Meta and Microsoft Unveil Llama 2 AI Models

Meta and Microsoft Unveil Llama 2 AI Models for Azure and Windows Platforms

20th July ’23, New Jersey: Meta and Microsoft made an unprecedented move today. The tech giants Meta and Microsoft have announced their joint support for the revolutionary Llama 2 family of large language models (LLMs) on Azure and Windows. A groundbreaking development will soon unleash a new era of transformative possibilities, captivating the world’s imagination and reshaping industries, revolutionizing our work.

At the highly anticipated Microsoft Inspire event today, Meta and Microsoft made a historic announcement, introducing Llama 2. A powerful AI tool empowers developers and organizations worldwide to build generative AI-powered tools and experiences.

Meta and Microsoft are committed to democratizing AI and making its benefits accessible. With Meta adopting an open approach with Llama 2, developers will now have greater freedom in the types of models they build on, supporting available and frontier models for cutting-edge innovation. 

Azure customers can now seamlessly fine-tune and deploy the Llama 2 models with varying parameters. These include 7B, 13B, and 70B on the Azure platform. Azure is the world’s most widely adopted frontier and open model platform. It provides a robust and secure environment for developers to harness the full potential of Llama 2 models.

Meta will optimize Llama 2 to run natively on Windows, making it easy for Windows developers to leverage the power of AI. By targeting the DirectML execution provider through the ONNX Runtime, developers can effortlessly integrate Llama 2 into their applications, offering users an unparalleled AI-driven experience.

The Meta-Microsoft collaboration has a rich history, with previous successful joint efforts, such as integrating ONNX Runtime with PyTorch on Azure. This latest announcement further solidifies their partnership, signaling their dedication to accelerating innovation in the era of AI.

Azure’s purpose-built AI supercomputing platform stands at the forefront of AI research and development. It boasts a facility, hardware, and software uniquely designed to support the world’s leading AI organizations. Integrating Llama 2 models into Azure AI unlocks a world of possibilities for developers, providing powerful tools for model training, fine-tuning, and inference while ensuring top-notch AI safety capabilities.

Llama 2’s inclusion in Windows is a significant milestone. It has positioned Windows as the ultimate destination for developers to craft personalized AI experiences for their customers’ needs. With cutting-edge tools like Windows Subsystem for Linux (WSL), Windows Terminal, Microsoft Visual Studio, and VS Code, developers can now fine-tune LLMs directly on their Windows PCs. All of this will further advance the AI revolution.

As part of Microsoft’s dedication to responsible AI, they have adopted an iterative, layered approach to mitigate potential risks associated with large language models. Azure AI customers can test Llama 2 using their data samples to evaluate its performance for specific use cases. Prompt engineering and retrieval augmented generation (RAG) techniques are utilized to create safer and more reliable end-user experiences.

The collaboration between Meta and Microsoft exemplifies a shared vision of building a responsible, open, and inclusive AI ecosystem. The Azure AI model catalog serves as a hub for foundation models. And the Llama 2 models’ integration into the Azure AI model catalog marks another significant stride towards democratizing AI innovation.

The world stands on the brink of an AI revolution. Llama 2 on Azure and Windows empowers developers and organizations to unlock AI’s potential to transform industries and human-technology interaction.

Leave a Reply