Google Cloud announced a strategic collaboration today with Hugging Face, the leading AI model repository, to advance the development and accessibility of generative AI. The partnership provides Hugging Face developers streamlined access to Google's optimized AI infrastructure for model building.
Over the last two years, Hugging Face has emerged as a vital nexus of the open AI community, with over a million models, datasets and apps hosted on its eponymous model hub. Its expressed mission is to "democratize AI" by enabling any developer to leverage state-of-the-art models.
Google Cloud becomes Hugging Face's preferred cloud partner under the new alliance. Developers on Hugging Face's platform can harness Google's tensor processing units (TPUs), top-of-the-line NVIDIA GPUs, and its Vertex AI machine learning suite.
For Google Cloud customers, this partnership brings new experiences. They'll benefit from training and deploying Hugging Face models within Google Kubernetes Engine (GKE) and Vertex AI, utilizing unique hardware capabilities like TPU instances and NVIDIA H100 Tensor Core GPUs.
Vertex AI integration will allow for seamless model deployment from Hugging Face to Google Cloud. While Google Kubernetes Engine will facilitate containerized workloads at scale. Additionally, the collaboration grants open source developers access to cutting-edge hardware like the NVIDIA H100 at lower cost.
It's truly an exciting time for the AI community, with the promise of continued innovation and growth driven by open source research and development. Overall, this partnership offers exciting possibilities for developers, researchers, and businesses alike.