Accelerate your AI journey with Azure AI model catalog
summary
Azure AI Model Catalog shows live deployments of Mistral, Phi3, Llama3 and the new Time Gen 1 forecasting model, offers serverless APIs and managed GPU hosts, and claims data never trains provider models while HiddenLayer scans catalog models for vulnerabilities. It runs RAG with vector indexes, provides benchmarking and side‑by‑side evaluations (example: swapping a 5 three mini SLM for a larger model drops cost and latency with small accuracy loss), supports LoRA fine‑tuning, on‑device Phi3 Vision, and cites customers — EY’s chat used by 275,000 employees and Bridgestone reporting a 30% reduction in forecasting errors.
AI-written summary · about
chapters (15)
- 0:00Introduction to Model Choice
- 2:58Azure AI Model Catalog
- 5:41Catalog Features and Usage
- 7:31Model Playground and Rag
- 9:48Deployment and API Consistency
- 12:24Model Deployment Demonstration
- 16:20SDK and Tooling Integration
- 19:57Building with Prompt Flow
- 26:38Evaluation and Small Models
- 29:41Enterprise Security and Rules
- 31:47New Time Series Model Announcement
- 33:29Customer Success Stories
- 37:22Lenovo and Fine-tuning
- 40:43On-Device and Vision Capabilities
- 43:38Conclusion and Resources
related talks
48mDeploy large language models responsibly with Azure AI
48mUsing Azure AI Foundry to manage all your Large Language Models
23mLlama V2 in Azure AI for Finetuning, Evaluation and Deployment from the Model Catalog
30mNVIDIA AI Enterprise Registry for AzureML deployments of AI Workflows
01h 01mAI and ML imbedded in Azure Data Platform
31mAzure AI Innovations at Microsoft
40mCloud-based Machine Learning Offerings - Beginner's Perspective
01h 00mLearn Live: Get started with AI on Azure