The collaboration enables developers to harness Google Cloud’s robust infrastructure seamlessly for all Hugging Face services, facilitating the training and deployment of Hugging Face models on Google Cloud platforms.

This partnership represents a significant stride towards Hugging Face’s mission to democratise AI and underscores Google Cloud’s commitment to fostering the growth of the open-source AI ecosystem. By leveraging Google Cloud’s advanced AI-optimised infrastructure, including compute capabilities, tensor processing units (TPUs), and graphics processing units (GPUs), developers gain enhanced agility and cost-efficiency in training and deploying open models, thus unlocking new possibilities for generative AI applications.

Key highlights of the partnership include:

• Streamlined Model Training and Deployment: Developers can now effortlessly train, fine-tune, and deploy Hugging Face models using Vertex AI directly from the Hugging Face platform, leveraging Google Cloud’s purpose-built MLOps services to expedite the development of cutting-edge AI applications.
• Support for GKE Deployments: Through Google Kubernetes Engine (GKE), developers utilizing Hugging Face services can employ “do it yourself” infrastructure, scaling models efficiently with Hugging Face-specific Deep Learning Containers on GKE.
• Enhanced Performance with Cloud TPU v5e: Access to Cloud TPU v5e offers developers up to 2.5 times more performance per dollar and up to 1.7 times lower latency for inference, enhancing the efficiency of model inference tasks.
• Future-Ready Infrastructure with A3 VMs: The partnership aims to introduce support for A3 VMs, powered by NVIDIA’s H100 Tensor Core GPUs, promising three times faster training and ten times greater networking bandwidth compared to the previous generation.
• Simplified Management via Google Cloud Marketplace: Developers can seamlessly manage and monitor Hugging Face’s managed platform, including Inference, Endpoints, Spaces, AutoTrain, and more, through the user-friendly interface of Google Cloud Marketplace.

The alliance marks a significant leap forward in democratising access to advanced AI capabilities while empowering developers to innovate and scale their ML initiatives on a global scale. As the partnership evolves, it is poised to reshape the landscape of generative AI and ML development, driving innovation and accelerating the pace of AI adoption across industries.

“Google Cloud and Hugging Face share a vision for making generative AI more accessible and impactful for developers,” said Thomas Kurian, CEO at Google Cloud. “This partnership ensures that developers on Hugging Face will have access to Google Cloud’s purpose-built AI platform, Vertex AI, along with our secure infrastructure, which can accelerate the next generation of AI services and applications.”

“From the original Transformers paper to T5 and the Vision Transformer, Google has been at the forefront of AI progress and the open science movement,” said Clement Delangue, CEO of Hugging Face. “With this new partnership, we will make it easy for Hugging Face users and Google Cloud customers to leverage the latest open models together with leading optimized AI infrastructure and tools from Google Cloud including Vertex AI and TPUs to meaningfully advance developers ability to build their own AI models.”

Vertex AI and GKE will be available in the first half of 2024 as deployment options on the Hugging Face platform.

Nigle Walsh, Managing Director of Insurance at Google, also commented, describing the partnership as an: “Exciting development that provides more choice and velocity to builders at the critical phase of moving beyond model testing to refinement and solution creation.”