Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Model Mondays S1:E7 - Hands-on With Open Source...

Model Mondays S1:E7 - Hands-on With Open Source Models

Join us for Model Mondays – a weekly show where we round up all the latest AI model news and put one model in the spotlight for a deeper dive.

In this episode, we’re joined by Pamela Fox (Principal Cloud Advocate at Microsoft) as we talk about open source models for generative AI and put the spotlight on the Llama 4 herd of models from Meta. Learn what capabilities these models offer and get a sense for their multimodal (vision) capabilities with an interactive hands-on demo.

Session Resources:
https://aka.ms/model-mondays
https://aka.ms/model-mondays/discord
https://aka.ms/model-mondays/chat

Explore The Models
https://aka.ms/model-mondays/collection

Watch The Replay
https://developer.microsoft.com/en-us/reactor/events/25358/

Nitya Narasimhan, PhD

April 24, 2025
Tweet

More Decks by Nitya Narasimhan, PhD

Other Decks in Technology

Transcript

  1. Azure AI Foundry PM, Microsoft Alexander Hughes Pamela Fox Principal

    Cloud Advocate, Microsoft Spotlight This Week: Open Source Models Build Your Model IQ https://aka.ms/model-mondays
  2. Senior AI Advocate Nitya Narasimhan, PhD Sharmila Chockalingam Senior Product

    Marketing Mgr. Welcome to S1:E07 Model Mondays Build Your Model IQ https://aka.ms/model-mondays
  3. ❶ · Try Command A & Embed 4 https://aka.ms/model-mondays Agentic

    RAG Deploy high-performance LLMs and semantic search engines directly into production with Command A and Embed 4 on Azure AI Foundry.
  4. ❷ · Try MAI-DS-R1 https://aka.ms/model-mondays MAI-DS-R1 A new open-weights DeepSeek

    R1 model variant with improved responsiveness on blocked topics while maintaining its reasoning capabilities.
  5. ❸ · Try OpenAI Models https://aka.ms/model-mondays GPT 4.1 and o

    Series GPT 4.1 models represent something special—a generational advancement in coding, reasoning, and understanding context o3 — Their most advanced model yet, strong in coding, math, science, and visual tasks o4-mini — Smaller, faster, and remarkably capable in technical domains, at lower cost.
  6. ❺ · Try Azure MCP Server https://aka.ms/model-mondays Azure MCP Server

    Connect AI agents to Azure services like Storage, Cosmos DB, AppConfig, Monitor and more - with natural language.
  7. Welcoming the Llama 4 Herd to Azure AI Foundry What

    is Llama 4? What are the model's capabilities and advantages? How can you get started with it? Alexander Hughes Product Manager II – Azure AI Foundry Models
  8. ❶· What is Llama 4? Llama 4 herd of state-of-the-

    art natively multimodal models on Azure AI Foundry We’re sharing the first models in the Llama 4 herd, which will enable people to build more personalized multimodal experiences. According to Meta, Llama 4 Scout is one of the best multimodal models in its class and is more powerful than Meta’s Llama 3 models, while fitting in a single H100 GPU.
  9. ❸· What is MoE architecture? Mixture of Experts (MoE) Architecture

    achiev es strong performance without sacrificing compute cost Llama models are fully open source, which allows for their cross-industry benchmarking and use What makes Llama 4 special?
  10. ai.azure.com/explore/models/Llama-4-Scout-17B-16E/version/1/registry/azureml-meta ❹· How can I use Llama 4? How to

    get started with Llama 4 How to use inferencing code in your application To get started: 1. Navigate to Azure AI Foundry or GitHub Models 2. Search for Llama 4 Scout or Llama 4 Maverick in the model catalog 3. In Foundry, select 'Deploy' and then the 'Serverless API' option. In GitHub, search for the model and test it in the playground. You can select 'Use this model' to view the code you can use in your application.
  11. Learn about open source like Llama models and how to

    deploy the latest Llama4 model and its uses cases. Hands on with Open-Source Models Pamela Fox, Principal Cloud Advocate - Microsoft Models
  12. Hands-on with Forecasting Models Learn about the Nixtla TimeGEN-1 model

    for time-series forecasting with applications to finance, retail & more. Khuyen Tran, Developer Advocate - Nixtla Next Week
  13. Watch the Livestream Join the Conversation Explore the Resources Mon

    1:30pm ET Microsoft Reactor https://aka.ms/model-mondays/RSVP Fri 1:30pm ET Azure AI Discord https://aka.ms/model-mondays/Discord Watch or Star GitHub Repo https://aka.ms/model-monday #ModelMondays