Microsoft and Hugging Face Deepen Collaboration: Access 10K AI Models on Azure

Hugging Face and Microsoft have deepened their collaboration to make open-source AI more accessible everywhere, announcing a broad set of new features and collaborations at Microsoft Build. The expanded partnership puts over ten thousand Hugging Face models at the fingertips of Azure developers, enabling immediate deployment of popular open models on the cloud platform.

What Happened

The collaboration between Hugging Face and Microsoft has been ongoing for several years, with a shared goal to democratize AI by making it easier for developers and enterprises to access, deploy, and scale the latest open models on Azure. The expanded partnership builds upon previous collaborations, including the introduction of the Hugging Face Collection in the Azure Model Catalog last year.

Today, Microsoft and Hugging Face announced two major new experiences to build AI with open models on Microsoft Azure: Expanded HF Collection in Azure Model Catalog and Build AI with the new AMD MI300X on Azure. The expanded collection includes some of the most popular open Large Language Models, such as Llama 3 from Meta, Mistral 7B from Mistral AI, Command R Plus from Cohere for AI, Qwen 1.5 110B from Qwen, and some of the highest performing fine-tuned models on the Open LLM Leaderboard from the Hugging Face community.

The new virtual machines (VMs) based on the latest AMD Instinct MI300 GPUs will enable developers to achieve amazing performance and cost/performance for Hugging Face models. This work leverages the deep collaboration between Hugging Face, AMD, and Microsoft, with optimization, ROCm integrations, and continuous testing of Hugging Face open source libraries and models on AMD Instinct GPUs.

Background and Context

Hugging Face has emerged as a central hub hosting millions of open models, serving as the default launchpad for open-source AI innovation. The company's platform allows organizations and developers to access and share models, datasets, and applications, with over 8 million AI builders using the Hugging Face platform.

The collaboration between Hugging Face and Microsoft aims to bring open-source AI innovation to Azure through a secure, scalable, and reliable integration between the Hugging Face hub and Azure AI Foundry. This integration enables developers to incorporate more modalities in their applications, including text generation, feature extraction, fill-mask, translation, identifying sentence similarity, image classification, image segmentation, text-to-image generation, image-to-text conversion, automatic speech recognition, and audio-classification.

Microsoft is also working with Hugging Face to bring in additional models enabling tasks such as drug research and computer vision. The collaboration has the potential to accelerate open-source AI innovation on Azure, making it easier for developers and enterprises to access, deploy, and scale the latest open models.

Why It Matters

The expanded partnership between Hugging Face and Microsoft is significant for several reasons. Firstly, it puts over ten thousand Hugging Face models at the fingertips of Azure developers, enabling immediate deployment of popular open models on the cloud platform. This will make it easier for developers to access and deploy AI models, accelerating innovation in various industries.

Secondly, the collaboration highlights the growing interest in transformers among smaller companies, including those without in-house machine learning talent. Hugging Face's vision is to become the GitHub for machine learning, providing a broad range of machine learning services, including off-the-shelf transformer models.

The partnership also has implications for market consolidation and potential acquisitions in the future. As Hugging Face continues to grow and expand its offerings, it may face challenges from larger tech companies, potentially leading to market consolidation or acquisition.

What Comes Next

The expanded collaboration between Hugging Face and Microsoft is just the beginning of a broader partnership that will enable more models in the Azure AI Studio model catalog and introduce new features and experiences in the months to come. The companies will continue to work together to make open-source AI more accessible everywhere, accelerating innovation and democratizing access to AI.

Key Facts

• Hugging Face and Microsoft have deepened their collaboration to make open-source AI more accessible everywhere.
• The expanded partnership puts over ten thousand Hugging Face models at the fingertips of Azure developers.
• The new virtual machines (VMs) based on the latest AMD Instinct MI300 GPUs will enable developers to achieve amazing performance and cost/performance for Hugging Face models.
• Microsoft is working with Hugging Face to bring in additional models enabling tasks such as drug research and computer vision.
• The collaboration has the potential to accelerate open-source AI innovation on Azure, making it easier for developers and enterprises to access, deploy, and scale the latest open models.