From Data Exploration to Modeling to Deployment
summary
Shows an end-to-end Azure notebook workflow that goes from exploration to production: uses Azure Notebooks, Data Science / Deep Learning VMs, HDInsight or Databricks Spark for scale, and Azure Machine Learning services to log cell-level runs, track parameters and metrics, and version artifacts. Demonstrations include an Iris classification and an energy-demand time series with grid search and feature importance, saving models to outputs and auto-generating a Swagger schema, then packaging as a Docker image and deploying to Kubernetes with routing and load balancing handled.
AI-written summary · about
chapters (7)
- 0:00Start
- 0:05Example of a Typical Data Science Workflow
- 2:19Challenges
- 3:42How Do I Deploy My Model to Production
- 9:42Apache Spark
- 20:31Train a Model
- 21:11Hyper Parameter Tuning
related talks
40mThe data scientists are gone. how do I deploy their work to prod?
37mPracticing Data Science in the cloud
48mStreamlining your Machine Learning projects with Azure Machine Learning
10mGet Started with Azure Machine Learning
45mMLOps and ML-Flops: going from dev to prod
49mBuilding Advance Analytics pipelines with Azure Databricks
57mDevOps for Machine Learning
19mFast Focus: Azure Notebooks for Artificial Intelligence