AWS Partners with Hugging Face for cheaper generative AI deployment
Most companies struggle to implement the quality and talent required to successfully adopt appropriate governance models, data collection, and generative AI models. Vendors are a workaround to help companies with tight budgets acquire the right skills.
Amazon Web Service (AWS) has partnered with Hugging Face to make it easier for IT teams to deploy and fine-tune large-scale language models for generative AI applications.
According to the announcement, Hugging Face offers researchers and data scientists 100,000+ machine learning models. This is a large language model trained on data containing 13 programming languages and 46 natural languages.
The partnership mimics similar agreements between other AI startups and big tech companies. Hugging Face uses AWS to develop its products, and AWS customers can adapt and fine-tune the technology to their business needs.
One of the biggest barriers to commercial adoption of generative AI models is the cost of building, customising, and fine-tuning the technology. This is where big tech companies have an edge.
Amazon and Hugging Face target the two biggest performance gaps for business cost and talent. Amazon announced that companies using Hugging Face models on AWS can save 50% on training costs and experience 4x higher throughput and up to 10x lower latency.
AWS customers can use Hugging Face mode on AWS via the AWS Tranium or AWS Inferentia tutorials, as well as the SageMaker JumpStart or Hugging Face AWS Deep Learning containers.
Whether a company runs its systems on AWS, Google Cloud, Azure, or other cloud provider, the potential for generative AI applications to trickle down from providers through partnerships and acquisitions of AI startups is available.
Oracle partnered with NVIDIA last October to bring its computing stack to Oracle's cloud infrastructure. Earlier this month, Google Cloud made an investment in AI startup to give users access to its AI capabilities. Microsoft Azure has a multi-year, multi-billion dollar partnership with OpenAI that will bring OpenAI models to enterprise and consumer products.