NVIDIA AI Enterprise Registry for AzureML deployments of AI Workflows
summary
Azure ML hosts an NVIDIA AI Enterprise registry of environments, components and pipelines that run GPU‑optimized toolkits like Tao transfer‑learning toolkit and Rapids, with pipelines that download data, fine‑tune pretrained models, prune/quantize, package to a Triton‑ready model repository and deploy as Azure ML REST endpoints. Demo jobs show object‑ and face‑detection reuse the same pipeline by swapping data and base model, expose training metrics and sample inference images, and are currently available by invitation in preview.
AI-written summary · about
chapters (19)
- 0:00Introduction
- 1:37NVIDIA AI Enterprise overview
- 3:06Azure ML basics
- 4:10Azure ML registries and NVIDIA
- 6:01NVIDIA registry in Azure ML
- 8:14Key components in registry
- 10:22NVIDIA components in Azure ML
- 13:10Pre-built pipelines
- 14:36Customizing pipelines
- 16:22Running pipelines
- 17:26Demo: object detection
- 18:52Training metrics
- 19:57Face detection pipeline
- 21:05Deploying models
- 23:56Production AI benefits
- 24:17Customer success
- 26:28Pipeline differences
- 27:12Q&A
- 29:03Next steps
related talks
43mA lap around Azure Machine Learning services
22mQuickly deploy NVIDIA optimized generative AI on Azure AI Studio
47mAccelerate your AI journey with Azure AI model catalog
50mAccelerate development and deployment of Gen AI with NVIDIA on Azure
58mDeep Learning in Microsoft Azure: CNTK, CaffeOnSpark and Tensorflow
48mStreamlining your Machine Learning projects with Azure Machine Learning
52mDevOps for AI in the Microsoft World
50mAzure AI, Power new possibilities for every organization