Build and deploy production ready PyTorch models
summary
Shows how to take a PyTorch image classifier from local notebook to a Kubernetes-backed REST endpoint using Azure Machine Learning, training a transfer-learning ResNet18 on ~50 images per class, registering the model, and wiring CI/CD with GitHub Actions + Event Grid and a Logic App to trigger deployments. Demonstrates YAML pipeline components, curated environments, NC6 GPU clusters (≈€1.17/hr), online endpoints with traffic routing, and a deployed model version (v14) that returns JSON labels — pipeline runs take ~10–15 minutes.
AI-written summary · about
chapters (10)
- 0:00Start
- 3:49Programming
- 5:00What is Deep Learning
- 14:19What is Azure Machine Learning?
- 15:33Azure ML service Key Elements
- 23:51Training Options
- 24:24What is Automated Machine Learning?
- 25:17Azure ML Designer
- 39:45ML Life Cycle
- 48:34Choosing an inferencing target
related talks
43mBuild and deploy PyTorch models with Azure Machine Learning
46mDeploying scalable Machine Learning models with Kubernetes
57mDevOps for Machine Learning
34mManaging Machine Learning in Production with Kubeflow and DevOps
33mDeploying and Scaling Spark ML and Tensorflow AI Models
44mGetting started with Azure Machine Learning services
42mDeploying ML at Scale with Kubernetes and TFX
39mMachine Learning Made Easy on Kubernetes. DevOps for Data Scientists