workspace. Provisioning a GPU cluster takes ~15 minutes, experiments run via notebooks or Python SDK, and deployment produces a Docker image runnable on ACI, AKS or a Raspberry Pi. trained a custom vision model in one minute from ~50 images exported Docker image and deployed to ACI and Raspberry Pi
AI-written summary · about
chapters (13)
- 0:00Introduction and speaker intro
- 2:04What is machine learning
- 3:53Azure ML services overview
- 7:05Defining your ML project
- 9:14Preparing your data
- 13:10Training and evaluating models
- 15:59Computer vision services demo
- 23:25Custom vision training
- 29:17Setting up Azure ML workspace
- 32:50Running experiments in the cloud
- 37:10Deploying models to production
- 42:27Testing deployed models
- 43:54Summary and Q&A
related talks
01h 00mLearn Live: Get started with AI on Azure
16mHow to get started with Computer Vision AutoML in Azure
10mGet Started with Azure Machine Learning
43mA lap around Azure Machine Learning services
45mFrom Data to AI on The Edge
01h 02mAI on the Edge
24mDon't want to miss out on the AI wave?
48mStreamlining your Machine Learning projects with Azure Machine Learning