Hugging Face on Google Cloud Google
Google and Hugging Face have introduced a strategic partnership geared toward advancing open AI and machine studying growth.
This collaboration will combine Hugging Face’s platform with Google Cloud’s infrastructure, together with Vertex AI, to make generative AI extra accessible and impactful for builders. The partnership will allow Hugging Face customers and Google Cloud prospects to deploy fashions simply for manufacturing on Google Cloud with Inference Endpoints, speed up purposes with TPUs on Hugging Face Spaces and handle utilization by means of their Google Cloud account.
Developers will be capable of prepare, tune and serve open fashions shortly and cost-effectively on Google Cloud, leveraging AI-optimized infrastructure comparable to TPUs and GPUs. The partnership additionally helps Google Kubernetes Engine deployments, permitting for the creation of recent generative AI purposes.
This transfer is seen as a big step into AI for Alphabet, Google’s dad or mum firm, and is in comparison with the collaboration between Microsoft and OpenAI, though Hugging Face’s head of product, Jeff Boudier, commented that the Google-Hugging Face partnership is sort of totally different.
Google’s Tensor Processing Units are specialised {hardware} developed to speed up machine studying duties, significantly these involving giant matrix operations. Unlike General Purpose Graphics Processing Units, that are designed for parallel computing and are versatile throughout numerous compute workloads, TPUs are purpose-built for AI and ML workloads, specializing in tensor operations to attain increased speeds and power efficiencies. TPUs are additionally extra energy-efficient than GPUs. They are designed with a give attention to decreasing energy consumption per operation, which may result in decrease power prices and a smaller carbon footprint. With this partnership, Hugging Face customers will be capable of make the most of TPUs accessible by means of Google Cloud.
Vertex AI is Google’s machine studying and MLOps platform accessible in its cloud surroundings. With the mixing between the 2, Hugging Face customers can goal Vertex AI because the deployment platform to host and handle open fashions. They may also select GKE, the managed Kubernetes service for internet hosting fashions, which delivers fine-grained management and customization capabilities.
Hugging Face has attracted vital investments from tech giants, together with Google. In a Series D funding spherical, Hugging Face raised $235 million, with participation from Google, Amazon, Nvidia, Intel, AMD, Qualcomm, IBM, Salesforce, and others, doubling the startup’s valuation to $4.5 billion. With its dedication to open supply and open fashions, Hugging Face has quick develop into the popular platform for internet hosting fashions, datasets and inference endpoints. Almost all of the open mannequin suppliers, like Meta, Microsoft and Mistral make their fashions accessible on Hugging Face Hub.
Google has basis fashions which might be completely accessible on its public cloud platform. Gemini, one of many top-performing giant language fashions, was introduced final month. Other fashions like Imagen, Chirp and Codey are a part of the Vertex AI providing. With the mixing of Hugging Face with Google Cloud, prospects get the selection of proprietary fashions and open fashions for constructing and deploying generative AI purposes within the cloud.
The partnership between Google and Hugging Face is predicted to democratize AI by making it simple for corporations to construct their very own AI utilizing open fashions and applied sciences. As Hugging Face turns into the central hub for open-source AI software program, this collaboration will doubtless double the dimensions of its repository of AI-related software program.
The new capabilities, together with Vertex AI and GKE deployment choices, are anticipated to be accessible to Hugging Face Hub customers within the first half of 2024.
https://www.forbes.com/sites/janakirammsv/2024/01/28/google-and-hugging-face-partner-to-advance-generative-ai-with-open-models/
